CA3I7 
 
 ,?*• 
 

 STEEL SHIPS. 
 
BY 
 
 THOMAS JEALTON, 
 
 NAVAL architect; author of "know tour own ship. 
 
 iFiLUSTRATED WITH PlATES, FoLDING DIAGRAMS REDUCED FROM WORKING 
 
 Drawings, and Illustrations in the Text. 
 
 LONDON: 
 
 CHARLES GRIFFIN & COMPANY, LIMITED. 
 
 PHILADELPHIA: J. B. LIPPINCOTT COMPANY. 
 
 190L 
 
 [All Eights Reserved.'] 
 
 (T 
 
 \STEEL SHIPS !/ ,^ 
 
 THEIR CONSTRUCTION AND MAINTENANCE. 
 
 A MANUAL FOB SHIPBUILDERS, SHIP SUPERINTENDENTS, 
 STUDENTS, AND MARINE ENGINEERS. 
 
PPJNTKD BY 
 
 IJEILL AND COMPANY, LIMITED, 
 
 EDINBURGH. 
 
PREFACE. 
 
 The present volume is, in a large measure, the outcome of the 
 gratifying reception with which the smaller work, "Know your 
 Own Ship," published in the Nautical Series, met on its first and 
 subsequent editions. The success of that book emboldened its Author 
 to embark, at the suggestion of the publishers, upon the preparation 
 of a larger and more important undertaking, the result of which is 
 now presented to the reader in the hope that it will be found, not 
 less than the former work, to merit the approval of that section 
 of the shipbuilding world for whose needs it has been specially 
 devised. A sketch of the plan of the book will be found at the end 
 of this preface ; it is therefore unnecessary here to do more than 
 briefly outline the circumstances under which it has been written. 
 The work has taken four years to complete, the Author having 
 been unable to devote more than his leisure hours to its com- 
 position. A careful study of some years' duration, carried out in 
 that centre of the steel trade, the Cleveland district, has afforded the 
 necessary basis for the first two chapters, while the subsequent 
 chapters are the result of the Author's daily experience in the pro- 
 fession of ship construction and maintenance. The book has been 
 copiously illustrated, and no expense has been spared in the 
 preparation and execution of diagrams intended to amplify and 
 elucidate the text. These have been placed in close juxtaposition to 
 those portions of the work to which they refer, the Author conceiving 
 that they will thus prove more readily available for purposes of 
 reference than if they had been published as a separate volume, 
 as is sometimes done in works of this class. The whole subject has 
 been treated from a practical point of view, and the requirements 
 of students, ship superintendents, shipbuilders, and marine engineers 
 have been carefully studied. The Author desires to express his 
 
VI PREFACE. 
 
 thanks to the Cimard Steamship Co., Liverpool, for the permission 
 to reproduce illustrations and details of their Eoyal Mail Steamers 
 ' Campania ' and ' Lucania ' ; to Messrs. Alfred Holt & Co. (Ocean 
 Steamship Co., Liverpool), for permission to illustrate one of their 
 latest cargo steamers; and also to the following shipbuilders, — 
 Messrs. Armstrong, Whitworth, & Co., Wm. Doxford & Co., Eopner 
 & Son, J. Priestman & Co., and Messrs. Sir Eaylton Dixon & Co., 
 Middlesbrough, for permission to illustrate particular vessels built 
 by them, and some details of their methods of construction. 
 
 Further, he desires to record his best thanks to his friend, 
 J. Bolam, Esq., Head Master, Government Navigation School, Leith, 
 for his kindly assistance in reading over part of the manuscript. 
 
 THOMAS WALTOK 
 
 London, June 1901. 
 
PLAN OF BOOK. 
 
 Chapter I. is a condensed description of the processes of the manu- 
 facture of steel and iron, from its crude state in the form of ore, to the 
 finished product in the form of ship plates, forgings, hars, etc., particularly- 
 noting those constituents of the material which are essential to the production 
 of good ship steel or iron, and those which, if in excess, introduce objection- 
 able quahties into the metal. 
 
 Chapter II. treats of the strength and quality of ship steel and hon as 
 a result of the proportions in which the various constituents referred to 
 above are present in the metal, and the particular processes through which 
 the material passes in the course of manufacture. 
 
 A description is also given of the tests applied in order to definitely 
 ascertain both the strength and quality. 
 
 Chapter III. explains what is meant by a vessel being 'Classed,' and 
 the nature of the work of those Societies empowered to assign loadhnes. 
 
 Chapter IV. is a general introduction to the subject of ship construc- 
 tion, drawing attention to the principal structural features, and the alter- 
 native modes in which a vessel may be built. 
 
 Chapter Y. deals with the various forces which are exerted upon the 
 hulls of ships, tending to strain them and produce deformation ; and shows 
 also how to estimate the maximum stresses endured by the material under 
 the worst of such conditions. 
 
 Chapter VI. — Section I. gives a structural description of i\iQ fundamental 
 types of vessels known as ' Full Scanthng ' (one, two, and three deck), Spar- 
 Decked, and Awning-Decked vessels ; and modifications of these types. 
 
 Section II. describes the construction of typical vessels. Among these 
 are : — The Eoyal Mail Steamers ' Campania ' and ' Lucania ' ; the ' Great 
 Eastern'; an Ocean Steamship Co.'s cargo steamer; 'Turret,' 'Trunk,' 
 and other 'Self -Trimming' steamers; and, in addition, special arrangements 
 for carrying water ballast for long over-sea voyages. 
 
 Chapter VII., which is the largest section of the book, deals in detail 
 with the construction and combination generally of the various parts which 
 go to make up the Avhole ship structure — framing, plating, stern frames 
 and rudders, riveting, pumping, ventilation, etc., and includes also remarks 
 upon launching. 
 
 Chapter VIII. describes the causes of decay and deterioration generally 
 in a vessel, particularly noting those parts especially liable to rapid corrosion, 
 and the best means of combating the causes of such corrosion, and of pre- 
 serving and maintaining the structure in a state of efiSciency. 
 
CONTENTS. 
 
 CHAPTER I. 
 
 Iron and Steel. 
 
 PAGES 
 
 Iron and Steel supersede Wood in Shipbuilding— Composition of Iron and 
 Steel— Pig Iron, its Nature and Manufacture— The Blast Furnace- 
 Castings— Malleable Cast Iron— Malleable or Wrought Iron ; its Nature 
 and Manufacture— The Puddling Furnace — Quality and Classification of 
 Wrought Iron— Steel, its Nature and Composition— Siemens Steel, Manu- 
 facture of— Bessemer Steel, Manufacture of— Basic Steel— Cementation 
 Steel— Case-hardened Steel— Crucible Steel— Steel Castings— Forgings— 
 Iron and Steel Sections used in Shipbuilding 1-18 
 
 CHAPTER II. 
 
 Strength, Quality, and Tests of Steel for Shipbuilding Purposes. . 
 
 Definitions of Important Terms ; Tensile Strength, Stress, Ductility, Elasticity, 
 Elastic Limit— Value of Nickel— Fatigue— Tests of Plates and Angles- 
 Remarks upon the Reduction in Thickness of Steel Plates— Tests for 
 Steel Castings— Rivet Tests— Treatment of Plates and Bars in Shipyard . 19-25 
 
 CHAPTER III. 
 
 Classification, 
 
 Purpose for which Classification Societies exist— Societies empowered to assign 
 Load Lines — Government the supreme Authority for Assignment of Load 
 Lines, and responsible for Seagoing Condition of Vessels leaving British 
 Ports— Standard of Strength upon which Load Lines are assigned— Load 
 Lines of Three Deck, Spar Deck, and Awning Deck Vessels— Grades of 
 Class — Maintenance of Class— Unclassed Vessels 26-32 
 
 CHAPTER IV. 
 
 Outline of Principal Features and Alternative Modes of 
 Ship Construction. 
 
 Transverse and Longitudinal Framing — Form and Function of Parts — Butts 
 in Transverse Framing— Framing in Double Bottoms — Regulations for 
 increasing the number of Tiers of Beams — Compensation for dispensing 
 Avith Hold Beams — Necessity of thorough combination of Transverse 
 and Longitudinal Framing— Structural Value of Shell Plating— Alter- 
 native Modes of Construction— Numerals for Scantlings .... 33-50 
 
CONTENTS. IX 
 
 CHAPTER Y. 
 Stress and Strength. 
 
 PAGES 
 
 Preliminary. — Forces exerted upon Ships— Water Pressures— Pressure per 
 Frame Space — Estimated Pressure upon a Bulkhead and Centre of 
 Pressure — Pressure upon a Tank Top — Tendency to Transverse Deforma- 
 tion — Longitudinal Stresses in still water and tendency to Longitudinal 
 Deformation — Bending Moment on a Loaded Bar — Possible effect of a bad 
 disposition of Weights in a Ship — Longitudinal Stresses in Wave Water 
 and tendency to Longitudinal Deformation — Local Stresses, Panting 
 Stress, Stress due to Propulsion by Steam, Stress due to Propulsion by 
 Sail, Stress caused by loading Heavy Cargoes on deck and the shipping 
 of Heavy Seas, Rudder Stresses, Strains from loading aground, Stresses 
 upon Vessels which are temporarily only partially waterborne, Launching 
 Stresses — Stress and Strain — Dnctility and the Elastic Limit — Moment of 
 Inertia — Curve of Loads — Curve of Shearing Stresses — Curve of Bending 
 Moments — Stress per Square Lich — The Equivalent Girder — Computa- 
 tion of Moment of Inertia of Compound Girder — Stress per Square Inch 
 upon Upper and Lower Surfaces of foregoing Girder — Bending Moment of 
 an Actual Ship ; Curve of Weights, Curve of Buoyancy, Curve of Loads, 
 Curve of Shearing Forces, Curve of Bending Moments — Moment of 
 Inertia of Actual Ships — Comparison of Vessels — Disposition of Material 
 — Eff'ect of Modification in Depth of Transvei-se Frames — Value of Stress 
 Calculation — Calculation for Position of Neutral Axis and Moment of 
 Inertia of an Actual Ship — Estimate of Bending Moment for an Actual 
 Ship, and Stress per square inch — Value of Registration Societies to Ship- 
 owner — Comparison of Stresses on Vessels increasing in size — Further 
 Remarks upon the Value of Stress Calculations — Working Stress — 
 Erections on Deck — Board of Trade Instructions for Comparing the 
 Strength of Vessels for Freeboard Purposes — Deductions — Calculation to 
 find Neutral Axis — Calculation of Moment of Inertia of a Vessel when 
 supported upon Wave Crest at Middle of Length, and thus subject to 
 Hogging Stresses 51-106 
 
 CHAPTER VI. 
 
 Types of Vessels. Section I. 
 
 Fundamental Types and Modifications of same — Relation between Deck 
 Erections and Deadweight — No Reduction in Freeboard for Excessive 
 Strength in a vessel with Full Scantlings — Determination of Type— Three 
 Deck, One and Two Deck, Spar and Awning Deck Vessels — Illustration 
 of Principal Scantlings of foregoing types — Vessels of Intermediate Grades 
 between Three Deck and Spar Deck, and Spar Deck and Awning Deck- 
 Raised Quarter Deck Vessels — Maximum Stress — Partial Awning Deck, 
 Shelter Deck, Well Deck, Shade Deck Vessels, etc 107-131 
 
 Section II. 
 
 Description and Illustration of the Principal Structural Features of some 
 noted, and also some novel Types of Vessels — ' Campania ' and ' Lucania ' 
 — 'Great Eastern' — Turret Steamers (Doxford's) — Trunk Steamers 
 (Ropner's) — Self-Trimming Steamers (Priestman's) — Some advances in 
 Modern Shipbuilding — Steamers for carrying Oil in Bulk — Water-ballast 
 Arrangements 132-166 
 
CONTENTS. 
 
 CHAPTER VII. 
 Details of Construction. 
 
 PAGES 
 
 Rivets and Riveting — Butt Straps and Butt Laps — Keel Blocks and Launch- 
 ing Ways — Frames, Reverse Frames, and Floors — Beams — Pillars — 
 Keelsons and Stringers — Bulkheads — Decks — Outside Shell Plating — 
 Stern Frames and Rudders — Miscellaneous Details : Continuity of 
 Strength, Engine and Boiler Space, Masts and Derricks, Panting, 
 Hatches, Deck Houses, Poop and Bridge Front Bulkheads, Tunnel and 
 Casings, Breast Hooks, Bilge Keels— Ventilation— Pumping — Launching 167-269 
 
 CHAPTER VIII. 
 
 Maintenance 270-286 
 
 Index 287-290 
 
PLATES AND ILLUSTRATIONS. 
 
 Frontispiece. Royal Mail Twin Screw Steamer ' Liicania. 
 
 Plate I. ' Turret Deck ' Cargo Steamer, . 
 
 II. ' Truuk' Cargo Steamer, . 
 
 III. A Typical Cargo Steamer, 
 
 IV. Royal Mail Twin Screw Steamer ' Oceanic,' 
 
 To face page 138 
 ,, 142 
 ,, 146 
 ,, 268 
 
 DIAGRAMS : AND FIGURES IN TEXT. 
 Fig. 
 
 1. Section of a Cleveland blast furnace, . 
 
 2. Cold bending test. 
 
 3. Hot angle tests, 
 
 4. Hot tee tests, .... 
 
 5. Rivet tests, .... 
 
 6. Midship section showing transverse framing, 
 
 7. ,, ,, showing transverse framing with a cellular double bottom, 
 
 8. ,, ,, longitudinal and transverse framing with ordinary floors, 
 
 9. ,, ,, merchant steamer, 
 
 10. ,, ,, two tiers of beams, . 
 
 11. ,, ,, hold beams dispensed with, . 
 
 12. ,, ,, compensation deep framing, . 
 
 13. Length between perpendiculars, 
 
 14. Types of upper deck, spar deck and awning deck vessels, 
 
 15. Water pressures on variously shaped objects, 
 
 16. ,, ,, internal and external, 
 
 17. Total side pressure, 
 
 18. Calculation of amount of pressure, 
 
 19. Pressure upon watertight bulkhead, . 
 
 20. Direction of water pressures upon hull, 
 
 21. Tendency to transverse deformation, . 
 
 22. ,, vertical elongation, 
 
 23. ,, to sag through deck weights, etc. , 
 
 24. Tendency to deformation with racking strains, 
 
 25. Alternating buoyancy and weight in light loaded ships, 
 
 26. Uniform buoyancy and weight, 
 
 PAGE 
 
 7 
 22 
 23 
 23 
 25 
 34 
 37 
 40 
 43 
 44 
 45 
 46 
 48 
 49 
 55 
 56 
 56 
 57 
 59 
 61 
 62 
 62 
 63 
 64 
 65 
 65 
 
xu 
 
 PLATES AND ILLUSTRATIONS. 
 
 Pia. 
 27. 
 28. 
 29. 
 30. 
 31. 
 32. 
 33. 
 
 Bending moments in ship loaded fore and aft, 
 
 ,, ,, in bar loaded each end, 
 
 Tendency to ' hog,' 
 
 ,, fracture at ends of bridge, 
 
 ,, buckle at ,, ,, 
 
 ' Stressed ' bar, 
 
 34. Elongation and compression in bent bar, 
 
 35. Variation of stresses in bent bar, 
 
 36. Girder and bar sections, 
 
 37. ' Moment of inertia,' . 
 
 38. Curve of weight for loaded bar, 
 
 39. Curve of shearing stresses, 
 
 40. ' Compound Girder,' . 
 
 41. ' Equivalent Girder,' . 
 
 42. The equivalent girder in application, . 
 
 43. Bending moment in actual ship, 
 
 44. Section of deep framing, 
 
 45. Midship section : single deck vessel, 
 
 46. 47. Sectional profiles of two-deck and raised quarter deck vessels, 
 
 48. Midship section : three deck shelter deck steamer, 
 
 49. Profile section, ,, ,, ,, 
 
 50. Raised quarter deck vessel, 
 51-55. Commonly recurring types of vessels, 
 
 56. Bridge front bulkhead stiffened with channel bars, 
 
 57. Royal Mail Screw steamers, ' Campania ' and ' Lucauia ' type, 
 
 58. Midship section, ' Campania' and ' Lucauia' type, 
 58a. Detail of hydraulic riveting in upper deck sheer strake, 
 
 59. Sectional profile of ' Great Eastern,' . 
 
 60. Midship section, 'Great Eastern,' 
 
 61. * "Whaleback ' steamer, 
 
 62. 'Turret steamer,' two deck type, 
 
 63. ,, ,, with one laid deck, 
 
 64. ,, ,, with two decks, 
 
 65. ,, ,, with one deck and widely spaced beams, 
 
 66. ' Trunk steamer,' profile and deck plan, 
 
 67. „ ,, midship section, 
 
 68. Trunk for coal cargoes, 
 
 69. Trunk for rice cargoes, 
 
 70. Self-trimming steamer, 
 ,, ,, ,, midship section, 
 
 Some new features in modern shipbuilding, 
 
 midship section. 
 
 Detail sketches of fig. 72, 
 Stem arrangements and rudder (fig. 72), 
 Oil carrying steamer, elevation and deck plans (engines aft), 
 ,, ,, elevation (engines amid ships), 
 
 78. ,, ,, ,, midship section, 
 
 79. Enlarged view of coffer-dam, . 
 
 80. Transverse bulkheads, 
 
 81. Vertical and horizontal stiffeners, 
 
 82. Side stringer cut and bracketed to bulkhead, 
 
 83. Liner, and angle connection to shell, . 
 
 84. Bulkhead liner, .... 
 
 85. Cellular double bottom. 
 
 PAGE 
 
 67 
 67 
 69 
 70 
 
 71 
 71 
 
 76 
 
 79 
 
 80 
 
 82 
 
 83 
 
 85 
 
 86 
 
 87 
 
 87 
 
 88 
 
 91 
 
 97 
 
 110 
 
 to face 112 
 
 to face 122 
 
 to face 122 
 
 . 122 
 
 127, 128 
 
 . 130 
 
 to face 132 
 
 to face 134 
 
 136 
 
 to face 136 
 
 to face 138 
 
 . 139 
 
 to face 140 
 
 to face 140 
 
 to face 140 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 to face 144 
 
 ". 147 
 
 . 151 
 
 to face 152 
 
 to face 154 
 
 to face 156 
 
 . 156 
 
 to face 158 
 
 to face 158 
 
 . 169 
 
 . 159 
 
 . 159 
 
 . 163 
 
PLATES AND ILLUSTRATIONS. 
 
 Xlll 
 
 Fig. 
 
 86. Side water ballast tanks, 
 
 87, 88. Steamer with side water ballast tanks, profile and deck plans, 
 
 89. Steamer with side water ballast tanks, midship section, 
 
 90. Forms of rivets, 
 
 91. Tap rivet, 
 
 92. Methods of riveting, . 
 
 93. Butt lap and strap connections, 
 
 94. Riveting of outside shell plating, 
 
 95. Angle bar, 
 
 96. Keel blocks and launching ways, 
 
 97. Security of keel blocks, 
 
 98. Beam camber mould, . 
 
 99. Stern frame connection to solid bar keel, 
 
 100. ,, ,, to side bar keel, 
 
 101. Frame bending, 
 
 102. Bevelling of frames, . 
 
 103. Floor plate, .... 
 
 104. Floor at bilge, 
 
 105. Lugs upon floors for keelson connection, 
 
 106. Bulb angle frames, 
 
 107. Inadvisable method of fitting reversed frame, 
 
 108. Bulb angle beam, witli welded knee, . 
 
 109. ,, ,, with bracket knee, 
 
 110. Side elevation of hatch coaming plate, 
 111-113. Bulb plate, knee and bracket, 
 
 114. Beam cut for companion or hatchway, 
 
 115. Section through hatchway, 
 
 116. Strong beam welded knee, 
 
 117. Semi-box beam, 
 
 118. Butterly bulb welded knee, 
 
 119. 120. Methods of making connections, 
 121. Supporting steel decks, . . , 
 122-124. Connections of pillar heads, hold beams, etc., 
 125-128. ,, ,, tie beams lie, 
 
 129. Pillar foot on inside plating, . 
 
 130. ,, connection with steel decks, 
 
 131. 132. Arrangements of pillar feet, 
 
 133. Centre keelsons on ordinary floors, 
 
 134. Intercostal centre keelsons, 
 
 135. Centre through plate keelsons, 
 
 136. Side keelsons, . . . 
 
 137. Side stringers, 
 137«. Stringer cut at a bulkhead, . 
 
 138. Steel bulkhead and detail sketches, 
 
 139. Connection of bulkhead to shell, 
 
 140. Bulkhead diamond liner, 
 
 141. Bulkhead watertight door, 
 
 142. Manipulation of watertight doors, 
 
 143. Plan of upper deck of vessel (figs. 12 and 46), 
 
 144. Expansion of outside shell plating, 
 
 145. Mode of obtaining close work at bulbs, 
 
 146. Stem frame for single screw streamer, 
 
 147. 148. Fitted pintle, 
 
 149. Pintles forged on rudder post, 
 
 150. Locking pintle, 
 
 to 
 
 PAGE 
 164 
 
 to face 164 
 to face 164 
 169 
 170 
 171 
 172 
 ^face 176 
 177 
 179 
 179 
 181 
 182 
 184 
 187 
 187 
 189 
 190 
 191 
 192 
 193 
 195 
 195 
 197 
 198 
 199 
 200 
 201 
 201 
 202 
 204 
 204 
 205 
 206 
 207 
 207 
 208 
 209 
 210 
 211 
 211 
 212 
 213 
 to face 216 
 218 
 218 
 220 
 221 
 to face 224 
 to face 226 
 229 
 232 
 235 
 236 
 237 
 
XIV 
 
 PLATES AND ILLUSTRATIONS. 
 
 Fig. 
 
 151. Bottom pintle, 
 
 152. Rudder stopper, .... 
 
 153. Connection of struts with stern frames, 
 
 154. 155. Modes of strengthening ships at after end and attaching struts 
 
 plating, ..... 
 
 156. Struts carried through shell plating, . 
 
 157, 158. Stei'u frames, .... 
 
 159. Frame rudder, .... 
 
 160. Plate rudder, ..... 
 
 161. Cellular bottom discontinued under boilers, . 
 
 162. Steel mast, ..... 
 
 163. Telescopic mast, .... 
 
 164. Hinged top mast, . . . . 
 
 165. Derrick upon ventilator, 
 
 166. Derrick tables on masts, 
 
 167. Outreaches on masts, . . . . 
 
 168. Cast iron socket for derrick, . 
 
 169. Composite deck house, 
 
 170. Insulation, ..... 
 
 171. Ventilators, ..... 
 171a. Ship's hand pump, .... 
 
 172. Sluice valve, ..... 
 
 173. Mouthpiece for suction pipe, . 
 
 174. Pumping arrangement, 
 
 175. Bilge suction valve chest, 
 
 176. Diagram showing straight and cambered launching ways, 
 
 177. Launching diagram, .... 
 
 178. Plate shoe on bar keel, 
 
 179. Hinged bilge shutters, . . 
 
 PAGE 
 
 237 
 238 
 239 
 
 to shell 
 
 240 
 241 
 242, 243 
 
 244 
 
 245 
 
 249 
 
 249 
 
 251 
 
 252 
 
 253 
 
 to face 254 
 
 to face 254 
 
 . 254 
 
 . 254 
 
 . 256 
 
 . 258 
 
 to face 260 
 
 . 261 
 
 . 262 
 
 to face 262 
 
 264 
 
 . 266 
 
 . 268 
 
 . 275 
 
 . 283 
 
STEEL SHIPS. 
 
 CHAPTER I. 
 IRON AND STEEL. 
 
 Iron and Steel supersede Wood in Shipbuilding — Composition of Iron and Steel — Pig 
 Iron, its Nature and ManuPacture — The Blast Furnace — Castings — Malleable Cast 
 Iron — Malleable or Wrought Iron ; its Nature and Manufacture — The Paddling 
 Furnace — Qaalit}' and Classification of AVrought Iron — Steel, its Nature and Com- 
 position — Siemens Steel, Manufacture of — Bessemer Steel, Manufacture of — Basic 
 Steel — Cementation Steel — Case-hardened Steel — Crucible Steel — Steel Castings — 
 Forgings — -Iron and Steel Sections used in Shipbuilding. 
 
 Iron and Steel supersede Wood in Shipbuilding. — The day of wooden 
 ships has practically gone. Travelling through our modern shipbuilding 
 districts, one can scarcely fail to be struck by the conspicuous absence of 
 this material. Whereas, sixty years ago, wood was the principal con- 
 structive element in shipbuilding, it is now a rare sight indeed to see in 
 this country the hull of a seagoing craft being built of this material. Even 
 then it is only used in the construction of the smallest types of vessels, or 
 else only as a composite part of larger ones. 
 
 This great and rapid transformation has all been brought about by the 
 introduction in shipbuilding, first of iron, and subsequently of steel. The 
 manufacture of iron has been carried on for thousands of years, and great 
 skill evinced in its production, while the uses of steel have been understood 
 and appreciated for centuries. It is only since about the year 1860, 
 however, that the latter could be produced in sufficient quantities and of 
 the requisite quality for its adoption in the construction of ships. Even 
 then it was not until after the year 1880 that steel became extensively 
 employed in the building of ships for the mercantile marine. At the 
 present time over 99 per cent, of the vessels built in this country are 
 constructed of this material. 
 
 While, generally speaking, iron has given place to steel, it will be shown 
 more definitely in a subsequent chapter that it still finds favour for certain 
 purposes on account of important qualities which experience has shown it to 
 possess. Hence, to a limited extent, it is still employed as a constructive 
 element in certain parts of a ship. 
 
 1 
 
2 STEEL SHIPS. 
 
 It is in strength, toughness, and malleability, that steel, as now manu- 
 factured, shows its vast superiority over wood, and the possession of these 
 qualities accounts for its having almost entirely superseded that material in 
 the art of shipbuilding. 
 
 While the principal aim of this work is to describe the construction 
 and the means which may be adopted for the maintenance or preservation 
 of steel ships, it will not be out of place to convey some information 
 respecting the process of the manufacture of steel, particularly noting those 
 constituents of the metal whose presence or absence confers, in a marked 
 degree, the qualities of malleability, ductility, weldability, hardness, soft- 
 ness, brittleness, toughness, cold-shortness and red-shortness and strength. 
 
 Although in a steamer the hull is principally composed of mild steel 
 plates and bars, made by the Siemens process* (a process which takes its 
 name from Sir William Siemens, the inventor of a system of producing 
 steel which has proved especially suited to the requirements of ship con- 
 struction), yet iron, in practically every form in which it is known com- 
 mercially, is employed, to a greater or less extent, in the construction and 
 equipment of a ship. We thus find, in addition to mild steel and steel 
 castings, that iron castings, malleable iron plates, bars and forgings are also 
 employed. The great bulk of plates and angles, bulb-plates and bulb-angles, 
 tees, channels, and Z bars, etc., used in shipbuilding, are, however, made of 
 Siemens steel. 
 
 To embrace the whole subject of the metallurgy of steel and iron in what 
 must necessarily be a few pages in a work of this kind, would be an utter 
 impossibility. However, notwithstanding the enormous dimensions of the 
 subject, and the complicated processes attending the manufacture of iron 
 and steel, we shall endeavour to convey to the reader unacquainted with the 
 subject some knowledge of the manufacture, composition, etc., in as intelligible 
 0, manner as the brevity of the treatment will permit. 
 
 Composition of Iron and Steel. — As is v/ell known, steel and iron are 
 products of iron ore which is obtained, like all other minerals, from the 
 earth by luining, and which, after being subjected to a course of treatment 
 whereby most of the associated impurities are separated, yields iron or 
 steel according to the nature and amount of the foreign elp-ments still 
 remaining in combination with the pure iron in the final product. There are 
 many substances whose natures can be easily described by a simple defini- 
 tion. Such is not the case, however, with iron and steel as they occur com- 
 mercially. N'either of them are pure metals. Indeed, absolutely pure iron, 
 in addition to the practical impossibility of its production upon a large scale, 
 is worthless for any industrial purpose. It is to the presence and pro- 
 portions of the foreign elements or impurities which are introduced or 
 found naturally in combination with the pure metal, that the special 
 qualities which distinguish the alloys known as cast iron, malleable or 
 
 * A small proportion of chequer plates, boiler plates, channels, etc., is still made from 
 Bessemer steel. (Seep. 14.) 
 
IRON AND STEEL. 3 
 
 wrought iron, and. steel from each other, are really due. Iron ore, in its raw 
 state, contains from about 30 to 70 per cent, of iron, according to the 
 nature of the ore. Pig iron, the result of the first stage of the manufacture, 
 contains from about 90 to 95 per cent, of pure iron ; malleable or wrought 
 iron over 99 per cent. ; mild steel ship plates about 99J per cent. ; while 
 other varieties of hard steel may contain considerably less of the pure metal, 
 and a larger proportion of other elements. It will thus be seen that while 
 neither iron nor steel are absolutely pure, the associated elements, Avhether 
 metallic or non-metallic, are exceedingly small in proportion to the pure 
 metal, and yet it is to the influence of these elements that the vast differ- 
 ences between cast iron, malleable iron, and steel, as to tenacity, ductility, 
 and malleability, are attributable. 
 
 The chief foreign elements found in commercial iron and steel are 
 carbon, sulplmr, 'phosplionis, manganese, and silicon, though infinitesimally 
 small quantities of other substances may sometimes be present. Any one 
 of the above substances, if in excess, may produce a deleterious influence, 
 and confer an undesirable quality on the metal ultimately produced. The 
 most important of these elements is carbon. 
 
 In pig iro7i, and likewise in cast iron, carbon is found in relatively 
 large proportions, reaching sometimes to 5 per cent., and constituting, as a 
 rule, not less than 3 per cent, of the total. Such metal may be hard and 
 sometimes exceedingly brittle, or soft and tough, but these qualities depend 
 less upon the amount of carbon present than upon the condition in which 
 it occurs. Such iron readily melts at a high temperature, and can be 
 moulded into various forms. In its crude form it is chiefly used in the 
 production of malleable iron aud steel. 
 
 Malleable or tcrouglit iroji may contain traces only of carbon, and 
 at the most it rarely contains more than about •12 per cent. At one 
 time it was customary to define wrought iron as, chemically, the purest 
 iron that could be produced for commercial jiurposes, but with the great 
 improvements which have taken place in recent years in the manufacture 
 of steel, it is outrivalled in this respect by mild steels produced in the 
 processes introduced by Sir W. Siemens and Sir Henry Bessemer, so 
 that the distinction between wrought iron and mild steel lies more in 
 the mode of production than in chemical composition. Wrought iron is 
 comparatively soft, malleable, ductile, and tenacious. It is excellently 
 adapted for welding purposes, its malleability increasing as the temperature 
 is increased. Fusion is only effected at very high temperatures. When 
 heated, and suddenly cooled it retains its softness. It may be produced 
 direct from the ore, but is usually made from pig iron. 
 
 Mild Steel contains sometimes as little as "06 to "15 per cent, of 
 carbon, along with small proportions of other elements. Such steel is 
 malleable, ductile, and tenacious, and when low in carbon and comparatively 
 free from other impurities, it is equal to wrought iron in its welding 
 qualities, Siemens mild steel ship plates contain from '12 to '2 per cent. 
 
4 STEEL SHIPS. 
 
 (not more) of carbon. Steel containing over '2 per cent, of carbon is not 
 adapted for the manufacture of such plates, being too hard. The greater the 
 proportion of carbon, the lower the temperature at which fusion occurs. 
 Thus malleable iron requires a higher temperature for welding purposes 
 than all except the very mildest qualities of steels. 
 
 A most important property of steel containing more than a certain 
 percentage of carbon is that it is capable of being hardened or softened by 
 the processes of tempering and annealing. In order to harden the metal, it 
 is heated and suddenly cooled in oil or water, its final hardness depending 
 largely upon the amount of carbon it contains. By reheating it and allowing 
 it to cool slowly, almost any degree of softness required may be obtained. 
 
 Sulplmr, even in infinitesimally small quantities, assumes, in combination 
 with iron or steel, great importance. It is commonly found in iron ores, 
 and combines with the metal on being subjected to heat. When present in 
 excess it produces a most objectionable feature termed red-shortness. Such 
 metal is useless for forging purposes, being defective in malleability, and 
 Avhen over '12 per cent, of sulphur is present, it is 'rotten' at a red heat, 
 and will, therefore, neither stand rolling nor hammering. Its presence in 
 iron may result from using sulphurous ores, or it is introduced by the 
 fuel employed in the blast furnace.* Small quantities of sulphur are 
 supposed to be of some advantage in some kinds of castings, producing 
 increased tensile strength, but for both malleable iron and steel plates 
 and forgings the smaller the proportion present the better the metal.. 
 The connnoner kinds of steel contain on an average "05 to '08 per cent, 
 of sulphur. 
 
 Phosphorus, when present in iron to the amount of '125 per cent., 
 produces cold-shortness, that is to say, the metal is unable to withstand 
 working and hammering at ordinary temperatures, owing to its reduced 
 tenacity. It then cracks at the edges and breaks readily. Phosphorus 
 unites readily with iron when subjected to a high temperature. It is 
 present in most ores. When heated to redness, cold-shortness disappears, 
 and the metal may be hammered or rolled with facility. When malleable 
 iron or steel contains very small quantities of phosphorus, the metal is 
 harder, while the tenacity may be scarcely affected. There are very few 
 purposes for which more than -1 per cent, is permissible, the tendenc}^ where 
 large proportions are present, being to reduee the tensile strength. 
 
 The effect of phosphorus may be modified by the presence and amount 
 of other impurities. One advantage produced by a considerable proportion 
 of phosphorus in cast iron is that it adds fluidity to the metal in a molten 
 state, and is thus well adapted for very thin castings or for light ornamental 
 work which has no stresses to bear; but for large heavy castings, where 
 strength is required, such a metal would be very brittle and hence most 
 unsuitable. Phosphorus is naturally found in greatest proportion in pig 
 
 * The large furnaces in whicli the ores, fuel, and fluxes are treated preparatory to 
 the production oi pig iron. (See fig. 1.) 
 
IRON AND STEEL. 5 
 
 iron, but some grades, sucli as hematite, largely used in the manufacture of 
 Siemens steel, contain very little. 
 
 While phosphorus increases tlie tenacity of steel, its tendency to brittle- 
 ness renders it less reliable in resisting sudden stresses caused by impact or 
 shock. Mild steel plates for shipbuilding purposes usually contain from 
 about '04 to '06 per cent, of phosphorus. 
 
 Manganese is always found in pig iron. Steel usually contains from "2 to 
 1 per cent., — ship plates, "3 to '6 per cent. Like carbon, it produces hardness, 
 and increases the elasticity and tensile strength, but reduces the ductility. 
 To some extent, manganese counteracts the red-shortness of sulphur, though, 
 when itself in excess, it produces brittleness. It is most valuable as a 
 neutralizing agent when such impurities as sulphur, phosphorus, silicon, etc., 
 are present. When manganese is present in unusually large proportions in 
 iron or steel, it is supposed to promote or hasten corrosion by sea-water. 
 Specially prepared iron alloys, rich in manganese, and known as spiegeleiseu, 
 ferro-manganese, etc., are most valuable in the manufacture of mild steels, 
 for the reasons already given. 
 
 Silicon is not an invariable constituent of steel. When present in 
 considerable proportions it induces hardness, and also, like phosphorus, 
 cold-shortness, which shows itself in the brittleness of the metal. At 
 ordinary temperatures, such steel will neither stand rolling nor hammering. 
 The usual percentage of silicon in steel is from about "02 to "06, with 
 a limit of about "OS per cent. Steel ship plates contain from '03 to "06 
 per cent. It is found in a greater or less degree in all ores. The 
 higher the temperature the more readily does silicon combine with the 
 metal. At a white heat, silicon is very liquid. Its greatest value is 
 in foundry work. While iron comparatively free from silicon is hard 
 and brittle, the introduction of a moderate proportion of specially pre- 
 pared iron, rich in silicon, induces softness, strength, and fluidity. 
 When the silicon is added iu excess, it again produces hard, brittle, 
 and weak castings, with reduced tensile and compressive strength. The 
 result of an adjusted percentage of silicon is the production of sound 
 castings remarkably free from blow-holes, and soft enough to be workable 
 with tools. 
 
 Nickel is a metal whose presence is of the greatest value in steel for 
 certain purposes, though its importance has only been established in com- 
 paratively recent years. It will doubtless be more extensively used in the 
 future, especially in the manufacture of engine forgings and shafting. Like 
 carbon, it increases the hardness and adds greatly to the tensile strength 
 and toughness of the metal. These qualities are further improved by the pro- 
 cess of tempering with subsequent annealing. For marine engines, where high 
 speed combined with lightness and strength are the most desirable factors, 
 nickel steel is being more and more i;sed. For such forgings it is usually 
 introduced iu the proportion of from 3 to 5 per cent. For ordinary tramp 
 steamers of low speed it is not as yet much used, the expensiveness of 
 
O STEEL SHIPS. 
 
 the metal being against its general adoption. Nickel is not a constituent 
 in the composition of the steel employed for ship plates and bars. 
 
 The increased toughness of nickel steel makes it specially suitable for 
 armour plates, as it offers great resistance to cracking by shock, or impact 
 from shot, and also to perforation. 
 
 Pig Iron, Cast Iron, and I^Ialleable Iron. 
 
 Pig Iron, — Pig iron differs from malleable iron and from steel in being 
 unforgeable, and therefore nnweldable ; in its extremely low ductility ; and 
 chiefly in the large amount of impurities it contains. 
 
 Production. The Blast Furnace. — The actual smelting, or reduction of 
 iron ore to a liquid state suitable for casting into the form of ' pig,' is 
 effected in the blast furnace. The shape of these furnaces varies in 
 different localities, as also do their heights, diameters, and capacities. In 
 the Cleveland district in Yorkshire, where a very large proportion of the 
 iron and steel used in shipbuilding in this country is manufactured, they 
 have increased rapidly in size. Fig. 1 shows a Cleveland blast furnace. 
 It consists of an outer shell of wrought iron plates riveted together, and 
 resting upon a ring which is supported upon cast iron pillars. It is lined 
 with fire-brick slabs, and closed at the top with a ' cup and cone ' or ' bell- 
 cone ' arrangement. The waste gases are led away at the top of the furnace 
 by means of a vertical pipe, or 'downcomer.' In some cases blast furnaces 
 reach as much as 95 ft. in height. When the blast furnace is in full working- 
 order, or 'in blast,' it is filled to the top with a mixture of fuel, ore, and fluxes. 
 
 A flux is a material introduced into the blast furnace for the purpose of 
 combining with the earthy matter attached to the ore, — much of which is 
 otherwise almost infusible at the usual furnace temperatures. By this 
 means a fusible substance known as slag is produced, which is one of the 
 chief agents in carrying off the impurities. Thus, in order to effect the 
 reduction of charges in which much silicon is contained in the ore, an 
 abundance of limestone is generally introduced as a flux. 
 
 The furnace charges are always introduced at the top, additions being 
 made frequently to keep the~ furnace full. These charges are taken up 
 to a platform at the top of the furnace, and dropped on to the bell lid, 
 which is so balanced as to sink beneath the superposed minerals. As soon 
 as the material slides into the furnace, the lid rises again, and closes np 
 the furnace mouth. The flare of flame so commonly seen at the summit 
 of blast furnaces is caiised by the escape of combustible gases whenever 
 the bell is lowered. While the charge at the top of the blast furnace is 
 thus composed entirely of solid matter, all the materials introduced pass 
 away either in the form of gas, slag, or molten metal. The chief aid to 
 combustion in the furnace is the introduction of a powerful air blast which 
 may be hot or cold.* 
 
 * As at present practised, these terms, hot and cold, must be regarded as relative. 
 In hardly any case is the cold blast introduced without some preliminary heating. 
 
IRON AND STEEL. 7 
 
 This blast enters near the bottom of the furnace, and blows incessantly, 
 excepting when fresh charges are being introduced into the furnace, and 
 also during the time occupied in running out the molten metal, or, as it 
 is called, ' tapping.' 
 
 The very high temperature produced in the blast furnace causes the 
 molten iron to readily unite with a considerable amount of the carbon from 
 
 Fio. 1.— Section of a Cleveland Blast-Fiirnace. 
 
 the incandescent fuel in the furnace. The melted iron descends towards 
 the base or hearth of the furnace, along with slags produced by the 
 combination of the earthy substances in the ore with the fluxes. In its 
 downward course the iron absorbs more carbon by contact, and finally 
 accumulates in the health of the furnace, a compound of iron, carbon. 
 
8 STEEL SHIPS. 
 
 sulphur, phosphorus, silicon, and other impurities. Owing to its greater 
 specific gravity, the molten metal sinks to the bottom and the slag floats 
 on the top. The passage of the blast into the furnace near the bottom, 
 laden as it is with oxygen, would decarburize the metal, were it not for 
 the protecting influence of the slag floating above it. As the slag continues 
 to increase on the surface of the steadily accumulating metal, it at last 
 reaches a height where there is an aperture in the furnace side (the slag 
 notch), out of which it flows. By a variety of means it is collected, and 
 taken away at intervals. The process of smelting in the blast furnace 
 results in the production of highly carburized metal. 
 
 The more phosphorus the ore contains, the greater the proportion 
 which passes into the slag. However, many ores contain less than 1 per 
 cent., and then it practically all passes into' the iron. If manganese be a 
 constituent of the ore, Avhich is generally the case, it is found partly in 
 the pig iron, and partly in the slag. Pig iron containing about 5 per 
 cent, of manganese and upwards is called spiegeleisen, but when above 
 40 per cent, is present, it is known as ferro-manganese. This is largely 
 used in the subsequent manufacture of mild steel. Sulphur is to a con- 
 siderable extent eliminated in the preliminary process of calcination.* 
 However, a quantity may still find its way from the ore and fuel into the 
 product. As silicon is only reduced at the very highest furnace tempera- 
 tures, a comparatively small proportion only of the total passes into 
 the metal, the greater proportion passing into the slag. It follows, 
 therefore, that a more siliceous metal is produced from hot-blast than in 
 cold-blast furnaces. By means of a basic slag (one rich in lime), silicon 
 is still further eliminated. Highly siliceous iron, which has special uses 
 in foundry work, is only produced at the expense of a large expenditure 
 of fuel and at a high temperature. 
 
 The richer the ore, the less limestone is necessary in the charge, and, 
 consequently, less slag is produced, and the fuel consumption is smaller. 
 
 The blocks of metal called pigs, of SM//Mk section, and about 3 ft. in 
 
 length, are obtained by preparing a damp bed of sand in front of the 
 furnace, in which grooves are made by wooden ' patterns ' corresponding 
 to the shape of the pigs. By means of a channel in the sand the molten 
 metal is led from the furnace to these' grooves. 
 
 Castings. — While pig iron is extensively used in the manufacture of 
 wrought iron and steel, a large quantity is also used in the production of 
 castings in the foundry. 
 
 Though rolled iron and steel preponderate as the constructive elements 
 in modern ships, still, a survey of any vessel will reveal many fittings made 
 of cast iron — bollards or mooring bits, hawsepipes, boat-davit sockets, 
 
 * Calcination is a process of roasting to wliicli certain kinds of ores are subjected, by 
 which means, combined with the effect of the atmosphere, water, sulphur, carbonic 
 acid, and other volatile matters are wholly or in part driven off. 
 
IRON AND STEEL. 9 
 
 fair leaders, the columns supporting engines, cylinders, and many other 
 fittings on the hull and in the engine room. In smelting pig iron for 
 castings, special care is required in selecting and combining the most 
 suitable brands or grades. 
 
 Several reasons may be given for the extensive use of cast iron. It can 
 be melted at a comparatively low temperatur-e as compared to that required 
 in working the metal in some of its other forms. It is tlie cheapest form in 
 which iron is produced. Iron is specially adapted for castings, being very 
 fluid and capable of receiving clear, definite, and exact impressions from the 
 mould. It possesses enormous crashing strength. The tensile strength of 
 cast iron varies from about 5 to 15 tons per square inch of section, the average 
 of good cast iron being about 8 tons. The transverse and torsional strength 
 varies between 1 "5 and 4*5 tons per square inch. Its average shearing strength 
 is about 12 tons to the square inch. The crushing strength of ordinary case 
 iron lies between 25 and 60 tons ; the average being from about 40 to 45 tons. 
 The tensile strength averages about one-sixth of the crushing strength. 
 It will thus be clear that while cast iron is admirably adapted by its great 
 strength under a compressive stress for such purposes as columns and 
 struts, it is totally unfit for use where sudden and severe torsional, tensional, 
 or transverse stresses, such as experienced by the hulls of ships, have to be 
 borne. 
 
 As already stated, cast iron is superior to wrought iron for compressive, 
 but greatly inferior for tensional stresses. It is therefore unfit for ship- 
 hold stanchions or pillars which have to act both as stmts and ties. 
 
 In choosing pig iron for castings, it is necessary to take into account 
 the nature of the casting and the purpose for which it is required. There 
 will thus be a vast difference in the quality of pig iron used for engine 
 cylinders, machinery, girders for structural purposes, etc., and that required 
 for light ornamental work. For the latter, fluidity would be an important 
 quality, and thus phosphorus and silicon would be welcome constituents 
 of the metal. But phosphorus, when in excess, produces brittleness, cold- 
 shortness, and weak castings, and therefore while this feature would make 
 no practical difference in the ornamental castings, it would be an objection- 
 able constituent in iron for structural purposes, where severe, and, probably, 
 sudden stresses would be experienced. 
 
 In remelting pig iron for important castings, it is the practice to 
 combine iron of several grades, and by this means a stronger metal is 
 obtained, the aim generally being to get a compact grey iron. The metal 
 may be further increased in strength and toughness by melting with it 
 either wrought iron or steel scrap. A most important element in the 
 production of good castings is silicon, when not present in quantities ex- 
 ceeding 2'5 per cent. Its value in foundry work has already been described 
 on page 5. Remelting soft cast iron makes it harder and stronger, but 
 remelting hard iron makes it harder and more brittle, weak, and unwork- 
 ahle with tools. 
 
lO STEEL SHIPS. 
 
 Malleable Cast Iron. — Castings made from pig iron are both hard ^aiid 
 brittle, and are workable with neither file nor chisel. This may be 
 remedied by a process somewhat resembling annealing, and by means of 
 which the metal is made to approximate more to malleable iron. The 
 castings are embedded in powdered red hematite in cast iron ' boxes ' and 
 placed in furnaces. In the operation which follows, silicon, sulphur, 
 manganese and carbon are removed from the surfaces. Such castings are 
 capable of being, to some degree, forged and chiselled, and are rendered 
 soft, tougher, and stronger, though they are incapable of being welded. 
 This process is sometimes applied to machinery and propellers, rendering 
 the latter especially less liable to breakage by blows or accidents. In 
 large castings, only the outer surface is affected by this treatment, hence 
 the necessity of choosing pig iron as free as possible from phosphorus, 
 sulphur, silicon and manganese. 
 
 Malleable or Wrought Iron. — The distinctive features of malleable or 
 wrought iron used to be its small percentage of carbon and its almost 
 complete freedom from phosphorus, sulphur, silicon, and other foreign 
 elements. But with the vast improvements which have been made in the 
 production of mild steel in recent times, the special qualities of 
 malleable iron, viz., malleability, ductility, weldability, and tensile strength, 
 are found to an almost equal extent in mild steel. 
 
 Considerable difficulty, therefore, presents itself in attempting to define 
 the difference between malleable iron and mild steel. The difference, as 
 was stated on an earlier page, must be attributable to the actual method 
 of production, malleable iron being the terra applied to the production of 
 iron chiefly by the process of puddling. A distinct difference between 
 wrought iron, cast iron and cast steel is, that while the former is fibrous, the 
 two latter are not, but present granular appearances, when fractured. Cast 
 iron, moreover; is hard and brittle Avhen cold, while wrought iron is soft 
 and ductile. It differs from the harder kinds of steel in the fact that it 
 does not temper well, i.e. if heated in a furnace and plunged into water 
 or oil, it shows little sign of hardening. This is a distinct characteristic of 
 hard steels, though mild steels, such as ship plates, resemble wrought iron 
 in being practically incapable of being tempered. It is scarcely necessary to 
 add that only iron in its malleable form can be of use to the ship's smith. 
 
 Malleable iron is obtained indirectly from pig iron by ^ puddling,^ 
 whereby most of the impurities are got rid of, and the metal is delivered 
 from the puddling furnace in a pasty condition, ready to be hammered and 
 rolled into the required sections or shapes. 
 
 Production. Puddling Furnaces. — What is known as the ordinary single 
 puddling furnace is a long horizontal furnace at one end of which is the 
 fire and at the other end the flue from whence the smoke and heat escape 
 to the chimney, or else are economized by being led away for reheating 
 purposes. Between the fire and the flue is the puddling bed, upon which 
 the charge is laid, being introduced through an opening in the side of the 
 
IRON AND STEEL. I I 
 
 furnace. In this system the pig iron is charged into the bed of the furnace 
 entirely away from the fire grate upon which the fuel is consumed, but in 
 such a position as to be exposed to the action of the flame and draught. 
 After the charge has been made and the five replenished, the furnace is 
 entirely closed. In a short time the metal begins to soften, and the 
 puddler then introduces a long iron bar through an opening in the charging 
 door, and 'puddles' or turns over, or works the pigs so that they are 
 uniformly heated. 
 
 Ultimately the iron is reduced to a more or less fluid state, and spreads 
 itself out over the furnace bed, where it is stirred in order to expose it 
 uniformly to the action of the heat. After this operation has been con- 
 tinued, so that the carbon of the iron has been more or less eliminated by 
 the influence of the oxidizing atmosphere passing through the furnace, the 
 iron begins to stiffen, and spongy masses of malleable metal begin to 
 appear amongst the slag, which has gathered on the surface of the fluid 
 metal. The puddler works up the spongy masses of iron into balls, 
 weighing about 60 or 80 lbs., by means of his iron bar. When these have 
 been made, and are sufficiently coherent, the balls are drawn out of the 
 furnace separately, and carried to the forge hammer, where they are 
 'shingled' or hammered. This operation again rids the iron of much 
 impurity, which is seen to fly out as slag, with every blow of the hammer. 
 When the iron has been formed into a consolidated block, or 'bloom,' it is 
 taken to the rolls, where it is rolled into lengths of bar iron. In the 
 several phases through which the metal passes between entering and 
 leaving the furnace, silicon, manganese, phosphorus, carbon, etc., are all 
 more or less eliminated. 
 
 Quality and Classification. — The quality and grade of the iron is 
 largely determined by the subsequent mechanical treatment it receives. 
 Thus the long length of rolled bar made from the bloom is termed No. 1 
 quality, or merchant bar. To obtain ordinary malleable iron of ISTo. 2 
 quality suitable for smiths' use, No. 1 quality is cut up into short lengths, 
 arranged into piles, and placed in a reheating furnace, from which it is 
 taken and again rolled. To obtain No. 3, or best iron, a pile of iron is 
 again formed for reheating. It may consist entirely of short lengths of 
 No. 2, or of No. 1 with the top and bottom plates only of No. 2, and after 
 reheating, is again passed through the rolls. This quality is known as best 
 merchant iron, or treble best, and is admirably adapted for welding, 
 possessing the qualities of toughness and ductility in a more marked degree 
 than any of the lower grades. Ordinary ship-hold stanchions, or pillars, 
 are usually made of No. 2 quality.* 
 
 Just as repeatedly remelting cast iron may, in the earlier stages, improve 
 the metal, but ultimately produce weak and brittle castings, so on 
 reheating wrought iron the strength and ductility of the metal is improved 
 
 * The names vary greatly in different districts. Thus, in some, No. 1 is known as 
 best, No. 2 as best best, and No. 3 as treble best. 
 
12 STEEL SHIPS. 
 
 by the first few operations. Yet wlien the reheatings exceed five or six 
 times, the metal shows distinct deterioration in quality. Hence, for 
 ordinary purposes, one reheating after the puddled bar has been cut uj) is 
 found sufficient, and when the iron is for smiths' use, the subsequent 
 reheatings tend to the further improvement of the metal. Best best iron is 
 specially suitable for anchors, chains, cables, rivets, etc. 
 
 For the production of iron plates for ships' use, a similar method of 
 piling is employed. A pile is formed by placing the lengths of bars in 
 alternate layers — the bars in one layer being placed across the bars of the 
 layer beneath, the top and bottom being covered by a simple thick slab 
 suitable for the dimensions of the plate it is intended to produce. 
 In a like manner, in order to produce iron tee bars, butterly bulbs, and 
 girder sections used in shipbuilding, the operation of piling is carried out. 
 Special care is required in the building of these piles, for unless the butt 
 or end joints of the bars in the pile be well covered by the next layer, they 
 do not weld properly. Thus it is customary in building the piles for the 
 sections of iron just mentioned, to make the top and bottom flanges of 
 these sections of a single slab of iron usually of a superior quality to the 
 bars used in the centre of the piles. 
 
 The quality of the iron produced depends not only upon the quality of 
 the pig iron first charged into the puddling furnace, but also upon the 
 manipulation of the furnace in regulating the heat and gases. The 
 production of either very soft, or hard, or steely iron, or of iron lacking in 
 tenacity, is due to a considerable extent to the skill or otherwise of the 
 puddler. The more impurity the iron contains, the longer does it take to 
 carry out the puddling operation, and there is, in consequence, a greater 
 waste of metal. 
 
 It need scarcely be repeated that the aim of puddling is to almost 
 eliminate carbon, phosphorus, sulphur, silicon, manganese, and other 
 impurities. To such an extent is this eff"ected in good wrought iron, that 
 the finished metal contains over 99'5 per cent, of pure metal, sometimes quite 
 as much as 99*7, with usually only a trace of carbon,* and correspondingly 
 small proportions of the other~ elements mentioned. To absolutely purify 
 the metal and produce pure iron is not desired, and, moreover, such a pro- 
 duct would be commercially worthless. 
 
 The conversion of bar iron into the required sections for shipbuilding 
 and other uses is effected by rolling in cast iron rolls, which are specially 
 cast and ' turned ' to suit the numerous sections of iron required. 
 
 Steel. — Nature and Composition. — Considerable difficulty attends 
 the attempt to accurately define the milder kinds of steel, as dis- 
 tinct from malleable iron. Malleable iron, as we have seen, is delivered 
 from the furnace in a soft, spongy state. Steel, however, is delivered in 
 a state of fusion, the metal being cast directly from the furnace into a 
 
 * The purest iron, sometimes termed piano wii'e qiiality, will, at times, contain as 
 much as 99*85 per cent, of pure iron. 
 
IRON AND STEEL. 1 3 
 
 mould, thus forming a malleable ingot of sufficient weight to produce the 
 bars or plates required. Cast iron is also delivered from the furnace in a 
 state of fusion, but it differs from steel in its total lack of malleability and 
 ductility. 
 
 An absolute essential in steel is carbon, though it may be present 
 in exceedingly small quantities, ranging from -06 to 1'5 per cent. For 
 Siemens steel ship plates the percentage is from about -09 to -14. While 
 phosphorus, sulphur, silicon, manganese, etc., may also be present, they 
 are all looked upon as impurities, though, under certain conditions, the 
 presence of some of them, in small proportions, may exercise a beneficial, 
 or at least a non-injurious effect upon the steel. An idea of the composi- 
 tion of Siemens steel ship plates will be obtained from the foUoAving figures. 
 
 Siemens I Cavbon. I Silicon. | Sulphur. I Phospliovus. I Manganese. 
 Steel Plates. | '09 to -14 | '03 to -06 | -04 to -06 | '04 to -06 | "30 to '60 
 
 Mild steel is both malleable and forgeable at a full red heat, but owing 
 to the percentage of carbon present, it should be worked at a lower 
 temperature than wrought iron, except in the harder qualities, there being 
 a danger of 'burning' the metal. As previously stated, the property of 
 tempering is possessed only by the harder kinds of steels. Malleable iron 
 is not affected by such treatment, and mild steel, with its small percentage 
 of carbon, is incapable of being tempered to any marked degree. The 
 hardness produced by tempering is in proportion to the carbon in the 
 steel, and the rapidity with which it is cooled. If steel be too much 
 heated, it is said to be burnt, and becomes unweldable. 
 
 Steel, in a greater or less degree, unites the qualities of cast and 
 malleable iron. It may be produced in a number of ways, although 
 ship plates and bars are almost exclusively manufactured by the 
 Siemens process. 
 
 Siemens Steel. — The success which has attended the manufacture of 
 mild steel by the Siemens process, has given it an unrivalled importance 
 for the purposes of shipbuilding. The principal feature in the production 
 of Siemens steel, lies in the reduction and treatment of pig iron and selected 
 ores in what is known as the Siemens open hearth furnace. 
 
 This gave rise to a modification of the process, whereby steel was pro- 
 duced by first melting pig iron in the hearth of the Siemens furnace, and 
 then charging into this bath, wrought iron in the form of scrap, croppings, 
 etc. This process was ultimately known as the Siemens-Martin process. 
 
 The Siemens open hearth process, as now practised and generally 
 adopted in this country, consists practically of a combination of the two 
 fore-mentioned methods. Suitable pig iron is first charged into the hearth 
 of the furnace and reduced to a lii^uid state. Into this, scrap metal from 
 the rolling mills and iron ore of good quality are introduced. Finally, 
 when the metal in the furnace has reached the necessary degree of 
 purification, it is found that it has become almost completely decarburizerl, 
 
14 STEEL SHIPS. 
 
 that is, the action of the furnace has eliminated the carbon, and it is 
 necessary that this essential element, together with a proportion of 
 manganese, should be restored. Hence, spiegeleisen, or ferro-manganese, 
 both of which are rich in carbon and manganese, are charged into the 
 furnace just before the conclusion of the process. 
 
 AVhen the metal is ready to be discharged from the furnace, it is tapped 
 into a ladle, and cast from thence into moulds, visually of cast iron. These 
 castings are known as ingots. In order to convert them into plates or bars 
 of any particular section, they have to be rolled. If the ingots were 
 exposed to the air and allowed to cool they would solidify, but the 
 solidifying operation would not go on uniformly throughout the ingot, as 
 obviously the outside Avould cool and ' set ' more rapidly than the interior. 
 It therefore becomes necessary to either place the ingots in pits specially 
 heated, in which they can come to the solid state suitable for rolling 
 purposes, with a uniform temperature ; or else they are allowed to cool, 
 and are reheated in a special furnace. The ingots are, in either case, 
 taken to the mills, and finally finished by rolling into whatever sections 
 (plates, bars, etc.) may be required. Charges amounting to from 30 to 50 
 tons may be melted in the Siemens open hearth furnace at once, and in 
 America even larger charges are worked. 
 
 Siemens steel is more uniform in quality than that obtained by the 
 Bessemer process, and it is largely owing to this fact that it has become so 
 universally adopted for ship construction. Mild Siemens steel for ship 
 plates and bars possesses a tensile strength averaging 30 tons. It is also 
 very ductile, showing an elongation of at least 20 per cent, in a length of 
 8 in., and is as malleable as wrought iron. The Siemens process is 
 specially suitable for the manufacture of steel for castings. 
 
 Bessemer Process. — In the early days of steel ships, Bessemer steel 
 (named after Sir Henry Bessemer, the inventor of the process of 
 manufacture) was to some extent employed, though now it has almost 
 entirely given place to Siemens steel. It is, however, extensively used, 
 and admirably adapted for the manufacture of rails. The pi'ocess is 
 conducted in a pear-shaped vessel, called a converter, which, in the system 
 most widely adopted, is swung upon an axis, and can be tilted at any angle. 
 This arrangement greatly facilitates the work of discharging the metal 
 from the mouth of the converter. 
 
 Bessemer steel is produced from pig iron which is conveyed in a molten 
 state direct from the blast furnace in a huge ladle, and charged into the 
 converter. 
 
 The predominant feature in this system is, that the pig iron is 
 decarburized, i.e. purified to a greater or less degree from its over-supply 
 of carbon by the action of the oxygen of the air, which is blown from the 
 bottom of the converter through the fluid metal. Ey this means the 
 greater part of the carbon, silicon, and manganese of the pig iron are 
 oxidized and consumed. 
 
IRON AND STEEL. 1 5 
 
 The passage of the air blasts through the molten metal causes violent 
 ebullition, owing to the escape of carbonic oxide resulting from the oxidation 
 of tlie carbon in the pig iron and the oxygen of the blast, combined with the 
 intense heat developed. By the nature of the flame it can be judged when 
 the operation of blowing has been carried on sufficiently, that iss, when the 
 almost complete decarburization of the metal has been attained. 
 
 Towards the conclusion of the ' blow ' in the Bessemer converter, a 
 metal w'ould be produced practically free from carbon, silicon, and 
 manganese, but still containing any sulphur or phosphorus which was 
 present in the original metal. Such a metal would be exceedingly soft 
 and unweldable, and probably both red-short and cold-short, so that when 
 rolled it would crack round the edges, and even crumble to pieces. It 
 therefore becomes necessary, just before the termination of the blow, to 
 furnish the metal in the converter with the necessary proportion of carbon 
 and manganese in the form of spiegeleisen, or ferro-manganese, the 
 manganese counteracting the effect of the sulphur and to a less degree 
 neutralizing the effect of phosphorus. When the metal is finally ready 
 for discharging from the converter, it is cast into ingots in a way similar 
 to that described for Siemens steel. 
 
 When the metal is overblown, it is burnt, and possesses the quality 
 known as ' rottenness.' 
 
 Basic Steel. — The method known as the ' Basic ' process of producing 
 steel may be conducted in either the ordinary Siemens open hearth furnaces, 
 or in Bessemer converters, usually the former. The modification in the 
 furnace for the Basic process lies chiefly in the nature of the lining, which 
 is made of dolomite or magnesian limestone. During the working of the 
 charge of pig iron, a considerable quantity of limestone is introduced into 
 the furnace. This combines with the silicon and phosphorus, and passes off 
 in the form of slag. By this means, phosphoric ores, which would other- 
 wise be unsuited for the manufacture of steel, can be employed. 
 
 The principle of the Siemens open hearth and of the Bessemer process 
 is practically identical in each case. The molten metal is first decarburized, 
 and then, by the addition of spiegeleisen, or ferro-manganese, it is recar- 
 burized to the desired extent. It is decarburized because of the difficulty 
 of ascertaining the precise amount of the impurities present. Then the 
 foreign ingredients are added, in the proportions found by experience to be 
 necessary. 
 
 While steel of good quality may be jiroduced by either of these 
 processes, Siemens steel is more uniform in quality, and better meets 
 shipbuilding requirements. The Bessemer process is admirably adapted 
 for the luanufacture of rails, hut mild steel for ship plates and bars is made 
 almost exclusively by the Siemens process. 
 
 Cementation Steel, Case-hardened Steel, and Crucible Steel. — As 
 cementation steel and case-hardened steel do not play any part in the 
 construction of the hulls of merchant ships, a very brief description of these 
 
1 6 STEEL SHIPS. 
 
 processes must suflfice in passing. (' Harvey ized' and kindred hard steels, 
 used particularly for armour plating in war vessels, may, however, be regarded 
 as forms of ' case-hardened ' steel). 
 
 In producing steel by the cementation process, bars of malleable iron 
 are embedded in charcoal and placed in a furnace, where a high and long- 
 continued temperature is maintained. By this means the carbon of the 
 charcoal permeates the malleable iron. By a somewhat similar process, 
 known as case-hardening, carried out at a great heat, a thin layer of steel 
 may be produced upon the outside of the iron. When a hard surface is 
 required over part of tlie steel, the tempering is done by immersing such 
 parts in cold water or oil after reheating, or, in order to liarden the whole 
 surface, the steel is totally immersed in water. The interior of the case- 
 hardened metal remains wrought iron. AVhere, however, the process of 
 cementation is long-continued, the entire bar of malleable iron may be 
 converted into steel. A peculiarity of the process of cementation is the 
 blistering of the surface of the metal bars, hence this kind of steel is known 
 as 'blister steel.' 
 
 The bars are ultimately cut up into short lengths, subjected to one or 
 more reheatings in piles, welded by hammering, and rolled. By this 
 treatment the qualities of tenacity and strength are considerably improved. 
 The best quality of steel, however, can only be obtained by breaking up 
 the blister steel and melting it in a crucible or pot. This renders it more 
 uniform in its nature. It is then called crucible steel, and is the most 
 valuable steel produced. It is chiefly used for tools, cutlery, and also for 
 special forgings and castings. 
 
 Steel Castings. — On examining steel ingots, we invariably find that 
 they are honeycombed with cavities or holes, probably owing to the escape 
 of gas. This would naturally be a most detrimental feature in important 
 steel castings. Various devices have been proposed to overcome these 
 defects, but probably the introduction of silicon is one of the simplest 
 and most effective. By this means, it would appear that the formation 
 of such cavities and unsoundness is prevented. 
 
 We have seen, however, that silicon in large proportions exerts a 
 detrimental influence on the steel produced, reducing its malleability, 
 ductility, and tensile strength, as well as its adaptability for being 
 hammered. It therefore becomes necessary to somewhat counteract the 
 effect of the silicon, and thus manganese is introduced also. 
 
 The combined silicon and manganese is called silico-spiegel, and is 
 introduced into the molten metal just before the operation of casting. 
 Siemens steel is also used for castings. 
 
 Steel castings should be annealed, that is, they should be heated to a 
 red heat, and. then allowed to cool down slowly, generally embedded in 
 ashes or other substance, which will make the cooling process gradual. 
 This operation renders the casting softer in its nature, and more uniform 
 in strength. Cast steel stern frames are now commonly employed, though 
 
IRON AND STEEL. 1/ 
 
 some difference of opinion exists as to whether they are preferable to 
 wrought iron forgings. 
 
 Forgings. — Among the moving parts of the engines, and also in the 
 hulls of steamers, we find numerous forgings (crauk shafts, stern frames, 
 etc.), which were at one time made exclusively of wrought iron, but are 
 now commonly made of steel. The excellent welding properties of wrought 
 iron make it possible for most satisfactory results to be obtained in the 
 production of large forgings, by the process of welding together the several 
 pieces which go to make up the whole structure. 
 
 . In very large and heavy wrought iron forgings, however, there is always 
 the possible danger of unsatisfactory welding, which, moreover, may not be 
 very apparent on the surface, and only fully revealed after the forging has 
 fractured, owing to severe stresses. 
 
 In the case of mild steel forgings, the operation of 'forging' consists 
 purely in the rolling and compression to which the steel is subjected ; 
 and the metal from which crank shafts, etc., are made, is practically 
 identical with that of any other mill products, such as angles, tees, etc. 
 Honeycombing is eliminated in the rolling and under the steam hammer. 
 As a matter of fact, the lower the percentage of carbon the less homo- 
 geneous the steel. Very 'hollow^' ingots usually result when the carbon 
 falls below "07 per cent. Nevertheless, as long as such ingots can be rolled 
 at all, the resulting finished bar will be sound. It is not possible to roll 
 a shaft with cranks at various angles. Large shafts are built up ; the parts 
 being shrunk on and keyed. Crucible steel is unsuitable, owing to the 
 small quantities produced at a time, for anything but the smallest castings, 
 and is not used to any great extent for ship work. 
 
 In important steel forgings the qualities of tensile strength and 
 toughness are further improved by tempering in oil with subsequent 
 annealing, and the thinner the section of the forging, the greater the tem- 
 pering effect. Hence the advantage, in such forgings as propeller shafts, 
 of having a hole through the centre, as this not only reduces the weight 
 of the forging, by removing that part of the material in the shaft which 
 contributes least in affording strength, but facilitates and improves 
 the results of tempering. It may be noted that shafts are often cast 
 hollow. 
 
 The qualities of forgings have been still further improved by intro- 
 ducing other metals in combination with steel, notably nickel, which 
 produces a remarkable improvement both in the tensile strength and 
 elastic limit of steel (see page 21), and also increased power to 
 resist fracture by impact or shock. Nickel also adds to the efficiency 
 of tempering. 
 
 Wrought iron forgings which have been built up by welding several 
 
 pieces together, should always be annealed, for otherwise there may be 
 
 considerable irregularity in the strength, owing to variation in the heats 
 
 and irregularity in cooling. 
 
 2 
 
i8 
 
 STEEL SHIPS. 
 
 Iron and Steel Sections used in Shipbuilding. 
 
 The following is a list of tlie principal sections of steel and iron 
 which are rolled in the ' mills,' and are suitable for the construction of 
 ships. 
 
 L 
 
 Plain plates, used for shell, 
 decks, tanks, stringers, bulk- 
 heads, etc. 
 
 Cliequer plates, for decks, plat- 
 forms, etc. 
 
 Plain angle bars, used for 
 frames, reversed frames, 
 beams, stringers, keelsons, 
 bulkhead stiffeners, and for 
 connections in every part of 
 the vessel. 
 
 Bulb angles. These may be used 
 for frames, beams, bulkhead 
 stiffeners, keelsons, stringers, 
 etc. 
 
 T 
 T 
 
 1 
 
 I 
 
 : 
 
 5. Plain tee bars, used for beams, 
 stiffeners, stanchions, etc. 
 
 6. Butterly bulbs or tee bulbs, 
 
 used chiefly for beams and 
 hold stringers. 
 
 7. Bulb plates, used in conjunc- 
 
 tion with angles for beams, 
 keelsons, stringers, stiff'eners, 
 etc. 
 
 8. H bars or girders, used for 
 
 strong beams and keelsons. 
 
 9. Z bars, used for frames and 
 
 bulkhead stiflfeners. 
 10. Channel bars, for frames, 
 beams, bulkhead stiffeners 
 and stanchions 
 
 11. Half rouiul, convex iron, and flat bar, for mouldings and 
 sometimes for stiffeners of casings in passage-ways where sharp 
 angle stiffeners might be dangerous. 
 
 ,»- 12. A patent section for hatches, combining in efficiency both the usual 
 rest iron for hatch covers and moulding. 
 
 i! 
 
 ■"!r> 
 
 ju' 13. Best iron, which is attached to the inside of hatch coamings to 
 support the covers. 
 
 •o 
 
 14. Bound iron, for pillars, stanchions, etc. This may be solid or 
 hollow. 
 
CHAPTEE II. 
 
 STRENGTH, QUALITY, AND TESTS OF STEEL FOR SHIP- 
 BUILDING PURPOSES. 
 
 DefiuitioDS of Important Terms ; Tensile Strength, Stress, Ductility, Elasticity, Elastic 
 Limit — Value of Nickel —Fatigue — Tests of Plates and Angles — Remarks upon the 
 Reduction in Thickness of Steel Plates — Tests for Steel Castings — Rivet Tests — 
 Treatment of Plates and Bars in Shipyard. 
 
 Definitions of Important Terms. — Having briefly traversed the steps in 
 tlie process of the manufacture, and noted the constituents of iron and steel, 
 it is now proposed to inquire into the special qualities essential for purposes 
 of ship construction, and the tests to which the metal is subjected in order 
 to ensure the possession of such qualities. 
 
 It is advisable at this stage to clearly understand certain terms and 
 expressions which are constantly recurring in treating this aspect of the 
 subject. 
 
 Tensile Strength. — By the tenacity of steel is meant the property of 
 adhesion, and the tensile strength is the measure of the greatest stretching 
 force, applied in the direction of the length of the material, necessary to 
 produce fracture ; or, in other words, it is equivalent to the maximum stress 
 required to produce fracture, and is usually expressed by the number of tons 
 required to break a bar one square inch in sectional area. The remarkable 
 tenacity of steel is one of its most valuable and important characteristics. 
 
 Stress may be defined as the resistance offered by any material to defor- 
 mation caused by the application of an external force. Stress is also equal 
 to the intensity of the force which produces deformation, providing that it 
 is not so great as to cause rupture or fracture (see also page 75, Chapter 
 V. on ' Stress and Strength '). 
 
 Ductility is the property of being permanently elongated by the applica- 
 tion of a tensile force. This property in steel is well illustrated by the fact 
 that it can be drawn into wire. , 
 
 Elasticity and Elastic Limit. — Within certain limits, if a tensile force 
 be applied to a bar of steel, it will show distinct elongation, but immediately 
 upon the removal of this stretching force the material returns to its original 
 dimensions. 
 
20 STEEL SHIPS. 
 
 This elastic quality is termed elasticitij. Elasticity may also be 
 defined as that property whereby, after the metal has been subject to a 
 certain pressure, whether of tension or compression, at a given temperature, 
 it seeks to regain and retain its original volume and shape. If steel, how- 
 ever, be stressed beyond this limit, permanent elongation, or ^ set,'' takes place. 
 This limit, or point, at which elasticity ceases, and permanent set or dis- 
 tortion takes place, is called the ' elastic limit.^ The elastic strength is 
 thus the tensile strength up to the elastic limit. So long as steel is not 
 subject to a stress beyond the elastic limit, the elongation is directly pro- 
 portional to the stress, but as soon as the elastic limit is passed, the elonga- 
 tion takes place at a more rapid rate. It will now be clear that in very 
 ductile materials, such as wrought iron and mild steel, very considerable 
 distortion and elongation beyond the elastic limit may occur before actual 
 fracture takes place. 
 
 Most of the harder steels possess less ductility, though their elastic limit 
 and ultimate tensile strength may be greater. It is obvious to everyone 
 how important these qualities of tenacity, ductility, and elasticity must be 
 in such a structure as a ship. In a large iron or steel vessel rolling and 
 pitching in a seaway, enormous and innumerable stresses are set up in every 
 part of the structure, and owing to the elasticity of the material, though 
 imperceptible to ordinary observation, there is a certain amount of give and 
 take — elongation and retraction — constantly going on, which must somewhat 
 relieve the severity of the stresses which are experienced by rivets and con- 
 nections generally. True, this is not, nor should it be, so serious as to be 
 perceptible, nevertheless, knowing what we do of the numerous stresses borne 
 by ships at sea, and also of the elastic quality of iron and steel, it follows 
 that this actually takes place to a greater or less degree. 
 
 The value of ductility in steel is often manifested in cases of collision, 
 or fouling, or grounding. Steel plates, in such cases, have been known to 
 buckle and bend, and to be greatly twisted and distorted, and yet show 
 no signs whatever of rupture. 
 
 Importance of Elastic Limit. — In considering the strength of steel for 
 ship construction, important as tensile strength may be, the elastic limit 
 is of still greater importance. In fact, in dealing with structures from a 
 purely strength point of view, it is the elastic limit and not tensile strength 
 which is of foremost consideration. And while ductility is of great import- 
 ance, and it might be possible to stretch the steel in a ship considerably 
 beyond the elastic limit but well within the tensile strength, such permanent 
 elongation, excepting under such exceptional circumstances as grounding, 
 collision, etc., is decidedly most objectionable, and could only be productive 
 of ultimate disaster. It can easily be imagined what horrible distortion in 
 the form of a steel ship would ensue, if vessels were built to anything 
 approaching their maximum tensile strength, or even in such a way as to 
 be subject to stresses in excess of the elastic limit. 
 
 Thus the aim of the designer of ships, bridges, etc., after obtaining all 
 
STRENGTH, QUALITY, AND TESTS OF STEEL. 21 
 
 other necessaiy qualities and properties, is to get the maximum limit of 
 elasticity ia the material required for the structure in question. 
 
 Nickel. — 111 recent years, considerable improvement lias been made in the manu- 
 facture of steel forgings for marine engines and also for boiler tubes, by introducing 
 suitable proportions of nickel during the process of manufacture. This has the effect of 
 increasing both the tensile strength and the elastic limit. But as nickel induces hard- 
 ness, and is also very costly, it is not used in the manufacture of steel for ship plates 
 and bars. Nickel in boiler tubes has been found to retard corrosion. 
 
 The tensile strength, ductility, and elasticity of steel are further increased by 
 elongation, produced by cold rolling and wire-drawing at suitable temperatures. This 
 is amply illustrated in the case of steel wire. Steel rods drawn into wire will increase 
 as much as 80 to 100 per cent, and over, in tensile strength. So that steel wire with a 
 tensile strength of 100 tons per square inch is a common production, while even this 
 has been greatly exceeded. 
 
 Fatigue. — -Of course, it is clearly understood that where elongation of 
 the nature just described is carried out, the stretching force is applied 
 most gradually. Suddenly applied irregular stresses, which vary greatly 
 in magnitude, produce what is known as Fatigue, that is, diminished 
 resistance to rupture. This probably accounts for the fracturing of tail- 
 end shafts, which experience severe jerking strains as the propeller blades 
 strike the water after emersion in pitching movements. A similar reason 
 could be found for many other breakages. Such irregular strains un- 
 doubtedly cause disarrangement in the natural molecular disposition in 
 both iron and steel, and microscopic examinations of such fractures show 
 the effects of this disturbance. 
 
 In ordinary mild ship plate steel the elastic limit ranges at about 
 50 per cent, of the tensile strength. Tempered nickel steel may even 
 have a tensile strength of 45 tons and over, with an elastic limit of 30 
 tons, while hard drawn steel wire may reach an elastic limit of 50 tons. 
 
 Tests. — Classification Societies such as Lloyd's, Bureau Veritas, and 
 the British Corporation, agree very closely as to the qualities of iron and 
 steel for shipbuilding purposes, and consequently their tests are generally 
 very similar. Samples of the steel are tested at the steel works (under the 
 personal inspection of the society's surveyors) before the material leaves for 
 the premises of the shipbuilder. These societies require that samples be 
 tested from every charge or cast employed in the manufacture of the 
 material ; and moreover, that the whole operation and process of testing 
 be witnessed by their surveyors. If they are satisfied with the manner 
 in which the steel stands the various tests imposed upon it, then every 
 plate, beam, and angle is required to be clearly and distinctly stamped 
 by the manufacturer. The brand upon the steel so tested is as follows: 
 
 1^ 
 
 (Lloyd's). R C^ British Corporation. 
 
 and indicates that a shearing from the plate or bar has satisfactorily passed 
 through the whole of the tests made upon it. 
 
2 2 STEEL SHIPS. 
 
 When these samples fail to fulfil the test requirements, the plates or 
 angles from which they were cut are rejected. These tests are made so 
 as to ascertain and measure the tenacity or tensile strength of the steel, 
 its ductility and elasticity, and also its transverse or bending strength. 
 Such tests usually comprise hot and cold forge tests, tempering, and, in the 
 case of forgings, the sudden impact caused by either allowing the forging to 
 fall on hard ground, or by dropping weights upon the forging. In testing 
 steel, the classification societies require that strips from a plate, angle, 
 or bulb plate or bar, cut lengthwise or crosswise, should have an ultimate 
 tensile strength of not less than 28 and not exceeding 32 tons per square 
 inch of section, with an elongation equal to at least 20 per cent, on a 
 length of 8 in. before fracture in samples 2"%*'^^^ of an inch and above 
 in thickness, and 16 per cent, in samples below this thickness. The 
 elongation test, it is scarcely necessary to add, is to ensure ductility, 
 and the limit of 32 tons as the maximum tensile strength is to obtain 
 a metal which is soft, and as workable in the hands of ships' platers 
 and smiths as the best wrought iron. Indeed, where sharp bends are 
 required, as in garboard strakes for bar keels, and in the plates round 
 the stem and counter and stern frame (boss and oxter plates), steel is 
 greatly to bo preferred to wrought iron, as it can be manipulated with 
 better results. 
 
 So easily can steel plates be bent, that it is now a very common 
 practice with shipbuilders to make the corners of engine and boiler 
 casings and deckhouses with a single plate bent to a radius of 2, 3, 
 or more inches, thus dispensing with the usual connecting corner angle 
 bar. Steel plates are also preferable for the round corners of hatch- 
 coamings. So fully recognized is this quality in steel, that even where 
 iron is employed for houses on deck, etc., it is a very common practice 
 for shipbuilders to adopt steel plates for the bends at the corners, etc. 
 
 Classification societies also require that steel angles for the frames of 
 vessels, and bulb steel for beams, may have a maximum tensile strength 
 of as much as 33 tons per square inch of section. 
 
 Strips cut from the plate, angle, or bulb steel, after being heated to 
 a low cherry red, and cooled in water of 82° Fahr., must stand bending 
 double round a curve of which the diameter is not more than three 
 times the thickness of the plate tested. In addition, samples of plates 
 and bars should be subjected to cold bending tests. 
 
 Among the numerous tests adopted, the following may be taken as 
 examples. 
 
 Mild steel cut into strips of about 1|- in. 
 
 ' ) ^ in width, should stand being bent cold in a 
 
 hydraulic press to a curve the diameter of Avhich 
 
 'Pig,.' 2. Cold Bendinc does, not exceed twice the thickness of the plate 
 
 Test. (or inner radius of bend equal to thickness 
 
 of plate). (See fig. 2.) 
 
STRENGTH, QUALITY, AND TESTS OF STEEL. 23 
 
 A plain angle should bend hot as shown in Nos. 2 and 3, fig. 3. 
 
 12 3 
 
 Fig. 3.— Hot Angle Tests. 
 
 A 7 bai- should bend hot as shown in N'os. 2 and 3, fig. 4. 
 
 L 
 
 N r 
 
 J 
 
 Fig. 4.— Hot Tee Bar Tests. 
 
 Turning to wrought iron, we find that the classification societies require 
 that, for vessels classed by them, it must be of good malleable quality, 
 capable of standing a tensile strain of 20 tons per square inch with, and 
 18 tons across, the grain, and of standing certain hot and cold foro-e tests.* 
 
 Remarks upon the Reduction in Thickness of Steel Plates. 
 
 The advantage of mild steel over wrought iron is at once apparent. 
 The tensile strength is increased from 40 to 50 per cent., while the metal 
 shows very decided superiority in elasticity and ductility. As a natural 
 result, steel ships are lighter in weiglit than iron ones, a reduction of about 
 20 per cent, being usually allowed by the classification societies in the 
 thickness of steel plates and angles, and as a more lightly constructed ship 
 permits of a greater deadweight being carried, and thus increased freights 
 being earned, the reason for the universal adoption of steel is obvious, and 
 more especially as the price of steel is usually as low as that of iron. But 
 even with so much to recommend it, there are still certain minor dis- 
 advantages associated with its adoption. This applies to all vessels, but 
 more especially to very small craft. As just noted, a steel vessel with 
 plates and angles 20 per cent, lighter than an iron one is equivalent to 
 the latter in strength. (Plates and angles which would be, say, j^g-ths thick 
 in an iron vessel, become /^ths in a steel ship.) But apart from strength, a 
 certain objection is attached to reducing the thickness of plates. Thickness 
 gives rigidity, and thus in some classes of very lightly constructed steel 
 
 * Should any of the samples tested show signs of failure by cracking or breaking 
 in any way, and not fulfilling the test requirements, the whole of the material repre- 
 sented by such samples is rejected. 
 
24 STEEL SHIPS. 
 
 vessels, the effect of the water pressure upon the immersed surface is 
 sufficient to cause the plating to buckle in between the transverse frames, 
 so that the position of every frame is distinct on the outside surface. 
 Severe pressure upon the bows in high-speed vessels makes this objection- 
 able feature even more pronounced, and while it may be argued that this 
 may not inflict any serious effects upon the stieugth, it certainly detracts 
 from the appearance of the hull. 
 
 In some quarters, steel has been sadly abused for its rapid deterioration 
 owing to corrosion. Probably, in some measure, there is truth in such a 
 complaint against the metal. But it should be remembered that taking 
 the same amount of corrosion iu two plates of equal strength to, say, Lloyd's 
 requirements, one of iron and the other of steel, the steel plate would only 
 at first sight appear to be the worse of the two. But then the thinness of 
 the steel plate is entirely against it in forming judgment. A yV^^^ ^^ * 
 plate xV^hs thick would appear more striking than xV*''^ ^^ ^ plate y^ths 
 thick. It seems, therefore, that while some qualities of steel may, in some 
 measure, show more signs of decay than wrought iron, it should also be 
 remembered in judging the results that the steel was probably thinner to 
 begin with. 
 
 Tests for Steel Castings. — The tests for steel castings in the hulls of 
 ships are generally as follows : — Cast steel stern frames, rudders, steering 
 quadrants, and tillers, must be subjected to percussive, hammering, and 
 mechanical tests, in the presence of one of the society's surveyors, so as to 
 ensure the material being of ductile quality. A tensile test is to be made 
 on a piece taken from each casting, and the extension on a length of 8 in. 
 is not to be less than 8 per cent., and the tensile strength not less than 28 
 tons, nor more than about 35 tons, per square inch. A cold bending test 
 must also be made corresponding to each tensile test, and the sample must 
 bend cold before fracture through an angle of at least 90°. Large stern 
 frames cast in one piece must be allowed to fall on a hard flat ground 
 (excavations being made to take the boss part and other projections) after 
 being raised through an angle of 45°. Stern frames cast in more than one 
 piece, and rudders, must be dropped from a height of from 7 to 10 ft,, 
 according to the design, shape, and weight of the casting. The casting in 
 such case must subsequently be slung up, and well hammered with a sledge 
 hammer, not less in weight than 7 lbs., to satisfy the surveyors that the 
 castings are sound and without flaws, existing either originally, or developed 
 as the result of the application of the preceding percussive tests. 
 
 Rivet Tests. — An equal tensile strength per square inch of section is 
 required in rivets as in plates and angles. They should be of special 
 quality, both soft and ductile, and capable of standing the following tests. 
 One rivet per hundred should be subjected to a forge test. 
 
 1, Temper Test. — The rivet bar should be heated to a low cherry red, 
 cooled in Avater of 82° Fahr., and then bent double round a curve the 
 diameter of which is equal to the diameter of the bar. 
 
STRENGTH, QUALITY, AND TESTS OF STEEL. 25 
 
 2. Forge Tests. — To bend cold without fracture as shown in No. 2, 
 fig. 5, ;f = diameter of the rivet. 
 
 3. To bend hot without fracture, as shown in No. 3, iig. 5, so that 
 tliough the rivet has been nicked with a chisel at x, no signs of tearing 
 are apparent. 
 
 4. The head to be flattened when hot as shown in No. 4, fig. 5, until 
 the diameter is two and a half times the diameter at x. 
 
 Treatment of Plates and Bars in the Shipyard. — Even after the steel 
 has been delivered in the shipyard, a hard biittle plate may often be 
 detected in the operation of punching, the plate cracking round the rivet 
 hole. The nature of the outer surfaces of the punchings also often expose 
 bad material; a cracked creviced surface, together with a severely torn 
 elge, point to the same defect. This is specially noticeable in bad iron 
 plates. 
 
 In rolling a plate cold, the objectionable feature of cold-shortness (cold- 
 brittleness) is sometimes discovered, and in the same operation with a 
 heated plate, hot-shortness (hot-brittleness) may be detected. 
 
 The manipulation of a plate in the hands of the ships' platers may tend 
 
 2 3 4 
 
 Fig. 5.— Rivet Tests. 
 
 to cause deterioration in the quality of the material unless it be subjected 
 to proper treatment. Thus, plates (such as counter plates and boss plates 
 round the tail-end shaft) which have been severely hammered in order 
 to shape them for their particular positions on the hull, and are, moreover, 
 probably subjected to several heatings and coolings, should be annealed 
 before being fixed in position in the vessel, in order to regain the uni- 
 formity in strength which any such severe processes as hammering and 
 uneven cooling destroys. 
 
 Similarly, the severity of the operation of punching the holes for 
 rivets in a steel or iron plate, so destroys the molecular arrangement of 
 the material in the neighbourhood of the holes that, in order to regain the 
 original uniformity of strength, the plates should either be annealed after 
 punching, or else the holes should be rimed out. These features are fully 
 recognized by the classification societies, and we find that important 
 structural items in the material, such as stringer plates, sheer strakes, 
 garboard strakes, and all buttstraps Avhen about i^ths of an inch in 
 thickness and above, are to be carefully annealed, or else the holes are to 
 be rimed after punching. 
 
CHAPTEE III. 
 
 CLASSIFICATION. 
 
 Purpose for which Classification Societies exist — Societies empowered to assign Load 
 Lines — Government the supreme Authority for Assignment of Load Lines, and 
 responsible for Seagoing Condition of Vessels leaving British Ports — Standard of 
 Strength upon which Load Lines are assigned— Load Lines of Three Deck, Spar 
 Deck, and Awning Deck Vessels — Grades of Class — Maintenance of Class — Un- 
 classed Vessels. 
 
 Purpose for which Classification Societies exist. — Al at Lloyd's is a 
 phrase oftener used than fully understood, though every seaman knows that 
 it conveys an idea of the good quality of, and seaworthiness of a ship. 
 
 Such an expression brings us into contact with a subject of paramount 
 importance, viz., ' Classification,' and as a large proportion of seagoing 
 vessels are ' classed,' and the structural strength in new vessels and the 
 manner in which such structural strength has been maintained in 
 old vessels determines the ' class ' to which they belong, it will be 
 well at the outset to obtain a clear idea of this matter of ' classification.' 
 
 A steel or iron ship is often compared, from the point of view of 
 strength, to a beam or girder, and in many respects such a comparison 
 may fairly be made. When iron or steel girders are used in the construction 
 of bridges, buildings, etc., a very close approximation can be made to the 
 nature and amount of the severest strains which may have to be borne. 
 And thus the necessary strengths and dimensions of these girders can be 
 arrived at almost entirely by calculation, provided that the quality of the 
 material is thoroughly understood. But when we come to the actual ship 
 girder, we find ourselves confronted by considerable difficulties and com- 
 plications, which make the determination of the scantlings and disposition 
 of the material in order to ensure sufficient strength, a matter which 
 cannot be ascertained purely by calculation. The innumerable varying 
 and sudden stresses which are experienced by ships in a seaway, when 
 rolling and pitching in light, loaded, or ballast conditions, render accurate 
 mathematical treatment impossible, although, as we shall show, by no means 
 rendering mathematical approximation useless. 
 
 The other factor necessary in arriving at an adequate knowledge of 
 
CLASSIFICATION. 27 
 
 required strength is experience. Thus we find throughout the history of 
 iron and steel shipbuilding, continual changes have been made in the 
 generally accepted rules, as experience has indicated their necessity. 
 
 All ships are built to carry. Some, as in the case of steam yachts, 
 though two or three hundred feet in length, have little else to bear than 
 the owner, his family, a few guests, the crew, bunker coals, stores, and 
 provisions. Other vessels, such as tramp steamers, are built to carry cargo 
 of great specific gravity. The former are so light that, in order to 
 sufficientl}' immerse them, considerable quantities of permanent ballast 
 are usually carried. In the latter, on the other hand, if the holds were 
 entirely filled with cargo of the nature referred to, the vessels would be 
 almost, if not entirely, immersed, and apart from considerations of strength 
 — assuming them to be able to float — the question of stability, with little 
 or no freeboard, would probably be a serious matter. 
 
 The purposes for which ships are built are innumerable, and to build all 
 vessels which are similar in size and proportion, of equal structural strength, 
 regardless of the work they have to do, would be absurd. What is more 
 reasonable and sensible is to ensure that each vessel is sufficiently strong 
 to satisfactorily perform her own work. To introduce as much structural 
 strength into a cross-channel passenger steamer which has to carry little 
 else than passengers, mails, and luggage, besides her own equipment, stores, 
 and bunkers, as into a vessel of equal dimensions which has to carry, say, 
 1000 or 1500 tons of cargo to any part of the world, would be foolish 
 in the extreme. Or even in the case of two purely cargo carrying vessels 
 of identical dimensions, if one is intended to be solely employed in carrying 
 a cargo of great density, such as coal, while the other is intended to carry 
 a cargo of much less density, such as wool, the one amounting to say 
 3000 tons, and the other to only about 2000 tons, obviously it would be 
 absurd to build each of these vessels to exactly the same scantlings, for 
 either the one would be excessively strong, or the other dangerously weak. 
 But, as stated, it is necessary that each vessel should be strong enough to 
 do her ow-n work. 
 
 The strength of ships, therefore, should be regulated by their proportions 
 and maximum displacements, provided that enough freeboard remains to 
 ensure sufficient stability and a condition of general seaworthiness. 
 
 It is clear that to study and tabulate the scantlings for every new 
 ship lies outside the sphere of the shipowner. That shipbuilders, by 
 their wide experience, Avould be vastly more capable of adequately dealing 
 with the matter, is true, but then, without a fixed standard of strength, 
 no two of them would either arrange their material alike, or arrive at 
 the same strength in the finished vessel, even for similar vessels under 
 similar conditions. Moreover, no such method of ship construction 
 would be either satisfactory or acceptable to underwriters. From both 
 owners' and underwriters' standpoint, standards of strength, both for 
 maximum displacements and minimum freeboard, are absolutely essential. 
 
28 STEEL SHIPS. 
 
 Or in other words, a guarantee is required that the vessel is strong enough 
 to carry with safety, and without injury to herself when experiencing the 
 various and probable demands which may be made upon her strength, a 
 certain load, which may not, under any conditions, be exceeded, and also that 
 her design is such, that, with a minimum reserve buoyancy when properly 
 loaded, she runs no risk of capsizing through deficient stability. With 
 this aim in view, there exist several societies, which, by scientific and 
 mathematical investigation coupled with long experience, have drawn up 
 rules and tables of scantlings suited to all types of cargo and passenger 
 vessels. The best known of these societies are Lloyd's Register, the 
 British Corporation, and the Bureau Veritas. These societies save both 
 shipowners and shipbuilders an immense amount of labour and trouble, at 
 the same time providing general uniformity in strength, and a satisfactory 
 guarantee of efficiency to the insurance societies for the vessels constructed 
 under their rules. To carry out this system, the committees of these 
 societies employ considerable numbers of surveyors, whose training and 
 experience have specially fitted them for the work. The quality of the 
 material used in the construction, the efficiency of the workmanship, the 
 carrying out of their societies' rules and requirements, and the periodical 
 survey for the renewal or alteration of the 'class' of the vessels placed in 
 their hands, is their sole responsibility. 
 
 The Load Line Act of 1890.— Before the Load Line Act came into force 
 in the year 1890, the overloading of ships, which was a source of danger 
 from both a structural and stability aspect, was attended with loss of life at 
 sea as well as loss of ships. The necessity for Government interference so 
 impressed itself upon this country, that the passing of the Load Line Act 
 was the final result, by which the Government became supremely responsible 
 for the seagoing condition of all British vessels and vessels leaving British 
 ports. 
 
 Societies empowered to assign Load Lines.— Hence, while the 
 British Government has sanctioned Lloyd's Register, the British Corpora- 
 tion, and Bureau Veritas Classification Societies, to assign load lines to 
 vessels classed by them, the Board of Trade still remains the supreme 
 authority for such assignment. It is certainly true that these societies can 
 build ships to whatever scantlings they please, but it is the work of the 
 Board of Trade to ensure that a maximum load line be fixed strictly in 
 accordance with the strength of such vessels with a reasonable percentage of 
 reserve buoyancy. 
 
 Standard of Strength upon which Load Lines are assigned. — While 
 considerable latitude is thus allowed in the disposal of material in the ship 
 structure, the necessity for a minimum standard of strength is obvious. 
 This is, moreover, essential in order that uniformity in the strength of the 
 different types of vessels built be secured, while it does not necessarily 
 follow that uniformity in the modes of construction will be a result. The 
 standard of strength laid down for the guidance of all classification societies 
 
CLASSIFICATION. 29 
 
 is that embodied by Lloyd's Rules for the year 1885. This gives the 
 minimum structural strength for a minimum freeboard for all classes of 
 vessels. Any society is at liberty to demand greater structural strength in 
 vessels classed by them than that of the Board of Trade standard, and, 
 naturally, they are individually responsible for efficient local strengthening. 
 
 Societies for the classification of vessels are thus at liberty to arrange 
 and formulate their own methods of construction, and determine the 
 scantlings of the material used in ships built to their particular classes. 
 With uniformity of strength, uniformity in assigning the freeboard of every 
 class of vessel is secured, since it mast be in accordance with the Freeboard 
 Act of 1890. 
 
 To the shipowner this subject of classification is of the greatest 
 importance. If he wishes to add a new vessel to his fleet, and he is 
 desirous of ensuring her being built to the highest class, he inserts a clause 
 in his specification, or in the contract between himself and the builders, 
 to the effect that she must be classed 100 Al at Lloyd's, or to the 
 corresponding highest class in any other society. 
 
 Load Line of Three Deck, Spar Deck, and Aioning Deck Vessels. — There 
 are four principal types of steel vessels considered in the Board of Trade 
 Freeboard Tables, viz.. Three Decked, Spar Decked, Awning Decked 
 Vessels, and Sailing Ships. A vessel may belong to the highest class in 
 any one of these types. The three decked type is the strongest vessel 
 built, and is thus allowed the minimum freeboard with consequently the 
 maximum immersion, displacement, and carrying power. Awning decked 
 vessels are the liglitest types of vessels built for over-sea voyages, and con- 
 sequently are required to have the greatest freeboard, with smaller inimersion,- 
 displacement, and carrying power. Supposing that into such an awning 
 decked vessel, more structural strength than is required by Lloyd's Rules 
 for 1885 is introduced, a comparison would be made between her increased 
 strength and that required for a spar decked vessel, and a proportionate 
 reduction made in the freeboard. Or, if a spar decked vessel were built in 
 excess of the standard strength, a comparison would be made between her 
 increased strength and thafc^of a three decked vessel, and a proportionate 
 reduction made in her freeboard also. But as the three decked vessel has 
 already a minimum reserve buoyancy, additions of strength beyond that 
 required by rule would obtain no concessions in the matter of diminislied 
 freeboard. 
 
 Grades of Class. — While it is customary for shipowners to liave new 
 vessels built to the highest class of their respective types, it does not follow 
 that such vessels will always maintain the highest class ; for example, at 
 intervals of four years, Lloyd's require that vessels classed with them should 
 be subject to special surveys. These special surveys are designated K'o. 1, 
 No. 2, No. 3, and No. 4 respectively, and as long as a vessel maintains her 
 structural strength, she maintains her class: but as soon as she begins to 
 deteriorate, and suffer reduction in structural strength, her class may be 
 
30 STEEL SHIPS. 
 
 reduced to 90a or 80a, which simply means that as she is now less able to 
 carry the same deadweight as originally, her freeboard is increased until her 
 carrying power is in accordance with her strength. 
 
 Maintenance of Class. — It is, however, possible, by carrying out certain 
 repairs, to restore the vessel to her original strength and class. " If a vessel 
 is at a port in the United Kingdom after the expiration of the prescribed 
 period for survey, and is not subjected to the special survey then due 
 before leaving the United Kingdom, the word 'Expired', is inserted 
 against her character in the Register book ; and in no case will a vessel 
 be allowed to retain her class, if she has not been' subjected to the whole 
 of the requirements of the requisite special survey within twelve months 
 from the date when the survey became due." (Lloyd's Rules.) 
 
 On the shipowner contracting with the shipbuilder for the building of 
 his vessel to class, say, 100 Al at Lloyd's, 3/3.1.1. Bureau Veritas, or B. S. 
 of the British Corporation, tbe first work of the shipbuilder is to have 
 drawn a set of plans upon which is shown the structural arrangement of the 
 vessel in accordance with the instructions contained in the Classification 
 Societies' Book of Rules, and the scantlings of each structural part, whether 
 of iron, steel, or wood, are figured upon these plans. They are next sent 
 to the classification society's registry through the local surveyors, where 
 the scantlings are checked, and the special structural requirements of this 
 particular vessel are considered. The plans are then returned to the ship- 
 builder with the necessary corrections or additions required clearly marked 
 upon them. This is the first introduction the ship has to the classification 
 society, but from that day to the time of her completion, she is the 
 object of their constant attention. Indeed, throughout her existence, so 
 long as she is 'classed' she is periodically subject to inspection by their 
 surveyors. The steel must also be of the quality required by the 
 particular society, and therefore such instructions accompany the specifica- 
 tions sent to the steel mills where the plates and bars are rolled, and the 
 forgings or castings made. Before these are sent back to the shipbuilder, 
 they must be tested by the classification society's testing surveyors, and 
 the stamp of the society's approval must be upon them. As the work of 
 building the ship proceeds, and the various parts are united, the ship 
 surveyors are ever watching that the steel or iron has the test stamp upon 
 it, that the plates and bars are to the dimensions indicated upon the 
 corrected original plans submitted by the builder, that the connections of 
 plates and angles, and the size and spacing of the rivets, are strictly in 
 accordance with the rules of their society, and that the general workmanship 
 is satisfactory. In this manner, the ship is daily under the eye of these 
 surveyors. 
 
 Simultaneously with the building of the ship, the engines and boilers 
 are constructed, and here again -the work is under the superintendence of 
 the society's engineer surveyors. Even the thicknesses of the plates and 
 the minimum diameters for steel masts, size of wire for the rigging, weight 
 
CLASSIFICATION. 3 1 
 
 of anchors and cables, are all determined in the Classification Societies' 
 Rules, and the last mentioned are subjected to their tests. Appliances 
 for life-saving, boats, belts, etc., are, however, all regulated by the Board 
 of Trade. 
 
 When at last the vessel is completed, and the watertightness of ballast 
 tanks, peak bulkheads, decks, etc., have been tested, and the whole work 
 carried out to the satisfaction of the surveyors, the minimum freeboard for 
 vessels of this class is permanently marked upon the ship's side. The 
 shipowner is then provided with a Certificate of Registry, and the vessel 
 is entered in the Register of the Society with which she is classed, and 
 denominated according to her class and type. Thus, by having his vessel 
 classed, not only has a shipowner the satisfaction of knowing that she is 
 well built, but as she is periodically surveyed, he is further assured as to 
 the maintenance of her strength, while in the event of his desiring to sell 
 her, her class is a guarantee to the purchaser as to her structural coiidition. 
 
 Moreover, as the vessel will probably be insured when she proceeds to 
 sea, her class is the only guarantee the underwriters have as to her strength 
 and the condition in which she has been maintained. 
 
 N'ote. — Although it is customary for a shipowner who intends having a new ship 
 classed, to have her built to the rules and under the supervision of the surveyors of the 
 .society with whom he purposes to have her classed, yet this is by no means an absolute 
 necessity. A vessel may be built independently of all society rules, and still obtain 
 a class. Before any society, however, would grant this, the vessel would require to 
 be thoroughly surveyed by their surveyors, and after an accurate estimate had been 
 made of her structural worth in comparison with their own particular standard, a class 
 corresponding to one of the grades of the classification society in question would be 
 assigned to her, with a maximum load line in accordance. 
 
 Undctssed Vessels. — That a ship should be classed with any society in order to 
 receive a load line in accordance with her strength is not essential. Thus, many 
 large shipowners, who employ thoroughly capable naval architects, design their own 
 vessels, and determine the scantlings themselves. When such a vessel is completed 
 and ready for sea and a load line is required, application is made to the Board of 
 Trade (or to one of the societies empowered to assigu load lines), who emjiloy a large 
 staff of ship, engineer, and nautical surveyors, by whom the vessel is thoroughly 
 surveyed, and who fully report upon the workmanship and scantlings ; and finally, 
 a freeboard is assigned in accordance with the vessel's strength judged by the standard 
 previously described. Such vessels are distinguished by the term 'unclassed,' 
 though it should be clearly understood that an unclassed vessel is by no means 
 necessarily inferior in any way to a classed one. 
 
 That Lloyd's present-day Rules may differ from their Rules of 1885, 
 which is the standard of strength laid down by the Board of Trade for 
 the guidance of all classification societies empowered to assign load lines, 
 is true, yet it should be remembered that this difference is due to extended 
 experience and fuller knowledge of the structural requirements of ships, 
 and may be taken as a more perfect form of the older rules. It is also 
 remarkable that this increased experience has rarely been followed by any 
 reduction in scantlings, but rather by their augmentation. The result 
 is, that Lloyd's own standard for classification is to-day higher than ever 
 it was. 
 
32 STEEL SHIPS. 
 
 JS'otwithstanding the various criticisms which from time to time have 
 been passed upon the rules for the construction of ships formulated by 
 Lloyd's Registry, the fact remains that by far the vast majority of British 
 vessels have been built to their requirements and classed with them, and, in 
 addition to these, large numbers of foreign vessels are included also. 
 
 This, in itself, is sufficient to conclusively show how great is the 
 confidence placed in Lloyd's by shipowners, and the general satisfaction 
 which their rules have given. 
 
 With this fact in view, and remembering that in these days most new 
 vessels are built to ' class,' it seems advisable to state, that in the treatment 
 of ship construction in the succeeding sections of this book, the methods 
 followed will be generally in accordance with the practice of Lloyd's 
 Registry, though where any particular purpose may be served, we shall not 
 hesitate to deviate from their rules, and describe other methods and 
 practice. 
 
CHAPTEE IV. 
 
 OUTLINE OF PRINCIPAL FEATURES AND ALTERNATIVE 
 MODES OF SHIP CONSTRUCTION. 
 
 Transverse and Longitudinal Framing — Form and Function of Parts — Butts in Trans- 
 verse Framing — Framing in Double Bottoms— Regulations for Increasing the 
 Number of Tiers of Beams — Compensation for Dispensing with Hold Beams — 
 Necessity of thorough combination of Transverse and Longitudinal Framing — 
 Structural Value of Shell Plating — Alternative Modes of Construction — Numerals 
 for Scantlings. 
 
 Possibly the knowledge of ship construction possessed by many readers 
 may be of a very limited character, and in this chapter it is simply pro- 
 posed to briefly enumerate the principal parts in the structural arrangement 
 of ships, so that the names of such parts may become familiar ; and to 
 give a general idea of their functions. This plan, it is believed, will 
 simplify the course pursued in this work, and render the chapter on 
 * Stress and Strength ' more intelligible, besides curtailing elaboration and 
 explanation in the longest division of the book dealing exclusively with 
 ' Construction.' 
 
 Framing. 
 
 Steel and iron ships are built on a combination of two systems of 
 framing, viz., longitudinal and transverse. 
 
 Longitudinal Framing includes all girder forms of material which run 
 in a fore and aft direction, whose function is to afford longitudinal strength. 
 
 Transverse Framing embraces all girder forms which cross the longi- 
 tudinal framework at right angles, affording transverse or athwartship 
 strength. 
 
 The strongest structure is obtained only when these two systems of 
 framing have been intelligently woven together, the strength of the one 
 co-operating with the strength of the other — that is, in relation to the work 
 which, conjointly, they have to do. When this is accomplished, the whole 
 is then covered by a skin, in the form of ' shell plating ' and decks, which 
 not only stiffens and strengthens the skeleton or framework, but adds 
 enormously to the total strength of the ship considered as a compound 
 girder. 
 
 3 
 
34 
 
 STEEL SHIPS. 
 
 Transverse Framing. — In order to preserve the transverse or athwartship 
 form of a ship under all the conditions of stress to which ships are subject 
 in carrying their various loads in smooth and wave water, a girder or frame 
 is placed at intervals of from 20 to 30 or more inches apart, all fore and aft. 
 
 Fig. 6 gives an example of the simplest form of transverse framing. 
 Here we have a half section of a comparatively small vessel showing such 
 a transverse frame with what are known as oudinary floors. It consists of 
 
 Fig. 6. — Midship Section showing Transverse Framing with Ordinary Floors. 
 
 a frame bar, a reverse frame bar, a floor plate, a beam, and a pillar or 
 stanchion. The efficiency of the complete frame depends upon the thorough- 
 ness of the combination of the various parts into one whole. 
 
 Frames. — The /rame bar, in this system of framing, is continuous from 
 the top of the keel to the gunwale, extending from the top of the keel to the 
 bilge along the lower edge of the floor j)late to which it is riveted. 
 
MODES OF SHIP CONSTRUCTION. 35 
 
 Reverse Frames. — The reverse bar also extends continuously from the 
 middle of the upper edge of the floor plate on the opposite side to the frame 
 (shown by dotted lines on the section) up the bilge, then on to the frame 
 bar to the gunwale except in the smallest vessels, though the alternate 
 reverse bars do not usually extend to this heiglit. 
 
 Floor Plates. — The floor plate extends from bilge to bilge, either in one 
 plate, or in two plates butted (joined end to end) alternately on either side 
 of the centre line. The frame and reverse bars are riveted to its lower 
 and upper edges respectively, converting the plate into a girder with top 
 and bottom flanges, which strengthen it to resist athwartship buckling. 
 (See fig. 6.) 
 
 The heiglit of the bilge ends of the floor plate above the base line is 
 usually abont twice the depth of the floor at the middle line of the ship ; 
 and at three-quarters the half-breadth out from the middle line it should be 
 at least half the depth of the floor plate at the middle line. 
 
 The advantage of carrying the floor plate round the turn of the bilge 
 must be obvious, strengthening what is liable to be a weak place in the 
 framing, especially in very square-bilged ships. At such a corner, ' working ' 
 is more liable when the vessel is subject to the severe stresses which are 
 experienced among waves, tending to produce alteration in the transverse 
 form. 
 
 Beams. — The beam is a steel or iron bar, uniting (in a single-decked 
 ship) the uppermost extremities of the frame, and preventing, by its own 
 tensile strength and rigidity, the tendency of the frame heads to open wider 
 apart from each other, or to approach each other when the vessel is subject 
 to stresses consequent upon loading, or from the pressure of the water upon 
 the immersed skin of the ship. Beams thus perform the function of both 
 struts and ties. 
 
 Here, again, much depends upon the efficiency of the means of con- 
 nection, and also in thoroughly supporting the angular connection of the 
 beam to the frame. Hence it is necessary to form a web of plating (beam- 
 knee) which should extend at least two and a half or three times the depth 
 of the beam down the frame to which it is securely riveted. We shall refer 
 more fully to this in Chapter VII. Owing to the depth of the vessel 
 illustrated in fig. 6, additional transverse support is required to the sides 
 between the weather deck and the floors. An additional tier of widely 
 spaced beams is thus introduced. 
 
 Pillars. — It is scarcely necessary to say that the shorter a bar is, the 
 more rigidity does it possess, and also, that the fullest efficiency of any 
 structure whose strength is made up of a number of parts, can only be 
 fully developed when such parts are combined in the most perfect manner 
 so as to cause the entire combination to act as one. Hence, in order 
 to develop the full efficiency of the transverse frame, its several parts 
 must be 'tied,' in order to prevent their acting independently of one another. 
 Pillars are therefore introduced, uniting the beam at the centre with the 
 
36 STEEL SHIPS. 
 
 floor at the centi-e. These cact both as struts and ties, preserving the 
 distance relationship of these opposite parts of the structure. Moreover, 
 without pillars, any crushing strain upon the sides of a ship might tend to 
 cause the beam to spring up at the centre (see fig. 22), its great length 
 reducing its rigidity and resistance to any such strain ; on the other hand, 
 they are needed to support the numerous and varying loads, — both 
 stationary, as Avinches, windlasses, etc., and temporary, as deck cargoes. 
 Where the beam is very great, additional pillars, termed quarter pillars, are 
 introduced between the centre ones and the sides of the ship, or else a 
 substitute of some kind is required. 
 
 Butts of Frame and Reverse Frame. — Though both the frame and the 
 reverse frame have a break in their lengths at the centre line above the 
 keel, in each case it is intended that the strength of these bars be continuous. 
 To unite these ends by such means as will ensure this is therefore 
 necessary. In the case of the frame, this is done by fitting a heel piece 
 (a piece of angle iron about 3 ft. long, of the same size as the frames), 
 on the opposite side of the floor and covering the butt, and in the case of 
 the reverse bar, by fitting a short covering piece on the opposite side of the 
 floor also. (See figs. 6 and 9.) These butt-covering bars also perform other 
 services in the ship structure, as will be shown later. 
 
 Transverse Framing in Double Bottoms. — The system of so construct- 
 ing the bottoms of vessels as to make them capable of carrying water for 
 trimming purposes or as ballast has become more and more universal during 
 recent years. 
 
 The earlier forms of these double bottoms were called M'Intyre tanks, 
 after the name of the inventor. At first these tanks were so built 
 that the transverse framing was maintained in the usual way, the inner 
 bottom being formed by laying the plating upon fore and aft girders stand- 
 ing on top of the ordinary floors. In some cases the inner bottom plating 
 extended horizontally out to the ship's side, where it was made watertight 
 by fitting angle collars round the frames. In the most common form of 
 M'Intyre tank, however, the tank side in each wing is formed by 
 turning the inner bottom plating down perpendicularly to the bilge. 
 Here, again, the frame bar was sometimes continuous through the tank 
 side, or, as it is more usually called, the tank margin plate. Bat 
 generally, the frame bar, especially in vessels now fitted with these tanks, 
 is cut at the tank margin plate, and a continuous angle bar fitted, 
 making the connection between the tank side and the outer shell plating 
 watertight. 
 
 The continuity of the transverse frame strength in such cases is main- 
 tained by connecting the frame legs to the tank side by large bracket plates. 
 Though M'Intyre tanks are less adopted than formerly, they are still 
 fitted in some cases. Such a tank is illustrated in a midship section, 
 fig. 71. 
 
 The double bottom water-ballast tank now most commonly adopted 
 
MODES OF SHIP CONSTRUCTION. 
 
 is that known as tlie ' cellular double bottom.' The transverse framing 
 of this system (and longitudinal framing dotted) is illustrated in fig. 7. 
 
 Examination of the diagram shows the arrangement of the transverse 
 framing to be somewhat interfered with, though, in effect, this modified 
 arrangement is identical in principle with that previously described for 
 vessels with ordinary floors. 
 
 The water ballast tank extends from margin plate to margin plate, and 
 
 €unm!eb3r tlPP^ <i^^^ '''"><'^'' P'^ 
 
 Jdnk wargn plate 
 
 Keel plate 
 
 Fig. 7. — Midship Section showing Transverse Framing in a Vessel with a 
 Celhilar Double Bottom, also the Longitudinal Framing (dotted). 
 
 from the keel to the top of the floors. Here we usually find that a 
 continuous deep plate runs fore and aft through the centre of the tank 
 (centre keelson or centre through-plate). 
 
 The continuity of the floor plate is necessarily interrupted, but the 
 continuity of the transverse strength is maintained by connecting the 
 floors to the centre through-plate by means of angles. The floors at 
 
38 STEEL SHIPS. 
 
 the middle line are much deeper than required for ordinary floors, thus 
 permitting a good connection, and somewhat compensating for the break 
 in the continuity of the frames and reverses. The outer ends or 
 extremities of the floors, called bracket plates, extend to exactly the 
 same height as in similar ships with ordinary floors. But again, at the 
 bilge, we observe that in order to form the outer boundary of the cellular 
 bottom tank, the floors have to suffer interruption in continuity in order 
 to allow the margin plate to pass continuously fore and aft, and fit hard 
 on to the shell plating. However, by means of angle connections, and a 
 considerable depth of floor at this place, a minimum for which is definitely 
 stated in Lloyd's Rules (margin plate width), a good connection is made 
 to the bracket plate, or tank knee as it is sometimes called, and the shape 
 of the floor is preserved to the height prescribed. 
 
 In order to ensure the watertightness of the tank at this place, a 
 continuous bar is fitted to the lower extremity of the margin plate, to 
 connect it to the outer bottom. This necessitates another break in 
 the continuity of the frame bar. The depth of the floor plate, however, 
 enables a satisfactory connection to be made between the transverse framing 
 inside and outside of the tank, by means of the bracket plate already men- 
 tioned. In exceptional cases the frames have been preserved continuously 
 from keel to gunwale, and the watertightness at the margin plate effected 
 by fitting angle collars round the frames for the whole length of the 
 tank. The reverse bar is also intercostal between the centre keelson and 
 the margin plate, and then again commencing outside the margin plate, is 
 continued along the upper edge of the bracket plate, and up the frame bar 
 to its prescribed height. 
 
 In all other respects, the upper framing, beams, pillars, and their 
 connections, are identical with those in the ordinary system previously 
 referred to in fig. 6. 
 
 Eegidations for increasing the number of tiers of Beams. — So 
 far, we have only dealt with a small type of vessel requiring but two 
 tiers of beams. As vessels increase in size, however, not only for con- 
 venience and adaptation for- carrying cargo, but for reasons of structural 
 strength, more tiers of beams, which may or may not be sheathed with 
 a wood or steel deck, are introduced. If the vessel is classed, these are 
 regulated by the rules of the classification society. Thus, we find that 
 vessels classed at Lloyd's, when over 15 ft. 6 in. from top of keel to top of 
 upper deck beam at centre, require in addition to the weather deck, an 
 extra tier of beams in the hold, which may be widely spaced — on every 
 tenth frame — (see fig. 6) ; and Avhen 24 ft. is exceeded, still another tier 
 of beams is required. 
 
 Compensation for dispensing with Hold Beams. — The introduction 
 of an extra tier of beams in a ship of, say, 16 ft. in depth, may, 
 however, prove a source of inconvenience to a shipowner in the stowage 
 of certain kinds of cargo, and a similar inconvenience may also arise 
 
MODES OF SHIP CONSTRUCTION. 39 
 
 ill the introduction of the third tier of beams in a vessel of, say, 
 25 ft. depth. 
 
 In such cases, by modifying the transverse framing, these lowermost 
 tiers of beams could be dispensed with, in the manner illustrated later in 
 this chapter. First of all, we observe that these additional tiers of beams 
 are required for purposes of strength. With ships of greater depth, and 
 naturally we infer, of greater breadth, the increased immersion produces 
 vastly increased stress from the external pressure of the water, tending 
 to crush in the sides of the vessel. 
 
 Moreover, the loading of heavy cargo tends to distort the transverse 
 form. Thus additional tiers of beams, supported from keel to uppermost 
 deck by pillars, become necessary, unless a satisfactory method of com- 
 pensation can be provided. This can be done by increasing the strength 
 and rigidity of the transverse framing, the details of which method are 
 full)"" described a little later. 
 
 Longitudinal Framing. — We have already observed that the special 
 function of transverse framing is to preserve the transverse form, when 
 experiencing the numerous and ever- varying stresses to which ships are 
 subject. In a like manner, and to a much greater extent, especially in 
 vessels of great length, there is the tendency to alter in longitudinal 
 form ; as in their pitching movements they are subject to immense 
 differences and sudden changes in the buoyant support afforded by 
 the water, and to sudden twisting stresses when crossing skew seas. 
 Moreover, such stresses may be greatly increased, especially under 
 certain conditions, by the filling of large peak or other ballast tanks 
 at the ends of a vessel, or in stowing very heavy cargo towards the 
 extremities. 
 
 Of what paramount importance does an efficient arrangement of 
 longitudinal framing become, will therefore be obvious. Thus, we find 
 that along the bottom, on the bilge, and up the sides of a vessel, are 
 a number of girders of various forms extending all fore and aft, the 
 continuity of whose strength is most rigidly maintained. 
 
 Keelsons and Stringers. — Fig. 8 shows the same section as fig. 6 with 
 the longitudinal framing introduced. We may notice that all longitudinal 
 girders of whatever form, on the bottom of the vessel between bilge and 
 bilge, are termed keelsons, and those on the sides above the bilge, stringers. 
 The name given depends, not so much upon the form of the girder section, 
 as upon the locality in which it is placed. The most important keelson girder 
 is that standing upon the top of the floors at the middle line. It is 
 composed of a vertical plate with two large angles on the top, and two on 
 the bottom, the whole being mounted by a thick plate called a rider plate. 
 This girder is termed the centre keelson, and really forms the backbone 
 of the ship. 
 
 In order to prevent the floor plates tripping, that is, inclining either 
 forward or aft, and to afford strength and stiffness to the shell, it is usual 
 
40 
 
 STEEL SHIPS. 
 
 to introduce what is termed an intercostal girder or side keelson, so named 
 because it is composed of plates fitted intercostally between the floors in a 
 continuous line all fore and aft, and connected to them by angles. In 
 very small vessels, somewhat similar plates are fitted, whose height termi- 
 nates at the upper edge of the floors. These are named ivasli plates, their 
 function being to prevent any water which may have drained into the 
 bilges, dashing from side to side when the vessel rolls. This is also one of 
 the functions of side intercostal keelsons. But in larger vessels, the plates 
 
 Sunwdk bar 
 
 Upper deck stringer plite 
 
 Fig. 8.— Midship Section showing the Longitudinal and Transverse Framing 
 combined, in a Vessel with ordinary Floors. 
 
 are carried above the floors between two angles as shown in fig. 8, and 
 thereby converted into a continuous keelson. Keelsons may be built up in 
 a variety of ways according to the dimensions and proportions of the vessel. 
 These are more fully dealt with in Chapter VII. 
 
 The next girder is on the bilge, and is called a bilge keelson. 
 
MODES OF SHIP CONSTRUCTION. 4 1 
 
 Furtlier up the vessel's side is a small girder, known as a side stringer 
 (or it may be an upper bilge stringer). 
 
 Deck Stringers. — At the gunwale, and on the lower tier of beams, are 
 broad continuous plates fitted on the ends of the beams. These are deck 
 stringer plates, and form most valuable girders in conjunction with the 
 beams. Wherever a tier of beams is fitted, whether a wood or steel deck 
 is, or is not laid, this thick stringer plate is always to be found. 
 
 Centre Through- Plate. — In fig. 7 the longitudinal girders are shown 
 (dotted lines) in conjunction with the transverse framing of the double 
 bottom. Extending continuously down the middle line, and standing on 
 the plate keel, is a deep plate, of the depth of the floors and tank. Indeed, 
 the depth of the floor is regulated by the depth of this plate, the minimum 
 width of which is fixed by the classification society. Two large angles 
 on the top and bottom, mounted by a thick plate forming part of the 
 tank top, make up the combination forming the centre keelson. It is also 
 known as the centre through-plate, or centre girder. 
 
 Intercostal Keelsons. — The next girder is usually intercostal, the floors 
 being continuous, as a general rule, in merchant vessels. This is known as a 
 side or intercostal keelson. Its continuity is maintained by connection to 
 the floors by means of angles. 
 
 Margin Plate. — At the bilge is the tank side, or margin plate, 
 extending continuously fore and aft. This naturally forms an efficient 
 bilge keelson, though not known by that name. Above the bilge, the 
 arrangement of stringers is similar to that for the ship with ordinary floors. 
 
 Necessity of thorough conibination of Transverse and Longitudinal 
 Framing. — Even viewed separately, although both the transverse and longi- 
 tudinal framing are absolutely essential for the work to be done, yet neither 
 could possibly perform its own duty without the aid of the other. 
 "Whenever a longitudinal girder crosses a transverse frame, it ought to be 
 carefully and thoroughly connected with it, if necessary, by the aid of small 
 pieces of angle iron, or lugs, as they are often termed. For instance, in 
 fig. 8 we find that the centre keelson is riveted to every reverse frame 
 that it crosses on the top of the floors by means of its own bottom bars. 
 But in order to get a doubly secure connection, a short piece of reverse 
 bar (lug piece) is riveted to the other side of the floor, through the 
 horizontal flanges of which the bottom bars of the keelson are riveted 
 also. This same lug forms the covering piece for the reverse frame butt. 
 The same principle of connection is applied to all the other keelsons and 
 stringers wherever practicable. 
 
 Structural Value of Shell Plating. — When the skin, or shell plating, is 
 worked over this framework, the effect is an enormous contribution to both 
 transverse and longitudinal strength by more effectually binding together 
 the whole structure into a single compound girder. Such a mode of con- 
 struction, where one part is interdependent upon another for the utmost 
 development of its own strength and rigidity, must commend itself. 
 
42 STEEL SHIPS. 
 
 Fig. 10 shows the vessel whose framing has been illustrated in figs. 6 
 and 8 with the shell plating in addition, and similarly fig. 11 shows the 
 vessel whose framing has been illustrated in fig. 7, with the modification in 
 the framing caused by the hold beams being dispensed with and web frames 
 and stringers introduced as compensation. 
 
 The transverse framing stiffens and assists the longitudinal, and the 
 longitudinal stiffens and assists the transverse, in doing each its own 
 work. The whole system of the structural work of a ship is based upon the 
 principle that unity is strength. Each of the innumerable parts performs its 
 work in conjunction with the adjacent parts to which it is closely related, 
 thereby contributing to the perfection of the efficiency of the ship girder. 
 Thus many severe stresses, which would at first appear to be of a local 
 character, become general, simply because the particular part upon which, 
 perhaps, they are first experienced, communicates a share of the stresses 
 through the other girders which cross and are attached to it, to the area 
 around. In this manner the stress is distributed over the structure, its 
 local severity reduced, and the possibility of evil results minimized. 
 
 Fig. 9 is a midship section of a large vessel, showing the longitudinal 
 and transverse framing and the shell and deck plating. It also furnishes the 
 name of each structural part, which no doubt may be of assistance to many 
 readers in traversing the following pages. 
 
 Alternative Modes of Constkuction. 
 
 Assuming, in the case of a vessel to be classed 100 Al at Lloyd's, that all 
 the preliminary work of fixing the dimensions, and getting out and approv- 
 ing of the design and general arrangement has been satisfactorily concluded, 
 before any steps can be taken in the work of construction, structural 
 plans must be prepared by the shipbuilder, and submitted for the 
 approval of Lloyd's Committee. These plans usually consist of a midship 
 section, which gives a transversei view of the structure of the vessel amid- 
 ships ; a profile, showing a longitudinal sectional elevation (see figs. 48 and 
 49) ; and sometimes a deck plan ; and upon these plans, the sizes or scant- 
 lings of all iron, steel, or wood, forming structuial parts of the vessel, are 
 distinctly marked according to the requirements of Lloyd's Rules. 
 
 As we have previously observed, in the case of a classed vessel the 
 responsibility of structural strength, in relation to the weight to be carried 
 (or the freeboard), is borne by the classification society ; hence both builders 
 and owners are relieved from this, so long as the workmanship is of a 
 thoroughly satisfactory nature. However, as the classification societies 
 provide alternative modes of arriving at the required structural strength, the 
 owner has the option, under the guidance and advice of his naval architect, 
 of making the choice of the system most suited to his requirements. 
 
 A ship is a girder composed of a host of smaller girders which comprise 
 the framework. In the nature or the design of the framing the shipowner 
 
MODES OF SHIP CONSTRUCTION. 
 
 43 
 
 ONTINUOUST I ROM 
 RUNNER 
 
 TYIACKS PATENT 
 •^ RAIL SECTION 
 
 MOULOINC 
 
 GUNWALE 
 
 MBLEHOMC 
 
 fWmOATION PLATE 
 
 KEEL--^ ' CARBOARO 
 h STRAKE 
 
 Fig. 9. — Midship Section of Merchant Steamer, 
 
44 
 
 STEEL SHIPS. 
 
 is free, to some extent, to exercise his choice. This can probably be best 
 illustrated by the aid of the diagrams figs. 10, 11, and 12. 
 
 With Hold Beams. — Fig. 10 is a midship section of the vessel whose 
 transverse framing we have already examined in fig. 6. She has a bar keel, 
 and is built on the ordinary floor system. Being over 15 ft. 16 in. in depth 
 
 Fig. 10. — Midship Section of Vessel requiring two Tiers of Beams 
 with ordinary Frames and Floors. 
 
 from the top of the keel to the top of the upper deck beams at centre with 
 the normal amount of camber, this vessel requires two tiers of beams. 
 But while the beams in the upper tier are placed on every frame (2 ft. 
 apart), the beams in the lower tier, which are exceptionally strong, are only 
 
MODES OF SHIP CONSTRUCTION. 
 
 45 
 
 required on every tenth frame (20 ft. apart). The transverse framing, 
 supporting the side shell plating, is made up of two angle bars (the frame 
 and reverse frame) fitted back to back, and the longitudinal frame girders 
 are as shown in the section. 
 
 Web Frames in Keu of Hold Beams. — Fig. 1 1 is a midship section of 
 
 Fig. 11. — Midship Section showing Hold Beams dispensed with, and Compensation 
 in the form of Web Frames and Stringers introduced. 
 
 the same vessel as shown in fig. 7, and is of exactly the^same dimensions 
 and external form as fig. 10. In this case, the vessel j has a side bar 
 keel, and is fitted with a double bottom. The hold]](beams are dispensed 
 with, and in lieu of them, transverse web frames are spaced on every sixth 
 
46 
 
 STEEL SHIPS. 
 
 frame throughout the vessel's length. A section at a web frame is 
 shown. It consists of a web of plating 15 in. wide, extending from the 
 tank side to the upper deck beams. It is attached to the shell plating by 
 
 Fig. 12. — Midship Section showing Hold Beams dispensed with, and Compensation 
 in the form of Deep Framing introduced. 
 
 an angle of ordinary frame size, and has two smaller angles fitted on its 
 inner edge, or a single bar of larger size (equivalent sectional area) may be 
 
MODES OF SHIP CONSTRUCTION. 47 
 
 substituted. (When there is no double bottom, the web frame blends into 
 the ordinary floor, and forms a complete continuous girder from gunwale to 
 gunwale, and the transverse frames between the web frames are similar to 
 those shown in fig. 10.) The longitudinal framing consists of two web 
 stringers fitted intercostally between the Aveb frames. When the beams at 
 the head of the web frames are continuous from side to side, they are fitted 
 of extra strength, but elsewhere they are similar to the upper deck beams 
 shown in fig. 10. 
 
 Deep Framing in lieu of Hold Beams. — In fig. 12 the hold beams 
 are again dispensed with, and compensated for by adopting what is known 
 as deep framing. These deep frames, unlike the web frames, are upon 
 every frame, and are made up of two large angles fitted together as 
 shown on the midship section, fig. 12, which, in combination with the side 
 stringers, possess strength and rigidity fully compensating for the omission 
 of the hold beams. 
 
 Several alternative forms of side stringers may be adopted, one of which 
 is shown in the diagram fig. 12. It may also be pointed out that when 
 the deep frame system is adopted, the tank bracket knees should be of 
 extra depth so as to provide ample support to the bilge, and the beam 
 knees to the lower tier of beams should be three times the depth of the 
 beam. 
 
 The adoption of any one of the foregoing systems of framing would in 
 no way interfere with a vessel's obtaining the highest class with any 
 classification society. On coming to vessels of larger dimensions than the 
 one with which we have been dealing, it is natural to expect that the depth 
 and thickness of the transverse framing should increase, and that a greater 
 number of side stringers should be required. But as soon as the depth 
 exceeds 24 ft., we have seen that an extra tier of beams is required, 
 and the vessel, assuming that full scantlings have been adopted, becomes 
 of three deck type. 
 
 However, that the vessel will have three tiers of beams does not neces- 
 sarily follow ; for, just as in a two deck vessel the lower tier of beams 
 may be dispensed with, in like manner, the lowest tier in a three deck vessel 
 may be dispensed with by fitting in lieu, either web frames up to the 
 middle deck, or deep framing. 
 
 Fig. 48 shows the transverse framing of a three deck vessel of 34 ft. 
 depth from keel to upper deck (not shelter deck) beam at centre (having 
 the normal camber), with deep framing in place of the omitted hold 
 beams. 
 
 Nmnerals for Scantlings. — In order to ascertain the scantlings of the 
 material shown on the plans submitted to Lloyd's Committee, their rules 
 give the following instructions : — 
 
 Assuming that the vessel is to be of full strength with, therefore, maxi- 
 mum rule scantlings, — add together (measurements being taken in feet) the 
 girth of the half midship frame section of the vessel, measured from the 
 
48 STEEL SHIPS. 
 
 centre line at the top of the keel to the upper deck stringer plate ; half the 
 moulded breadth; and the depth (Lloyd's). The sum of these numbers 
 gives what is known as the 1st numeral. By multiplying the 1st numenil 
 by Lloyd's length, the 2nd mwieral is obtained. By means of the 1st 
 numeral, the sizes and thicknesses of all frames, reverse frames, depth and 
 thickness of ordinary floors, thickness of bulkheads, and diameter of pillars 
 are found from the tables in Lloyd's Book of Rules. 
 
 The greatest breadth of the vessel at each deck regulates the size of the 
 beams for each respective deck. 
 
 The depth of the vessel regulates the number of tiers of beams, and 
 the number of stringers. The 2nd numeral, both alone and in conjunction 
 with the proportion of length to depth, regulates all the remaining 
 scantlings, and the number of complete or partial steel or iron decks. 
 It should be noted, however, that for vessels of over 24 ft. in depth 
 (Lloyd's) — 'three deck' — notwithstanding the fact that the lowest tier 
 
 Fig. 13. — Length between Perpendiculars. 
 
 of beams may be dispensed with by making compensation in some form 
 or other, the 1st numeral is obtained hy deducting 7 from the sum of the 
 half girth, half moulded breadth and depth ; and the 2nd numeral by 
 multiplying the 1st numeral thus obtained by the length in the usual way. 
 For the lighter types of vessels, known as ' spar ' and * awning ' deck, 
 the 1st numeral is obtained by adding together the half girth and Lloyd's 
 depth taken to the main deck (see fig. 14) together with the half 
 moulded breadth, and the 2nd numeral is the product of the 1st numeral 
 hy the length of the vessel. (See fig. 9 for Lloyd's depth.) 
 
 In the book of Lloyd's Rules, these numerals are all arranged in a 
 graduated tabular form, and adjoining them is given the scantling of the 
 material, or structural item or items which they govern. By referring to 
 these rules, and the notes accompanying them, the particulars of scantlings 
 given along with the midship section (fig. 10) are found, as are also the 
 tables of scantlings on pages 111 and 112. 
 
MODES OF SHIP CONSTRUCTION. 
 
 49 
 
 After considering the special features of the vessel in question, and 
 
 adjusting the scantlings Avhere 
 necessary, the plans are returned 
 to the shipbuilder, who sees to 
 the ordering of the steel and 
 iron, and whose draughtsmen 
 proceed with the work of draw- 
 ing the detail structural plans 
 for the workmen in the ship- 
 yard. 
 
 Definitions of Important 
 Terms. 
 
 1. Length between Perpen- 
 diculars. — This is the length 
 usually agreed upon by ship- 
 owner and shipbuilder when 
 contracting for a new vessel, 
 and is that generally understood 
 when speaking of the length 
 of a vessel. For vessels with 
 straight vertical stems, it is 
 taken from the foreside of the 
 stem bar to the aft side of the 
 stern post. When the stem is 
 raked, that is, inclined forward 
 from the foot, the length is 
 measured from the fore side of 
 the stem bar at the upper deck. 
 Should the vessel have a clipper 
 or curved stem, the length is 
 measured from the place where 
 the line of the upper deck 
 beams would intersect the fore 
 edge of the stem if it were 
 produced in the same direction 
 as the part below the cutwater. 
 (See fig. 13.) 
 
 2. Length over all is the 
 length measured from the fore- 
 most tip of the stem bar to the 
 aftermost tip of the overhang 
 of the stern. For vessels with 
 straight stems it is practically 
 the extreme length. 
 
 03 
 
 ti 
 
 Cl 
 
 Qi 
 
 CM 
 
 •CI 
 
 c: 
 
 c 
 
 
 Ct3 
 
 <c 
 
 
 Moulded Depr/i 
 
 Jilii' 
 
 o 
 
 I 
 
 -4 
 
 Li 
 
 o 
 <o 
 Q 
 
 t~ 
 
 CD 
 CI. 
 00 
 
 o 
 Q 
 
 CO 
 
 Moulded Depth 
 
 o 
 
 ■>4 
 
 ^-iV 
 
 Qi 
 
 Moulded Depfh- 
 
 o 
 
 
 O 
 
 O 
 -~4 
 
 O 
 
 OJ 
 
 CO 
 
 cr» CO 
 
 o 
 
 0> 
 
 en 
 
 t. 
 
 CO 
 
 Q-. 
 CO 
 
 CO 
 
 ^ 6 
 
 03 
 
 03 
 03 
 
50 STEEL SHIPS. 
 
 3. Lloyd's length is the same as the length between perpendiculars, 
 except that the length is taken from the after side of the stem to the fore 
 side of the stern post. 
 
 4. Registered length is measured from the fore side of the tip of the 
 stem bar to the after side of the stern post. 
 
 5. Extreme breadth is measured over the outside plating at the greatest 
 breadth of the vessel. This is also the Registered breadth. 
 
 6. Breadth moulded is taken over the frame at the greatest breadth 
 of the vessel. 
 
 7. Depth moulded is measured in one, two, and three deck vessels at 
 the middle of the length, from the top of the keel to the top of the 
 upper deck beams at the side of the vessel. In spar and awning decked 
 vessels, the depth moulded is measured from the top of the keel to the top 
 of the main deck beams at the side of the vessel (see iig. 14). 
 
 8. Lloyd's Depth. — This is somewhat diflFerent. It is required that 
 all upper decks, and all main decks in spar and awning decked vessels 
 (see fig. 14) have a round upon them, the height of which at amidships 
 amounts to a quarter of an inch for every foot of the greatest breadth of 
 the deck. This is termed 'camber,' 'round up,' or 'round of beam.' 
 This is added to the moulded depth, and gives Lloyd's depth. "With this 
 modification, it is otherwise the same as No. 7. 
 
 9. Depth of Hold. — This is measured from the top of the ceiling in 
 vessels with ordinary floors, or from top of ceiling on double bottoms, if 
 ceiling is laid ; or from the tank top plating when no ceiling is laid, to the 
 top of the beams of upper, spar, or awning decks. This is also the 
 Registered depth (about 2|- in. is the usual allowance for ceiling). 
 
 10. Extreme Proportions. — A vessel is said to be of extreme propor- 
 tions when her length exceeds eleven times her depth (moulded depth plus 
 camber). Under such circumstances, additional longitudinal strength is 
 required over and above that necessary when of ordinary proportions. 
 
CHAPTEE V. 
 
 STRESS AND STRENGTH. 
 
 Preliminary. — Forces exerted upon Ships — Water Pressures — Pressure per Frame Space — 
 Estimated Pressure upon a Bulkhead and Centre of Pressure — Pressure upon a Tank 
 Top— Tendency to Transverse Deformation — Longitudinal Stresses in still water and 
 Tendency to Longitudinal Deformation — Bending Moment on a Loaded Bar — 
 Possible effect of a bad disposition of Weights in a Ship — Longitudinal Stresses in 
 Wave Water and Tendency to Longitudinal Deformation — Local Stresses, Pant- 
 ing Stress, Stress due to Propulsion by Steam, Stress due to Propulsion by 
 Sail, Stress caused by loading Heavy Cargoes on deck and the shipping of 
 Heavy Seas, Rudder Stresses, Strains from loading aground. Stresses upon 
 Vessels which are temporarily only partially waterborne, Launching Stresses — 
 Stress and Strain — Ductility and the Elastic Limit — Moment of Inertia — Curve of 
 Loads — Curve of Shearing Stresses — Curve of Bending Moments — Stress per Square 
 Inch — The Equivalent Girder — Computation of Moment of Inertia of Compound 
 Girder — Stress per Square Inch upon Upper and Lower Surfaces of foregoing Girder 
 — Bending Moment of an actual Ship ; Curve of Weights, Curve of Buoyancy, Curve 
 of Loads, Curve of Shearing Forces, Curve of Bending Moments — Moment of Inertia 
 of Actual Ships — Comparison of Vessels — Disposition of Material — Effect of Modifi- 
 cation in Depth of Transverse Frames — Value of Stress Calculation — Calculation 
 for Position of Neutral Axis and Moment of Inertia of an Actual Ship — Estimate of 
 Bending Moment for an actual Ship, and Stress per square inch — Value of Regis- 
 tration Societies to Shipowner — Comparison of Stresses on Vessels increasing in 
 size — Further Remarks upon the Value of Stress Calculations — Working Stress — 
 Erections on Deck — Board of Trade Instructions for Comparing the Strength of 
 Vessels for Freeboard Purposes — Deductions — Calculation to find Neutral Axis — 
 Calculation of Moment of Inertia of a Vessel when supported upon AVave Crest at 
 Middle of Length, and thus subject to Hogging Stresses. 
 
 Preliminary. — Having become acquainted in the previous chapter -with. 
 the general structural features of ships and the usual disposition of both 
 longitudinal and transverse framing, it follows in the natural order of 
 things in traversing the wide area of the field of ship construction, that at 
 the outset it is of the highest importance that a comprehensive knowledge 
 of the numerous and severe stresses which ships have to endure under 
 varying circumstances in following their various trades be possessed. And 
 not only should the strength of individual pieces of steel or iron when sub- 
 ject to tensile, compressive, torsional or shearing stresses be fully under- 
 stood, but the strength value of girders and combinations of girders variously 
 
52 STEEL SHIPS. 
 
 disposed, such as are found iu the structure of ships, should be comprehended. 
 Indeed, without such knowledge as this, it is absolutely impossible to in- 
 telligently determine the structural arrangement of ships, so as to obtain 
 anything like the highest degree of efficiency. 
 
 The stresses to which ships are subject are usually classed under three 
 principal heads, and arranged in order of their importance as follows : — 
 
 I. Longitudinal Stresses. 
 II. Transverse Stresses. 
 III. Local Stresses. 
 
 However, although it is intended to observe and enumerate the chief 
 stresses which come under each of these divisions, to adhere rigidly to 
 this order in dealing with them is neither necessary nor advisable. 
 
 It was pointed out in an earlier chapter, that when a vessel is classed 
 at Lloyd's, or with any other classification society, the responsibility of 
 structural strength and efficiency of the workmanship is borne by the 
 society in question. Their special work is to consider the numerous 
 and varied stresses to which ships are liable iu smooth and wave water, 
 and to specify in their rules the nature and disposition of material to 
 ensure against structural damage arising as the result of the continuous 
 or occasional presence of these stresses, when vessels are properly, or 
 rather, intelligently loaded and ballasted. 
 
 These societies, however, whose business is conducted and superintended 
 by men of great scientific and practical experience, are always willing to 
 accept, and substitute, any proposed mode of construction, providing that 
 its efficiency is equal, or superior, to that prescribed in their rules. 
 
 A huge accumulation of material does not ensure a strong vessel, and 
 therefore the most heavily constructed ship is not necessarily the best. 
 Ships are not built to be marvels of ponderous massiveness, strength, or 
 weight. Far from it. Cargo or merchant ships are built to earn 
 money, and in these days of keen competition, economy in design, in 
 working, and in maintenance, are primary considerations to every 
 shipowner. The lightest ship is therefore the best ship, that is, if 
 she possesses sufficient strength for the demands which will be made 
 upon her in carrying her maximum load in all the probable conditions 
 of weather she will experience. 
 
 Every ton of unnecessary iron, or steel, or wood, or other constructive 
 element in the ship, means a ton less freight, and reduced earning power. 
 Too much attention, therefore, cannot be given to the careful study of every 
 detail in the design, and the alternative modes of construction. 
 
 Forces exerted upon Ships. — It is thus proposed at this stage to con- 
 sider, as briefly as is consistent with clearness, the varying forces which 
 exert themselves upon ships, and to observe the nature of the resistance 
 which the structure, as a whole, offers to deformation and fracture. 
 
 Water Pressures. — To attempt to dispose of the structural material in a 
 
STRESS AND STRENGTH. 53 
 
 vessel, before fully comprehending the nature and magnitude of the various 
 stresses which are experienced, could only result in most imperfect design 
 and misplacement of strength. 
 
 To those who have given little consideration to this aspect of the 
 subject, many peculiar and sometimes amusing ideas present themselves 
 as reasons why ships float. But, on careful investigation, there is nothing 
 mysterious after all. All ships, no matter what their weight may be, are 
 subject to the same gravitation influence as all other structures, and, like 
 them, have to be supported if they are to maintain a condition of 
 equilibrium. Thus, if it were possible to suspend a ship weighing, say, 
 2000 tons, on a spring balance, in mid air, and slowly lower her into 
 the water, we should find that on first touching the surface she would 
 still weigh 2000 tons. But immediately she began to be partly immersed, 
 the balance would register an alteration in the weight, decreasing gradually 
 in a proportion strictly in accordance with the amount or volume of 
 immersion from 2000 tons down eventually to nothing, when the ship 
 would remain stationary, with not the least tendency to become either 
 more or less immersed. Indeed, she now offers determined resistance 
 to any alteration in draught so long as neither additions nor deductions 
 are made to her original weight. Though the balance registers tons, 
 reason forbids our assuming that the ship has decreased in, or lost, 
 weight. What it does prove is, that the balance has gradually been 
 relieved of, or freed from, the effort of supporting the original 2000 
 tons, the support having been gradually obtained from an evidently 
 upward pressure from the water, which increased as the ship continued 
 to be further immersed, until it was able to support the whole of the 
 2000 tons, at which moment the balance showed tons. 
 
 So that when we talk about a ship floating, we simply mean that 
 she is being supported by the water. To balance this ship in mid air 
 by means of supports, say, pillars of iron, would be no easy matter. 
 A ship is a huge, bulky object ; her weight is enormous, and a very 
 careful estimation of the strengths of the supports, and their number 
 and positions, would have to be made in order to safely effect this. 
 But in water, the support and balance are perfect. There are exactly 
 2000 tons of upward pressure from the water to exactly support the 
 2000 tons downward pressure of the weight of the ship, the balance 
 being perfectly effected by the centre of the support (or the resultant of 
 all the supports) termed the Centre of Buoyancy, and the centre of the 
 weight, termed the Centre of Gravity, being exactly in the same vertical 
 line, which condition is absolutely essential in order to preserve an exact 
 balance, or, as we say, equilibrium. The net result is, that we have 
 two equal and opposite forces exactly neutralizing each other, and 
 producing no result whatever. 
 
 A ship, or any object capable of floating, on being launched, or by any 
 means placed in water, immediately, by these natural laws, places herself in 
 
54 STEEL SHIPS. 
 
 a position such as we have desciibed ; indeed, she must do so before a 
 condition of rest can be maintained. 
 
 We have seen that any object floating in water is supported by the 
 pressure or thrust, derived from the water itself. Upon every part of the 
 immersed surface of a floating object, the pressure is exerted in lines of 
 action perpendicular to the surface. This is indicated in diagram 15 
 by the lines of water pressure, in the various immersed, or . partially 
 immersed objects. 
 
 As the direction in which water pressures exert themselves is of very 
 great importance in determining the structural strength of immersed, or 
 partially immersed objects, a little experiment, confirming the preceding 
 statement that they act in lines perpendicular to the surface immersed, will 
 be of interest. 
 
 Suppose a vessel of the form shown in diagram 16, perforated all over 
 its surface with small holes, to be filled with water. It would be found 
 that the water would squirt out of the holes in directions perpendicular to 
 the surface of the vessel at each part, thus indicating that the pressures 
 on the internal surface of the vessel are perpendicular to the surface 
 acted upon. See fig. 16, a, page 56. 
 
 Or again, suppose the same vessel to be suddenly immersed in water. 
 In this case, water would squirt into the vessel in directions perpendicular 
 to the surface, thus again proving that the external pressures are exerted 
 in lines perpendicular to the immersed surface. See fig. 16, b. 
 
 rigs, a, b, c, d, e, /' and g on the next page will enable us to make a 
 simple analysis of these forces, and in some degree estimate their amount. 
 
 First, then, we will take fig. a. This is a rectangular box vessel, 
 whose bottom surface has an area of exactly 1 sq. ft., the height of the 
 box being, say, 2 ft. 6 in. Let the water in which this vessel has to be 
 immersed be fresh water. The box is so weighted as to produce a total 
 weight of 62| lbs. On placing the box in the water, we should find that 
 it would sink until its draught, or immersed depth, is 1 ft., proving that 
 at no less depth or draught was there 62|- lbs. of upward pressure from 
 the water. As all water pressures act perpendicularly to the immersed 
 surface, only those on the bottom of the box act vertically, the others on 
 the sides acting horizontally. The horizontal pressures afford no support 
 whatever, their force being only exerted to crush in the sides of the vessel. 
 The whole of the pressure for the support of the vessel is therefore provided 
 by the vertical pressures on the bottom of the box, and their total pressure 
 must equal 621 lbs. in order to support the 621 ibs. of weight of the box. 
 This upward pressure is termed buoyancy. ISTow these upward pressures 
 upon which a floating object depends for its maintenance upon the surface 
 of the water, increase in force in direct proportion to the depth from the 
 surface. This we can easily prove. Suppose another 62^ lbs. be placed in 
 the box vessel, a. We should find, instead of the depth of immersion 
 remaining at 1 ft., that it had increased to 2 ft. exactly, as shown by fig. b. 
 
STRESS AND STRENGTH. 
 
 55 
 
 That is, the same area 
 of bottom surface is 
 exposed to exactly 
 twice the amount of 
 pressure ; and so on, 
 thepressure increasing 
 proportionally to the 
 depths of immersion. 
 
 Thus the total area 
 of immersed surface 
 has no relation what- 
 ever to the amount of 
 support given. 
 
 Fig. h has not twice 
 the total immersed 
 surface of fig. a, 
 though the draught 
 is exactly twice as 
 much, but it {h) cer- 
 tainly experiences con- 
 siderably more than 
 twice the total pres- 
 sure upon its total 
 immersed surface. 
 
 Owing to the in- 
 crease in pressure with 
 increase in draught, 
 we have seen tliat on 
 h there is twice the 
 upward pressure on 
 the bottom. Then ' 
 the sides of h for 1 ft. 
 down have the same 
 external pressure as 
 in fig, a, but for the 
 second foot down, the 
 pressure is twice as 
 great as on the sides 
 for the upper foot. 
 To sum up, there- 
 fore, the bottom of h 
 experiences twice as 
 much upward pressure 
 as the bottom of a 
 and the sides four 
 
 ^ 
 
 I'lii.'iiliii iiiiiil hi 
 
 ^ 
 
 .x-,\>''""',. 
 
 
 '//, '"'ill , ,\\"\v^ 
 
 '""i!,,'A«'' 
 
 [Ml ill l l|'li' " il||l l. i n '|i M | i | H 
 
 lii|iii|ii|'iMi"i|i|ii'iiiii;iiiili,|||ii,i 4 
 
 l|lilJlllllll|i||llll|l>llllll>ll<llllll<lil>>Ml>l 
 
 1 
 
 Iji!lil']lil''l'llii<ii''i"l>iiii"M"' 
 "iMilllnlllillllll jllllhlll I 
 
 <!i!|il!'l|ll|l'llH!MlllllM|lill|lr| 
 
 'MlMlillilijl. 
 
 :-f «3 
 
 PU| 
 
 Plh 
 
56 
 
 STEEL SHIPS. 
 
 times as much. This will be made clear by fig. 17 and the accompanying 
 notes. 
 
 The ratio of the increase in these horizontal pressures is most important, 
 for they show that on sach a surface as a watertight bulkhead to a peak 
 tank, or a deep ballast tank, in order to get maximum efficiency, the distri- 
 bution of strengthening materials should not be made uniformly over the 
 area of the bulkhead, but rather in accordance with the stresses endured 
 
 Fig. 16. — Water Pressures. 
 
 locally over the bulkhead owing to the water pressures. Did area of im- 
 mersed surface regulate the amount of support afforded by the water, then 
 such an object as d, fig. 15, ought to have exceptional support. But this is 
 not so, for the only support is again from the vertical pressures, which have 
 a very small area upon which to exert themselves, and consequently, in order 
 to receive the support necessary, a considerable depth has to be reached. 
 Had the same object been placed horizontally in the water as in g, fig. 15, 
 
 averag e Sl^lbs 
 
 averaqe SE^Itu 
 
 125 lbs 
 
 Fig. 17. — Total side pressure upon « = 1 sq. ft. x 4 sides x 31;| lbs. = 125 
 „ „ „ „ 6 = 2 „ „ X 4 „ X 621 ,^ . = 500 
 
 then, though the total immersed surface would be greatly reduced, yet the 
 much greater surface exposed to vertical pressure would require only slight 
 immersion in order to obtain the amount of support necessary. 
 
 Thus, in two vessels of different proportions but equal displacements, 
 the one whose draft is greatest will experience most crushing pressure on its 
 external surface. 
 
 Fig. 15, c, shows the nature "of the pressures upon a cylindrical object 
 floating in water. Here, although all the pressures are perpendicular 
 to the immersed surface, yet only a small proportion are truly vertical 
 
STRESS AND STRENGTH. 
 
 57 
 
 and horizontal. The others are more or less oblique, and afford support 
 in proportion to the degree of vertical tendency they possess, so that the 
 resultant of all the upward pressures is equal to the Aveight of the cylinder. 
 The total of the external pressures experienced in this case also, represent 
 a force exceeding the weight of the cylinder. 
 
 Fig. 15 e is a spherical object suspended as shown, its weight being 
 such that the water pressures are incapable of supporting it in any partially 
 immersed condition. The result is that it would sink. But if it is true that 
 the vertical pressures increase with the depth, surely it ought to be able 
 to receive a sufficient support at some definite depth below the surface. 
 At any rate, this is sometimes the reasoning of younger students. But 
 suppose the sphere is allowed further immersion, as in/, we shall see that 
 
 Water 
 
 Z 
 
 eve/ 
 
 
 
 
 \ A 
 
 
 
 A. 
 
 1 
 o 
 
 k 
 
 1 
 
 1 
 
 1 
 1 
 
 1 o 
 
 
 
 a> 
 
 1 
 
 " o 
 
 1 
 
 1 ' w 
 
 t— 
 
 1 
 1 1 
 ' • 1 
 
 
 
 
 i 
 
 1 
 1 
 1 
 
 — 12 
 
 Fig. 18. — Diagram illustrating the Calculation for amount of Water Pressure 
 on a single Frame of a floating Vessel. 
 
 it has really received no more ultimate support than at e. In these cases, 
 where there is total immersion, we have vertical pressures downwards upon 
 the upper surface of the sphere, as well as upwards on the lower surface, 
 and thus, just in the same proportion as the upward pressures increase with 
 depth, the downward pressures increase for the same reason also. There 
 is therefore no position of equilibrium for totally immersed objects between 
 the surface and the bottom of the water, unless, as in the case of submarine 
 vessels, some means exist for increasing the displacement, or of propelling 
 the vessel vertically. 
 
 Thus, on account of this increasing pressure as greater depths are 
 reached, objects which could easily endure the external pressure at or near 
 the surface, collapse entirely at greater depths from the excessive crushing 
 
58 
 
 STEEL SHIPS. 
 
 strain experienced. While fresh water exerts a vertical pressure of 62| lbs. 
 per square foot at 1 ft. depth, sea water (salt) registers 64 lbs. owing to its 
 greater density. 
 
 Pressure per Frame Space. — Bat suppose the designer of a vessel, in 
 considering the strength of the transverse framing, is desirous of knowing 
 the amount of water pressure upon the area of the space between any two 
 consecutive frames, or, what amounts to the same thing, the amount of 
 pressure per frame, the result may be arrived at in the following manner. 
 
 Let fig. 18 represent a section of the vessel up to the water level at 
 which the pressure is required. Assume the frames to be spaced 2 ft. 
 apart. Divide the half girth, which is 28 ft. from keel to waterline, into 
 divisions similar to those shown in the fig. As the pressures increase in 
 direct proportion to the depth, the pressure in lbs. upon the area of each 
 division is found by multiplying its area in square feet by 64 lbs. (salt 
 water), and this product by the distance of its centre of gravity below the 
 water level. 
 
 The calculation would be as follows : — 
 
 Spaces. 
 
 Areas sq. ft. 
 
 lbs. 
 
 Distance of C.G. 
 
 below water level. 
 
 ft. 
 
 lbs. 
 pressure. 
 
 2' X 10' 
 2' X 6' 
 2' X 12' 
 
 20 
 12 
 24 
 
 I 
 
 64 
 64 
 64 
 
 5-00 
 12-75 
 15-00 
 
 6,400 
 
 9,792 
 
 23,040 
 
 'ressure upon one side of vessel = 
 
 : 39,232 lbs. 
 2 
 
 
 
 78, 464 lbs. pressure upon 
 whole frame space for 
 both sides. 
 
 Estimated Pressure upon a Bulkhead and Centre of Pressure. — Or 
 
 again, suppose it is desired to determine the amount and distribution of the 
 pressure upon a bulkhead, assuming that a compartment be perforated, and 
 partly, or just filled with water ; and also how to apportion the structural 
 strength to best withstand such pressure. 
 
 Let fig. 19, a, represent the front elevation of the bulkhead up to the 
 height to which the water has risen. Being rectangular, the centre of its 
 area is obviously 6 ft. below the top. Then the total pressure upon the 
 bulkhead is 10 x 12 x 64 x 6 = 46,080 lbs. 
 
 But as we have seen, the water pressure increases directly in proportion 
 to the depth. The distribution of this pressure can be illustrated as 
 follows : — 
 
 Let AB (&) represent the bulkhead in section. Make BC equal to 
 A B, or, for the matter of that, it may be drawn to any scale. Join A C. 
 It follows that all horizontal lines from A B to AC will be proportional to 
 
STRESS AND STRENGTH. 
 
 59 
 
 the pressures at the various depths. The area of the triangle of pressure 
 ABC represents the total pressure upon the bulkhead A B, and its centre 
 of gravity represents the vertical height of the ' centre of pressure ' upon 
 the bulkhead, or the point at which the bending moment upon the 
 bulkhead attains its maximum. In the case of a triangle, the centre of 
 gravity is at one-third the height from the base, hence the centre of 
 pressure is situated 4 ft. from the bottom of the bulkhead. The shearing 
 stresses, on the other hand, increase with depth, and therefore reach a 
 maximum at the lowest depth, or, in other words, at the bottom of the 
 bulkhead. 
 
 Theoreticallj', the structural strength arranged vertically on a bulkhead 
 should gradually increase in dimensions down from the top to the point at 
 which the bending moment reaches its maximum (at the centre of pressure) 
 
 a 
 
 c P 
 
 -10 >J 
 
 Fig. 19. — Pressure upon Watertight Bulkhead. 
 
 and be again reduced towards the bottom. Similarly, structural strength 
 arranged horizontally should be distributed in the same manner. In actual 
 practice, however, as shown in Chapter VII., theoretical deductions often 
 require modification. It will be obvious, however, how essential it is that 
 the structure of the bulkhead be firmly secured at its bottom extremity. 
 How this is done we shall see more clearly when we come to deal with 
 the actual structure of bulkheads. 
 
 In the case of an actual ship bulkhead which is not absolutely 
 rectangular, the calculation for centre of pressure would be arranged as 
 follows : — 
 
 First divide the bulkhead into a number of parallel areas of equal 
 breadth, as shown in fig. 19, each in this case being 2 feet wide. Next, 
 determine the number of square feet in each area, and the distance of its 
 
6o 
 
 STEEL SHIPS. 
 
 centre of gravity below the top of the bulkhead. Then find the pressure 
 upon each area, and finally, the centre of the application of the whole of 
 these pressures, thus : — 
 
 i 
 
 
 
 Distance of 
 
 Lbs. 
 
 Approximate 
 Leverage for 
 
 Moments : 
 
 < 
 
 Area 
 
 Lbs. 
 
 Centre of Area 
 
 Pressure 
 
 foot lbs. 
 
 ■s 
 
 sq. feet. 
 
 below Top of 
 
 (Product of 
 
 (Product of 
 
 
 
 
 Bulkhead. 
 
 last three 
 
 Feet.* 
 
 last two 
 
 S 
 
 
 
 Feet. 
 
 Columns). 
 
 Columns). 
 
 1 
 
 20 
 
 64 
 
 1 
 
 1,280 
 
 1 
 
 1,280 
 
 2 
 
 20 
 
 64 
 
 3 
 
 3,840 
 
 3 
 
 11,520 
 
 8 
 
 20 
 
 64 
 
 5 
 
 6,400 
 
 5 
 
 32,000 
 
 4 
 
 20 
 
 64 
 
 7 
 
 8,960 
 
 7 
 
 62,720 
 
 5 
 
 20 
 
 64 
 
 9 
 
 11,520 
 
 9 
 
 103,680 
 
 6 
 
 20 
 
 64 
 
 11 
 
 14,080 
 
 11 
 
 154,880 
 
 
 
 
 
 46,080 
 
 46,080)366,080 
 
 
 
 
 Centre of Pressure below top of Bulk 
 
 head =7-94 ft. 
 
 
 The result is nearly 8 feet, practically f depth down from top of 
 bulkhead. 
 
 Pressure upon a Tank Top. — As further illustration of water pressure, 
 take a water ballast tank, filled to its utmost capacity by means of a pump 
 whose discharge pressure is, say, 10 lbs. per square inch, and no means of 
 escape provided. 
 
 Supposing the tank be 30 feet long and 20 feet wide, and pumping to 
 proceed after the tank is full, the pressure upon the top of the tank will 
 amount to 10 lbs. per square inch, or upon the whole surface : — 
 
 30x20x144x10 
 
 2240 
 
 = 385- tons. 
 
 Or, to take another example, suppose that a pipe of, say, 1 square inch 
 
 sectional area, 20 ft. long, and open at the top end, be placed vertically in 
 
 the top of a double bottom tank, and that water be pumped into it until it 
 
 overflows at the upper extremity of the pipe. The pressure at the 
 
 bottom of the pipe would be equal to the weight of water in the pipe 
 
 20 X 12 62i 
 = ~T*70Q~x~i = 8'7 lbs. (fresh water taken), that is, the pressure upon 
 1728 1 
 
 the inner surface of the top of the tank is 8*7 lbs. per square inch, or, upon 
 
 20x20x144x8-7 
 
 an area 20 ft. by 20 ft. the pressure would be '^'^ ^ ""^ T.^^^"^ = 223*7 
 
 •^ ^ 2240 
 
 tons. 
 
 This is the method usually adopted in testing the watertightness and 
 general efficiency of cellular, deep, and other tanks. 
 
 The value of sufficient and properly disposed air pipes, apart altogether 
 
 * The horizontal strips being taken narrow enough, this approximate leverage is 
 sufficiently accurate for pr-actical purposes. 
 
STRESS AND STRENGTH. 
 
 6i 
 
 from any consideration of ventilation, will be obvious, and the possibility of 
 causing damage to an inner bottom or other tank top, through pumping up 
 these tanks with air pipes closed, and no other means of exit for air or 
 overflow water, will also be apparent. 
 
 Tendency to Transverse Deformation, — Having shown by the fore- 
 going examples the mode of calculating the amount of pressure upon an 
 immersed surface, it is proposed to illustrate, by the following series of 
 sketches (figs. 20-24), the tendency to deformation of transverse form 
 in the hulls of ships caused by these pressures, as well as by other crushing 
 forces which are experienced under varying circumstances. It should be 
 understood, however, that though the tendency of any such deformation as 
 may be illustrated undoubtedly exists, yet in a well-constructed vessel, such 
 tendency never betrays its existence, simply because the structural strength 
 has been intelligently distributed to resist it, that is to say, to render it 
 practically imperceptible. 
 
 Fig. 20. — Showing direction of Water Pressures upon Immersed Hull of Ship. 
 
 Here the transverse form of the vessel floating in an upright condition is 
 shown, and upon its immersed surface are indicated the lines of water 
 pressures. 
 
 The resulting tendency, owing to the application of the afore-mentioned 
 pressures (see page 54), is illustrated in fig. 21. To resist such deformation, 
 the vessel must have efficient transverse girders or frames, and especially so 
 along the bottom where the pressures are greatest. The topsides are kept 
 in place by the beams, which are connected to the frame heads, and act as 
 both struts and ties. The strength to the sides is assisted at intervals of the 
 depth by transverse struts in the form of tiers of beams. The bottom is 
 stiffened by means of deep transverse floor plates placed vertically 
 (usually on every frame), which are further assisted by the keelsons, and by 
 being tied to the decks by means of the pillars, numbering at least one in 
 the breadth. 
 
62 
 
 STEEL SHIPS. 
 
 In fig. 22 we see the same transverse section of the vessel subject to 
 the same water pressures, but, in addition, it is supposed that a severe local 
 stress is caused by the concentration of some very heavy deadweight such 
 as cargo of great specific gravity; deep water-ballast tanks; engines, 
 
 
 I'lii,;: 
 mil 
 
 M.r.v^ 
 
 
 Fig. 21. — Tendency to Transverse Deformation due to Water Pressures. 
 
 boilers, and bunkers (especially in a light ship). The tendency is to elon- 
 gate the vessel vertically, and to cause the insufficiently supported bottom 
 to droop ; while the tendency for the topsides to contract causes the deck to 
 spring in the middle. Eesistance is offered to such deformation by the 
 
 ;\tvv^ 
 
 Fig. 22. — Sliowiug tendency to produce Vertical Elongation. 
 
 same transverse framing as enumerated for fig. 21, assisted by the same 
 longitudinal girders. Here, especially, is seen the great value of pillars 
 acting as ties, binding together the top and bottom of the vessel, and hold- 
 
STRESS AND STRENGTH. 
 
 63 
 
 ing the beams and floors in their true relative positions. The value of 
 efficient means of connection betvreen the side framing and beams, and side 
 framing and floors in the form of webs, will also be very evident. 
 
 Neglecting the tendency to deformation caused by the water pressures, 
 in fig. 23 we have assumed the deck to be subject to a heavy weight placed 
 upon it, producing the tendency to droop at the middle. On most vessels 
 there is considerable deck weight in the form of winches, windlass, houses, 
 bollards, and other deck furniture. In addition, many vessels carry very 
 heavy deck cargoes in the form of timber, etc. Bat apart from intentional 
 deck loads, huge heavy seas are sometimes shipped upon deck, producing 
 enormous and sudden stresses, which, even in large new vessels, built to the 
 highest class, have been known to sheer the rivets in the beam knees, 
 damage the hold stanchions, carry away hatch coamings, and effect other 
 
 -';;;-';;Miv::::'M^^;:;i;;iii;:;'l'^>;:-- 
 '"•■'ii'iiiM:;!;;;;i;ii!:;::;;;i;iiii'i"''" 
 
 Fig. 23. — Showing tendency for deck to sag at the middle, owing to heavy- 
 Deck Weights or Shipped Seas. 
 
 damage, in addition to producing considerable sinkage in the deck, as 
 illustrated in the above figure. 
 
 The principal structural parts resisting such deformation as described, 
 are deck beams connected to the side framing by efficient knees well riveted, 
 and supported by thoroughly efficient pillars well connected at heads and 
 heels. 
 
 In fig. 24 the same vessel is supposed to be subject to violent rolling in 
 wave water, which movements, aggravated by ^he inertia of cargo in the holds, 
 or water in large ballast tanks, tend to severely rack the transverse form of 
 the vessel, and to produce such distortion as illustrated. To prevent this 
 deformation, the localities where such working is most likely to be 
 experienced — viz., at the junction of the uppermost deck with the sides, 
 and at the bilge — should be well strengthened. This is done by making 
 the connection between the beams and the side framing by webs of plating 
 (knees), and connecting the floors to the side framing by similar, but larger 
 
64 STEEL SHIPS. 
 
 webs, — turned up floors in the case of ordinary floors, and brackets where 
 double bottoms are fitted. See figs. 6 and 7.* 
 
 There are certain other severe local stresses, chiefly aff"ecting the 
 transverse form, to which ships are sometimes subject. These, however, 
 are generally the result of ignorance on the part of those responsible, and 
 neither the naval architect nor the rules of any classification society pretend 
 to fully provide for them. 
 
 The first of these is the case where a vessel is in dry dock with 
 an insufiicient number of blocks supporting the keel. Consequently 
 enormous pressure is experienced on the bottom in the way of each block, 
 with the accompanying tendency to crush in the bottom and produce great 
 distortion locally in the transverse form^ and serious damage to the bottom. 
 
 The second is that of a vessel in dry dock, when an ample number of 
 blocks has been placed under the keel, but insufficient support afi'orded at 
 the bilges owing to the bilge blocks being too widely spaced. Here the 
 tendency is for the bilges to droop. Such straining is all the more severe 
 
 Fig. 24. — lUustratiiig tendency to deformation due to Racking Stresses. 
 
 and complicated when the vessel has cargo in the holds, or, in addition to 
 the weight of the engines and boilers, side or cross bunkers are full, or 
 deep water-ballast tanks happen to be full. An approximation to the 
 amount of the transverse bending moment may be obtained by multiplying 
 the weight (W) in any locality on one side between the keel and side, 
 by the distance of its centre of gravity out from the centre of the keel 
 (.r). 
 
 "VV X a; = bending moment. 
 
 And lastly, wherever there is very excessive preponderance of weight 
 over buoyancy, or buoyancy over weight, in any locality in a vessel, though 
 the effect may not be to produce a longitudinal bending moment, yet it tends 
 to produce local bulging. 
 
 * Transverse complete or partial bulkheads in holds and 'tween decks, in addition 
 to providing excellent support to tlie longitudinal framing, are of great value in resisting 
 such deformation as is liable to be jDroduced by severe transverse racking stresses. 
 
STRESS AND STRENGTH. 
 
 65 
 
 Longitudinal Stresses in still water and Tendency to Longitudinal 
 Deformation. — While such stresses as illustrated from fig. 20 to 24 may 
 be very severe both in smooth and wave water, they are probably the most 
 easily provided against in the construction of a ship. They are known as 
 ' Transverse Stresses.' When we come to deal with fore and aft, or 
 longitudinal stresses, much more difficulty, however, is experienced. 
 
 In fig. 25 we have a longitudinal elevation of a ship floating in smooth 
 water, at the water line indicated. The stresses borne by the hull in con- 
 sequence of the water pressures on every part of the immersed surface, tend- 
 
 IT 
 
 I » I 
 
 I V 
 
 /. I 
 
 M I I P^ 
 
 J I L 
 
 Fig. 25. — Diagram illustrating the excesses of Buoyancy and Weight at internals 
 throughout the length of a vessel in a li(/ht condition. 
 Upward arrowed lines represent buoyaiacy in excess, and downward arrowed lines 
 represent weight in excess. Where both arrowe<l lines appear, the weight and 
 buoj-ancy are equal. The thick vertical lines A B C D represent bulkheads. The 
 space between B and C = engine and boiler space. Tlie spaces between A B and 
 C D = cargo spaces. JVote. — The draught on the diagram has been exaggerated, 
 in order to more clearly illustrate the weight and buoyancy by means of the arrow. 
 
 ing to crush in the bottom, sides, and ends, are unavoidable. But even 
 though lying motionless u[)on the water, it is quite possible that other stresses 
 of considerable importance may exist. 
 
 Let fig. 26 be a box-shaped vessel which, though drawn to a smaller 
 
 ;iriiiii':i;iii,iii;i iiiii"iiiii|i|i|iir:i' 
 
 i.'inn |l,M'iiiiiii|ii,iiiiii|i|| ,111 
 
 ">Mt> ihhlllll|llllll<llllilMl|i|llll||lll 
 
 i i 'ii!! !' | i !i '"ii i' i l '" ' ;' '! i l 'Hi| i' : i ii l i i! i !'i| iM' i 'll li! | | 'ii;i i in iiii i i i i i iii i i ii iiii|i iii i i iii iiiM i |i i |. 
 
 hiiiii'i|ri'»i'<iiMi';;,;;Mi|,.i,;i;ii,!i|!i'i,.,Mi Iinti: :""""i;'::;"::'::ill;M:in";:i!' 
 
 iiii|liiiiiiiiiMiiiiiiiiiii,iiiiiiiii,ii|iiiiiiiiiii|ii |iiiiiii|iiii;,:!;iii:;ii::'iiM!'-'.i'!i'i.lMi..'iii!ii.i 
 
 Fig. 26. — Box-shaped vessel uniformly loaded all fore and aft with weight and buoy- 
 ancy exactly counterbalancing each other. No tendency to deformation in a 
 longitudinal direction exists. 
 
 scale, is supposed to be of the same extreme dimensions as fig, 25. We 
 will assume that the weight of the cargo taken on board has been uni- 
 formly distributed throughout the length, thereby producing no alteration to 
 the trim. Then the uniform upward pressure of water along the bottom 
 of the vessel is exactly counterbalanced by the uniform downward pressure 
 of weight throughout the length. Such uniform counterbalance of forces 
 does not produce any stress tending to longitudinal deformation. 
 
 But on returning to our actual ship in fig. 25, no such uniformity is 
 
 5 
 
66 STEEL SHIPS. 
 
 found. If it be assumed that the ship floats light, we find considerable 
 difference between the buoyancy and the weight locally throughout the 
 length, though for a ship, as for every other floating object, the law must 
 hold good that the total weight and the total buoyancy equal each other. 
 Thus, at amidships, or wherever the engines and boilers are situated, there 
 is a great concentration of weight, and in many cases this is increased by 
 the erection of bridges, houses, etc. Then, again, at the ends of vessels 
 there is usually a concentration of heavy weight. At the fore end there 
 is often a forecastle, also the anchors, chains, windlass, solid iron stem bar, 
 anchor crane, heavy mooring arrangements, and in addition, special 
 stiffening to prevent 'panting' (see page 72). At the after end a poop is 
 often built. A heavy soHd steel or iron stern frame, propeller, tail end shaft- 
 ing, steering gear, mooring arrangements, etc., are also found in this part. 
 
 Turning from weight to buoyancy, we see where and how the difference 
 previously mentioned arises. Wherever we have greatest immersed bulk, 
 we have greatest buoyancy, simply because where there is this bulk there 
 must be in some shape or other a large surface exposed to the upward or 
 buoyant pressures. Hence, keeping before our minds the usual form ot 
 a vessel's immersed hull, we see that the greatest buoyancy is at amidships, 
 and, in cargo vessels, this may extend for a considerable proportion of the 
 vessel's length, with practical uniformity, fore and aft of midships. As the 
 ends, however, are approached, a rapid change of form is observed as the 
 vessel fines down towards the stem and stern, causing an enormous reduction 
 in immersed bulk, and hence in buoyancy. Thus, in traversing the length 
 especially of a steam ship (see fig. 25), we are struck by the unequal 
 distribution of the weight and the support, or buoyancy. The heavy after 
 end has little and most inadequate support from the water, owing to the 
 fineness of its underwater form. Bat as we come forward to the hold space 
 the ship rapidly increases in fulness, and is, moreover, comparatively free 
 from exceptional weight. The consequence is that the buoyancy rapidly 
 increases, and is eventually in excess of the weight. At amidships, where 
 the machinery is situated, the weight is again in excess of the buoyancy in 
 a light ship. Forward, similar conditions to those found at the after end 
 exist. At first, after leaving the machinery space, when the heavy engines 
 and boilers are amidships, the buoyancy is considerably in excess of the 
 weight, then it gradually diminishes as the vessel fines down, until 
 eventually the weight of the fore end preponderates considerably over 
 the buoyancy. Thus throughout the length of a ship, a series of stresses 
 are experienced which tend to strain and produce deformation in the 
 longitudinal direction.* 
 
 * When a bar, loaded at each end, and supported at the middle of its length, droops 
 towards the ends owing to the bending moment experienced, the visible deformation 
 produced is popularly termed 'hogging.' 
 
 "When the bar is supported at each end, and loaded at the middle of the length, 
 causing the bar to droop in way of the weight, the visible deformation produced is 
 popularly termed 'sagging.' 
 
STRESS AND STRENGTH. 
 
 67 
 
 While it is owing to the existence of great excesses of weight over 
 buoyancy, or vice versd, throughout the length of a ship, that longitudinal 
 bending moments are created, tending to produce longitudinal deformation, 
 it must not be imagined, as shall be shown, that every such excess 
 necessarily produces a corresponding longitudinal bending moment. For 
 instance, a small excess caused by an isolated concentration of weight in 
 the middle, say, of a huge bulk of excess of buoyancy would not necessarily 
 produce a longitudinal bending moment, though the proportionately small 
 excess of local weight would certainly tend to produce local bulging. 
 
 Let fig. 27 represent a vessel floating in a light condition with, say, 
 
 And 
 Boilers 
 
 Fig. 27. — A small Steam Vessel with fore and after peak tiinks full, and bunkers full. 
 
 engines and boilers situated amidships, bunkers full, and a large peak 
 ballast tank at each end full. If the holds were completely empty, it is 
 clear that in an ordinary cargo vessel the greatest excesses of support over 
 weight would be in way of the two hold spaces. The excesses of weight 
 over buoyancy at the ends and amidships, owing to the peak water-ballast 
 tanks and the machinery and bunkers, at first incline one to conclude that 
 considerable straining will in consequence be produced towards the ends and 
 in the region of amidships, and ' sagging ' in these localities possibly ensue. 
 
 jr-a 
 
 H 
 
 X-->^--X-->^ X- - >^ 
 
 m. 
 
 A 
 
 W2 
 
 -X ^ 
 
 '^ Pi 
 
 Fig. 28. — Showing the nature of the bending moments upon a bar loaded 
 somewhat similarly to the ship in fig. 27. 
 
 Bending Moment on a Loaded Bar. — But let us consider the same 
 features in a simpler form (see fig. 28). 
 
 Let AB be a bar of the same length as the ship. W and W3 are 
 weights suspended at the ends, representing, in some measure, the peak 
 ballast tanks, and "W^ and Wg (one weight) at the middle of the bar, 
 representing the weights of machinery and bunkers. Let the supports 
 P and Pi be placed midway between the middle and the ends of the bar. 
 
 Now W, "Wp Wg, and Wg are by hypothesis, in this case, supposed to be 
 
68 STEEL SHIPS. 
 
 equal to one another in value, and, instead of the double weight W^ and 
 Wg producing a 'sagging' moment at the middle of the bar, no bending 
 moment whatever exists there. 
 
 Suppose the bar to be perfectly rigid all fore and aft. The pressure upon 
 P will be equal to the weight of W + W^ and the pressure upon the support 
 P^ will be equal to W., + Wg. 
 
 JS'ot only are the pressures on the supports exactly equal to the weights, 
 but they exactly poise their respective weights in equilibrium, W and "W^ 
 each being X distance from P, and "Wg and Wg each being the same X 
 distance from P-^. 
 
 The bending moments, therefore, instead of being greatest at the middle 
 of the bar, are greatest at the points of support P and P^, where they are 
 W X A". This quantity being set off to any arbitrary scale of moments 
 above the bar at the point of support, and the point so obtained joined to 
 the ends of the bar (see dotted lines), a graphic representation of the 
 bending moment on the beam is produced. A similar maximum bending 
 moment would have been obtained by W^ x X, or W., x X, or Wg x X. Or 
 again, for the sake of example, the same result will be got by taking the 
 moments up and down about any other point, say A. 
 
 Bending moment about A = 
 
 {(W X no distance) + (W^ + W^) 2x + (Wg x ix)} - {P/ + (P^ x 3/)} 
 
 In like manner, the actual bending moment at the point ' a ' could be 
 found by taking the algebraic sum of the moments up to, or about any 
 point in the bar, or even outside of the bar. 
 
 The moment of any force acting upon the bar at any point is obtained by 
 multiplying such force by the distance of its line of action (or position of its 
 centre of gravity) from such point. 
 
 Prom the foregoing example, it is further evident that had the weight 
 Wj + Wg been greater than either twice W or Wg, there would have been at 
 the middle of the length a stress tending to a perceptible 'sagging' 
 moment, and that had W^ + Wg been less than either twice W or Wo, 
 there would have been a '^hogging' moment over the whole length, tend- 
 ing to cause the beam to droop from the middle of the length towards 
 the ends. 
 
 The more Wi + Wg exceeds twice either W or W3, the further does the 
 sagging moment extend on each side of the middle of the length of the bar 
 before a hogging moment is experienced. 
 
 Possible effect of a had disposition of Weight in a Ship. — But while a ship 
 is a much more complicated object with which to deal than a bar, loaded 
 and supported as described, and the nature of the actual bending moments 
 more difiScult to ascertain, because the loads and supports are more 
 distributed and more complicated, yet the principles involved are identical, 
 though a greater amount of labour is entailed in arriving at the results. It 
 will, nevertheless, be obvious how enormously the longitudinal bending 
 
STRESS AND STRENGTH. 69 
 
 moments, even in smooth Avater, may be increased in the operation of 
 ballasting and loading. For example, we have seen how little support is 
 aflforded to the fore and after ends of a vessel, the excess of the weight 
 over the buoyancy of these parts practically hanging upon the super- 
 abundant buoyancy of the adjacent parts of tlie hull. Suppose, as is often 
 the case, that a large peak tank is filled, capable of containing 80 or more 
 tons of water, it can easily be imagined how the stress we have observed to 
 exist may be greatly aggravated. 
 
 While, ttndouhtedly, the hrealdng of propeller shafts, ivhicJi is so common 
 among steamers in light or ballast conditions, is attributable in the majority 
 of cases to the fact that they are under- bal lasted ivith their propellers only 
 partially immersed, and in bad loeather are subjected to violent racing and 
 sudden jerhing and checking as the propeller blades strike the ivater, yet it is 
 not at all unlikely that, in many cases, damage to shafting may arise from 
 longitudinal bending or twisting caused by the enormous iveights concen- 
 trated, often in ignorance, at the aftermost extremity of ships, and especially 
 when, in addition, poops are used for heavy cargo or bunker coal. 
 
 While the tank in a continuous double bottom is capable of carrying a 
 very large amount of water, its capacity throughout its length is generally 
 governed by the form of the ship, and it is not therefore calculated to 
 produce severe stresses in smooth water. 
 
 Fig. 29. — Showing tendency to ' hog.' 
 
 Then in regard to loading, it is evident that excessive straining would 
 be produced by the disposing of heavy cargo where the buoyancy is 
 naturally least. 
 
 Longitudinal Stresses in loave water and. Tendency to Longitudinal 
 Deformation. — Severe as such stresses as have been enumerated may be in 
 still water, they only assume their true proportions when in wave water. 
 Probably by the aid of the following diagrams showing a ship among 
 waves, we shall best be able, in a simple, graphic manner, to give 
 prominence to the principal tendencies to deformation, which the longi- 
 tudinal bending moments exert upon a ship's structure. 
 
 In fig. 29 we have a flush decked vessel supported at the middle of her 
 length upon a wave, producing great excess of buoyancy at amidships, and 
 increased deficiency towards the ends. Any vessel, waveborne in this 
 manner, experiences her severest stresses on each side amidships. The 
 tendency is, therefore, for the ends to droop, or as we saj', for the vessel to 
 hog. Sometimes, in looking along the gunwale of an old ship, and especially 
 a wooden one, the drooping of the ends is distinctly seen, and the vessel is 
 said to be hog-backed. 
 
70 STEEL SHIPS. 
 
 Fig. 30 represents exactly the reverse of the foregoing example. Here, 
 each end of the vessel is supported upon a wave crest, and a great deficiency 
 is produced in the buoyancy at amidships. In this case, the tendency is 
 for the vessel to sag amidships, owing to the sagging moment experienced. 
 
 To resist such deformation as illustrated in the two foregoing diagrams, 
 it is necessary that a vessel be of reasonable depth and possess a sufficient 
 number of properly disposed longitudinal girders or I'rames. Thus along 
 the bottom we find the girders, called ' keelsons,' varying in number and 
 form with the size of the vessel. In all cases, whether these keelsons be 
 fitted on the top of or between the floors, they should be well connected 
 to the transverse framework. 
 
 Along the sides of the vessel somewhat similar girders are found, though 
 no longer called ' keelsons,' but * stringers.' These also vary in number 
 and size according to the depth and size of the vessel. (See figs. 7 to 12.) 
 
 While it is necessary to cover the framing with a skin, the fact should 
 not be overlooked that the shell plating forms au enormous contribution to 
 the longitudinal strength. For example, each of the vertical sides, stiffened 
 and preventfid from buckling by the transverse framing, forms a huge girder 
 in itself of very great strength, while these side girders are united, and the 
 
 Fig. 30. — Showing tendency/ to ' sag.' 
 
 ship girder further strengthened by the bottom shell plating and decks. 
 Stringers and decks render additional service in resisting longitudinal 
 twisting when a vessel is crossing skew seas and rolling heavily. In con- 
 sidering the actual strength of a ship girder at a later stage, we shall see how 
 additional strength is obtained by inner bottom plating in vessels with 
 double bottoms, by steel or iron decks, and by longitudinal middle line 
 bulkheads. 
 
 While it is often necessary to consider the structural values of the 
 longitudinal and transverse framing independently, yet, how utterly 
 dependent the one is upon the other for the fullest development of its effici- 
 ency, is easily seen. While strong, well constructed transverse Avatertight 
 bulkheads are of great value from a safety point of view, in the event of 
 perforation of the shell plating in way of any compartment, yet their 
 value in stiffening the longitudinal framing and holding it to its work is 
 of the greatest importance. This was especially illustrated in the case of 
 the ' Great Eastern,' which was built upon the longitudinal girder system 
 and contained very little transverse framing (see fig. 60, and Chapter VL). 
 This remarkable vessel depended in a great measure for her transverse 
 
STRESS AND STRENGTH. 
 
 71 
 
 strength upon tlie enormous strength of her well-arranged longitudinal 
 framing, covered by au inner and outer watertight shell, assisted by 
 numerous transverse bulkheads. 
 
 Then again, a vessel in a seaway does not always float in the upright 
 condition, but often heels to very considerable angles of inclination, when 
 a noticeable difference takes place in the structural resistance offered to the 
 bending and twisting moments experienced, the stringers being called upon 
 to perform in a measure the function of keelsons, and keelsons the function 
 of stringers. 
 
 Fig. 31 illustrates a vessel with a poop, a bridge of some length, and 
 
 Fig. 31. — Showing tendency to fracture at ends of bridge. 
 
 a forecastle, supported upon a wave in the middle of her length. The 
 stress experienced is similar to that in fig. 29, bat the locality of likely 
 damage resulting from the sudden reduction in longitudinal strength in 
 the midship region is more marked. 
 
 The erection of a bridge, increasing the depth of the ship girder, adds 
 greatly to the longitudinal strength of the ship over that part covered by 
 the bridge, to resist damage (i.e. permanent deformation). But however 
 valuable bridges may be, the sudden termination of these erections produces 
 a weak section in the ship girder, which, unless specially provided for in some 
 such manner as we shall observe at a later stage and when dealing more 
 intimately with construction, would undoubtedly result in serious damage. 
 
 Fig. 32. — Showing tendency to buckle at ends of bridge. 
 
 Fig. 32 is similar to fig. 31, excepting that the vessel is now waveborne 
 at the ends with the same tendency to deformation, due to the same cause 
 as indicated in the previous example, fig. 31. 
 
 Similarly in vessels with raised quarter decks terminating somewhere 
 on the middle length, without or with bridge houses to which they are 
 connected (fig. 50) ; vessels with partial awning decks (fig. 51) ; or 
 in vessels where the poop and bridge are combined (fig. 53) ; weakness 
 in longitudinal strength caused by the sudden termination of such 
 erections, necessitates careful attention to the structural arrangement 
 at these parts. 
 
72 STEEL SHIPS. 
 
 Before any of the forms of deformation just illustrated could pos- 
 sibly be effected in a well designed ship, damage of a serious nature 
 must have been wrought upon the structure. In a bad design, or in an 
 old vessel, the nature of such damage would probably show most in the 
 butts of the upper deck stringer plates and butts of topside plating (sheer 
 strake, etc.) evincing signs of working, or opening, or in the rivets in these 
 butts shearing, and, under exceptional circumstances, even these plates 
 shearing through some line of rivet holes. 
 
 It is scarcely necessary to add that in the event of such damage as 
 just described, the material must have been stressed far beyond its limit 
 of elasticity, and if actual shearing of plates occurs, beyond the ultimate 
 shearing strength. Damage of this nature could only arise through ignorance 
 of the principles of loading, if the ship be well designed, or, where the 
 vessel is properly loaded, to defective design combined probably with too 
 light scantlings for the deadweight carried. 
 
 Local Stresses. — There are still several stresses, chiefly of a local 
 character, which should be enumerated. 
 
 1. Panting Stresses. — These stresses are caused by the pressures 
 upon the immersed fore end of vessels as they are propelled through 
 the water, with the tendency to make this part Avork in and out in a 
 manner resembling the action of ' panting ' — hence the term. 
 
 It follows, then, that the higher the speed the greater will be the 
 pressure, and the severer the stress. Small vessels of considerable speed 
 with thin shell plating often show the result of this pressure by the 
 plating between the frames bulging inwards. Bluff fore ends also expose 
 a greater surface to such pressure. 
 
 All or part of the following means may be adopted to strengthen 
 vessels to resist damage from such stresses : — Thicker shell plating Vf'ith 
 a closer spacing of transverse frames, deep floors, extra tiers of beams, 
 and stringers, with large breast-hooks.* 
 
 2. Stresses due to Propulsion hij Steam. — As already shown, there may 
 be considerable inequality in the pressure of weight and buoyancy in 
 way of the engines and boilers in certain light, load, and ballast 
 conditions. But the most important stresses produced by the machinery 
 in many vessels are those of vibration, though at the same time it must 
 
 * In addition to panting stresses, a vessel is subject to severe thumping under the 
 bows as she drives ahead, and dives into head seas. These stresses are likely to be 
 most severe in vessels of high speed and at a light draught, especially if they be 
 of full form under the fore foot, or ' club-footed ' ; as is the term in shipyard practice. 
 These stresses have in some instances been so severe as to bulge in the fi'ames and 
 shell plating in the locality indicated. To prevent this and to afford the necessary 
 stiffening, it is advisable that the frames be doubled from margin plate to margin 
 plate in double bottom tanks and from bilge to bilge where no such tanks exist, 
 for about one-fifth of the vessel's length from the stem, and, in addition, to fit 
 intercostal plates between the floors forward of the collision bulkhead and extending 
 well down on to the keel plate, and for such a distance forward towards the stem as 
 the form of the vessel shows to be necessary. See figs. 46, 47, 49. 
 
STRESS AND STRENGTH. "] 2) 
 
 be remembered that vibration is not necessarily an indication of 
 weakness, nor even that stresses are being endured. All bars, beams, 
 and girders bave their natural vibratory periods, and it is when the 
 natural period of vibration of the engines (revolutions) coincides with, 
 or synchronises with that of the ship girder, that vibration may become 
 very prominent and dangerous. 
 
 The position of the machinery in a vessel may considerably influence 
 the degree of vibration experienced, and therefore from this cause alone 
 a strong vessel may sufifer an unusual amount of vibration.* 
 
 Vibration may also be the result of weakness in the ship structure, 
 and when this is the case, measures should immediately be taken to stiffen 
 up the hull and add the necessary strength, t 
 
 Excessive vibration from whatever cause should receive careful attention, 
 for its continuance must ultimately seriously strain the structure, and tend 
 to loosen rivets and open butts and produce leakage. To provide against 
 this, the following precautions may be adopted : additional keelsons under 
 engines and boilers ; strong and well connected engine seating carefully 
 scarphed forward and aft of engine space ; well-built thrust seating and shaft 
 stools ; deep floors aft, and the opposite sides of the transverse framing and 
 sides of the ship aft well tied by means of beams and stringer plates, 
 or transverse plate webs; stout shell plates connecting the stern frame 
 to the hull ; additional transverse stiffening in the engine and boiler 
 space in the form of deeper framing, web frames, strong beams, 
 and double reverse frames under engines and boilers, as well as efficient 
 pillaring. 
 
 3. Stresses due to Propulsion hy Sail. — These stresses are experienced 
 mostly by sailing vessels, whose lofty masts, heavy yards, and large sail 
 area, when exposed to wind pressure, transmit severe racking stresses to 
 the hull. To prevent damage from this source, the masts should be firmly 
 secured at the heel and deck, and further supported by a good spread of 
 shrouds, stays, etc. To transmit and distribute the stresses at the deck, 
 if no steel or iron deck is laid, the mast should pass through a stout deck 
 plate called a ' mast partner,' which in turn should be brought into intimate 
 association and connection with the neighbouring beams and stringer plates, 
 by means of tie plates carried diagonally across the deck. Owing to the 
 severity of the transverse stresses experienced particularly by sailing vessels, 
 combined with their lack of bulkheads, it follows that the strength of the 
 
 * It will thus be apparent that vibration in many instances depends to a considerable 
 extent upon the speed at which the engines are working. For instance, some vessels 
 are remarkably steady, and no perceptible signs of vibration are experienced when the 
 engines are going full speed. But the same vessels will give indications of slight 
 vibration when the number of revolutions is reduced. Should the speed be still further 
 reduced, and synchronism obviated, the vibration will diminish or disappear altogether. 
 
 t Increased depth conduces to minimise vibration, especially in long vessels. The 
 importance, therefore, of canying the strength well up to the topsides in vessels with 
 continuous erections is apparent. 
 
74 STEEL SHIPS. 
 
 transverse framing (including beams) and the depth of the beam knees 
 should exceed that required for steamers. 
 
 4. Stresses caused by loading heavy cargoes on deck, and the shipping of 
 heavy seas. — What is to be hoped for is the complete stoppage of what is 
 now the unusual and dying-out practice of taking out part of the hold 
 pillars when loading timber cargoes in the hold, and afterwards piling up 
 similar cargo on deck. Such a practice is quite sufficient to account for the 
 shearing of beam knee rivets, and other damage. As a matter of fact, 
 under such conditions of loading, a vessel requires all the pillar and beam 
 support that can be given her. 
 
 Those who know little of the sea experience of ships, seldom dream 
 what enormous weights of water are sometimes, in the most sudden 
 manner, shipped on deck in boisterous weather. This alone on many an 
 occasion has been sufficient to inflict severe damage upon decks, hold 
 stanchions, beam knees, hatch coamings, bulwarks, etc., not to mention the 
 tremendous force of such seas as they dash against bridge and poop front 
 bulkheads sometimes with very disastrous results, and in some authenticated 
 cases, even causing the foundering of a ship after being struck by a single 
 sea. Such bulkheads should be well strengthened with good coaming plates, 
 and vertical stiffeners kneed to the decks at the top and bottom. Hold pillars, 
 with good heads and heels well connected, or some equivalent structural 
 support, are most essential to support decks under such trying circumstances. 
 Whenever heavy permanent deck weights are carried, good supports or 
 equivalent beams should be provided. 
 
 6. Rudder Stresses. — The enormous pressure upon the surfaces of 
 rudders, especially in large, heavy, high-speed vessels when turning, throws 
 great twisting stress upon the rudder stock. The necessity for adequate 
 diameter of stock in proportion to the stress experienced is obvious. 
 
 Where frame rudders plated on each side are adopted, it is most 
 important that the space between the two plates be well filled with timber 
 so that the whole is one solid mass, with no possibility of panting, for even 
 with this precaution, rudder rivets sometimes work loose. 
 
 6. Stresses from loading aground. — The trades pursued by some vessels 
 cause them to enter rivers and tidal harbours where, when the tidfe has 
 fallen, they lie aground. Especially where the bottom is of an uneven 
 stony nature, the weight of cargo which is being loaded or discharged, in 
 addition to the vessel's own weight, tends to crush or bulge in the bottom, 
 and if the bottom of the vessel is only supported at intervals in her length, 
 even bending may occur. The possibility of this latter deformation should 
 be avoided. To strengthen the bottom in order to resist bulging, the bottom 
 plating should be of increased thickness, and be well supported with keelsons, 
 which come down to and are connected with the bottom plating. 
 
 7. Stresses upon vessels lohich are temporarily only partially Water- 
 borne. — One of the worst examples of this condition is that in which 
 a vessel runs aground, say, upon a sandbank at high tide, and remains fast 
 
STRESS AND STRENGTH. 75 
 
 at one end while the other is afloat. The stress endured assumes its worst 
 proportions when the tide has fallen, and the midship portion loses buoyant 
 support in a greater or less degree. This stress is sufficient to so strain 
 many a ship as to break her back. Equally bad is the condition of a vessel 
 stuck at the middle of her length upon a sandbank, so that when the tide 
 has fallen, the ends are left wholly or partially unsupported. 
 
 Stresses of such an extreme nature as these the naval architect does 
 not pretend to cover, and he does not hold himself responsible for the 
 results w'hich may accrue. 
 
 8. LauncMng Stresses. — In the operation of launching a vessel, and 
 thereby transmitting her weight from terra firma to her own element, she 
 experiences considerable bending moment. This, however, if the vessel be 
 well constructed, is not more than she is able to endure without any danger 
 of damage. Unfortunately, it sometimes happens that vessels stick when 
 they have travelled nearly half way down the launching ways. Under such 
 circumstances, serious damage sometimes results ; the great pressure upon tlie 
 bottom at the end of the ways, and the enormous bending moment which 
 is created by the unsupported overhanging part, is sufficient to crush in 
 the bottom and to severely strain the structure, and produce permanent 
 longitudinal deformation. 
 
 Stress and Strain. — In designing the structural arrangement of a 
 vessel, it is necessary that the designer should be thoroughly able to grasp 
 the nature, and to make an estimate of the amount of the severest bending 
 moments likely to occur, and to understand what are the various agencies 
 at work tending to produce deforn\ation in any form whatever. 
 
 Forces acting upon any part of a ship and producing deformation, are 
 properly termed ' stresses ' Avith resulting ' strains,' and straining is, 
 therefore, correctly speaking, the measure of the alteration of form. A 
 structure may be extremely strong, but so long as no external force acts 
 upon it, it remains perfectly rigid and inert. Immediately, however, any 
 external forces are experienced, tending to produce either elongation 
 or bending, or other alterations of form, the material in the structure 
 simultaneously arouses itself, as it were, and offers resistance to any such 
 deformation. These internal forces or resistances, equal to the external 
 forces exerted, and tending to restore the structure to its original shape, 
 are 'stresses.' Any deformation is called 'strain,' its amount being 
 proportional to the original stress. Thus, while these terms, ' stress ' and 
 ' strain,' are commonly, though inaccurately, used interchangeably by 
 ordinary land folk (ship folk use them correctly when they say ' stress ' 
 of weather 'strains' the ship), in the strict mathematical sense thei'e is 
 a distinct difEerence. Strain is the deformation or change of shape measured 
 in terras of the original dimensions, produced by an external tensile, com- 
 pressive, shearing or twisting force, though, perhaps, to an imperceptible 
 extent, while stress is what produces the deformation, or is the effort 
 to offer resistance to change of shape. Such stress is measured by the 
 
76 STEEL SHIPS. 
 
 intensity of the force which produces deformation, whether or not it be 
 so great as to cause permanent set, or rupture, or fracture. An old, but 
 nevertheless a very good illustration of strain and stress is that obtained 
 by stretching a piece of india-rubber. The total physical force which is 
 required to produce a certain elongation is, when reduced to rate, say 
 lbs. per square inch, a tensile stress ; the elongation, when measured as a 
 ratio of the original dimensions, is the strain ; and the internal resistance 
 which is exerted to reduce the elongation and restore the rubber to its 
 original length, is equal to the stress, and may therefore also be called 
 the stress, provided it be measured in the proper way, namely, as a rate, 
 say, lbs. per square inch. 
 
 Every 'stress ' produces a 'strain,' that is, a deformation, however slight, 
 and (in well-designed structures) quite within the bounds of complete safety. 
 When a bending moment is being experienced which is well within the 
 margin of safety, it must follow that at the same time it is being opposed 
 by a stress of an exactly equal amount. We have, therefore, an absolutely 
 essential condition in order to resist any material permanent and therefore 
 unsafe deformation due to bending or elongating or shearing or twisting 
 forces, namely, that under the severest action of such foi-ces, the material 
 
 40 Tons 
 
 Fig. 33. 
 
 in the structure must be able to develop an internal resistance of at least 
 an equal amount, and well inside of the limit of permanent set. In 
 constructing a ship, it is necessary that such a condition exist, and the 
 material must be so disposed as to secure this. 
 
 Before considering in greater detail the strength of actual ships, which 
 are in effect huge hollow girders made up of combinations of numerous 
 smaller girders, perhaps a little investigation into the streugth of simpler 
 sections of material may lead to a clearer conception of the strength of the 
 more complex nature of the ship form. 
 
 Let fig. 33 represent a bar of iron or steel of any length by 12 in. deep 
 and 1 in. thick. Suppose it to be firmly secured at one end A, and at the 
 other, B, to be gripped, and a tension of, say, 40 tons, applied in the 
 direction shown by the arrow. Then the 40 tons would be borne by the 
 12 sq. in. of sectional area, and the tension, or pull, being evenly distributed 
 over the sectional area, would mean a stress of fA = 31 tons per square inch of 
 section. So long as j?rac^*caZZ?/ no extension of the material or permanent 
 set is found to have taken place after the tension is removed, the material 
 has not been stressed beyond its ' elastic limit.' 
 
 As shown in Chapter I., elasticity of a body is that property whereby. 
 
STRESS AND STRENGTH. 'J "J 
 
 after it has been subject to a certain pressure, whether of tension or com- 
 pression, shearing or twisting, at a given temperature, it seeks to regain and 
 retain its original volume and shape at the same temperature. The limit at 
 which the material ceases to resist 'permanent extension or compression is 
 termed the 'limit of elasticity.' 
 
 As a matter of fact, tension, even within the elastic limit, produces 
 permanent extension in most materials, though it is found to be so extremely 
 slight that it may, in practice, be ignored altogether. 
 
 Under tension, the bar in fig. 33 will stretch (in this case well within 
 the elastic limits) in direct proportion to the force applied, or, in other 
 words, in proportion to the stress. Thus, with a tension of 4 tons, the 
 elongation will be twice that of 2 tons, and so on. However, as soon 
 as a certain point has been reached, viz., the 'elastic limit,' increased 
 tension, and therefore increased stress, produces much more rapid 
 elongation, until eventually the stresses so increase that fracture is 
 produced. The stress necessary to produce this fracture determines 
 the ^ultimate strength' of the bar. In determining the scantling of 
 a steel or iron bar, or a girder, or a ship which is subject to re- 
 peatedly applied tension and compression throughout its length, or locally 
 in its structure, it is absolutely essential, if the strength is to be 
 maintained, that the material be preserved from stresses approaching 
 the elastic limit, or, in other words, an ample margin of safety is 
 necessary. The ultimate tensile strength of good mild steel plates, 
 such as are used for the construction of ships built to Lloyd's 
 requirements, is not less than 28 nor more than 32 tons per square 
 inch of section, with or across the grain, and for iron an ultimate tensile 
 strength of at least 20 tons per square inch with, and 18 tons across the 
 grain, is demanded. 
 
 Ductility and the Elastic Limit. — Since brittleness is a most 
 objectionable feature in ship steel or iron, a standard for ductility is 
 imposed. Steel must stand an elongation of at least 16 per cent, on 
 a length of 8 in. before fracture. Of course, such extension is only 
 demanded in testing. It is never supposed that any such extension 
 will take place in the actual ship. Indeed, as is evident, such stress as 
 is experienced in producing these elongations is far beyond the elastic 
 limit, and, as previously pointed out, the material should not be subject to a 
 stress which is likely even to approach that producing permanent elongation. 
 The elastic limit of ship steel, whose ultimate tensile strength has been 
 given, would be from about 17 to 20 tons, and of iron, 10 to 13 
 tons. This at once shows the unmistakable superiority of steel over 
 iron, and it explains why steel ships are so much lighter than iron 
 ones, the reason being, as has been shown, that a square inch sectional 
 area of good mild steel will develop a strength equal to at least 1| 
 square inches of good iron. 
 
 However, it must not be reasoned from this that steel ships are 
 
yS STEEL SHIPS. 
 
 therefore 33 per cent, lighter than iron ones. Tliere are more things 
 than merely tensile strength and elastic limit to be considered in 
 determining the scantlings of plates and bars. 
 
 Thickness gives stiffness, and were steel plates to be unreasonably 
 reduced in thickness, in some cases buckling of a very serious nature 
 would inevitably ensue. Moreover, this reduction is practically only to 
 be found in the plates and angles, and does not materially affect the 
 equipment, ship fittings, stem bar and stern frame, keel bar, etc. 
 
 But supposing, as is quite reasonable, that a saving of 100 tons 
 weight is effected in a 3000 ton ship, owing to the adoption of steel 
 instead of iron for her hull construction, this to the shipowner is an 
 important consideration, for his ship will now carry 100 tons more 
 deadweight than if she had been made of iron, and deadweight means 
 freight. As a result, steel has almost entirely superseded iron for ship 
 construction, excepting in certain localities in the vessel. These localities 
 v;e shall observe at a later stage. 
 
 The working stress of steel or iron should not exceed 50 or 60 per 
 cent, of the elastic limit. Hence, taking the elastic limit of ship steel at 
 18 tons, the maximum working stress should not exceed 9 or 10 tons, with 
 the severest stresses tire material may under exceptional circumstances have 
 to sustain. The ordinary stress will of course be well within this extreme 
 limit. 
 
 Again, let fig. 34 represent a bar of iron or steel 12 in. deep by 1 
 in. thick. It is supported at the points A and B at the ends. The 
 bar is weighted with a load uniformly distributed over its length. The 
 tendency now is for deformation to take place, and the bar to bend as 
 shown. So long as this condition exists, and the stress does not exceed 
 the elastic limit, the bar bends into a curve whose shape and amount 
 of deflection can be ascertained by methods laid down in books on 
 practical mechanics. Assume here that X is the centre of a circle of 
 which Aj B;^ is the arc. It is evident that before the bar could have 
 assumed this shape, considerable extension or stretching must have taken 
 place on the side A;^ Bp and an equal amount of compression on the 
 side PY. It is also clear, if the compression and extension are identical 
 in amount, that midway between these two surfaces of the bar there 
 must be a layer where neither extension nor compression has taken 
 place. This layer is termed the 'neutral surface,' and passes through 
 the centre of gravity of the section of the bar, which in this case is at 
 half its depth. A transverse line passing through the centre of gravity of 
 the bar is called the neutral axis (see fig. 34). When the bar was straight, 
 G H and J K were two transverse sections perpendicular to the sides P Y 
 and A B ; now, however, in the bent bar they have become inclined as in 
 G^ Hj and J^ K^ intersecting the parallel lines E F and C D at the points 
 Ml and Ni in the neutral surface (note M N is equal to M^ N^). So that 
 F D has stretched to H^ Kj and E C has been compressed to G^ J^. 
 
STRESS AND STRENGTH. 
 
 79 
 
 The elongation and compression for any layer in the bar varies directly 
 in proportion to the distance of the layer from the neutral axis. 
 Within the elastic limits we have previously pointed out that the 
 elongation varies directly in jjroportion to the stress. Similarly, 
 the compression varies in direct proportion to the stress. Hence it 
 follows that as the upper and lower layers in the bar are further 
 from the neutral axis than any intermediate one, the stresses must 
 be greater towards the upper and lower edges, and this is exactly the 
 case. The stresses vary in direct proportion to their distance from the 
 neutral axis. 
 
 J 
 
 Neutral 
 
 M 
 
 N Surface 
 
 
 . Neutral Axis 
 
 '<3 
 
 H, F D K, 
 
 Fig. 34. — Illustrating Elongation and Compression in a bent bar ; 
 also Neutral Surface. 
 
 Moment of Inertia. — Let fig. 35 represent an enlarged side view of the 
 bent bar shown in fig. 34 (12 in. deep by 1 in. wide), and let A, B, C, D E 
 F, G be iufinitesimally small units of area in the cross section of tlie bar X 
 Y. The bar is still supported at the ends as before, and the weight dis- 
 tributed over the length, producing the tendency to bend as shown. The 
 resistances ofi'ered by every unit of area in the cross section above the 
 neutral axis to compression, and below the neutral axis to elongation are the 
 stresses. Supposing the distance between the units of area to be equal to 
 one another, and A be on the neutral axis, the stress on A will be nil • at B 
 
8o 
 
 STEEL SHIPS. 
 
 it will be, say, x ; at C, 2x ; at D, 3a; ; at E, 4a; ; at F, f>x ; at G, 6x. 
 Moreover, these resistances are exerted in lines of action parallel to the 
 upper and lower surfaces of the bar at their respective leverages from the 
 neutral axis as shown by the arrowed lines. Thus, while the stresses increase 
 in proportion to the distance of the unit of area from the neutral axis, 
 the moments of these stresses, or moments of resistance, increase in pro- 
 portion to each stress multiplied by its respective leverage from the neutral 
 axis. 
 
 The sum of all these moments of resistance gives a total moment of 
 resistance equal to the total bending moment. 
 
 It will be observed that the depth of the bar in fig. 35 is divided by the 
 arrowed lines into twelve strips, each 1 in, in depth. 
 
 Fig. 35. — The arrowed lines in the upper half of the bar represent the resistance of 
 the material to compression, or compressive stresses. The arrowed lines in,; the 
 lower half represent the resistance to elongation, or tensile stresses. 
 
 The calculation should be : — 
 
 Let R = stress at distant edge of 1st strip. 
 
 Then, stress = an average of - R lbs. per square inch. 
 
 + R \ , . . , „ 
 
 ;e — - — 1 on strip of area A B, 
 
 Then on strip B 0, average stress is — - — = - R 
 
 a 
 
 CD 
 DE 
 E F 
 F G 
 
 2R + 3R _ 5 
 
 2 ~ 2 ^ 
 3R + 4R 7 ^ 
 
 ^2 - 2 ^ 
 
 4R + 5R _ 9 ^ 
 
 ~~2 - 2 ^ 
 
 5R+ 6R ^11,. 
 2 ~ 2 
 
 I. 
 
 n. 
 in. 
 
 17. 
 
 V. 
 
 VI. 
 
 Of 
 
STRESS AND STRENGTH. 
 
 Leverage in each case respectively 
 
 is approximately : — 
 
 
 inches. 
 
 
 I 
 
 i 
 
 
 II 
 
 s 
 2 
 
 D 
 
 Ill 
 
 6 
 
 D 
 
 IV 
 
 i 
 
 
 V 
 
 9 
 
 
 VI 
 
 1_1 
 
 Moments (A x B) are :- 
 
 i X 1 R = 
 
 \ R inch lbs. 
 
 f X 1 R = 
 
 9 T? 
 
 
 1 X f R = 
 
 "^ R ., 
 
 
 1 X 1 R :^ 
 
 " R .. 
 
 
 f X * R = 
 
 -V- R „ 
 
 
 ^ X J^R = 
 
 H^R ,. 
 
 
 Sum 
 
 -4^ R inch lbs. 
 
 il axis double = 
 
 : ^f ^ R = 
 
 143 R inch lbs 
 
 k'^) 
 
 gives exactly 
 
 [Worked out by accurate integrational formula 
 
 144 E.] 
 
 Now, if we multiply every little area patch in the cross sections we are 
 dealing with by the square of its distance from the neutral axis, and add 
 them up, we get the same number as before, viz. (accurately) 144. This 
 number has a special and well-known name. It is called the ' Moment of 
 Inertia of the section,' about the centre of gravity of the section. 
 
 Practical men may justifiably assume that the quantity called 'Moment 
 of Inertia,' which, to begin with, is a purely mathematical quantity, has a 
 physical meaning and a practical use in practical shipbuilding and other 
 constructive work. 
 
 We have seen that the moment of resistance to bending in a bar or 
 girder — whether it be plain, thus X) o*" complicated, as shown by the 
 midship section of a ship, which shows the ship girder in section — equals 
 the sum of the moments of the stresses about the neutral axis of each 
 sectional area of the girder. 
 
 In working through the foregoing illustration, we commenced by 
 assuming a certain stress on a unit of area at a unit of distance from the 
 neutral axis. But in actual practice, when dealing with a loaded bar, as 
 in fig. 34, it is the maximum stress [i.e. at the outermost layers) that has 
 to be ascertained in order to see whether the material is being strained at 
 these layers beyond the elastic limits, when, of course, permauent elongation, 
 or ' set,' or perhaps even rupture would occur. It is therefore evident that 
 the calculation will be carried out in office work by dealing with the value 
 of R at the outermost layers. 
 
 However, the stress at ' unit ' distance from the neutral axis enables us 
 to arrive at a useful quantity by which the desired result can be obtained. 
 
 This quantity, the moment of inertia of the whole cross section of the 
 
 ^n 6 
 
82 STEEL SHIPS. 
 
 bar, and represented by 144 in this example, equals the sum of the 
 products of every unit of area in the cross section multiplied by the square 
 of the distance of each unit from the neutral axis. 
 
 As we have seen, the moment of resistance for the whole cross section 
 is the sum of the moments of all the stresses in the cross section of the bar. 
 So that the moment of resistance at the section in question equals the 
 moment of inertia multiplied by the stress at a unit of distance from the 
 neutral axis. 
 
 Hence the total moment of resistance divided by the moment of inertia 
 equals the stress per unit of area at a unit of distance from the neutral axis. 
 But so long as the material in the bar is not subjected to stresses beyond 
 the elastic limit, the moment of resistance is equal to the bending moment. 
 
 Therefore the bending moment divided by the moment of inertia equals 
 the stress per unit of area at a unit of distance from the neutral axis. 
 
 Let I = Moment of inertia of a cross section, 
 
 and M = Bending moment of same cross section ; 
 
 M 
 
 then -^ — Stress per unit of area at unit of distance from neutral axis, 
 
 in the same cross section. 
 
 But, as we have seen, the stresses increase from the neutral axis towards 
 the upper and lower surfaces in proportion to the distance. 
 
 Let P = distance of upper surface from neutral axis, 
 
 „ K = „ lower „ „ 
 
 M X P 
 Then — ^ — = maximum stress per unit of area upon upper suface of bar ; 
 
 , MxK 
 and — I — = „ „ „ lower „ 
 
 In practice, the %mit of area is the square inch, and the stress obtained 
 is therefore stress per square inch. 
 
 Of course in the bar of the section shown in figs. 34 and 35, the neutral 
 axis passes through the centre of gravity of the bar, which is at half the depth, 
 because it was assumed that the bar would resist extension and compression 
 
 equally, and therefore P = K. However, in a bar of I section, or indeed any 
 
 other section, P and K would not necessarily be equal. 
 
 The value and reasims for making girders and bars of the following 
 sections will now be obvious. 
 
 r T T I T in 
 
 Fig. 36. 
 
 As the stresses increase in magnitude as the distance increases from 
 the neutral axis, by adding flanges, or bulbs or plates, to the upper and 
 
STRESS AND STRENGTH. 
 
 83 
 
 lower extremities, the moments of inertia of the sections are greatly 
 augmented, and the stresses are more readily met.* 
 
 Before proceeding to show how the bending moment is ascertained, it 
 is hoped, by means of another simple graphic illustration, to make this 
 subject of moment of inertia still more comprehensive, and to dispel a 
 mistaken idea which has grown among many young students that it is 
 almost too difficult and mysterious and intricate to understand. 
 
 Here is a section of a bar of iron 10 in. deep, and 2 in. thick, x y = the 
 neutral axis. The section of the bar is divided into units of area of 
 
 J- 'o ^- 
 
 
 ^T y i 
 
 '■y 
 
 Fig, 37. — Illustratiug Moment of Inertia. 
 
 1 square inch, the unit of area adopted in practice. The total amount of 
 inertia of the bar about the neutral axis is found by multiplying each unit 
 of area by the square of its distance from the neutral axis. 
 
 The moment of inertia of the upper half of the section of the bar = 
 
 2 X -o- = 0-5 
 2 X 1-52= 4-5 
 2 X 2-5"-^ = 12-5 
 2 X 3-52 = 24-5 
 2 X 4-52 = 40-5 
 
 82-5 
 
 And for the lower half of the bar section it will be 82'5 also, making a 
 total of 165. 
 
 Supposing the bending moment to be 100 inch- tons, the stress on 1 square 
 
 * Tensile and compressive stres83S are always greatest in that part of the material 
 which is farthest from the neutral axis. 
 
84 . STEEL SHIPS. 
 
 incli at 1 in. from the neutral axis would be \^i, and at the layer on the upper 
 
 100 X 5 
 and lower surfaces of the bar (G) would be — t-oK — = Tef = about 3 tons 
 
 per square inch, this being the maximum stress. 
 
 The method just adopted to calculate the moment of inertia of the 
 
 section of this bar is exactly similar to that adopted in ship calculations, 
 
 but it must not be supposed that it is absolutely correct, though it is 
 
 sufficiently accurate for all practical purposes. To begin with, our units 
 
 of area are not infinitesimally small (we have taken 1 square inch). Hence 
 
 there is a slight error in the moment of inertia. The correct moment of 
 
 inertia of this bar (true for rectangular sections only) would have been 
 
 thickness of bar x depth^ 2 x 10^ ■,/./>/. Tro ^ ^ /• 
 
 ; T-r: — - — = — p^; — = loD'o, a difterence ot l*b. 
 
 constant 12 12 
 
 We can get as near this as we like by making the strips narrower 
 
 and narrower. 
 
 Reverting now to diagram 35 (bar supported at ends) the actual moment 
 
 1 X 12^ . 
 
 of inertia of its section (12 in. x 1 in.) will be — — — =144, as previously 
 
 shown. 
 
 We also know that the centre of gravity of such a section would be at 
 half the depth, and that through the centre of gravity passes the neutral 
 axis of the bar. The stress, therefore, upon either of the extreme surfaces 
 on the top or bottom of this rectangular bar will be : — 
 
 Bending moment x 6 ^ g^^^^^ per square inch at most severely strained 
 144 
 part (outermost edge). 
 
 Bending Moment of Girder. — Having, it is hoped, made reasonably 
 clear the value of the moment of inertia of a girder section, and explained 
 the principles involved in arriving at a computation of the same, our next 
 step will be to show how the actual bending moment is arrived at, though 
 in a simple form this has already been treated (see page 67). 
 
 Curve of Loads. — Let A B, fig. 38, be a bar somewhat similar to that 
 dealt with in fig. 28, and loaded with a continuous load as indicated by the 
 hatched lines. Let the bar be, say, 20 ft. long. 
 
 At 2 ft. mtervals on A B, set up ordinates, calculate the weight of the 
 bar per foot with its load, and set up the results to scale, each upon its 
 respective ordinate. Through the points so obtained draw the curve A C B. 
 The total area of this curve represents to scale the total weight resting upon 
 the supports at A and B, and the centre of the area (G) enclosed by A B 
 and the curve A C B, represents the fore and aft position on the line A B of 
 the centre of gravity of the bar and its loading. The curve ACB is 
 termed a curve of loads. 
 
 Now it is evident that the upward pressures of the supports upon the 
 ends of the beam must be equal to the downward pressure of the total 
 weif^ht of the bar and its load ; but as the centre of gravity of the load is 
 
STRESS AND STRENGTH. 
 
 85 
 
 not, ill this example, in the middle of the length of the bar, the pressure of 
 the supports at the ends will be unequal. 
 
 To find the suj>porting pressures at the ends of the bar. — Let L = total 
 load, and P and Y the supporting pressures. 
 ThenL = P + Y. 
 
 (PxAB) + (YxAB) = Lx AB 
 PxAB 
 
 and 
 
 L 
 YxAB 
 
 = k 
 
 = d 
 
 Thus, so far, we have obtained the total load, and the pressure at each 
 end of A B supporting the same load. 
 
 Curve of Shearing Stresses. — By a process of graphic integration, we 
 now proceed to ascertain the bending moment. This is first done by 
 constructing a curve of shearing stresses. 
 
 Fig. 38. — Curve of Weiglit A C B for a loaded bar. 
 
 Let AB, fig. 39, represent the bar, with the curve of loads A C B upon 
 it, with ordinates at 2 ft. intervals. Make A M (set oif above A B) equal to 
 the pressure P, and B^N" equal to the pressure Y (set oS below AB) The 
 shearing stress at any point in the length A B is the difference between the 
 supporting pressure on that side of the centre of gravity, and the weight 
 represented by the area of the curve of loads from the end of the bar up to 
 that point. 
 
 Thus the shearing stress at 
 
 cZ = P - (weight represented by area A A' fZ.) 
 and at e it equals P - ( ,, ,, ,, „ AA^'g.) 
 
 „ / „ P - ( „ „ „ „ A/"/) 
 
 These differences are set off from A B on their respective vertical lines to 
 the same scale as adopted for A M and B IS". Naturally, as we travel from 
 A towards B, the difference between the pressure P and the weight repre- 
 
86 
 
 STEEL SHIPS. 
 
 sented by the successive areas AX d, AX' e, etc., grows less and less, until, 
 at a certain point 0, the pressure and weight are exactly equal, and if we 
 proceed beyond this point towards B, the weight preponderates, and the 
 difference, which is now negative, is set off below AB. Similar results 
 would have been obtained had we commenced from B and deducted from 
 B N" the successive weights, represented by the areas of the curve from B 
 (except that in the drafting, the curve would have been above at the B end 
 and below at the A end). 
 
 Let M O IS" be the curve of shearing stresses so obtained. The effect of 
 these opposite, or negative and positive, shearing stresses, is to produce the 
 tendency to rack or distort the beam by bending, and the point is known 
 as the point of reverse racking, or the point at which the racking action is 
 reversed. In addition, the curve of shearing stresses shows the tendency to 
 shear (i.e. cut through) at any cross section. 
 
 A d e f g h 
 
 Fig. 39. — Curve of Shearing Stresses M N, Curve of Bending Moments A R B 
 (AC B Curve of Weights). 
 
 From this curve of shearing stresses we are able to ascertain the bending 
 moment at any point. 
 
 As previously shown, the shearing stress at any point in the length of a 
 beam or bar is simply the difference between the pressure (reaction) at one 
 end, and the weight upon the bar up to that point, from the same end ; or, 
 in other words, it is the resultant of the forces of the downward weight 
 upon the bar and the upward pressures at the ends reckoned at any cross 
 section in the length of the bar. 
 
 Curve of Bending Moments. — In addition to the tendency to shear 
 between the points of support, there is a tendency to bend also. Now the 
 bending moment for any 07ie load at any section in the length of a bar 
 supported at the ends is equal to the pressure (reaction) at either end, 
 
STRESS AND STRENGTH. 
 
 87 
 
 multiplied by the distance of the section from that end of the bar ; and, in 
 general, for any loading, it is the algebraic sum of all the shearing stresses 
 on the bar between the cross section and the end of the bar. 
 
 The area, therefore, of A M represents the bending moment at 0, and, 
 similarly, the area B N represents the bending moment at 0, since A M O 
 = B N. Thus at the point O where the racking reverses, the maximum 
 bending moment is experienced. In order to graphically represent the 
 bending moment at any section in the length A B, the area of the curve of 
 shearing stresses would have to be calculated at intervals from either end of 
 the bar, and set up to scale at their respective positions throughout the 
 length of A B. For example, the area of A M Z of would be set up at d, 
 and A M Z' e at e, and so on up to 0, where the bending moment attains its 
 maximum. The shearing stresses now become negative, and are subtracted 
 from A M instead of added. Through the points so obtained the curve 
 A R B is drawn. By means of this curve, the bending moment may be 
 found at any cross section in the length of the bar by measuring the 
 perpendicular distance from A B to the curve. 
 
 ]1 
 
 t^ 
 
 -4— 
 
 
 Neufral 
 
 ^ 
 
 Axis 
 
 if-'^. jf _.L_i. 
 
 lA 
 
 ? 
 
 ff - - 5 . 
 
 Fig. 40. — Compound Girder. 
 
 -4 
 
 Fig. 41. — Equivalent Girder. 
 
 Stress per Square Inch. — It has now been shown how each part in the 
 equation for stress in the bar example is obtained. Thus (for units using 
 inches and either lbs. or tons as most convenient) : — 
 
 Bending moment Stress per square inch of section at 1 in. from the 
 Moment of Inertia ~ neutral axis. 
 
 ! distance of upper or lower -i 
 surface of bar from the V 
 neu ra ^.xis. =MaximutD stress per square inch. 
 
 Bending moment x 
 
 Moment of Inertia. 
 
 The Equivalent Girder. — So far we have only been dealing with the 
 very simplest foi'm of bar, viz., of rectangular section. But supposing it to 
 have been of a more complicated form of section similar to that shown in 
 fig. 40, a little more trouble would have been entailed in obtaining the 
 moment of inertia. 
 
88 STEEL SHIPS. 
 
 By closing up the material in any compound girder section into one 
 compact girder, we have a consolidated girder of equivalent sectional area 
 and with similar vertical distribution of such area. This is usually called an 
 equivalent girder. (See fig. 41 which shows fig. 40 contracted horizontally). 
 
 An equivalent girder is rarely constructed before making the calculation 
 for the moment of inertia of a section of a ship. But while the diagram 
 affords a graphic way of presenting the material in the sectional area, it 
 serves little other purpose, and, as a result, the equivalent girder is seldom 
 constructed in mercantile shipyard practice, though, in making the 
 calculation, precisely the same process of combining the material is 
 carried out. Fig. 42 illustrates a section of a vessel with its correspond- 
 ing equivalent girder. 
 
 Computation of Moment of Inertia of Compound Girder. — Perhaps 
 
 i 
 
 . 
 
 Neutral Axis 
 
 JU 
 
 Fig. 42. — Midship Section of a Vessel showing the structural material embraced in the 
 Equivalent Girder, and dealt with in the Calculation for Moment of Inertia (see 
 fig. 12.) 
 
 it may aflford a useful example to compute the moment of inertia of the 
 section of such a girder as shown in fig. 40, relatively to its neutral axis, 
 as the calculation will be identical in every way with that for the moment 
 of inertia of the section in the more complicated ship-girder. The distance 
 of the neutral axis (which passes through the position of the centre of 
 gravity of the area of the section) from the base line A B is found by 
 multiplying the area of each part composing the section by its distance 
 from A B and dividing the sum of the moments by the sum of the areas. 
 Or it could be found by assuming a neutral axis and dividing the difi'erence 
 of the moments above and below the assumed neutral axis by the total 
 area of the section. The result is the amount of correction necessary for the 
 
STRESS AND STRENGTH. 89 
 
 actual position of the neutral axis, and the side upon which the moments 
 preponderate indicates the direction in which the correction has to be made. 
 Keference to fig. 37 will enable us to more easily follow the steps in the 
 process of calculation. In that example it was shown that the moment of 
 inertia about the neutral axis of any unit of area was obtained by multiplying 
 the area by the square of its distance from the neutral axis, — one square 
 inch being taken as the unit of area, — and that the sum of the products of 
 all the units of area in the section multiplied by the squares of their 
 respective distances gave total moment of inertia. 
 
 Exactly the same process is followed in dealing with the girder section, 
 fig. 40. In all sections of material disposed horizontally {i.e., parallel to 
 neutral axis) see Xos. 1, 2, and 4 flanges of girder, and of small depth or 
 thickness, the areas in inches, which represent sums of units of areas, 
 multiplied by the squares of their distances from the neutral axis, will 
 give their respective moments of inertia about the neutral axis. But on 
 coming to vertical webs (JS'o. 3) or areas disposed vertically, while the 
 area is again multiplied by the square of the distance of its centre from the 
 neutral axis, this only treats of the units of area as though they were all 
 ranged exactly parallel to the neutral axis of the girder, which is not the case. 
 We have already considered the value of the moment of inertia of a vertical 
 web about its own neutral axis, which, in a case like this under considera- 
 tion, would produce a total moment of inertia about the girder neutral axis 
 in excess of what would have been obtained had the same material been 
 disposed horizontally.* 
 
 Proceeding with the calculation : — 
 
 1st. To find the neutral axis of the whole girder. It is obvious that the 
 
 * In the accompanying illustration W and Z are similar in size and of equal area, 
 the difference being that one is placed horizontally and the other vertically in relation 
 to the line a h, while the centre of gravity of each is at the same distance from a b. The 
 moment of inertia of Z about the line a b is equal 
 to the moment of inertia of W, plus its own 
 moment of inertia about its own neutral axis, 
 which latter in actual calculations may be expressed 
 
 as i\th of depth cubed x breadth ; or, more 1 I - 
 
 usually, as y^tli of area x depth squared, which is 
 exactly the same. 
 
 It is theoretical!}' true that horizontal flanges 
 have a moment of inertia about their own neutral 
 axis also, but in practice it is so insignificant as 
 
 to be entirely ignored. Decks, and inner and ^ 
 
 outer bottoms, are similar horizontal flanges in 
 actual ship girders. 
 
 Assuming the dimensions of W and Z to be 10" x 1" and the distance x to be 20" 
 
 Then the moment of inertia of 
 
 /lO X 1^\ 
 W about a ft = (10 X 1 X 20'-) + ( —^ J = 4000 '8, 
 
 /]0xl0-\ ,„o<> o 
 for Z about « & = (10 x 1 x 20^) + ( ^^ j = 4083-3. 
 
 and 
 
90 
 
 STEEL SHIPS. 
 
 centre of gravity of each of the X paits of the girder is at half the depth, 
 that is, 4| in. above AB. 
 
 The centre of gravity of the top plate (which is | in. thick) will 
 
 therefore be at 9^ in. above A B. 
 
 The common centre of gravity, through which the neutral axis passes, 
 will be obtained as follows : — 
 
 The two X parts of the girder contain together 18 square inches in 
 sectional area, and the top plate 6 square inches. 
 
 2. X Parts of Girders, 
 Top Plate, . 
 
 Sectional Area. 
 
 Distance of Centre 
 
 of Areas of Girder 
 
 above A B. 
 
 Leverage. 
 
 sq. ins. 
 
 18 
 
 6 
 
 ms. 
 4-50 
 9-25 
 
 Moment. 
 
 81-0 
 55-5 
 
 24 
 
 136-5 
 
 24 
 
 136-5 
 
 5 '69 inches above A B. 
 
 The neutral axis of the whole girder is 5*69 inches above the line AB. 
 
 2nd. To find the moment of inertia of the section of the girder about 
 the neutral axis. 
 
 The following is the usual form of the calculation : — 
 
 
 
 Distance 
 
 
 
 
 
 Particular 
 Items in 
 
 Sectional 
 Area 
 
 of Centre of 
 Gravity from 
 
 Products. 
 A X A^ 
 
 Depths of 
 Vertical 
 
 d^ 
 
 xV X A X c?2 
 
 Girder. 
 
 = A 
 
 Neutral 
 
 Webs = d 
 
 
 
 
 
 Axis = h 
 
 
 
 
 
 
 sq. inches. 
 
 inches. 
 
 
 inches. 
 
 
 
 Item 1 
 
 6 
 
 3-56 
 
 76-02 
 
 •5 
 
 •25 
 
 
 2 
 
 5 
 
 3-06 
 
 46-80 
 
 •5 
 
 •25 
 
 
 „ 3 
 
 8 
 
 1-19 
 
 11-28 
 
 8 
 
 64-00 
 
 42-66 
 
 ,, 4 
 
 5 
 
 5-44 
 
 147-95 
 
 -5 
 
 •25 
 
 
 282-05 
 42-66 
 
 42-66 
 
 324 ^71 Moment of Inertia. 
 
 Stress per square inch upon upper and lower surfaces of foregoing 
 girder. 
 
 The total depth of the girder =9^5 in. 
 
 From neutral axis to top of girder = 3^81 ,, 
 „ „ „ bottom „ = 5-69 „ 
 
 Suppose the girder to be subject to a maximum bending moment of 
 14 foot-tons = 168 inch-tons. 
 
 The tensile stress per square inch upon the upper surface of the girder 
 will be 
 
 168 X 3-81 
 324-71 =l'^^t»"^' 
 
STRESS AND STRENGTH. 
 
 9T 
 
 and the compressive stress per square inch upon the bottom surface of the 
 girder will be 
 
 168 X 5 69 
 
 324-71 
 
 = 2-94 tons. 
 
 Having briefly shown how to obtain the moment of inerti:i, and the 
 bending moment of a loaded bar or girder of plain section, it is found, 
 on turning to an actual ship, that though the labour of calculation is 
 considerably increased, the principles worked upon, and the steps in the 
 calculation, are practically identical in arriving at the results. 
 
 Bending Moment of an actual Ship. 
 Curve of Weights. — The first operation is to determine the distribution 
 of the weights in the vessel. This curve may show great variation, 
 
 Fig. 43. 
 
 ACB = Ci 
 
 irve 
 
 of Weights. 
 
 EDB 
 
 !> 
 
 Buoyancy. 
 
 AGJKB = 
 
 J) 
 
 Loads. 
 
 A N L B = 
 
 J) 
 
 Shearing Forces. 
 
 A MB 
 
 
 Bending Moments 
 
 especially in a cargo-carrying vessel, according as to whether the vessel 
 is ballasted, light, or in a loaded condition. For the sake of example, 
 let us suppose that fig. 43 represents the profile of the vessel for which the 
 
92 STEEL SHIPS. 
 
 calculation of beading moment has to be made. As in the case of the bar, 
 we lay down a base line A B equal to the length of the ship. Assuming 
 that we have a lines plan which enables us to obtain the girth and section 
 of the vessel at any point in the length of the hull, and assuming also that 
 we have a constructive plan showing in profile and section the scantlings, 
 arrangement, and disposition of all the material and distribution of all 
 permanent weights, we are fully equipped with the information necessary, 
 in order to proceed with the determination of the curve of weights for the 
 light condition. 
 
 Divide the length of the vessel into a number of equal intervals, say 
 15, and, at each division, calculate the weight of the material in the vessel 
 for one foot of the length. This will mean 1 ft. of all material which runs 
 continuously fore and aft ; but for transverse framing, beams and pillars, 
 which are jDrobably spaced at intervals of more than 1 ft., the proportion 
 of the weight which belongs to 1 ft. of length will be taken. Exceptional 
 weights, such as the stern frame, stem bar, etc., should be added later, only 
 the longitudinal and transverse framework and plating being calculated at 
 this stage. These ' weights per foot ' are then set ofi" to scale from the base 
 line A B on their respective ordinates. 
 
 Assuming the vessel to be in a light condition, propelled simply by 
 sail, a few additions will still have to be made to the curve of weights. 
 Masts and rigging, winches, windlass, anchors and cables, and whatever 
 stores and other equipment and spare gear there may be on board, should 
 all be included in the vessel's weight, and appear in their proper localities 
 in the curve. Had the vessel possessed a poop, bridge, or forecastle, these 
 would also have added to the local weight. If she had been a steamer, 
 there would have been the further additional local weight of engines, 
 boilers, bunkers, shafting, propeller, funnel, etc. 
 
 Suppose, again, the vessel to be in a ballast condition, with water 
 ballast in a double bottom, extending fore and aft along the bottom, with, 
 in addition, as is common, peak tanks at the ends, and perhaps a large 
 deep tank somewhere in the region of amidships. The weight per foot of 
 length of all such ballast should be calculated, and addeil to the curve 
 of weights. 
 
 In the loaded condition, however, a vast increase takes place in the area 
 of the curve of weights. Fully loaded with a miscellaneous cargo, which 
 probably changes on each succeeding voyage, it is obvious that no accurate 
 curve of weights could be obtained for the loaded condition on any 
 particular voyage. 
 
 The practice is therefore to assume that the vessel's holds are filled 
 with a homogeneous cargo which exactly puts her down to the load line. 
 Knowing the total volume of the holds, and the deadweight of the 
 ^^gg^j cubicjee^holdcapacity ^^^^.^ ^^^^ ^^^ ^^^^ stowa^ge. Therefore, 
 
 deadweight 
 the number of cubic feet of hold space per foot of length, at each 
 
STRESS AND STRENGTH. 93 
 
 of the respective stations in the length, divided by the number of cubic 
 feet per ton of stowage, gives the tons addition to be made to the curve of 
 weights in the light condition. 
 
 In short, the curve of weights accurately represents the weight per foot 
 at any point in the length, and its total area represents the total weight of 
 the vessel for a specified condition. This will generally be for the fully 
 loaded condition, when the greatest bending moments may be experienced in 
 wave water, or even for a ballast condition when the concentration of large 
 volumes of water in certain localities in a vessel's length may also produce 
 excessive bending moments at sea. With the increasing adoption of large 
 deep tanks, and even hold spaces for water ballast, it seems that ship- 
 builders will be required to give very special attention to the proper 
 strengthening of ships, so as to effectively resist strain, which the con- 
 centration of great weight is liable to produce. 
 
 Assuming that the process of calculating the weights of a sailing 
 vessel for the loaded condition has been completed, let ACB (fig. 43) 
 represent the curve of weights. 
 
 In reality, the method described for calculating the weights of a 
 vessel throughout her length, would not produce a curve at all, but rather 
 an uneven, and, in some places, a very jagged line bounding the area repre- 
 senting the total weight. Sometimes the irregular boundary of the area 
 is dealt with just as it is produced by the computation of the actual 
 Aveight; but it is more commonly the practice to run a mean curve 
 embracing, as far as possible, all the features of the irregularly bounded 
 area first obtained. This new curve, if correctly draAvn, should accurately 
 represent the total weight of the vessel. The calculation just described, 
 as is evident, entails a considerable amount of labour, though of a very 
 simple character. 
 
 Curve of Buoyancy. — Having ascertained the amount and distribution 
 of the weight as shown by the curve of weights, the next step is to deter- 
 mine the distribution of the support given by the water in the form of a 
 curve of buoyancy. 
 
 Whatever be the condition of loading, we are aware that the total 
 support given by the water (buoyancy) is exactly equal to the total weight 
 of the ship and all in it. This condition is fulfilled whenever a vessel floats 
 in equilibrium. It is needless to say that (in ascertaining any particular 
 result) the calculation is based upon a certain condition of loading, which 
 means a fixed draught, the mean of which, in either smooth or wave water, 
 remains practically constant. 
 
 The volume of displacement per foot of length, at each of the stations 
 upon AB, fig. 43, is calculated; this volume divided by 35 (salt water), 
 gives tons of buoyancy per foot. Set these off upon their respective 
 ordinates to the same scale as used for the curve of weights. A curve 
 through the points so obtained will give the curve of buoyancy. By means 
 of this carve, the buoyancy per foot can be ascertained at any point in the 
 
94 STEEL SHIPS. 
 
 length of the vessel, and its whole area represents the total buoyancy of the 
 ship. 
 
 For either a light, hallasted, or loaded condition, the curve of buoyancy 
 would be worked out in identically the same manner, — the volume in each 
 case being taken up to the water-line, at whatever trim for the particular 
 condition. Let the curve EDB, fig. 43, represent the curve of buoyancy 
 for the fully loaded condition. If the calculation has been correctly worked, 
 the total area of the ' curve of weights ' should exactly equal the total area 
 of the ' curve of buoyancy ' for the same condition of loading. 
 
 Curve of Loads. — Curves A C B and EDB show exactly where, 
 and to what extent, the weight preponderates locally over the buoyancj^, 
 and the buoyancy over the weight. By carefully measuring the difference 
 upon the ordinates between the two curves, and setting off the excess of 
 buoyancy above the line A B and the excess of weight below A B, we have 
 what is called the ciirve of loads. The points F and H indicate where the 
 buoyancy and weight are exactly equal to each other and are known as 
 water-borne sections, and in the next stages of the calculation, these being 
 the only points in the length where the buoyancy exactly sustains the 
 weight, they are considered as the points of support. 
 
 As a necessary condition for any vessel to float in equilibrium is that 
 the centre of buoyancy and the centre of gravity be in the same vertical 
 line, it follows that the centre of gravity of the area enclosed by the curve 
 of weights A C B and the curve of buoyancy EDB, and also of the sum 
 of the three parts of the area enclosed by the curve of loads A G F, F J H, 
 H, K, B, will be in the same vertical line. 
 
 Curve of Shearing Forces. — Following the same graphic method 
 adopted in dealing with the plain bar, we proceed to obtain the curve of 
 shearing forces. 
 
 Calculate the area of the 'curve of loads,' commencing from A to each of 
 the succeeding ordinates. Set these areas off above A B. The ordinates in 
 the curve will continue to grow up to the point F. But as the curve of 
 loads at this point crosses A B, the areas of the curve of loads above A B will 
 now be deducted from the foregoing area, until, at X, the shearing curve 
 crosses A B, when the ordinates now begin to extend below A B up to the 
 point H in the curve of loads, when deduction of the succeeding areas of 
 the curve of loads produces the curve of shearing forces as shown. The 
 shearing force at any point in the length of a vessel may be expressed 
 as the algebraic sum of all the stresses caused by the excess of weight and 
 buoyancy from either end. 
 
 Curve of Bending Moments. — Finally, the bending moment at any 
 section in the length of the vessel is represented by the area of the curve 
 of shearing forces from A or B. From A, the ordinates for the curve of 
 bending moments will continue to grow in height up to the point x in the 
 curve of shearing forces, where it attains its maximum. Beyond x from A, 
 the areas of the curve of shearing forces should be deducted from the 
 
STRESS AND STRENGTH. 95 
 
 area obtained up to x. By calculating the bending moment at intervals 
 along A B, a curve showing the whole range of the bending moments is 
 obtained, A M B, fig. 43. The length of any ordinate from A B to the 
 curve AMB gives the bending moment at that point in the length of 
 the vessel for the fully loaded condition, and floating in smooth icater at 
 F and H, where the curve of shearing forces attains its greatest height, are 
 indicated the points of maximum shearing force. The bending moment 
 at any point in the length of a vessel may be expressed as the algebraic sum 
 of all the shearing stresses from either end. 
 
 So far, we have simply considered a sailing vessel fully loaded and 
 floating in still water at her maximum draught. But the curve of bending 
 moments, for such a condition, in no w^ay represents the extremes of the 
 bending moments experienced by the ship ; for, while in wave water the 
 distribution and the amount of the weight itself remain constant, great 
 alterations take place in the distribution of the buoyancy, causing great 
 augmentation in the excess of weight over buoyancy and buoyancy over 
 weight in certain localities in the length, as the waves change in form and 
 position, thereby producing vastly increased bending moments compared with 
 what is experienced when floating in smooth water. But not only have 
 these greatly augmented bending moments to be provided for in the 
 structure of the vessel, but, owing to the rapid passage of the waves, the 
 suddenness with which an excess of weight over buoyancy is succeeded 
 by an excess of buoyancy over weight, and vice versa, conduces to the 
 severity of the stress experienced by the hull, and must receive due 
 attention from the naval architect. In making the calculations necessary 
 in order to arrive at a fair approximation (absolute accuracy being impossible) 
 to the bending moment experienced in wave water, two conditions which 
 probably produce the greatest extreme bending moment to which a ship 
 is likely to be subject, are usually taken. 
 
 (1) When supported upon a wave at the middle of the length with 
 consequently greatly reduced buoyancy at the ends, the tendency is for 
 the vessel to ' hog,' as illustrated in fig. 29, p. 69. (2) When the ends 
 are supported upon wave crests, with consequently greatly reduced buoyancy 
 at the middle of the length. In this case, the tendency is for the vessel 
 to * sag,' as illustrated by fig. 30, p. 70. 
 
 In making the calculation in each of these cases, the designer constructs 
 upon the profile of the vessel, as accurately as possible, a geometrical wave 
 form, the height of which is usually taken as -^^th. of the length measured 
 from crest to crest. Having satisfied himself as to the form of the wave, he 
 must now carefully ascertain that the displacement is exactly equal to that 
 for the still water condition. This must be rigidly observed, for whatever 
 be the contour of the wave water, the volume inmiersed must remain 
 unchanged as the weight of the ship remains unaltered. 
 
 The steps in the calculation are now similar to those followed for the 
 bending moment in still water. 
 
gb STEEL SHIPS. 
 
 Moment of Inertia op Actual Ships. 
 
 Comparison of Vessels. — As stated at an earlier stage, the standard 
 of strength required in any vessel in order to obtain the freeboard 
 stipulated in the Board of Trade Freeboard Tables, must be at least 
 equal to the structural strength of ships as defined and specified in 
 Lloyd's Eules of 1885. 
 
 Fig. 42 is a midship section of a vessel which fully complies with 
 such requirements. 
 
 Now while it is demanded that this vessel be at least equal in strength 
 to the standard, it does not follow that the scantlings will be identical 
 in every detail. The true comparison of vessels of similar dimensions 
 and form should be made upon the moment of inertia of the weakest 
 section in the neighbourhood of amidships, assuming that the material 
 has been intelligently distributed. 
 
 Disposition of Material. — Any addition of material to the section of 
 a girder will produce an increase in its moment of inertia. But we have 
 already seen that such increase is most rapidly obtained by making the 
 addition to the upper and lower extremities. 
 
 On turning to the ship girder, the same principle is equally true. 
 For example, in designing the structural arrangement of a vessel, the 
 thickness of the shell plating might be kept uniform from keel to 
 gunwale, and the necessary moment of inertia obtained. But it is 
 plain that an undue amount of material may have been introduced into 
 the structure, for probably the same moment of inertia could have been 
 obtained by reducing the side plating from bilge to sheer strake, and 
 perhaps in some degree from keel to bilge, by making a substantial 
 addition to the thickness of the sheer strake, or by doubling it in 
 certain cases, and perhaps increasing the thickness of the strake below 
 the sheer strake, say ^V'. It might also be advisable, in large long 
 vessels, to increase the thickness of bilge plating in one or two strakes. 
 What we have proposed is simply to rearrange the material in the 
 girder section, by adding to the sectional area of the upper and lower 
 extremities, which is the most effective means of increasing the moment 
 of inertia, and permits of a reduction in the thickness of the remaining 
 plating which is nearer to the neutral axis. By this means, an ultimate 
 saving in weight is obtained without any loss in strength or general 
 efticiency. This is a most important consideration to the shipowner, 
 for reduced weight of hull means to him increased deadweight capacity 
 and freight. 
 
 In vessels of special design, with particularly light shell plating, 
 caution would be necessary in carrying out any such plan as proposed 
 above, for the absurdity of unduly reducing the thickness of the plating 
 between the sheer strake and bilge in order to effect a saving in 
 weight, and at the same time to promote the possibility of the plating 
 
STRESS AND STRENGTH. 
 
 97 
 
 L 
 
 Shell Plating 
 
 collapsing in the region of the neutral axis when subject to racking 
 stress, is obvious. 
 
 A fairly good motto to follow in ship designing is * Minimum weight 
 to obtain maximum efficiency and strength.' So long as the standard of 
 strength is secured, both the classification societies and the Board of 
 Trade are satisfied, no matter how material may have been wasted, and 
 unnecessary weight introduced by an incapable designer. It must not 
 be overlooked, however, that exceptional strength is required in certain 
 localities to provide for local requirements. 
 
 Effect of modification in depth of Transverse Frames. — The vessel 
 in fig. 42 (see also fig. 12) is built on the deep frame system by which 
 hold beams are dispensed with. The transverse framing is made up, 
 over three-fifths of the middle of the length, of two angles each 
 5x3x/jy. (See fig. 44.) 
 
 Now suppose the designer or shipbuilder finds that for some reason 
 or other it is necessary to adopt 
 angles of different scantling, which 
 may affect the thickness, or perhaps 
 alter the depth of the frame girder. 
 Though the common system is to 
 secure the same sectional area of 
 material as in the original section, 
 it must be evident that with such 
 variations in form as just described, 
 a difference in the capacity of the 
 frame to resist bending may have 
 arisen, owing to reduced moment of 
 inertia. Strictly, then, the com- 
 parison should not be made upon 
 
 the sectional area of the material, but upon the moment of inertia of the 
 section. 
 
 Further, the more correct method is to calculate the moment of inertia of 
 the frame girder about an axis passing through the heel of the bar upon the 
 shell plating, for the frame being assisted in its stiffness to resist bending 
 by the shell plating for half a frame space upon each side of the frame, 
 the neutral axis of the shell plating, together with the frame girder, comes 
 nearer to the heel of the frame bar than the true neutral axis of simply the 
 two angles forming the frame girder. 
 
 Value of Stress Calculation.— It is impossible to make any strength 
 calculation which will give as a result the actual maximum tensional, 
 torsional, or compressive stresses endured by any vessel even where accurate 
 information is furnished regarding the distribution of the cargo or the 
 weight. First of all, the severest bending moments are only experienced 
 in wave water, and while it is possible to arrive very closely at an accurate 
 curve of weights, the process of defining the buoyancy is purely tentative. 
 
 7 
 
 Fig. 44.— Section of Deep Framing. 
 
98 STEEL SHIPS. 
 
 A curve showing the total buoyancy is easily constructed, but a curve 
 showing the distribution of buoyancy as actually experienced in a seaway, 
 is not so easily obtained, and at best is only an approximation. However, 
 the distribution of buoyancy usually chosen in making the calculation {i.e. 
 vessel supported at amidships upon a wave of her own length, and the wave 
 height -^th of the length, crest to crest ; or upon a wave of her own length, 
 the hollow being at amidships) is probably worse for the ship than the most 
 extreme condition actually encountered, and the bending moment so 
 obtained is the severest ever likely to be experienced. 
 
 On coming to the moment of inertia at a section of the vessel where 
 the bending moment is greatest, and the section weakest — which will 
 generally be in the neighbourhood of amidships — though in ordinary 
 practice no deduction is usually made for rivet holes in the weakest 
 section, still, whether such deductions are or are not made in order to 
 obtain the truest results, the assumption that the workmanship is of a 
 thoroughly reliable and trustworthy character is always made. It must 
 also be assumed that an intelligent disposition of butts has been adopted 
 in the strakes of shell and deck plating and other parts, and that the butt 
 connections are of a thoroughly satisfactory nature. 
 
 Taking the calculation, therefore, in its entirety, the tensional or 
 compressive stresses deduced may be other than accurate. Then the 
 question may be asked — Of what use is the performance of so laborious a 
 calculation 1 In the first place, it is a fair approximation to accuracy, and 
 most probably does not err on the side of leniency, but produces a calculated 
 stress which is larger than that actually experienced under the most trying 
 circumstances at sea. 
 
 But its best use lies in its efficiency as a means of comparison. For 
 example, having a standard of strength for a particular freeboard — which 
 at present is that provided by Lloyd's Kules for 1885 — in designing a"new 
 vessel independently of Lloyd's or any other rules, by a comparison of the 
 maximum tensile and compressive stresses from the two sets of scantlings, 
 the efiiciency of the designer's scantlings as compared with the strength 
 standard may be estimated. 
 
 Assuming that it is equivalent, it will be further advisable to compare 
 the weight of one foot of midship length for each vessel, as it will be 
 clear that equivalent strength obtained by means of a greater weight 
 of longitudinal material is a most serious objection, resulting, as 
 it does, both in increased cost of vessel and reduced deadweight, 
 while there is nothing to recommend it. The best and most scientifi- 
 cally designed vessel is that in which the lightest material is so disposed 
 and combined as to produce the greatest strength, that is, the least 
 stress. 
 
 Or again, where minimum freeboard, as specified by the Board of Trade 
 Freeboard Tables, does not form an item in the designer's consideration, as, 
 for instance, in the case of light high-speed channel or river passenger 
 
STRESS AND STRENGTH. 
 
 99 
 
 steamers, the comparison may be made apart altogether from any such 
 standard of strength as imposed by Lloyd's 1885 Rules. 
 
 Thus, having a light type of high-speed passenger steamer which, 
 after years of life at sea, has proved herself thoroughly efficient in 
 structural strength, the maximum calculated stresses may reasonably 
 be used in making a comparison of strength for a new similar type of 
 vessel. 
 
 Calculation for Position of Neutral Axis and Moment of Inertia of 
 an actual Ship. — The following calculation is for the moment of inertia 
 of the section of the vessel shown in fig. 42. It is scarcely necessary to 
 explain that only continuous longitudinal material is dealt with in such a 
 calculation. 
 
 Calculation to find Neutral Axis. 
 
 Name of Part. 
 
 Sectional Area 
 
 Distance of Centre 
 of Gravity above 
 
 Moment. 
 
 
 in sq. in. 
 
 Keel in feet. 
 
 
 Keel plate, .... 
 
 12-6 
 
 0-0 
 
 
 
 Garboard strake, 
 
 
 
 19-8 
 
 0-0 
 
 
 
 Bottom plating, 
 
 
 
 60-0 
 
 ■3 
 
 18 
 
 Bilge 
 
 
 
 49-8 
 
 1-8 
 
 89-6 
 
 Side ,, 
 
 
 
 84-0 
 
 11 
 
 924 
 
 Sheer strake, 
 
 
 
 22-6 
 
 18-6 
 
 420-3 
 
 Deck stringer plate, 
 
 
 
 17-0 
 
 19-3 
 
 328-1 
 
 Gunwale bar, 
 
 
 
 4-0 
 
 19-3 
 
 77-2 
 
 Deck plating, 
 
 
 
 33-1 
 
 19-6 
 
 648-7 
 
 Centre keelson, . 
 
 
 
 8-1 
 
 1-5 
 
 12-1 
 
 Centre keelson bottom angle, 
 
 4-0 
 
 •15 
 
 •6 
 
 ,, ,. top ,, 
 
 3-2 
 
 2-8 
 
 8-9 
 
 Side keelson, .... 
 
 10-9 
 
 1-6 
 
 17-4 
 
 Margin plate. 
 
 
 7-5 
 
 20 
 
 15-1 
 
 Tank side angle, 
 
 
 2-8 
 
 •8 
 
 2-2 
 
 Tank top centre strake, 
 
 
 7-2 
 
 3-0 
 
 21-6 
 
 ,, remainder, . 
 
 
 50-4 
 
 3-0 
 
 151-2 
 
 Upper side stringer plate. 
 
 
 8-0 
 
 120 
 
 96-0 
 
 ,, bulb angles. 
 
 
 16-9 
 
 12-0 
 
 202-8 
 
 Lower side stringer plate. 
 
 
 8-0 
 
 5-8 
 
 46-4 
 
 ,, bulb angles. 
 
 
 16-9 
 
 5-8 
 
 98-0 
 
 
 
 
 446-8 
 
 3178-2 
 
 3178-2 
 446-8 
 
 = 7-1 ft. Distance of neutral axis above keel. 
 
 It will be seen that the calculation for moment of inertia of any section 
 of a vessel does not involve any serious amount of labour. It is divided 
 into two parts. In the first, the neutral axis is found, the moments being 
 computed relatively to a base line passing through the top of the keel. 
 This is necessary, as explained in the foregoing pages, in order to find the 
 moment of inertia in the second stage of the calculation. 
 
 Estimate of Bending Moment for an actual ship, and Stress per 
 square inch. — But before the stress upon either the top or bottom sides 
 
lOO 
 
 STEEL SHIPS. 
 
 Calculation of Moment of Inertia of a vessel when supported upon wave crest 
 at middle of length, and thus subject to Hogging Strain. 
 
 
 
 Distance ol 
 
 
 
 
 
 
 
 Sectional 
 
 Centre of 
 
 
 
 Depths 
 
 
 
 Particular Item in 
 
 Area 
 
 Gravity 
 
 h- 
 
 A X h- 
 
 of 
 
 f/2 
 
 tV X A 
 
 Ship Girder. 
 
 = A. 
 
 from 
 
 Vertical 
 
 X cP 
 
 
 Sq, in. 
 
 Neutral 
 Axis=7i. 
 
 
 
 webs = d. 
 
 
 
 
 
 ft. 
 
 
 
 ft. 
 
 
 
 Keel plate, .... 
 
 *12-6 
 
 7 
 
 1 
 
 50-4 
 
 635-0 
 
 
 
 
 
 Garboard strake, 
 
 
 19-8 
 
 7 
 
 1 
 
 50-4 
 
 997-9 
 
 
 
 
 
 Bottom plating, . 
 
 
 60-0 
 
 6 
 
 8 
 
 46-2 
 
 2772-0 
 
 
 
 ..• 
 
 
 Bilge 
 
 
 49-8 
 
 5 
 
 3 
 
 28-0 
 
 1394-4 
 
 4-2 
 
 17-6 
 
 73-0 
 
 Side ,, 
 
 
 84-0 
 
 3 
 
 9 
 
 15-2 
 
 1276-8 
 
 14-0 
 
 196-0 
 
 1372-0 
 
 Sheer strake, 
 
 
 22-6 
 
 11 
 
 5 
 
 132-2 
 
 2987-7 
 
 2-9 
 
 8-4 
 
 15-8 
 
 Deck stringer plate, 
 
 
 17-0 
 
 12 
 
 2 
 
 148-8 
 
 2529-6 
 
 
 
 
 Gunwale bar, 
 
 
 4-0 
 
 12 
 
 2 
 
 148-8 
 
 595-2 
 
 
 ... 
 
 ^.. 
 
 Deck plating, 
 
 
 331 
 
 12 
 
 5 
 
 150-2 
 
 5170-2 
 
 
 
 
 Centre keelson, . 
 
 
 8-1 
 
 5 
 
 6 
 
 31-3 
 
 253-5 
 
 3-0 
 
 '9-0 
 
 "e-o 
 
 Centre keelson bottom angle, 
 
 4-0 
 
 7 
 
 
 
 49-0 
 
 196-0 
 
 
 
 
 ,, top ,, 
 
 3-2 
 
 4 
 
 3 
 
 18-4 
 
 58-8 
 
 
 
 
 Side keelson, 
 
 10-9 
 
 5 
 
 5 
 
 30-2 
 
 329-1 
 
 '2-6 
 
 '6-7 
 
 "e-o 
 
 Margin plate, 
 
 7-5 
 
 5 
 
 1 
 
 26-0 
 
 195-0 
 
 1-8 
 
 3-2 
 
 2-0 
 
 Tank side angle, . 
 
 2-8 
 
 6 
 
 3 
 
 39-6 
 
 110-8 
 
 
 
 
 Tank top centre strake. 
 
 7-2 
 
 4 
 
 1 
 
 16-8 
 
 120-9 
 
 
 
 
 
 ,, remainder, . 
 
 50-4 
 
 4 
 
 1 
 
 16-8 
 
 846-7 
 
 
 
 
 
 Upper side stringer plate, . 
 
 8-0 
 
 4 
 
 9 
 
 24-0 
 
 192-0 
 
 
 
 
 
 ,, bulb angles, 
 
 16-9 
 
 4 
 
 9 
 
 24-0 
 
 405 6 
 
 
 
 
 
 Lower side stringer plate, .. 
 
 8-0 
 
 1 
 
 7 
 
 2-8 
 
 22-4 
 
 
 
 
 
 ,, bulb angles, 
 
 16-9 
 
 1-7 
 
 2-8 
 
 47-3 
 
 
 
 
 
 
 
 
 
 21136-9 
 
 
 
 1474-8 
 
 
 
 
 
 1474-8 
 
 
 
 
 
 
 
 
 22611-7 
 
 Moment 
 
 of inertia for 
 
 
 
 
 
 
 half se 
 
 ction. 
 
 ^ 
 
 
 
 
 
 2 
 
 For botli 
 
 sides of vessel. 
 
 
 
 
 
 45223-4 
 
 Total m( 
 
 )ment 
 
 f inertia. 
 
 of an actual ship can be ascertained, the bending moment has to be found. 
 This, as previously shown, is a very laborious and lengthy process of 
 calculation. But happily, for ordinary types of vessels of usual proportions, 
 whether cargo or passenger, ^^^th of the length on the load water-line 
 multiplied by the displacement in tons is found to give a fairly accurate 
 approximation to the maximum bending moment experienced, and which 
 is developed when the vessel is supported upon the crest of a wave at amid- 
 ships, and enduring a ' hogging ' strain. This is the bending moment 
 usually adopted by designers in making stress calculations. For certain 
 
 * The sectional areas in the above calculation are for one-half of the section of the 
 vessel. It is thus necessary to multiply the moment of inertia for the half section 
 by 2 in order to obtain the total moment of inertia. 
 
STRESS AND STRENGTH. lOI 
 
 special types of vessels, the fraction of the length may drop as low as 
 ^Vtli or even -jVth. But, as just stated, for tlie comparison of ordinary 
 types of vessels, -^th of the length is generally taken. 
 
 The vessel whose midship section is shown in fig. 42, and moment of 
 inertia calculation on previous page, is 240 ft. long, 35 '5 ft. extreme 
 breadth, and 16 ft. draught. Coefficient of fineness "72. 
 
 The displacement is : — 
 
 240 X 35-5 X 16 X -72 „^^, ,. , 
 OK = 2804 tons displacement, 
 
 , 240 X 2804 ,^^,^ . , ,. 
 
 and OK = 19227 maximum bending moment. 
 
 The neutral axis being 13'1 ft. from top of sheer strake, and 7*1 ft. 
 from the keel, tlie maximum tensional stress on the top of the sheer strake 
 Avill be — 
 
 19227 X 13-1 ^ ^ 
 
 — Aroo-\-± — = 5"5 tons per square inch, 
 
 and the maKimum comjpressive stress on the bottom plating will be- — 
 
 19227 X 7-1 . 
 
 — .--^^^ , — = 6 tons per square inch. 
 
 On account of the heavier massing of longitudinal material along the 
 bottom and bilge of a vessel, the neutral axis is generally considerably 
 nearer to the bottom of the vessel's section than the top. Consequently 
 the stresses are usually severest on the topsides. The great value of steel 
 upper decks is not only that they increase the moment of inertia of the 
 section, but that they raise the neutral axis also, and thus reduce the 
 excess of stress upon the topsides. 
 
 Value of Registration Societies to Shipowners. — The custom is so 
 prevalent in these days for shipowners to class their vessels in one or other 
 of the registration societies, that the shipbuilder, not being responsible 
 for the structural strength, seldom finds it necessary to calculate the 
 stresses, except in the case of special types of vessels not built to any 
 class. 
 
 It naturally follows that the most extensive, varied, and valuable of 
 such information must be held by the registration societies, who are 
 directly responsible for the strength of the vessels built to their require- 
 ments. With their wide and diversified experience of vessels of all types, 
 including the enormous value of jDeriodically coming into contact with 
 vessels which have to be surveyed and carefully examined in order to 
 maintain their class, no all-round judgment can be more authentic or 
 valuable than theirs. 
 
 Comparison of Stresses on Vessels increasing in Size. — The late 
 i\Ir. John, of Lloyd's Register, as the result of most extensive investigation 
 into the strength of ships, gave the following particulars of calculated 
 
I02 
 
 STEEL SHIPS. 
 
 stresses for vessels increasing in size, and which, are known to have proved 
 thoroughly efficient. 
 
 Register Tonnage of 
 
 Maximum Tensile Stress per square 
 
 Vessel. 
 
 inch upon Upper Part of Section. 
 
 100 
 
 1-67 
 
 200 
 
 2-36 
 
 400 
 
 3-55 
 
 600 
 
 3-72 
 
 800 
 
 4-59 
 
 1000 
 
 5-19 
 
 1500 
 
 5-34 
 
 2000 
 
 5-9 
 
 2500 
 
 7-08 
 
 3000 
 
 8-09 
 
 The proportions of the above vessels were about eight breadths and 
 eleven depths in length, and they were each equal to Lloyd's highest class 
 at that time. The maximum bending moment, when supported upon a 
 wave crest at the middle of the length, was assumed to be -^-gth. of the 
 length multiplied by the displacement. As the proportion of length to 
 breadth and to depth increases, the stresses increase, and similarly, when 
 the proportion decreases, the stresses decrease also. 
 
 Though Mr. John's stress results were published in 1874 for iron 
 vessels then in service, subsequent calculations made upon modern vessels 
 do not widely differ from these results, which still remain both valuable 
 and interesting. 
 
 In 1892, Mr. A. Denny, in a short, but very valuable paper read before, 
 the Institute of Naval Architects, showed that, for a typical vessel, 
 350' X 44' X 32', with a displacement coefficient of fineness varying from 
 •56 to "78, loaded with homogeneous cargo, and bunker coal consumed at 
 the end of a voyage of 5000 nautical miles, the maximum hogging stresses 
 upon waves equal to the length of the vessel, and of a height equal to -^Qth. of 
 their length, I'anged between 7'93 and 9-76 tons per square inch. For a 
 smaller vessel, 250' x 30' x 23', Mr. Denny also showed that with displacement 
 coefficients of fineness of "637 and "795 under similar conditions, excepting 
 that the length of the waves was equal to the length of the vessel, and their 
 heights equal to -^jih. of their length, the maximum calculated hogging 
 stresses were 4*28 and 5*57 tons per square inch respectively. 
 
 It is observed from the foregoing results of calculated stresses, that 
 the maximum stresses upon small vessels are very much smaller than 
 the maximum stresses upon large vessels of similar proportions. From 
 a purely strength of girder point of view, such wide disparity in stresses 
 should not exist, but while ships may, in some respects, be treated as, and 
 compared to girders, they are very much more than ordinary girders. The 
 exceedingly wide diversity of the demands made upon the strength and 
 stiffness of ships, which in many instances have to be met in small vessels 
 
STRESS AND STRENGTH. IO3 
 
 to au almost equal extent as in large vessels, necessitates the introduction of 
 much more material than the demands upon a simple girder would entail, 
 and as a result, the calculated stresses in small vessels are comparatively 
 small. 
 
 Further Remarks upon the Value of Stress Calculations. — As we 
 have pointed out and explained at an earlier stage, the calculated stresses 
 must not on any account be taken as the actual stresses experienced by 
 ships in performing their work at sea. As a matter of fact, we have no 
 means of ascertaining what the actual stresses are which a ship endures. 
 At the same time, all such calculations should, as far as it is possible, 
 approach the truth. In most stress calculations made for vessels 
 supported amidships upon a wave crest, and upon wave crests at their 
 ends with the wave hollow at amidships, it is usual to consider the water 
 pressure at equal depths in all parts of the wave crest and hollow as 
 though it were exactly the same as in smooth water. But this is not the 
 case. The upward heave of the wave reduces the weight of the particles of 
 water in the waves, and consequently varies their still water upward 
 thrust, while the water in the wave hollow, being below the normal level 
 (remembering that in waves it is the form only that travels, the water 
 itself only rising and falling), exerts a pressure varying in the opposite way. 
 These variations of pressure in wave water naturally have their effect upon 
 the buoyant supports given to the vessel, and certainly modify the results 
 obtained by considering the vessel to be among wave water, the pressure of 
 which at similar depths below all points of the surface is assumed to be 
 equal. 
 
 To introduce all such modifying agencies as these into the calculation 
 ■would mean enormous labour, and while it is very unusual to consider 
 them in making an actual calculation, it is well to remember that they do 
 exist. 
 
 As previously shown, the great value of all calculated stresses lies in 
 their value as a means of comparing vessels of similar types. 
 
 Although in a steel structure, where a stress is gradually and steadily 
 applied and as gradually removed, to expect the material to stand a working 
 stress of 9 or 10 tons per square inch would be quite reasonable when 
 it is of good uniform quality similar to that used for ship construction, 
 and the workmanship is of the most efficient character, yet in many steel 
 structures, and particularly in ships, the stresses experienced are most 
 irregular, and often for considerable periods when labouring in a seaway, 
 applied in the most sudden manner. 
 
 How true this is must be very clear when it is remembered that in 
 the rapid transit of waves during the period of a few seconds, a severe 
 tensile stress is converted into a compressive sagging stress, and this goes 
 on indefinitely. Under such circumstances, it is most important that the 
 material be not taxed to any such limit as may be considered safe under a 
 steadily applied stress. 
 
I04 STEEL SHIPS. 
 
 Working Stress.— Taking the factor of safety of steel as 5, the maximum 
 
 T . . , , 30 or 32 
 ■working stress should not exceed ^ = 6 to 7 tons. But as vessels 
 
 with a calculated maximum stress of 8 tons and over are doing service 
 without showing any signs of weakness, the natural inference is, that 
 calculated stresses are in excess of those actually experienced, which is 
 undoubtedly the case, for probably part of the assumptions made for 
 extreme conditions, in calculating the maximum stresses, are never actually 
 experienced. 
 
 However, whenever stresses work out, in cargo vessels, to between 8 and 
 10 tons, and in some cases even more than this, special attention should 
 be given to the strengthening of the top and bottom flanges of the ship 
 girder. 
 
 lErections on Deck.— Though short poops, short bridges, and short 
 forecastles increase the depth of the ship girder locally, they add no 
 material longitudinal strength to a vessel. Isolated strength may be 
 necessary to meet a local strain, and short erections may serve a certain 
 purpose, but these do not improve the strength of the ship girder as a 
 whole. Indeed, where short bridges are built, the sudden termination of 
 strength at their ends makes a decided weakness in the longitudinal 
 strength, especially when subject to the sudden and severe stresses ex- 
 perienced in wave water (the maximum bending momenta, it is remembered, 
 being greatest over the middle of the length), and if no provision were made 
 to cover this, signs of severe straining might be expected at the ends of such 
 bridges, — in the sheer strake and side plating and the deck stringer. 
 This would show itself in opened butt joints, sheared rivets, or, perhaps, 
 serious deflexion as illustrated in an exaggerated manner in figs. 29, 30, 
 31, and 32. Therefore the extra strength afforded by the bridge, or any 
 other erection terminating on the midship half length, must be blended into 
 the hull proper. This is usually done by doubling the sheer strake for some 
 distance at each end of the bridge, by doubling or thickening the deck 
 stringer in the same locality (or both methods), and by carrying the lower 
 strake of the bridge side plating, which is generally thicker than the 
 bulwark plating, for some distance beyond the bridge at each end, thereby 
 forming part of the bulwark plating. Neither slwrt poops, bridges, fore- 
 castles, nor the extra means for strengthening vessels at the ends of bridges 
 or other midship erections are taken account of in calculating the moment 
 of inertia of the section, and in estimating the stress upon the upper and 
 lower parts of the section. 
 
 Where, however, a long bridge is fitted, covering at least fths of the 
 midship length of the vessel, it affords very material longitudinal strength. 
 Certainly it does not cover the whole length of the vessel, but it extends 
 over that part of the length upon which the severest bending moments are 
 experienced. It is right, therefore, that such a bridge should be included in 
 calculating the longitudinal strength of the vessel, provided that care be 
 
STRESS AND STRENGTH. IO5 
 
 taken to avert the sudden termination of strength at the bridge ends in a 
 manner similar to that described for the ends of short bridges. 
 
 Similar precautions must be taken in dealing with all other erections 
 which terminate anywhere on the middle length of the vessel. Thus, raised 
 quarter decks which terminate on the middle length, and are not connected 
 with any other erections, require special attention at the end. Not only are 
 the precautions necessary as explained for short bridges, but the two lengths 
 of the ship of different depths are strengthened at the break by overlapping 
 their decks for several frame spaces, and uniting the overlapping parts by 
 diaphragm plates. (See fig. 47.) 
 
 Without such preventive means, injury such as described at the ends of 
 short bridges might be anticipated. Similar steps are necessary when the 
 raised quarter deck terminates against a bridge. The broad principle to be 
 kept in view is, that all sudden terminations in strength must be rigorously 
 avoided. In making calculations for strength, a section should be chosen 
 where the bending moment is greatest and the transverse section weakest ; 
 and only such longitudinal material as contributes substantial structural 
 value to the ship girder should be taken account of in the calculation. 
 
 Board of Trade Instructions for Comparing the Strength of Vessels 
 for Freeboard Purposes. — In making calculations of strength in order to 
 make a comparison fur freeboard purposes, the Board of Trade instructions 
 state that steel or iron decks which cover not less than f ths of the midship 
 length of the vessel are to be considered in the calculation just as they would 
 be if of the full length. 
 
 "Wood weather decks, if continuous throughout the midship portion of 
 the ship, are to be considered as equivalent to steel of Trg-th the sectional 
 area of the wood. 
 
 For the purpose of comparison of strength, the breadth of the hatchways 
 in the standard vessel is deemed to be |^rd the breadth of the deck, and 
 the tie plates are assumed to be fitted at the sides of the hatchways. A 
 deduction of ith is to be made for rivet holes in steel, and -^th in iron for 
 the parts under tension. 
 
 Deductions. — After the consideration we have given in the foregoing 
 pages to the strength of ships, a brief enumeration of a few deductions 
 from facts which must now be fairly obvious will appropriately conclude 
 this section. 
 
 1st. Vessels of great length require more longitudinal strength (larger 
 moment of inertia of cross section) than vessels of less length, even though 
 of identical transverse dimensions, simply because increased length, and, 
 in all probabiUty, increased displacement, produce an augmented bending 
 moment. 
 
 The classification societies provide for this by demanding additional 
 longitudinal strength in vessels of extreme proportions. 
 
 2nd. Long shallow vessels with their small depth of girder, and hence 
 small distance from neutral axis to upper and lower extremities, are 
 
I06 STEEL SHIPS. 
 
 greatly handicapped, as it were, in producing a satisfactory moment of 
 inertia of cross section sufficient to provide for the demands which may 
 be made upon their strength when subject to maximum hogging and 
 sagging bending moments. 
 
 While long shallow vessels of comparatively light scantlings may be 
 safely used for passenger service in rivers and channels, it is obvious 
 that, without extraordinary additions to their scantlings, they are totally 
 unfit for over-sea trade, and more especially so, if required to carry 
 miscellaneous cargoes. In such vessels, continuous longitudinal middle 
 line bulkheads afford most valuable longitudinal strength. 
 
 3rd. In all vessels, more transverse strength is required in the region of 
 amidships than elsewhere. This is apparent when it is remembered that the 
 girth is greatest here, with consequently maximum pressure from the water. 
 Transverse framing should therefore be of extra strength over the middle 
 length (usually iths length). 
 
 The maximum hogging and sagging bending moments also are greatest 
 over the middle length of ships, and all longitudinal material should there- 
 fore be of extra strength over this locality (usually the half length). 
 
 The special weakness caused by hatchways and other deck openings, 
 and openings in side plating, and the means which may be adopted for 
 providing the necessary strength, is considered in the chapter on * Structure.' 
 
CHAPTEK VI. 
 
 TYPES OF VESSELS. SECTION I. 
 
 Fundamental Types and Modifications of same— Relation between Deck Erections and 
 Deadweight— No Reduction in Freeboard for Excessive Strength in a vessel with 
 Full Scantlings— Determination of Type— Three Deck, One and Two Deck, Spar and 
 Awning Deck Vessels— Illustration of Principal Scantlings of foregoing types- 
 Vessels of Intermediate Grades between Three Deck and Spar Deck, and Spar Deck 
 and Awning Deck — Raised Quarter Deck Vessels — Maximum Stress— Partial 
 Awning Deck, Shelter Deck, Well Deck, Shade Deck Vessels, etc. 
 
 Fundamental Types of Vessels. — The great variety of purposes for 
 which ships are used, and the widely different kinds and densities of 
 cargoes which may be carried, have had the very natural effect of producing 
 vessels specially adapted for specific purposes. And thus, while ordinary 
 mercantile steamers may generally be classed under the heading of cargo 
 or passenger steamers — wholly or partially — much greater subdivision of 
 type is adopted. Hence ship-folk are in these days familiar with the three 
 principal or fundamental types of vessels — 
 
 (1) Vessels of full scantlings, knoion as Single, Tico, or Three Deck, 
 
 (2) Spar Dech Vessels, 
 
 (3) Atoning Dech Vessels, 
 
 and also with modifications of these types, such as Raised Quarter Deck 
 Vessels, Partial Atoning Deck Vessels, Shelter Deck Vessels, Shade Deck 
 Vessels, and in addition, vessels of more novel type, such as ' Turret,' 
 ' Trunk,' and other * Self-trimming ' steamers. 
 
 In this section it is purposed, by means of description and illustration 
 where necessary, to point out the special features in either design or 
 structure which produce the distinctions in these different types. 
 
 We have previously shown, in dealing with the subject of classification, 
 that the load line is assigned to a vessel strictly in proportion to her 
 structural strength with a minimum percentage of reserve buoyancy. 
 This, it will be understood, applies to every kind of vessel to which 
 a minimum freeboard is assigned. 
 
 Relation between Deck Erections and Deadweight. — Now, assuming 
 that a flush deck vessel has been built to the highest standard of structural 
 
I08 STEEL SHIPS. 
 
 strength, it will follow that no reduction can be made in the freeboard 
 unless additional reserve buoyancy, suitably situated, can be added to 
 the vessel. But as valuable additional buoyancy can be obtained in the 
 form of forecastles, bridges, poops and other superstructures covering either 
 the whole, or part of the length of the vessel, reductions in freeboard, 
 which may considerably augment the deadweight carrying capacity of 
 the vessel, are obtained. 
 
 No reduction in freeboard for excessive strength in a vessel of full 
 scantlings. Again, taking the case of a flush decked vessel (with no 
 erections whatever) built in excess of the highest standard of structural 
 strength, no concession could be obtained in the matter of freeboard, 
 simply because, as we have pointed out, such a vessel would have a less 
 percentage of reserve buoyancy than required by the freeboard tables 
 ■of the Load Line Act of 1890. And as seaworthiness embraces 
 other features in addition to structural strength and equally important, 
 to jeopardise the safety of a vessel by allowing to her the maximum 
 immersion which a consideration of her structui'al strength alone would 
 permit, would be exceedingly unwise. One of these important features 
 which is vastly influenced by the amount and disposition of the reserve 
 buoyancy is the stability, and in drawing up the ' Freeboard Tables,' a 
 minimum percentage of reserve buoyancy was specified for flush decked 
 vessels of ordinary type and proportions, such as, it was assumed, would 
 ■ensure as far as possible, favourable stability conditions when properly 
 loaded. 
 
 With these points kept in view, together with what has already been 
 said upon ' Classification,' we shall better be able to see how the diff'erent 
 types and modifications of types of vessels have arisen. 
 
 The Determination of Type. — Every reader knows that there are 
 many bulky cargoes of such comparatively small density that a vessel 
 could be loaded up to her hatches, and every available space in deck 
 erections occupied, and the total deadweight would not be sufficient to 
 immerse her to the maximum draught ; while, on the other hand, there 
 are other cargoes of vastly greater density with which the same vessel 
 could be brought to her load draught long before the total hold space had 
 become occupied. To such variations in the specific gravity of cargoes are 
 principally due the types known as awning deck, spar deck, and vessels 
 ■of full scantlings. 
 
 Three Deck Vessels. — Let us first suppose that a shipowner con- 
 templates the addition of a new vessel to his fieet. We now see that he 
 has more to consider than merely external dimensions and speed. He must 
 be satisfied in his own mind, to some extent at any rate, as to what is the 
 most probable kind of cargo this vessel will carry. If the probable cargo 
 be of considerable density, it will naturally follow that it will be 
 impossible, as with many of such cargoes, to fully occupy the whole 
 hold space before the vessel has reached her load water-line. He will 
 
TYPES OF VESSELS — SECTION I. IO9 
 
 therefore want that type of vessel which will carry the greatest dead- 
 weight in relation to her size, and as we have seen, the strongest type of 
 vessel built and that which permits of the greatest immersion (?'.<?., 
 to which the least freeboard is assigned), is that known as the ' three 
 deck ' vessel, though it is not actually necessary, as before shown, that 
 the vessel should possess three actual tiers of beams, for the lowest 
 tier may be dispensed with by introducing compensation in the form 
 of deep framing, web framing, etc., and, under some circumstances, even 
 the middle deck (if only required to be of wood) may be dispensed with, 
 the beams only remaining. 
 
 One and Two Deck Vessels. — Vessels of smaller size, though built 
 to the fullest scantlings, may only require two, or even one tier of 
 beams. 
 
 The amount and nature of erections on vessels of this kind, as for 
 Three Deck vessels, would probably depend chiefly upon the requirements 
 for the accommodation of the crew, protection of steering gear, and, as 
 in the case of Three Deck vessels, upon the proportions chiefly in relation 
 to the depth to length. Even in the case of erections such as those 
 just mentioned, considerable reduction may be obtained in the freeboard 
 (which means increased deadweight) according to the value of the reserve 
 buoyancy afi"orded by such erections. 
 
 Awning Deck Vessels. — Supposing, again, that the shipowner required 
 a vessel to carry cargo of comparatively light density, or say, to carry cargo 
 in the lower holds with passengers in the 'tween decks and other super- 
 structures, it is obvious that the three decked vessel of full scantlings 
 would be unsuitable for the purpose, for not only would he find it im- 
 possible to bring the vessel down to her load draught line with ordinary 
 miscellaneous cargo, but the excessive strength of the vessel would have 
 entailed much unnecessary outlay in the first cost, for excessive and 
 therefore unnecessary structural material is both useless and waste. Under 
 such circumstances the light awning deck type of vessel would probably be 
 adopted, which, possessing large freeboard, that is, high side out of water, 
 would have the additional advantage of being drier on deck, and there- 
 fore better adapted to the comfort and requirements of passengers. 
 
 Spar Deck Vessels. — Intermediate between the three deck vessel of 
 full scantlings and the awning deck vessel of greatly reduced scantlings, 
 comes the spar deck type of vessel, which is particularly suited for cargoes 
 of moderate density. In recent years this type has been extensively 
 adopted. All other vessels, whether they be called shelter deck vessels, 
 shade deck vessels, raised quarter deck vessels, partial awning deck vessels, 
 or vessels of special design for self-trimming purposes, may, from a 
 structural point of view, be assigned to one of the fundamental types, 
 though modified in certain of their structural features in a greater or less 
 degree. 
 
 The external view of the structure of a vessel, such as may be obtained 
 
no 
 
 STEEL SHIPS. 
 
 as she lies afloat unloaded, or in dry dock, affords little indication whether 
 she be of three deck, spar deck, or awning deck type. However, in 
 a fully loaded condition, the amount of freeboard would present a more 
 certain clue to the determination of her type. 
 
 Closer Examination of Types, — It will be necessary at this stage to 
 examine the structural features of these different types of vessels in more 
 minute detail. 
 
 Vessels op Full Scantlings. 
 
 First. One and Two Deck Type. — See fig. 45, which is a midship 
 section of a single deck vessel. 
 
 See also fig. 10, which is a midship section of a two deck vessel, built 
 
 M-^ 
 
 X 
 
 
 
 
 ^ 
 
 ,p. 
 
 8-7 
 
 ZO 
 
 S-7 
 20 
 
 m 
 
 20 
 
 
 Fig. 45. — Midship Section of a Single Deck Vessel. 
 
 with ordinary frames and widely spaced hold beams; fig. 11, which is a 
 midship section of the same vessel with web frames compensating for hold 
 beams; and fig. 12, which is also the same vessel built with deep framing, 
 likewise dispensing with hold beams. 
 
 The sectional profile of the two deck vessel shown in fig. 46, is 
 for the same vessel to whose midship section we have just referred 
 in fig. 12. 
 
 The following Tables of Scantlings for figs. 10, 11, and 12 (and 46) 
 will give to those readers unacquainted with the rules of the classification 
 
TYPES OF VESSELS — SECTION I. 
 
 Ill 
 
 societies, some idea of the sizes of material required for vessels of this 
 size. The scantlings are such as would comply with Lloyds' present 
 requirements for 100 Al class. 
 
 Scantlings for Fig. 10. 
 
 Scantlings for Fig. 11. 
 
 Scantlings for Fig. 12. 
 
 Lloyd's Numerals* — 
 
 J Girth, . . 34-4 
 
 I Breadth moulded, 1775 
 
 Depth, . . 20-00 
 
 1st number, . 72*15 
 
 Length, . . 241 
 
 2nd number, 
 
 17,388-15 
 
 Depth in length = 12-05. 
 
 8 — 7 
 Frames, 4^ x 3 x • 
 
 Frame spacing, 24 in. 
 Reverse frames, 3 x 3 x -Jj^. 
 Floors, 22| x ^%. 
 
 Centre keelson, 17 x — — — . 
 
 ,, angles, 5 x 4 x -^V 
 Rider plate, 11 x hf. 
 
 Side intercostal, ^%. 
 
 „ angles, 5 x 4 x ^%. 
 Bilge keelson bulb plate, 
 
 Bilge angles, 5 x 4 x ijV 
 Side stringer, ^ 
 
 I 5 X 4 x ^9^. 
 
 Two angles, j 
 
 Same as for fig. 10. 
 
 Frames outside tank, 4^ x 3 
 8-7 
 
 O '7 
 
 Web frames, 15 x 
 
 20 
 
 angles, 3 x 3 x /^ six 
 
 frame spaces apart (12 ft.). 
 
 Frames in double bottom, 
 
 Same as for fig. 10. 
 
 Floors, 35 in. deep x -^V 
 
 Centre through-plate, 44 x ^\. 
 
 Top angles, 4 x 4 x -^^. 
 Tank margin plate, 2i x ^. 
 Tank top centre strake, 36 
 8-7 
 
 ^"20- 
 Tank top under engines and 
 
 boiler, ^\. 
 Tank top in holds, •/^. 
 Side intercostal, ^. 
 
 ,, angles, 3 x 3 x -i^. 
 Formed by margin plate. 
 
 rTwo side or web stringers, 15 
 
 I 8-7 
 
 I ^ 20 ■ 
 
 I Angles, 3 X 3 X T^. 
 
 Same as for fig. 10. 
 
 Depth of frames 7 in., 
 composed of two angles, 
 ' - 7 
 
 5 X 3 X 
 
 20 
 
 Same as for fig. 10. 
 
 See frames. 
 
 Same as for fig. 11. 
 
 Centre through - plate 
 
 35 X ■^\. 
 Top angles, 4 x 4 x ■^^. 
 Bottom ,, 5 X 4 X 1^. 
 Same as for Fig. 11. 
 
 ame as for fig. 11. 
 
 Two side stringers. Plate, 
 
 Bulb angles, 9x3^ 
 
 13 - 12 ,^ 
 
 X — — — , or, alter- ^ 
 
 nate method, 
 
 Plate, 20 X ^-^• 
 
 ' 20 
 
 Angles, 5 J X 4 X ^. 
 
 See page 47 for explanation of numerals. 
 
112 
 
 STEEL SHIPS. 
 Tables op Scantlings — continued. 
 
 Scantlings for Fig. 10. 
 
 Upper deck beams, 5J 
 
 'X' O X Tj-jy. 
 
 Bulb angle. 
 
 J 
 
 Lower deck beams, 9| 
 
 -{ Bulb plate. 
 
 I Angles, 4 X 4 X ^'u. 
 L Covering plate, ^%. 
 Upper deck stringer, 34^ x 
 Lower ,, ,, 32 x 
 
 - 25 X ^V 
 Pillars, hold, 3|. 
 
 ,, 'tween decks, 2|. 
 Bulkheads, four in number, 
 
 thick. 
 Bar keel, 9 x 23. 
 
 Complete steel deck, ■^■^. 
 Reverse angles to upper and 
 
 lower decks alternately. 
 See fig. 10 for scantlings of 
 
 shell plating. 
 
 Scantlings for Fig. 11. 
 
 Same as for fig. 10. 
 
 Whole beams at head of web 
 frame, 8 x ^V bulb plate and 
 two angles, 3 x 3 x Z^,. 
 
 None. 
 
 Same as for fig. 10. 
 None. 
 
 Pillars, 3g. 
 
 None. 
 
 Same as for fig. 10. 
 
 Keel slabs on sides of centre 
 through-plate 1 in. thick. 
 
 Same as for fig. 10. 
 
 Reverse angles to upper deck 
 and stringer below alter- 
 nately. 
 
 Scantlings for Fig. 12. 
 
 Same as for fig. 10. 
 
 None. 
 
 Same as for fig. 10. 
 None. 
 
 Same as for fig. 11. 
 
 None. 
 
 Same as for fig. 10. 
 
 Flat plate keel, 3& 
 14 - 11 
 ^-20-' 
 
 Same as for fig. 10. 
 
 All reverse angles to- 
 upper deck, except in 
 the peaks, where ordin- 
 ary framing is adopted. 
 
 A single deck steam vessel, according to Lloyd's Rules, is one having 
 a depth from the top of the keel to the top of the beam at the centre (with 
 the normal round up) of less than 15 ft. 6 in., while a two deck vessel is one 
 of less than 24 ft. depth. As previously shown, the two deck vessel need 
 not necessarily possess two complete decks, for while the upper deck must 
 be a complete deck laid and caulked, the lower deck may only consist of a 
 tier of widely-spaced strong beams, or even these may be dispensed with, 
 if compensation in the form of web frames and stringers, or deep framing, 
 be introduced. 
 
 The 1st numeral, by means of which is obtained the spacing of the 
 frames, and the scantlings of frames, reversed frames, and floor plates, the 
 thickness of bulkheads, and the diameters of pillars, is determined by 
 adding together in feet the half-moulded breadth of the vessel amidships, 
 the depth at the middle of the vessel's length from the upper part of the 
 keel to the top of the upper deck beams with the normal round up, and 
 the half girth of the midship frame section measured from the centre line 
 at the top of the keel to the upper deck stringer plate. 
 
 By multiplying the 1st numeral by Lloyd's length,* the 2nd numeral 
 
 * See page 50. 
 
Fig. 46. 
 
 SECTIONAL PROFILE OF A TWO-DECK STEAM VESSEL, WITH A POOP, BRIDGE, AND FORECASTLE. 
 
 For Midship Section, see Fig. 12. 
 NOTE. — The lower tier of Beams is dispensed with, and compensated for by Deep Framing. 
 
 ti-m 1 1 1 1 \Mr\Bkim^^k\ 1 1 III I w\Mii^s^u\ 1 1 1 Tini I u I [m^vA-xxr 
 
 Tunnel Well 
 
 Water Ballast Tank Well under 
 Boilers 
 
 Fig. 47. 
 sectional profile of a raised quarter-deck vessel, having a poop, raised quarter-deck, and bridge house combined, and a forecastle. 
 
 This Vessel is built on the Web Frame system in lieu of a tier of widely-spaced Hold Beams. 
 
 Tunnel Well Hold Well 
 
TYPES OF VESSELS — SECTION I. II3 
 
 is obtained, by means of which are ascertained the scantlings of the keel, 
 stem, and stern post, keelson and side stringer plates and angles, the 
 thickness of the outside shell plating; and this 2nd numeral, in conjunc- 
 tion with the proportion of the vessel's length to depth, determines the 
 scantlings of deck stringer plates, thickness and extent of deck plating, 
 and special additions to structural strength which are required when the 
 vessel is of extreme proportions, that is, of over eleven depths to length. All 
 bulkheads extend to the height of the upper deck. When there is an 
 erection in the form of a bridge, poop, or forecastle, all the frames extend 
 to the height of the stringer plate upon the deck of such erections. 
 
 In vessels with two tiers of beams, the reversed angles should extend 
 alternately to the upper deck stringer plate, and to the top of the 
 continuous angle bar on the hold beam stringer. Should the hold beams 
 be dispensed with and web frames substituted, the reversed frames will 
 extend to the upper deck stringer plate and to the top of the angle bar on 
 the stringer plate next below it. With deep framing both the frame and 
 reversed frame extend to upper deck. In small single deck vessels, where 
 the numeral is less than 57, the reversed frames may extend to a less height 
 than that just given. As such erections as poops, bridges, and forecastles 
 are only considered as superstructures, damage to which it is not supposed 
 woald endanger the safety of the vessel, the thickness of side plating and 
 the scantlings of their beams and stringers are less than are required for 
 the main hull of the vessel. 
 
 Second. Tliree Deck Vessels. — In vessels of this type the upper deck 
 is the strength deck. By this we mean that the vessel depends for 
 structural strength upon the material up to and including the upper deck, 
 when she is subject to the various stresses which are experienced in smooth 
 and wave water, in a ballast, or fully or partially loaded condition. The 
 upper deck, moreover, mtcst be considered as the strength deck, because, 
 in the event of damage accruing in any part of the structure up to, or 
 including that deck, the safety of the vessel may be greatly imperilled. 
 A three deck vessel is one of 17 ft. or over in depth from the top of the 
 keel to the middle deck at centre (with the normal round up), or 24 ft. 
 or more to the upper deck at centre. 
 
 Here, again, these vessels need not necessarily possess three tiers of 
 beams, for tbe lowest tier may be dispensed with in a manner similar to 
 that described for the two deck vessel. The numerals for scantlings are 
 obtained somewhat differently from the method described for one and 
 two deck vessels. The 1st numeral is obtained by adding together the 
 half-moulded breadth, and the depth and half girth of the midship frame 
 section measured to the itjJper deck. From this sum 7 is deducted, and 
 the result is the 1st numeral. This, multiplied by the vessel's length, 
 gives the 2nd numeral. All bulkheads extend to the upper deck. All 
 frames extend to the upper deck stringer plate, and where poops, bridges, 
 or forecastles are fitted to the height of their respective stringer plates. 
 
I 14 STEEL SHIPS. 
 
 When ordinary framing is adopted, the reversed frames extend alternately 
 to the upper deck, and to the angle bar on the middle deck stringer 
 plate. If the vessel has deep frames, each alternate reverse frame may 
 stop at the middle deck angle, unless, as in figs. 48 and 49, the height 
 of 'tween decks exceeds 8 ft., in which case all the reversed frames 
 extend to the upper deck. The sheer strake is placed at the upper 
 deck. 
 
 While the original idea of a three deck vessel was one possessing two 
 or more complete decks laid and caulked, and a tier of hold beams, or a 
 substitute for these hold beams, it is now common for some of these vessels 
 to have not more than one laid deck, and whether a steel deck be required 
 will depend upon the size of the vessel's 2nd numeral, and the proportion 
 of depth to length. For example, a three deck vessel, with a 2nd numeral 
 of less than 23,000 and under ten depths in length, or 19,000 and under 
 eleven depths, would require no steel deck whatever ; and only when the 2nd 
 numeral had reached 25,000, if there be less than ten depths in the length, 
 would one complete steel deck be required. Where, however, a steam three 
 deck vessel requires not more than one steel deck, the wood middle deck 
 may be dispensed with by making an addition to such freeboard as would 
 be assigned in accordance with the Load Line Act to an ordinary three deck 
 vessel with, the middle deck laid. Should, however, the shipowner object 
 to any increase in freeboard, the middle wood deck may be dispensed with 
 if the frames and reverse frames are of a size such as would be obtained 
 from the 1st numeral without the deduction of 7. Moreover, all reverse 
 frames would be required to extend to the upper deck, or other compensa- 
 tion introduced. 
 
 The three deck vessel illustrated in figs. 48 and 49 has a shelter deck 
 erection extending all fore and aft, and on the shelter deck a bridge erection. 
 So valuable may the shelter deck be in affording longitudinal strength and 
 buoyancy, especially if it be covered with a steel deck as shown in the 
 diagram, that by estimating its value to the whole ship structure, it may be 
 found^to merit important consideration in the determination of the load line. 
 See further remarks upon Shelter Deck Vessels, jj. 127. 
 
 The dimensions of this three deck vessel are : — Length between per- 
 pendiculars, 443 ft. ; moulded breadth, 51 ft. 6 in. ; depth moulded, 33 ft. 
 3 in. ; 'tween decks, 9 ft. 4 in. Lloyd's 1st number is 108"55, and 2nd 
 number 47,870. There are 12*8 depths in the length. Some idea of the 
 sizes of the material for such, a vessel of the highest class, say 100 A 1 at 
 Lloyd's, may be obtained from the following particulars : — 
 
 10 — 9 
 The depth of the deep framing is 10| in., made up of angles — ^ — 
 
 in thickness. 
 
 Frame spacing, 26 in. 
 
 The longitudinal stringer intercostal plates are 20 in. broad x -|i 
 and the bulb angles 9 x 3| x — . 
 
TYPES OF VESSELS — SECTION I. I I 5 
 
 Upper deck stringer plate 67 in. x l^ for half length amidships to 
 51 in. X ^jj at ends. 
 
 9-8 
 Upper deck plating .^^ 
 
 ,, ,, beams, tee bulb, 11 in. x i-^. 
 Middle ,, stringer 67 in. x ii to 51 in. x J^^. 
 
 „ „ plating -^ 
 
 „ ,, beams, 12 in. x i^ in., tee bulb. 
 
 Shell plating. Sheer strake |^ in. ( + -^jj as part compensation for sheer 
 strake doubling) to |-§- at ends. 
 
 Keel 1 in. ( + -^jj as part compensation for keel plate doubling). 
 
 Garboard strake ^ in. (+0^0 ^^ P^'^^ compensation for keel plate 
 doubling). 
 
 ^, , ., ,14-11 
 
 Elsewhere, outside strakes — ^ — 
 
 . ., 13-10 
 
 „ inside „ ~20~ 
 
 11-9 
 
 Centre through plate in double bottom 48 x .-^ 
 
 9-8 
 1 loors —:jjr 
 
 Shelter deck ^ steel, or ^^ iron. 
 
 „ beams, 8 x 3 x tPjj. Bulb angles on every frame. 
 
 Shelter deck side ^^^ ( + 4jj as part compensation for sheer strake 
 doubling) to -j^- ends. 
 
 By reason of the extreme proportions, a doubling is required to the 
 sheer strake for its whole width. This has been dispensed with, and 
 compensation introduced into the upper side plating, as noted in brackets. 
 Moreover, the shelter deck stringer plate has been made of exceptional 
 strength — |-|ths, for the same cause. 
 
 In addition to the foregoing, ^ths are required upon the inside strake 
 of shell plating below the sheer strake, and ^%ths to the upper deck stringer 
 plate. 
 
 It may again be pointed out, that in all vessels of full scantlings, and of 
 very extreme proportions (Lloyd's say 13 depths in length), a substantial 
 erection (bridge) should cover the half length amidships. The structural 
 value of this has been fully treated in Chapter V. 
 
 Spar Deck Vessels. 
 
 First, of ordinary Rule Requirements. — In spar deck vessels, the 
 uppermost or spar deck is the strength deck as in the case of three deck 
 vessels, and for the same reason. As we have previously pointed out, spar 
 deck vessels are of lighter construction than vessels of the same dimensions 
 built to Three Deck rule. 
 
I 1 6 STEEL SHIPS. 
 
 These vessels require to have three tiers of beams, though as shown for 
 vessels of full scantlings requiring two and three decks, the lower tier of 
 beams may be dispensed with, providing that compensation be introduced in 
 lieu of the same. 
 
 Though a spar deck vessel may be of the same length and depth to spar 
 deck as for a similar three deck vessel, both the 1st and 2nd numerals are 
 considerably less. The reason for this is, that the half-girth and depth are 
 taken to the main or middle deck instead of to the upper deck as in three 
 deck vessels, with the result that a much smaller 1st numeral is produced, 
 which, multiplied by the length, gives a smaller 2nd numeral. But the 
 proportion of depth to length is also taken to the main deck, and con- 
 sequently, these vessels are of greater extreme proportions than the 
 corresponding three deck vessel, which may mean considerable additions to 
 structural strength.* However, as it is found that taking the depths to 
 length to the main deck would too severely penalise the spar deck vessel if 
 the ]'ule requirements for additional strength on account of extreme pro- 
 portion were given, while the vessel has in reality the benefit of the depth 
 between the main and spar decks, a deduction of two depths is allowed. 
 And thus a vessel of, say, sixteen depths to main deck in the length, would 
 be considered in taking out her scantlings as though she were only of fourteen 
 depths. 
 
 In spar deck vessels, all the bulkheads required by rule extend to the 
 height of the spar deck. The spacing of the frames, and the scantlings of 
 frames, reverse frames, thickness of bulkheads, depths and thickness of 
 floors, size of bar and flat plate keels, stems, stern posts, keelson and 
 stringer angles, constructive material in double bottoms, thickness of shell 
 plating up to and including the main deck sheer strake, main and lower 
 deck beams, are exactly the same as are required for a two decked vessel of 
 the same dimensions up to the main deck. In short, the vessel up to the 
 main deck is practically similar in construction to a two decked vessel. 
 Having, however, a valuable erection above the main deck, which is not 
 only an indispensable but also an integral part of the ship structure, a reduc- 
 tion of two depths is allowed in the proportion of depths to lengths in 
 consequence as previously described. 
 
 All the frames extend to the height of the spar deck stringer plate. 
 The reverse frames should extend to the upper part of the frames and to the 
 top of the angle bar on the main deck stringer alternately. In addition to 
 the spar deck, these vessels must have a complete main deck laid and 
 caulked. 
 
 The beams to main and lower decks are of the same scantling as would 
 
 * In no case, however, does the structural material in the upper parts of these 
 vessels, and the number and thickness of steel decks, require to be in excess of that 
 specified for the three deck vessel of the same dimensions. In the case of the bilge and 
 bottom, extra strength is to be introduced strictly in accordance with the actual pro- 
 portion of their length to depth to main deck. 
 
TYPES OF VESSELS — SECTION I. I I 7 
 
 be required for a two decked vessel. The spar deck beams are lighter than 
 required for the upper deck in three decked vessels. For example, a three 
 deck vessel of 40 ft. in breadth at the upper deck would require beams 
 under a steel deck of bulb angle 7| x 3 x ^^, while the beams for a 
 spar deck of similar breadth would be 7 x 3 x ^^^ . 
 
 While the plating immediately above the main deck sheer strake in 
 spar deck vessels is less in thickness than the platii:g below, and thus 
 considerably less than what would be required for a three deck vessel 
 between the main and upper decks, it is found to be necessary, in order 
 to provide structural efficiency, to increase the thickness of the topmost 
 strake of side plating (the spar deck sheer strake). Hence we find in 
 Lloyd's Rules that both a width and thickness is specified for the spar 
 deck sheer strake according to the 2nd numeral. In order to exemplify 
 the difference in thickness of the side plating between the main and 
 upper decks in a three deck and in a spar deck vessel, it is found on referring 
 to Lloyd's Rules, where the former is 12, 13 and 14 twentieths in thick- 
 ness, the latter need only be 9 twentieths. 
 
 While the foregoing are the principal features in the construction of ordinary 
 spar deck vessels, it will now be evident that the smaller numerals will 
 generally cause a reduction in the scantlings of the material throughout these 
 vessels, as compared with three deck vessels of the same dimensions. 
 
 Second, Spar Deck Vessels in excess of Rule Requirements. — By 
 carrying up all the reverse frames to the spar deck, and by introducing 
 additional material in such localities as will undoubtedly increase the 
 structural strength of the ship girder, concessions may be obtained in the 
 matter of freeboard. Such additions to strength might be gradually made 
 so as to ultimately bring the ordinary spar deck vessel through every stage 
 of structural strength, until an equivalent in strength to the three deck 
 vessel of full scantlings is produced, and naturally, corresponding allowances 
 in freeboard would be secured; or, in other words, the vessel being now a 
 three decker, a three deck freeboard would be assigned. 
 
 The freeboard for a spar deck vessel is measured downwards from the 
 spar deck. 
 
 The standard height of a spar deck 'tween decks is 7 ft. Should 
 this height be exceeded or he less than the standard, a modification must 
 be made in the freeboard assigned.* 
 
 Awning Deck Vessels. 
 
 Structural Features. — In awning deck vessels, the main deck is the 
 strength deck, it being assumed that damage to the erection above the 
 main deck would not necessarily endanger the safety of the vessel. 
 
 * In spar decked steam vessels requiriug not more than one steel deck, the wood 
 ' main deck ' may be dispensed with, provided the beams, frames, and reversed frames 
 be of the sizes required for three deck vessels, and all reversed frames be extended to 
 the spar deck. A special freeboard may then be assigned. 
 
I 1 8 STEEL SHIPS. 
 
 However, although the awning deck erection is only a comparatively 
 light superstructure erected on the main deck of the vessel, it aflfords both 
 valuable reserve buoyancy as well as additional strength to the ship girder, 
 and these are fully taken into account in the Freeboard Tables. 
 
 The freeboard for awning deck vessels is measured from the main deck 
 downwards. 
 
 The 1st and 2ud numerals for awning decked vessels are obtained in 
 the same manner, and are in every way identical with those of a spar 
 decked vessel of the same dimensions to the main deck — the girth and 
 depth being taken to the main deck. 
 
 Thus, the frame spacing, the size of frames, reverse frames, thickness 
 of bulkheads, depth and thickness of floors, size of bar and flat plate keels, 
 stem and stern posts, keelson and stringer angles, constructive material in 
 double bottom, thickness of shell plating from keel up to and including 
 the main deck sheer strake, main deck beams, hold beams, are all exactly of 
 the same scantlings as for a spar decked vessel of similar dimensions up 
 to the main deck, excepting that the addition for extreme proportions, as 
 we shall see, will in all probability be somewhat more excessive than in 
 the spar deck. Indeed, all the afore-mentioned scantlings are identical 
 with what would have been required had the main deck been the upper 
 deck, and the vessel, therefore, a simple two decker. 
 
 The beams for the awning deck are less in size than required for a 
 spar deck of the same breadth. For example, under a steel spar deck 
 40 ft. in breadth, bulb angle beams 7 x 3 x ^V would be required upon 
 every frame, while for an awning deck of similar breadth, bulb angles 
 6 X 3 X 2% would be sufficient. 
 
 The awning deck being simply a light superstructure, the sheer strake 
 is placed at the main deck (the strength deck). The plating above the 
 sheer strake is comparatively very light. For example, where in a three 
 deck vessel whose plating below the sheer strake and between the middle 
 and upper decks is required by rule to be ^f ths or l^l^ths in thickness, and 
 in a spar deck vessel -rfuths between the main and spar deck sheer strakes, 
 ■^ths only is required by rule for an awning deck vessel. 
 
 The awning deck erection being of much less structural value than the 
 structure of a spar deck vessel above the main deck, vessels of the former 
 type are more dependent for structural strength upon the material up to 
 the main deck, and in effect come to be of more extreme proportions than 
 the spar deck, and vastly more so than the three deck vessel. This fact 
 must therefore be taken into account and the main hull strengthened accord- 
 ingly. This is provided for by taking the ratio of length to depth strictly as 
 obtained, that is, exactly as though the vessel were of two deck type, the 
 erection being entirely ignored. Thus, unlike the spar deck vessel, no 
 reduction whatever is permitted in the proportions. Hence, more additional 
 strength is introduced into an awning deck vessel than into either a spar 
 or three deck vessel of similar external dimensions. 
 
TYPES OF VESSELS — SECTION I. II9 
 
 All bulkheads in awning deck vessels extend to the height of the main 
 deck excepting the collision bulkhead, which should reach to the height of 
 the awning deck. All the frames extend to the awning deck stringer 
 plate. All the reverse frames extend to the top of the angle bar which 
 runs continuously along the main deck stringer plate on the inside of the 
 frames. 
 
 Awning deck vessels must have a complete main deck laid and caulked, 
 and the coamings round all hatches should be built as though the main 
 deck were the weather deck. The importance of this will be evident when 
 the lightness of the superstructure above the main deck, and consequently its 
 more damageable nature, is considered. To vessels having strength in excess 
 of the awning deck requirements, but below the spar deck standard, a load 
 line is determined in proportion to their structural strength as compared 
 with a spar deck vessel. 
 
 The following Table will give some idea of the comparative scantlings 
 of a three deck, a spar deck, and an awning deck vessel, as required by 
 Lloyd's Eules. 
 
 [Table 
 
I20 
 
 STEEL SHIPS. 
 
 
 S=u 
 
 +3 nd 
 
 pi 
 
 
 
 CD 
 
 ^ 
 
 o 
 
 -H 1^ 
 
 
 o 
 
 
 B 
 
 ^ 
 
 H 
 
 O) 
 
 ^ 
 
 O) 
 
 
 
 
 02 
 
 o 
 
 o 
 
 +3 
 
 w 
 
 
 
 
 fi 
 
 '^ 
 
 
 CD 
 
 
 fTi 
 
 u 
 
 O) 
 
 < 
 
 
 <B 
 
 M 
 
 
 w 
 
 5 T3 't; d 
 I— I CO t> o 
 
 fl ••-I Sh 
 
 It'" 
 
 J <I> •r-l 
 <3 t3 
 
 .So CO 
 
 c-tJ 
 
 
 2 SOQ^ 
 
 03 
 
 .-S 
 
 
 !^ ^ t) 45 
 
 ^ g oo 
 
 r 5 -W kO . 
 §Q CO- 
 
 § ^ ni -S CO 
 
 o 
 
 o 
 
 o 
 
 « 
 
 be 
 
 '3 
 
 n3 o 
 • S ^ cu 
 
 '^ a 
 ?3 
 
 
 o o o.:^ 
 
 a °2 
 
 - s^ o 
 
 'Si) a. 
 
 
 -^ X 
 
 ^^ X -*< 
 
 J X a 
 
 '« ^ ^: 
 
 
 ■i^fl 
 
 tJ 
 
 rQ 
 
 
 S 
 
 
 
 •a 
 
 tS 
 
 6f) 
 
 ,jS 
 
 S 
 
 • »-* 
 
 ^ 
 
 ^ 
 
 trt 
 
 C5(3 
 
 
 r^lK 
 
 Cli 
 
 X 
 
 "te-'S 
 
 .3 £0 .« 
 
 
 +2 rrt ril 
 
 
 '^ 3 '^, 
 
 -C! S^ 
 
 fep5 
 
 60 q 
 
 .is cS 
 
 CO t^ & 
 
 a> a) (D 
 
 ^ " rg rH 
 
 ^■> "5 o 
 
 ,ry as v_-- aj 
 CO 5J a; ^ 
 
 -^3 J3 S 
 
 
 
 . 05 a 
 
 wScc 
 
 pR 
 
 2'o 
 
 O « 
 
 M (>] Til (M ( 
 
 ^"-^ 
 
 ' o 
 ) <I> 
 
 :.as 
 
 ) o 2-2 
 
 1-. o 
 
 1 'Sea, 
 
 2 ■> p » 
 
 X 
 
 '^ ^ .s 
 
 .2 <m;^ 
 
 ..-I i-f]^'.^ 
 
 M 
 
 X 
 
 CO tp 
 
 X a 
 
 o .5 
 
 S £ aj 'kiS =^ 1 
 
 K^lPaQ W Or 
 
 <" n a 
 
 <I> C <U ■» ri -S 
 T3 S '^ 13 '5 -^ 
 
 
 So ^Mfi; wS 
 
 .^ . CO -g to 02 -J^ 
 
 to ^ ^^ =- t« rg 2 ' 
 
 ^ r- 2 a S . 'te 
 jSa-uraiH-^s-it? 
 
 OJ j3 s 
 
 s S 
 
 to TO 
 
 13 
 
 a> 
 
 t§ -a 
 
 . o 
 
 P. 
 
 a 
 
 so 
 
 a 
 
 CD 
 
 ■n 
 
 a 
 
 
 05 
 
 c/T 
 
 a 
 
 cS 
 
 tu 
 
 T3 
 
 ^3 
 
 pi 
 
 o 
 ^3 
 
 
 a> 
 
 ,co 
 
 »!=x 
 
 -* 
 
 i^"" 
 
 a.B 
 
 H?! OT^ C^ n 
 
 X 
 
 grO.S 
 
 X . 
 
 X 
 
 Kte of 
 X g 
 
 2 s ^a 
 
 7 03 -"^ O tfcl 
 
 a ioH 
 
 S tj a? 
 
 'sba^- ^ _ 
 
 ^ -(J a -w 
 
 — ' C^ r— ( p. 
 
 
 s M 
 
 «0 o,KWto -^ 
 co.S^co CO " 
 
 • rHo -XtD-rJl— I'TSra 
 
 ^^^5^,-,^ a .2 " 
 
 „t»o a:>rtSo 
 ft^ S — c? ^ tfn.2 ^ 13 
 
 I S^ 8-^-^:52^3^^ sis. 
 
TYPES OF VESSELS — SECTION I. 
 
 121 
 
 -M CD 
 
 '-l-'^ 
 
 
 
 
 
 -si 
 
 
 
 
 -^-^ 
 
 'c3 
 
 -s^ 
 
 a 
 
 o o _ .S 
 
 O 
 
 
 toio ,^ 
 
 o 
 
 
 :; Tj< X „ 
 
 r3 ?; 
 
 ■* . OD^ 
 
 0=2 
 
 - rd^ '^ 
 
 
 "^ ^ . X 
 
 ^-^^ 
 
 1^ ^^ -IM 
 
 rG 
 
 -p, s^°° 
 
 , +J 
 
 
 t"^ 
 
 C 'S be cS 
 
 c3,-r 
 
 ck stri 
 
 er mid 
 z strin 
 eck be 
 
 
 
 -5 g g-^ 
 
 ^^ 
 
 
 
 Lowe 
 Awni 
 Main 
 li 
 Lowe 
 Awni 
 
 4-t 
 
 X 8 
 
 
 |x -^ 
 
 Qj q; id 
 
 
 ^ r^ 
 
 
 J3 
 
 CS ^ 
 'P.bp 
 
 +3 
 
 X S3 
 X'g 
 
 
 CO ^" 
 
 
 oi a> 
 
 
 
 
 c: 
 
 a g « 
 
 „ 
 
 
 
 a; 
 
 '^ 
 
 -»-5 
 
 ^ 
 
 -^ 
 
 
 
 CO 
 
 •* 
 
 •^ 
 
 o3 pi^ 
 
 
 ^ 
 
 
 • — 1 
 
 
 
 
 
 
 Cm a;> <i> 
 
 
 
 
 
 S'^ 
 
 
 •TIJ 
 
 
 tH-^ 
 
 
 a 
 
 
 P 
 
 
 
 
 
 (O c8 
 
 cS 
 
 c3 
 
 cfl 
 
 CC M 
 
 s 
 
 ^^ 
 
 B 
 
 
 >-> a oj X .2 
 
 
 f^' bc: ■ 
 
 o 
 
 ^-t; i^ a^ *^ 
 
 "f^V 
 
 -!= O (U 
 
 j3 0^.^ => te 
 
 CO -'^ •— ' 
 
 "J r- -- -- 
 O 00 53J3 ? O g 
 
 cS O 
 
 
 CO S 5; 
 
 a> 
 
 M 
 
 -g e3 to 
 
 3 
 O 
 
 11 
 
 P.'-r " 
 
 
 X ^ 
 
 X X 
 
 10 iin 
 
 «o X 
 
 ^co 
 
 
 X 
 
 -^ :s -^ 
 
 PS 
 
 ^%> 
 
 
 -& 
 
 -g 
 
 P.X 
 
 X 
 
 X 
 
 
 HUl . 
 
 
 
 •^,a 
 
 -* ,s 
 
 
 
 
 ^ « 
 
 X bO 
 
 X bO 
 
 rS S 
 
 -1?) s 
 
 ^^a 
 
 
 50 <U 
 
 igles, 6 
 r half le 
 42 X A§. 
 
 c3 tS 5 -: 
 
 _Q © ^ aj cu CO 
 
 - ,^ g ^ § , 2 
 
 t, t^ <u ^ oi t^ -H; 
 
 S a) t> rr! > S 5 
 
 ^ d^ <Ur5 OJ ^ R 
 
 o ^ vj o 
 
 := :3 
 
 P'S <D g S Ch 
 
 i^i te o S " 
 
 'be c« 
 
 p'& 
 
 ca OS 
 bD '^ 
 
 &C* 
 
 
122 STEEL SHIPS. 
 
 Modifications op Principal Types. 
 
 Raised Quarter Deck Vessels. — See figs. 47 and 50. A considerable 
 difference exists between an erection called a raised quarter deck, and such 
 erections as complete or partial awning decks, poops, bridges, and fore- 
 castles. Indeed, in the truest sense, a raised quarter deck is not an 
 erection, bvit a hona-fide integral part of the ship's hull — the term ' erection ' 
 being used for purposes of convenience. The raised quarter deck is in 
 reality an increase in the depth of the vessel over part of her length — 
 usually the after portion of length. (See figs. 47 and 50.) The height of 
 a raised quarter deck above the main deck seldom exceeds 4 ft., though it is 
 often of less height. It will scarcely be necessary to say that the raised 
 quarter deck erection is alioays an extension in the depth of the hull of the 
 vessel immediately above the main deck, and is never found above spar or 
 awning decks, where it would be practically useless. So valuable is the 
 increased buoyancy of the space provided by a raised quarter deck 4 ft. 
 in height, that it receives as much credit in the assignment of the freeboard 
 as would be given to a long poop erection of similar length and sub- 
 
 FiG. 50.* — Raised Quarter Deck Vessel. 
 
 stantially closed with a bulkhead at the fore-end, and of at least 7 ft. in 
 
 height, t 
 
 While it is customary to speak of the raised quarter deck being of a 
 certain height above the main deck, it must be clearly understood that 
 neither decks nor beams are fitted to the main deck immediately under the 
 raised quarter deck excepting for a short distance (a few frame spaces) at 
 the termination of the raised quarter deck or ' break,' as it is usually termed. 
 What is really done is to interrupt the continuity of the main deck by 
 lifting it up for part of its length to a certain height, and then term the 
 raised part, a raised quarter deck. It is therefore constructed exactly as 
 though it were the main deck carried continuously all fore and aft. 
 
 The more detail work of obtaining an efficient connection between the 
 main and raised quarter decks is dealt with a little later in this volume 
 (see page 123). It is customary to adopt raised quarter decks in one and 
 two deck vessels — seldom in three deck vessels, and it may safely be said 
 never when two laid decks are absolutely required. Perhaps the principal 
 reason which led to the wide adoption of the raised quarter deck type of 
 vessel may be given as follows : — 
 
 * The heavy black lines in figs. 50 to 55 divide the hull proper from the erections, 
 t When the raised quarter deck is of less than 4 ft. in height above the main deck, 
 the allowance for freeboard suffers reduction. 
 
TYPES OF VESSELS — SECTION I. I 23 
 
 In comparatively small vessels — say up to about 250 ft. ia length — it 
 was found that when the engines and boilers were placed amidships, so 
 much hold capacity was lost owing to the space occupied by the shaft 
 tunnel, in combination with the fact that most cargo vessels are designed 
 with the after body somewhat finer than the fore body in order to get a 
 form as consistent with speed requirements as possible, that when laden 
 with homogeneous cargo, to bring such vessels to a satisfactory trim was 
 practically impossible — the tendency being to trim by the head. By 
 increasing the depth of the hold space abaft of the engines by raising the 
 main deck (constructing a raised quarter deck), this objectionable feature 
 was obviated. Thus, a vessel with a raised quarter deck represents in some 
 degree parts of two separate vessels united into one. The structure of that 
 part without the raised quarter deck is in strict accordance with what would 
 be required for a vessel of the same depth throughout to the main deck. 
 In way of the raised quarter deck, however, all the frames extend to the 
 raised quarter deck stringer plate. The alternate reverse frames extend 
 to the same height, while the intermediate reverse bars are carried to a 
 height such as would be required for a vessel whose 1st numeral was 
 computed by the depth and half girth being taken to the raised quarter 
 deck. The construction of the raised quarter deck, as we have already 
 stated, is a continuation of the main deck. While the number and 
 arrangement of hold beams, beam stringers, and stringers in hold in that 
 part of the vessel uncovered by the raised quarter deck is in accordance 
 with the requirements for the depth taken to the main deck, in the after 
 part of the vessel, which comes in way of the raised quarter deck, these 
 structural items are regulated by the depth, which is now increased by the 
 height of the raised quarter deck above the main deck. It will now be 
 seen that the disposition of the stringers below the raised quarter deck 
 may differ considerably from that in the other part of the vessel. 
 Indeed, it is possible that the depth, which has now been increased 
 by the height of the raised quarter deck above the main deck, may 
 have actually produced the necessity for an additional tier of beams, 
 or at any rate, additional side stringers. These, of course, must be fitted, 
 unless in the event of an additional tier of hold beams being required, 
 compensation in the form of deep framing, or web frames, be fitted in lieu 
 thereof. 
 
 Now we have previously seen how highly important it is that the 
 continuity of all structural strength should be rigorously maintained. 
 Hence, in vessels of this type, great attention must be bestowed upon 
 the union of that part of the vessel covered by the raised quarter deck 
 and the fore length. This is effected by overlapping the main and raised 
 quarter decks for 2, 3, 4, or 5 frame spaces according to the size of the 
 2nd numeral and the proportions of the vessel ; while in the case of the side 
 stringers these also are overlapped, or, in some instances, scarphed into one 
 another. The overlapping lengths of the main and raised quarter decks are 
 
124 STEEL SHIPS. 
 
 further blended together by means of a number of fore and aft webs of 
 plating called diaphragm plates,* 
 
 In these vessels, the main deck sheer strake is usually carried 
 continuously right aft to the stern, while the side plating above the 
 sheer strake and in way of the raised quarter deck may be oVth of an inch 
 less in thickness than the shell plating below, if such is ^^^ths of an inch in 
 thickness or more. 
 
 As the stresses produced by longitudinal bending moments are always 
 severest wherever any sudden diminution or increase takes place in 
 longitudinal strength, it is very important that special strengthening be 
 introduced at the break, so as to blend, as it were, the strength of the 
 raised quarter deck with its increased depth into the other length of the 
 vessel. Hence it is usual to double part of the topside plating for a 
 reasonable distance before and abaft of the 'break.' This doubling is 
 usually fitted to the sheer strake, or when the vessel has a bridge house 
 connected to the raised quarter deck the better method is to double the side 
 plating of the raised quarter deck for a similar distance. In any case, the 
 raised quarter deck side plating must be strengthened in way of the break, 
 and thus, if it is not doubled, it must be increased in thickness, in addition, 
 of course, to the doubling of the sheer strake. 
 
 The thickness of the bulwark plating should be increased at the ends 
 of such erections, thus further tending to blend the strength of the one 
 part into that of the other. Great care must be exercised in arranging 
 the butts of the side plating and stringers in way of the break, and ample 
 strength should be introduced into the butt connections whether they 
 be strapped or overlapped. Usually all the butts are treble riveted 
 in the principal structural parts in way of the break, and the thickness 
 of butt straps increased. 
 
 As the break bulkhead is a most vital part in the structure, it is 
 essential that it be of satisfactory thickness (usually same as bridge side 
 plating), well constructed, and connected and stiffened. To this bulkhead 
 the raised quarter deck should be connected by means of double angles.! 
 
 Though distinguished by the name 'raised quarter deck' vessel, 
 this type belongs to the ' full scantling ' class, no such erection as a 
 raised quarter deck ever being built, as has already been stated, upon 
 
 * The length of these diaphragm plates is equal to the length of tlie overlap of the 
 raised quarter deck over the main deck, and the width to the distance between the 
 overlapping decks. The diaphragm plates should be well connected to these two decks 
 by double angles. For such a vessel as illustrated in fig. 47 about four diaphragm plates 
 would be required. 
 
 + In order to obtain the mininmm freeboard for well-deck vessels, it is necessary : — 
 1. When the crew is berthed in the forecastle, that the ' well ' between the bridge 
 and the forecastle be bridged by a gangway supported on stanchions (which may be 
 made portable if desired) so as to facilitate the passage of the sailors and firemen to their 
 respective quarters. Where, however, the well is 80 feet or more in length, or the vessel 
 is under 150 feet in length, the gangway just referred to is not required. 
 
TYPES OF VESSELS — SECTION I. I 25 
 
 spar or awning deck vessels. They must therefore be considered upon 
 this understanding. 
 
 Maxivium Stress. — Now what is a severe stress for the material in 
 a vessel of the spar or awning deck t3'pe is also a severe stress for one 
 of full scantlings. The idea must not be entertained that because these 
 types diflFer in their adaptation or efficiency for carrying deadweight, that 
 there is any varying degree of maximum stress to which they are respec- 
 tively capable of being subject, when fully and properly loaded to their 
 properly assigned load lines. Such an idea is erroneous. That is not 
 the manner in which ships are designed and classified. Indeed, the 
 absurdity of such an idea must be obvious after very little thought. 
 The only intelligent and scientific method which can be adopted in 
 considering the deadweight carrying capacity of any vessel is to assign 
 to her such a draught that when fully loaded with a homogeneous 
 cargo, the material in the structure subject to the greatest stresses — 
 tensiona), tortional, and compressive — will be able to endure the same 
 without the possibility of rupture or distortion occurring, and also with 
 no possibility of the material situated in any intermediate locality 
 between the regions of maximum stresses collapsing. 
 
 Thus, assuming that a three deck, a spar deck, and an awning deck 
 vessel were built to Rule Scantlings, and it were desired that when 
 each was fully loaded with a homogeneous cargo, not more than 7 
 tons stress per square inch of material should be experienced on the top 
 of the upper deck sheer strake, it would follow, after the comparison 
 ■we have made of the scantlings of these respective vessels (see table 
 on p. 120) that the spar deck vessel must have more freeboard than 
 the three deck type, and the awning deck more again than the 
 spar deck type, simply because the structural material is less in midship 
 sectional area in the spar than in the three deck vessel, and less in the 
 awning than in the spar deck type. And with less sectional area of 
 material in the midship sections of vessels of similar depth and 
 breadth, and disposed according to the usual practice for these respective 
 types, the moment of inertia must be less in the spar deck type, and 
 least in the awning deck type as compared with the three deck type. 
 And less moment of inertia with equal bending moments must produce 
 greater stress. But this must be obviated of absolute necessity. Hence 
 
 2. That the enguie and boiler openings be protected and entirely covered by an 
 efficient bridge house. 
 
 3. That the total area of the freeing ports in the bulwarks in the well be sufficient 
 to speedily rid the deck of water. This area is determined by the Board of Trade 
 according to the lengtli of the well. 
 
 4. That iu well deck vessels of the highest class having erections extending over 
 3^ths of their length, the bridge front bulkheads must be specially strengthened and 
 the hatch coamings at least 30 in. high above the deck, while in order to facilitate the 
 means of getting rid of seas shipped into the well, the freeing port area must be 25 per 
 cent, in excess of that referred to in the foregoing paragraph (3). 
 
126 STEEL SHIPS. 
 
 smaller bending momeat is obtained by carrying smaller deadweight, which 
 means less draught or more freeboard. Hence the varying draughts of three 
 deck, spar deck, and awning deck vessels. 
 
 However the design of a vessel may be altered, or whatever pecu- 
 liarities may be introduced into her construction, she must finally, 
 in receiving her statutory freeboard, come under such test as we 
 have just described, and must satisfy that with a fully loaded homogeneous 
 cargo of a certain density, she is not unduly stressed, and that she 
 possesses at least the minimum percentage of reserve buoyancy required 
 by the Load Line Act, 1890. 
 
 However, as we have shown in Chapter V., there is no such thing 
 as a standard maximum stress for all types and sizes of vessels ; for local 
 requirements, especially as seen in small vessels, may demand certain 
 structural strength in order to resist strains apart altogether from those 
 produced by longitudinal bending moments. Hence small vessels, when 
 fully loaded, are rarely stressed beyond 4 or 5 tons per square inch, while 
 large vessels may reach as much as 8 tons and over. 
 
 In the Freeboard Tables and Rules fixed by Act of Parliament, and 
 from which all assignments for British Load Line Certificates are made, 
 the three types of vessels we have thus far considered — 'full scantling,' 
 'spar deck,' and 'awning deck' — are all that are mentioned for steam 
 vessels. So that to what society's rules a vessel may be constructed and 
 classified, or to what special design she may be built matters not. If 
 she is required to have the minimum freeboard, she must belong, 
 from a structural strength point of view, to one of these three types, or 
 to some modification or intermediate grade between them, and upon the 
 strength so ascertained will she receive her statutory freeboard. 
 
 Deck Erections other than Awning and Kaised Quarter Decks. — 
 Having noted the value of the complete awning deck erection from both 
 a buoyancy and a structural strength point of view, and also the raised 
 quarter deck, which, it may not be unwise to reiterate, is much more than 
 an erection, or it may be still more correct to say that it is not an erection 
 at all, but an integral and vital part of the vessel's hull, we shall briefly 
 examine the structural features and value of other erections which rank of 
 less importance. Vessels having such erections may be described and 
 illustrated as follows : — 
 
 L Partial Aioning Deck Vessels. — With long raised quarter deck and 
 partial awning deck covering the whole length. See fig. 51. 
 
 After the description already given for raised quarter deck and awning 
 deck vessels, little more needs to be said about this type. Such vessels 
 are built to scantlings from numerals obtained by taking the girth and 
 depth up to the main deck, the erections agreeing with the rule require- 
 ments for raised quarter decks and awning decks respectively. The part 
 awning deck erection, which, as has been shown, is of comparatively light 
 construction, is well suited for light cargo, sv;ch as cattle or even passengers. 
 
TYPES OF VESSELS SECTION I. I 27 
 
 It affords protection to the engine and boiler casings, and forms, with the 
 raised quarter deck, a continuous erection providing valuable reserve 
 buoyancy, for which due allowance is made in determining the freeboard. 
 The main and raised quarter decks are scarphed (overlapped) as for vessels 
 with raised quarter decks unconnected to bridges or part awning decks, 
 
 Fig. 51. — Vessel having long raised Quarter Deck and partial Awning Deck together 
 completely covering the whole length (with or without a poop). Known as 
 partial Awning Deckers. 
 
 except in the very smallest vessels, where a system of knee bracketing is 
 sufficient.* 
 
 2. Shelter Decked Vessels. — See fig. 52. 
 
 The great development which has taken place in recent years in the 
 cattle carrying trade is chiefly responsible for the evolution of the ' shelter 
 deck ' vessel. This is usually a three deck vessel with a complete erection 
 
 ^ 
 
 Fig. 52.— Shelter Decked Vessel. 
 Note. — Dotted lines in Shelter Deck at A denote openings (not hatchways). 
 
 all fore and aft, entirely closed from the sea, with the exception that it is 
 fitted with one or more openings through the shelter deck, which may 
 extend from side to side of the deck, or simply be an opening resembling 
 
 * It maj' be pointed out that the Board of Trade Freeboard Eules allow even a 
 less freeboard in vessels of this type than for plain rule awning deck vessels if certain 
 structural conditions are fulfilled whereby the partial awning deck erection is brought 
 up to a state of structural efSeiency more in harmon)' with the strength of the raised 
 quarter deck. It is also assumed that the partial awning deck covers the engine and 
 boiler openings. These structural conditions, the rules state, must be in accordance 
 with the requirements described in section 44 of Lloyd's Rules for the year 1889, and 
 include increased thickness of side plating, stronger stringer plates, a steel or iron deck, 
 and increased riveting in the side plating. Where, however, the partial awning deck 
 is built to the ordinary rule requirements for awning deck vessels, the allowance for 
 freeboard will be similar to that for a vessel with a complete awning deck. 
 
128 
 
 STEEL SHIPS. 
 
 a hatchway, excepting that no coamings are fitted, and no permanent 
 means provided for battening down the opening with planks. With such 
 a deck opening or openings, it is evident that, in heavy weather, it is not 
 improbable that in shipping seas on deck, water may find its way into the 
 'tween decks between the upper and the shelter decks. Hence, waterports 
 and scuppers are fitted in shelter deck spaces. 
 
 r 
 
 Fig. 53. — "Well Deck Vessel, first type, long poop and bridge combined, 
 and topgallant forecastle. 
 
 Well Deck Vessel of second type is that previously described and illustrated in 
 figs. 47 and 50. 
 
 A principal reason why this type of erection is generally adopted for 
 the carrying of cattle is that all unenclosed space in a shelter deck 'tween 
 decks is exempted from the gross tonnage of the vessel, while according 
 to its value in affording reserve buoyancy and strength to the ship as a 
 whole, it receives consideration in the assignment of the freeboard. 
 
 Fig. 54. — Vessel with poop, bridgehouse, and forecastle. 
 
 There are no definite rules for the construction of shelter deck erections. 
 But, with the exception that the shelter deck is an erection upon a three deck 
 vessel (in arriving at whose first numeral a deduction of 7 is made, while no 
 such deduction is made for the awning deck numeral), it is very similar in 
 construction to an awnin^ deck erection. Often the shelter deck is of steel. 
 
 Fig. 55. —Shade Deck Vessel. 
 
 Against its value from a strength standpoint there is to be considered 
 the diminished value of the buoyancy of such 'tween decks owing to the 
 openings already mentioned. By fitting complete iron or steel bulkheads 
 to the poop, bridge, and forecastle spaces in the shelter deck 'tween decks, 
 its value as an erection is considerably enhanced. 
 
 Well Deck Vessels. — With long poop and bridge combined, and top- 
 gallant forecastle. See fig. 53. 
 
TYPES OF VESSELS — SECTION I. I 29 
 
 The particulars of the principal structural features of sucli erections may 
 be enumerated as follows : — 
 
 (1) All frames extend to poop, bridge, and forecastle decks. 
 
 (2) Poop and bridge beams of same size as awning deck beams. 
 
 (3) Forecastle beams similar to spar deck beams. 
 
 (4) Side plating of same thickness as required for awning deck vessels. 
 
 (5) As the forecastle sides are particularly exposed to the full force of the 
 sea, extra stiffness to the plating should be introduced, especially in large 
 vessels. This may be done by carrying up the alternate reverse frames, 
 or by fitting a double angle stringer midway between the forecastle and 
 upper deck. 
 
 (6) The bulkhead at the fore end of the combined poop and bridge 
 house should be of same thickness as the side plating — the coamings being 
 ■ffVth of an inch thicker. This bulkhead should be well stiffened with 
 bulb plates and angles or channel bars or any other equivalent, spaced about 
 30 in. apart. These bulb stiffeners should be thoroughly connected to the 
 coaming plates and the decks between which the bulkhead extends. If 
 these decks are not plated, athwart-ship plates should be fitted to their 
 beams, so that an efficient connection by means of bracket knee plates may 
 be eflected, (See fig. 56.) 
 
 The importance of this will be evident when it is remembered that the 
 well between the forecastle and the bridge is specially subject to the incur- 
 sions of heavy seas, which often dash furiously against the bridge front 
 bulkhead, damage to which might seriously endanger the safety of the 
 vessel should water find its way into the deck openings over the engines 
 and boilers, w'hich are enclosed by comparatively light casings. 
 
 (7) "We have previously drawn attention especially to the necessity of 
 considerable additions to the strength of the topsides in the way of the main 
 deck in awning deck vessels of extreme proportions, that is, over eleven depths 
 to length. The same necessity for additional strength exists where a con- 
 siderable proportion of the length of a vessel (f ths or over) is covered by 
 erections. Thus, in well deck vessels of this type, whenever either of these 
 conditions exist, strength to the topsides over the middle half length of the 
 vessel should be introduced either by doubling the sheer strake or increasing 
 the thickness of the bridge side plating and stringer plates, and the efficiency 
 of the butt connections of these parts. In any case, the sheer strake should 
 be doubled for about 20 ft. at the fore end of the bridge, and the raised 
 quarter deck side plating at the break for a similar distance. 
 
 Vessel ivith Poop, Bridge, and Forecastle. — See fig. 54. All the 
 frames extend to the height of the stringer plate on the deck of each of 
 these erections. In neither the poop nor the bridge is it necessary to carry 
 any reverse frames above the upper deck, unless tlie 'tween deck height of 
 such erections be unusually great, when additional support would necessarily 
 be required. In the case of the forecastle, whose sides are exposed to the 
 full force of head seas, the necessity for ample strength and stiffening is 
 
 9 
 
I30 
 
 STEEL SHIPS. 
 
 evident. This can be obtained by extending the alternate reverse frames to 
 the forecastle deck, or by fitting a double angle or tee bar stringer along the 
 inside of the frames midway between the two decks. When these stringers 
 meet at the stem bar, they should be connected by means of a plate bracket 
 or breast hook. 
 
 While the side plating of poops, bridges, and forecastles is usually 
 comparatively thin, it may be advantageous in certain cases to increase 
 the thickness of the forecastle side plating for the reason just given 
 
 A stringer plate is fitted to the ends of the beams on each of these 
 erections. 
 
 Bridge Deck 
 
 r 
 
 o o 
 o 
 
 o 
 o 
 
 o 
 
 o 
 
 b o 
 
 r 
 
 \ , Upper Deck 
 
 Fig. 56. — Bridge front Bulkhead stiffened with Channel Bars. 
 
 In determining the load line for vessels of this kind, an allowance is made 
 for the erections strictly in accordance with the vahie of the buoyancy which 
 they afford. Thus, if the bridge be closed with a thoroughly efficient water- 
 tight bulkhead at each end, and the poop with a bulkhead at its fore end, 
 the allowance will naturally be more than would have been made had these 
 spaces been open at the ends or pierced by alleyways or other permanent 
 openings. The bridge front particularly should have a good coaming 
 plate well connected to the deck, and the stiffeners should be at least as 
 
TYPES OF VESSELS — SECTION I. I3I 
 
 large as the main frames, spaced about 30 in. apart, with bracket 
 connections to the decks at their upper and lower extremities. 
 
 Shade Deck Vessels. — The principal function of a shade deck is to 
 afford a shelter and provide a promenade for passengers. As the sides 
 are open, the 'tween-deck space affords no buoyancy, and as the construction 
 is usually of a very light character, because of the light duty it has to 
 perform in the economy of the ship structure, it is practically valueless 
 in contributing to the general structural strength of the vessel. It should, 
 however, be sufficiently strong for the purpose for which it is erected, so 
 as in no way to endanger the lives of the passengers or crew. Such decks 
 may be constructed in a variety of ways ; frame angles, tee bars, or round 
 iron stanchions may be used to carry the shade deck beams, while the shade 
 deck itself may be built of light beams with light stringer and tie plates 
 upon which the wood deck is laid (see fig. 55). 
 
 Steel Sailing Ships. — Owing to the severity of the transverse stresses 
 experienced by sailing ships, tending to rack and distort the form in that 
 direction, additional transverse strength is introduced into large vessels 
 
 (1) by carrying the reversed frames up to the gunwale on every frame, 
 
 (2) by making the beams deeper than would be required for steam vessels 
 of the same size, at any rate to the lower and orlop decks. Special care 
 should be taken in the formation of the beam knees. When the knees are 
 not made of bracket plates, the beam ends should be turned down to form 
 the knees as shown in fig. 111. All beam knees should be three times the 
 depth of the beam. (3) When no steel deck is fitted, there should be, in 
 addition to the fore and aft tie plates, diagonal tie plates laid upon the 
 deck, so as to effectively transmit the severe stresses from the masts and 
 the adjacent beams to the stringer plates. 
 
132 STEEL SHIPS. 
 
 SECTION II. 
 
 Description and Illustration of the Principal Structural Features of some noted, and 
 also some novel Types of Vessels — ' Campania ' and 'Lucania' — ' Great Eastern ' — 
 Turret Steamers (Doxford's)— Trunk Steamers (Ropner's)— Self Trimming Steamers 
 (Priestman's)— Some advances in Modern Shipbuilding — Steamers for carrying Oil 
 in bulk — Water-ballast Arrangements. 
 
 ' CAMPANIA ' AND ' LUCANIA ' (see figs. 57 and 58). 
 
 In these magnificent twin screw Transatlantic passenger steamers, we 
 have an example of two of the finest specimens of naval architecture afloat. 
 They belong to the Cunard Line, sailing regularly between Liverpool and 
 New York, and were built in 1893, by the Fairfield Shipbuilding and 
 Engineering Co. Ltd., Glasgow. Particulars of their principal dimensions, 
 etc., are as follows : — 
 Length over all, . 
 
 „ between perpendiculars, . 
 Breadth extreme. 
 Depth moulded to upper deck, . 
 ,, to boat deck, 
 
 About 14 depths in length. 
 JEngines — Twin Screw. Ten cylinders. 
 4 high-pressure cylinders, 
 2 intermediate „ . . . 
 
 4 low pressure ,, 
 
 Stroke ..... 
 Boilers — 12 double ended, 17' 6" diameter x 18' 9" long. 
 1 single „ „ „ 
 
 1 donkey boiler, 11' , 
 
 Natural DraugJtt. 
 
 Total number of furnaces, 102. 
 Working steam pressure, 165 lbs. per square inch. 
 Indicated horse power on trial, 31,000. 
 Speed on trial, 23-18 knots. 
 
 Average speed per hour on fastest passage, 22-01 knots. 
 Gross tonnage = 12,950 tons. Nett register tonnage, 4973. 
 Passenger accommodation — 
 
 
 
 622' 6" 
 
 
 
 600' 0" 
 
 
 
 65' 3" 
 
 
 
 41' 10" 
 
 
 
 59' 6" 
 
 
 
 37 in. diam 
 
 
 
 79 „ 
 
 
 
 98 „ 
 
 
 
 69 „ 
 
 18' 
 
 9' 
 
 long. 
 
 10' 
 
 9' 
 
 )> 
 
 9' 
 
 3' 
 
 )> 
 
 First Class, 600. 
 Second Class, 400. 
 
 Third Class, 1000. 
 Crew, 450. 
 
Fig 67. 
 ROYAL MAIL TWIN-SCREW STEAMER 'CAMPANIA' AND 'LUCANI/ 
 
TYPES OF VESSELS — SECTION 11. 1 33 
 
 Most of the structural features are distinctly shown in diagrams 57 and 
 58, wliicli have been kindly furnished for this work by the Owners ; hence, 
 more than briefly drawing attention to the chief items in the construction 
 is unnecessary. 
 
 The vessels are classed 100 Al at Lloyd's. They fully comply with the 
 rales of the Bulkhead Committee, in being subdivided into watertight com- 
 partments by transverse watertight bulkheads, in such a manner that any two 
 (and in some cases, even three) of these compartments may be flooded with- 
 out endangering the safety of the vessel. The bulkheads are spaced more 
 closely towards the ends of the vessel Avhere loss of buoyancy would cause 
 most change of trim, and where the danger from collision is greatest. The 
 bulkheads are stiffened by deep channel bars, and connected to the tank top 
 and decks by bracket plates. Where passage-ways are required through 
 these bulkheads, watertight doors are fitted, but these are of the least 
 possible ntmiber. 
 
 Unlike some large passenger steamers, no central longitudinal bulkhead 
 is fitted, excepting through the engine room. Longitudinal bulkheads, 
 apart from structural strength, may be a source of danger to a vessel 
 under some circumstances, for in the event of the shell being perforated and 
 the sea finding access, the loss of buoyancy on the one side only may 
 cause the vessel to take a very considerable list. The necessity for the 
 longitudinal bulkhead through the engine room, however, is so apparent 
 as to need little comment. The two engines driving the two propellers 
 are kept in separate watertight compartments, and thus, in the event 
 of one of them being flooded, the other remains intact, and the 
 vessel has the advantage of propulsion by means of one propeller, at 
 any rate. 
 
 The ' Campania ' and ' Lucania ' are on the Admiralty list of mercantile 
 armed cruisers, and hence satisfy all requirements for such. Among other 
 features, the boiler compartments are protected against shot at the sides 
 and over the top by watertight coal bunkers. A cellular double bottom 
 extends continuously throughout the whole length. The keel is of the 
 flat plate type, 60 in. broad by over an inch in thickness (see midship 
 section, fig. 58). It has a doubling on the inside of over fths of an inch 
 in thickness. The double bottom is about 56 in. in depth, excepting under 
 the engines, Avhere it is increased in depth to about 7 ft. 6 in., in order to 
 form the engine seating. The centre through-plate is continuous all fore 
 and aft with double angles at top and bottom. Between the centre 
 through-plate and the tank margin plate, there is a continuous longitudinal 
 girder, also with double angles at the top and bottom. Between this last 
 mentioned continuous girder and the centre through-plate, and also between 
 the same girder and the margin plate, intercostal girders are fitted as shown 
 in fig. 58. Under the engines, however, where much greater strength and 
 rigidity is required owing to the enormous power of the engines, two con- 
 tinuous and three intercostal girders are fitted on each side as shown in the 
 
134 STEEL SHIPS. 
 
 section of the machinery sjDace, fig. 58. The frames are of channel bar 
 section, spaced 30 in. apart. 
 
 In order to give increased stiffness to the transverse framing, there are 
 introduced at intervals, in addition to the numerous transverse bulkheads, 
 deep web frames or partial bulkheads. These web frames are spaced more 
 closely together in the engine and boiler space, where special precautions 
 are obviously necessary. 
 
 There are four complete steel decks laid throughout the entire length of 
 the ship, viz. : — Promenade deck, upper deck, main deck, and lower deck, 
 all of which are wood sheathed. The boat and orlop decks have wide 
 stringer and tie plates, and are laid with wood decks. The beams to ail 
 the decks are of the butterly tee bulb section, with the exception of those 
 at the ends of the promenade deck, which are bulb angle. The promenade 
 deck amidships is supported on T section steel stanchions, which are 
 brought down to the sheer strake, and form stififeners for the bulwarks 
 (see midship section, fig. 58). 
 
 The frames extend continiiously from the tank margin plate to the upper 
 deck. The centre through-plate and tank margin plate and some of the 
 side girders being continuous fore and aft, the floors must be inter- 
 costal between these longitudinals. At the fore end, where, on account 
 of the high speed, severe panting stresses are encountered, the frames are 
 doubled for a considerable distance abaft of the stem. Doublings are also 
 fitted to the frames for a considerable distance forward of the stern post, 
 where great stiffness is required to carry the propeller brackets through 
 which severe strain is transmitted, especially when the vessel travels at a 
 high speed. Double angle bar frames are fitted to all watertight bulkheads. 
 
 In all steam vessels, owing to interruptions in transverse strength caused 
 by the omission and cutting of beams in the machinery space, too much 
 attention cannot be given to the matter of compensation for such weakening 
 of the structure, and to so strengthening this locality that no signs of 
 weakness may be developed, and vibration reduced to a minimum. It is 
 also necessary to ensure that the foundation upon which the engines and 
 boilers rest is of the most substantial character. In such vessels as the 
 'Campania' and ' Lucania,' with their enormous engines developing 30,000 
 indicated horse power, the strengthening of the machinery space naturally 
 became one of the most important subjects in the construction of these 
 vessels. 
 
 As we have stated, the floors under the engines were increased con- 
 siderably in depth, and additional longitudinal girders were introduced. 
 
 The section through the engine space (fig. 58) well illustrates this. 
 In order that the various parts might be connected in the most efficient 
 manner, all rivet holes in the structure under the engines were rimed and 
 made perfectly true, and the angles in the double bottom Avere welded 
 and joggled, so as to thoroughly combine the whole and make the 
 transverse framing as continuous as possible. 
 
Fig. 58. 
 MIDSHIP SECTION, 'CAMPANIA' AND 'LUCANIA.' 
 
TYPES OF VESSELS — SECTION II. 
 
 135 
 
 As the orlop beams had necessarily to be dispensed with in the 
 machinery space, compensation was provided by a semi-box stringer 
 (see section, fig. 58), which is itself furthermore strengthened by the 
 introduction of diaphragm plates at intervals. 
 
 The stringer plates on beams are connected to the shell by double 
 intercostal angles instead of by a single angle as commonly adopted. 
 
 The shell plating is arranged on the overlap system from the keel to the 
 under side of the main deck sheer strake. 
 
 The butts of the upper and main deck sheer strakes, and the strake of 
 shell plating between, are connected by double butt straps quadruple 
 riveted over the middle length and treble riveted at the ends. The straps 
 are placed one inside, and one out. All the other shell butts are overlapped 
 and quadruple riveted over the midship length, and treble riveted before 
 and abaft this. As shown in the section, fig. 58, the upper deck sheer 
 
 This row countersunk 
 
 (p.cp. ([)(t)0(P(l)c|)(|)cl)(:p - 
 
 This row (Countersunk — 
 
 Snap rivets only shown 
 
 J aside View 
 
 Fig. 58a. — Detail of H3'draulic Riveting in Upper Deck Sheer Strake. 
 
 strake and the strake below are doubled over the middle length of the 
 vessel. In long vessels it is a common practice to double one or more 
 strakes of bilge plating, but in the case of the ' Campania ' and ' Lucania,' 
 increased strength was obtained by increasing the thickness of the bilge 
 strakes, and adopting treble riveting at the edges or seams of these 
 strakes. 
 
 In order to obtain the fullest efficiency, and to ensure the soundest 
 workmanship, hand riveting was dispensed with in the connection of the 
 upper deck sheer strake to the deck stringer plate, as it was practically 
 impossible to obtain satisfactory workmanship by hand riveting so many 
 plates of great thickness. (See fig. 58a.) 
 
 Hydraulic riveting was therefore adopted, and, as a consequence, the 
 rivets are snap-headed. Though the exposed heads of such rivets may 
 be considered by some to detract a little from the general appearance, 
 
136 STEEL SHIPS. 
 
 this consideration was outweighed by the vastly improved character of 
 the work, and the satisfaction felt iu having the various thicknesses of 
 plating thoroughly closed up. All rivets are of steel, and in way of the 
 double bottom, the riveting is all done by hydraulic machinery. The shell 
 plates range from | in. to 1 in. in thickness. 
 
 The deck houses are of the composite type, having steel coamings, and 
 frames, and beams (see fig. 169). 
 
 The upper deck is sheathed with teak, and the others with yellow pine. 
 The stern frames and propeller brackets are of cast steel. 
 
 By' adopting the system of bossing the frames at the after end (see 
 figs. 157, 158), not only is thorough support given to the propeller shafts, 
 but ready means of access is obtained to the shafts and their bearings 
 throughout the whole length excepting the tail shafts. 
 
 When the ships were built, they were fitted with plate rudders having 
 specially designed arms on either side. The plates for the rudders were made 
 by Messrs. Krupp, of Essen, in Germany, and measured 22' 0" x 11' 6" x 1^" 
 thick. The necessity of going abroad for these plates was afterwards called 
 in question by British makers, and the matter was severely commented 
 upon at the time by some of the leading journals. However, after the 
 ships had run a few voyages, the plates were found to be cracked, and the 
 owners refitted the ships with new cast-steel rudders made up in three 
 sections. 
 
 'GREAT EASTERN' 
 (See figs. 59 and 60). 
 
 While the 'Campania' and 'Lucania' are two of the largest vessels 
 afloat, their size was considerably exceeded by Scott Russell's great work, 
 the famous ' Great Eastern,' and a few particulars about this vessel, to- 
 gether with the profile and midship section, figs. 59 and 60, will doubtless 
 be interesting, if not instructive, for comparison.* 
 
 The ' Great Eastern ' was 692 ft. in length over all. 
 
 Length between the perpendiculars, 680 ft. 
 
 Breadth moulded, 82 ft. 2 in. 
 
 Depth moulded to upper deck, 58 ft. in. 
 
 Depths in length, about 11|^. 
 
 This vessel was propelled by a screw propeller, and also by a pair of 
 paddle-wheels. The total indicated horse-power was about 8000, the screw- 
 engines being somewhat more powerful than the paddle, indicating about 
 4500 horse-power. 
 
 The maximum speed obtained was from 14 to 14J knots per hour. 
 There were ten tubular rectangular boilers 18 ft. long, 17 ft. 6 in. wide, and 
 
 * The * Great Eastern ' has, however, since been exceeded in size by the White Star 
 steamer ' Oceanic,' built by Messrs. Harlan d & Woolf, 
 
Fig. 59. 
 SECTIONAL PROFILE OF 'GREAT EASTERN.' 
 
TYPES OF VESSELS SECTION 11. I 37 
 
 14 ft. high, with 112 furnaces in all, and a working steam-pressure of 
 
 about 20 lbs. 
 
 Each set of engines had four cylinders witli a diameter and stroke of 
 7 ft. and 4 ft. respectively in the screw-engines, and 6 ft. 2 in. and 14 ft. 
 respectively in the paddle engines. 
 
 The propeller had 4 blades with a diameter of 24 ft. with 44 ft. 
 pitch. The paddle-wheels were about 56 ft. in diameter. The vessel was 
 designed to carry about 800 first class, 2000 second class, and 1200 third 
 class passengers. It has sometimes been stated that the ' Great Eastern ' 
 was a huge failure. If being far ahead of the times constitutes a failure, 
 she was one. Certainly she was never profitable commercially, but she 
 was an advance in naval architecture such as has never been equalled 
 since, and her designer (Brunei) must ever rank as one of the ablest and 
 most ingenious naval architects that ever lived. 
 
 The fact is apparently often overlooked, that the experience provided 
 by the ' Great Eastern ' has directly, or indirectly, been of immense value to 
 modern naval architects, and in no small way has contributed to the success 
 of our largest ocean liners. 
 
 The system of her construction was unique, and although it has not been 
 rigidly folloAved in the building of succeeding mercantile vessels, yet it has 
 doubtless influenced the construction of ships generally. 
 
 She was built exclusively upon a system of longitudinal framing which 
 was composed of plate girders spaced about 30 in. apart from the keel 
 to the lower turn of the bilge, and from thence to the lowest deck at about 
 twice this distance. (See fig. 60.) 
 
 These girders w^ere entirely covered with an inner skin of watertight 
 plating in addition to the outer shell. Such an arrangement not only added 
 immensely to the strength, but contributed in a great measure to the safety 
 of the vessel, and but for which, on at least one occasion, she might have 
 had a shorter career. The longitudinal framing received great support and 
 assistance from numerous watertight transverse bulkheads. 
 
 Unlike modern shipbuilding practice, with its multiplicity of sizes 
 and thicknesses of plates and bars, the inner and outer shell plating of the 
 * Great Eastern ' was | in. in thickness ; the longitudinal plate girders, 
 ^ in. in thickness ; the rivets, |^ in. in diameter spaced 3 in. apart ; 
 and the angles, 4 in. x 4 in. x f in. 
 
 While the beams to both the uppermost and lowermost decks are of 
 exceptional strength, as shown by figs. 59 and 60, the upper deck especially 
 is exceedingly strong. 
 
 In addition to the strong upper deck beams, further stiffness and rigidity 
 is given by intercostal plates fitted between them longitudinally, and both 
 the upper and lower surfaces of the beams are plated, making in their 
 entirety a double iron deck. 
 
 A general idea of the construction of this marvellous vessel can be 
 obtained from the profile and midship section given. 
 
138 STEEL SHIPS. 
 
 TUREET DECK CARGO STEAMER (see figs. 62, 63, 64, and 65). 
 
 This novel type of cargo steamer first appeared in about the year 1891, 
 and the very fact that since that time over sixty of these vessels have been 
 built, some of them reaching to 450 ft. in length, and also that owners 
 who have had vessels of this type built have in many cases given farther 
 orders for new vessels of the same kind, proves that ' Turrets ' have shown 
 themselves at least equal, if not decidedly preferable, to the ordinary type 
 of cargo steamers for certain trades. 
 
 The first turret steamer was designed and built in the shipyard of 
 the patentees, Messrs. W. Doxford & Sons, Sunderland. Though very 
 different and vastly superior to the notorious whaleback steamer, it is 
 probable that the idea of the turret steamer was first conceived from this 
 vessel. As is now well known, the whaleback steamer was an American 
 invention, and although it was prophesied that this peculiar vessel would 
 revolutionize ship construction, yet the first visit of the first whaleback 
 steamer, the ' Charles W. Wetmore,' across the Atlantic to this country in 
 1891 was also her last, and she has had remarkably few successors. Fig. 
 61 shows the main outline and form of a whaleback steamer. The main 
 object in the design was to provide a sea-going vessel with absolutely clear 
 decks, so that if heavy seas broke aboard, there would be no deck erections 
 to carry away, while the peculiar form of the rounded deck would break 
 the force of the sea, and allow it to escape over the sides again above the deck. 
 
 The great difficulty of getting along such a weather deck in heavy 
 weather is obvious, the only means facilitating this being a gangway above* 
 the deck supported upon turrets at intervals in the length. 
 
 The hatchways are simply holes in the deck, having no coamings, and 
 closed by means of plates which are bolted through the deck. This is 
 objectionable, as there are no means of feeding the holds when carrying bulk 
 cargoes. 
 
 The spoon-shaped bow and the form of the bottom, from the stem for 
 some distance aft, makes the hull specially liable to damage when the 
 vessel is pitching in a seaway, owing to the ponuding action produced 
 as the vessel falls into the seaway. 
 
 In one or two features, the turret steamer is similar to the whaleback. 
 Neither vessels have any fore and aft sheer, and the gunwale is of rounded 
 form in each case. But beyond these two features they are widely different, 
 as is seen by comparing the diagrams of turret steamers, figs. 62 to 65, 
 with the whaleback steamer, fig. 61. 
 
 The following are the principal features in the design and construction 
 of the turret deck steamer. 
 
 First of all, it may be noted that the form of the vessel below the load 
 water-line is similar to that of any ordinary steamer, the points of difference 
 being above the load line. 
 
UJ 
 
 < 
 
 LJ 
 I- 
 (D 
 
 o ^ 
 o s 
 
 O ^ 
 
 O :. 
 
 UJ '^ 
 
 Q ^ 
 
 h- 
 
 cc 
 en 
 
 D 
 I- 
 
Fig. 60. 
 GREAT EASTERN' MIDSHIP SECTION. 
 
 r- ^ 1 "' T"*T^ ^ f ' 1 ' f 
 
TYPES OF VESSELS SECTION II. 
 
 139 
 
 The most striking feature is the turret 
 erection, which extends continuously from stem 
 to stern. The sides of the turret are blended 
 into the harbour deck, and the harbour deck 
 into the vertical side plating, by well-rounded 
 corners ; see figs. 64 and 65. 
 
 All hatches, deck openings, and deck houses 
 are placed on the turret deck, which at sea is 
 the working deck of the vessel. The harbour 
 deck is practically clear of deck fittings with 
 the exception of a storm rail for purposes of 
 protection of crew while working mooring bits 
 and fairleaders and watertight coaling ports. 
 As shown in fig. 62, the accommodation for 
 officers and crew is on the turret. 
 
 As to the strength of these vessels, the very 
 fact that they have been given the highest class 
 by the principal classification societies, leaves no 
 doubt as to their structural merits. Indeed, 
 when the fact is considered that the frames 
 extend continuously from the margin plate of 
 the inner bottom up the vertical side, round the 
 gunwale and harbour deck, and up the turret 
 side to the turret deck, and also that the thick- 
 ness of the shell plating is maintained round 
 the gunwale and harbour deck to the turret 
 deck, it is at once apparent that such an erection 
 as the turret, carried continuously all fore and 
 aft, and blended so completely into the main 
 hull of the vessel, must afford a most valuable 
 addition to the longitudinal strength in resisting 
 longitudinal bending. And although the depth 
 of the vessel is increased by the turret erection, 
 yet the transverse form of this upper part, and 
 the assistance given by the thick harbour deck 
 (which in ordinary vessels is usually of com- 
 paratively thin plating) and the plate brackets 
 in the turret, between the turret deck and 
 harbour deck beams, prevents any chance of 
 transverse weakness. The turret sides and 
 harbour deck really form a pair of huge angle 
 girders, one on each side, extending all fore and 
 aft. 
 
 Owing to the increased stifi"ness which the 
 curved sides give to the deck and the strength 
 
 d 
 
 J 
 
 >jl 
 
 eu 
 
 c/) 
 
140 ^ .STEEL SHIPS. 
 
 afforded by the large angle girders just referred to, it is quite admissible 
 that the hold pillars, if of increased strength, should be more widely spaced 
 than is usual in ordinary vessels. The advantage thus given for better 
 stowage is evident. The pillars are usually made of two channel bars 
 riveted back to back, and bracketed at their upper and lower extremities 
 to the deck beams and inner bottom plating. 
 
 The turret erection being of a thoroughly substantial character, it is 
 needless to say, in making the strength calculation, that the sectional area 
 of all continuous longitudinal material up to the turret deck is used in 
 arriving at the moment of inertia, hence, with an increased depth of girder, 
 the resulting large moment of inertia is accounted for. At the same 
 time, there being no sheer, and, consequently, relatively less hold space, 
 towards the ends of the vessel, a reduction in bending moment would be the 
 result, were there no turret erection suitable for carrying cargo. Any actual 
 reduction, therefore, in bending moment, as compared with an ordinary 
 vessel, may be ignored, but with increased moment of inertia, it follows that 
 a reduction in the stress per square inch on the topsides may result. 
 
 These vessels are usually built with a cellular double bottom all fore 
 and aft, which is of the ordinary construction. The frames are generally of 
 channel or Z section. 
 
 Other advantages possessed by the turret steamer may be enumerated 
 as follows : — 
 
 There is no possible means of heavy shipped seas finding lodgment on 
 either the turret or the harbour decks, as no closed bulwarks are fitted to 
 any part of the vessel. The turret erection is an ideal feeder to the lower 
 holds in vessels carrying grain and other bulk cargoes, and therefore the danger 
 of a cargo shifting, which is of such frequent occurrence in ordinary vessels, 
 is rendered impossible in the lower holds, while any shift of cargo that 
 might take place in the turret erection itself would have so little effect 
 upon the stability of the vessel as to be unnoteworthy. The rounded form 
 of the gunwale also contributes to the better stowage of these cargoes. By 
 the wide spacing of the hold pillars, a further important advantage is 
 obtained in the matter of stowage. The height of the turret deck (10 to 12 
 ft. above the load line) on which are situated all the hatches and ventilators, 
 also the engine and boiler casings, companion-ways, etc., is a source of 
 safety and protection to these important parts. 
 
 It will be noticed, in looking at the midship sections, figs. 64 and 65, 
 that as far as possible, all connections of plates to one another have been 
 effected by flanging or joggling, thereby dispensing with the usual connecting 
 angle bar and packing. The midship sections show that the floor plates 
 are flanged as well as also the intercostal girders in the double bottom at 
 their upper and lower edges. The edges of the shell plating and inner 
 bottom plating are joggled. By adopting this system, all the plating bears 
 close upon the surface of the material to which it is attached. This 
 confers the great advantage of improving the efficiency of the riveting by 
 
Fig. 63. 
 
 turret steamer with one laid deck, having special arrangement for 
 making oversea voyages in ballast (water). 
 
 .--p- 
 
 ^r= 
 
 I i-tj:^nti 1 1 ! 1 1 1' 1 1 1 1 1 1 1 1 1 iT£ki 1 1 fi 11 i^krt rrri rrrn 1 1 1 ri l-^cittti i 
 
 dlspace --=^°\ 
 
 and 
 Waier Sfllast 
 
 ,3Webs ^ 
 
 Ballast Water Line 
 
Fig 62. 
 TURRET STEAMER WITH TWO DECKS. 
 
Fig. 64. 
 
 MIDSHIP SECTION OF TURRET STEAMER WITH 
 TWO DECKS. 
 
TYPES OF VESSELS SECTION II. I4I 
 
 having less thicknesses to rivet, not to mentiou the great saving by 
 relieving the vessel of the weight ot packing. Owing to the strength 
 aftbrded by the sides of the turret, very long hatchways may be made in 
 these vessels, which is an especial advantage in the loading of certain 
 kinds of cargo. 
 
 The additional depth and strength provided by the continuous turret 
 erection makes less necessary any addition to structural strength, on accoiint 
 of the proportion of length to depth. 
 
 The turret deck is fitted with the usual loading and discharging gear — • 
 winches, derricks, etc, — and following the general tendency in these days,, 
 the masts are short, and make no pretence of carrying much sail. Indeed, 
 their chief function is to support the derricks. Objection has been lodged 
 against tlie turret steamer because of its un-shiplike appearance, but the 
 turret vessel is essentially a cargo steamer, and a deadweight carrier.. 
 Moreover, it only resembles, in a more advanced form, the tendency of the 
 ordinary cargo vessel, which, by its full formed hull, shorter masts, and 
 conspicuous smallness of sail, together with its elaborate discharging and 
 loading geai', is fast losing all the features which originally conveyed the 
 idea of a ship. However, as cargo steamers are built to earn freio-hts, 
 any argument against appearance is unreasonable, so long as a satisfactory 
 condition of seaworthiness is assured. Figs. 62 and 64 give a profile 
 and midship section showing tlie general and structural arrangement of 
 a two-decked turret steamer, and figs, 63 and 65 are similar illustrations of 
 a single-decked turret steamer. 
 
 Recognizing the importance of making provision for the carriage of 
 ballast when making long over-sea voyages without cargo, which is so 
 usual an occurrence in these days of keen competition, fig. 63 shows an 
 arrangement for carrying water-ballast proposed by the patentees of the 
 vessel, whereby the draught is increased from the light line shown to a 
 deeper water-line allowing considerably greater immersion and trimming 
 the vessel reasonably by the stern, iu order to secure the most favourable 
 conditions for the propeller. Of course, where a huge, deep tank is fitted, 
 such as shown amidships in fig, 63, the bulkheads bounding such tanks 
 require to be most efficiently stiffened, and in the profile are shown the 
 deep vertical webs which the builders fit to these bulkheads, in addition to 
 other stiffening. 
 
 The space on the harbour deck at the sides of the turret erection lends 
 itself and is easily utilized for the carriage of deck cargoes, such as 
 timber, etc, 
 
 "While the freeboards of all ordinary ships which fully comply with the 
 structural requirements of their respective types as laid down by the Board 
 of Trade standard may be deternuned on a minimum percentage of reserve 
 buoyanc}', exclusive of erections which are treated and allowed for separately, 
 the ' turret ' and other self-trimming steamers may be treated in a like 
 manner, for although they may differ in design, they still belong to full 
 
142 STEEL SHIPS. 
 
 scantling or spar deck types, or some modification of them, in respect to 
 strength. 
 
 It is patent to every intelligent observer that valuable as the turret erection 
 is in affording strength and buoyancy, yet the buoyancy of the turret erection 
 could not reasonably receive the same credit in the matter of freeboard as 
 the same amount of buoyancy distributed over the whole width of the 
 harbour deck as in an ordinary vessel. Hence the Board of Trade Free- 
 board Tables and Instructions allow as much as 60 per cent, of such buoy- 
 ancy as being effective in determining the freeboard which is measured from 
 the turret deck, a fact in itself testifying to the value of the turret erection. 
 
 A few extracts from the Board of Trade Freeboard Kules regulating 
 the depth of loading of turret deck vessels and vessels of similar types, 
 will no doubt greatly assist in conveying a clear idea of the basis upon 
 which such vessels are treated in the assignment of their load-lines. 
 
 "A 'turret' is a strongly constructed continuous erection at the middle 
 line of a vessel, forming with the main or harbour deck an integral part of 
 the hull, and of a breadth not exceeding six-tenths the greatest breadth of 
 the vessel." (Should this breadth be exceeded, the case must be submitted 
 to the Board of Trade.) 
 
 " The freeboard, calculated according to the rule given, is in considera- 
 tion of the turret being continuous. Where the turret is separated within 
 the six-tenths length amidships, additional strength will be required to 
 compensate for this." 
 
 '* Hatch coamings at least 2 ft. high, and casings to engine and boiler 
 openings at least 4 ft. 6 in. high, to be fitted above the ' turret ' deck." 
 
 " The volume of the turret to be estimated from a normal beam line 
 drawn through the point where a vertical line at the quarter breadth of vessel 
 cuts the upper surface of the vessel's deck. "Where the turret is nearly one- 
 half the breadth of the vessel, and its transverse section is of rounded form 
 at its base, the base line of the turret to be drawn through the point where 
 the vertical line at the quarter breadth cuts the upper surface continued in 
 the same curve as the normal line of beam." 
 
 " The reserve buoyancy required by the Tables to be estimated by 
 taking 60 per cent, of the volume of the turret. The height of the turret 
 allowed for is not to exceed 25 per cent, of the moulded depth." 
 
 " The moulded depth of the vessel to be taken to be the depth at side 
 from the beam line, as before defined, to the top of the keel." 
 
 " The transverse and longitudinal strength of the vessel to be regulated 
 by that required for a ' three-deck ' or ' spar-deck ' vessel of the same length, 
 breadth, moulded depth, and coefiicient of fineness." 
 
 TRUNK STEAMERS. 
 
 Figs. 66, 67, 68, and 69 are plans illustrating in profile and midship 
 Section what is known as the 'Trunk ' steamer. 
 
 The inventors of this type of vessel are Messrs. Ropner & Son, the 
 
TYPES OF VESSELS — SECTION II. I 43 
 
 well known shipbuilders of Stockton-on-Tees. The trunk steamer is 
 essentially a cargo vessel which possesses special facilities for the stowage 
 of bulk cargoes, and advantages in the navigation of the vessel. On referring 
 to the midship section, fig. 67, it will be seen that, up to the gunwale, this 
 vessel is in every respect similar to the ordinary type of cargo vessel. She 
 has a cellular double bottom, and is framed on the web frame system, 
 thereby dispensing with a tier of hold beams, and leaving the whole space 
 perfectly clear, with the exception of the hold pillars, which are widely 
 spaced. Instead of web-frames, the deep-frame system may be adopted. 
 In addition, these vessels have a poop, bridge and forecastle constructed 
 in the usual way. 
 
 The special feature of this vessel is the trunk erection. This trunk is 
 strongly framed and plated, and can be made thoroughly substantial. It 
 extends from the poop to the bridge, and from the bridge to the forecastle, 
 into each of which the sides of the trunk are scarphed. The height of the 
 trunk is about 7 ft. — that is, the height of poop, bridge and forecastle — with 
 which erections the top of the trunk forms a continuous deck. The breadth 
 of the trunk is about half the beam of the vessel. Strong, through beams 
 are placed where necessary, at intervals at the base of the trunk (see figs. 66 
 and 67). These strong beams greatly assist in maintaining the transverse 
 strength. Apart from these strong beams, the trunk space is entirely open 
 to the lower hold, and, as the midship section shows, the ordinary main deck 
 beams are cut at the sides of the trunk. This apparent weakness, however, 
 is fully overcome by the strong connection made between the trunk sides, 
 the beams, and the pillars. (See fig. 67.) First of all, the deck plating 
 is flanged into the trunk side. The hold pillars are of channels, or tee 
 bars, fitted back to back, one of which bars extends some distance up the 
 trunk side plating to which it is riveted, and further, the pillars are 
 bracketed to the deck beams and deck plating, as shown. Thus the trunk 
 sides, being thoroughly connected to and scarphed into the poop, bridge 
 and forecastle, and supported as they are by extra strong hold pillars, 
 afford ample strength for carrying the beam ends and supporting the 
 main deck. 
 
 The trunk side frames, and the beams supporting the trunk deck, are all 
 of about the size of the main frames of the vessel, and connected by brackets 
 as shown in figs. 67, 68, and 69. 
 
 As all the principal hold hatchways are situated upon the trunk deck, the 
 trunk deck beams are necessarily cut at the hatch sides, but, as the midship 
 section shows, the hatch sides are well supported by large web plates fitted 
 between the trunk deck and main deck strong beams. The deck area of the 
 trunk deck is increased by extending the deck plating about 1 ft. beyond the 
 trunk sides, and this being practically the navigating deck, hand rails are 
 placed along each side for protection as shown. 
 
 The great advantage of having the navigating deck 9 or 1 ft. above the 
 load water-line is a feature which every seaman can appreciate. 
 
144 - STEEL SHIPS. 
 
 Moreover, as all hatchways, ventilators, and deck openings are situated 
 either on the trunk deck, or on the poop, bridge, or forecastle, the protection 
 which is afforded to these important parts, including the engine and boiler 
 openings, is very apparent. 
 
 In the loading of bulk cargoes, the trunk forms an excellent feeder to 
 the main holds in addition to possessing the facility for self trimming. The 
 main deck is either fitted with open rails, or closed bulwarks of the usual 
 form, as may be desired. Owing to the substantial nature of the hold 
 pillars, and the excellent support they give to the trunk side and main deck 
 beams, they are spaced at wide intervals apart, and by this means the 
 possibilities of broken stowage are minimised. The main deck outside of 
 the trunk is well adapted for the carrying of timber, cattle, or other deck 
 cargo. Fig. G8 shows the trunk of one of these vessels specially arranged 
 for coal cargoes, the sloping trunk sides farther facilitating the self-trimming 
 propensities of the vessel. In some cases, cargo skids or platforms at the 
 sides of each hatch are fitted, extending from the level of the trunk deck to 
 the sides of the vessel, for convenience in discharging cargo. Fig. 69 illus- 
 trates another special arrangement of trunk particularly designed for the 
 purpose of carrying rice cargoes. As rice cargoes are very liable to damage 
 owing to the vessel sweating, the diagram shows gutter-ways at the base 
 of the trunk (G in fig. 69), by means of which the moisture which drips 
 down the trunk sides is collected and led off by scuppers down into the 
 bilges. 
 
 The cargo ports, P, shown on the sides of the trunk are sometimes 
 fitted for the purpose of facilitating either the loading or unloading of 
 such cargoes as can be readily passed in or out of such ports. They also 
 serve the useful purpose of assisting ventilation. The other principal 
 structural features of this vessel are clearly illustrated in the accompanying 
 diagrams. 
 
 PEIESTMAN'S SELF-TRIMMING STEAMER. 
 
 Fig. 70 is a profile, and fig. 71 a midship section illustrating another 
 type of self-trimming steamer. The design is the patent of Messrs. 
 J. Priestman & Co. of Sunderland. The principal feature is an erection 
 about 5 ft. in height extending from stem to stern. The top breadth of 
 this erection is about half the moulded breadth of the vessel, and it forms 
 the navigating deck. The hull from the keel up to the gunwale is of 
 the ordinary form and construction. The diagrams illustrating this vessel 
 show her to be built upon the web-frame system, thereby dispensing with 
 hold beams, but if preferred, deep framing might be adopted instead, 
 or the ordinary frame and reverse bar (or some equivalent) together with 
 widely -spaced hold beams. But as it is necessary for trimming purposes 
 that the erection be in free communication with the main hold, the 
 transverse strength of the vessel in way of the main deck is maintained 
 
Fig. 65. 
 
 MIDSHIP SECTION OF A TURRET STEAMER WITH ONE 
 DECK AND A TIER OF WIDELYSPACED HOLD BEAMS. 
 
Fig. 66. 
 TRUNK STEAMER. 
 
 Dimensions — 
 Langlh B.P., 800 ft. Breadth « 10., 4S ft. 
 
 Depth of Hold, HO ft. 5 ins. Depth H lO, H ft. S i 
 
Fig. 67. 
 
 MIDSHIP SECTION OF TRUNK 
 
 STEAMER. 
 
 Dimensions— 300 ft. by 46 ft. IttiO. by 22 ft. 2 ins. Mi-D. 
 
Fig. 68. 
 
 TRUNK 0€CK 
 
 TRUNK \ FOR 
 COAL CARGOES 
 
 STRONQ BZmS RTINTERVHLS 
 
Fig. 69. 
 
Fig. 72. 
 ILLUSTRATING SOME NEW FEATURES IN MODERN SHIPBUILDING. 
 
 As described in pages 146-152. 
 
Fig. 73. 
 midship section of vessel illustrated in fig- 72. 
 

TYPES OF VESSELS — SECTION II. 1 45 
 
 by means of widely-spaced bulb plate beams well kneed to the Aveb frames, 
 and each supported by two hold pillars formed of double channel bars 
 fitted back to back, which are bracketed at the top to the beams, and at 
 the bottom to the tank top-plating. 
 
 The sides of the erection slope from the navigation deck to 
 about 2 ft. from the gunwale, in order to increase the facility for self- 
 trimming. The main frames from the tank side to the gunwale are 
 channel bars. 
 
 The self-trimming erection also is framed by means of channel bars 
 which are kneed on to the main frames, and extend continuously from one 
 side to the other, excepting in way of the hatches. Support is afforded to 
 the sloping sides of the erection by means of large bracket plates on every 
 strong beam. (See fig. 71.) 
 
 The navigation deck is supported by large web plates fitted between 
 this deck and strong main deck beams immediately above each hold pillar, 
 (See fig. 71.) It is also further supported by means of large bracket 
 plates placed in a fore and aft direction at the angle of the navigation 
 deck with the sloping sides. (See the profile, fig. 70, and the section, fig. 
 71.) All hatches, ventilators, deck openings, including the engine and boiler 
 casings, are upon the navigation deck, and thus considerably more elevated 
 than they would be in a vessel of the ordinary type. The erection forms a 
 natural feeder to the hold, with which it is in free communication, and is 
 therefore specially adapted for the carriage of bulk cargoes. The sloping 
 sides of the erection obviously greatly facilitate the self-trimming of such 
 cargoes. An ordinary open rail is fitted round the main deck for purposes 
 of protecting the crew. As the sloping side of the erection offers little re- 
 sistance to the inroad of beam seas which are liable to sweep over the vessel, 
 a closed iron bulwark is fitted to the navigation deck, well stayed by means 
 of bars of tee iron section or something equivalent. This not only affords 
 protection to the seamen in their duties in navigating the vessel, but keeps 
 the upper deck drier than it otherwise would be. The erection being 
 continuous from stem to stern with the top and sides blended into the main 
 hull by the stringer plate being flanged as shown in the midship section, 
 and the erection side plating flanged on to the navigation deck, there is no 
 doubt as to the substantial nature of the self-trimming erection. Further, 
 the hold pillars being widely spaced, the possibility of broken stowage of 
 cargo is greatly reduced. 
 
 While the diagrams show this vessel to be built with a M'Intyre 
 double bottom all fore and aft, excepting under the engines and boilers 
 where ordinary floors are fitted, a cellular double bottom might have been 
 adopted equally well had such been preferred. The sides of the self- 
 trimming erection are specially arranged for the carriage of timber on deck ; 
 special provision for securely lashing the same from side to side of the 
 vessel is provided by bar iron posts bracketed to the deck, spaced at 
 intervals throughout the length as shown in the profile. 
 
 10 
 
146 STEEL SHIPS. 
 
 Some New Features in Modem Shipbuilding. 
 
 "While several smaller changes and improvements have taken place in 
 recent years in many of the details of ship construction, probably tbe 
 greatest and most startling of these changes, and, it may be truly said, 
 improvements, have taken place in a number of vessels designed by 
 Mr. H. B. Wortley for the ' Ocean Steamship Company ' of Liverpool, 
 managed by Mr. Alfred Holt. When first these innovations were pro- 
 pounded, they were looked at in consternation by many experts, but 
 the fact that a number of these vessels have been built, and have sailed 
 several round voyages to China and back in all weathers and seasons, and 
 have had unstinted praise lavished upon them by the men who have 
 handled and manoeuvred them, in addition to the compliment that some 
 of the features have so commended themselves to practical shipowners 
 that they have reproduced them in their new vessels, proves that the 
 advance is in the right direction. The principal innovations referred to 
 are as follows : — 
 1st. No sheer. 
 
 2nd. More widely spaced transverse framing. 
 
 3rd. The substitution in the main hold of two pillars instead of the 
 usual numerous stanchions, and a modification of the same system in the 
 other holds. 
 
 4th. The usual form of stern frame for single screw vessels dispensed 
 with, thereby enabling the shipbuilder to get rid of much useless material 
 (and hence weight) in that locality known as the dead-wood in wooden 
 vessels, at the same time improving the steering qualities, while a special 
 arrangement has been introduced to carry the rudder, which also is of 
 special design. 
 
 5th. Cement. The ^bottom of the vessel inside the ballast tanks is 
 not cemented. 
 
 Eeference to figs. 72, 73, 74, and 75, graphically illustrates most of the 
 points just mentioned. It is advisable, however, to look more particularly 
 into the features or innovations just enumerated in order to determine 
 more fully what results are entailed in their adoption. 
 
 1st. Sheer. — That sheer is not given to a vessel for purposes of strength 
 should be clearly understood, for while increased depth over the midship 
 length undoubtedly contributes to increased effective longitudinal strength, 
 yet the increased depth which sheer gives towards the extremities is utterly 
 useless for this purpose, as the longitudinal bending moments decrease 
 from midships towards the ends. 
 
 The value of sheer lies in the increased reserve buoyancy which it 
 provides towards the ends of a vessel, affording increased lifting power, 
 when she is labouring among head or following seas, which in turn means 
 a drier deck, and greater protection to the crew. Consequently vessels with 
 less than the Board of Trade standard sheer are penalized by having 
 
TYPES OF VESSELS — SECTION II. 
 
 Coaming 
 
 147 
 
 Upper 
 kL, Deck 
 
 
 
 Main Deck 
 
 HalcH" 
 Coaminq 
 
 
 7 
 1 
 
 r 
 L 
 
 Section through Main Hold Hatch 
 Side in way of Pillars. 
 
 Plan of Top 
 
 ^ -^1 
 
 Plan of Bottom 
 
 Top and Bottom of Pillars in After 
 Main Hold. 
 
 I 
 
 i ■■ : -l-\ 
 
 JL 
 
 Section at A B (see fig. 72). 
 
 nzizi 
 
 Section at C D (see fig. 72). 
 
 Tank top. 
 
 Section through Stool under 
 Main Hold Pillars. 
 
 Fig. 74. — Detail Sketches of Vessel illustrated in fig. 72. 
 
T48 STEEL SHIPS. 
 
 the freeboard increased, while vessels with sheer in excess of the standard 
 obtain a reduction in freeboard. Spar and awning deck vessels are exceptions 
 to this rule. 
 
 Now if by any other means a vessel can be provided with ample reserve 
 buoyancy at her ends, the same purpose for which sheer is given will be 
 served, and it will be better from a structural point of view to take the 
 increased depth which a certain sheer would liave provided, and add it 
 uniformly all fore and aft to the depth of the ship. By this means not 
 only is the total amount of reserve buoyancy above the upper deck 
 maintained, but, as previously pointed out, the longitudinal strength of the 
 ship is effectively increased. 
 
 In these special vessels with which we are dealing, the necessary reserve 
 buoyancy at the ends is abundantly supplied by means of a high poop and 
 forecastle (about 9 ft.), which cover a considerable length at each end. 
 Sheer is therefore unnecessary. 
 
 In addition to the aforementioned erections, a well-constructed and 
 efficient bridge covers a considerable portion of the middle length, including 
 the engine and boiler openings. In order to avoid the appearance of 
 being 'hog-backed,' which might result from a horizontal deck, a sheer 
 is given to the bulwark line only, the sight edge of which is continued 
 from stem to stern. 
 
 ^nd. Widely-spaced TvMisverse Framing. — Whether a load line be 
 assigned to a British vessel by a registration society or the Board of 
 Trade, as previously stated, she must come undsr the test of Lloyd's 1885 
 Rules. It is therefore reasonable to expect that if the transverse frames 
 be more widely spaced than are specified in these rules for any particular 
 size of vessel, compensation, or, in other words, equivalent strength must 
 be introduced, and this is done in the vessels we are describing in the 
 following manner : — 
 
 First. — As the transverse frames are spaced 36 in. apart instead of 
 27 in. — the maximum given in Lloyd's present Rules — the required 
 transverse stiffness is obtained by increasing the depth of the framing ; 
 and as these ships are biiilt on the deep frame principle, this is easily done 
 by increasing the depth of the athwartship flanges of both the frame and 
 the reversed bar. As fully described in Chapter V., in dealing with 
 the strength of girders, an increase in the depth of the deep framing not 
 only increases the sectional area, but the moment of inertia also in a 
 greater proportion than the sectional area. So that although the individual 
 frames are heavier, their decrease in number ought to result in a lighter 
 total weight of framing. 
 
 Second. — If widely spaced transverse frames be adopted in combination 
 with the ordinary longitudinal framing, greater areas of unsupported shell 
 plating would exist over the whole of the immersed surface of the hull, and 
 the tendency would be for the shell plating to fall hollow, owing to the 
 enormous pressure to which these areas would be subject. Something must 
 
TYPES 0B\ VESSELS SECTION II. 1 49 
 
 therefore be doue to provide the necessary rigidity. This can be effected by 
 somewhat reducing the sectional area of the large cumbersome side stringers 
 required by rule, and compensating for this reduction by fitting additional 
 side stringers of a more compact form. In this way the longitudinal 
 strength would be preserved, while, at the same time, such strength would 
 be more uniformly distributed, thereby reducing the area of unsupported 
 shell plating. Fig. 73 shows an excellent arrangement of side stringers. 
 
 Every strake of shell plating has a stringer upon it, extending through- 
 out its whole length, thus converting each strake with its accompanying 
 stringer into a combined girder. The excellency of such an arrangement 
 needs no further comment. 
 
 Third. — When the foregoing system of adopting additional longitudinal 
 girders in combination with widely spaced transverse framing is not carried 
 out, the required stiffness to the shell plating can only be obtained by 
 increasing its thickness. Indeed, in the Board of Trade Freeboard Tables 
 and accompanying Rules, the following paragraph is found relating to 
 this aspect of a vessel's strength : — 
 
 " If the frame spacing be increased one-fourth, the thickness of all the 
 plating, excepting garboard and sheer strakes, should be increased by oV th of 
 an inch over the thickness required in the standard ship. Other increases 
 in spacing should be dealt with in the same proportion." 
 
 3rd. Hold Pillars. — Pillars, or some substitution for the same, are an 
 absolute necessity in all ships. In vessels of comparatively small beam, 
 middle line pillars alone are sufficient, but in larger vessels having greater 
 breadth, the beams (decks) i-equire additional support between the middle 
 line and their connection to the frames. In figs. 72 and 73, which illustrate 
 a vessel of 52| ft. in breadth, both middle line and side or quarter pillars are 
 required by Lloyd's Rules, but as shown in the diagram, these are entirely 
 omitted with the exception of those shoAvn in the profile, Avhich are of 
 exceptional construction and strength. In the main hold, which is the 
 longest in the vessel, two hollow plate pillars, 21 in. in diameter, support 
 the deck, the supporting strength being distributed throughout the length 
 of the hold space by a very strong box girder w^ell connected to the deck 
 and beams and to the bulkhead by means of large brackets. 
 
 The thrust of these two pillars is distributed over the bottom of the 
 vessel by means of a stool to which they are connected by angle rings — the 
 stool in its turn being well connected to the inner bottom. (See figs. 72, 
 73, and 74.) 
 
 The main 'tween-deck pillars are 12 in. in diameter. They are 
 connected to the deck at their lower extremity by an angle ring, and to 
 the upper deck by a girder made up as shown in fig. 74 of plates and 
 channel bars. This girder comes alongside the upper deck hatch coaming 
 plate, which itself forms one web of the girder, and is extended along the 
 upper deck over the full length of the main hold, so as to rest upon the 
 bulkheads at the extremities of the space. A couple of angle bars along the 
 
150 STEEL SHIPS. 
 
 top edge of this plate (excepting in way of the hatch) add additional 
 stiffness to the girder. The continuity of these deck girders or stringers is 
 preserved all fore and aft, though somewhat modified in form on account of 
 the shorter lengths of the remaining holds. For a similar reason, a modi- 
 fication is made in the construction of the pillars. These as shown in 
 figs. 73 and 74, are made up of two channel bars, assisted at intervals in 
 their height by plate ties, and well connected to the deck girders and tank 
 top by means of large brackets at the upper and lower extremities. 
 
 The 'tween-deck pillars for these holds are made of solid round iron, 6 in. 
 in diameter, with large palms to both their feet and heads, thereby permitting 
 a good rivet connection. 
 
 Jfili. Stern Arrangement. — The great innovation in the arrangement of 
 the stern for the efficient support of the propeller, and in the construction 
 of the rudder and the manner in which it also is supported, is clearly 
 illustrated in fig. 75. Unlike the ordinary stern frames in single screw 
 steamers, both forgings and castings are conspicuous by their almost entire 
 absence, the only item of this kind being a comparatively small steel casting 
 for the boss. This extends, as shown in fig. 75, from a watertight flat 
 immediately above the boss to the lowermost extremity of the aftermost 
 watertight bulkhead, and is riveted to both the watertight flat and the 
 bulkhead through palms on the casting. 
 
 The remainder of the propeller aperture arch is constructed of a 
 (J -shaped Siemens-Martin steel plate 1 in. in thickness. By this 
 arrangement, both the longitudinal and transverse framing are thoroughly 
 connected to the stern plating as shown by the detailed sketches in fig. 75. 
 Such an arrangement at once dispenses with a very considerable amount 
 of weight, and makes the structure of the stern a thoroughly combined 
 and compact one, and experience has proved these vessels to be remarkably 
 free from vibration. Loose rivets, so prevalent in ordinary stern frames, 
 have so far been practically unknown. 
 
 The stern post proper, upon which the rudder is shipped, consists of a 
 wrought iron tube, 1 in. in thickness and 21 in. in diameter; and to afford 
 additional strength to the lower part of the post where most of the stress 
 from the rudder is sustained, a doubling or liner of the same thickness is 
 fitted. 
 
 The rudder also is of unusual design. It contains no forgings nor 
 castings excepting in the neighbourhood of the coupling, where a steel 
 casting is necessary in order to obtain an efficient connection with the 
 rudder stock. Apart from this, the rudder is entirely made of plates 
 and angles (see fig. 75), and is designed to offer the least resistance to 
 propulsion. 
 
 This rudder possesses the further advantage, owing to its form, of 
 having the interior accessible in every part. The socket in the rudder 
 into which the bottom of the stern post fits, is made of Siemens-Martin 
 steel, with a white metal lining for a working surface. 
 
TYPES OF VESSELS — SECTION II. 
 
 151 
 
152 STEEL SHIPS. 
 
 For this construction of stern frame and rudder, it is claimed that a 
 saving of at least 10 tons of weight is effected in vessels of this size. 
 
 These ships have proved exceptionally easy to handle, and thus the 
 arrangement just described, in addition to its structural efficiency, enhances 
 manoeuvring powers. 
 
 5th. Cement. — A great saving of weight is effected in these vessels 
 by the disuse of cement as ordinarily applied in cellular double bottoms. 
 The inner surfaces of these tanks are well preserved by being coated with 
 cement wash only, which must naturally be renewed periodically. 
 
 The use of cement has been more fully discussed in Chapter VIII., which 
 deals particularly with the subject of ' Maintenance.' 
 
 Steamers for carrying Oil in Bulk. 
 
 The great development which has taken place in comparatively recent 
 years in the exportation of petroleum from Black Sea and American ports, 
 has naturally led shipowners to consider what was the most economical way in 
 Avhich this cargo could be carried, and notwithstanding the prejudice and 
 suspicion of danger which first existed in the minds of many interested 
 shipowners and others against the carrying of oil in bulk, as well as tlie 
 difficulties which presented themselves, yet the great advantage over that of 
 carrying oil in cases and in barrels led to the experiment being made. The 
 fii'st ship in which oil was carried in bulk was fitted out at the shipyard of 
 Messrs. Craggs & Co., of Middlesbrough. The vessel was an ordinary cargo 
 steamer, into the holds of which huge oil-tight tanks were fitted, shaped to 
 the form of the vessel. In these tanks the oil was carried. Since then, a 
 large number of vessels have been specially built to engage in this trade, 
 and great developments have been made in the construction and adaptation 
 of ships for this purpose, due in no small measure to the Bureau Veritas 
 Classification Society, whose experience was probably unequalled in the 
 early days of tank steamers. 
 
 Figs. 76 and 77 are profile illustrations of two modern oil steamers built 
 by Messrs. Armstrong, Whitworfch, & Co. of Newcastle-on-Tyne, who 
 have earned a great reputation in building this class of vessel. Fig. 78 is 
 a midship section showing the disposition of the material in the construction. 
 Before briefly observing the principal features in the arrangement and 
 construction of oil steamers, a few points especially peculiar to vessels 
 engaged in carrying oil in bulk, which affect the design to a considerable 
 extent, and need to be well remembered in the process of construction, may 
 be first noted. 
 
 1st. The gas which arises from petroleum, especially crude oil, in 
 combination with the atmosjihere becomes highly explosive at a certain 
 point. 
 
 2n(d. If the temperature of the oil is increased, it is subject to expansion. 
 
 3rd. While, in an ordinary cargo vessel, the cargo bears directly upon 
 
Fig. 76. 
 
 ELEVATION AND DECK PLANS OF STEAMER FOR CARRYING OIL IN BULK. 
 
 WITH ENGINES AFT. 
 
 Dimensions— Length B.P. 335 ft.; breadth 46 ft. ex.; depth mid. 29 ft. 6 ins. 
 
 Plan of Second 
 Deck 
 
 'Cofferdam 
 
 Cofferdam 
 
TYPES OF VESSELS — SECTION II. 153 
 
 the transverse and longitudinal framing, the weight chiefly Leing taken by 
 the floors, in a modern oil steamer the oil extends out to the outside shell 
 plating, and, like all fluids, exerts its pressure square to the surface against 
 which it comes into contact. 
 
 In addition to this, there is the increased pressure due to the inertia of 
 the cargo itself, as the vessel rises and falls in her pitching movements, and, 
 to a less extent, in rolling. 
 
 4th. With liquid cargoes, like oil or water carried in bulk, there is 
 always a certain amount of danger in the process of filling up the tanks, if 
 the vessel is deficient in stability. Evidently, should the vessel, through 
 lack of metacentric height in any condition, take a list, the free surface of 
 the liquid in a partially filled tank may tend only to increase the heeling, 
 for the more the vessel inclines the more will the liquid shift. 
 
 From the principal points just enumerated, it follows that oil-tightness, 
 structural strength, and stability are of the utmost importance in vessels 
 carrying liquid cargoes in bulk, and suitable provision must be made for the 
 expansion of the oil. Nothing is more essential than that the riveting be 
 most efiiciently performed, and the workmanship generally of the very highest 
 character. Bad riveting spells doom to an oil steamer, for one can easily 
 imagine, taking one aspect of the danger alone, what terrible results might, 
 and indeed have, accrued owing to leaky bulkheads, and the oil or gases 
 finding their way into the engine and boiler compartments. More disasters 
 from explosion and fire have happened to oil steamers than from any other 
 cause. A loose rivet should never be caulked, but the rivet removed, and 
 the hole re-riveted. It is also necessary in order to obtain and maintaui 
 oil- tightness, that the very best form of rivet should be adopted. Two kinds 
 of rivets have proved very satisfactory for oil-tight work, the pan-head rivet 
 with the swollen neck, and the plug-head rivet (see fig. 90). In adopting 
 either of these, the plating ought to be countersunk, so as to better receive 
 the swelled neck. The drift punch must be rigorously forbidden. Elind 
 holes should be rimed fair and not torn open, as so often is done with the 
 drift punch. The spacing of rivets for oil-tight work should never exceed 
 3 to 3^ diameters of the rivet. 
 
 In all vessels carrying liquid cargoes in bulk in the hold spaces, the 
 strain upon shell plating and bulkheads is very great, owing to the 
 pressure of the liquid increased by the inertia of the cargo when pitching and 
 rolling. The bulkheads should therefore be amply stiff"ened and supported. 
 The rivets through the flanges of the frames connecting the shell, which 
 largely bear the strain due to pressure on the shell plating, should be spaced 
 not more than six diameters apart. 
 
 In all hold spaces in which liquid cargoes are to be carried, the 
 middle line bulkhead should be fitted all fore and aft through such spaces, 
 extending from the keel to the deck forming the crown of the tank, or to the 
 top of the expansion trunk. This bulkhead should be strictly oil- and water- 
 tight. It is valuable because it contributes strength and unites and supports 
 
154 STEEL SHIPS. 
 
 the deck and bottom plating, but it is most valuable from a stability point 
 of view, its principal function being to minimise the shifting of cargo in the 
 event of the vessel taking a slight list in the process of filling up the tanks. 
 It ought never to be required for this purpose when the vessel is at sea, 
 for all tanks ought to be completely filled before proceeding upon any sea 
 voyage, however short it may be. To proceed to sea with an oil or a water- 
 ballast tank only partially filled is to invite disaster, for such display of 
 ignorance has not infrequently resulted in bulkheads being torn down and 
 other extensive damage wrought. A longitudinal bulkhead also reduces the 
 inertia of the cargo when rolling in a seaway. Long hold spaces in a fore 
 and aft direction are most undesirable in any vessel carrying liquid cargo in 
 bulk because of the rapidity with which such cargoes shift should the tank 
 not be completely filled, and the great inertia which huge volumes of such 
 cargo possess. Such danger as described can easily be imagined if by any 
 chance a vessel began to pitch in a seaway, with a very long hold only 
 partially filled. Under such circumstances, for any bulkhead to bear 
 the strain would be practically impossible, not to mention the additional 
 danger which would arise from vastly increased change of trim and heavy 
 diving which would undoubtedly ensue. Transverse bulkheads should 
 therefore never be spaced at greater intervals than between 24 and 28 ft., 
 and should be thoroughly oil- and water-tight. Vessels carrying liquid cargoes 
 are especially subject to severe racking stresses, and strength to encounter 
 such is largely provided by these numerous transverse bulkheads. 
 
 As before stated, all bulkheads, both transverse and longitudinal, should 
 le very strong and well supported. Oil steamers should possess a reasonable 
 metacentric height at every intermediate draught between the light and load 
 lines, which is passed through as the tanks are being filled. In the fully 
 loaded condition, from 1 to 2 ft. should prove a satisfactory metacentric 
 height in vessels of usual proportions. ISTo class of cargo vessel is so easy to 
 manipulate in the process of designing, in order to arrive at a desired con- 
 dition of stability, as the bulk oil-carrying steamer, because in no other 
 cargo vessel is the seagoing condition so constant. 
 
 In order to provide for the increased bulk in the event of expansion 
 taking place, a trunk-way is fitted in the 'tween decks continuously all fore 
 and aft over the main hold tanks. This is shown clearly in fig. 78. The 
 transverse bulkheads must extend to the top of the expansion trunk. 
 
 In order to minimise the danger of leakage, so as to prevent either oil or 
 gas finding its way either into the engine and boiler space or cross bunkers, 
 or into any hold space used for other purposes than carrying oil, double 
 bulkheads are fitted at a distance of not less than two frame spaces apart. 
 Both these bulkheads are oil-tight, and extend from side to side, and from 
 the keel to the top of the expansion trunk. In fig. 76, where the engines 
 and boilers are situated at the after end of the vessel, the two oil-tight bulk- 
 heads just mentioned are shown between the cross bunker at the fore end 
 of the boiler room and the aftermost oil tank. The space between these 
 
Fig. 77. 
 
 ELEVATION OF STEAMER FOR CARRYING OIL IN BULK, WITH ENGINES AMIDSHIPS. 
 
 Dimensions-Length B.P. 366 ft. ; breadth 49 ft. 6 ins. MLD.; depth 30 ft. 6 ins. MLD. 
 
 Store \'\% ^^^ Expansion Trunks 
 
 Reserve Expansion 
 Bunker Trunk 
 
 Oil Compartment 
 Shaft Tunnel 
 
 Pumj)Room\\'^^ fng,„e 
 Room 
 
 X'Doiijiey b.oiler 
 
 Cofferdam 
 
 ■-■-• y, ^»«:^^-l-l-(-I-r^^ '~rrr[-i-F' 7 tT-»=93-si ;' : 
 
 rrr^-i-r rrrrrrflrg;!!! : : ! 1^ ! ; ! ! i ; I : ! i : ; i i igj-tTT; 
 
 ft 
 
 Oil Compartments i/ F^ a lu^ 
 
 ^ , 7 2 Water Ballast ■ 
 
 I Cofferdam I ^ _ ^,-'.>; 
 
 Oil 
 Compartment 
 
 Cofferdah 
 
 Ballast Tank 
 
 Cofferdam 
 
TYPES OF VESSELS — SECTION II. I 55 
 
 bulkheads is called a coffer-dam. In this vessel, as the whole compartment 
 abaft of the collision bulkhead is used for general cargo, a cofFer-dam is 
 again found between this compartment and the foremost oil tank. It is 
 most unusual to carry oil in the endmost compartments of any steamer. 
 Indeed, for vessels passing through the Suez Canal, this is strictly enforced ; 
 and another regulation is that the coffer-dams be filled with water to their 
 utmost extent, when carrying oil in bulk. This, it is hoped, will form an 
 effective barrier against any oil finding its way through the cofFer-dam. But 
 even with these precautions, owing to bad riveting, or excessive straining 
 of the bulkheads arising from defective stiffening, oil and gas have been 
 known to percolate through the coffer-dam bulkheads into cross bunkers 
 and into engine and boiler rooms. When this has happened in vessels 
 "where it has been customary to fill the coffer-dams with water, this serious 
 condition of affairs has generally been brought about by part of the water 
 leaking from the coffer-dam, its place having been taken by oil which has 
 leaked into the coffer-dam and floated on the water, and eventually found 
 its way into the adjoining spaces. Hence that these coffer-dams be 
 frequently inspected during the voyage and their condition ascertained, is 
 most essential. If oil is found in the space, it should be drawn off. 
 Ample ventilation to the coffer-dams is extremely necessary. Some owners 
 prefer to keep the coffer-dams empty, and by frequent inspection, the ships' 
 officers are kept cognisant of their condition. In the event of oil having 
 found its way into the space, it is immediately pumped back into the main 
 tank. In this case again, thorough ventilation is essential. Fig. 79 shows 
 an enlarged view of a coffer-dam in a vessel where the engines are situated 
 amidships. It wall be seen in the coffer-dam diagram that the strength 
 of its two bulkheads is united and great additional support given by means 
 of the diaphragm plates shown. In a transverse direction, these diaphragm 
 plates are spaced about 3 ft. apart. In fig. 77, Avhich shows the engines 
 and boilers amidships, it will be observed that a coffer-dam is situated at 
 the aft end of the engine room, another at the fore end of the cross bunker 
 forward of the boiler room, another at the fore end of the foremost tank, 
 and another at the after end of the aftermost tank. While there is much 
 to be said in favour of placing the engines and boilers amidships, the great 
 advantage of placing them at the after end is that they are more entirely 
 shut off from the oil tanks, and the danger rising from leakage minimised. 
 To ensure safety, the donkey boiler and galley must be thoroughly isolated 
 from the oil tanks. 
 
 The transverse frames should preferably be of bulb angle, or channel, or 
 Z bars, instead of the ordinary frame and reverse bars. This reduces the 
 riveting and secures greater transverse strength, which is a highly important 
 factor in bulk oil-carrying vessels. In two decked vessels, the second deck 
 forms the crown of the main oil tanks. It should therefore be a thoroughly 
 oil-tight flat. To secure this most effectively, all the frames are usually cut 
 in way of this deck, and the deck plating carried without a break right out 
 
156 
 
 STEEL SHIPS. 
 
 to the shell of the vessel, to which it is connected by a continuous angle bar. 
 This is clearly shown in the midship section, fig. 78. The frames in the 
 
 Fig. 79. — Enlarged View of Coffer-dam, showing Diaphragm Plates, Trunk-way 
 to Tunnel, and Ventilating Pipe. 
 
00 
 
 o 
 
 Li. 
 
 iL S 
 
TYPES OF VESSELS — SECTION II. I 57 
 
 'tween decks are connected to the beams at their upper extremity, and 
 to the deck at the lower extremity, by means of bracket knees. The 
 expansion trunk (see fig. 78) is fitted all fore and aft in the 'tween decks, 
 and its sides are supported by webs of plating at intervals as shown. 
 All the oil hatches on the upper deck must be strongly constructed and 
 oil-tight. The spaces in the 'tween decks at the sides of the expansion 
 tanks are used as bunkers. The plan of the upper deck (fig. 76) shows the 
 arrangement of all the oil and bunker hatches, and the plan of the second 
 deck (fig. 76) shows the continuous expansion trunk-way, and the longitu- 
 dinal and transverse bulkheads, including the coffer-dam bulkheads. Where 
 lines indicating bulkheads are shown in full, they extend to the upper 
 deck ; and wliere such parts of them are dotted, they terminate at the 
 second deck. 
 
 In oil vessels of a depth such as to require three tiers of beams, the 
 lower deck is usually dispensed with as shown in the midship section, fig. 
 78, compensation being made by means of web frames and web stringers. 
 Two web frames are usually placed in each oil tank. An extra deep beam 
 extends across the vessel at the head of each web frame, to which it is 
 connected by a large plate knee, as dotted upon the midship section. The 
 longitudinal bulkhead, which is oil-tight, extends from the keel plate to the 
 top of the expansion trunk. The vertical stifEeners are of bulb angles one 
 frame space apart, lapped on to the floor plates, and further connected to the 
 floors and to a narrow width of deck plating through the expansion trunks 
 by plate brackets. At intervals, deep web plate stiffeners are fitted. 
 These are spaced to agree with the web frames on the ship's side, th& 
 bulkhead web plate stiffeners being bracketed to the deep plate beams 
 which unite the heads of the web frames. They are also lapped on to and 
 bracketed to the floor plates (see fig. 78). On the other side of the 
 longitudinal bulkhead, channel plate horizontal stifl^eners are fitted, spaced 
 so as to be at the same distances above the keel as the web stringer 
 plates. 
 
 The transverse bulkheads are continuous from side to side of the vessel.. 
 The longitudinal bulkheads are therefore fitted intercostally, as it were,, 
 between them. 
 
 The stiffening of the transverse bulkheads is very similar to that just 
 described for the longitudinal bulkhead. (See figs. 80 and 81.) The 
 vertical stiffeners, which are of bulb angles, are spaced about 2 ft. apart,. 
 and the horizontal channel plate stiffeners agree with the spacing of the 
 web stringers and horizontal stiffeners in the longitudinal bulkhead. 
 
 The vertical stiffeners are bracketed, top and bottom, as shown in figs. 
 80 and 81, and the horizontal stiffeners of both the transverse and 
 longitudinal bulkheads are thoroughly united by large brackets as shown in 
 fig. 81. 
 
 Transverse bulkheads, exceeding 42 ft. in breadth, should have four 
 vertical web plate stifteners in addition to the longitudinal bulkhead, and 
 
158 STEEL SHIPS. 
 
 where the breadth is less, two vertical webs, one on each side of the 
 longitudinal bulkhead. 
 
 Attention is drawn to the splendid stiffening which is afforded to the 
 oil tanks in the arrangement just described by having all horizontal 
 stiffeners and web stringers practically in the same horizontal plane, each 
 tank being circumscribed by this web system of stiffening. 
 
 As the oil-tightness, particularly of the transverse bulkheads, is of such 
 paramount importance, it is usual to cut all stringers and keelsons which 
 would pass through this bulkhead, and preserve the continuity of strength of 
 these important longitudinal girders by large bracket plates, as shown in 
 fig. 82. Sometimes, however, the side stringers are carried uninterruptedly 
 through the transverse bulkheads, as in fig. 81. In this case, most efficient 
 angle collars have to be fitted metal to metal to preserve watertightness. 
 
 The connection of all seams and butts in the plating of bulkheads should 
 be by means of laps, double riveted throughout. A most important point 
 in all oil-tight work is to avoid three-ply riveting, as the tightness of such is 
 not nearly so reliable. Thus, that the connection of the transverse bulk- 
 heads to the shell plating and deck and to the longitudinal bulkheads be 
 by means of large single angle bars, with flanges broad enough to take two 
 rows of zigzag riveting, is strongly recommended. (See figs. 80, 81, 82, 
 83, and 84.) 
 
 The bulkhead liners fitted to alternate strakes of shell plating are 
 usually of sufficient width, so as to get one row of rivets on each side of the 
 bulkhead shell bar. This, it is found, is more easily caulked and made 
 watertight. (See fig. 84.) "When the shell plating is flanged, as shown in 
 the midship section, fig. 78, each strake of shell plating therefore bearing 
 close upon the frames, and no packing being required, the bulkhead liners 
 may be fitted on the outside of alternate strakes of the shell plating (see fig. 
 80), but more usually the liners are dispensed with, and by making the 
 bulkhead shell bar of extra thickness, a wider spacing of shell rivets is 
 obtained, so as to weaken the shell as little as possible. 
 
 That the bulkheads be caulked upon both sides is strongly recom- 
 mended. In order to ensure thorough tightness at all connections, this 
 should be secured by the surfaces of the metal bearing hard upon each 
 other. Felt packing should, under all circumstances, be avoided in oil-tight 
 work, and white lead as far as possible. The latter, however, cannot 
 always be entirely dispensed with, white lead injections being at times 
 necessary. 
 
 The seams of the shell strakes should be double riveted, and the 
 butts preferably overlapped and treble riveted with three complete 
 rows of rivets. Two rivets should pass through the seams upon every 
 frame instead of the single rivet in ordinary cargo steamers, in order to 
 ensure oil-tightness. If the butts are not lapped, they should be connected 
 by double butt straps treble riveted in way of all oil-tanks. When the 
 engines are situated amidships, as in fig. 77, the shaft tunnel is entered by 
 
Fig. 80. 
 transverse bulkheads. 
 
 Vner 
 
 A 
 
 Co 
 
 5 
 
 
 T 
 
 Co 
 
 4} 
 
 Co 
 
 :5i 
 
Fig. 81. 
 
 plan showing vertical and horizontal stiffeners 
 upon a transverse bulkhead, and connections 
 to longitudinal bulkhead and side stringers, 
 (oil steamer.) 
 
 Stringer continuous 
 through Bulkhead 
 
 T — pr r 
 
 Transverse BuHf' 
 
 Angle Collars on Bulkhad 
 round continuous Stringer 
 
 Connection of Vertical Webs 
 to Keelson 
 
 ^^ 
 
 fv\ 
 
 '^ 
 
TYPES OP VESSELS — SECTION II. 
 
 159 
 
 means of an oil-tight trunk-way from the upper deck at each end of the 
 tunnel. In fig. 77 this trunk-way passes through the coffer-dam at each 
 
 Section 
 through 
 
 AB 
 
 Fig. 83.— Showing Bulkhead 
 Liner and Angle Connection 
 to Shell. 
 
 '^ ^ o 0000 o o o 
 Fig. 84.— Bulkhead Liner, 
 
 Fig. 82. — Showing a Side Stringer cut and bracketed 
 to a Bulkhead (oil steamer.) 
 
 end of the tunnel and is itself independently oil-tight. An enlarged view 
 of the fore trunk- way is shown in fig. 79. Xeedless to say, the tunnel 
 
l6o ^ STEEL SHIPS. 
 
 is thoroughly oil- and gas-tight. It should, however, be well ventilated 
 by cowl ventilators into each trunk-way. 
 
 The gases which are emitted from the oil are of greater specific gravity 
 than the atmosphere ; they are, therefore, apt to lodge about the lowermost 
 recesses of all spaces into which they enter. A system of drawing off any 
 such gases, and assisting in the ventilation of the tunnel, is shown in fig. 79, 
 where a large perforated pipe passes along the tunnel and up the trunk-way 
 to a fan which acts as a suction. 
 
 Even when all the oil has been pumped out of the oil tanks, great 
 danger still exists for the reason just mentioned, viz., that the heavy and 
 inflammable gases hang about the bottom of these spaces. To rid the tanks 
 of these objectionable and dangerous vapours, several means are used, the 
 chief being to blow out the gases by means of steam injections, and to draw 
 off the gases by using the oil suction pipes with a fan. Steam injections 
 are sometimes used when necessary in the tunnel. Ample ventilation 
 should be provided to all oil tanks. In order that the oil may readily 
 reach the pump suctions, sufficient limber holes should be fitted through 
 floors and side intercostal plates, but on no account through the middle 
 line bulkhead. 
 
 Oil vessels have been built with an inner skin, but there is little to 
 recommend this system, the space between the inner and outer skins being 
 a harbourage for gas, "When such a system is adopted, the importance of 
 fully ventilating these spaces is obvious. 
 
 Water-ballast Arrangements. 
 
 It is not intended in this chapter to enumerate and discuss the reasons 
 and necessity for arrangements in the modern cargo steamer for the 
 carriage of water for ballasting purposes. This has already been done in 
 " Know Your Own Ship." Suffice it to say, as every seaman knows, 
 that in these days of fluctuating freights, large vessels frequently make 
 long over-sea voyages without cargo. And with the full bilge and small 
 rise of floor so commonly given to present-day steamers, and remembering, 
 too, the great increase whichhas taken place in the dimensions of such 
 vessels, it follows that the draught in light condition is greatly less in 
 proportion than it used to be for the finer formed and smaller-sized cargo 
 vessels then in use. All this enhances the necessity for ballast when in a 
 light condition, and as water has been found to be the most economical 
 ballast that can be carried in steamers, it is proposed to illustrate the 
 principal arrangements for the carriage of water-ballast. 
 
 M'Intyre Double Bottom Tanks. — The first occasions upon which 
 water-balList was introduced into vessels were to enable colliers carrying coal 
 from the Tyne to London to make the return voyage without the necessity 
 and expense of loading, and at the end of the return voyage discharging dry 
 ballast. The necessary alterations in the first vessels fitted for this purpose 
 
TYPES OF VESSELS — SECTION II. l6l 
 
 were carried out by Messrs. Palmer & Co., of Jarrow-on-Tyne. In the 
 system then adopted, the water ballast was carried along the bottom and 
 contained between the outer and an inner bottom, which latter was laid upon 
 fore and aft girders standing upon the ordinary floors, and made water- 
 tight by being connected to the outside plating,* Fig. 71 illustrates the 
 midship section of a vessel fitted with the modern arrangement of such a 
 double bottom tank, in some of the smaller details the construction having 
 been slightly modified from what it was originally. For instance, in some of 
 the earlier cases, the tank top plating was carried in a horizontal, or almost 
 horizontal direction from the top of the centre keelson right out to the bilge 
 or shell plating, and made watertight at each of its side extremities by being 
 connected to the shell by means of angle collars. Another system of 
 making this connection is by flanging the tank side or margin plate 
 similarly to the system illustrated in tig. 71, so as to meet the bilge plating 
 at right angles. The watertightness is here again secured, and the 
 connection made by means of angle collars. In both of these cases the 
 reverse frames are cut at the tank side, and compensation made by doubling 
 the frame bar (which is continuous from keel to gunwale) for about 3 ft. 
 This greatly facilitates the fitting of the watertight collars. 
 
 But the method now most visually adopted for the connection of the 
 tank side to the shell plating is that shown in fig. 71, where the tank side 
 or margin plate is arranged to meet the shell at right angles. In this 
 case the main frames are cut at the tank side and the margin plate is 
 connected to the shell by a continuous angle bar. This is probably the 
 simplest way of ensuring the watertightness of the tank. The efficiency 
 and continuity of the transverse framing is effected by connecting the frame 
 legs outside the tank to the tank side plating by means of large bracket 
 plates or tank knees. Hence the tank side plate must have sufficient 
 breadth, so as to get a sufficient number of rivets through to ensure a good 
 connection. 
 
 Cellular Double-Bottom Tanks. — However, since the inauguration of 
 the M'Intyre tank, a very considerable development has taken place, and 
 now a type of tank which is more embodied in the main hull of the vessel 
 has come to be extensively adopted, though not to entirely supersede the 
 M'Intyre tank. This tank is that illustrated in figs. 12, 48, and 58, and is 
 known as the cellular double-bottom water-ballast tank. In the M'Intyre 
 double-bottom tank the floors are upon every frame all fore and aft, and in 
 the earliest form of this tank the ends of the floors were turned up the 
 bilge in the usual way. But in later M'Intyre tanks, the floors inside 
 the tank have been carried in a horizontal line from the centre keelson 
 to the intersection with the frame or tank side, and strengthened at their 
 narrow outward ends by bracket plates fitted to the tank side. (See fig. 71.) 
 If desired, this tank may extend over only part of the length of the vessel, 
 
 * This is known as the M'Intyre tank, as the idea originated with Mr. John M'Intyre, 
 the manager at that time of Palmer's shipyard. 
 
 11 
 
1 62 STEEL SHIPS. 
 
 the remainder of the bottom being of the usual construction. The tank is 
 commonly dispensed with under the engines, and especially under the boilers, 
 on account of the rapid corrosion, due to damp heat, which is known to take 
 place under the boilers. 
 
 In the cellular double-bottom form of construction, the centre keelson 
 or the centre through-plate is continuous fore and aft, with two large con- 
 tinuous angles on its upper and lower edges. The floors in mercantile 
 steamers are usually solid from the centre through-plate to the tank margin 
 plate, lightened by manholes as shown in figs. 12, 48, and 58, which man- 
 holes are also the only means of passage through the tank. The inner 
 bottom plating is laid on the top of these floors, and the margin plate is 
 arranged at the tank side so as to meet the shell plating practically at 
 right angles. The floor plates are stiffened and prevented from buckling 
 by fitting one or more intercostal girders, according to the size of the 
 vessel, which are connected to the inner and outer bottom plating and to 
 the floor plates by angle bars. (See fig. 48.). Or the angles taking the 
 inner and outer bottom may be dispensed with by flanging the plates. 
 <See fig. 64.) 
 
 Where, however, the maximum longitudinal strength is desired, these 
 side girders are commonly fitted continuous all fore and aft, the floor 
 plates and reverse bars inside the tank between them, therefore, being 
 intercostal owing to the top angles on the side girders being continuous. 
 On account of the depth and strength of the floor plates, the large 
 frame bar, such as is fitted outside the tank, is not required, and, as 
 shown in fig. 48, a sinaller bar, whose flanges are of about the size of the 
 smaller flange of the ordinary frame, is used. As shown in fig. 48, the 
 reverse bar is fitted, as in the case of ordinary floors, on the opposite side of 
 the floor plate to the frame bar. The tank side is made watertight by 
 fitting a continuous bar on the lower outside edge of the margin plate, so 
 as to bear hard against the bilge shell plating to which it is riveted. Both 
 frames and reverse bars are severed at the tank side, this being necessary in 
 order to simplify the rendering of the inner bottom watertight. However, 
 as the transverse strength must be uninterruptedly maintained, the frame 
 legs are attaclied to the tant side by large bracket plates carried well up the 
 bilge. These are connected to the tank side by double angle bars for at 
 least half the length in large vessels. It may be noted that the reverse 
 bars in the double bottom are doubled under the engines and boilers where 
 special rigidity is required, and for the same reason, additional intercostal 
 side girders are fitted. The cellular double-bottom ballast tank is usually 
 divided into numerous separate watertight compartments, not only for the 
 purpose of subdivision in the event of the perforation of the outside plating, 
 but in order to ballast the vessel so as to obtain a desired condition of trim. 
 
 Fig. 85 shows a modification in the construction of the cellular double- 
 bottom tank, as previously shown. 
 
 In this system, the longitudinal side girders are more numerous, and 
 
TYPES OF VESSELS — SECTION II. 
 
 163 
 
 the floor plates are only solid upon alternate frames. On the intermediate 
 frames, brackets are fitted to the centre through-plate and the tank side. 
 The connection of the tank side to the frame legs is identical with 
 that previously described for the more ordinary cellular double bottom, 
 brackets being fitted to every frame. A point to be noted in this form 
 of construction is, that whereas in Board of Trade standard vessels with 
 cellular double bottoms having solid floors to every frame, when the shell 
 plating in way of the double bottom is ^ths or more in thickness, a reduc- 
 tion of o^ijth is permitted ; no such reduction is allowed where the floors are 
 not solid upon every frame. All double-bottom tanks are entered by means 
 of manholes in the tank top plating, which should be sufficiently numerous 
 and situated in such positions as will give the easiest access to every cell or 
 compartment in the tank. At least one manhole should therefore be cut in 
 every solid floor plate between the longitudinal girders, whether intercostal 
 or continuous, and, at intervals, manholes should be cut in the side girders 
 also ; but as the centre through-plate is a most important item in the longi- 
 tudinal strength of the vessel, upon no account should it be perforated by 
 manholes, as this would seriously interrupt the continuity of its strength. 
 
 Fig. 85. — Cellular Double Bottom, constructed with Solid Floors upon alternate 
 frames, and Bracket Plates upon intermediate frames, as shown in this sketch. 
 
 Indeed, in some vessels the centre through-plate is perfectly watertight, 
 for part, if not for its whole length. 
 
 Fore and After Peak Tanks (see figs. 46, 47, and 49). — As the lower 
 spaces on the fore side of the collision bulkhead, and on the after side or the 
 aftermost bulkhead, are both inconvenient and unsuited for the carriage of 
 cargo, and ill adapted even for ship stores, it has become common in steamers 
 to use one or both of these spaces in which to carry water for trimming 
 purposes. It is scarcely correct to call them ballast tanks, for their capacity 
 is comparatively so small that the weight of water they contain produces 
 very slight increase in the mean draught, but the filling of one of them 
 produces considerable change in the trim, situated as they are so far 
 from the centre of gravity of the ship. Hence, in a light condition, the 
 filling of the after-peak tank may considerably contribute to the immersion 
 of the propeller, though there follows the natural objectionable result of the 
 bow being lifted correspondingly out of the water. The fact must not be 
 
164 STEEL SHIPS. 
 
 overlooked, that when these tanks are made very large, containing, as they 
 do in some cases, over 100 tons of water, they are liable to produce straining 
 when the ship pitches in a seaway in the light condition. In any case, the 
 bulkhead separating these tanks from the holds will be watertight, though, 
 when the peaks are intended for water-ballast, these bulkheads should be 
 especially strengthened. In addition, the deck forming the top of such 
 a tank must also be watertight. it must therefore be constructed of 
 iron or steel, and if the space complies with the conditions necessary for 
 its exemption from the British tonnage, it must be entered only by means 
 of a manhole in the deck plating. As there is always the danger of these 
 tanks being only partially filled (a condition which ought to be rigorously 
 striven against), wash plates are fitted down the middle of their length. 
 (See fig. 49.) 
 
 Deep Midship Tanks. — Especially in recent years has the practice been 
 adopted of constructing one or more deep water-ballast tanks, somewhere 
 in the middle length of the vessel, usually at one or both ends of the 
 engine and boiler space. Fig. 72 illustrates the position of these tanks. 
 They are in reality hold spaces so constructed as to be available for 
 the carriage of cargo when required, and water as ballast when in a light 
 
 Shade Deck _ 
 nam UeclT 
 
 Engines & Boilers 
 
 3B 
 
 Cross Bunker 
 
 Fig. 86. — Illustrating Side Water-ballast Tanks in shallow high-speed 
 Passenger Steamers. 
 
 condition. They are bounded on each end by watertight bulkheads specially 
 strengthened, and are enclosed on the top by a watertight iron and steel deck 
 or flat. When these tanks extend from side to side of the vessel, a water- 
 tight bulkhead must be fitted down the middle of the tank in a fore and aft 
 direction. These tanks are entered by means of small hatchways through 
 the watertight deck, which hatchways are themselves made watertight by 
 means of an iron or steel plate bolted on to their coamings. The steamer 
 in fig. 49 has a deep water-ballast tank at the after end of the engine 
 room. 
 
 Side Deep Tanks. — Instead of carrying midship deep tanks from side to 
 side of the vessel, they are sometimes fitted only in the wings. (See fig. 86.) 
 Carrying less water than they would do if fitted from side to side, they do 
 not produce as much increase of immersion, but from a theoretical point of 
 view, they possess a decided advantage in easing the vessel in her rolling 
 motions, and producing a steadier ship. They are therefore better adapted 
 for shallow high-speed passenger vessels, which carry little or no cargo. In 
 such cases, an ordinary manhole is sufficient as a means of entrance, but 
 
Figs. 87 and 88. 
 PROFILE AND DECK PLANS OF STEAMER WITH SIDE WATER BALLAST TANKS. 
 
 Manholes in Sinnger Plate to Side tanks 
 
05 
 
 00 
 
 o 
 11. 
 
 < 
 
 OQ 
 Ui 
 
 I- 
 < 
 
 UJ 
 Q 
 
 a. 
 
 UJ 
 
 < 
 
 UJ 
 
 I- 
 
 (f> 
 
 I 
 
 O 
 3 
 O 
 
 oc 
 
 I 
 \- 
 
 z 
 o 
 
 o 
 
 UJ 
 CO 
 
 UJ 
 CO 
 (T 
 
 UJ 
 
 > 
 
 CO 
 
 z 
 < 
 
TYPES OF VESSELS — SECTION II. 1 65 
 
 where it is desired to use these tanks for other purposes than water-ballast, — 
 bunker coal, stores, or cargo, — watertight hatches may be fitted. 
 
 M'Glashan's Patent Ballast System. — The inventor of this system of 
 ballasting, realizing the uusnitability of the ordinary double bottom tank as 
 a sole means of ballasting, has devised the system illustrated in figs. 87, 88, 
 and 89, whereby, in addition to a cellular double bottom which is of the 
 ordinary mode of construction, side tanks are built into the vessel's structure 
 on the sides over the middle length. The extent and construction of these 
 tanks are fully illustrated in the profile, midship section, and deck plans, 
 together with the notes given upon the same, in figs. 87, 88, and 89. 
 
 In the vessel we illustrate, the side tanks extend from the cellular 
 double-bottom tank margin plate to the upper deck. The breadth of each 
 side tank ii about 27 in., and they extend in a fore and aft direction over at 
 least one-half the vessel's length amidships. The framing in the way of 
 these tan cs is somewhat similar to the web frame system, solid plate web 
 frames, lightened by manholes, being fitted to the alternate frames. The 
 longitudinal framing on the vessel's sides is composed of web stringer plates 
 fitted in- ercostally between the web frames. The upper deck beam knees 
 aie riveted to the inside tank plating by means of angle bars. In addition 
 to the increased and imj^roved facility for the carriage of water-ballast in this 
 vessel, several other advantages are gained. 
 
 1st. Over the midship half length, the vessel possesses an inner skin 
 from gunwale to gunwale. This may prove of immense value to the safety 
 of the ship in the event of the outer skin being perforated, owing to collision 
 or other cause. 
 
 2nd. The additional strength which is manifestly given to the structure 
 by two vertical skins of plating on each side, supported and united as they 
 are by the transverse and longitudinal web framing, fully compensates for 
 the omission of the lower deck beams in the way of these tanks. This, 
 in its turn, improves the conditions for stowage of cargo. The capacity 
 of the hold space is somewhat reduced, but it should be remembered that 
 the tonnage also suffers reduction, as it is only measured to the inner tank 
 side plating, and it is also noteworthy that a vessel built on the ordinary 
 web frame system suffers from broken stowage of cargo, owing to the 
 web framing, which projects 15 or 18 in. from the shell plating into the 
 hold. The construction of the vessel forward and abaft of the side tanks 
 is of the usual kind. A most important point in such tanks as these 
 is that free access be readily gained to every part of the interior, hence 
 manholes fitted with watertight plate covers are cut at intervals through 
 the deck plating. (See plan, fig. 88.) And in order that access may be 
 obtained from the hold or machinery space, manholes with similar water- 
 tight covers are cut through the inner tank side plating. In addition, as 
 previously pointed out, the transverse web plates are lightened by numerous 
 manholes which facilitate the means of passage from one part of the tank 
 to another, and the longitudinal web stringers also have manholes cut 
 
I 66 STEEL SHIPS. 
 
 through them for a similar reason. This is absolutely essential in order 
 that the tanks may be frequently, and as easily as possible, examined 
 inside, and measures taken for their maintenance. In other respects, 
 these vessels are similar to an ordinary cargo or passenger steamer. 
 
 The foregoing are the principal arrangements for water-ballast in use 
 at the present day. 
 
CHAPTER VII. 
 
 DETAILS OF CONSTRUCTION. 
 
 Rivets and Riveting — Butt Straps and Butt Laps — Keel Blocks and Launching "Ways — 
 Frames, Reverse Frames, and Floors — Beams — Pillai's — Keelsons and Stringers — 
 Bulkheads — Decks — Outside Shell Plating — Stern Frames and Rudders — Mis- 
 cellaneous Details : Continuity of Strength, Engine and Boiler Space, Masts and 
 Derricks, Panting, Hatches, Deck Houses, Poop and Bridge Front Bulkheads, 
 Tunnel and Casings, Breast Hooks, Bilge Keels — Ventilation — Pumping — 
 Launching. 
 
 Rivets and Riveting. 
 
 Quality of Riveting. — However carefully and perfectly the ship designer 
 may determine the scantlings and arrange the material for the construction 
 of a vessel, and however excellent may be the quality of the material, 
 unless the innumerable plates and bars which go to make up the whole 
 hull are most efficiently united to one another, it is impossible for the 
 structure to develop its full strength without fracture at some of the most 
 heavily strained connections supervening. Thus, while thoroughly efficient 
 butt and edge connections should be arranged for in so far as this is 
 obtained by the amount of overlap or size of straps and number and 
 diameter of rivets, everything depends entirely upon the quality of the 
 riveting. Bad riveting utterly frustrates all ingenuity displayed in the 
 design. In short, good workmanship is of the very highest importance in 
 shipbuilding, and too great a stress cannot be laid upon this feature. Bad 
 riveting has put many a shipowner to enormous expense, owing to the 
 delay and cost of effecting repairs, and it is not altogether new to hear of 
 huge ocean liners being put into dry dock in order to have their shell 
 plating partially or entirely re-riveted. Carelessness in marking off and 
 spacing rivet holes, and also in the operation of punching, invariably 
 produces what are known as Uind holes. These occur where the rivet 
 holes, in overlapping plates or bars, are only partially over one another. 
 A most objectionable practice under such circumstances is to force a 
 passage for the rivet by means of a drift punch. Such a method should 
 be rigorously forbidden. The proper way in which to clear away 
 
1 68 STEEL SHIPS. 
 
 any obstruction caused by blindness should be by means of a rimer (a 
 drill). 
 
 The tightness of rivets is tested by tapping the side of the rivet head 
 with a small hammer, while two fingers of the other hand rest against the 
 opposite side of the rivet head. An experienced workman can immedi- 
 ately detect a loose rivet. Loose rivets should never be caulked tight, 
 but renewed and re-riveted. Sometimes, in testing a tank, or a water- 
 or oil-tight compartment, a tight rivet may show signs of leakage ; the 
 only rivet which under such circumstances admits of being satisfactorily 
 caulked watertight is the plug-head rivet, all others should be re-riveted. 
 It is always more difhcult to obtain tight work where the riveting is 
 more than two-ply. Thus, as far as practicable, two-ply riveting should 
 not be exceeded. 
 
 A number of different forms of rivets are used in the work of 
 shipbuilding for watertight or oil-tight work. Rivets Nos. 6, 5, and 3 
 in fig. 90 cannot be surpassed for general efficiency. Nos. 6 and 5 are 
 known as pan-head rivets, and ISTo. 3 as the plug-head rivet. Pan-head 
 rivets with snap points (No. 4) are not so reliable for watertight work. 
 
 Pan-head Rivets. — It will be observed that the pan-head rivet is of 
 conical form just under the head, or, to use the common phrase, it has 
 a swelled neck. Tlie reason for this is that when a plate or bar is 
 being punched by a punching machine, the hole so punched is of conical 
 form, increasing in diameter in the direction in which the punch passes, 
 and as plates and bars are always punched from the faying surfaces 
 (surfaces which have to lie against one another) the neck of the rivet 
 is swelled so as to completely fill the hole in the plate or bar. In 
 such places in the construction of the vessel where the work must 
 be watertight, but the appearance of the rivet point is not of great 
 importance, undoubtedly the best result can be obtained by beating 
 down the point as shown in l^o. 6, fig. 90, especially if the rivet hole 
 be countersunk. 
 
 A countersunk rivet hole is one which has more taper than is 
 produced simply by punching. It is formed by taking the punched plate 
 or bar to a machine with a countersinking drill, which gives the bevel 
 required. 
 
 In some parts of the ship structure, however, it is necessary that the 
 rivet point be as near flush as possible with the plating, as, for instance, 
 ill the outside shell plating, steel or iron decks, and top of double- 
 bottom tanks, In such cases, the rivet holes must be well countersunk 
 and the rivets beaten down, any surplus material in the rivet being 
 chipped 00^, and the point finally beaten so as to present a slightly full 
 or concave appearance. This is shown in Nos. 2, 3, and 5, fig. 90. The 
 pan-head rivet with the swollen neck and countersunk point can be 
 highly recommended for great holding power. The heads are laid up 
 with facility, while the rivets themselves are well adapted for entirely 
 
DETAILS OF CONSTRUCTION. 1 69 
 
 filling the holes. When the riveting has been satisfactorily performed, 
 reliable water- or oil-tight results are assured. 
 
 Plug-head Rivets. — Plug-head rivets are also capable of producing 
 highly satisfactory oil- or water-tight work, though greater care is 
 requisite in performing the riveting. As shown in No. 3, fig. 90, the head 
 of this rivet is of conical form, and it is intended to fill a rivet hole 
 which has been countersunk. In laying up the rivet head, though the 
 hole should be completely filled, the rivet head must i)roject at least 
 •^ths or ith of an inch beyond the plate. 
 
 When this rivet has failed, it has generally been brought about by 
 
 NP 4 N°5. /V*? e. 
 
 Fig. 90.— Forms of Rivets. 
 
 allowing too little projection, with the result that when heavy stresses 
 have been experienced, the rivet head has not possessed sufficient holding 
 power, and it has been torn through the rivet hole. However, where 
 care is exercised, and rivets with insufficient head projection are 
 condemned and renewed, excellent results are obtained. It is evident 
 that this rivet is more successful where the thicknesses of the plates or 
 bars which it connects are considerable, as it depends for its holding 
 power very largely upon the depth of the countersinking. Where the 
 material connected is thin, the pan-head rivet is preferable. The point 
 of the plug-head rivet may be formed similarly to the methods described 
 for the point of the pan-head rivet. Pan-head rivets have been used in 
 
I70 
 
 STEEL SHIPS. 
 
 practically every part of iron and steel ships. The plug-head rivet has 
 found special favour for the inner bottom plating of double-bottom 
 tanks, certain parts of the shell plating, oil- and water-tight bulkheads. 
 
 Snap-head Rivets. — Another form of rivet used is that known as the 
 snap-head rivet. (See Nos. 1 and 2, fig. 90.) As the semi-spherical 
 head is very neat in appearance, this rivet is usually adopted in such 
 internal parts of the vessel as are exposed to view, and where it is 
 desired to give a finished appearance, such as in engine-room bulkheads, 
 engine and boiler casings, etc. It is preferable that the rivet neck be 
 swelled as shown in Nos. 1 and 2, fig. 90, thus better filling the hole. 
 The point of the rivet may be either beaten down, or finished off flush 
 with the rivet hole well countersunk. Sometimes, however, to form the 
 point like the head is desired. This is done by placing a 
 snap cup over the heated rivet point, after it has been 
 put into place, and hammering the same until the head 
 is clenched, and the semi-spherical form obtained. When 
 this work can be performed by machinery (hydraulic, 
 etc.), satisfactory results are obtained. But where the 
 riveting is done by hand, the results are not always 
 efficient, the tendency being, in hammering the snap cup 
 over the rivet point, to press the edges of the rivet 
 close, and to leave a hollow all round under the newly- 
 formed head. At first detection is not easy, as the 
 riveting appears to be quite tight, but should water 
 eventuallj'' find its way under the newly-formed rivet 
 head, corrosion takes place round the edge of the rivet 
 and in the plate upon which it bears, and the result 
 is that in the course of time the rivet becomes loose. 
 Owing to the uniform and enormous pressure obtained 
 in machine riveting, the objectionable tendency just 
 referred to does not exist. In any case, snap riveting 
 is greatly inferior to that previously described for pan 
 and plug-head rivets. As is apparent, it is much more difficult to lay 
 up a snap rivet head, especially if performed by hand. 
 
 By laying up a rivet head is meant the operation of holding on to the 
 rivet head by means of a heavy hammer, while the point is being beaten up 
 and finished off, during which time the holder-on works his heavy hammer 
 round the rivet head, thus thoroughly bringing it into close contact with 
 the plate or bar. 
 
 Tap Rivets. — Another form of rivet used in ship work is the tap rivet. 
 Its form is shown in fig. 91. It will be seen that it has a conical head with 
 a thread turned upon the length of the rivet below. Upon the top of the 
 head there is a rectangular projection. This rivet is used in places where it 
 is difficult to get to the back of the material being riveted, in order to hold 
 on, and finish off the point; or where it is desired to connect a compara- 
 
 FiG. 91. 
 
 Tap Rivet. Plan 
 
 of Rivet Head. 
 
DETAILS OF CONSTRUCTION. 
 
 171 
 
 tively thia plate to a thick bar or forging (stern frame, stem, etc.). The 
 outer plate is countersunk to receive the rivet head, while a thread is put 
 upon the lower portion of the rivet hole. The rivet is screwed into this by 
 means of a key, and when thoroughly closed up, the projection on the head 
 is chipped off. 
 
 Butt Straps, Butt Laps, etc. — The connection between two plates end 
 to eud is made either by fitting single or double butt straps, or by over- 
 lapping the plates. 
 
 Single Riveted 
 Butt Lap. 
 
 N° I. 
 
 Double Riveted 
 Butt Lap. 
 
 N°2. 
 
 o o 
 
 o o 
 
 o o 
 
 o o 
 
 o o 
 
 Treble Riveted 
 Butt Lap. 
 
 N9 3. 
 
 o ' 
 I 
 
 o I 
 
 I 
 
 o , 
 
 I 
 
 o I 
 
 IT 
 
 N9 4- 
 
 N? 5. 
 
 N9 6. 
 
 c 
 
 B, 
 
 A 
 
 A B 
 
 C 
 
 
 lo 
 
 
 
 
 
 
 
 ] 
 
 
 1 
 lo 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 lo 
 
 1 
 
 
 
 
 
 
 
 1 
 
 \ 
 
 ^ 
 
 ' 
 
 
 
 B A 
 
 B C 
 
 Single Riveted 
 Butt Strap. 
 
 Double Riveted 
 Butt Strap. 
 
 Fig. 92. 
 
 Treble Riveted 
 Butt Strap. 
 
 Fig. 92 illustrates single, double, and treble riveted butt lap and butt 
 strap connections. 
 
 Comparison of Butt Lap and Butt Strap Connection. — Rows or strakes 
 of plating such as are found in the shell, decks, etc., are usually con- 
 nected at their edges by overlapping the plates, excepting in the case of 
 yachts where, for the sake of appearance, a flush, smooth surface is 
 
172 
 
 STEEL SHIPS. 
 
 preferred. In all cases, however, the overlap connection, when there are 
 sufficient rivets properly spaced, is more efficient than the single strap 
 connection. An attempt has been made to illustrate this in fig. 93, A 
 and B. In B is shown a double riveted single butt strap ; in A, a double- 
 riveted overlapped butt. Suppose, in each of these cases, a tensile 
 stress is experienced. The direction of the line of stress vpill be that 
 shown by the dotted line vs^hich passes in the direction of the strake of 
 plating between the butts, and at the butts necessarily through the butt 
 strap, which is the binding agent. In any case, when such a stress is 
 borne, the tendency is that it should be exerted in a straight line, which 
 means that the tendency is to bring the butt strap into the same plane as 
 the strake of plating. This necessarily produces an amount of pressure 
 
 Frame 
 
 
 Frame 
 
 r-\ I \ 
 
 M=^ 
 
 u_\ 
 
 :\ 
 
 J -/--x 
 
 C3 r~\ ^-^ r-\ 
 
 3 
 
 1 
 
 Fig. 93. — Butt Lap and Butt Strap connections. 
 
 varying vpith the degree of stress at the back of the butt, which, though 
 thoroughly well caulked before these stresses are experienced, tends to press 
 forward the plating at the butt, and open the butt (as illustrated in an 
 exaggerated form, see D, fig. 93), and produce leakage. Such results 
 have actually taken place in many vessels, especially in the shell plating at 
 the bilge, and it has not infrequently happened that the only way in which 
 this damage could be effectively repaired has been by fitting an extra butt 
 strap on the outside. In this case, the stress divides itself between the two 
 butt straps, producing a pressure upon each side of the butt which 
 neutralizes itself and thereby causes deviation in the plating impossible. 
 This also explains why, in many of the important structural parts of every 
 large vessel (sheer strake, bilge plating, upper deck stringer plate, etc.), 
 double straps are fitted. The advantage of an overlapped butt over a single- 
 
DETAILS OF CONSTRUCTION. 1/3 
 
 strapped butt will be apparent on examining the butts in fig. 93. Hers (A 
 and C), though an equal amount of stress may have to be borne, no tendency 
 to open or destroy the efficiency of the caulking exists. It may be noted, 
 that under both tension and compression, the stress is borne by the rivets in 
 the overlap butt, while, in the case of the strapped butt, the plate ends at 
 the butt bear hard upon each other under compression, and the rivets are 
 thus relieved of the stress, only being stressed under tension. 
 
 Strength of Butt Connections. — To observe a few of the considerations 
 which must be kept in view in arranging the rivets in butt connec- 
 tions will be advisable at this point. Let fig. 94 represent part of 
 the shell plating of an ordinary mercantile steamer whose butts are 
 connected by means of single straps in the three upper strakes, and, for 
 the sake of example, by butt laps in the three lower strakes of plating. 
 In way of every frame, the shell plating must necessarily be perforated 
 with a line of rivet holes round the Avhole girth from gunwale to gunwale, 
 spaced according to the usual practice, 7 to 8 diameters of the rivet apart. 
 Especially is it essential that no butts in adjacent strakes of the outside 
 shell plating should come nearer to each other than at least two frame 
 spaces, nor in alternate strakes than one frame space. This rule has there- 
 fore been observed in fig. 94. 
 
 Now, assuming that a severe shearing stress be experienced, and that the 
 strakes of plating are composed of continuous material without any butts, 
 it follows that rupture could only take place through one of the lines of 
 frame rivet holes, say, through S S. But supposing the butt connection 
 at Y Y to be exceedingly weak, the rupture would then most likely occur 
 by shearing the plate through S X, and shearing all the rivets on one side 
 of the butt strap, and so on down the line of the nearest frame rivets as 
 shown. From this, it will be seen that to attempt to make the butt 
 connection as strong as the unpunched plate would be absurd, for if such 
 were possible, which obviously is not in the application of continuously 
 increasing stress, as we have shown, the plating would ultimately shear 
 through one of the lines of frame rivet holes. The efficiency, therefore, of 
 the butt connection depends upon its strength relatively to that of the 
 strake of plating in which it comes through a line of frame rivet holes. 
 
 The minimum shearing strength necessary for the rivets on one side of the 
 butt Y Y should equal the strength of the plate through the line of nearest 
 frame rivet holes, say, P P. Had the frame spacing been wider, or the butt 
 straps only double riveted, there would have been one or more pairs of 
 rivets in the seams between the frame and the edge of the butt strap. 
 (See, for example, butt laps, fig. 94.) In such a case, before rupture could 
 occur at the butt, not only would the rivets on one side of the butt have 
 to be sheared, but these additional seam rivets also between the frame 
 and the butt strap. The shearing strength of all the rivets on one side of 
 the butt plus the additional seam rivets just referred to, should equal the 
 shearing strength of the plate through the line of the nearest frame rivet 
 
174 STEEL SHIPS. 
 
 holes, or in other words, the minimum shearing strength of the rivets on 
 one side of the butt strap should equal the strength of the plate through 
 the line of nearest frame rivet holes less the shearing strength of the 
 additional seam rivets. Having arrived at the minimum number of rivets 
 required for each side of the butt, they must now be disposed in 
 such a manner as to produce no unduly weakened section in either the 
 plate or the butt strap. A row of closely-spaced rivet holes, however, 
 cannot be avoided on each side of the butt, A A, in fig. 94, this being 
 necessary in order to ensure watertightness. The usual disposition of the 
 rivets in treble riveted butt straps is shown in the three upper strakes in 
 fig. 94, and also in fig. 92, No. 6. 
 
 There are five different ways in which fracture may occur at such 
 connections. (See figs. 94 and 92, No. 6.) 
 
 (1) By the butt strap shearing through A A. 
 
 (2) By the plate shearing through C C. 
 
 (3) By the plate shearing through B B, and shearing the rivets in the 
 row farthest from the butt, viz., C C. 
 
 (4) By the bufct strap shearing through B B, and shearing all the rivets 
 in the line A A. 
 
 (5) By all the rivets shearing on one side of the butt strap. 
 
 Were the row of rivets farthest from the butt spaced as they are in the 
 other two rows, another particularly weak section would be created, for, as 
 we have already stated, by the plate shearing through C C, total separation 
 would occur. These rivets, however, are always more widely spaced. For 
 moderately small vessels, Lloyd's require that every alternate rivet be 
 omitted in the back row. For large vessels, however, a somewhat closer 
 spacing is required, usually about 5 to 5| diameters of the rivet from centre 
 to centre. Were the butt strap of the same thickness as the plating, it is 
 clear that by far the weakest section would be through the line A A, 
 fracture of the butt strap through this line producing total severance of 
 the plate. However, the necessary strength can be provided by increasing 
 the thickness of the strap, and as a result the sectional area of the material 
 through this line. 
 
 The tensile strength of the butt strap through the line of rivet holes 
 A A should equal the strength of the plate through the line of rivet holes C C. 
 The strength of a butt-strapped connection cannot exceed the strength of 
 the plate through the line of rivet holes farthest from the butt. 
 
 Thus, classification societies demand one or several twentieths of an 
 inch more thickness in the butt straps than in the plates they connect, over 
 the vessel's midship length. For a similar reason, double butt straps should 
 be each more than half the thickness of the plates they connect, while at 
 the same time almost double the shearing strength is obtained in the rivets. 
 When overlapped butts are adopted (see three lower strakes in fig. 94), 
 the minimum shearing strength of the rivets is determined as previously 
 described for the rivets on one side of the butt strap. As the rivets must 
 
DETAILS OF CONSTRUCTION. 
 
 175 
 
 be closely spaced in the row nearest to tlie caulking edge of the butt, no 
 advantage whatever is gained by increasing the spacing of the rivets in the 
 row farthest from the caulking edge, whether the butt be double or treble 
 riveted, as the strength of an overlapped butt joint cannot exceed the 
 strength of the plates through- the line of rivet holes nearest to the caulking 
 edge of the butt. In a deck stringer plate, the strength of the plate in 
 way of a line of beam rivet holes regulates the minimum shearing strength 
 of the rivets in a butt joint. 
 
 Clasdfication Societies' Rules for Riveting. — While it is natural that 
 every student of naval architecture should desire to possess an intelligent 
 reason for the methods adopted in practice in ship construction, it may be 
 pointed out that the registration societies give very full and complete 
 information, not only as regards the scantlings of vessels classed with them, 
 but also as regards the size, spacing, and number of rivets to be used in the 
 various connections. And even in the case of vessels which are not classed 
 with any registration society, the practice of the societies, at any rate in 
 regard to the matter of riveting, is, as a rule, very closely followed. Indeed, 
 that is a natural consequence, as Lloyd's Eules for the year 1885 is 
 the standard upon which all British vessels are judged in being assigned 
 a load line. 
 
 For the sake of illustration, we will notice some of the principal 
 requirements of Lloyd's society. 
 
 A rule that should be observed in all rivet work is, that no rivet should 
 come nearer to the edge of any plate or bar than its own diameter. In 
 connecting two plates or two angle bars, or a plate and an angle bar in 
 any important structural part, the greater thickness regulates the diameter 
 of the rivet to be used. The corresponding diameters of rivets for plates 
 or angles increasing in thickness is given in the adjoining table. 
 
 Table of Riveting. 
 
 Thickness of "j 
 Plates in V 
 inches. J 
 
 5 
 
 6 
 
 6-7 
 
 7 
 
 8 
 
 9 
 
 9-10 
 
 10 
 
 11 
 
 12 
 
 12-13 
 
 13 
 
 14 
 
 14-15 
 
 15 
 
 16 
 
 20 
 
 20 
 
 20 
 
 20 
 
 20 
 
 20 
 
 20 20 
 
 20 
 
 20 
 
 20 
 
 20 
 
 20 
 
 20 20 
 
 20 
 
 Diameter ofj 
 Rivets in >- 
 inches. J 
 
 5 
 
 8 
 
 1 
 5 3 3 
 
 8 ; 4 4 
 
 3 
 
 4 
 
 3 
 
 4 
 
 3 
 
 4 
 
 7 
 8 
 
 7 7 
 
 8 8 
 
 7 
 8 
 
 7 
 8 
 
 1 
 
 1 
 
 1 
 
 1 
 
 When the outside shell plating is orr^hs of an inch in thickness and 
 above from the keel to the upper turn of the bilge, and oV^^^ ^^ ^'^ "^^^ 
 and above from the upper turn of the bilge to the sheer strake, the seams 
 or landing edges of the strakes of plating are to be double riveted. The 
 seam on the lower edge of the sheer strake must, in all cases, be double 
 riveted ; for less thicknesses, the edges may be single riveted. In double 
 edge riveting, the space between the rows of rivets must be at least once 
 
176 STEEL SHIPS. 
 
 and a half their diameter. The distance between the centres of the rivets 
 in a fore and aft direction should not exceed 4 to 4^ diameters. 
 
 When the butts of the shell plating are overlapped, the butt must be 
 treble riveted with three complete rows of rivets for at least one-half of 
 the vessel's length amidships. When the 2nd numeral or the plating 
 number is less than 16,000, the overlapped butts at the ends may be 
 double riveted, but when above 16,000, the lap butts are to be treble 
 riveted throughout. 
 
 Between each row of rivets in the butt there must be a space equal to 
 twice the diameter of the rivet. In the other direction, the rivets in the 
 butts must be not more than 3^ diameters apart from centre to centre. 
 
 When the shell butt connections are effected by means of butt straps, 
 and the 2nd numeral is above 28,000, the whole of the butt straps all fore 
 and aft are to be treble riveted. When above 24,000 and not exceeding 
 28,000, they are to be treble riveted for three-fourths the length amidships, 
 and when above 20,000 and not exceeding 24,000, the butts are to be treble 
 riveted for half the vessel's length amidships. Below 20,000, treble 
 riveting is only adopted in the butts for the more important strakes of 
 shell plating, such as the sheer strake, and one or more bilge strakes. In 
 very small vessels the butts need only be double riveted. The rivets 
 through the shell plating which take the frames are spaced from 7 to 8 
 diameters apart, and in order that the frame may not be unduly weakened, 
 when the seams are double riveted by there being two rivets close together 
 through the frames on every seam, one of these is omitted, the one left being 
 the rivet nearest to the caulking edge. (See fig. 94.) 
 
 The butts of deck stringer plates should be at least double riveted, 
 whether they be strapped or overlapped. As vessels increase in size, it 
 becomes necessary to treble rivet the butts, and even to fit double butt 
 straps. The spacing of the rivets in the butts is similar to that previously 
 given for shell butts. The butts of a steel deck are to be double riveted 
 for half the vessel's length amidships, and the seams to be single riveted 
 4 to 4I7 diameters apart. The rivets connecting a steel deck to the beams 
 are spaced 7 to 8 diameters apart. 
 
 In inner bottom plating, the butts and edges of the middle line strake 
 all fore and aft, and also the butts of the inner bottom plating in the 
 engine and boiler space, should in all cases be double riveted. Elsewhere, 
 where the 2nd numeral is 20,000 and under 30,000, the butts of the inner 
 bottom plating should be double riveted for half the vessel's length 
 amidships. In larger vessels it becomes necessary to double rivet both 
 butts and seams for at least half the vessel's length. The butts of outside 
 plating, deck plating, inner bottom plating, are chain riveted. In bar 
 keels, stem bars, and stern frames, double zigzag riveting should be 
 adopted in all vessels. Rivets through keelsons, bilge and side stringei\s, 
 floors, frames, reverse frames, beams, are spaced about 7 diameters from 
 centre to centre. The butt connections of angle bars are effected by fitting 
 
Fig. 94. 
 RIVETING OF OUTSIDE SHELL PLATING. 
 
 (VIEW INSIDE OF SHIP.) 
 
DETAILS OF CONSTRUCTION. 1 77 
 
 either a bosom piece, as shown in fig. 95, or a butt covering bar, as shown 
 in fig. 9, where a heel piece or butt covering bar covers the frame butts on 
 top of the bar keel. 
 
 Keel Blocks and Launching Ways. 
 
 IMethod is a most important factor in the successful management of a 
 shipyard, where possibly several vessels of different sizes are in various 
 stages of construction at the same time. Without a rigorously followed 
 system, a state of chaos would undoubtedly supervene, and great loss 
 would ensue, simply because of the time which would inevitably be 
 wasted. Only those who have had actual shipyard experience know 
 how true this is. 
 
 Though a shipyard be well equipped with first-class plant and 
 machinery, and thoroughly capable workmen and officials, there is a large 
 amount of very important and responsible work to be completed before a 
 single frame of the ship can be hoisted into position on the blocks. 
 
 Bosom Piece. 
 
 /BUTT 
 
 L 
 
 ■^ B OSOM ^ PI ECE 
 
 Fig. 95.— Angle Bar. 
 
 First of all, a suitable berth has to be chosen in which to build the 
 proposed vessel. If the ground has a gentle natural declivity, all the better, 
 as is soon discovered when the work of laying the keel blocks is about to 
 be done. The ground must be firm, so that no sinkage or shifting of the 
 earth takes place when the enormous weight of a large vessel is upon it. 
 To ensure a good foundation, in some cases piles of timber are driven into 
 the earth, and in other cases enormous blocks of timber, 12 in. or more 
 square, are embedded in the earth, so that the uppermost surface is level 
 with the ground. The baulks are laid in a fore and aft direction paralled 
 to the keel line. Having secured a sure foundation, the blocks are 
 arranged upon which the keel is to be laid. (See fig. 96.) 
 
 Everyone acquainted in any degree with ships understands that when a 
 vessel is so far completed as to be ready for launching, what are called 
 * launching ways ' are laid under each side of the bottom between the 
 bilge and keel. (See fig. 96.) 
 
 And naturally, as a vessel is generally launched by the impetus created 
 by her own weight, the launching ways must be laid on such a declivity as 
 will cause the vessel to move by her own weight when she is freed. The 
 declivity given to launching ways is usually about | in. to 1 ft., while 
 
 12 
 
178 STEEL SHIPS. 
 
 the declivity of the keel blocks is made slightly less, say, ^ in. to 1 ft. 
 Consequently, where the ground has only a very slight natural slope, 
 the stem of the vessel gets nearer and nearer to the earth every foot it 
 travels in its journey down the launching ways. It is therefore of vital 
 importance that a sufficient height be given to the foremost keel block, to 
 ensure that the stem will clear the ground just before she makes her final 
 plunge into the water. Considerable damage is sometimes done by the stem 
 striking the dock or riverside in the operation of launching, simp]y because 
 sufficient care was not taken to allow for ample clearance. Suppose the 
 distance from the water edge to the foremost block to be 300 ft. and it is 
 intended to make the declivity of the launching ways |ths of an inch, with 
 a clearance of, say, 1 ft. at the water edge (keel blocks J in. per ft.), the 
 height of the foremost keel block will be (300 x f") + 12" = 16 ft. 7| in. 
 above a horizontal base line passing through the surface of the ground at 
 the water's edge. 
 
 But supposing the ground to have a natural declivity of fths of an inch 
 to 1 ft., then 16 ft. 7J in. less 300 x f" = 7 ft. 3 in., the height of the fore- 
 most block above the ground line. (See fig. 97.) 
 
 Thus the declivity given to the keel blocks, and the height of the 
 foremost block, must be determined in conjunction with the declivity of 
 the launching ways.* See further remarks upon 'Launching,' p. 265. 
 
 Laying Off. — After the plans of a vessel have been prepared in the 
 drawing office, the first man who takes her in hand is the loftsman. His 
 domain is the mould loft, where, after being provided with the ' lines ' plan 
 from the drawing office, or else carefully measured or calculated particulars 
 
 * With a ground declivity of § in. per ft., to lay the keel at a greater inclination 
 than J in. per ft. is quite unnecessary, for if such were done it would make a very 
 high, and consequently a very expensive line of keel blocks and staging. 
 
 If the keel blocks were laid at J in. per ft., the height of the keel blocks would give 
 ample head room for riveting all fore and aft. 
 
 The proper method of determining the slope of keel depends on two things : — 
 
 1st. Launching the ship at such an angle with the water that the buoyancy moment 
 should exceed the tipping moment, viz., Bxd should always slightly exceed GxD. 
 (See fig. 177.) 
 
 2nd. As tlie ground slope is usually less tlian the keel slope, the keel should always 
 be sufficiently high at the after end to permit of good riveting. 
 I in. is a very common 3''ard slope. 
 4 ill- ., u keel ,, 
 
 § in. ,, ,, launching ,, 
 
 These are in no way imperative, but are the average of ordinary practice for ships 
 of about 300 ft. in length. 
 
 Short boats (about 100 to 150 ft. in length) are often launched with way slopes of 
 1 in. to IJ in. to prevent them tipping off the ways. 
 
 Fine-lined ships, large yachts, and small war ships, often have | to 1 in. way slope 
 to prevent tipping, and, in addition, the ways are run further into the river than usual. 
 
 Very often the launching ways are above the bottom of the keel at the fore end. 
 This is done so as to keep the fore cradle as shallow as possible. 
 
 In' this case the foreshore or ground between the ways is dug out. This is a 
 particularly common practice in north-east coast shipyards. 
 
DETAILS OF CONSTRUCTION. 
 
 179 
 
 to 
 
 
 ^ 
 
 
 ^ 
 
 
 :^.r 
 
 to 
 
 ^■^ 
 
 •^.^ 
 
 ^1 
 
 t: "^ 
 
 
 C3 -t^ 
 
 •^S 
 
 <3 ^ 
 
 
 5 
 
 •^^ 
 
 1 
 
 
 !, 
 
 
 •^i 
 
 ^ 
 
 t*^ 
 
 t^ 
 
 ^ 
 
 1^ 
 
 •^ 
 
 
 
 o-is 
 
 .J: 
 
 
 6X3 <n 
 
 
 'O -iJ 
 
 ■^ 
 
 
 <o eS 
 
 1 
 
 03 
 
 Oi r— C 
 
 
 !U 
 
 
 
 CU 
 
 
 § 
 
 ^ 
 
 S 9 
 
 
 \ 
 
 ^ -IS 
 
 S I 
 
l8o STEEL SHIPS. 
 
 taken by the draughtsman from the ' lines ' plan, he proceeds to reproduce 
 the ' lines ' plan upon the loft floor to actual size.* 
 
 In addition to these plans, the loftsman is furnished with numerous 
 detail plans, such as the midship section, giving the scantlings of the 
 material, and drawings of the stern frame and stem bar. 
 
 It is obvious, even assuming that the ship draughtsman has most carefully- 
 designed his ' lines ' plan in the drawing ofiice, and endeavoured to secure 
 perfect agreement between his 'buttocks,' 'sections,' and ' waterplanes,* 
 that when the loftsman ' lays off ' the ship to full size upon the loft floor 
 (in chalk lines upon a blackened floor), numerous discrepancies and 
 unfairnesses may be discovered. The loftsmau's work is to rectify any 
 such irregularities, and to produce perfect harmony and fairness in all 
 lines which make up the form of the hull. 
 
 When this is done, he proceeds to prepare the results of his labour for 
 the workmen who are to manipulate the material for the various parts of 
 the structure. His chief work is to prepare what is called a scrieve hoard. 
 This consists of a rectangular area large enough to take a full-sized midship 
 section of the vessel, made up of stout planks 9 in. or 11 in. wide, which 
 are tightly clamped together. He then transfers from the mould loft floor 
 to the scrieve board the midship section of the vessel (to outside of frame), 
 and afterwards every frame between the stem bar and stern post. By 
 means of carefully-prepared supple pine battens, which he bends round each 
 frame on the scrieve board fixed by means of long steel pins driven into 
 the board alternately on each side of the batten, he cuts or scrieves every 
 frame into the scrieve board with a sharp-edged tool called a scrieving 
 knife. 
 
 He also scrieves in all decks, stringers, keelsons, and floors. The shell 
 plating seams or edge laps are also shown on the scrieve board, but these 
 are usually painted on in white. 
 
 When the scrieve board is completed, the planks composing it are 
 undamped, and carried to the shed or shop in the shipyard near to where 
 the frames and reversed frames are bent and the floor plates prepared. 
 Here the scrieve board is again put together, and secured as before. 
 
 But the loftsman does much more than this. From detail plans 
 supplied to him by the drawing office, he draws the stem bar, stern frame,. 
 
 * The 'lines' plan, which is usually drawn to a scale of ^ in. to 1 ft., consists of 
 three principal plans. 
 
 1. A profile showing the sheer, the form of the stem and stern, the decks and 
 stringers, and the form of longitudinal vertical sections of the vessel at certain intervals 
 from the fore and aft middle line. These sections are called ' buttocks ' in the after body, 
 and bow lines in the fore body. 
 
 2. A body plan showing the transverse sectional form of the vessel at the outside 
 of the frames, taken at intervals in her length from stem to stern. Upon this plan 
 are marked all decks, stringers, keelsons, the floors, and the edges of the strakes of 
 shell plating. 
 
 3. , A plan showing the form of the rail lines, and horizontal sections of the vessel 
 (termed waterplar es) from the keel upwards. 
 
DETAILS OF CONSTRUCTION. l8l 
 
 rudder frame, etc., to full size, and from these, full-sized wooden templates 
 are made, and sent to the makers of these parts whether they be forgiugs 
 or castings. 
 
 In a similar manner, the loftsman is responsible for the making of 
 templates for other special parts of the vessel, and lays off or expands 
 certain parts which need carefully dealing with in detail, such, for instance, 
 as the stern plating round the counter, etc. A very important mould 
 made by the loftsman is the beam mould. The standard ' camber ' or 
 'round up' for midship beams upon weather decks is | in. for every foot 
 in the length of the beam. Thus, for a beam 40 ft. long at amidships 
 the camber at the centre would be 40 x ;| in. = 10 in. 
 
 Let fig. 98 represent the curve of the midship beam. As we travel 
 towards the stem and stern, the length of the beam gradually becomes 
 less and less, until at, say, x feet from the stern it is only 30 ft. The 
 camber is not 30 x ;^ in. = 7| in., for, as is easily seen, this would be an 
 entirely diiferent curve from that at amidships. The curve of every beam 
 on any particular deck is obtained from the beam mould by setting off half 
 its length on each side of the middle of the mould. By this means a uniform 
 curve is preserved from stem to stern. 
 
 Decks below the weather deck may either be perfectly horizontal or have 
 
 Fig. 98.— Beam Camber Mould. 
 
 the usual camber. Generally speaking, camber is given to all decks, though 
 not necessarily of the regulation ' round.' 
 
 With the keel blocks laid, and the scrieve board prepared, the workmen 
 are in a position to proceed with the manipulation of the steel and iron, and 
 the erection of the vessel bit by bit upon the blocks. By the time this stage 
 has been reached, quantities of steel plates and bars will have been received 
 by the storekeeper, and stacked in the yard ready to be handed out as 
 required. 
 
 As the only practicable method is to build from the keel upwards, our 
 attention will naturally be first called to the keel. 
 
 As a comparison of the structural features of the section and profile, figs. 
 12 and 46, with the section and profile, figs. 48 and 49, the former about 
 242 ft. vessel, and the latter 443 ft., it will be seen that large vessels, 
 though possessing a greater amount of material in their structure, are 
 remarkably similar — practically identical, indeed, with small vessels in the 
 detail work of construction. 
 
 While the description and illustration of the details of the construction 
 of the vessel illustrated in figs. 10, 11, 12, and 46 will, to a considerable 
 extent, serve our purpose, we shall not confine ourselves to this particular 
 
1 82 STEEL SHIPS. 
 
 vessel, but when found advisable, turn to vessels of different size or type or 
 design. This, we believe, will be the simplest and most comprehensive 
 course to pursue. 
 
 KEELS. 
 
 The Bar Keel. — The rule dimensions of a bar keel for this vessel 
 (fig. 10) are 9 in. x 2| in., the 9 in. representing the depth and the 
 2f in. the thickness. It is made_ of forged iron, and is ordered in separate 
 lengths, varying from 20 to 50 ft. or more. These lengths are united so 
 as to form a continuous bar extending from the stem bar to the stern post, 
 to both of which it is connected. The connections of the various lengths 
 are made by means of scarphs. The scarph connection of a bar keel to a 
 stern frame is illustrated in fig. 99. The scarphs connecting the several 
 lengths of keel bars to one another, and the stem bar and stern frame to the 
 keel, are in all respects identical. 
 
 Tack Rivets 
 
 °ooo OO O OO lOOOOO'O 
 
 oooooo ooo'*o oo'o o. 
 
 o o o o I \ 
 o o o o dj 
 
 ;-<_ Scarph — ^ 
 
 Lip End 
 Fig. 99.— stern Frame Connection to Solid Bar Keel. 
 
 The length of the scarph (by Lloyd's Rules) is nine times the thickness 
 of the keel, viz., 2 If in. The thin half length of each scarph is called 
 the lip end. Tlie extremities of the lip ends which check into the adjoining 
 lengths are about fths in. or | in. thick. The faces of the scarphs are planed 
 and fitted as closely as possible, otherwise it would be difficult to caulk the 
 bottom edge of the scarph to produce watertightness, which is absolutely 
 essential. 
 
 Bar Keel Eivets. — The diameter of rivets used in the keel is \ in. more 
 than the rivets required for plates of the thickness of the garboard strakes. 
 The rivet holes are drilled Jg-t^i of an inch larger than the diameter of the 
 rivet. This is to allow for the expansion which takes place when rivets are 
 heated, which provision is specially necessary when the diameter is consider- 
 able. According to Lloyd's Rules, the garboard strake for this vessel should 
 be |-iths in., and the table on page 175 shows that the diameter of the rivets 
 required for this thickness is |ths in. The rivet therefore for the keel is 
 
DETAILS OF CONSTRUCTION. 1 83 
 
 I in. + ^ in. = i^ in. diameter, and the rivet hole is 1^ in. + y^ ^"' 
 = If^g- in. diameter. 
 
 The keel is always double riveted, that is, there are two rows of rivets. 
 It is usual for these rivets to be arranged zigzag, as shown iu fig. 99, though, 
 on rare occasions, chain riveting has been adopted. 
 
 The rivets in each row are spaced 5 diameters apart from centre 
 to centre. There should be about twice the diameter of a rivet between 
 the edges of the rows of riveting. A most important rule in arranging 
 the spacing of all rivets is, that no rivet should come nearer to the edge 
 of a plate or bar than a distance at least equal to the diameter of the 
 rivet. Now the garboard strake is never brought to the bottom of the 
 keel, but is kept up about ^ or f ths of an inch. This is in order to permit 
 of more satisfactorily caulking the bottom edge, and also to preserve the 
 caulking in the event of the keel chafing over sandy bottoms or river bars. 
 In some cases, it is found to be necessary to fit chafing plates or shoes over 
 the bottom of the keel to preserve the lower edge of the garboard strake 
 from wear from this cause. It also reduces the possibility of severe strain- 
 ing of the rivets which might arise in the event of the vessel grounding upon 
 the keel with a list, and one of the garboard strakes taking the weight 
 of the vessel. The bottom rows of rivets must be kept at least one 
 diameter of the rivet hole from the bottom edge of the garboard strake. 
 At the same time, to keep the rivet too far from the edge of any plate 
 which has to be caulked is a distinct objection, as it is difficult to satis- 
 factorily close up the seam. 
 
 When the space between the rows is made more than two diameters of 
 the rivet, it sometimes throws the top row of rivets so high that sound 
 riveting is almost impossible. If the top rivets are upon the bend of the 
 garboard strake, the work cannot be closed up, and on looking down 
 between the floors, the rivet necks can be distinctly seen. Such work, 
 needless to say, is most objectionable. 
 
 AU rivet holes in the keel bars are drilled before the bars are fixed in 
 place upon the blocks, excepting those in the lip ends of the scarph, which 
 should be drilled after the keel has been fixed in position on the blocks, 
 so as to avoid any blindness in the holes, which would be particularly 
 objectionable in this locality. Before the frames are erected, and the 
 garboard strakes are put on, the keel lengths are held in position by riveting 
 tliem together by means of about half a dozen f ths in. tack rivets through 
 each scarph, spaced where found most convenient, and as far away from the 
 other rivets as possible. 
 
 The scarphs are caulked before the garboard strakes go on, so as 
 to ensure that no water finds its way into the vessel through the 
 connections. 
 
 The bar keel is held in position on the blocks by means of chocks as 
 shown in fig. 96, until the frames are erected and the garboard strakes can 
 be put on. 
 
184 STEEL SHIPS. 
 
 Side Bar Keel. — la a vessel with a side bar keel, the centre through- 
 plate (centre keelson) is continuous all fore and aft, and extends from the 
 bottom of the keel to at least the height of the top of the floors. (See figs. 
 11 and 135.) The keel proper is usually made the same depth as required 
 for a solid bar keel ; and the total thickness of the centre girder and side 
 slabs combined is equal to a solid bar keel. A most important condition 
 which should be aimed at in the construction of side bar keels is, that the 
 butts of the side bars and the centre through-plate, as well as the butts 
 of the garboard strakes, be kept as far from each other as the lengths of 
 the plates and bars will permit. For the reasons given, in dealing with bar 
 keels the garboard strakes should be kept at least ^ in. from the bottom 
 of the keel. The connection of the side bar keel to the stern frame is 
 illustrated in fig. 100 ; and in an identical manner, the stem bar also is 
 connected to the side bar keel. The riveting in all respects is similar to 
 the riveting previously described for solid bar keels. 
 
 or 
 
 Sfern Frame ' '"<:'" 
 
 Scarph >| ^ide Bar 
 
 Cenfre 
 
 I 
 
 H^ss^ — Through 
 Side Bar ^^^/e 
 
 Fig. 100. — Stern Frame Connection to Side Bar Keel. 
 
 Flat Plate Keels. — A flat plate keel resembles an ordinary strake of 
 outside shell plating, with the exception that it is of much greater thickness 
 than the adjacent strakes. (See figs. 12, 48, and 58.) This is necessary, not 
 only because of the wear and- tear to which this plate is often subject, but 
 because, in conjunction with the centre through-plate, which stands upon and 
 is strongly attached to it, it assists in giving longitudinal strength. A flat 
 plate keel is usually about 36 in. in width for ordinary vessels. The angle 
 bars, by means of which the connection to the centre through-plate is made, 
 are also of exceptional size and thickness. The spacing of the rivets through 
 the plate keel and these angles is about 5 diameters apart. The keel 
 plate has the further assistance of the garboard strakes, which, though 
 less than the keel plate thickness, are greater than the adjacent plating. 
 
 In some cases, a horizontal bar keel, or rubbing piece, as it is sometimes 
 called, is attached to the centre of the keel plate by large rivets passing 
 through both the keel plate and the horizontal flanges of the angles at the 
 bottom of the centre through-plate. (See fig. 48.) 
 
DETAILS OF CONSTRUCTION. I 85 
 
 Frames, Reverse Frames, and Floors. 
 
 Frames. — The frames wliich, in figs. 6 and 10, extend continuously 
 from the top of the keel to the gunwale, or to the poop, bridge, or fore- 
 castle deck stringer plate, are ordered in straight bars, and bent to the 
 required shape in the shipyard by the frame benders. The rivets connecting 
 the frame bar to the reverse bar, and the frame bar to the shell, are spaced 
 7 to 8 diameters apart. All the rivet holes in the frame bar are punched 
 before it is bent to its required shape, excepting those we shall now 
 mention. 
 
 First. — The rivet holes on the turn of the bilge, which are apt to elongate 
 up, owing to the amount of bending which takes place here. 
 
 Second. — The rivet holes in way of all the edge laps or landings or 
 seams of the strakes of the outside shell plating, for as only one rivet is put 
 through the frame in the way of shell landings (excepting in oil steamers ; see 
 p. 158) so as not to unduly weaken the frame, it is most essential that the 
 hole be -perfectly fair, in order to ensure the best workmanship. Although the 
 positions of the laps are marked upon the frames before they are erected, 
 it often happens, in checking the fairness of the fore and aft sight edge lines 
 of these laps upon the ship herself, which is done by pinning a batten upon 
 the frames after they are erected, a slight deviation has to be made 
 from the marks upon the frames. Hence the necessity of omitting the 
 punching of these holes, and drilling them at a later stage when all is fair. 
 
 Third. — The holes for beam knees are left unpuuched also, with the 
 exception of one hole, as, in fairing up the beam or sheer line, one or more 
 beams might need slightly raising or lowering after being temporarily 
 placed in position. The single hole punched in the knee is further 
 necessary in order to attach the tackle for hoisting the beam up. The 
 positions of the shell landings are obtained from the scrieve board by 
 bending a light wooden strip round the particular frame being dealt with, 
 and marking the laps upon it. On allowing the strip to spring straight 
 again, it is laid upon the straight frame bar, and the position of the laps 
 transferred. The space between the rivet hole in one lap and the rivet 
 hole in the next lap is divided off into equal spaces representing, as 
 nearly as possible, about 7 or 8 diameters of the rivet. 
 
 Punching. — It is of the greatest importance in punching, not only the 
 frames, but all plates and bars throughout the vessel, that this be done 
 from the faying surfaces. 
 
 The faying surfaces are those which have to bear against each other 
 when riveted together, as, for instance, the faces of the frame and reverse 
 angle flanges. The reason for this is, that in punching a plate or bar, the 
 action of the punch is to leave a rag edge on the opposite side from that on 
 which the punch enters the material. This would be a hindrance to any 
 other plate or bar bearing close against it, which is absolutely essential to 
 secure watertight and, in all respects, satisfactory work. Moreover, the 
 
t86 
 
 STEEL SHIPS. 
 
 conical form of the punched rivet hole is better filled by the swelled neck 
 of the rivet (see p. 169, Riveting). 
 
 Having punched the frame bar, the next step is to get it bent to shape. 
 Generally, under the same shed in which the scrieve board is laid, there 
 are what are known as the frame furnaces. In front of these furnaces 
 there are laid large rectangular slabs of cast iron, 5 or 6 ft. square, and 
 about 5 in. thick, which are perforated all over their surfaces with round 
 or square holes about 1^ in. diameter. These iron blocks are so arranged 
 as to form a perfectly horizontal surface of sufficient area to take the 
 largest frame in the vessel. 
 
 In order to get the shape of the particular frame to be bent, a mould 
 is made by means of a long strip of iron, called a set iron, which usually 
 measures about 1| in. wide by | in. or fths in. thick. When this has been 
 bent to the shape of the toe of the frame on the scrieve board, it is taken 
 to the cast iron blocks, and pinned down. (See fig. 101.) 
 
 However, not only have the frame bars to be bent to the shape of the 
 section of the vessel, but they have to be bevelled also. Fig. 102 will 
 illustrate this. 
 
 The frame angle has a long and short flange. The size of the frame 
 angle for the vessel in figs. 6 and 10 is 4i x 3 x ^V The short flange 
 always goes against the shell, and the long flange points towards the 
 interior of the vessel, perpendicular to the fore and aft middle line. It 
 will thus be seen that over part of the midship length of ordinary cargo 
 vessels, where the form of the section is constant, the angle produced by 
 the flanges of the frame bar is a perfectly right angle. Towards the ends of 
 the vessel, however, where the hull tapers in to the stem and stern, the 
 angle produced by the flanges of the frames must necessarily change. Did 
 the shell flanges of all the frames point in the same direction, then at one 
 end of the vessel the angles produced would become very obtuse, and at 
 the other, very acute. But with an acute angle, or as it is called, a closed 
 bevel, to get good riveting is most difficult, and in some cases impossible, 
 as the head of the rivet cannot be properly laid up. It thus becomes 
 necessary, in order to get opfen frame bevels at the ends of the vessel, to 
 reverse the frames at one end, that is, the shell flanges of the frame bars 
 point towards amidships from both ends of the vessel. The reversing of 
 the frame takes place at amidships. (See fig. 102.) 
 
 The bevelling of the frame bars can be done either by hand or by 
 machinery. In small vessels, both the bevelling and the bending can be 
 done in one heat, but for larger vessels two heats may be necessary. 
 The amount of bevel to be given to the frames is supplied by the loftsman 
 on a bevel board. 
 
 A possible danger in hand bevelling is tliat the shell flange may be made 
 somewhat concave between the heel and the toe of the bar. This could 
 only conduce to unsatisfactory riveting, hence it is necessary to carefully 
 guard against it, and in some cases to chip off the projecting heel of the bar. 
 
DETAILS OF CONSTRUCTION. 
 
 187 
 
 In machine bevelling, the bar is bevelled from the heel, and this objection 
 exists only to a very small extent. 
 
 To bend the frame, it is drawn out of the furnace after it has reached 
 a full heat, and one end is pinned down to the end of the set iron, 
 while the other is worked round to its shape, and pinned down and 
 kept in position on the blocks by means of iron dogs and pins. (See fig. 
 101.) An important feature in the bending of angle bars is, that in the 
 
 « 
 
 
 Fig. 101. — Frame Bending. 
 
 process of cooling, more contraction takes place at the heel of the bar where 
 the material is thickest than anywhere else. The result is, that in a bar 
 hent from the heel (as is the frame bar), the contraction causes the bar to 
 somewhat straighten itself, producing a reduction in the curvature. Know- 
 ing this, the frame benders, guided by experience, give more curvature to the 
 frame when it is in a heated state than the scrieve board indicates, with 
 
155 STEEL SHIPS. 
 
 the effect that when the frame bar is cool, it has approximately come to the 
 desired shape, — so near ndeed, that a few blows from a hammer, when it is 
 cold, brings it to the correct curvature. It is carefully tested on the scrieve 
 board before any riveting is done. The corresponding frame bar for the 
 other side of the vessel is made in a similar manner, and the two sides 
 tested by laying one upon the other. 
 
 We observed in the last chapter, that the amount of external water 
 pressure on the outside of a vessel is greatest where the immersed girth 
 is greatest, and it naturally follows that greater weight is likely to be 
 carried where the hold capacity is greatest. Therefore more strength 
 is required in the transverse framing in the full middle body of a vessel 
 than towards the ends, where she gradually fines down into the stem and 
 the stern. Lloyd's Rules thus stipulate that for three-fifths of the vessel's 
 length amidships, the frames are to be ^V^^^ thick (fig. 10), and for the 
 remaining one-fifth at each end, they need only be -oVths thick. Similarly, 
 the floors, which are ^V^^is thick for three-fifths length amidships, are 
 gradually reduced down to -^ths at the ends. 
 
 Reverse Frames. — Like the frames, the reversed frame angles come into 
 the shipyard in straight bars. They are considerably smaller than the frame 
 angles, being, for the vessel in figs. 6 and 10, only 3 x 3 x -Jq. The holes 
 for rivets connecting the reversed frames to the frames and floors, and those 
 which take stringers and keelsons, are all punched after the bar is bent. 
 The reversed frame is bent and bevelled in a manner very similar to that 
 described for the frames. The set iron is made to the shape of the toe of 
 the frame from its uppermost extremity down to the bilge where it leaves 
 the frame and takes the form of the upper edge of the floor plate. 
 
 Unlike the frame angle, the heel of the reverse frame comes against the 
 set iron, and in cooling — most contraction taking place there — the tendency 
 is for the bar to bend more than is given to it on the blocks. In this, again, 
 the experience of the frame bender guides him in making due allowance, and 
 when the bar has cooled, and is finally tested upon the scrieve board, a few 
 blows from a hammer is generally sufficient to make any necessary correc- 
 tion. The holes in the unpuiiched flange taking the frame are transferred 
 from the frame by laying the frame bar upon the reverse bar, and marking 
 the rivet holes by means of a whitened wooden cylinder which is pushed 
 through each rivet hole. The remainder of the holes connecting the reverse 
 frame to the floors are spaced 7 to 8 diameters apart, and punched. Holes 
 for keelsons and stringers are always omitted, and punched or drilled at a 
 later stage. (For heights to which reversed frames extend in one, two, 
 three deck, spar deck, and awning deck vessels, see Chapter VI., pages 
 110 to 123, 'Types of Vessels.') 
 
 In the engine and boiler space of steam vessels, double reversed frames 
 should be fitted to every floor, and extend at least from bilge to bilge. In 
 figs. 6 and 10 the reversed frames extend alternately to gunwale and to 
 the top of the hold beam stringer angle ; and in the way of the engine and 
 
DETAILS OF CONSTRUCTION. 1 89 
 
 boiler space, the double bars extend from upper bilge stringer to upper 
 bilge stringer. 
 
 The spacing of the rivet holes in the flange connected to the frame 
 angle will naturally be the same as the spacing in the frame (7 to 8 
 diameters of rivet) from which they were transferred. The rivets in the 
 same flange taking the floors are spaced 7 to 8 diameters also. The rivet 
 holes in the other flange taking the ceiling and sparring must be spaced to 
 suit the width and arrangement of the battens. 
 
 Floors. — In the case of very small vessels, the floors are usually in one 
 piece from bilge to bilge. But in larger vessels this is impracticable, and 
 the plates forming the two parts of the floor are lapped on one and the 
 other side of the centre line alternately. The lap must be treble riveted 
 (see fig. 6) or, if the butt be strapped, double straps treble riveted should 
 be used. 
 
 As we have previously observed, the floors extend up the bilge in a fair 
 curve to a height above the top of the keel of twice the depth of the floors 
 at the middle line. But as it adds very materially to the cost of plates to 
 have them cut hollow to shape by the manufacturer, the custom in all 
 shipyards is to order all floor plates so that the top edge is a perfectly 
 
 Fig. 103.— Floor Plate as ordered. 
 
 straight line. Indeed, only rarely are any plates ordered with a hollow 
 in them, for the reason given. What the draughtsman does in ordering 
 floor plates is to expand the floor with the top edge perfectly straight, 
 and in order to minimise waste, he orders the floor plate somewhat 
 like the plate shown in fig. 103. The first operation, then, in dealing 
 with the floor plate, after it has arrived on the shipbuilder's premises, 
 is to curve its top edge to the shape of the particular floor on the 
 scrieve board. Here, again, a set iron is bent to the required shape, and 
 pinned on the iron bending slabs. The floor is put into the frame furnace 
 and heated to a white heat. It is then brought out of the furnace, bent 
 round the set iron, and pinned down with dogs. It is hammered when 
 necessary, and care taken to prevent buckling. As the floor plate is 
 inserted between the frame and reversed frame at the bilge — that is, 
 just where the frames and the reversed frames diverge — the floor plate is 
 heated again at the ends, and hammered out so as to produce a gradual 
 wedge-shaped taper. (See fig. 104.) 
 
 After the floor has cooled, the next step is to curve its outer edge to 
 suit the frame. It is taken to the scrieve board, its upper edge is adjusted to 
 its own floor curve, and the centre line over the keel is marked upon it. Its 
 own frame, which is already bent, is then laid upon it, and the curve of 
 
190 STEEL SHIPS. 
 
 its outer edge marked with chalk. At the same time, the rivet holes are 
 transferred from the frame by the whitened wooden cylinder. In a 
 similar way, by laying the reverse frame upon the floor, the rivet holes are 
 transferred. The floor plate is then carried away and sheared — not, how- 
 ever, to the exact frame edge line, but about ^ in. from the frame edge, so 
 that when the floor attached to its frame is in position in the ship, there 
 is no doubt about the frame flange bearing hard upon the shell (see fig. 
 104). According to the 1st numeral of this vessel (figs. 6 and 10) the depth 
 of the floors on the keel should be 22|- in., and at three-fourths of the half- 
 
 221 
 breadth from the middle line the minimum rule depth -~ = 11|, and the 
 
 Reverse Frame. 
 
 J ^■■■_ ITTTl 
 
 Floor Framed 
 
 Fig. 104.— Showing Floor at Bilge. 
 
 height from the base line — top of keel — to the top of the floor at the 
 extremities should be Vl\ x 2 = 45 in. 
 
 Both the frame bars and the reversed frame bars for the two sides 
 of the vessel butt at the middle line — the frames on the top of the keel, 
 and the reversed frames immediately under the centre keelson. On 
 account of the unavoidable interruption in the continuity of these parts 
 of the transverse framing, butt straps must be fitted. Thus, covering 
 the frame butt, we have a piece of angle bar 3 ft, long, of the frame 
 size, .riveted on the opposite side of the floors (figs. 6 and 9). This 
 butt strap, or heel piece, as it is called, not only affords strength at 
 
DETAILS OF CONSTRUCTION. 
 
 191 
 
 the butt, but offers an additional means of securing a good connection to 
 the garboard strakes, wliich in turn carry the bar keel. This is an 
 important point, as the only connection which the bar keel has to the 
 vessel itself, is by means of the garboard strakes to which it appears to 
 hang. Hence the bar keel is sometimes termed a 'hanging' keel. The 
 reversed frame butt strap, or lug piece, is also fitted to the opposite 
 side of the floor plate (figs. 6 and 9). Not only does it perform the 
 function of a butt strap, but it also provides a double means of 
 connection between the largest and the chief keelson in the vessel to the 
 transverse framing (see figs. 9 and 10). Indeed, wherever keelsons or 
 stringers cross over the transverse frames, the connection should be made 
 by means of these double reversed bars or lug pieces, even though no 
 butt occurs in the ordinary reversed frame, and the lug should be long 
 enough to take three rivets. (See fig. 105.) 
 
 Like the frames, and for the same reason, the floor plates are thicker 
 over the three-fifths length amidships (from ^th to o^^yths) than at the ends. 
 
 r4 
 
 t 
 
 A 
 
 l ojlioj Keelson \0\ 
 
 W, 
 
 Angles |o||i o! 
 
 Ji 
 
 -Floors- 
 
 [1 
 
 lOi 
 
 D 
 
 IT 
 
 O'Mo' I 
 
 o!oi i 
 
 D 
 
 Fig. 105. — Showing Lugs upon Floors for Keelson Connection. 
 
 The special requirements of transverse framing in way of the engine 
 and boiler spaces will be dealt with at a later stage under that heading. 
 
 Erection of Transverse Framing. — When the frames, reverses, floor- 
 plates, and beams have been prepared, and their accuracy tested upon 
 the scrieve board, they are carried to a staging at the stem of the 
 vessel, and riveted together with or without the beams. They are then 
 hoisted into their respective positions indicated upon the keel, and first 
 of all (remembering the declivity of the blocks) given such a rake 
 (angle of inclination) as will bring them perpendicular to the keel. 
 This operation is termed ' plumbing.' The process of adjusting the 
 frames, so that they cross the keel at perfect right angles, is called 
 'horning.' In a vessel with a double bottom, this exi)editious system 
 of erecting complete transverse frames cannot be carried out, the double 
 bottom having to be erected first in the order of keel, centre through- 
 plate, floors, tank margin plate, and the frame legs hoisted into position 
 after the framing of the double bottom has been practically completed. 
 
 Double-bottom Framing. — The details of the transverse framing in 
 
192 STEEL SHIPS. 
 
 vessels having double bottoms with continuous centre through-plates standing 
 upon flat plate keels have already been fully described in Chapters IV. 
 and VI, (see Water-ballast Arrangement). It might be noticed in 
 passing, however, that Avhere the floors are solid from centre through- 
 plate to tank margin plate, the angle bar upon the lower edge of' the 
 floor plate by means of which the connection is made to the shell, need 
 not have flanges larger than the smaller of the ordinary frame flanges. 
 But where floors come upon alternate frames, as shown in fig. 85, with 
 brackets only upon alternate frames, the frame angle in the tank should 
 be of the full size. Both frame and reverse frame angles upon solid 
 floor plates are sometimes dispensed with, the connection to shell and 
 inner bottom being made by flanging the floor plates. (See fig. 64.) 
 
 Where side bar keels with continuous centre through-plates from 
 bottom of keel to tank top are adopted, as shown in figs. 11 and 135, 
 it is common for the frame angle in the tank to be continuous from 
 margin plate to margin plate. In such case, a triangular hole must 
 be notched out of the centre through-plate immediately above the 
 keel, to allow the frame bar to pass through. In the erection of this 
 form of double bottom upon the blocks, the centre through-plate and 
 
 1 .4 . *" ■ «' III ! 
 
 r T 
 
 Fig. 106. — Bulb Angle Frames with large Lug for Stringer Connection. 
 
 side bars are first fixed in position, followed by the tank frame bar 
 which is held up by means of shores, and after this, the floors with 
 reverse bars and angles for the connection to the centre through-plate 
 and tank margin plate are hoisted up and riveted to the frame. 
 
 Z Frames. — While the Z frame section of frames is preferable to the 
 ordinary frame and reverse, its costliness, and the increased difficulty in 
 bending and bevelling, has made its adoption comparatively rare iu 
 merchant shipbuilding, but in the Royal Navy it is extensively used. 
 Fig. 64 is framed with Z bars. 
 
 -Bulb Angle Frames (in lieu of frame and reverse). — When bulb angle 
 frames are adopted (and also Z frames), it is usually in vessels with 
 double bottoms ; the shorter lengths of frames terminating, as they do, 
 at the tank, renders the operation of bending and bevelling more easy 
 of accomplishment. One noteworthy point about the bulb angle frames 
 is that they have no fore and aft fiange to which the stringers and 
 sparring in the holds can be attached. This difficulty, however, is got 
 over in the case of the stringers by fitting to that side of the transverse 
 flange away from the bulb, a lug with a flange large enough to take 
 double rivets. (See fig. 106.) The sparring is fixed by means of cleats. 
 
DETAILS OF CONSTRUCTION. 1 93 
 
 Channel Bar Frames. — This excellent section, combining both frame 
 and reverse bar, is now extensively adopted, especially in large vessels with 
 a long uniform midship body for which the channel frames require no 
 bevelling. 
 
 The difficulty of both bending and bevelling channel bars for frames, 
 and the still greater difficulty of performing satisfactory riveting where 
 much bevelling takes place, is sufficient reason for dispensing with them at 
 the ends of vessels. Though channel frames may be used in vessels with 
 ordinary floors, yet, like Z frames, they are usually adopted for the frame 
 legs only in vessels with double bottoms. In some cases, extra deep 
 channel frames have been used as compensation for an omitted tier of 
 hold beams. 
 
 Deep Framing. — While the two large angles which go to make up what 
 are termed deep frames take the place of the ordinary frame and reversed 
 frame, their chief function, as has been 
 pointed out in Chapter IV., is to com- 
 pensate for a tier of hold beams which are , 
 dispensed with. Instead of fitting together Lug foRj Frame I '^'^Shell 
 the two angles in the same fashion as is ^^^i^g^i^ ^s — 
 usual for the ordinary frame and reverse — /?e verses/ [ 
 the toe of the frame being flush with the Frame I 
 heel of the reverse frame — the connection 
 is made by means of a 3 in. lap only, in Fig. 107.— Inadvisable method of 
 which case it is preferable to fit the angles fitting the Reversed Frame in 
 together as shown in the detail section in ^^^P Framing, 
 fig. 12. The lug piece which is usually fitted in order to get a good, 
 connection between the stringers and the transverse frames, is placed on 
 the back of the reversed frame (see fig. 12). This is better than fitting the 
 frame and reverse as shown in fig. 107, where not only is a piece of 
 packing iron required, but satisfactory riveting is with difficulty obtained. 
 Figs. 12 and 48 are framed on the deep frame system. 
 
 Like channel frames, deep framing is usually adopted for the frame legs 
 in vessels with double bottoms. The bending and bevelling is done as 
 described for ordinary frames. 
 
 Weh Frames. See also Chapter IV. 
 
 The width and spacing of web frames in lieu of a tier of hold beams is 
 governed by the depth of the vessel, while the thickness of the plates is 
 usually similar to that of the ordinary frames. In conjunction with web 
 frames, web stringers of the same width as the web frames are fitted, 
 intercostally between the web frames. The number of these also depends 
 upon the depth of the vessel. 
 
 A web frame is made up of one or more plates connected by treble 
 riveted laps or straps. The connection to the shell is made by an angle 
 of the frame size, and on the inner edge, by two angles of about the 
 reversed frame size or one equivalent angle. 
 
 13 
 
194 STEEL SHIPS. 
 
 A section through the web stringers is shown upon fig. 11. It will be 
 noticed that they are connected to the shell by an intercostal angle between 
 the frames, and to the reverse frames by an angle which is intercostal 
 between the web frames. On the inner edge of the web stringer plate 
 are two small angles which may be substituted by a single angle as for 
 web frames. The connection of the web stringer to the web frames is made 
 by double angles, and also by diamond plates. (See midship section, fig. 
 48,) The connection of the web frames to the tank side is made by 
 double angles also. 
 
 In order to make the transverse strength afforded by web frames as 
 continuous as possible right round the transverse circumference of the hull, 
 extra strong through beams with deep knees are fitted to the head of the 
 web frames where practicable. 
 
 Figs. 11, 67, and 71 are midship sections of vessels built upon the web 
 frame system. The sectional profile in fig. 70 will afford a good idea of the 
 general arrangement of web frames and stringers. 
 
 Beams. 
 
 As a general rule, it may be stated that under iron or steel decks it 
 is preferable to fit beams to every frame ; that is, when the frame spacing 
 is somewhat similar to what would be required according to Lloyd's Eules. 
 They may consist, in small vessels, of plain angles, and in larger vessels 
 of bulb angles or channel bars. The reason for this is, that if, with an 
 average frame spacing of, say, 24 in. in a vessel 250 ft. in length, the 
 beams were placed upon the alternate frames (4 ft. apart), the comparatively 
 thin iron or steel deck, which structurally would be sufficient for such a deck, 
 would possess the tendency to fall hollow between the beams, which is most 
 objectionable and unsightly. To place beams upon alternate frames under 
 steel decks should only be attempted in very large vessels with thick deck 
 plating. Indeed, Lloyd's Rules distinctly state that while it is preferable 
 that beams be fitted to every frame under steel decks, it is only when the 
 steel deck exceeds -^ths of an inch in thickness, that beams on alternate 
 frames are permitted in their classification. 
 
 The steel upper deck of the vessel shown in fig. 12 is 2^0 ths in. in 
 thickness, and the beams under it are placed upon every frame (24 in. 
 apart). In figs. 48 and 49 the uppermost or shelter deck is only -^QtYiS, in. 
 in thickness, and the beams under it are again upon every frame, but the 
 upper and main decks, which are of steel ^ths in. in thickness, have their 
 respective beams on every alternate frame. Where a system of widely 
 spaced frames is adopted, as has been pointed out in figs. 72 and 73, in which 
 the spacing is 36 in., the beams under steel decks must of necessity come 
 upon every frame. 
 
 As the tendency to fall hollow between the beams in vessels with wood 
 decks only practically does not exist, it is usual to fit beams of stronger 
 section on alternate frames. They may consist of large bulb angles, tee bars, 
 
DETAILS OF CONSTRUCTION. 
 
 195 
 
 bulb tee bars or butterly bulbs, bulb plates with double angles on top, or 
 channel bars, varying in dimensions according to the greatest breadth of 
 the vessel. 
 
 Like the frame and reversed frame angles, the beams come into the 
 shipyard in straight bars. But, as we have seen, all beams on weather 
 decks, at any rate, should have a ' camber ' or ' round up ' of ^ in. to every 
 foot of midship beam. This result is obtained by cold bending in a 
 ' Squeezing Machine.' 
 
 The length of every beam in the ship's deck is obtained from the 
 mould upon which they are indicated. Having cambered the beams, and 
 cut them to their respective lengths, the next step is to prepare the knees. 
 
 Lloyd's require that the depth of beam knees be at least two and a half 
 times the depth of the beam. The knees may be either welded or 
 bracketed. 
 
 The welded knees are generally made by cutting off part of the bulb, 
 
 /C 
 
 Plan showing the Beam fitted 
 into the Bosom of the Frame, 
 
 Fio. 
 
 108.— Bulb Angle Beam 
 with Welded Knee. 
 
 Deep 
 Framing 
 
 Plan. 
 
 Fig. 109.— Bulb Angle Beam with 
 
 Bracket Knee. 
 
 and welding to its extremities a piece of plate of the thickness of the beams 
 (see fig. 108), where the welded plate is shown by hatched lines. When 
 the welding is well done, a stronger knee is formed by carrying the bulb 
 round the edge of the knee. 
 
 Bracket knees, as shown by hatched lines in fig. 109, consist simply of 
 bracket plates riveted to the back of the vertical flange of the beam. In 
 both these cases, the knees fit into the bosom of the frames (see plan in 
 sketches). Beam knees of this size (14 in. in depth) should have, at least, 
 four |-in. rivets connecting them to the frames, but not more than two rivet 
 holes should be punched in any beam knee, or in the frame before the beam 
 has been hoisted into place and adjusted. 
 
 Knees for plain angle beams, such as would be used for upper decks in 
 smaller vessels, and for the poop and bridge of the vessel illustrated in 
 figs. 12 and 46, are made in exactly the same manner as described for bulb 
 angles. 
 
196 STEEL SHIPS. 
 
 Having prepared the kneea, the spacing of the rivet holes in the horizontal 
 flange of the beam taking the deck is the next operation. This is a more 
 important work than one would at first imagine, and requires very consider- 
 able care. The general rule is, that rivets connecting an iron or steel deck 
 are spaced from 7 to 8 diameters apart, but, as we shall show, the regularity 
 of any such spacing may be greatly interfered with. Reference to the deck 
 plan of the vessel in fig. 143 will help to make this clear. On the deck 
 stringer plate, and round all deck houses, hatches, casings, etc., there are 
 angle bars forming the connecting means between the deck and the sheer 
 strake, or deck houses, or hatch coamings, etc., as the case may be. The 
 positions of all such bars as these are supplied to the outside workmen by 
 the ship's draughtsman, either on a deck plan, or on what is called a beam 
 list. Then, again, care has to be taken to set off all the rivet holes in the 
 edge laps of the strakes of deck plating. After precautions have been 
 taken to carefully mark off the positions of these particular rivet holes, 
 the remaining rivet holes are spaced, as nearly as possible, 7 or 8 diameters 
 apart. The holes are then punched. 
 
 The midship section, fig. 48, shows a section of the deck through a 
 hatchway, and will further illustrate the points just mentioned. See also 
 fig. 115 (channel beam). 
 
 When a wood deck is laid upon the beams, then, in addition to marking 
 off the positions of deck angle bars, tie plates, etc., the holes in the deck 
 flange or flanges of the beam through which the wood deck bolts pass, 
 must be arranged to suit the width of the deck planking. Wherever 
 beams are fitted to every frame or alternate frames, they must be tied 
 together so as to preserve the proper spacing between them, and thus keep 
 them in their correct positions relatively to one another. When a steel 
 or iron deck is laid upon the beams, the deck itself admirably performs 
 this function, but when only a wood deck is laid upon the beams, continuous 
 iron or steel tie plates, laid in a fore and aft direction across the beams, 
 assisted when necessary by diagonal tie plates, are required. These tie 
 plates vary in width from 6 in. to 30 in. or more, and in addition to 
 serving the useful purpose just mentioned, act like a steel deck in 
 distributing stresses thrust upon the deck in any locality (from masts, etc.) 
 from beam to beam, and to the stringer plates. 
 
 Plates somewhat similar to these tie plates should be fitted to the beams 
 along the sides and round the ends of the hatchways, engine and boiler 
 openings, companion ways, and all other deck openings, so as to get a 
 better connection for the angle bars which go round the coamings of all 
 these openings. 
 
 In the way of all hatchways, the beams necessarily only extend from 
 the frame to the hatch coaming. When the hatchways are very long, and 
 when they come in the middle length of the vessel, the result must be a 
 very serious reduction in transverse strength, unless adequate measures are 
 adopted to compensate for the interruption in the continuity of these trans- 
 
DETAILS OF CONSTRUCTION. 
 
 197 
 
 verse girders. In any case, it is necessary to carefully protect all deck 
 openings, especially when situated upon the weather deck, — these being 
 most exposed to the attacks of shipped seas, etc. This protection is 
 afforded to hatchways by fitting vertical side plates, or hatch coamings, 
 along the sides and ends of the openings, and carrying them to a height 
 of at least 2 ft. 6 in. from the deck. These hatch coamings not only 
 afford protection to the hatch openings, but a means of binding together 
 the beam ends, and provide the compensation previously referred to. 
 
 First of all, we notice in the profile, fig. 46, that at all hatch ends 
 an extra strong deep beam is fitted, consisting of a bulb plate and an 
 angle (or it may be a strong channel bar or some equivalent), instead of 
 the ordinary bulb angle beam. At the hatch ends, the coaming plate is 
 carried down over this beam to the top of the bulb, and securely riveted 
 to it with a double row of rivets as shown in the section of the beam, 
 fig. 110. 
 
 Instead of the side or fore and aft hatch coaming plates extending 
 from the deck only to their prescribed height, they are carried down well 
 
 Match Coaming 
 
 a 
 
 Hatch Battens 
 
 Forei After 
 
 ^ 
 
 I 
 
 Section 
 through 
 Hatch end. 
 
 Deck 
 
 r 
 
 Fig. 110. — Side Elevation of Hatch Coaming plate, showing angle connection to 
 deck plating and collars on beam ends. 
 
 over the ends of all the half beams, or flanged round the bottom of the 
 beams, as shown in figs. 48 and 115. When the hatch corners are round, 
 which is the more preferable method of construction in every respect 
 (except for cheapness), the coamings are continuous all round the hatchway, 
 the separate plates composing the whole coaming being connected by double 
 riveted butt straps. These hatch coamings should be made of good stout 
 plates, -^ths in. or -|^ths in. thick, in vessels of average size. It will now 
 be seen what an excellently strong and suitable arrangement the hatch coam- 
 ings become, upon which to hang, as it were, the ends of the beams which 
 are necessarily cut in way of all hatches. By the hatch end coaming 
 plates being well riveted to the strong hatch end beams, the side coaming 
 plates become exceedingly strong fore and aft girders well able to carry 
 the beams and the deck at the hatch side, that is, of course, in conjunction 
 with the assistance afforded by the hold stanchions, which are fitted along 
 each side of all large hatchways (unless special compensation is introduced 
 
198 
 
 STEEL SHIPS. 
 
 111.— Bulb Plate Turned 
 Knee. 
 
 ■- , - ' , - / 
 
 / ^ y 
 
 
 ^' 
 
 %\ 
 
 X- 
 
 \ ^' 
 
 -' 
 
 
 -^ / ' 
 
 
 (-^^ 
 
 / 
 
 
 \/ ^ 
 
 y 
 
 
 \/^ ' 
 
 
 
 \^/ 
 
 
 for dispensing with them), and well connected to the beams. Indeed, 
 the hatch coamings are formed into a strong, rigid girder framework, 
 well connected to the main structure of the ship at the hatch ends. 
 
 It is when labouring at sea that the sec- 
 tion of the vessel in way of the hatches is 
 liable to develop weakness, but as shown in 
 the further reference to hatchways (p. 250), 
 transverse stiffness is afforded by the intro- 
 duction of portable transverse web plates 
 between the coamings. (See figs. 46 and 
 115.) 
 
 "When the hatch coamings have square 
 corners, the side and end plates are connected 
 by means of a corner angle bar. 
 
 The half beams in way of the hatch- 
 ways are connected to the hatch coamings 
 by means of angle collars, as shown in 
 fig. 110. 
 
 When the strong beams at the ends of 
 hatches are made up of a bulb plate and 
 angle, as shown in figs. 46 and 110, the beam 
 knee is made in one of the following ways : 
 either the beam is heated and the end turned 
 down, and a piece of plate welded on the 
 corner as shown in fig. Ill, or else the bulb 
 is chipped off the end of the bulb plate and 
 a piece of plate or bulb plate welded on, as 
 shown in fig. 112. 
 
 It is also common to fit bracket plate 
 knees to bulb plate beams, as shown in 
 fig. 113. 
 
 In order that the bracket plate may fit 
 close against the bulb plate, the bulb on one 
 side must be chipped off. 
 
 The beam is necessarily cambered, and, 
 like the ordinary beams, fits into the bosom 
 of the frame as shown in fig. 112. In 
 welded knees, preferably the bulb should be 
 preserved right round the knee, and should 
 finish upon the outside of the frame, as 
 shown in fig. 48, as it affords both stiff- 
 ness and strength to resist collapse under 
 severe, transverse, racking stresses. 
 When we come to the upper deck beams in way of the engines 
 and boilers, again the ordinary bulb angle beams are cut in way of the 
 
 k 
 
 b 
 
 Plan. 
 
 Fig. 112.— Bulb Plate 
 Welded Knee. 
 
 Fig. 113.— Bulb Plate 
 Bracket Knee. 
 
DETAILS OF CONSTRUCTION. 
 
 199 
 
 openings, with only an occasional specially strong beam run continuously 
 from side to side of the vessel. These half beams are supported in a way 
 similar to that explained for beams at hatch sides. Stout steel coaming 
 plates are fitted down over the beam ends, to which the half beams are 
 attached by angle collars or lugs. 
 
 In the decks of most steam vessels, it is very common to see small hatch 
 openings, say 2 or 3 frame spaces (about 4 or 6 ft.) in length, and 2 or 
 3 ft, in breadth. (See fig. 143, deck plan.) These are used for trimming 
 coal into the bunkers, or, when at the ends of the vessel, as entrances to 
 store-rooms (boatswain's, sail, etc.). When they open into coal bunkers only, 
 the beams usually run through them without interruption, and the coal is 
 trimmed through the spaces between the beams. In such cases, the beams 
 are usually protected from damage by falling coal, by riveting a piece of 
 
 A 
 
 Hatch or 
 Companion 
 
 B 
 
 r=^ 
 
 D 
 
 D 
 
 Section through A.B. 
 
 Fig. 114.— Beam cut for Hatchway or Companion-way. 
 
 convex iron to the top flange of the beam. In such cases, the coamings 
 are simply erected upon the deck, there being no necessity to extend them 
 further down. 
 
 Where the deck openings are entrances to store-rooms, the space is 
 generally left entirely open, and the beams are therefore cut at the sides of 
 the hatchways, as shown in fig. 114. 
 
 The half beams may be supported by being attached to fore and aft 
 coaming plates carried down below the beams, as previously described for 
 large hatchways, or else a strong bulb angle or channel bar may be made to 
 form the ' fore and after ' to carry the half beams, as shown in fig. 114. 
 
 A reduction is sometimes made iu the thickness of the beams towards the 
 vessel's ends. This reduction is usually made upon all beams which are less 
 
200 
 
 STEEL SHIPS. 
 
 i4 
 
 .^^ 
 
 
 c>i 
 
 ,— , 
 
 
 
 
 u:: 
 
 
 H 
 
 ^ 
 
 u 
 
 "^^ 
 
 r 
 
 ^ 
 
 -| 
 
 r 
 
 
 
 -«3 
 
 
 i? 
 
 rn 
 
 ^ 
 
 
 ^ 
 
 t3 
 rl 
 
 -tS 
 
 
 
 ce 
 
 -S5 
 CO 
 
 
 
 rS 
 
 ^ 
 
 
 
 ^ 
 
 "=>-) 
 
 
 
 ^ 
 
 c= 
 
 
 +J 
 
 i 
 
 
 
 ^\ 
 
 
 be 
 
 (3 
 
 V 
 
 
 fe 
 
 =1 
 
 [=* 
 
 
 O 
 
 -Ci 
 
 
 
 >-, 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 :i^ 
 
 
 
 
 than three - fourtlis of the 
 length of the midship beam. 
 The reasonableness of this is 
 easily explained. The shorter 
 a beam or strut is made, pre- 
 serving the same depth and 
 thickness, the more rigidity 
 does it possess, and the more 
 resistance it offers to buck- 
 ling. Hence the shorter end 
 beams may be reduced -^th 
 in. in thickness, and still 
 retain comparatively the 
 efficiency of the beams at 
 amidships. 
 
 Before leaving the upper 
 deck beams, we may note 
 that had the vessel been of 
 larger type, and channel bar 
 beams been adopted, the 
 knees would generally be 
 formed of bracket plates, 
 and would be constructed 
 as shown in fig. 115, the 
 bulb being, of course, sub- 
 stituted by the lower flange 
 of the channel bar. In all 
 other respects the knee for- 
 mation would be as usual. 
 
 Turning to the forecastle 
 (see fig. 46) we see that the 
 beams consist of a bulb plate 
 and two angles fitted to 
 alternate frames. The knee 
 to the bulb plate is made 
 as described in figs. Ill, 
 112, or 113. In the fore 
 peak, the beams are similar 
 to the forecastle beams. 
 The plan in fig. 112 shows 
 how the two beam angles 
 fit against the frame and 
 reverse frame. 
 
 The widely-spaced hold 
 beams, as shown in the mid- 
 
DETAILS OF CONSTRUCTION. 
 
 20I 
 
 ship section, fig. 10, consist of bulb plates with two top angles and a 
 covering plate. The larger flanges of the angle bars are placed horizontally, 
 which necessitates that the covering plate be of the same width. The knee 
 is formed like all other bulb plate beam knees. 
 
 The hold beams might have been as shown in fig. 116, in which case the 
 knee may be either welded or bracketed. 
 
 Against the bulkheads we observe (fig. 46) that what are called semi-box 
 beams are fitted. Fig. 117 will show more clearly how this strong beam is 
 composed. 
 
 Semi-box beams are principally intended to act as stifi'euing to the 
 bulkheads. We shall again refer to these bulkhead stiflfeners when dealing 
 with bulkheads. Suffice it for the present to say that the bulb plate knee 
 is made by one of the methods previously described. 
 
 On coming to the engine space, however, we find a number of extra 
 strong through beams, which are continuous from side to side of the vessel. 
 
 <^ 
 
 L 
 
 One frame space 
 
 A 
 
 Fig. 116.— Strong Beam with Turned Knee. 
 
 Fig. 117. — Semi-box Beam. 
 
 These are specially necessary when we remember that a large number of the 
 ordinary beams are stopped in the way of the engine opening — these strong 
 beams in some measure compensating for the loss of transverse strength. 
 They consist mostly of bulb plates with four angles. The knees are of the 
 usual construction. In elevation, the beam appears as shown in fig. 116, 
 the plan being similar to fig. 112. 
 
 Sometimes two of these strong beams are fitted upon adjacent frames, 
 and combined by plating them over — see strong beams in figs. 46, 47, and 49. 
 We observe in the profile, fig. 46, that a steel deck is fitted all fore and aft 
 to the upper deck. According to Lloyd's Eules, this vessel does not require 
 more than a steel deck for half length amidships for structural purposes. 
 Had the upper deck been sheathed with wood, the steel deck could have been 
 gradually tapered off from the half length amidships, and the subsequent 
 beams might have been of stronger section, and spaced upon alternate 
 
202 
 
 STEEL SHIPS. 
 
 frames. However, tlie usual practice for cargo vessels has been followed 
 in this case, and the steel deck made continuous all fore and aft, and 
 unsheathed with wood. In many respects, this is preferable in cargo 
 vessels. 
 
 When beams under wood decks are of tee bar, the knees are usually 
 formed of bracket plates riveted on as for plain angles. When they consist 
 of bulb plate and two angles, the knee is formed as already illustrated 
 (see figs. Ill, 112, and 113). When the beam is of butterly bulb section, 
 the knee is usually made as shown in fig. 118. 
 
 The extremities of the beam are cut through 
 
 ■^^^^^^ I the middle of the depth, and the lower part 
 
 I g; ^^^'?;"^^ . -"77i is turned down. A piece of plate is then 
 
 \ ' \. ^ '^ic ^ ^/l welded into the space completing the knee 
 
 form. Bracket knees might have been adopted, 
 
 in which case the bulb would have to be 
 
 chipped off one side. When steel decks are 
 
 laid upon beams to alternate frames, beams of 
 
 smaller section should be fitted to every frame 
 
 at the sides of all hatchways and engine and 
 
 boiler openings, 
 
 Lloyd's Rules give the following table for number and size of rivets in 
 
 beam knees : — 
 
 Fig. 118.— Butterly Bulb 
 Welded Knee. 
 
 Depth of Knee. 
 
 Unc 
 
 ler 17 
 
 inches, 
 
 17 and under 21 
 
 21 
 
 
 „ 24 
 
 24 
 
 
 „ 28 
 
 28 
 
 
 „ 32 
 
 32 
 
 
 ,, 36 
 
 36 
 
 
 „ 40 
 
 No. of Rivets, 
 
 Diameter of Eivets, 
 
 i of an inch. 
 
 The following beam knees should be at least three times the depth of 
 the beam : — 
 
 1. In all steamers or sailing ships having one tier of beams, 
 
 2, To all beams on deep framing or web frames, 
 
 3, To all beams on deep tanks, peak tanks, or other water-ballast tanks. 
 
 4. To all beams of over 50 ft, in length. 
 
 Pillars, 
 
 A variety of kinds of bar iron or steel may be used to perform the tie 
 and strut functions of pillars in ships (between decks and floors), and 
 thus in addition to the plain round bar of circular section which is most 
 commonly used, tee bars either single or double, fitted back to back, chamael 
 
DETAILS OF CONSTRUCTION. 203 
 
 bar single or double, and iron of I section, are now frequently adopted. 
 One of the most recent types of ship pillars is made of stout steel plates 
 bent into circular form and riveted like a mast. These, as we shall see, are 
 fitted in combination with other structural features, to dispense with the 
 numerous round iron pillars ordinarily required. 
 
 Ordinary Bound Iron Pillars. — Round iron pillars are made of 
 malleable iron, and may be either solid or hollow. Comparing sectional 
 area with sectional area, the hollow pillar is stronger and much more 
 efficient than the solid one, though it has the objection of occupying more 
 space, and being more costly to produce. The heads and heels of hollow 
 pillars are solid. 
 
 According to Lloyd's requirements, all beams upon alternate frames (when 
 the frames are spaced to rule) for at least three-fourths of the vessel's length 
 amidships shoukl be pillared. The beams for the remaining one-eighth length 
 at each end need only have pillars to beams upon every fourth frame. The 
 pillars should be placed, as far as practicable, in the middle line of the 
 vessel. In vessels with several decks or tiers of beams, in order that the 
 pillars develop their full efficiency, they should extend from keelson or tank 
 top to the upper deck, as nearly as possible in a vertical line, so as to form 
 a continuous tie or strut. 
 
 The foregoing are only general rules, for it is impossible to rigidly follow 
 them throughout the length of the vessel from stem to stern. For instance, 
 in the engine and boiler spaces, such an arrangement is necessarily greatly 
 interrupted, and stanchions have to be fitted where convenient. Again, in 
 way of all hatches in the midde of the deck, the stanchions are placed on 
 each side of the hatchway, generally somewhat more widely spaced (about 
 six frames), as the deck has now two supports instead of one at the middle 
 line. Under all permanent heavy deck weights, such as windlass, capstan, 
 winches, deck houses, etc., additional pillars should be introduced, two or 
 more generally being placed athwartships under such deck weights. Where 
 hold beams are dispensed with — compensation in some form or other being 
 introduced — the lower pillars, obviously, should be of greater diameter than 
 would have been necessary had the hold beams been fitted, the greater 
 length reducing the resistance to buckling, and increasing the danger of 
 collapse. \ 
 
 In vessels of great breadth, naturally the beams require support between 
 the ship's side and the centre line pillars, hence quarter pillars are intro- 
 duced. (See fig. 48.) Lloyd's Rules require that for steam vessels of over 
 43 ft. and under 55 ft. in breadth, the quarter pillars be fitted to beams 
 upon every fourth frame for half the vessel's length amidships, in addition 
 to the centre line pillars. 
 
 In pillaring a vessel, it is of the greatest importance to see that the 
 heads and heels are efficiently constructed. Indeed, owing to the bad 
 formation of the heads especially of some pillars, they never get the 
 chance, as it were, of developing their fullest efficiency. In the making of 
 
204 
 
 STEEL SHIPS. 
 
 pillar heads especially — which, by the way, together with the feet, are made 
 separately, and afterwards welded on to the pillar bar — the formation should 
 be such that any thrust or compressive stress will be transmitted directly 
 through the middle of the pillar, and not on one side (see figs. 119, 120, etc.) ; 
 and the pillar head should bear hard against the beam or girder, so that 
 the stress is borne by the material in the pillar, and not the rivets only. 
 With such a formation of heads and feet, the rivets are only subject to 
 stress when tensile strains are endured, i.e. when the pillar becomes a tie. 
 
 Fig. 119. — Connection of Pillar Head to 
 Bulb Plate Beam. 
 
 ''s-'J\ tk ^ "^"' Ik Xii 
 
 Fig. 120. — Connection of Pillar Head 
 to Channel Bar Beam. 
 
 A great point to be kept in view in the pillaring of a ship is to see that 
 the pillars do not act as independe^it struts or ties, but that they distribute 
 their efficacy as far as practicable. Thus, especially in the case of vessels 
 with steel decks, and bulb angle beams to every frame, to fix the pillar from 
 the tank top or keelson to this comparatively flexible beam itself, would 
 
 Intercostal 
 Channel Bar 
 between Beams 
 
 Bean)^ 
 
 Intercostal 
 between Beam 
 
 'Beam 
 Intercostal 
 
 
 I r Continuous Fore ^ Aft Channel Bar 
 
 Section of Beams 
 Fig. 121. — Showing an excellent method of supporting 
 Steel Decks in combination with Pillars. 
 
 produce the tendency, when the pillar is subject to a compressive stress, to 
 bulge up the deck in way of the pillar. To obviate this, a common practice 
 is to fit a tee bar, or two angles back to back, to the underside of plain or 
 bulb angle beams, and make the connection by fixing a short lug upon the 
 bottom of the beam, as shown in figs. 122, 124, 125, and 49. The pillar 
 heads are then riveted to these fore and aft bars, under the alternate beams 
 
DETAILS OF CONSTRUCTION. 
 
 205 
 
 according to rule. The formation of pillar heads is shown in figs. 119 to 
 127. By means of the continuous longitudinal beam tie bars just referred 
 to (figs. 121, 122, 124, 125, 127), the work done by the pillars is distributed 
 to all the beams, instead of only to those under which the pillars are fixed. 
 Similarly, where channel bar beams are fitted to every frame, the usual 
 practice again is to tie the beams together, so as to distribute the pillar 
 support. This is sometimes done by fitting a continuous channel bar to 
 the under side of these beams (fig. 121). In this case, the pillars may 
 
 Deck 
 
 X 
 
 ra 
 
 Beam 
 
 \ — Beam W- 
 
 Beam 
 
 ,^ 
 
 ^ 
 
 Section of Beam 
 
 1 r° o! \ 
 
 Fig. 122. — Connection of Pillar Heads to a fore and aft Tee-bar tie under Beams. 
 
 be fitted alternately to one and the other of the continuous channel bar 
 flanges on the under side of the beams, but it is preferable to attach 
 the pillar head directly to the channel beam. As the section in fig. 120 
 points out, a much better supporting surface is obtained than could be 
 got by riveting the pillar head to one of the flanges of the continuous 
 channel tie bar. An even better system than this is seen in fig. 121, 
 where, in addition to this continuous channel tie bar, intercostal channel 
 
 9 
 
 nr 
 
 k 
 
 
 vny 
 
 •iSf>> 
 
 Fig. 123.— Another 
 form of Pillar 
 Head. 
 
 Fig. 124. — Arrangement for 
 fitting Hold Shifting 
 Boards between Pillars. 
 
 bars are fitted between the beams, to which, by means of lugs, they are 
 riveted; and also in fig. 127, where, in addition to the double angles on 
 the under side of the beams, intercostal plates are fitted between the beams, 
 and riveted to the deck. Such a system of fitting a substantial girder to 
 beams and deck, especially if bracketed to the bulkheads at the end of 
 each hold compartment, affords great support and stiffness to the deck, and 
 
206 
 
 STEEL SHIPS. 
 
 permits of the hold pillars (if of increased sectional area) being spaced 
 much more widely than is usual. The advantage of such an arrangement 
 from a stowage of cargo point of view is apparent. This applies equally 
 to fig. 121. 
 
 The formation of pillar heads for channel beams is shown in figs. 120 
 
 
 Fig. 125.— Double Angle 
 Beam Tie, and Pillar- 
 head Connection. 
 
 n 
 
 Head 
 
 ^ Foot 
 
 o 
 
 Plan of 
 Pillar Foot 
 
 Fig. 126.— Pillar in Cabins. 
 
 and 121. Where, however, the beams are fitted on alternate frames, 
 and are therefore much more rigid owing to their greater depth and 
 sectional area, than would be the case for beams on every frame, it is 
 usual to fit the pillar heads directly on to the beams, no continuous 
 tie being fitted to the under side of such deep beams. (See figs. 119 
 
 Intercostal plate 
 
 Fig. 127. — Another excellent method 
 of supporting Steel Decks in con- 
 junction with Pillars. 
 
 crni liD 
 
 Plan 
 
 Fig. 128.— Ordinary 
 Pillar Foot. 
 
 and 120.) When shifting boards must of necessity be fitted down the 
 middle of a hold to prevent cargo moving, the common practice is to 
 place them between round iron pillars arranged as shown in fig. 124. In 
 this case we have a continuous channel tie on the under side of the beams 
 
DETAILS OF CONSTRUCTION. 
 
 207 
 
 to which the alternate pillars are fitted to first one and then the other flange. 
 Instead of the channel bar, two angle bars are sometimes used for the 
 same purpose. 
 
 All pillar heads, such as those we have just described, should be con- 
 nected 1)y at least two rivets. A pillar head not very commonly adopted, 
 but which requires only one rivet, is seen in fig. 123. The head is formed 
 as shown in A. This head, on being heated and placed under the beam, is 
 closed up as in B, and secured by one rivet. Where pillars are required to 
 be portable, the heads are fastened by means of nut and screw bolts instead 
 of rivets. 
 
 The formation of pillar feet, connected to a centre keelson standing on 
 ordinary floors, is shown in elevation and plan in fig. 128. The connection 
 here is by means of two rivets. Where, however, the pillar feet come over the 
 centre through-plate in a vessel with a double bottom, the feet are formed and 
 connected as in figs. 129 and 130. In fig. 129 a socket is formed for the 
 
 Tank top centre strake 
 
 Centre through plate In 
 Cellular Double Bottom 
 
 li 
 
 'i^ 
 
 Plan 
 Fig. 129.— Pillar Foot ou Inner 
 Plating of Double Bottoms. 
 
 -^ K'\ 
 
 Fig. 130. — Showing connection 
 of Pillar Feet to Steel Decks, 
 or Inner Plating of Double 
 Bottoms. 
 
 pillar foot, either between two flange plates, or two angle bars riveted on 
 to a base plate through which the connection is made to the tank top. In 
 the alternative method, fig. 130, the frame foot is connected to a piece of 
 tee bar which is riveted through the tank top plating and upper bars of 
 the centre keelson. The pillar feet in a 'tween decks where a steel deck 
 is laid, should be riveted through the deck plating and the beam. But a 
 better arrangement is to fit the feet on to a short piece of tee bar, long 
 enough to take three rivets, somewhat similar to the sketch in fig. 130. 
 Where, however, no steel deck is laid, and the beams are upon alternate 
 frames, the pillar feet may be formed as in fig. 128, and attached directly to 
 the beam. Or better still, Vjy fitting a continuous tie bar, or two angle bars 
 riveted back to back, a more satisfactory distribution of the work of the 
 pillar is obtained. These angle bars, if made large enough in their 
 horizontal flanges, will form a substitute for the tie plates which in any 
 
208 
 
 STEEL SHIPS. 
 
 case would be necessary upon beams on alternate frames where no steel deck 
 is laid. Where pillars are required to be portable, which is often necessary 
 at hatch sides, the feet may be made in any one of the three ways shown in 
 figs. 131, 132 (A and B). 
 
 Fig. 131 shows the pillar foot dropped into a cast-iron ring, which is 
 riveted to the deck or tank top plating as the case may be. A half inch or 
 five-eighths nut and screw bolt is passed through the cast-iron ring and pillar 
 foot as shown. Fig. 132, B, shows the pillar foot placed in a horseshoe 
 angle bar, and is kept in position by a preventive nut and screw bolt. In 
 this method, the pillar only performs the function of a strut. As shown in 
 the diagram 132, A, the foot of the pillar is permanently riveted, the upper 
 part only being portable, with a nut and screw bolt means of connection 
 as shown.* 
 
 fft* 
 
 ^ 
 
 Plan 
 
 Fig. 131.— Portable 
 Pillar Foot. 
 
 Plan 
 /I Plan 
 
 B 
 
 Fig. 132.— Portable Pillars. 
 Arrangement of Feet. 
 
 The heads and heels of all ordinary pillars fitted in cargo vessels should 
 be connected to the structure by at least two well formed rivets. But those 
 fitted in deep water-ballast tanks should be connected by means of three 
 rivets. 
 
 Channel and Tee Bar Pillars. — Where pillars are constructed of tee bars 
 or channel bars (either single or double), or of I section, they are connected 
 to the beams at their upper extremities, and to the tank top at their 
 lower extremities, by means of bracket plates. (See figs. 64, 67, 70, 73, 
 and 74.) 
 
 Large Round Plate Pillars — substituting centre line hold and quarter 
 pillars — have already been dealt with in Chapter VI. Section 2, ' Some JSTew 
 Features in Modern Shipbuilding.' (See figs. 72, 73, and 74.) 
 
 * These portable pillars are not as efEcient as well formed fixed stanchions, but as 
 only very few are usually required (generally at the side of hatchways), no serious 
 objection need be offered to them. 
 
 When cabins are fitted in a 'tween decks, the ordinary round pillar is sometimes 
 objectionable, and in many cases half round or flat iron stanchions, as in fig. 126, are 
 substituted. 
 
DETAILS OF CONSTRUCTION. 
 
 209 
 
 Keelsons and Stringers. 
 
 Keelsons. — The most important keelson in the ship is the centre 
 keelson, and although, according to the particular mode of construction 
 adopted, it varies considerably in form, its function is always the same, viz., 
 to afford longitudinal strength along the bottom of the ship. The centre 
 through-plate keelson and side keelsons, in vessels with double bottoms, 
 have been referred to on several occasions and fully described already. 
 (See figs. 48, 58, etc.) 
 
 Fig. 133, A and B, illustrate what are known as centre keelsons 
 standing upon ordinary floors, the vessel in the diagrams having a bar 
 keel. The principal parts of the keelson, A, are a deep vertical plate 
 with two larcre ansrles on the bottom connected to the reverse frames and 
 
 Fig. 133. — Centre Keelsons standing upon Ordinary Floors. 
 
 the lug piece shown; and, on the upper edge, two similar large angles 
 with a flat plate, known as a rider plate, covering the full width of the top 
 flanges. The large keelson angles have one flange exceptionally large, 
 while the other is considerably smaller. The large flanges are arranged 
 vertically for the bottom bars, and horizontally for the top bars. The 
 butts of both centre plate and rider plate should be either strapped or 
 overlapped, and treble riveted. The butts of the angle bars should be 
 connected by bosom pieces Avith at least three rivets on each side of the 
 butt. The butts of vertical plate, rider plate, and angles, should be kept 
 as clear from each other as practicable. The centre keelson in B, fig. 133, 
 is identical with A, with the exception that being for a larger vessel, a 
 thick foundation plate is riveted to the floors immediately under the 
 keelson to which the bottom angles are connected. 
 
 14 
 
2IO 
 
 STEEL SHIPS. 
 
 Fig. 134, D, E, F, illustrates another type of centre keelson, known 
 as a middle line intercostal keelson. As the diagrams show, the vessel 
 has ordinary floors and a bar keel. The keelson in D, which is for a very 
 small ship, has two continuous angles on the top of the floors, with 
 intercostal plates fitted between the floors extending from the top of the 
 keel to the top of the keelson angles. The intercostal plates are connected 
 to the floors by single angles as shown. Figs. E and F are similar keelsons 
 for vessels of larger size. The intercostal plate in each case extends to the 
 top of the lower keelson angles. 
 
 Fig. 135, G, H, and J, illustrates the centre through-plate keelson in 
 vessels having side bar keels. Here, again, the floors are of the ordinary 
 type. In G, the centre through-plate extends from the bottom of the keel 
 
 JL 
 
 Jl 
 
 Showing Intercostal Plates fitted upon Flat Plate Keels. 
 Pig. 134. — Intercostal Centre Keelsons. 
 
 to the top of the floors. Along its upper edge, and riveted to the floors, 
 is a broad thick plate, connected to the centre through-plate by the two 
 large angles shown on its upper edge. It will be seen how admirably 
 this plate binds both floors and keelson together. Fig. H, which is for a 
 larger vessel, shows the centre through-plate extending to the top of the 
 keelson angles. The foundation plate upon the floors under the keelson 
 angles is necessarily in two pieces, one on each side of the centre through- 
 plate. Fig. J is identical with fig. H, excepting that being for a still 
 larger vessel, the part of the keelson above the floors is somewhat modified. 
 Side Keelsons (fig. 136). — All vessels should have keelsons upon the 
 
DETAILS OF CONSTRUCTION, 
 
 211 
 
 floors at the lower turn of the bilge. These may varj^ in section as shown 
 in fig. 136, P, Q, M. P is two plain angles fitted back to back. Q is 
 the same as P with a bulb plate between. M is the same as Q, excepting 
 that in addition to the bulb plate, an intercostal plate is fitted from the 
 
 JL 
 
 Showing Centre Through-plates fitted upon Flat Plate Keels. 
 Fig. 135. — Centre Through-plate Keelsons for Vessels having Side Bar Keels. 
 
 top of the angles to the bottom of the floors, and connected to the shell 
 by another angle bar as illustrated in fig. L. The bulbs and intercostals 
 are often extra strength introduced on account of a vessel's extreme 
 proportions. Between the centre keelson and the bilge in very small 
 
 nr 
 
 JL 
 
 jL 
 
 Fig. 136.— Side Keelsons. 
 
 vessels, a thin plate is fitted intercostally between the floors to which it is 
 connected. The principal function of this plate is to check the wash of 
 bilge water when the vessel is rolling. Hence it is called a wash plate 
 
212 
 
 STEEL SHIPS. 
 
 ■I m '°\ ° ° ° /. 
 
 J Jo o p' o_^_o_"L_lL 
 
 I 
 
 z:^ 
 
 ir 
 1 
 
 -:---l 
 
 (see K). In larger vessels a double angle side keelson is required, 
 extending across the top edge of the floors, with sometimes both a bulb 
 
 and an intercostal plate be- 
 tween them (fig., L and M), 
 In still larger vessels, or 
 where the proportions are 
 very extreme, a side keelson 
 with an intercostal plate, as 
 shown in JST, is sometimes 
 required. 
 
 In all cases wherever 
 
 centre, side, or bilge keelson 
 
 angles, or horizontal plates, 
 
 pass over the floors, the 
 
 connection is made by short 
 
 lug bars in addition to the 
 
 ordinary reverse frames. 
 
 « Stringers. — Stringers, in 
 
 Ic many respects, are very similar 
 
 'B to keelsons. 
 
 I Fig. 137, E, S, T, U, V, 
 
 j» illustrates a number of types 
 ^ which are regulated by the 
 I depth and proportions of the 
 ^ vessel. As already pointed 
 *j out for keelsons, angle lugs 
 '^. are again required in addition 
 ^ to the reversed frames, in 
 order to get a good connection 
 to the transverse framing. 
 Stringers for vessels with 
 deep frames are illustrated 
 in figs. 12 and 48. Keelsons 
 and stringers may be carried 
 uninterruptedly through all 
 the bulkheads, in which case, 
 Avell-formed angle collars are 
 necessary in order to ensure 
 watertightness (see fig. 138, 
 Y.) Where the stringers are 
 cut at the bulkheads, which 
 is the simpler way of obtain- 
 ing watertightness, the con- 
 tinuity of their strength is maintained by fitting large bracket plates. 
 
 Fig. 137, A, illustrates a side stringer such as is shown in the midship 
 
 \jf" ~fT_Ty\ h 
 
 1 
 
 Co 
 
 z 
 
 "-G 
 
 T 
 
 ^T 
 
 jLl-JZ^j-r^ 
 
DETAILS OF CONSTRUCTION. 
 
 213 
 
 section of fig. 48, cut at the bulkheads and connected by large brackets, 
 which extend to the heel of the bulb angles forming the stringers. 
 
 Stringers and keelsons should be carried as far forward and aft as 
 practicable, aiid their ends should be connected to a breast hook — a 
 horizontal plate binding the two sides of the vessel together. 
 
 Bulkheads. 
 
 Steel and iron bulkheads are walls of plating extending transversely 
 from side to side, or longitudinally, throughout the whole or part of the 
 length of a vessel. They are valuable as a means of dividing the large 
 volume contained in the hull of a ship into a number of separate compart- 
 
 r° 
 
 000 1 \~ 
 
 y^OOOo| 000 
 
 
 
 Jr-l- 
 
 1 N 
 _1_ 
 
 ^ Jl 
 
 !•" 
 
 ' 
 
 
 1 000 
 
 000 ffl ' 
 
 ! 
 
 Q. 1 
 
 " ./ 
 
 
 oooo 000 
 
 y 
 
 / 
 / 
 / 
 / 
 / 
 / 
 
 / 
 
 Bulkhead. 
 
 
 Fig. 137, A. — Stringer cut at a Bulkhead, and connected by large Brackets. 
 
 ments, and in entirely shutting off the engine and boiler spaces from the 
 hold spaces, a feature of very great importance where inflammable cargo is 
 carried in the adjoining compartments. 
 
 But their principal function lies in affording structural strength and 
 safety in the event of the shell being perforated, and water finding access. 
 Long shallow vessels are particularly liable to evince longitudinal weak- 
 ness. In such vessels, great additional strength and stiffness is obtained by 
 fitting a continuous longitudinal bulkhead all fore and aft down the middle 
 of the hull and connecting it to, or making it form part of, a centre keelson 
 on its lower extremity, and attaching it to the deck at its upper edge. 
 
 Transverse bulkheads aflFord great support to both the longitudinal and 
 the transverse framing of a vessel, and contribute in a large measure to the 
 
214 STEEL SHIPS. 
 
 general strength. When transverse bulkheads are well constructed, and 
 stiffened, and made perfectly watertight, they may render immense service 
 in case of collision or perforation of the shell plating by confining the water 
 to one of the compartments into which they divide the hull. They must, 
 therefore, be securely connected to the shell plating, decks, and bottom of 
 the vessel, and, in addition, the plating must be of sui3S.cient thickness, and 
 stiffened and supported in such a manner as to withstand the pressure 
 of water which would come upon it in the event of such a compartment 
 being perforated and flooded. A longitudinal bulkhead, while it may pro- 
 vide valuable and necessary strength to the hull, may also be a source of 
 danger if made watertight, excepting in the case of vessels carrying liquid 
 cargoes, for in the event of a compartment on one side of such a bulkhead 
 becoming flooded, the loss of buoyancy may produce a positive danger, 
 which would not have existed had the water been able to fill the compart- 
 ment from one side of the vessel to the other. It is therefore advisable, 
 under some circumstances, to so construct a longitudinal bulkhead that 
 water could find ready access through it. 
 
 All vessels classed at Lloyd's are required to have a certain number of 
 transverse bulkheads, which number depends in steamers upon the length. 
 The rules require that all screw steamers have at least four watertight 
 bulkheads, one at each end of the engine and boiler space, and one at a 
 reasonable distance from each end of the vessel. 
 
 The necessity and importance of these bulkheads is fairly obvious. The 
 importance of having the engine and boiler spaces thoroughly isolated 
 from the remaining portions of the hull has already been referred to. The 
 fore-end of vessels being likely to suffer most severely in the event of 
 collision, and, in addition, being specially subject to the thumping action 
 of the waves in steaming against head seas, as well as the tendency to 
 pant when driven at considerable speed, even in smooth water, all clearly 
 show the necessity of a watertight bulkhead at a moderate distance from 
 the stem, which, the rule states, should not be less than half the vessel's 
 moulded breadth in order to safeguard the vessel should leakage occur from 
 any of these causes. 
 
 At the after end of the vessel, especially in very large ones, we have an 
 enormously heavy stern frame attached by means of long rivets to what are, 
 in comparison, thin plates composing the shell, and which, in addition, is 
 subject to considerable vibration caused by the propeller, especially when 
 racing in heavy weather. Again, the breaking of a tail shaft often fractures 
 the stern tube and hence the possibility of leakage, and the necessity of a 
 watertight bulkhead at a reasonable distance from the stern post. Where 
 the shaft passes through this bulkhead, the watertightness is assured by 
 fitting a stuffing box. 
 
 When the length is 280 ft. and above in steamers, an additional bulk- 
 head, is required by Lloyd's midway between the collision and the foremost 
 engine space bulkheads, and when the length reaches 330 ft. and over, 
 
DETAILS OF CONSTRUCTION. 215 
 
 another bulkhead is required in the after hold, situated therefore between 
 the aftermost bulkhead in the vessel and the after engine room bulkhead.* 
 Tliese additional bulkheads not only assist in stiffening the framing of the 
 vessel, but they improve the chances of the vessel remaining afloat in the 
 event of one of these compartments becoming flooded. All transverse 
 bulkheads in one, two, and three decked vessels should extend to the height 
 of the upper deck, and in spar deck vessels to the height of the spar deck. 
 In the case of awning deck vessels, all the bulkheads should extend to the 
 height of the main deck with the exception of the collision bulkhead, 
 which should extend to the awning deck.f 
 
 As we have pointed out, the fore end of the vessel being the most 
 vulnerable, and the loss of buoyancy at the extreme ends of a vessel through 
 perforation of the shell plating producing greater change of trim than the 
 loss of as much and even more buoyancy nearer amidships, the necessity of 
 continuing the collision bulkhead, even in an awning deck vessel, to the 
 weather deck, will be clear. To carry a watertight bulkhead in a direct 
 uninterrupted vertical plane from the floors to the prescribed height, is not 
 always convenient nor is it absolutely necessary. It may be recessed or 
 stepped backwards or forwards at some place in its height, and then carried 
 to the required deck. No objection can reasonably be lodged against this 
 so long as the workmanship is thoroughly efficient, and the strength and 
 watertightness maintained. 
 
 In sailing vessels, only the collision bulkhead is required by Lloyd's, 
 though many owners prefer to introduce additional bulkheads for the sake 
 of the improved conditions of safety, and the convenience of having the hold 
 space divided into separate compartments for the stowage of cargo. 
 
 In arranging the plates of a bulkhead, they are disposed either with their 
 lengths vertical or horizontal. Fig. 138 is a sketch of a bulkhead with the 
 plates arranged horizontally. As the figure shows, the vessel has two steel 
 decks, and as the continuity of a steel deck is vastly more important than 
 the continuity of the bulkhead, the deck suffers no interruption, but the 
 bulkhead is severed at the lower steel deck, and continued again in the 
 'tween decks. 
 
 A watertight bulkhead is connected to the shell plating by means of 
 double angle bars of the size of the frames between which angles the bulk- 
 head plating is fitted. (See horizontal section of bulkheads, fig. 138, A and 
 B.) As the watertightness must be maintained from lower to upper ex- 
 tremity, a connection to the steel decks is usually made by means of double 
 angle bars of about the size required for reversed frames. (See E, C, and K, 
 
 * Vessels over 400 ft. require 7 bulkheads. 
 
 ,, 470 „ ,, 8 
 
 „ 540 ,, ,, 9 
 
 t In awning deck and shelter deck vessels, partial bulkheads or web frames for the 
 
 support of the topsides should be fitted in the 'tween decks immediately above the main 
 
 bulkheads. 
 
2l6 STEEL SHIPS. 
 
 fig. 138.) Should the vessel have a double bottom, the connection to the 
 tank top is also usually made by means of double angles of the size already 
 mentioned, D, C, K, fig. 138. While the overlapping of the edges and butts 
 of the plates gives a certain amount of stiffness to the area of plating, yet 
 this is totally inadequate to provide the strength required in carrying even 
 bulk cargoes, such as grain, much less to withstand the pressure which 
 would accrue, owing to a head of water, in the event of a hold space 
 becoming flooded to any considerable extent. An intelligent arrangement, 
 therefore, of well-disposed stifFeners is necessary. Lloyd's requirements for 
 the stiffening of transverse bulkheads are as follows : — On one side of 
 the bulkhead the stiffeners are angle bars of the size of the frames spaced 
 not more than 30 in. apart. If the vessel is built with ordinary floors, 
 these stiffeners should be continued well down over the floor plate. (See E, 
 fig. 138.) If the vessel have a double bottom, these stiffeners should be 
 connected to the inner bottom plating by means of plate brackets (D, K, 
 fig. 138.) Additional vertical stiffening is given to all bulkheads of large 
 breadth in the form of webs of plating extending from the centre keelson or 
 top of floors, or from the tank top in vessels with double bottoms to the 
 height of the lowermost tier of beams (A and C, fig. 138). When the bulk- 
 head is 36 ft. and under 45 ft. in breadth, one such vertical web stiffener is 
 required, and, in the case of vessels with ordinary floors, would be riveted 
 at its lower extremity to the top of the centre keelson. When the breadth 
 of the bulkhead is 45 and under 55 ft., two such vertical webs are required. 
 Bulkheads of 55 and under 60 ft. in breadth require three vertical webs. 
 In order to support the vertical stiffeners, and to give additional rigidity 
 and strength to the bulkhead, horizontal stiffeners are fitted to the other 
 side of the bulkhead, and spaced below the lowermost laid deck, about 48 
 in. apart. These stiffeners, in all vessels of less than 40 ft. in breadth, 
 may be of angle bars of the size of the main frames, but when the bulkhead 
 is of 40 ft. breadth and above, the horizontal stiffeners should be of bulb 
 angles of at least the size required for a steel or iron deck fitted to the same 
 vessel. 
 
 These bulb angle stiffeners should be well connected to the vessel's side 
 by means of plate brackets. - 
 
 In all vessels requiring two or more decks, where the lowermost deck 
 has been dispensed with by compensation in some form or other, or when 
 such a deck is not laid so as to give longitudinal support to the bulkhead, 
 Lloyd's require that a semi-box beam, E, C, and K (see fig. 138), be fitted 
 across the bulkhead where the lower deck would have afforded its support ; 
 or, if there be a side stringer in this vicinity, the semi-box beam may be 
 fitted so that its ends may be attached to the stringer. The breadth of the 
 semi-box beam is one frame space, and the bulb plate or channel bar or bulb 
 angle forming the side of the beam away from the bulkhead should be 
 connected to the frame upon which it comes by an efficient knee. (See also 
 fig. 117.) The fore peak, as we have already noticed, being especially liable 
 
watertight Collars or Joggled Bars 
 round Frames on Watertight Flat. 
 
 T~ 
 
 Fig. 138. 
 STEEL BULKHEAD AND DETAIL SKETCHES. 
 
 Bulkhead stiffeners 
 with ordinary floors 
 
 Deck 
 
 Wstertiqht angk 
 cellar round stringer 
 on bulkhead 
 
 f-l 
 
 Jt^iL^-i: 
 
 Rust- 
 Cast Metal 
 
 Chock , 
 
 Connection of stilTeners 
 y-ock in one Frame space i" ''"'* t^P 
 
 Section 
 Through 
 Chock 
 
 emi-box 
 beam 
 
 Section showing 
 
 verticals horizontal 
 
 stiffeners on bulkhead 
 
 Tank '"e^l 
 
 J 3 3 1 ] ] 1 1 ] ] 
 
 Illustrating another 
 H arrangement of 
 
 stiffening bulkhead 
 Deck 
 
 ^Y 
 
 -i J , L J L J U 
 
 JL 1 1 1 1- 
 
 -L 1 3. 3- 
 
 Plans showing Vertical Stiffeners 
 
 flans showing Vertical stiffeners 
 
DETAILS OF CONSTRUCTION. 217 
 
 to damage, and the compartment to being flooded, special attention is ahvays 
 given to rendering this bulkhead thorouglily efficient under all eircurastances 
 of emergency. Hence, Lloyd's require that the horizontal stiffeners should 
 be of bulb angle of the size of iron and steel deck beams, and connected to 
 the vessel's sides by plate brackets. Where, however, the fore and after 
 peaks are intended to carry water-ballast, and indeed in all deep tanks 
 situated either at the ends of the vessel, or in the region of amidships, the 
 vertical stiffeners on the bounding bulkheads should always be of bulb 
 angles or channels of at least the size just mentioned (K, fig. 138). The 
 stress upon these bulkheads is very severe when the tanks are full, but 
 where, through neglect or carelessness, vessels are allowed to proceed to sea 
 with these tanks only partially filled, the damage to the bulkhead may be 
 enormous. (See sketches.) 
 
 Owing to the pressure of water upon a bulkhead in a flooded compart- 
 ment being greater on the lower half than upon the upper, it is the practice 
 to make the plating in the lower half yV^^ 0^' uV^^^ ^^- thicker than in the 
 upper. The riveting of both lap edges and lap butts of bulkhead plating is 
 usually made by a single row of rivets, excepting in the connection of the main 
 bulkhead plating to the deep floor plate which forms the lower part of the 
 bulkhead, in vessels with ordinary floors, and the coaming plate in vessels 
 with double bottoms. (See fig. 138.) This latter connection should be made 
 by a double row of rivets. The spacing of bulkhead rivets should be about 
 4 diameters apart from centre to centre. The rivet holes in the transverse 
 flanges of the double angle bars connecting the bulkhead to the shell plating 
 should also be about 4 diameters apart from centre to centre. 
 
 All bulkheads must be caulked to ensure watertightness. Tiiis, however, 
 it is only necessary to do uj^on one side of the bulkhead. 
 
 In making the coimection of the double angle bars round the bulkhead 
 to the shell plating, the rivet holes in the fore and aft flange of that bar 
 fitted upon that side of bulkhead which is caulked should be spaced about 
 4 to 4j diameters apart. The rivet holes in the fore and aft flange of the 
 other bar which is upon the other side of the bulkhead .should be spaced to 
 the ordinary spacing in frames, viz., 7 to 8 diameters apart. 
 
 As will be shown Avhen we come to deal more particularly with shell 
 plating, the most perforated transverse section of the shell is in way of the 
 line of rivet holes, spaced, as we have stated, 4 to 4| diameters apart on 
 the caulked side of the bulkhead. Some compensation must therefore be 
 made for this excessive weakening of the shell plating. This is done by 
 fitting what are called bulkhead liners, or doubling plates. 
 
 The doubling plates are fitted in way of the outside strakes of shell 
 plates only, and, it is needless to say, are on the inside of the plating. 
 They may be either rectangular or of diamond shape. (See figs. 139 and 
 140.) 
 
 It will be noticed in fig. 139 that the liner extends from inner edge to 
 inner edge of the adjacent inside strakes of plating, and in a fore and aft 
 
2l8 
 
 STEEL SHIPS. 
 
 direction over two frame spaces from the heel of the first frame abaft the 
 bulkhead to the toe of the first frame forward of the bulkhead. Fig. 140, 
 
 blNER 
 
 ^n 
 
 c 
 
 Elevation of 
 Bulkhead. 
 
 Internal view of ship, showing connection of 
 Bulkhead to Shell and Liner. 
 
 ~1L 
 
 / BULKHEAD 
 
 M_ / 1L 
 
 /SHELL PLATING 
 
 B Plam 
 
 Fig, 139.— Connection of Bulkhead to Shell. 
 
 showing the diamond-shaped liner, is so clear as to need no comment. 
 The spacing of the rivets in these bulkhead liners is usually similar to that. 
 
 Fig. 140. — Bulkhead Diamond Liner. 
 
DETAILS OF CONSTRUCTION. 219 
 
 illustrated in the sketch. By means of these doubling plates, fitted from 
 gunwale to keel on the inside of all outside strakes of shell plating in vessels 
 with ordinary floors, and from gunwale to tank side in vessels with a double 
 bottom, the loss of strength caused by the severe perforation of the shell 
 plating in way of the bulkhead is regained. In all cases, the collision and 
 aftermost bulkheads should have their efficiency and watertightness tested 
 by filling the fore and after peaks w^ith w^ater to the height of the load 
 line. In order to detect more readily where the caulking is faulty, in the 
 event of there being leakage, the collision bulkhead must be caulked 
 upon the after side, and the aftermost bulkhead upon the fore side. This 
 same rule applies to all bulkheads bounding deep water-ballast tanks, that 
 is, the caulking should be upon that side of the bulkhead Avhich is the 
 outside of the tank. For other bulkheads, it is immaterial which side 
 is caulked, but it is always the side upon which the rivets have been 
 hammered up. 
 
 The usual forms of rivets adopted in bulkheads which are not exposed 
 to view, and where appearance is of no moment, are the pan- and plug-head, 
 with beat-up points, these being the most effective rivets for completely 
 filling up the rivet holes, and obtaining watertight results. (See fig. 90.) 
 
 "Where, however, .the bulkheads are exposed to view upon one side, as 
 in the engine room, for instance, it is more common to adopt the pan-head 
 rivet, and to form a snap head on the other side, the snap head being on 
 the exposed side of the bulkhead, or to use snap-head and snap-point rivets. 
 This certainly gives a better appearance, but being formed by hand, the 
 results are not always satisfactory from a watertight point of view. (See 
 further remarks upon 'Rivets,' p. 167). 
 
 When a 'tween-deck bulkhead is stepped one or more frame spaces 
 either backwards or furwards, which, by the way, is often done in collision 
 bulkheads, great care should be taken to ensure that the flat at the break 
 in the bulkhead is both well constructed and thoroughly watertight. This 
 can be done by cutting the frames in way of this watertight flat, and 
 running an angle bar continuously along the outer edge of the flat, 
 connecting the flat to the shell plating, and preserving the strength of 
 the transverse frames by connecting the upper frame legs to the watertight 
 flat by means of plate brackets. (See 'tween decks, fig. 78.) 
 
 Another method is that of cutting the reverse bars in way of the flat, 
 and doubling the frames for a length of about three feet to compensate for 
 the loss of transverse strength, and joggling an angle bar round the frames 
 throughout the length of the flat. (See fig. 138, L.) 
 
 Another method still, though one that is by no means extensively 
 adopted, is to fit cast metal chocks in each frame space between the 
 shell and the continuous angle bar on the inside of the frames. A space of 
 about three-quarters of an inch round these chocks is filled with fine waste 
 metal from drilling machines rusted with sal-ammoniac and thoroughly caulked. 
 
 In arranging the stiffeners of bulkheads bounding deep water-ballast 
 
220 
 
 STEEL SHIPS. 
 
 tanks, the horizontal stiffeners are often dispensed with entirely, with the 
 exception, sometimes, of a semi-box beam stifiener, and instead, stronger 
 vertical stiffeners are kneed to the deck at the top, and to the tank top 
 at the bottom when the double bottom is fitted. (See K, fig, 138.) 
 While the foregoing remarks upon bulkheads are in accordance with the 
 
 r' 
 
 
 mD 
 
 C ■'( - '', ■ §> 
 
 SZ3 
 
 ^^ 
 
 c=^ 
 
 c=J>. 
 
 I 
 
 V" 
 
 r" 
 
 INOIA-RUBBCR 
 STEEL WEOCE 
 
 STEEL WEDCE 
 
 INDIA-RUBeeR 
 
 Watertight Door fg 
 
 mm 
 
 ^ 
 
 India Rubber 
 Bulkhead 
 
 Fig. 141.— Bulkhead Watertight Door. 
 
 • Enlarged Sketches showing different methods of making Doors Watertight. 
 
 usual practice of plating and stiffening, in recent years other methods 
 have been to some extent adopted. !Fig. 138, B, is a horizontal sectional 
 plan, showing a system of flanging the alternate vertical strakes of plating, 
 as shown, and by this means the plating is made to form its own stiffening, 
 
DETAILS OF CONSTRUCTION. 
 
 2 2 1 
 
 assisted by deep webs for vessels of grerit breadth iti accordance with the 
 conditions previously stated. Where this system of flanging is adopted, 
 the plating should be increased in thickness -jV^^ ^o^' ^^®®^ ^^ tV^^^ ^'^^ 
 iron. In vessels with double bottoms, it is scarcely necessary to add, 
 plate brackets will be required at the bottom of the flange stiffeners as 
 for the ordinary system previously described. Fig. 138, G, shows the 
 tank top brackets of the system of stiffening just described, or the 
 brackets may be dispensed with by fitting a large angle continuous from 
 margin plate to margin plate riveted to the face of the stiffeners and 
 tank top as shown by the dotted lines. Fig. 138, J, shows still another 
 system of flanging the bulkhead plating so as to form stiffeners chiefly 
 for 'tween-deck bulkheads. Here, again, the plating should be J^^\i or 
 yY^h thicker. 
 
 Bridge Deck 
 
 
 
 
 
 
 Upper Deck 
 
 ■ T 
 
 ?- 
 
 
 L 
 
 T 
 
 , Middle Deck 
 
 ^ 
 
 1 — — ^ 1 
 
 7 
 
 
 
 \ w. 
 
 Water Line 
 
 \ 
 W.T Door . 
 
 ^ 
 
 Lower Deck 
 
 L 
 
 
 W.T.Door 
 
 
 Double Bottom 
 
 Fia. 142, — Watertight Doors manipulated by means of rods from Upper Deck. 
 
 Watertight Doors. 
 
 A golden rule to be kept in mind, in considering the bulkhead 
 arrangement of a ship, is never to make a doorway, however good the 
 system of securing watertightness may be, if it can possibly be avoided ; 
 for, as accidents from collision usually happen at unexpected moments, 
 such doors are often open at these times, and many a good ship, from no 
 other cause, has foundered. The aforementioned rule is rigidly observed 
 by many of the companies owning large mail steamers. However, 
 doorways through watertight bulkheads cannot always be avoided, 
 especially in the 'tween decks when used for passenger accommodation. 
 
 Fig. 141 illustrates a watertight door, and several different modes of 
 
2 22 STEEL SHIPS. 
 
 construction and fastening. Wherever watertight doors are required 
 through a bulkhead below the load waterline, they should be either of 
 the horizontal or vertical sliding type, and opened and closed by means 
 of rods worked from the upper deck. (See fig. 142.) 
 
 Decks. 
 
 Steel DecJiS. — Under iron and steel decks it is always preferable to fit 
 beams to every frame, and where the plating of such decks is thin, say, 
 less than -g^ths in., it is imperative that this be done, because thin plating, 
 when not supported by closely spaced beams, is liable to sag, or fall hollow 
 in the spaces between the beams. Such beams may either be of plain 
 angles, or bulb angles, increasing in size (particularly depth) for vessels of 
 greater beam. (See figs. 12 and 46.) When, however, the deck plating is 
 over -/o-ths in. in thickness, the beams may be fitted to alternate frames, 
 when the frame spacing does not exceed 24 in. or 26 in. Such beams 
 must naturally be of extra strength, bulb plates with double angles, or 
 tee bulbs usually being adopted. (See figs. 48 and 49.) Apart from the 
 beams, the most important structural item in the deck is the stringer 
 plate. This is usually of exceptional thickness, the reason for which will 
 be readily understood when its function in assisting the shell plating, 
 particularly the sheer strake, and in affording general longitudinal strength, 
 is comprehended. Thus, in order to combine the deck stringers with the 
 shell plating, it has become a common practice, especially in large vessels, 
 to make the connection by means of double angles. This is necessary, 
 principally in the upper or two upper decks, in large vessels with several 
 decks. Where the frames terminate at the upper deck stringer plate (there 
 being no erections), these angles connecting the stringer plate to the shell 
 may be continuous. Where the frames extend through the upper deck 
 stringer plate on account of an erection, and in the way of a lower deck 
 stringer plate, these connecting angles must necessarily be intercostal. 
 <See fig. 58.) 
 
 The butts of deck stringer plates should be at least double riveted, 
 and, in large vessels, treble. The deck plating is arranged in strakes fore 
 and aft across the beams. The edges are connected by a single riveted 
 lap, and the butts for half length amidships should be double riveted, and 
 single at the ends.* (Fig- 143.) 
 
 While steel decks are often fitted in preference to wood decks, they 
 are often an indispensable structural part, required on account of the 
 extreme proportions of the vessel. The longitudinal stresses, as we have 
 seen at an earlier stage, being most severe over the midship length, it 
 often happens that a steel deck may only be required for the half length 
 
 * The overlaps of the butts of deck plating should look towards amidships. This 
 prevents water lodging at the butt, and eventually injuring the caulking. 
 
DETAILS OF CONSTRUCTION. 223 
 
 amidships, in wliich case it should be gradually tapered off from the 
 middle line towards the stringer plates. For the same reason, when the 
 deck plating is -—yths of an inch or more in thickness, at amidships, it may 
 be slightly reduced in thickness towards the ends. Owing to the superiority 
 of iron in resisting the attacks of the weather, and therefore suffering less 
 from corrosion, such parts of deck plating as are exposed to the weather, 
 are commonly made of iron, and elsewhere (where covered) of steel. In 
 order to obtain equivalent strength, the iron plating should be somewhat 
 thicker than the steel. 
 
 Of necessity, numerous openings are cut through the decks of all 
 vessels, — engine and boiler openings, hatches, companions, etc. This 
 naturally reduces the strength of the deck, and compensation in some 
 degree must be provided by thicker deck plating, or doublings, in the way 
 of such openings. As the deck is no stronger than its weakest section, 
 a certain percentage of the maximum transverse strength should be 
 maintained. Thus, over the midship half length, the aim should be to 
 secure, especially in the strength deck (upper deck), 80 per cent, of the 
 maximum sectional area of the plating. 
 
 Fig. 143 shows the upper deck plan of the vessel whose midship 
 section and longitudinal profile are illustrated in figs. 12 and 46. The 
 diagram shows that, as far as possible, butts of adjacent strakes have been 
 kept at least two frame spaces clear of each other, and alternate strakes at 
 least one frame space. The plating of the deck at the sides of hatches 
 and engine and boiler openings has been increased in thickness ; doublings 
 have been fitted at the corners of the openings for all large hatchways 
 and for engine and boiler casings, as, especially where such corners are 
 rectangular, there is the tendency, owing to the severity of the racking 
 stresses in such localities, to rupture. Doublings have also been fitted 
 at the sides of small bunker hatches. When it is intended to sheathe a 
 steel deck with wood, the strakes of deck plating should be arranged 
 alternately, one in, one out. (See Wood Decks, fig. 58.) A slight reduction 
 necessarily takes place in the thickness of the wood deck over all outside 
 strakes of plating. The wood deck over the inside strakes should bear 
 hard down upon the plating, and upon no account should the objectionable 
 practice of fitting wooden packing pieces be permitted, in order to save 
 thicker planking, or the trouble of efficient fitting. A sheathed steel or 
 iron deck should be well coated with reliable bitumastic composition or 
 vegetable tar, for it will easily be seen what havoc may be wrought in the 
 deck plating in the event of water finding its way through defective 
 caulking in any seam or butt in the deck planking, an occurrence not at 
 all infrequent. When the deck plating is unsheathed, it is preferable to 
 arrange the strakes of plating each in and out like the slates upon a roof 
 as shown in the midship sections, figs. 12 and 48. By this arrangement, 
 water on deck naturally flows towards the waterways, and finds its way 
 through the scuppers. Every possible means should be taken throughout 
 
2 24 STEEL SHIPS. 
 
 the vessel to prevent the lodgment of water, if the structure is to be 
 preserved from corrosion and decay. 
 
 All steel decks should be thoroughly watertight, hence every seam and 
 butt must be properly caulked. In the case of upper decks uncovered by 
 erections, the watertightness against the shell plating is easily secured by 
 the gunwale bar, which also ought to be caulked against the shell and the 
 deck. In the way of the erections, and in 'tween decks, where the frames 
 pass through the stringer plate, the watertightness is obtained by fitting a 
 continuous angle bar along the toe of the transverse flange of the frame, to 
 which it is connected either by an angle lug, or the extended reverse bar, 
 and an intercostal angle between the frames against the shell ; the space 
 between these two angles is then filled with good cement. 
 
 Where the most satisfactory and reliable method of effecting water or 
 oil-tightness is desired in a deck through whicli the frames extend, either 
 joggled bars are fitted round all the frames, or else the frames are cut and a 
 continuous angle attached to the shell and deck stringer plate, as shown in 
 the 'tween decks of the midship section of the oil steamer, fig. 78. In this 
 case the continuity of the strength of the frames is maintained by means of 
 bracket plates to the decks. When the continuity of a steel deck is 
 interrupted, as in the case of the raised quarter deck type, where the main 
 deck extends for a certain length of the vessel, and is then raised several 
 feet, after which it is known as the raised quarter deck, compensation for 
 the break must be made. This is done by overlapping the decks (see fig. 47), 
 excepting in very small vessels, when the maiu and raised quarter decks are 
 bound together by large bracket plates well riveted to the bulkhead at the 
 break. 
 
 Wood Declis. — When wood decks are laid upon the beams, they ought to 
 be scored out on the under side in way of all Jtie and deck plates. No 
 packing pieces should be allowed. The butts of the deck planking should 
 be carefully arranged, there being at least three clear shifts between any two 
 butts in the same beam space. Beams under wood decks are usually fitted 
 on alternate frames, and are generally of bulb plate and double angles or tee 
 bulbs. The butts of the deck planks should always come on the centre of 
 the beam or on a deck plate. The butts are practically always of the 
 vertical kind as shown in fig. 143, A. 
 
 The fantastic butts sometimes illustrated are seldom, if ever, adopted in 
 mercantile shipbuilding practice. When the deck planks are 6 in. in width 
 or under, a single through-bolt through every beam is sufficient. When the 
 planks are above 6 in. and not more than 8 in, in width, there should be one 
 through-bolt and a short screw-bolt on the under side in every beam. All 
 planks over 8 in. in width should be fastened with two through-bolts 
 through every beam. Through-bolts should be nut and screw. It is 
 preferable that all bolts be galvanized. They should be properly sunk in 
 the deck with their heads well bedded in oakum and white lead, and their 
 heads should be carefully covered with turned dowels also bedded in white 
 
Fig 143. 
 PLAN OF UPPER DECK OF VESSEL ILLUSTRATED IN Fios. 12 and 46. 
 
DETAILS OF CONSTRUCTION. 225 
 
 lead or some suitable composition. (See fig. 143, A, for sketches of bolts, 
 dowels, etc.) 
 
 The beams should be well painted with good red lead pai-nt before the 
 decks are laid ; and in some cases, in order to preserve the appearance of the 
 deck, by preventing discolouring due to rust water oozing through the butts, 
 felt, dipped in red lead, is placed under the butts. In very high class work, 
 zinc slips are fitted on the beams under the butts. 
 
 It is always better that the wood deck, alongside of iron deck houses, 
 waterways, hatch coamings, etc., be of either teak or greenheart. All seams 
 and butts should be caulked with oakum, and filled in with marine glue or 
 some other suitable composition. When openings are cut through the deck 
 for companions or skylights, care should be taken that the deck planks next 
 to the openings are so secured as to prevent their yielding, due to the 
 pressure caused by caulking. An excellent method of fitting an angle bar 
 round such openings is illustrated in fig. 143, B. For the same reason, an 
 angle bar should be fitted alongside all waterways. 
 
 All wood for decks should be free from sap, shakes, and knots, as far as 
 possible. It should be well seasoned, and of good quality. Yellow pine, 
 pitch pine, and teak, are among the principal timbers used, and not less 
 than six months should be allowed for seasoning. 
 
 Teak decks, which are very durable, may be of less thickness than 
 would be required for pine decks. In laying a deck, care should be taken 
 that the weather side (the hard side) be uppermost, with the heart 
 downwards. 
 
 "When a wood deck is laid upon a steel deck, the butts and fastenings 
 may come between the beams. 
 
 In order to get rid of the water which on weather decks naturally drains 
 into the waterways, it is necessary that scuppers be fitted. Fig. 143, C, D, 
 E, and F, shows four different kinds of scuppers. 
 
 C shows a small. aperture, 3 in. or 4 in. in length, cut through the sheer 
 strake immediately above the deck stringer plate. The gunwale bar is 
 preserved continuously as shown. 
 
 In F, owing to the gunwale bar being cut at the scupper, a compensation 
 bar is fitted on the under side of the stringer plate. 
 
 D and E are forms of scuppers adopted in bridges, or poops, or 'tween 
 decks, water finding its way through these by a perforated brass plate fitted 
 into the stringer plate, while storm valves are usually fitted to the shell 
 exit to prevent water finding its way into the ship through the scuppers ; or, 
 in some cases, screwed lids are fitted on to the perforated deck plate for the 
 same purpose. 
 
 Outside Shell Plating. 
 
 TThile the skin or covering over the framing of all vessels is an 
 indispensable part of the structure, in iron or steel vessels the outside 
 shell plating, while performing the work of a watertight skin, is also 
 
 15 
 
2 26 STEEL SHIPS. 
 
 the principal structural item in the ship girder. A steel ship being 
 composed of a vast number of minor girders, each contributing its share to 
 the strength of the whole, it follows that to consider alone the isolated 
 eflficiency of any particular part is unfair. Thus, while the shell plating 
 affords enormous longitudinal strength to the vessel, it depends upon the 
 assistance of the transverse framing, as well as upon the keelsons, stringers, 
 and decks, in being held to its work, and developing its fullest efficiency. 
 Some parts of the outside plating develop more strength than others. This 
 we investigated in Chapter V. on ' Stress and Strength.' Thus, it is 
 usual to find in all seagoing vessels that the plating is of increased thick- 
 ness on the topsides, particularly in the sheer strake, and possibly in tlie 
 strake beneath. In some cases, especially in large long vessels, the sheer 
 strake may even be doubled. As the sheer strake is usually and preferably 
 an outside strake, such a doubling would generally be placed on the 
 inside of the sheer strake. Increased thickness of shell plating is often 
 found to be necessary in one or more strakes of the bilge plating, which is 
 the lower extremity of the ship girder, just as the sheer strake forms the 
 upper. As the longitudinal stresses are greatest over the middle length of 
 the vessel, and are gradually reduced towards the ends, a reduction in 
 structural strength is permissible. We therefore find that all classification 
 societies allow for this in their rules. The midship thickness is usually 
 continued over the half length amidships, reduction gradually taking 
 place towards the ends. However, when we come to the keel and stern 
 frame, the local requirements demand that thick plating be adopted. In 
 the case of the keel, this is easily understood when it is remembered 
 that there is the possibility of the vessel frequently lying aground, with 
 consequently severe wear and tear. The necessity for thick plating, 
 where flat plate keels are adopted, is obvious. When a keel of the bar 
 type is fitted, the strakes of shell plating which are attached to it on 
 either side (the garboard strakes) again require to be of exceptional 
 thickness. This is necessary in order to obtain plating somewhat 
 compatible with the thickness of the bar keel and its large, widely-spaced 
 rivets. The absurdity of riveting a thin plate to a thick bar with large 
 rivets widely spaced, is evident if watertightness is to be ensured. As 
 previously pointed out, the lower edge of the garboard strake should be 
 kept about a quarter of an inch from the bottom of the keel, not only 
 to facilitate caulking, and relieve the plating of strain in the event of 
 grounding, but to preserve the caulking and watertightness which would 
 suffer, should such vessels be accustomed to chafe on sandy bars of 
 bottoms. The shell plating which is connected to the stern frame should 
 be at least of the midship thickness for reasons just given, where plates 
 have to be connected to thick bars or forgings. Moreover, the vibration 
 which is suffered in this locality, owing to the action of the propeller, when 
 being immersed and emerged in heavy seas, demands that the parts of the 
 material connected have some relation towards each other. Thick plates 
 
DETAILS OP CONSTKUCTION. 227 
 
 are particularly required round the boss. As both boss and oxter plates 
 (shell plating over propeller arch) have to be furnaced and hammered 
 in order to be brought to their required shape, these plates must be 
 annealed before being fixed in position. By increasing the thickness of 
 the shell plating at the fore end for about 20 or 30 ft. from the stem, 
 ships which have frequently to break their way through ice may be 
 considerably strengthened. Such thickening of shell plating will also 
 enable the vessel to more effectively resist panting action, and minimise 
 the tendency in lightly-built vessels to fall hollow between the frames, 
 owing to the pressure upon the bows when driving ahead. This also makes 
 a more satisfactory connection to the stem bar. Where this increased 
 thickness is adopted only on account of the possibility of ice being 
 encountered, such increase of thickness need only extend between light 
 and load lines. One or two shell strakes on either side of the keel 
 should of necessity preserve their midship thickness to the stem, on 
 account of the severity of the thumping which ships experience when 
 driving ahead and plunging into head seas. 
 
 A most important consideration in the arranging of shell plating 
 should be to ensure a good disposition of butts. No butts in adjacent 
 strakes should come nearer to each other than two frame spaces, and in 
 alternate strakes than one frame space. Similarly, butts of deck 
 stringers, side stringers, tank margin plates, centre through-plates, 
 keelsons of bar keels, or butts of side bar keels, should be kept clear. 
 
 Fig. 144 illustrates an expansion of the shell of one side of a vessel, and 
 shows the disposition of butts, etc. 
 
 Where very short plates are adopted in the outside shell, to carry out 
 such an arrangement would be impossible, hence it is usual to stipulate 
 that no shell plates be less than 7 frame spaces in length, excepting the 
 plates at the extreme ends of ships. Indeed, with the adoption of larger 
 and improved machinery in recent years, shell plates of 10 and more 
 frame spaces in length are commonly adopted. The nature of butts and 
 seam cormections, including the riveting, has been fully dealt with and 
 illustrated in Chapter VII., p. 167, 
 
 Packing. — When the strakes of shell plating are arranged alternately, 
 one inside {i.e. bearing upon the frames) and one outside {i.e. the edges 
 resting upon the edges of the adjacent strakes and forming the seams), as 
 illustrated in figs. 12, 48, etc., it will be seen that something must be done 
 to secure a firm connection between the outside strakes and the frames. 
 This is effected by fitting what is called 'packing iron' on all frames in 
 way of all outside strakes of plating. This packing iron must necessarily 
 be of the thickness of the adjacent inside strakes. The riveting through 
 all outside strakes and the frames must therefore pass through this packing 
 iron. Packing iron is necessary, not only in the shell plating, but in decks, 
 tank top, etc, when this arrangement of plating is adopted. When a steel 
 deck is laid with the strakes of plating in and out, as shown in the three 
 
2 28 STEEL SHIPS. 
 
 &teel decks of fig. 48, and also on the tank top of the same fig., the packing 
 iron must of necessity be tapered to snit the shape of the space "between the 
 beam or reverse bar and the plate, as the case may be. Sometimes an in 
 and out strake is adopted in the shell plating (see the bilge strake in fig. 
 48), when taper packing is likewise necessary. All this packing obviously 
 adds greatly to the weight of a vessel ; but in recent years it has been 
 largely dispensed with by joggling the edges of what Avould be the out- 
 side strakes of shell plating, decks, tank tops, etc. (See figs. 64, 65, 
 and 78.) 
 
 By this means, not only is a saving of weight eflEected, but superior 
 workmanship is assured, for a standing rule ought to be, never to rivet 
 through more than two thicknesses if it can possibly be avoided,— the less 
 the number of thicknesses, the more reliable the rivetiu". 
 
 The strength of butt joints, and the size and spacing of rivets in both 
 seams and butts, has already been dealt with in Chapter VII., p. 167. In 
 order to obtain watertightness, and to thoroughly close up the shell plating, 
 the following two points should be noted : — 
 
 Where the seam over-lap of an outside strake passes over a butt lap in 
 the adjacent strake (see section through plate landing x y, fig. 145) the 
 plating of the butt under the landing is planed ofi" in a tapered form, as 
 shown in the diagram, so as to allow the outside strake to bear close on 
 the inside strake, without leaving a tapered aperture, which would otherwise 
 exist were this planing operation not adopted. In the event of this not 
 being done, a piece of tapered packing iron would have to be fitted. In 
 lapped butts of outside strakes, it is not practicable to so plane the butt, 
 and thus tapered packing pieces long enough to take at least three pairs of 
 rivets are fitted. 
 
 Wherever the shell plating is weakened owing to excessive perforation 
 by rivet holes, or by the cutting of doors or openings of any kind, compensa- 
 tion should be introduced. Thus, in way of all watertight bulkheads, the 
 shell is perforated by a row of closely spaced rivet holes. This weakness is 
 compensated for by the fitting of bulkhead liners in way of the outside 
 strakes. (See figs. 139, 140, 144, 83, and 84.) Bulkhead liners have already 
 been more fully dealt with in the section on 'Bulkheads,' as well as in the 
 section on ' Oil Steamers.' When side lights are cut through a sheer strake, 
 compensation for the weakening of the strake should be introduced, either by 
 increasing the thickness of the plating, or doubling the same, or fitting stout 
 angle bars for at least one frame space in length over all such openings. 
 When openings for doors of any kind are cut through the shell plating, 
 compensation must be made by fitting doubling plates above and below such 
 openings by framing the opening with stout angle bars, and by fitting 
 vertical webs upon each side of the opening, and a fore and aft girder upon 
 the shell above these openings extending a few feet fore and aft on each side 
 of the doorway. Such of these methods should be adopted as are in 
 accordance with the special requirements of the case. In very large 
 
DETAILS OF CONSTRUCTION. 
 
 229 
 
 vessels, and in vessels of extreme length in proportion to tlie depth, to 
 fit doublings to the sheer strake, and in some cases, even to the strake 
 
 
 'il 
 
 
 -Sp 
 
 c» 
 
 ^ 
 
 m 
 
 l^ 
 
 
 
 <tt 
 
 to 
 
 
 1 
 
 4- 
 
 
 :f5 
 
 Co 
 
 r 
 
 ?:■?= ^ !5 
 
 ^^ ^^ 
 
 ■Si "^•5' 
 
 •3:3 to ^i^s 
 
 ?= '^ ^ 
 
 £ to 
 
 to 
 .S o 
 
 Ph 
 
 below is sometimes necessary. When this is done, the butts are usually 
 
230 STEEL SHIPS. 
 
 connected by means of straps, these being necessary for the doublings as 
 v/ell as the other plating. 
 
 At the ends of bridges and raised quarter decks or any other erections 
 terminating on the half length amidships, doublings should be fitted to the 
 sheer strake, or, in the case of raised quarter decks connected to bridge 
 houses, the raised quarter deck side plating above the sheer strake should be 
 doubled, the doubling to extend from four or more spaces along the bridge 
 side to a considerable distance abaft of the break according to the size of the 
 vessel. Doubling plates should be fitted round hawsepipes. (See Doublings, 
 fig. 144.) 
 
 The bulwark plating at the ends of long poops and forecastles, and at 
 both ends of bridges, should be increased in thickness, so as to somewhat 
 reduce the sudden interruption in longitudinal strength, Avhich such erections 
 create. All openings through the bulwarks for waterport doors, or mooring 
 ropes, should have narrow doubling plates fitted round them. As the 
 bulwark is usually constructed of thin plating about ^ in. in thickness, it 
 obviously requires stiffening upon its upper edge. This may be done by 
 fitting special sections of bar iron, which are rolled for this purpose, or bulb 
 angles with half round iron on the outer edge, or even in some of the cheap 
 tramp types of vessels, a couple of half rounds, one on either side of the top 
 edge of the bulwark plating, may be fitted. (See figs. 10, 11, 12, 48.) 
 
 The bulwark plating also requires supports of some kind from the deck. 
 These may be specially forged round iron stanchions with a palm on the top 
 and a fork at half the height, also with a palm to take the bulwarks, the 
 lower end of the stanchion being riveted to the deck. Other forms of 
 bulwark supports are also made of tee bars and bulb plates. See figs, last 
 referred to. 
 
 These bulwark stays should not be spaced more than about 6 ft. apart. 
 
 When the butts in the sheer strake are strapped it is preferable to fit 
 the straps on the outside, covering the full width of the plate. But 
 as this is not only unsightly, and often inconvenient, the straps are 
 fitted on the inside, when they should extend from the edge of the strake 
 below to the extreme top of the sheer strake, in which case the gunw^ale 
 angle bar should be joggled round the straps. When double straps are 
 required to the sheer strake, the outer one must extend over the full width 
 of the plate, while the inside one may extend from the edge of the strake 
 below to the deck stringer plate, and be fitted in a separate piece from the 
 bosom of the gunwale bar to the top of the sheer strake. In this case, a 
 packing piece must be fitted above the top flange of the gunwale bar. 
 
 The sheer strake ought to extend sufficiently high so as to get two rows 
 of rivets through the straps above the gunwale bar. 
 
 All butts should be carefully planed and fitted close. All butts and 
 seams must be thoroughly caulked. 
 
 In preparing a shell plate for its position in the ship, the position of all 
 frame rivet holes (which are punched before the frames are erected) and 
 
DETAILS OF CONSTRUCTION. 23 I 
 
 rivet holes in the seams of adjacent plates, are transferred to the new plate, 
 by means of a template. A template is a framework made of thin wood 
 large enough to cover the area of the plate, with wooden bars across in way 
 of each frame. This is pinned on to the vessel in exactly the same position 
 which the plate itself will occupy, and by means of a whitened cylinder 
 the positions of the rivet holes are transferred to the template. The 
 template is then taken down and laid upon the plate, and by means of 
 what is called a reversing tool the rivet holes are re-transferred to the plate, 
 which is then carried away to be punched, countersunk, and sheared to 
 its exact size. It is then lifted into position in the ship by means of block 
 and tackle, and securely bolted in place. Indeed, the plate must be held 
 by a sufficient number of bolts so as to bring it absolutely close to all 
 surfaces upon which it must fay. In riveting such a plate, especially 
 where tlie length is great, the riveting is performed from the centre 
 toward the outer edges. By this means, if any slight expansion occurs, 
 due to hammering, and heat from the rivets, it is evenly distributed in 
 all directions, and is therefore practically unnoticeable. 
 
 In those parts of the shell where there is considerable change of form, 
 the exact curvature of the plate is ascertained by strips of bar iron (set 
 irons) bent to the correct shape at two or more positions in the length of the 
 plate. The greater number of shell plates can be bent to their required 
 shape in the ' rolls,' manipulated by skilful platers. 
 
 Plates of more irregular shape need to be furnaced and hammered into 
 the required form, and should be annealed. 
 
 Stern Frames and Rudders. 
 
 Stern Frames. — The function of the stern frame in a single screw steamer is 
 to afford support to the rudder and to the tail end shaft and propeller, and to 
 provide an aperture in which the propeller may revolve. The usual form of 
 stern frame for vessels with single screws is shown in fig. 146. Its 
 principal parts are an after or rudder post into the gudgeons of which 
 the rudder pintles are fitted, and a fore or propeller post through the 
 boss of which the tail end shaft passes to carry the propeller. It is 
 obvious, therefore, that with the enormous stress which must be borne by 
 the stern frame, especially in large high-speed vessels, that it must be 
 exceptionally strong in itself, and at the same time that the connection to 
 the main hull of the vessel must be of the most efficient character. 
 
 Stern frames may be of cast steel or forgings made by first forging parts 
 of the stern frame in separate pieces, and then welding the whole together. 
 Naturally, great care must be exercised in the manufacture of a stern frame. 
 The material should be as free as possible from any excess of those elements 
 which are known to produce brittleness. If castings are adopted, they 
 should be free from blow-holes and cavities of any kind, and in the case of 
 
232 
 
 STEEL SHIPS. 
 
 p -r-Tmsom plate 
 
 Check 
 
 n 
 
 I'iG. 146.— Stern Frame for a Single-Screw Steamer. Frame in two 
 parts connected by Scarphs. 
 
DETAILS OF CONSTRUCTION. 233 
 
 forgings, the utmost care is necessary in making the welds uniting tlie 
 several parts. The tests for stern frames have already been given on page 
 24. The classification societies definitely specify the dimensions of a 
 section through any part of the stern frame, based, as a rule, upon the 2nd 
 numeral. A safe rule to follow is to preserve the standard sectional area, 
 if the vessel is classed, through every section in the circumference of the 
 frame. When a vessel trims by the stern, it follows that if she comes into 
 shallow water, the aftermost extremity of the keel, whicli would be formed 
 by the base of the stern frame, would be the first to touch the bottom. 
 This has frequently happened, with the result that the stern frame has been 
 broken usually at the bottom of the aperture. Sometimes, by fitting a 
 shoe strap, the stern frame may be sufficiently repaired so as to continue 
 to serve its purpose. In other cases, where the damage has been more 
 extensive, the whole stern frame has had to be renewed. 
 
 When the whole stern frame is made of a casting or forging in one piece, 
 the enormous difficulty, and labour, and expense, not to mention loss of 
 time incurred in extricating the broken stern frames from the hull, and in 
 ordering and refitting a new stern frame, will be apparent. To obviate this, 
 the arrangement shown in fig. 146 whereby the stern frame is made in two or 
 more pieces connected by means of scarphs as illustrated is now commonly 
 adopted, especially in large vessels, and as the damage from grounding 
 usually takes place either at the bottom of the aperture, or in the after 
 post, it will be seen that this piece can be renewed with comparative ease. 
 A commendable method which may often save a stern frame in the event 
 of grounding, is to lift up the frame at the bottom of the aperture from the 
 propeller post to its aftermost extremity, as shown in figs. 146 and 157. As 
 this may, in some cases, bring the lower part of the stern frame round the 
 bottom of the aperture, in too close proximity to the propeller, the frame is 
 reduced in depth, but increased in breadth at the bottom of the aperture, 
 preserving the sectional area, as illustrated by the plan of the stern frame, 
 fig. 146. 
 
 The connection of the stern frame to the hull, which, as previously 
 stated, should be of the most thorough character, is effected — 
 
 1st. By carrying up the after, or rudder post, to a height sufficient to 
 enable a good connection to be made to a thick plats called a transome 
 floor, which extends from side to side of the vessel above the counter on 
 the fore side of the post. (See fig. 146, B and C.) Lloyd's require that the 
 depth of this transome plate be at least one and a half times the depth of 
 an ordinary midship floor. 
 
 2nd. In large vessels, and in all single screw vessels of high speed, the 
 fore or propeller post should extend to a height sufficient to enable a good 
 connection to be made to another deep transome or floor plate. Lloyd's 
 require the propeller post to be extended as just described in all screw 
 steamers whose plating number is 20,000 and above. 
 
 The rudder and propeller posts are connected to their respective transome 
 
234 STEEL SHIPS. 
 
 plates, either by means of large stout angles (see fig. 146, B), or by having 
 a projecting flange upon the post itself, through which the rivet holes are 
 drilled for the connection to the transome plate. (See C.) 
 
 3rd. The lower part of the stern frame which projects forward of the 
 propeller post, should be of sufficient length to enable a good connection 
 to be made to the keel. In screw steamers, shipbuilders usually make this 
 projection at least three frame spaces, Lloyd's require that the projection 
 be at least 2|- frame spaces before the propeller post in steamers, and IJ in 
 sailing vessels and paddle steamers. When the keel is of the bar type, the 
 stern frame is connected to the keel by a scarph, as shown in fig. 99, which 
 is in every way similar to the scarphs uniting the several lengths of the bar 
 keel. When the keel is of the side bar type, the scarph is formed as shown 
 in fig. 100. When the keel is of the flat plate type, the projection forward 
 of the propeller post is usually formed as shown in fig. 146. It will be 
 seen that it tapers in respect to its depth, and forward of the check on the 
 bottom of the projection, where the keel plate terminates, it is hollowed 
 out, and conforms to the shape of the bottom of the vessel. The keel 
 plate is thus riveted to the stern frame by independent rivets for each side. 
 
 4th. The stern frame is finally attached to the hull by means of the 
 shell plating, which is riveted to it by at least two rows of rivets arranged 
 in zigzag fashion. These rivets should be spaced not more than five 
 diameters apart, and should be about a quarter of an inch larger in diameter 
 than would be required according to the thickness of the plates, which 
 ought to be, at least, as thick as the garboard strakes. Lloyd's, however, 
 state that they need not exceed 1^ in. in diameter, as it is practically 
 impossible to get efficient hand riveting with larger rivets. 
 
 Although the shell plating is reduced in thickness from the half-middle 
 length to the ends, the plates which take the stern frame should be at least 
 equal to the midship thickness or garboard strake thickness, and especially 
 so round the boss of the propeller post, for the absurdity of connecting a 
 huge, heavy stern frame, weighing twenty or more tons, to comparatively 
 thin plating by large rivets must be apparent. 
 
 The rudder is supported and secured to the propeller post of the stern 
 frame, though with every facility to swing freely, by means of pintles, which 
 fit into a number of gudgeons on the rudder post of the stern frame. De- 
 tailed sketches of these gudgeons are shown in figs. 147-151. The spacing 
 of these gudgeons upon the stern post should not exceed 5 ft. 6 in. Whether 
 the stern frame is of cast steel or forged iron, the gudgeons should form an 
 integral part of the post. The weight of the rudder is usually borne upon 
 the bottom gudgeon, though, in some cases, the weight is taken upon one of 
 the decks, through which the rudder stock passes, by means of a collar forged 
 on to the post which rests upon a suitable casting firmly secured to the deck 
 and well supported. However, the after stern post is not so absolutely 
 necessary as has sometimes been supposed. In the profile of the turret 
 steamer (fig. 63) built by Messrs. Doxford & Sons, the after post is 
 
DETAILS OF CONSTRUCTION. 
 
 dispensed with in way of the screw aperture, the fore post being projected 
 aft in order to afford support to the lower extremity of the rudder, the 
 weight of the rudder in this case being carried at the deck. An advance 
 has even been made beyond this, and Messrs. Alfred Holt & Co., of 
 Liverpool, have shown in their latest single-screw steamers that the 
 huge, heavy, costly, and cumbersome stern frame can be dispensed with 
 altogether by the ingenious and well-thought-out arrangement illustrated 
 in fig. 75. 
 
 This system of stern arrangement has already been dealt with on p. 150. 
 Notwithstanding the radical change which has taken place in this stern 
 
 Plan of Nut 
 
 Fig. 147.— Fitted Pintle. 
 
 Fig. 148.— Fitted Pintle. 
 
 arrangement, and the suspicion with which such innovations are often 
 viewed, several of these vessels have performed with complete success the 
 round journey to China and back, carrying a deadweight of over 8000 tons, 
 while the increased facility for steering has proved a decided advantage. 
 Cutting away and dispensing with the fin of plating extending from below 
 the boss to some distance forward, has not only lessened the resistance to 
 steering, but reduced the severity of the straining and vibration which is 
 usually experienced in this neighbourhood. 
 
236 
 
 STEEL SHIPS. 
 
 The rudder, which is of special design (see fig. 75), is of such a 
 shape as to offer the least resistance to propulsion compatible with the 
 principle of its design. 
 
 Stern Arrangement for Tivin Screios. — In all vessels with twin screws, there 
 
 Rudder 
 Post 
 
 Stern 
 Post 
 
 Fig. 149, — Pintlets forged on Rudder Post (unbnshed). 
 
 comes a place towards the stern where the propeller shafts would emerge 
 from the hull, if the usual form of hull were in no way modified. The 
 length, from the point where the shafts so emerge to the propellers, will 
 vary according to the size and fineness of the lines of the vessels, therefore 
 
DETAILS OF CONSTRUCTION. 
 
 237 
 
 in all such cases the shafts require support at their aftermost lengths. The 
 old-fashioned method of affording this support was by bossing out a few 
 
 Fig. 150. — Locking Pintle. 
 
 frames round the shaft in the neighbourhood where it would emerge 
 through the shell plating, and at the aftermost extremity supporting the 
 
 Fig. 15L— Bottom Pintle. 
 
238 
 
 STEEL SHIPS. 
 
 shafts by means of struts such as shown in figs. 153, 154, 155, and 156. 
 These struts may be either forgings or castings. 
 
 In fig. 153 there is a stern frame somewhat after the pattern required 
 for a single screw. The aperture in this case is necessary, because the 
 propellers overlap to some extent, and, as the diagram shows, the stern 
 frame is manipulated to receive the palms at the upper and lower ex- 
 tremities of the struts. 
 
 In fig. 154 the propellers are amply clear, and no aperture is necessary. 
 In this case, the upper palms of the struts are riveted on to the shell plating, 
 which is doubled, and supported by a thick bulkhead with large, thick, 
 double angles through which the rivets of the upper palms pass. (See fig. 
 155.) The lower palms are riveted through the keel bar, or through the 
 keel part of the stern post. In fig. 156 the struts pass through the shell 
 plating, and the upper palms are secured to a fore and aft, and to a trans- 
 verse plate, as shown ; but by far the best method of supporting twin-screw 
 
 Fig. 152.— Rudder Stopper. 
 
 shafts at their aftermost extremities is to carry the framing round the 
 shaft right aft to the propellers, and to work the shell plating round this. 
 With a sufficient number of transverse ties and bulkheads, this arrange- 
 ment is exceedingly strong, ~ and much less liable to suffer damage than 
 the strut arrangements previously described. 
 
 Special support is provided for the extreme ends of the tail shaft by 
 means of what are commonly known as the spectacles. This is usually a 
 steel casting, in one piece, firmly bolted through the fore post of the stern 
 frame, and securely riveted to the bossed plating of the shell. The notes 
 upon the diagrams are sufiicient to give a general idea of this mode of 
 construction (figs. 157 and 158). 
 
 Rudders. 
 
 Eudders for merchant steamers may be divided into two kinds, the 
 ordinary frame rudder and the plate rudder. The frame rudder is usually 
 
DETAILS OF CONSTRUCTION. 
 
 239 
 
 SECTION THROUGH 
 A.B. 
 
 (&.) Transverse Section (enlarged). 
 
 Fig. 153.— Connection of Struts with Stern Frame in Twin-Screw 
 Steamers : a, Elevation : &, Transverse Section, 
 
240 
 
 STEEL SHIPS. 
 
 WATERTIGHT FLAT ^ 
 
 Y ; §1 ■ 
 
 CENTRE. OF SHIP 
 
 Fig. 154. — Mode of Strengthening Ship at the after end and attaching 
 Struts to Shell Plating. 
 
 f 
 
 l(^ 
 
 -i. 
 
 o '00 o 
 ; III 
 i III I 
 
 o I ; 1 o 
 
 o 
 
 ' D Cf U 
 
 L. 
 
 o 
 
 NiO — 
 
 i— , 
 
 n 
 
 V I 
 
 ENUARGED VIEW AT YJ 
 
 Fig. 155.— Mode of Strengthenirig Ship at the after end and attaching 
 Struts to Shell Plating. 
 
DETAILS OF CONSTRUCTION. 
 
 241 
 
 Sectional Elevation of Struts. 
 
 TOP OF ~- 
 WATERTIGHT CHAMBERS 
 
 Fig. 156.— Struts carried through Shell Plating into Watertight Chamber. 
 
 16 
 
242 
 
 STEEL SHIPS. 
 
 d 
 
 c 
 
 Y(^'7JeT sections ^ 
 
 Fore stern frame post 
 
 Bolts securing propeller 
 brackets or spectacles 
 to stern post 
 
 Spectacles for the support 
 of the tail end shafts 
 
 Fig. 157. — Stern Frame and Spectacles for Twin-Screw Steamer, 
 
DETAILS OF CONSTRUCTION. 
 
 243 
 
 2nd Position. Frames considerably bossed 
 for Shaft, 
 
 Fig. 158. — Showing Bossing of Frames in Twin-Screw Steamer at 3 intervals. 
 
244 
 
 STEEL SHIPS. 
 
 a forging, and in construction as illustrated in fig. 159. The rudder frame- 
 work is plated on each side, and the space between the two plates is filled 
 in with wood. The number of stays, stiffening the rudder in a fore and 
 aft direction, vary with the size of the rudder, but one is usually placed 
 opposite each gudgeon. Until within recent years, the rudder frame and 
 stock was generally all one forging, but the awkwardness and incon- 
 venience of unshipping a large rudder of this kind led to the adoption 
 of the method illustrated in fig. 159, in which the upper rudder stock is 
 
 Check for plate 
 
 Locking Pintle 
 
 r. 
 
 d 
 
 Fitted Pintles 
 
 ^ 
 
 ■u 
 
 R 
 
 S 
 
 Fig. 159. — Frame Rudder. 
 
 ■s^ 
 
 an entirely separate forging, being connected to the main rudder itself by 
 a bolted coupling, as shown. 
 
 The plate rudder is somewhat different. It consists of a main post 
 usually coupled to the stock, and instead of a plated frame, a very thick 
 plate formed to the required shape of the rudder is notched into the main 
 rudder post, and supported at intervals with arms alternately on each side, 
 
DETAILS OF CONSTRUCTION. 
 
 245 
 
 which may be either forged into the post or shrunk on, as shown in fig, 
 160. The rivets connecting the side plates to the rudder frames in the 
 case of frame rudders, and the thick plate to the arms in plate rudders. 
 
 Enlarged view 
 of coupling 
 
 JEQi 
 
 "Key 
 
 kO 
 
 triY 
 
 Groove in Rudder Post 
 for plate 
 
 Plan of Coupling 
 
 ffmfifei^ocH 
 
 Section ofPudder through an arm 
 
 " -— -^ 
 
 ^ 
 
 r^ 
 
 :. . .0 .» .! !■ , . -.1 
 
 c-:y.vTA--:>-.i"«-=v£] 
 
 
 U 
 
 ] 
 
 
 fli 
 
 (,...... .... ; . : , : 1 
 
 Rudder Plate 
 1 '{thick 
 
 
 u 
 1 
 
 
 \ 
 
 rPi 
 
 P Udder Post 
 
 Section through Ruddet' between tne 
 arms showing groove in post 
 Fig. 160. — Plate Rudder, showing Arms shnmk on and keyed. 
 
 are usually spaced about five diameters apart. In the case of classed 
 vessels, the size of rivets is specified by the registration society. 
 
 Stern Frame Gudgeons and Rudder Pintles. — As before stated, the 
 
246 STEEL SHIPS. 
 
 gudgeons are forged on to a forged stern frame, or form part of the casting 
 when of cast steel. The pintles on the rudder may either form part of the 
 forging of the rudder frame, as shown in fig. 149, or, as is much more 
 commendable, the pintles are fitted separately, and held in position by a 
 nut with a check pin, which bears upon the head of the nut. Fitted 
 pintles are illustrated in the rudder frames (figs. 159 and 160), and in the 
 detailed sketches, 147, 148, 150, and 151. 
 
 A considerable amount of wear and tear must eventually take place 
 owing to the incessant wearing action incurred by the rudder's movements 
 when the vessel is under way. Numerous devices are adopted to prevent, 
 or, at any rate, to make the wear and tear of such a nature that it can 
 be repaired with a minimum of inconvenience and expense. When the 
 pintles are forged to the rudder and fit into the gudgeons, as in fig. 
 149, with no other bearing surfaces than the bare metal of which they 
 are made, repair would be no easy matter. But when the pintles are 
 portable, and the gudgeons are bushed Avith lignum vitae or steel or 
 gun-metal, and the pintles bushed with brass after the manner illustrated 
 in figs. 150, 151, 147, and 148, the wear and tear being confined to the 
 bushes, these parts are renewed at comparatively little trouble and expense. 
 When the weight of the rudder is transmitted through the bottom pintle to 
 the stern frame, the wear and tear is naturally more rapid at this part. A 
 good plan is to fit a round-topped steel disc, which is easily extracted and 
 renewed, if a hole about one inch in diameter be drilled through the part of 
 the stern frame beneath it, as shown in fig. 151, or, instead of the steel 
 disc, a piece of lignum vitae with the grain up is fitted into the bottom of 
 this gudgeon hole, forming a good bearing for the bottom pintle. 
 
 To prevent the possibility of the rudder lifting by a heavy blow from 
 a sea, it is common to fit one of the upper pintles with a forged head upon 
 its lower end. This is called a locking pintle. (See fig. 150.) 
 
 As rudders are not usually allowed to turn more than an angle of 40° or 
 45° on either side, it is essential to have what are called stoppers, usually 
 upon the gudgeons, and, if necessary, upon the rudder as well. Fig. 152 
 illustrates a rudder stopper, and the manner in which the gudgeon is formed 
 so as to prevent the rudder turning beyond the desired angle. Two stoppers 
 are usually fitted to each rudder, one upon an upper and one upon a lower 
 gudgeon. (See figs. 159 and 160.) As the rudder stock tapers towards its 
 lower extremity, a projection is sometimes made upon the rudder for the 
 lower stopper, in order to get sufficient bearing surface for the check at the 
 maximum angle. 
 
 Miscellaneous Details. 
 
 Continuity of Strength. — Nothing is of more vital importance in ship 
 construction than that the continuity of all the principal structural parts be 
 rigidly maintained. Wherever interruption takes place in the mode of 
 
DETAILS OF CONSTRUCTION. 247 
 
 construction in any part, which in itself would produce weakness, compensa- 
 tion must be introduced in such a manner that the continuity of strength 
 will be assured. Thus : — 
 
 (1) "Whenever a deck has a break in any part of its length, the one 
 part being at a higher level than the other, as in the case of a main and 
 raised quarter deck, the strength of the two parts must be united by 
 overlapping the decks, side stringers, and increasing the thickness of the 
 shell plating in way of the break. (See fig. 47.) 
 
 (2) Sometimes the bottom of a ship is built with ordinary floors, one 
 part of the length, and the other with a double bottom tank. "Wherever 
 this occurs, the strength of the one must be scarphed into the strength of 
 the other by carrying the keelsons on the ordinary floors into the tank for 
 three or more frame spaces. 
 
 (3) Where decks or stringers are unduly weakened by openings through 
 the deck, the strength must be maintained by increasing the thickness of the 
 remaining plating, or by doublings. The same applies to the outside shell 
 plating. (Figs. 143 and 144.) 
 
 (4) If stringers or keelsons are cut in way of bulkheads, their strength 
 must be continued by connecting them to the bulkheads with large bracket 
 plates. 
 
 (5) Where the transverse framing suffers interruption in its continuity 
 from keel to gunwale, as at the margin plate of double bottoms, and some- 
 times where it passes through a watertight flat, the continuity of strength 
 must be maintained by fitting large bracket plates (tank knees). (Figs. 47 
 and 78.) 
 
 Engine and Boiler Space. — The structural strength of the engine and 
 boiler space needs special consideration. The great interruption which 
 takes place in transverse strength owing to the cutting of so many deck 
 beams in order to make the light and air openings through the decks, and 
 the omission of so many hold beams, necessitates efiicient compensation. 
 
 "We have already seen how the deck ends are bound together and carried 
 on the thick fore and aft coaming plates, but, in addition to this, a good 
 number of extra strong continuous beams should be fitted wherever 
 practicable in way of all the decks ; and by converting these, where possible, 
 into semi-box beams by uniting and covering them with plating, valuable 
 strength is introduced. (See figs. 47, 48, 49, and 57.) 
 
 Especially in high-speed vessels is there the tendency to excessive 
 vibration where this part is in any measure weak. To ensure against this, 
 and to make the vessel as rigid as possible, not only are the reverse bars 
 doubled on the floor plates under the engines and boilers, but the double 
 bars are sometimes carried to the upper deck. To compensate for the loss 
 of through-beams, it is also found to be very beneficial to the vessel to 
 introduce several web frames into these spaces, and to thoroughly connect 
 the stringers to them, as illustrated in fig. 48. In vessels having double 
 bottoms, one or more additional intercostal girders should be introduced. 
 
248 STEEL SHIPS. 
 
 at any rate under the engines, while the tank top plating in these spaces 
 should be increased in thickness. On account of the rapid corrosion 
 which shipowners find to take place, especially under boilers, the tank top 
 plating is sometimes entirely dispensed with, and a form of construction 
 similar to that shown upon fig. 161 is often adopted. 
 
 It will be noticed that the tank margin plate is carried continuously all 
 fore and aft, while the keelsons shown must be scarphed into the double 
 bottom. In some cases the double-bottom form of construction is carried 
 continuously all fore and aft, with the exception that large openings are left 
 through the tank top plating between the floors and the fore and aft girders, 
 which are so large as to permit of easy and ready access at all times. 
 
 Masts and Derricks. — Masts are nowadays most frequently made of 
 steel. In sailing vessels, where they are very long and of large diameter, 
 there may be three plates in the round, with angle or tee bar stiffeners 
 throughout their length. 
 
 In steamers, however, where the masts are usually short, carrying very 
 little sail, two plates in the round are usually adopted, with or without 
 stiffeners as the case may require. All masts should be doubled in way of 
 the wedging at the upper deck for a length of about 4 ft. above and below 
 the deck. The butts of the plates should be well clear of each other, and 
 connected by either straps or overlaps. The edge riveting is usually single 
 in steamers, while the butts above the upper deck should be treble riveted, 
 and below, double riveted. Masts are secured, at the deck and heel by 
 angle collars which may be riveted or wedged to the mast, and also by 
 means of a good disposition of shrouds well connected to the sheer strake. 
 See figs. 47 and 49. 
 
 In small vessels, or in vessels of moderate beam, the derricks for the 
 loading and discharging of cargo are usually pivoted to the mast, as shown 
 in fig. 66. Derricks always swing most easily when the masts are perfectly 
 vertical, having therefore no rake. In large vessels of great beam, the length 
 of derrick, to enable cargo to be swung clear of the vessel's side, would be so 
 extraordinary that some means have become necessary to curtail their length 
 within reasonable dimensions. This can he done by fitting special derrick 
 posts to the deck, or hy nrakiug the main ventilators of extra strength 
 and height to serve as supports upon which the derricks may swing. 
 (Fig. 165.) 
 
 But by far the best system of arranging the derricks is as shown in figs. 
 166 and 167. 
 
 In this arrangement, special tables are fitted to the mast at a height of 
 about 7 ft. from the deck, and supported by stanchions and brackets as 
 shown. The length of these tables transversely from the mast will depend 
 upon the beam of the ship and the minimum length of derrick required. 
 The advantage of relieving the derrick tables of as much thrust from the 
 derricks as possible, by fitting a post from the deck to the derrick heel 
 as showDj is obvious. Fig. 166 shows an arrangement of mast tops so 
 
DETAILS OF CONSTRUCTION. 
 
 249 
 
 designed that the pivoting point of the derrick topping lift is vertically 
 above the derrick heel, thereby greatly facilitating the easy swinging of 
 
 _^ 
 
 •^ 
 
 ;§^ Position 
 1 or Deck 
 
 ^ 
 
 Mast Heel 
 
 O 
 
 Section 
 
250 STEEL SHIPS. 
 
 the derrick. The derricks as illustrated iu figs. 166 and 167 are such as 
 are fitted to the steamer in fig. 72, which is designed for the rapid loading 
 and discharging of cargo. Three derricks, it will be seen, are fitted to each 
 hold. Fig. 166 shows the middle derrick resting upon a small table just 
 above the larger tables, while fig. 167b shows a method of securing these 
 derricks in a vertical position to the masts when the ship is at sea. The 
 other derricks are laid in a fore and aft direction, and may rest either upon 
 a poop, bridge, or forecastle end, or upon special stanchions. 
 
 The function of the middle derrick is to lift heavy cargo out of the hold, 
 while that of the two side derricks is for lighter cargo only. 
 
 Fig. 163 illustrates a telescopic mast and fig. 164 a patent hinged 
 topmast (Sidgwick's). Arrangements for lowering the topmasts are 
 necessary in vessels which are required to pass under bridges, and which 
 use such waterways as the Manchester Canal. 
 
 Panting. — The principal methods adopted to resist panting may be 
 enumerated as follows : — 
 
 1. A closer spacing of frames. 
 
 2. Double frames. 
 
 3. An extra tier or tiers of beams with stringer plates on their ends well 
 connected to the shell, or additional double angle, tee bar, or intercostal 
 panting stringer may be introduced, 
 
 4. Floors of extra depth. 
 
 5. An increase in the thickness of shell plating, especially under the fore 
 foot, where excessive thumping is experienced. 
 
 6. The middle-line keelson should be fitted intercostally as far forward 
 as practicable. 
 
 Hatches. — The construction of the sides and ends of hatches has already 
 been fully dealt with and illustrated in our remarks upon 'Beams,' p. 194. 
 There is no doubt that many a ship has foundered through no other cause 
 than the inefficient construction of hatches which have been unable to bear 
 the weight of huge volumes of water which so often fall upon them when 
 heavy seas are shipped. Not only should the coamings be of ample 
 strength, but the hatch covers should be at least 3 in. in thickness, supported 
 upon strong fore and aft, and, where necessary, transverse bearers. In 
 the case of short hatches, one fore and aft middle line, and, if necessary, 
 two side bearers, or ' fore and afters,' as they are technically called, should 
 be fitted. These may be of wood, 6 in. or 8 in. square, fitted into shoes 
 on the hatch end coamings, or they may consist of bulb angles, tee bulbs, 
 etc. Where the hatches are of greater length, transverse assistance is 
 given to the deck, and to prevent the hatch sides collapsing, by fitting 
 transverse web plates, which should be stiff"ened on their upper and lower 
 edges by angles or half-round iron, and fitted into slides, several methods 
 of which are illustrated in fig. 115. Tarpaulins fastened to the hatch sides 
 by, cleats, battens, and wedges, are fitted to all vessels. 
 
 Deck Houses. — Deck houses for accommodation, etc., are frequently built 
 
DETAILS OF CONSTRUCTION. 
 
 251 
 
 entirely of steel with steel frames and beams. This certainly makes the 
 strongest house. Where, however, they are built of wood, it is advisable 
 
 Jackstay \ 
 Iron 
 
 Wood 
 Topmast 
 
 
 // 
 
 
 /^^ 
 
 
 
 
 <i? 
 
 Steel 
 LowerMast 
 
 Fore and 
 Aft view 
 
 Athwart ship 
 view 
 
 Fig. 163. — Telescopic Mast. Showing arrangement for lowering Top Mast. 
 
252 
 
 STEEL SHIPS. 
 
 Fig. 164.— Hinged Top Mast. (Sidgwick's patent.) 
 
DETAILS OF CONSTRUCTION. 
 
 25, 
 
 that they be framed with steel or iroa bars, and have steel coamings well 
 attached to a deck tie plate or to the steel deck if such is fitted. This 
 
 Seco nd Deck 
 
 Fig. 165. — Derrick upon Ventilator. 
 
 affords a substantial framework upon which to lay the wood construction, 
 and is illustrated in fig. 169. 
 
 Pooi^ and. Bridge Front Bulkheads. — All steel poop and bridge fronts 
 
254 
 
 STEEL SHIPS. 
 
 Plan 
 
 Elevation 
 
 Fig. 168. —Cast Iron Socket on Deck for Strong Derrick. 
 
 c 
 
 TT 
 
 "I I I I I , :^ 
 
 II n 
 
 _^,- 1 
 
 Ikngle Beams 
 
 V 
 
 ■Wood Clead'inj 
 
 Angle Iron 
 frames 
 
 I M I I 
 
 J 
 
 Fig. 169. — Composite Deck House. 
 
Fig. 166. 
 DERRICK TABLES ON MASTS. 
 
 Plan of Derrick Table 
 
Fig. 167. 
 OUTREACHES ON MASTS FOR DERRICKS. 
 
 Plan of Outreaches 
 for Derricks 
 
 Showing Centre Derricks hoisted up 
 and secured to mast 
 
DETAILS OF CONSTRUCTION. 255 
 
 should Lave thick coaming plates well connected to the deck. The upper 
 plating may be somewhat thinner. The stiffening should be made by 
 angle bars of at least the size of the main frames, spaced not more than 
 30 ill. apart, with brackets top and bottom. Fig. 56 and remarks, pp. 129 
 and 130, further describe and illustrate the construction of these parts. 
 
 Tunnel and Casings. — The tunnel, by means of which the propeller 
 shaft is encased, should be thoroughly watertight fi'om the engine room 
 bulkhead to the aftermost bulkhead with stuffing boxes upon each. As 
 it is advisable to have the engine-thrust in free and easy communication 
 with the engine room, the after engine room bulkhead is usually recessed 
 for this purpose. The tunnel should be stiffened to resist the weight of 
 the cargo which may rest upon it, by means of angles bent transversely 
 round it on the inside. These should be spaced not more than 3 ft. apart. 
 As far as possible, hold pillars should be kept off the tunnel, by arranging 
 them zigzag one on each side. Where this cannot be avoided, double 
 angle stiffeners must be fitted under the pillars. The tunnel under the 
 hatchways should be sheathed with wood, or the plating increased in 
 thickness, to protect it from blows from cargo in the operation of loading 
 or unloading. It should be large enough inside to permit of easy access to 
 all the shaft bearings. When the tunnel stands upon a centre keelson, the 
 tunnel floor is usually laid on the top of the centre keelson, thus permitting 
 more easily of being made watertight. "(See fig. 79.) 
 
 Casings enclosing engine and boiler openings should be stiffened by 
 angles or half-round iron on the inside, and if exposed to the weather, the 
 strength must be sufficient to afford ample protection to the engines and 
 
 boilers. 
 
 Breast Hooks. — All stringers, wherever 
 
 practicable, should be carried all fore and 
 
 aft, and united at the ends of the vessel <^7'/'//7yf/^^>^. %v Brp^'ithook 
 
 by means of hreast hooJis thus. In large P^^^^ 
 
 vessels, an extra breast hook should be 
 
 fitted at the stem between each tier of 
 
 beams for the further support of the 
 
 shell plating. 
 
 Insulation. — Fig. 170 illustrates a system of insulation adopted in the 
 
 holds and 'tween decks of steamers engaged in carrying dead meat, etc. 
 
 Bilge Keels. — A bilge keel is a projecting fin of some kind usually 
 
 attached to the outside shell plating on the turn of the bilge, the resistance 
 
 of which, as it oscillates with the rolling movements of the vessel, conduces 
 
 to the desirable quality of steadiness or easiness in the vessel's motions, and 
 
 reduction of the angle of inclination. 
 
 It may be constructed in several ways, but in mercantile steamers it 
 
 usually consists of a bulb plate, 8, 10, or 12 in. in width according to the 
 
 size of the vessel, riveted between double angles. It may be either riveted 
 
 through the shell or connected by means of tap rivets. Another method is 
 
256 
 
 STEEL SHIPS. 
 
 to connect the bulb plate to the shell by means of a single angle, or by 
 means of a tee bar. But in the event of grounding upon the bilge keel or 
 coming into contact with any kind of obstruction, the first method generally 
 causes least injury to the ship, as the bulb plate simply buckles or bends 
 over the two angles, and no injury is caused to the rivets through the shell. 
 In either of the other methods, there is the danger of the tee bar, or angle, 
 being wrenched at the root, and the caulking or riveting damaged and 
 leakage ensuing. (See figs. 48, 64, and 71.) 
 
 Bilge keels should be placed so as to give the least possible resistance to 
 propulsion, and kept as clear as possible from shell landings, tank side 
 riveting, etc. 
 
 Charcoal or 
 Vegetable Silica 
 
 Jwo% Boards with I 
 
 brown paper between) 
 
 Charcoal or Vegetable Silica 
 
 Z Deck 
 
 Fig. 170.— Insulation. 
 
 Ventilation. 
 
 In every well-constructed vessel a great amount of thought is given 
 to the subject of efficient ventilation, and though the arrangement for 
 ventilation may be more complicated in a passenger vessel, yet it is of 
 equal importance in a purely cargo one. 
 
 In every ship where a thoroughly efficient system of ventilation has 
 been installed, there will be provided a free circulation of air in every 
 compartment, whether it be saloon, cabin, or lavatory, occupied by 
 passengers; in every crew space; in every hold space in which cargo 
 of any kind has to be carried ; in engine room, boiler room, and shaft 
 tunnel; in bunkers, in every compartment used solely or temporarily 
 for the carriage of water-ballast ; in peak spaces ; in store rooms, galleys, 
 
DETAILS OF CONSTRUCTION. 257 
 
 pautries, etc.; in short, every compartment in the vessel should be 
 ventilated. 
 
 This is necessary, not only from a sanitary point of view, for the 
 health and comfort of passengers, but because it is an absolute necessity 
 for the engineers and stokers shut up in the bowels of the vessel, for the 
 preservation of perishable cargoes, in order to rid hold and other spaces 
 of obnoxious or poisonous and inflammable gases, and because good 
 ventilation is as efficacious in the preservation of the material of which 
 the vessel is constructed (iron, steel, wood, etc.), as the best patent 
 composition ever put upon a ship. 
 
 Ventilation does not solely consist in making a certain number of 
 inlets to, or outlets from, any particular space, as one would sometimes 
 imagine by the way in which it is carried out, but it is the intelligent 
 arrangement of such inlets and outlets by which they are so situated 
 that fresh air is introduced and foul air is expelled. This is usually effected 
 in ships by means of natural ventilation, though, in some cases, forced 
 ventilation, by means of fan draughts, or even steam injections, is necessary, 
 in order to rid certain spaces of the foul air which gathers in them. 
 
 Innumerable systems and patent arrangements for ventilation have been 
 introduced and are in use, but in a work of this kind it is impossible to 
 •examine all these various methods, and we must therefore confine ourselves 
 to the older method of natural ventilation, which may be very efficacious if 
 well arranged. 
 
 The best known ventilator is the cowl-head (see fig. 171). It is 
 essential that all ventilators situated upon the weather deck be sufficiently 
 strong to endure without damage the force of heavy seas shipped on deck. 
 In bad weather it is not very uncommon for deck ventilators to be carried 
 away, and the cargo to suffer considerable damage, or discomfort to be 
 brought upon passengers or crew, not to mention the possibility even of 
 positive danger accruing from the continued ingress of water through such 
 openings. Thus, in erecting cowl ventilators upon the weather decks, the 
 coamings or lower plating should consist of thick plates (at least ^ in.) 
 connected to the deck by a correspondingly strong angle bar. These coam- 
 ings should be at least 30 in. high. The upper part of the ventilator, 
 which includes the cowl-head, is portable, and is usually made of thinner 
 plating than the coamings, as in very bad weather, when much water is being 
 shipped on deck, it is usual to unship this upper part, and fit into the top 
 of the coamings the plate lid, which is shown in fig. 171, A, in a vertical 
 position, where it is kept when not in use. To ensure that the ventilator 
 be now thoroughly watertight, a canvas cover is lashed over the top of 
 the coamings. 
 
 "When these ventilators are intended to ventilate the hold space of a 
 single deck vessel, the diameter ought to be regulated in accordance with 
 the capacity of the space to be ventilated. 
 
 The coamings should be made of welded steel or wrought-iron plates. 
 
 17 
 
258 
 
 STEEL SHIPS. 
 
 The diameter of the cowl mouth should be large enough and so shaped as to 
 take in as great a volume of air as possible (say, two or two and a half times 
 the diameter of the coamings). The portable upper part of the ventilator, 
 when fixed on to the coamings, rests upon an iron ledge upon the outside of 
 the coamings, as shown in fig. 171, A; or it may come right down and rest 
 upon the vertical flange of the deck angle. There should be at least two 
 cowl ventilators to each hold, one at each end of the space, and in such 
 positions that they may efficiently act, one for the iulet of fresh air, 
 and the other for the outlet of the hold space air, which is expelled. "When 
 these ventilators are situated near to, or against, forecastle fronts, bridge 
 
 DecK I 
 
 ^ 
 
 Cowl-head 
 yenfilafor 
 
 Ci Mushroom j Venti/afor- 
 
 DecK] 
 
 t If 
 
 D SwannecK 
 Venfilafor 
 
 DecK 
 
 Fig. 171.— Ventilators. 
 
 ends, poop fronts, or any other deck erection, they should extend to a height 
 such as will bring the cowl mouth above the top of these erections. 
 Their extreme length necessitates that they be supported by means of stays, 
 either to the deck or to the erections near which they are situated. "When 
 a vessel has two or more laid decks, it is usual to ventilate the hold and 
 main deck spaces in a manner similar to that shown in fig. 171, B. The 
 diameter of the ventilator coamings upon the uppermost deck is made 
 sufficiently large to ventilate the whole of the space below the upper deck, 
 supposing no 'tween decks existed. But as 'tween decks do exist, the area 
 
DETAILS OF CONSTRUCTION. 259 
 
 of the ventilating opening on the upper deck is reduced by the insertion of 
 a tube leading to the second 'tween decks, the diameter of which is 
 supposed to be large enough to ventilate the lower 'tween decks and hold 
 space. But as the hold space is an entirely separate compartment, the area 
 of tlie ventilating opening in the upper 'tween decks is again reduced by 
 the insertion of a ventilating tube from the hold space, the diameter of 
 which is in proportion to the volume of such space. It will be evident 
 that while one ventilator at the end of each hold space may be sufficient in 
 a small single deck vessel, a greater number of ventilators will be required 
 in large vessels with one or more 'tween decks and huge hold spaces. (For 
 the ventilation of steamers carrying oil in bulk, see Chapter VI., Section 2.) 
 
 While it may be advantageous to ventilate both forecastles occupied by 
 crew, and cabin accommodation in other parts of the vessel by means of cowl- 
 head ventilators where the spaces are of considerable size, sufficient ventila- 
 tion can be obtained in small cabin spaces, galleys, pantries, lockers, store- 
 rooms, lavatories, etc., by means of swan neck and mushroom top ventilators 
 (see figs. 171, D and C), assisted by louvre openings in the doors. Most 
 of these spaces have the additional advantage of hinged scuttles, which in 
 fine weather may provide an abundant flow of fresh air, and in bad weather 
 may be secured and made thoroughly safe by their solid hinged plate 
 dead-lights. 
 
 One of the parts of the vessel which needs particular attention in 
 designing the ventilating arrangements is the engine room. This space 
 is often sadly neglected, as those who have had any sea experience of 
 these vessels know. The recesses in these spaces caused by pocket 
 bunkers, engine room store, etc., and the fact that the opening through 
 the decks is often very narrow, with consequently large flat areas of 
 deck overhead, tends to the harbour of a considerable amount of 
 stagnant air, owing to the ventilators, as a rale, being inserted through 
 the top of the casings, and failing to rid the space of the obnoxious air 
 in the engine room wings. In addition, engine rooms often get unbearably 
 heated, owing to the proximity of the boilers. The boiler room also needs 
 ventilating, and this is largely effected by the iron gratings on the top of 
 the casings, and cowl-head ventilators of very large diameter. One or 
 more cowl-head ventilators ought to be fitted into the tunnel, one as far aft 
 as possible. The other spaces previously mentioned may be ventilated by 
 one such ventilator as illustrated in fig. 171, C and D. 
 
 ^ Pumping. 
 
 Nothing in the equipment of a vessel is of greater importance than 
 a well-planned and thoroughly efficient arrangement of pumping. This 
 is essential for many reasons, all of which ought to be thoroughly 
 considered and thought out in the design of such an arrangement. 
 
 In any vessel there is always a possibility of water finding its way 
 
2 6o STEEL SHIPS. 
 
 into the interior of the hull below the weather deck. This may be due 
 to damage from collision, leakage, damage to deck fittings in heavy 
 weather, drainage from wet cargo, heavy sweating in the inside of the 
 shell plating, decks, etc., especially where ventilation is inefficient, 
 drainage from scuppers, etc., etc. 
 
 The pumping installation should be capable of speedily ejecting all 
 such water, excepting in cases where collision has made an opening so 
 large that for any pump to cope with the inflow is impossible. 
 
 In sailing vessels where there is no steam power on board, all such 
 pumping has to be effected by manual labour. Where the vessel has 
 much rise of floor, so that even under considerable list to port or 
 starboard, the water still gravitates to the middle line, there should be 
 at least one hand pump to each hold compartment.* 
 
 In order to allow water to drain from one hold compartment to 
 another, it is permissible to fit sluice valves to each bulkhead (when 
 more than one bulkhead is fitted). On no account should a sluice valve 
 be fitted to the collision bulkhead. 
 
 No sluice valve should be fitted to any watertight bulkhead unless 
 it can be reached at all times, on account of the possibilitjr of chips or 
 rubbish getting jammed into the opening and preventing the valve from 
 closing. 
 
 To reach the sluice valves upon watertight bulkheads, either a wooden 
 ■or plated trunk-way must be fitted from the deck above, sufficiently 
 dust-tight to prevent dirt, grain, etc., finding their way from the hold to 
 the valves. All rods working the sluice valves should extend to the 
 upper deck, and be clearly marked with a brass plate cover, so as to be 
 easily found, and fitted with an indicator to show when shut and open. 
 
 To ascertain the depth of water in any compartment in any possible 
 condition, a sufficient number of sounding pipes should be fitted. The 
 depth of such water is ascertained by means of a sounding rod lowered 
 inside the pipe. As the frequent thumping of such rods would be likely 
 to eventually perforate the shell plating, a small doubling plate should be 
 fitted under each sounding pipe. 
 
 If a tube were inserted iti a basin of water, and the air sucked out of the 
 tube, the water inside the tube would rise to a considerable height above 
 the level water in the basin. This is due to the atmospheric pressure upon 
 the surface of the water in the basin, and the fact that this pressure has 
 been wholly or practically taken from the water at the bottom of the tube. 
 "Were the tube long enough, and the vacuum so produced absolutely perfect, 
 
 * It may be pointed out that in sailing ships only one bulkhead is usually 
 fitted, viz., the collision bulkhead, and this therefore greatly simplifies the pumping 
 arrangement. As a rule, one pair of pumps to this main hold space is considered 
 sufficient. 
 
 These suctions are led from the pump well, which is usually situated just abaft of 
 the main mast. Access is obtained to this well, whenever necessary, by a wooden 
 trunk-way fitted from the weather deck to the bottom of the ship. 
 
Fig. 171a. 
 ship's hand-pump. 
 
 By removing the plunger 
 fhe lower valve can be 
 lifted out by means of a 
 book for purposes of cleaning 
 
 ^ 
 
 Brass Valve 
 Weighted leatfier clack 
 
 Cast iron with brass 
 bush in way of the stroke 
 
 Brass washers to take heads of screws 
 
 Weighted leather Clack 
 
 Brass valve 
 
 Enlarged view of brass 
 
 valve showing flannel 
 
 attached by wire bands 
 
 lead 
 'Strum box at bottom 
 
DETAILS OF CONSTRUCTION. 
 
 261 
 
 the water in the tube would rise to a height of about 34 ft., the weight of 
 such a column of water being equal to the atmospheric pressure upon every 
 unit of area on the surface of the water in the basin, equal to the sectional 
 area of the tube. 
 
 The principle of the manipulation of the hand pump is identical with 
 the foregoing illurstration. First of all, there is a long suction tube from 
 the bottom of the vessel up to the pump chamber. The action of the 
 bucket, which is worked by a handle on the deck, is to first pump 
 the air out of the pipe, immediately upon which the water ascends. 
 Were a perfect vacuum produced, it would rise, as previously stated, to 
 a height of about 34 ft., but as it is practically impossible in such 
 
 Fig. 172. — Sluice Valve. 
 
 pumps to get an absolute vacuum, it is safer never to count upon the water 
 rising above 24 ft. As hand pumps are usually worked from the upper 
 deck, the suction chamber should be deep enough to ensure that the bucket 
 is not more than 24 ft. above the bottom of the hold. 
 
 A rose, strum box, or suitable mouthpiece is usually fitted to the lowest 
 extremity of the suction pipe, to prevent any solid matter being drawn into 
 the pump. As a considerable thickness of cement is commonly laid on the 
 inside of the bottom shell plating for purposes of preservation, the cement 
 ought to be dished out in way of the strum boxes so as to bring the bottom 
 
262 
 
 STEEL SHIPS. 
 
 of the suction pipe below the level of the cement, and thus as nearly as 
 possible eject all water from the compartments (see fig. 173). The necessity 
 for an ample number of drain holes through the floor plates (or limber holes), 
 to allow the water to reach the pumps, will now be apparent. 
 
 A pump is now in existence (Downton's) by means of which several 
 suctions may be led into one suction chamber, and water pumped from any 
 compartment. In sailing vessels having a donkey boiler, these pumps 
 should be arranged so that they may be worked, if necessary, from a steam 
 winch. 
 
 Steavi Vessels 'without Double Bottoms. 
 
 Screvj and paddle vessels have the great advantage of possessing steam 
 power, which is utilized in pumping from the various compartments. 
 Steam vessels without double bottoms should have at least one hand pump 
 and one steam pump fitted in each hold when the rise of floor is considerable. 
 
 Lead 
 
 O O 
 
 o o 
 
 o o 
 o 
 
 s/ 
 
 .Lead Suction Pipe 
 
 m 
 
 o o 
 o o 
 P o 
 o 
 
 -Cast Iron 
 
 Rose or Strum 
 Box. 
 
 About % clearance giving 
 an area equal at least to 
 the area of the suction pipe 
 
 FiG.j|l73. — Mouthpiece for Suction Pipe with three legs resting upon the cement, which is dished out as shown. 
 
 "When there is very little rise of floor, three steam-pump suctions — one 
 as near the centre line as possible, and one in each wing — and at least one 
 hand pump, should be fitted into each hold compartment. If the vessel has 
 fore and after peak water-ballast tanks, a separate steam-pump suction 
 should be led to each. When water is not carried as ballast in these 
 compartments, a hand pump only is required in the fore peak. 
 
 JSTo sluice valve must be fitted to the collision bulkhead under any 
 circumstances ; indeed, no sluice valve must be fitted to any bulkhead which 
 bounds a water-ballast tank. If the after peak is not intended for the 
 carriage of water-ballast, a sluice valve may be fitted in order to allow any 
 water to reach the pumps in the after hold. The engine and boiler space 
 should be fitted with at least three steam-pump suctions, one at the centre, 
 and one in each bilge, and the Board of Trade requires one hand pump. 
 
Fig. 174. 
 PUMPING ARRANGEMENT. 
 
t 
 
DETAILS OF CONSTEUCTION. 263 
 
 Steam Vessels with Double Bottoms. (See fig. 174.) 
 
 Before dealing with the pumps, the nature of all the spaces or compart- 
 ments which exist along the bottoms of vessels with double bottoms for the 
 carriage of water-ballast ought to be thoroughly understood. 
 
 The double bottom should at least be divided into watertight compart- 
 ments agreeing with the number of the watertight bulkheads. But this 
 division is often exceeded in practice. (See figs. 49, 66, and 174.) 
 
 Hence it is necessary that steam pumps should draw from each of these 
 compartments. But at the wing extremities on the outside of double 
 bottom tanks, there are the bilges into which drainage from cargo, scuppers, 
 sweating, etc, may find its way, and pumps must be provided to get rid of 
 such water. 
 
 It is usual and advisable to arrange for about two frames spaces at or 
 near the after end of the engine and boiler space to act as a well into which 
 bilge Avater may drain. A cock or sluice valve may be fitted through the 
 bulkheads in way of the bilges at each end of this space so as to allow 
 bilge water in the adjacent holds also to drain into the engine-room 
 well. As before stated, such sluice valves or cocks should be always 
 accessible. 
 
 The well is in all respects constructed like the double bottom elsewhere, 
 excepting that the floor at each end of it is watertight, and several holes, 
 3 in. or more in diameter, are punched through the tank margin plate 
 through which the bilge water flows into the well, or in lieu of this a better 
 arrangement is to fit a non-return valve on the margin plate. When such 
 a well exists, one steam-pump suction should be led as near the middle line 
 as possible, and one in each bilge. 
 
 The main and donkey pumps should be arranged to draw from the fore 
 peak, and the after peak also if used for water-ballast, from double-bottom 
 tanks, wells, bilges, and from hold spaces in vessels without double 
 bottoms. The donkey pump should have a separate bilge suction in the 
 engine room. The steam suction pipes from the various compartments, on 
 reaching the engine room, are led into combined valve boxes or chests, which 
 are in turn connected to the pumps. Fig. 175 shows such a valve chest, 
 and the donkey pump, which is of double acting type. As shown in the 
 diagram, when it is desired to draw from any particular compartment, the 
 valves to suctions for other compartments can be closed by the valves 
 shown. 
 
 A well, similar to that just described for the engine room, is usually 
 arranged at the after end of the after-hold space into Avhich bilge drainage 
 finds its way. A steam pump suction is therefore necessary for this space 
 also. 
 
 Adjacent to this well, another well is usually constructed exclusively for 
 drainage from the tunnel, which in its turn also requires a separate suction, 
 
264 
 
 STEEL SHIPS. 
 
 CK^^^..^ 
 
 which may be connected 
 up to the hold well suction 
 by a valve box somewhat 
 similar to that illustrated 
 in fig. 175, and which is 
 manipulated from the in- 
 side of the tunnel. 
 
 A sounding pipe should 
 be fitted on each side of 
 all ballast tanks with the 
 necessary doubling plate 
 under each, as previously 
 pointed out. 
 
 To allow for the exhaust 
 of air in all water-ballast 
 tanks when being filled 
 up, a sufiicient number 
 of air pipes should be 
 led to the upper deck, 
 which, if made of 5 in. 
 or 6 in. diameter, instead 
 of 2^ in. or 3 in. as usually 
 fitted, may conveniently, 
 and with great advantage, 
 add to the preservation of 
 the structure, if made to 
 act as ventilators. 
 
 Sufiicient limber holes 
 should be punched through 
 all floor plates and tank 
 knees, increasing in 
 number towards the suc- 
 tions, to allow for an 
 ample flow of water to 
 reach the j^umps. 
 
 All tank floor plates 
 should be perforated with 
 circular holes near the 
 upper extremity, to act 
 as air passages, and for 
 a similar reason the pack- 
 ing pieces under outside 
 strakes of tank top plat- 
 ing should be kept a 
 little short. 
 
DETAILS OF CONSTRUCTION. 
 
 265 
 
 Hand Pumps, Steam Suctions, etc. 
 
 The following tables taken from Lloyd's Rules will give some idea of the 
 minimum sizes required for pipes to hand pumps and steam suctions. 
 
 Tonnage under Upper Deck. 
 
 Hand Pump in Holds. 
 
 Diameter of 
 Barrel. 
 
 Diameter of Tail Pipe. 
 
 In vessels under 500 tons, 
 
 ,, ,, of 500 but under 1000 tons, 
 
 ,, ,, 1000 „ „ 2000 „ 
 ,, ,, ,, 2000 and above, . 
 
 in. 
 
 4 
 
 4i 
 
 5" 
 
 5^ 
 
 in. 
 
 2 
 
 2i 
 
 2i 
 
 2| 
 
 Tonnage under Upper Deck. 
 
 Engine Room 
 Centre Suction, 
 
 Separate 
 Donkey Suction, 
 
 and Hold 
 Centre Suctions. 
 
 Wing Suctions 
 
 in Holds 
 where no Centre 
 
 Suctions are 
 fitted, and Wing 
 
 Suctions in 
 E. and B. Room. 
 
 Wing Suctions 
 in Holds 
 
 where Centre 
 Suctions 
 are fitted. 
 
 In vessels under 500 tons, 
 ,, ,, 500 tons and under 1000 tons, 
 „ ,, 1000 ,, ,, „ 1500 „ 
 „ ,, 1500 ,, ,, ,, 2000 ,, 
 „ „ 2000 ,, „ ,, 3000 ,, 
 ,, ,, 3000 ,, and above, . 
 
 in. 
 
 2 
 
 2i 
 
 2i 
 
 3 
 
 3i 
 
 3i 
 
 in. 
 
 2 
 
 2 
 
 n 
 
 2| 
 
 3 
 
 3i 
 
 in. 
 
 2 
 
 2 
 
 2 
 
 2i 
 
 n 
 
 Launching. 
 
 One of the most important events in the history of a ship is the 
 launch, — important, not only because this is her introduction to the 
 element upon which the work of her life will be performed, but because, 
 in the transit from land to water, unless the greatest care and caution are 
 exercised, results of a most disastrous nature may accrue. Instances are on 
 record when the damage received in launching has distinctly contributed 
 to the loss of the vessel at sea, and many a ship has been so injured as to 
 cost her builders thousands of pounds to repair. 
 
 The strength of the launching cradle and standing ways, the earth 
 foundation as well as the declivity, length and camber of the standing ways, 
 are all features in the carrying out of a launch which contribute to the 
 success of this operation. 
 
 In shipyard practice the laying of the ways Avas, and is still, in 
 some of the smaller establishments, left in the hands of the foreman 
 shipwright, whose experience guides him in fixing the declivity, etc.. 
 
266 
 
 STEEL SHIPS. 
 
 CQ 
 
 >> 
 
 Cr, 
 
 
 c: 
 
 
 "5 
 
 
 c; 
 
 
 ,'=0 
 
 
 v; 
 
 
 CO 
 
 
 <»-, 
 
 
 C5 
 
 
 ta 
 
 
 Co 
 
 
 « 
 
 ■^ 
 
 Co 
 
 ■t3 
 
 ■^ 
 
 
 <1 
 
 
 5 
 <5 
 
 c2 
 
 CI, 
 
 Co 
 
 
 cS 
 
 ^ 
 
 
 
 tr, 
 
 
 ce 
 
 
 
 
 s 
 
 
 'cS 
 
 :3 
 
 
 ^ 
 
 CX3 
 
 
 
 ^J 
 
 5~ 
 
 CO 
 
 
 s 
 
 M 
 
 
 .9 
 
 ■C: 
 
 .^ 
 
 Q 
 
 Co 
 
 -~j 
 
 1o 
 
 <0 
 
 OQ 
 
 s 
 
 Q 
 
 -t3 
 
 S) 
 
 "==1: 
 
 ^ 
 
 s 
 
 necessary for any particular vessel. But 
 with the vast increase in the sizes of vessels 
 in recent years, the launching particulars are 
 fixed in the ship drawing office, where the 
 whole matter is considered and investigated, 
 and the calculations involved carried out. 
 
 One of the greatest dangers in launching 
 a ship is, that in the event of there being 
 too little water on the ends of the ways, 
 she may tilt after her centre of gravity has 
 passed the ends of the standing or fixed 
 launching ways. The fore end of the vessel 
 would then be lifted entirely clear of the 
 sliding ways, and the whole weight of the 
 vessel (less the buoyancy of the water-borne 
 portion of the ship) would then be resting 
 upon the extreme end of the fixed ways, and 
 these, being made of wood, might collapse 
 under the pressure. Moreover, there would 
 be nothing to prevent the vessel slewing 
 across the launching ways by any slight 
 wind or tide, and even leaving the launching 
 ways altogether, bringing about a catas- 
 trophe of the most serious nature. (See 
 also Chapter VIL, p. 177.) The enormous 
 pressure experienced by the bottom, when 
 a vessel tilts, is sometimes sufficient to crush 
 in the plating and do other damage. The 
 ways may be flat or hollow (see fig. 96), and 
 the breadth of the ways for a merchant 
 vessel about 300 ft. long would generally be 
 from about 15 to 18 in. 
 
 The usual declivity of the keel is about 
 ^ in. or y^g- in. per foot of length, and is a 
 perfectly straight line. The declivity of the 
 standing ways may be either perfectly straight, 
 or cambered in the direction of the length. 
 The latter method is commonly adopted. 
 
 Fig. 176 shows what is meant by 'camber.' 
 Let A B be a straight horizontal line, repre- 
 senting the length of the launching ways. 
 Upon it construct an arc of a circle, A d B, 
 having a round or camber of, say, 18 in. 
 Now lift AB up to AC, so that it has 
 the declivity desired for the launching 
 
DETAILS OF CONSTRUCTION. 267 
 
 ways. ADC will be the form of the surface of the standing ways. The 
 camber varies from 9 in. to 12 in., though for special types of vessels it may 
 be lower or higher than these limits. 
 
 It will now be seen that a tangent upon the upper half of the arc of the 
 launching ways must be at considerably less declivity than the tangent upon 
 the lower half of the ways. Hence, in providing particulars of the launch- 
 ing ways, two declivities are usually given when the ways are cambered. 
 The usual declivity of the standing ways runs about h in. per foot on the 
 fore part of the ways, to f in. or more towards the ends. 
 
 The mean pressure per foot upon the sliding ways ranges up to three tons, 
 seldom over, the average being two tons. 
 
 Experience is invaluable in shipyard practice, but the very best is liable 
 to fail under exceptional circumstances, hence launches are not always as 
 successful as they might have been had the vessel's condition been tested 
 by calculation before the operation of launching was performed. Everyone 
 who has seen a ship launched has observed that at the moment she begins 
 to move after the dog chocks have been knocked out at the fore end, her 
 weight over the whole of the length bears on the launching ways. But as 
 she travels, her after end passes over the end of the ways, enters the water, 
 and eventually becomes afloat, while the fore end bears upon the fore end of 
 the launching cradle. (See fig. 177.) 
 
 The point at which the after end of the vessel becomes water-borne (the 
 transition stage between 177, A, and 177, C, namely, 177, B) will occur when 
 the moment of the ship's weight about the fore end of the launching cradle 
 (the sliding ways), X in fig. 177, is exactly equalled by the moment of the 
 buoyancy of the immersed volume of the vessel about the same point X. 
 
 The moment of weight is obtained by multiplying the vessel's weight 
 in tons by the distance of her centre of gravity from the fore end of the 
 cradle (X, in fig. 177), and the moment of buoyancy is found by multiplying 
 the buoyancy in tons of the immersed volume by the distance of its centre 
 of buoyancy from the fore end of the cradle also. 
 
 Supposing a ship to be 300 ft. long, her w^eight on the ways to be 
 1500 tons, and the distance from her centre of gravity to the fore end of 
 the launching cradle to be 130 ft., then the moment of weight would be 
 1500x130=195,000 ft. tons. When this amount of buoyancy moment 
 exists, it is known that the after end of the vessel is afloat, making 
 allowance for the fact that slightly more immersion is necessary iu order 
 to overcome the ship's momentum, which causes her to travel a little 
 further down the ways before her stern lifts. 
 
 When the aft end of the vessel enters the water, her weight will be 
 partly borne by the water, and partly by the launching ways. The 
 amount of support afforded by the water is the displacement, and the 
 difi'erence between the displacement and the total weight of the ship 
 represents the weight borne by the launching ways. Through not realizing 
 the amount of this pressure upon the fore end of the sliding ways, some- 
 
268 
 
 STEEL SHIPS. 
 
 to 
 
 K 
 
 5^ 
 
 ^ 
 
o 
 
 < 
 
 
 
 UJ 
 
 
 
 O 
 
 
 
 O 
 
 
 ~Ov 
 
 cc 
 
 
 _•" 
 
 u 
 
 
 <-. 
 
 s 
 
 
 j:: 
 
 < 
 
 UJ 
 
 
 P 
 
 1- 
 
 
 ^• 
 
 (D 
 
 
 .'•' 
 
 cc 
 
 
 '^^ 
 
 UJ 
 
 
 ■^ 
 
 o 
 
 
 
 z 
 
 ■ 
 
 ^ 
 
 UJ 
 
 f/) 
 
 'J 
 
 rt 
 
 (f) 
 
 
 
 < 
 
 •< 
 
 *, 
 
 Q. 
 
 (- 
 
 ^ 
 
 III 
 
 ^. 
 
 ^ 
 
 CC 
 
 
 W 
 
 o 
 
 
 bo 
 
 (f) 
 
 
 u 
 
 < f. 
 
 < 
 
 >■ 
 o 
 
 QC 
 
DETAILS OF CONSTRUCTION. 269 
 
 times the cmdle has collapsed through weakness, and the vessel has come 
 heavily down upon her stem, thereby producing considerable damage. 
 Because of this fact, more declivity is given to the after end of the standing 
 ways than the fore end by cambering, the endeavour being to get the vessel 
 over this critical stage as soon as possible, remembering at the same time 
 that the resistance offered by the water increases as the immersion increases. 
 
 In calculating the condition of a ship with regard to tipping, the 
 moment of the vessel's weight and the moment of buoyancy are calculated 
 about the end of the standing ways. If, when the ship's centre of gravity 
 passes over the end of the standing ways, the buoyancy moment falls below 
 the moment of weight, tipping will, under such circumstances, inevitably 
 occur. Care should be taken that a vessel has a reasonable margin of 
 moment to ensure against tipping, i.e., preponderance in buoyancy moment. 
 
 If the calculated results show that tipping is likely to ensue, something 
 must be done to prevent this : either the launching Avays must be lengthened 
 so that the vessel becomes more deeply immersed when the centre of gravity 
 passes the end of the standing ways, or else weight of some kind must 
 be introduced into the fore end of the vessel ; for instance, the fore peak 
 might be filled with water. This will have the effect of reducing the 
 ship's weight moment. Such a case could be further helped by increasing 
 the declivity of the after end of the ways. 
 
 The danger of tipping is considerably less in full-lined vessels than in 
 vessels with fine lines, the buoyancy moment naturally being proportionately 
 much greater in the one case than in the other. 
 
 The danger of a ship coming to a standstill in her course down the 
 launching ways may be caused by unfairness in laying the ways, or by too 
 little declivity, or sinkage of the earth, or insufficient tallow in frosty 
 weather. 
 
CHAPTEE VIII. 
 
 MAINTENANCE. 
 
 "While too much care and attention cannot be given to the quality of 
 the material used and the efficiency of the workmanship in the building 
 of a steel or iron ship, yet, throughout her whole life, she demands 
 incessant attention in order that she may be saved from the effects of the 
 perpetual attacks of deteriorating agents, if she is to be maintained in a 
 state of preservation and seaworthiness. 
 
 There are large numbers of vessels which, notwithstanding their 
 enormous first cost, need extensive and expensive repairs before they are 
 more than a few years old, this being, in many cases, entirely the result 
 of carelessness and neglect. 
 
 As long as steel, iron, and wood are surrounded by a dry atmosphere, 
 whether it be hot or cold, neither corrosion nor decay can arise. But as 
 soon as damp air or water come into contact with them, unless protective 
 measures of some kind be adopted, evidences of decay soon appear in a 
 greater or less degree. 
 
 Wood,* as a rule, succumbs to the attacks of these forces by far the 
 most rapidly.! As to the durability of iron as compared with steel, | there 
 is still a divergence of opinion as the result of varied experiences, though 
 the preponderance of weight seems to be against the comparative durability 
 of steel. Perhaps, in many cases, this results from the fact being over- 
 looked that, when Steel is adopted, it is usually of less thickness than 
 would be required for iron, owing to its greater tensile strength (indeed, 
 
 * Teak, under favourable conditions, has been known to last 2000 years. 
 
 t It is quite true that there are wooden vessels afloat to-day of 50, 60, and even 
 100 years of age. But this in no way proves that wood is as durable as iron or 
 steel for shipbiiilding, for these aged wooden ships have been so often repaired or 
 partially rebuilt, and decayed timber, fastenings, and bolts renewed, that in point of 
 fact not much of the original wood remains, while in some localities the structure has 
 been entirely renewed several times over. Steel and iron vessels also undergo extensive 
 repairs in the course of their existence, but not nearly to the extent to which it is 
 common to subject wooden ships. 
 
 X Note the supposed eff'ect of manganese, which is proportionately more abundant 
 in steel than in iron ('Manufacture of Steel,' p. 5). 
 
MAINTENANCE. 27 1 
 
 the classification societies make due allowance in the reduction of the 
 thickness of steel plates and bars), and, as a result, after corrosion has 
 taken place, and the scale removed, the thinness or reduced thickness is 
 more perceptible. 
 
 Moreover, notwithstanding the fact that during recent years great 
 permanent improvements have been made in the manufacture of steels — 
 especially mild steels, such as are used for ship plates and bars — that 
 perfection of manufacture has not yet been reached which guarantees the 
 uniformity of quality even throughout a single plate or bar : for, subject to 
 exactly similar conditions, it is not unusual to find that corrosion, often in 
 the form of pitting,* proceeds more rapidly in one locality than another. 
 It is also found that the rate of corrosion varies according to the constituent 
 properties of the liquid in which the metal may be totally or partially 
 immersed. Hence it is more rapid when subject to the action of salt water 
 than it is with fresh. Then it must be remembered that the drainage from 
 certain cargoes into the bilges may produce enormous and rapid decay. We 
 thus see that corrosion may be accelerated either by the nature of the 
 attacking agent or by the presence of certain elements in the composition 
 of the metal itself. 
 
 Understanding the rapidity with which decay may progress in steel or 
 iron when exposed to attacking agents, it is natural that shipowners, and 
 all those responsible for the maintenance of steel and iron structures, should 
 be alert to the necessity of discovering the most efficient means of so 
 protecting those parts subject to deteriorating influences, as to preserve 
 them from what must inevitably mean ultimate ruin. This protection is 
 usually afforded, in a greater or less degree, by some kind of composition 
 applied to the metal or material in liquid form. 
 
 Multitudes of 'anti-corrosives,' preventives, etc., have been patented and 
 brought into the market, each claiming the qualities necessary to prevent 
 corrosion or decay. However, while to specialize any particular manufac- 
 ture lies outside the province of this book, to specify the quality which 
 any composition must possess in order to prevent decay is not difficult. 
 It must hermetically seal the surface upon which it is laid, making it 
 both air and water tight. Any substance which does this will maintain 
 and preserve the structure in question, and will continue to do so in 
 proportion to the endurance it possesses, and the constancy of its attachment 
 to the material it covers. The condition of the surface upon which the 
 application has to be made is as important as the quality of the application 
 itself. Lead and zinc paints, cements, vegetable and mineral tar, and zinc 
 applied in the process of galvanizing, are among the principal materials 
 employed in various localities in ships, subject to varying attacking forces, 
 in order to preserve the structural strength and efficiency of the material by 
 preventing decay. 
 
 When steel and iron plates and angles come first from the manu- 
 * Pitting — small pin -like holes in the steel or iron. 
 
272 STEEL SHIPS. 
 
 facturers, their surfaces are not always smooth and clean, but it may often 
 be observed that they are covered with a scale of oxide of iron, or ' mill 
 scale,' which at first very often clings tenaciously to the metal. To cover 
 this with paint, or any other composition, would be worse than useless, for 
 the only condition under which satisfactory results can be obtained is that 
 the applied substance must come into direct contact with the metal itself, 
 neither dirt nor corrosion nor moisture intervening. It is therefore 
 absolutely essential that all mill scale be totally removed before the 
 application is made. Otherwise, instead of corrosion being utterly 
 prevented, it is allowed to proceed, and, eventually, this scale shells off, 
 carrying its useless covering of paint or other matter with it. The wise 
 thing, therefore, is to see that all this scale is entirely removed before any 
 application is made. This is generally attempted in ordinary merchant 
 shipyards by allowing the steel or iron to weather some considerable time 
 — that is, to leave all the steel or iron work unpainted until the vessel is 
 ready, before applying paint, etc. Experience, however, proves that this 
 method is not altogether effective, and thus many owners, as well as the 
 Admiralty, specify that all iron oxide be removed before the material is 
 placed on the ship. This is usually done by means of a solution of 
 hydrochloric acid in the proportion of one part of acid to nineteen of water. 
 The plates are placed in this bath, and after remaining some time they are 
 taken out, washed with fresh water, and scrubbed with steel brooms. They 
 are then coated with raw or boiled oil, and allowed to bleach in the weather 
 for some days. This process leaves the plate clean and smooth, and per- 
 fectly adapted for an application of the selected anti-corrosive composition. 
 
 An important consideration before the application of any substance of 
 the nature of paint is, that the surface of the iron or steel be thoroughly 
 dry, as well as perfectly clean. Unless these conditions are fulfilled, the 
 result will be far from satisfactory. Damp or moisture at once destroys, 
 in a marked degree, the tenacity of such-like compositions. This considera- 
 tion is more important than is often realized. It is a common practice 
 amongst shipowners, trading on long ocean routes, to supply a quantity 
 of paint or patent composition to their vessels, to be applied during or at 
 some stage on the voyage. - Nothing is more remarkable than the results 
 produced. In one vessel it gives the utmost satisfaction : it neither cracks, 
 nor blisters, nor shells off. It hermetically seals the surface of the metal, 
 and only praise is bestowed upon it. Another ship will be supplied with 
 identically the same composition, but a wide difference is experienced in 
 the results. It blisters, cracks, and shells off, water is admitted and 
 retained underneath the loose sheets of dry and useless composition, and, 
 as a consequence, rapid and often extensive corrosion is set up, and the 
 ship suffers more than she would have done had she been free from the 
 covering. Such an experience is most markedly observed where weather 
 decks have been covered in this way. In this vessel, the composition is 
 condemned as worse than useless, and only complaints are lodged against 
 
MAINTENANCE. 273 
 
 it. Were the whole truth known, it is quite probable that, in many cases, 
 the fault lies not with the application, but with the mode and conditions 
 under which it is applied. Here, carelessness is as disastrous as neglect. 
 
 With increasing experience, the more impressed does one become 
 with the truth that construction and maintenance are inseparably 
 associated. To design the constructive part of a ship regardless of the 
 fact that she will have to be maintained and preserved for many years 
 after she is built, if she is to satisfactorily fulfil her calling, betrays 
 unpardonable ignorance. 
 
 Strength is undoubtedly a foremost feature in tlie quality of a ship, 
 but the maintenance of that strength is equally important, and to effect 
 this preservation it is absolutely necessary that easy access be made to 
 all parts of the vessel, in order to survey, and examine, clean, paint, etc., 
 whenever desired. The paramount question should be how much access 
 can be given to ballast tanks, bilges, peaks, etc., without unduly weakening 
 the structure, and not how little can be made to suffice. Whenever 
 this latter method is adopted, the more liable are such parts to neglect. 
 Those parts of a vessel which present the greatest difficulty and awkwardness 
 in the means of access, are likely to be avoided and to receive the least 
 attention at the hands of workmen appointed to deal with them. 
 
 The relation of construction to maintenance will be more fully 
 observed as we consider sections of the ship in greater detail. 
 
 Maintenance of outside of Shell Plating. — If the vessel be new, 
 and in the builder's hands, as has been pointed out, it is absolutely 
 essential that all mill scale be removed, and that the surface be clean, 
 and smooth, and dry before the paint is applied. 
 
 A common method of doing this is to leave the bottom uupainted when 
 the vessel is launched. As soon as she reaches the water, a direct galvanic 
 contact is set up between the mill scale and the steel. This action, which 
 obviously tends to loosen the mill scale, is allowed to continue until the 
 vessel is completed, when she is put into dry dock, thoroughly scraped, steel 
 broomed, and washed with fresh water. When all — or as much of the miJl 
 scale as possible — has been removed, the bottom is given about three coats 
 of quick-drying anti-corrosive composition. 
 
 The first of these should be applied in as thin a state as possible — 
 almost pure oil. The next coat should have a medium, and the third, a 
 full body. Ample time should be given between each coat to allow the 
 composition to dry perfectly. The last coat is usually an anti-fouling 
 paint. 
 
 When the vessel is not put into dry dock before leaving the builder's 
 hands, she is usually allowed to run for a period not exceeding six months. 
 She is then put into dry dock, and all paint that can be afi'ected by an 
 ordinary scraper is removed. The bottom should then be treated in the 
 same manner as previously described. 
 
 If the vessel be an older one, and is being overhauled in a dry 
 
 18 
 
2 74 STEEL SHIPS. 
 
 dock, and it is found necessary to paint her, all rust, old painty dirt, 
 and all fouling must be entirely removed by scraping or chipping. The 
 new paint to be applied must come directly on to the dry metal surface. 
 For a ship's bottom, below the load water-line, excellent results are 
 obtained by applying hot zinc and tallow. Not only is this anti- 
 corrosive, but anti-fouling, and in cases where this composition has been 
 carefully and properly used, it has lasted for several years without any 
 sign of decay in the metal. The disadvantage of this excellent 
 ■composition is the great length of time required to apply it, especially 
 if the weather happens to be wet or frosty. For the topsides above 
 the load water-line, probably it is difficult to surpass red lead and oil, 
 or red lead and white zinc mixed, applied with similar precautions in 
 the usual way. Of course, to get these results, the zinc must be zinc, 
 and the red lead must be red lead. We are inundated with so many 
 patent compositions in these days, that it is not a little difficult to get 
 the original or genuine article, unless a price is paid for it. And the 
 huge differences in price at which zincs, etc., are offered, proves that 
 much of it must be, as experience shows, pure rubbish, if such a 
 ■contradiction may be used. 
 
 "Vessels which are in the habit of lying aground in certain rivers or 
 harbours to load or discharge at low tides, are subject to special injury 
 from the unevenness of the stony bottom. The effect is, that the paint 
 may be scraped off, and the bare surface of the plates exposed to the 
 •corroding action of the water. Corrosion always proceeds more rapidly 
 wherever the surfaces of plates or angles are severely scratched.. 
 
 In addition, boulders or large stones bearing against the edge of a 
 butt or seam, may have the effect of springing the caulking, or straining 
 the rivets, so as to admit water. Hence the importance of frequent 
 examination of ships' bottoms, which are subject to such experiences 
 as described. 
 
 If the shell is lap butted, care should be taken, when the vessel is 
 huilding, that all the outside butt edges point aft, so that, in coming 
 alongside a quay or landing, there is no chance of the plates catching 
 against any projection. Some owners find it advisable also to chamfer 
 the lower outside edges of the shell seam laps (which therefore point 
 downwards) abreast all hatchways between the light line and the gunwale. 
 This is advantageous to both ship and cargo when the latter is hoisted 
 out of barges by the ship's winches and derricks, and dragged up the 
 ship's side. 
 
 It frequently happens that vessels with bar keels trading over 
 shallow sandy bars are found to have their garboard strakes and keel 
 very much wasted by this continual grinding action. A means of 
 safeguarding these important parts from destruction is to fit a shoe 
 upon the keel as shown in fig. 178. To do this — as the illustration 
 shows — a plate about fths thick is bent round the keel, and either 
 
MAINTENANCE. 275 
 
 tapped on (connected by means of tap rivets), or tlie rivets through the 
 garboard and keel are taken out, and fitted long enough to go through 
 the keel, garboard, and shoe. 
 
 In the old-fashioned system of making shell butt connections by 
 means of straps instead of overlaps, the bilge strake butts frequently 
 showed signs of working, and, in many instances, the caulking was 
 corroded to such an extent as to leave a decided crevice. 
 
 Experience has taught that the best way in which to arrest this 
 action is to keep the crevices well filled with metal stopping. Where 
 this precaution has been neglected, the butts have, in some instances, 
 become so wasted as to necessitate the fitting of outside straps. 
 
 In other cases, where weakness has developed at these butts, assistance 
 has been given to the straps by fitting additional angle iron straps at right 
 angles across the butt. 
 
 Topside of Steel or Iron Weather Decks. — Opinion and experience seem 
 to favour the durability of iron as compared with steel for weather decks. 
 But in both cases, corrosion, sometimes in the form of extensive pitting, 
 takes place. It is scarcely necessary to draw attention to the importance 
 •of so arranging the deck plating that it will rapidly rid itself of water. 
 
 Fig. 178.— Plate Shoe on Bar Keel. 
 
 The strakes of plating from the centre line to the stringer plate, in 
 unsheathed iron or steel decks, should be arranged in and out (see figs. 12 
 and 48), and not one in and one out, as is usually done under wood decks 
 {fig. 58). To many shipowners, this corrosion of decks has been a sore 
 trouble, and numerous attempts have been made to prevent it, with varying 
 success. As previously pointed out, where these decks have been covered 
 with an anti-corrosive composition, the effect in many cases has actually 
 been to assist and encourage corrosion, owing to the application failing to 
 grip the deck, and eventually providing a lodgment for water. The fault 
 very often lies entirely in the carelessness evinced in making the application. 
 Where a steel or iron weather deck is sheathed with wood, the metal may 
 be well preserved by coating the clean, dry deck with a good covering of 
 vegetable tar — Stockholm or Archangel, not mineral tar, unless refined so as to 
 eliminate the ammonia. The wood deck should be laid hard down upon this, 
 never upon battens in way of inside strake, and the seams well caulked with 
 -oakum. When the oakum has been properly inserted, the seam should 
 be filled up with warm marine glue or vegetable tar. Butts should be 
 dealt with in a similar manner. If this work is effectively carried out, 
 
276 STEEL SHIPS. 
 
 water will only with difficulty be able to find its way through to the iron 
 deck. But should such leakage through the wood sheathing happen from 
 any cause, the deck is protected by the covering of tar. The bolts which 
 pass through the deck and wood sheathing ought to be well bedded in white 
 lead with dowels tightly fitted. This, again, will prove effective against 
 the passage of water. When a wood deck is laid upon the beams, with no 
 steel or iron deck, similar precautions are necessary, and the beams should 
 be cleaned and thoroughly painted with red lead and oil. Sometimes 
 a wood deck is badly disfigured by rust-coloured streaks in the region of 
 the butts, owing to corrosion having taken place in the beams below the 
 deck, and oozing up through the butt, promoting decay as well as disfigure- 
 ment. It is specially necessary, therefore, to ensure the top surface of the 
 beams being well coated with paint. Indeed, some owners specify that felt^ 
 dipped in red lead, be placed on the beams under all. butts, and, in the 
 case of yachts, where the whiteness of the wood deck is an important part 
 of the beauty of the craft, zinc slips are sometimes fitted on the beams under 
 all butts. 
 
 Owing to the generally acknowledged superiority of iron over steel 
 in off"ering resistance to corrosion, it is becoming more and more customary 
 to make all exposed weather decks — that is, those parts of uppermost decks 
 unsheathed with wood, and not covered by poops, bridges, or forecastles — of 
 iron. 
 
 Whenever a leak is found in the seam of a wood deck, or in the wood 
 sheathing of a steel deck, the leak should never be treated locally with 
 a caulking tool, as such treatment is almost sure to spring the seam on 
 either side of the defective part, and perhaps make the leak worse than 
 it was originally. 
 
 In all cases where it is found necessary to caulk a seam, the caulking 
 should at least extend from butt to butt. 
 
 Occasionally, the seams of a wood deck show signs of creeping after 
 being exposed under a hot sun, or after the vessel has experienced heavy 
 weather. This may occur even though the deck may have been caulked 
 but a comparatively short time previously. It is usual, under these 
 circumstances, to note the seams where the pitch has a cracked appearance, 
 and to caulk the worst ones right fore and aft, or to take every third seam 
 and harden them down right fore and aft, and, if very soft, to put in an 
 additional thread of oakum. 
 
 An easy method of finding the leaking or soft places in the seams of a 
 wood deck is to watch the deck drying after it has been wetted. When the 
 seams are good, they will dry as fast as the deck, but if there exist any 
 defect, the water will linger in the seam for a considerable time after the 
 other parts of the deck have dried. 
 
 In a sheathed deck, water will frequently find its way between the 
 wood and the iron or steel deck below it. This is almost inevitable, 
 notwithstanding the fact that great care may have been exercised in 
 
MAINTENANCE. 277 
 
 caulking the deck seams. If this water is allowed to remain, serious 
 corrosion may result. As time goes on the rust will increase, the deck 
 bolts will waste ; it is highly probable that the wood deck will spring up, 
 and more than likely that greater deterioration will occur than would have 
 happened had the deck been left unsheathed. 
 
 To prevent water accumulating, a scupper pipe should be led to the 
 bilges at the lowest part of the deck near the gunwale under the sheathing, 
 so as to drain oflf the water. 
 
 This simple device may be the means of saving large and expensive 
 repairs. 
 
 Many expedients have been adopted in order to preserve wood deck:?. 
 
 The commonest is to coat the decks with 9i parts of raw oil to J part 
 of terebeue, and another is to use a solution of gum and turpentine. 
 
 Double Bottom Water-ballast Tanks. — Probably the structural parts 
 enclosed in double bottom water-ballast tanks are more subject to the 
 ravages of corrosion than any other part of the ship, and many reasons 
 may be enumerated to show why this is so. 
 
 The best conditions under which the pi-eservation of iron or steel may 
 be effected are : either to keep the metal entire! j'^ dry, or entirely immersed. 
 Both these conditions are impossible in regard to the material constituting 
 water-ballast tanks. As water is only carried as ballast when it cannot be 
 s;ifely or conveniently dispensed with, it follows that these tanks are alter- 
 nately full and empty, and the material forming them alternately wet and 
 dry, or partially so. This, therefore, is one cause promoting corrosion. 
 Moreover, to pump these tanks absolutely dry is practically impossible, 
 and, as a result, when they are supposed to be empty, the two or three 
 inches of water in them is continually splashing and wetting the steel or 
 iron, as the ship rolls and pitches. As most tanks are practically void of 
 all ventilation, the air in them must be, in a greater or less degree, charged 
 with moisture. Such conditions are found in a still more aggravated form 
 under the boilers, where the heat produces a damp, steamy atmosphere, 
 which is found to encourage corrosion to a most serious extent in a 
 remarkably short time, unless special measures are taken to prevent it. 
 
 On surveying these tanks, signs of corrosion are more evident in some 
 parts than in others. In M'Intyre tanks, excepting under the boilers, the 
 wear and tear seems to be most prominent in the angles at the bottom of 
 the fore and aft girders standing upon the floors. Signs of loose, and 
 sometimes broken rivets, are found, especially in the side keelsons, caused 
 probably by the movement of free water when the tanks are not full. 
 
 The under side of tank top plating, and, indeed, all the upper material 
 in these tanks, is more subject to corrosion than the lower material, probably 
 owing to the greater amount of condensation which takes place in this 
 locality. 
 
 In cellular-bottom tanks, the upper parts also are more subject to 
 corrosion than the lower, especially the under side of the tank top plating, 
 
278 STEEL SHIPS. 
 
 the reverse bars, and the upper fore and aft girder angles. Considerable 
 wear is found at the limber holes in the floor plates. This no doubt is due 
 to the wearing action of cinders, sand, or other rubbish finding its way 
 into the tanks. The presence of stray bolts or rivets, which have care- 
 lessly been left in these tanks, sometimes produces a startling amount of 
 damage, for, confined in a frame space, their action, when the tanks have 
 been pumped empty, and the ship is rolling and pitching, has been, in some 
 cases, to wear through a considerable thickness of cement, and actually 
 through a thick keel plate. 
 
 The rate of corrosion throughout the double bottom is comparatively 
 slight excepting under the boilers, where, if the greatest vigilance is not 
 maintained, disastrous results occur. The heat from the boilers strikes 
 through the tank top plating, and the warm stagnant air (efficient ventila- 
 tion being exceptional), charged with moisture, produces a condition 
 immensely favourable to rapid corrosion. For about a foot down from the 
 tank top, and also between the floor manholes, this decay is found to be 
 most excessive. Sometimes, vessels not more than four or five years old, 
 through sheer neglect, have had the upper parts of their floor plates eaten 
 entirely through. Double bottom tank margin plates often show rapid 
 signs of corrosion, probably owing to bilge drainage. 
 
 Understanding the nature of the disease to which these tanks are subject, 
 the important question is, to ascertain the most efficient means of prevention. 
 
 As previously stated, the freest and most convenient access possible 
 should be provided, for the simpler and easier the means of reaching every 
 part of the tanks, the more likely is it that due attention will be paid in 
 examining and dealing with them. Special attention should be given to 
 the fitting and placing of the tank suctions, the aim being to empty the 
 tanks as completely as possible, so that little or no water remains. The 
 area of the water passages through the floors from floor space to floor 
 space towards the suctions should be ample. The lowermost of these 
 watercourses should be cut through the vertical flange of the frame bar 
 
 and floor plate thus I f~ ) I and should not be more than 
 
 three-quarters of an inch from the bottom of the tank (see fig.), so as to 
 allow as much water as possible to drain towards the suctions. The sketch 
 in fig. 173 illustrates the best form of suction mouthpiece for most entirely 
 emptying a tank. It is made of galvanized cast iron, and is dropped 
 into a cavity formed in the cement. 
 
 As the under side of tank top plating is particularly subject to corrosive 
 action, accelerated largely by the prevalent conditions which inevitably exist 
 and are so conducive to extensive condensation, iron is preferable to steel for 
 this part of the structure, because of its greater required thickness and 
 admitted superiority in resisting corrosion. Under the boilers, where the 
 worst eff'ects are experienced, it is advisable to adopt considerably thicker 
 
MAINTENANCE. 279 
 
 tauk top plating than satisfies the requirements of Lloyd's Kules from a 
 structural point of view. This method would considerably postpone what, 
 under ordinary circumstances, would be extensive repairs to the inner 
 bottom, because of the good margin of thickness. 
 
 Any method of ventilating or inducing a current of air would immensely 
 reduce the obnoxious conditions existing in most of these tanks, and rid the 
 space of the perpetual presence (when not entirely filled with water) of 
 highly injurious atmosphere. 
 
 Some owners prefer to dispense with the tanks under boilers, and not 
 close in the space as a watertight compartment. As shown in fig. 161, the 
 tank top plating is dispensed with, allowing free communication at all 
 times between the inside and the outside of the space. This idea is 
 certainly all right from a purely ' preservation of material ' point of view, as 
 the space is greatly relieved from the injurious effects of damp and 
 stagnant atmosphere, but it forms a serious objection from the ship safety 
 standpoint, nullifying to a serious extent the immense value of a double 
 bottom all fore and aft. Should the outer bottom in this space by any 
 means be injured and the sea find ingress, the ship is most hopelessly 
 crippled, even assuming that she still possesses sufficient reserve buoyancy 
 to compensate for the loss of the boiler space buoyancy — which usually 
 includes the engine space also — and that she is able to float in comparative 
 safety. Her machinery is utterly useless, and steamers in these days are 
 remarkably devoid of sailing equipment. 
 
 A double bottom only satisfactorily fulfils this, one of its chief functions, 
 when it is absolutely continuous from the collision bulkhead to the aftermost 
 bulkhead. 
 
 A much better plan is to construct the inner bottom continuous all fore 
 and aft with a distinctly separate watertight compartment under the boilers, 
 but to strictly exclude all water, and thus keep it always dry. In this way 
 the tank could be wonderfully preserved by coating it with refined vegetable 
 tar sprinkled with cement. Certainly it would reduce the capacity for 
 water-ballast, as would also the previous method mentioned, but as the 
 boiler space is usually comparatively small in cargo vessels, the reduction 
 would not be very serious. 
 
 In any case, it is advisable to keep the boilers as high as possible above 
 the tank top, say 2 ft., or 2 ft. 6 in., and by this means the heat is less 
 seriously experienced. Some owners have found that considerable protection 
 from the heat of the boilers is ensured by covering the tank top under the 
 boilers with two or three inches of cement and tar.* 
 
 • A system adopted by Messrs. W. Doxford & Son.s, Limits J, Sunderland, for the 
 protection of the structural material under the boilers in double-bottomed vessels built 
 for the Clan Line, is that of galvanizing the tank top, floor plates, intercostals, 
 etc., in the region of the boiler room. This system is said to have produced highly 
 satisfactory results, and although many owners might be inclined to object to the 
 additional first cost, yet, if the intended object is accomplished, it cannot fail to bo 
 well worth the additional expense, and indeed prove in the end to be economical. 
 
2 8o STEEL SHIPS. 
 
 Whenever water is carried in these cellular-bottom tanks, perhaps the 
 best protection is afforded by coating every part of their interior surfaces 
 . with a liquid of Portland cement and water applied with a large brush 
 after the interior has been thoroughly dried and cleansed out and rid of all 
 oxide of iron. This method, while commonly adopted, has proved very 
 effective when conscientiously carried out, and repeated at such intervals as 
 found necessary by examination. 
 
 Cementing. — In past years, before the double bottom system became so 
 universal a mode of construction for cargo and passenger vessels, and the 
 older ordinary floor system was still in vogue, considerable water or liquid 
 found its way into the bottom of the vessel. This might be caused by 
 sweating in the holds, or drainage from the cargo. In most ships, and 
 especially in those carrying certain cargoes such as copper ore or sugar, 
 which are particularly injurious to iron or steel, it has long been the custom 
 to put a substantial covering of a mixture of Portland cement and sand or 
 brick dust over the inside of the bottom plating from bilge to bilge. This 
 cement was thick enough to amply cover the plating and rivet heads, and 
 the frame flanges upon the shell, and afforded excellent protection from 
 the effects of cargoes, such as we have indicated. 
 
 While this system of cementing is more desirable in vessels with 
 ordinary floors, where all drainage finds its way on to the bottom plating, 
 there seems little or no reason for practising it to the same extent on the 
 bottoms of vessels inside the tanks where double bottoms are fitted. 
 
 Indeed, there is much to be said against it. In the first place, no 
 bilge drainage of any sort can find its way inside a double bottom tank. 
 The bilges on each side of the outer wings of these tanks, and sometimes 
 specially prepared wells, are the reservoirs for all such drainage, and there- 
 fore the original reasons for using this cement no longer exist inside the 
 double bottom tank. The weight of this cement amounts to scores of tons 
 in vessels of moderate size, which necessarily represents an equivalent loss 
 in deadweight. The vessel is therefore handicapped with a perpetual and 
 unnecessary burden. Moreover, this cement, instead of being a source of 
 protection, may prove a source of actual injury to a ship, unless the most 
 rigorous inspection be maintained, and frequent surveys instituted. 
 
 As we have mentioned previously, considerable injury may be inflicted 
 upon the outside bottom of a vessel lying aground to load or discharge. 
 But even when little or no trace of damage is apparent from the outside, 
 owing to the utter inflexibility of the cement, it may have become cracked 
 and loosened, and while almost if not entirely undetectable by eye, it 
 permits water to ooze through and lodge against the shell of the vessel, and 
 the mischief is not discovered until serious damage has been done. Even 
 the slightest working of a vessel which never touches the ground with her 
 keel under any circumstances, may crack the cement. If this cementing is 
 confined to the keel plate, and at most over the garboard strake on each 
 side, where cinders or other foreign matter which accidentally get through 
 
MAINTENANCE. 251 
 
 the manholes may exert considerable friction against the bottom, it ought to 
 prove sufficient. All other parts of the interior of these tanks should be well 
 coated with cement wash, and re-coateil at intervals. 
 
 Outside of Inner Bottom Plating, Bilges, and Wells. — In order to 
 best preserve the hold surface of tank tops, as we observed iu decks, a 
 foremost consideration should be to render the lodgment of water an absolute 
 impossibility. In building a double-bottom tank according to Lloyd's 
 requirements, the depth of the centre girder or centre keelson, and the width 
 of the margin of the tank, are fixed by rule. It is not a little surprising, 
 however, what amount of ingenuity has been exercised (often purely 
 economic) in manipulating these two fixed dimensions, and in producing, 
 in similar ships, tanks of vastly different capacity and shape. The general 
 system which used to be followed was to set off the depth of the centre 
 girder, lay the tank top plating horizontally over this, and extend it in each 
 transverse direction, until, by striking a straight line perpendicular to the 
 bilge, the required depth of margin was obtained. By lowering the 
 transverse extremities of these tank tops two or three inches, so as to give 
 an inclination towards the bilge and thus assist the drainage of any water in 
 that direction, probably the most satisfactory double-bottom tank to all 
 concerned is produced. (See figs. 12, 48, etc.) 
 
 Until recently, the almost universal custom has been to cover this 
 tank top plating with broad deals of timber (ceiling) in a fore and aft 
 direction over wooden grounds arranged athwartships. Where this is done, 
 the tank top strakes of plating are usually arranged in the same manner as 
 steel decks under wood sheathing, — alternate strakes in and out, unless the 
 reverse bars are joggled, or the edges of the strakes of plating flanged. As 
 this ceiUng is not usually caulked, but simply clamped together, it is 
 necessary to protect the plating by some efficient covering. This is often 
 satisfactorily accomplished by applying a good coating of vegetable tar, or 
 specially refined coal tar sprinkled well with cement. Where this is carried 
 out with all due precaution, the tar will serve its purpose for years, and 
 leave the iron as clean as ever. However, as is now being discovered, under 
 many circumstances this wood ceiling is utterly useless in a ship's hold, 
 costing money, adding weight, occupying space, and often harbouring water 
 and dirt beneath it. It has now become a frequent practice to dispense 
 with it. 
 
 The strakes of plating are arranged like those of an unsheathed weather 
 deck, each strake in and out like the slates on a house roof, and, with the 
 slight declivity previously recommended, any water flows off" into the bilges. 
 (See fig. 48.) The continued friction of loading and discharging cargo keeps 
 the tank top plating clean and smooth, and generally neither paint nor 
 covering for purposes of protection is necessary. Sometimes, by giving the 
 tank top an abnormal declension towards the bilges, and a fair rise of floor, 
 it is surprising how small and useless the double bottom may become, for 
 the width of the tank side can only be obtained at an absurd distance from 
 
252 STEEL SHIPS. 
 
 the bilge. True, considerable saving can be effected in the weight of iron or 
 steel in the construction of the tank, but the capacity of the tank is seriously 
 reduced for the carriage of water-ballast, and its efficiency as a double 
 bottom greatly minimised owing to the tank covering only part of the 
 breadth of the ship's bottom. 
 
 When tank tops are left unsheathed with ceiling, they are usually 
 protected immediately under the hatches with a covering of wood battens, 
 which are surrounded and held in position by an angle bar riveted to the 
 plating. The reason for this is obvious. Sometimes, in lowering heavy 
 weights into the hold, they are allowed to bump on to the tank top, and 
 were it not for the protection already described, considerable injury might be 
 done. Even though the plating itself were not seriously damaged, caulking, 
 rivets, or butts might be sprung, so as to destroy the watertightness of the 
 tank. Sometimes, even where the precaution of protecting the tank is 
 adopted, or the whole area covered with ceiling, carelessness in loading 
 cargo, or the accident of a heavy weight falling, may cause the same 
 injury. 
 
 In cases where the rise of floor is very great, were a horizontal or 
 nearly horizontal top to be given to the tank by adopting the rule minimum 
 width of margin on tank side, the tank would be extremely narrow, and 
 thus only partially cover the bottom, and moreover, contain only a compara- 
 tively small quantity of water for ballast. In such a case, a method has 
 sometimes been adopted of raising the tank top from the top of the centre keel- 
 son, toward the bilges. By this means, any water finding its way on to the 
 tank top would flow towards the centre line (the lowest point) and thence 
 into a well at the after end of the hold space. Here again, the strakes 
 of plating should be arranged like the slates on a house roof, each in 
 and out. The idea is simply to get rid of all water in the most rapid 
 manner. 
 
 Bilges. — Few, if any, parts of a ship are of more vital importance than 
 the bilges. These form the drains for the holds. All sweat moisture, all 
 moisture from cargo, even rubbish, and sometimes water from scuppers, find 
 their way into the bilges. The result of the combination, depending much 
 upon the nature of the cargo, is often a liquid possessing tremendous pro- 
 pensity for attacking and corroding iron or steel. Where the material in 
 these bilges has not been adequately protected, vessels carrying cargoes such 
 as copper, sugar, rice, in a single voyage have evinced remarkable decay. 
 Hence the necessity for frequent inspection, and the readiest and simplest 
 means of carrying out this operation. The most prevalent arrangement of 
 covering in the bilges to prevent, as far as possible, cargo or dust finding its 
 way into the space, is to extend the ceiling over the bilges to the ship's side. 
 Entrance to the bilges is obtained by lifting a fore and aft plank. (See fig. 
 64.) A better arrangement is that shown in fig. 179, where the bilge 
 coverings consist of iron hinged shutters, occupying less space, and obviously 
 much handier, not to mention the better view obtained for examination when 
 
MAINTENANCE. 
 
 283 
 
 they are lifted, afforded by the full exposure of the bilges. The pump 
 suctions in the bilges should be carefully placed so as to rid the space of 
 water as effectively as possible. Here, perhaps, more than anywhere else 
 in the vessel, cement should be liberally applied. The shell and tank 
 side should be thoroughly coated, and the bilge filled with cement up to the 
 level of the drain holes in the tank bracket plates. The cement should 
 be very smoothly laid, so that the water may find its way without any 
 obstruction to the suctions. Only by carefully carrying this out, and 
 frequently surveying this space, can this part of the vessel be saved from 
 rapid decay. 
 
 Wells. — Sometimes the drainage from the holds is allowed to accumulate 
 
 Continuous plate making an 
 excellent addition to the 
 connection of the frame legs 
 to the double bottom 
 
 Fig. 179.— Hinged Bilge Shutters. 
 
 solely in the bilges, from whence it is pumped. Often, however, the bilge 
 water is drained into wells (see fig. 174), from which steam suctions are 
 led for its discharge. These wells are usually situated at the after end 
 of a hold space, or engine and boiler space, and extend from bilge to bilge 
 with a length of about two frame spaces (4 ft.). The two ends forming the 
 extremities of the adjacent water-ballast tanks are, of necessity, strictly 
 watertight. This well space is a watertight compartment, excepting that 
 a few (about three) holes are punched in the tank side plate near the 
 bottom of the bilge, which allow the bilge water to drain into the well. 
 Being subject to the same wasting forces as are found in the bilges, these 
 wells require similar treatment for their protection. The chief objection to 
 them is that they interrupt the continuity of the watertight double bottom, 
 
284 STEEL SHIPS. 
 
 and thus, should damage happen to the outside plating in the way of these 
 wells, free access is opened to the large hold spaces, and the advantage of 
 a continuous inner bottom is lost, and its value therefore decreased. 
 
 Hold and 'Tween-deck Spaces. — On examining the under side of 
 steel or iron decks, the shell plating, and the framing in the holds of steel 
 vessels, one is sometimes surprised to discover to what extent corrosion may 
 be found in a place so exempt from weather. The under side of decks, 
 beams, beam knees, frames, and especially reverse frames, as well as the 
 shell, are liable to extensive corrosion. Keenly alive to this fact, many of 
 those responsible for the maintenance of iron and steel vessels, make 
 vio'orous and commendable efforts to save their vessels from the disastrous 
 effects which, unless effectively dealt with in time, inevitably ensue, utterly 
 ignoring the cause of the trouble. Everyone knows that to close up the doors 
 and windows in a room, and permit little or no ventilation, will result in 
 damp and sweating on walls and windows, or any other cold surface, and 
 especially will this be noticeable if the atmosphere be subject to any slight 
 heating. To get rid of this undesirable state of affairs, a current or draught 
 of air must be obtained by some means or other. Or, in other words, good 
 ventilation is necessary. Ships' holds produce similar results under similar 
 conditions. These spaces in hosts of cargo vessels are badly and inefficiently 
 ventilated, and with such a large area of naturally cold surface, condensation 
 or sweating upon beams, decks, frame, shell, etc., to an enormous extent, is 
 the result, often causing damage to the cargo. In ships where ventilation 
 has received adequate and intelligent attention, sweating is almost unknown, 
 and little or no injury is inflicted upon cargo and hull. It is not so much 
 continual scraping and painting that is required, but perpetual and efficient 
 ventilation in these spaces to dispose of this evil. 
 
 The more that air is induced to circulate through the holds, the less ill 
 effect is likely to be experienced by the ship and cargo. For miscellaneous 
 cargo, oil, and coal-carrying vessels, good and abundant ventilation is of the 
 very highest importance. By thoroughly cleaning the surfaces of all plating 
 and angles, and coating with red lead and oil, repeating the operation as the 
 wear and tear of the vessel indicates, there ought to be little difficulty in 
 throughly maintaining and ^^reserving all iron and steel exposed in the 
 holds. When sweating does occur, it should be seen that no pools of water 
 lodge upon hold stringers, or any such like projecting material. Small 
 holes drilled or punched in these places will allow the water to find its 
 way down into the bilges. Crevices or corners which are likely to harbour 
 water or dirt should be filled with cement. 
 
 The shell plating in the hold or 'tween deck spaces in the neighbourhood 
 of all side scuttles should be subject to periodical inspection, as leakage 
 may possibly occur, and thus promote decay. Hence the cleading on the 
 ship's sides in the 'tween decks of passenger steamers should be partially 
 removed at intervals, where the danger of such leakage exists. 
 
 Bunkers. — Bunkers are another locality in a ship which should receive 
 
MAINTENANCE. 285 
 
 adequate attention. The gas emitted from coal seems to have a specially 
 injurious effect upon iron and steel, particularly the latter, if the surface is 
 not protected by some suitable covering. Various compositions are found to 
 do this eifectively, while other owners get every satisfaction by coating their 
 bunkers with refined vegetable tar, and others by bituminous cement. 
 Ordinary paint rarely proves efficacious for any length of time against 
 the special conditions existing in bunkers. 
 
 Fore and After Peaks. — Fore and after peaks, when only used for cargo 
 or stores, should be kept as dry as possible. The deep recesses between 
 the floors would render this very difficult, were it not customary to fill the 
 lower extremities of these spaces with cement. Indeed, wherever awkward 
 recesses exist, in which water would be likely to lodge, or whatever small 
 spaces are found to be difficult to get at for examination, the better plan 
 is to fill up these crevices with cement. If the peaks are used for water- 
 ballast, probably a thorough coating of thick cement wash, or refined 
 vegetable tar, will prove as good a means of preservation as anything. 
 "When the space is only used for stores, or perhaps cargo, good paint will 
 give the necessary protection. Here, again, too much importance cannot be 
 laid upon the ventilation of these spaces. No preventive or anti-corrosive 
 will stand as well where the atmosphere is foul and stagnant, and those who 
 know anything about peaks will be aware that so utterly polluted, and 
 filled with poisonous gases, are these places in some vessels, that when first 
 the manholes or hatches are removed, neither lamps nor light will burn 
 inside, and to enter them would mean certain death. Indeed there are cases 
 on record of such disasters occurring. 
 
 Deep Water-ballast Tanks, also used as cargo spaces. — Like double- 
 bottom ballast tanks, these compartments are subject to alternating wet 
 ami dry. The importance of an abundant and well-arranged system 
 of ventilation must be obvious in order to get rid of all moisture and 
 damp atmosphere as early as possible, after the tanks have been pumped 
 empty. No doubt much damage is done to all kinds of water-ballast tanks, 
 owing to the fact that they are not always filled to the uttermost when used 
 for ballast. "When completely filled, the mass of water has no more 
 injurious effect than a well-stowed cargo, but where tanks are carelessly left 
 partially filled, it is difficult to estimate what damage may not be done. 
 Bulkheads may be ruptured, rivets loosened and broken, caulking destroyed, 
 and cargo in adjoining hold spaces seriously damaged. These tanks can be 
 made perfectly watertight, thoroughly reliable, and most satisfactory, and 
 of tremendous advantage to vessels going to sea without cargoes. And 
 yet more than one owner has dispensed with them, because of the trouble 
 they have caused. Such trouble can only arise from workmanship and 
 construction, or else gross carelessness on the part of those who have charge 
 of them — most probably the latter. Deep ballast tanks are being more 
 adopted than ever in ocean steamers, and with due care and attention in 
 construction and in treatment, they can be maintained with comparative 
 
286 STEEL SHIPS. 
 
 ease. Kefined vegetable tar or refined coal tar has proved to be a good 
 protecting agent in the maintenance of these tanks. 
 
 Corrosion owing to Galvanic Action.— The proximity of certain metals 
 to iron or steel may set up galvanic action, and, consequently, promote 
 corrosion. Hence when bronze propellers are fitted, unless suitable pre- t 
 
 cautions are taken, severe pitting is created in the stern post. This galvanic V 
 
 action may be counteracted by tapping thin zinc slips to the aperture surface i 
 
 of the stern frame. 
 
 A peculiar feature in cast iron propellers after being in use for any con- 
 siderable time is the excessive pitting which takes place on the back of the 
 propeller blades. Opinions are numerous as to the cause of this, but no 
 definite explanation can be given. However, by riveting thin brass plates 
 on to the back of the blades of cast iron propellers, protection is afforded and '^ 
 
 such injury obviated.* ' 
 
 * Shipowners might save themselves enormous expenditure for repairs, etc., if 
 more attention, systematically carried out, were conscientiously given to their vessels. i 
 
 Much of the deterioration and decay is entirely the result of neglect which might | 
 
 be greatly prevented, if, in addition to the ship superintendent, whether he be a 
 naval architect, an engineer, or a ship's oiBcer, another appointment were instituted 
 in large companies by which a man, thoroughly acquainted with ship construction 
 and the particular localities in ships liable to decay, could give his whole attention to 
 this matter. The ordinary ship superintendent of a fleet of vessels has too much 
 other work to attend to in the performance of his duties to give the attention which 
 is necessary for the preservation of the vessels entrusted to him. Where this vitally 
 important work is left in the hands of ill-paid or irresponsible persons, neglect to 
 hidden parts to which it is difficult to obtain access ensues. 
 
INDEX. 
 
 Al at Lloyd's, 30. 
 
 Awning Deck Vessel, 49, 109, 117-119. 
 ,, ,, ,, scantlings for, 120. 
 
 Bar keels, 182-3. 
 ,, ,, riveting, 183. 
 ,, ,, scarphs, 182. 
 Basic steel, 15. 
 
 Beam, camber mould of, 181, 195. 
 ,, knees, 195-202. 
 ,, knee riveting, 202. 
 Beams, 35, 194-202. 
 
 ,, connections to hatch coamings, 
 
 196, 200. 
 ,, number of tiers, 38. 
 Bending moment, extreme, 95, 98. 
 
 ,, ,, estimate of, in actual 
 
 practice, 99. 
 of girder, 84. 
 ,, ,, on loaded bar, 67. 
 
 Bessemer steel, 3, 14. 
 Bilge keels, 255-6. 
 
 ,, keelson, 40. 
 Blast furnace, 6-8. 
 Blind holes, 167. 
 Bosom piece, 177. 
 
 Bossed plating for twin screws, 238. 
 Breadth moulded, 50. 
 Breast hooks, 255. 
 
 Bridge, front bulkheads of, 130, 253-255. 
 ,, tendency to fracture at ends of, 71. 
 British Corporation, function of, 27, 28. 
 Bulb augle frames, 192. 
 Bulkhead doors, 220. 
 
 liners, 217-219. 
 ,, ,, in oil steamers, 158. 
 
 ,, rivets, 219. 
 ,, spacing of rivets of, 217. 
 Bulkheads, 213-220. 
 
 ,, caulking of, 217. 
 
 ,, caulking of, in oil steamers, 
 
 158. 
 ,, connection to shell and decks, 
 
 215. 
 ,, height of, 215. 
 
 ,, in oil steamers, stiffening of, 
 
 157. 
 ,, in 'tween decks, 64. 
 
 ,, number of, 214, 215. 
 
 ,, plating of, 215. 
 
 ,, recessed or stepped, 215, 219. 
 
 stiffening of, 216, 219-220. 
 
 Bulwark plating, 230. 
 Buoyancy, 54. 
 
 ,, curve of, 93. 
 
 Bureau Veritas, function of, 27, 28. 
 Butt connections, strength of, 173. 
 
 ,, laps and straps, 171. 
 
 ,, ,, ,, ,, comparison of, 172. 
 
 ' Campania,' structural features of, 132. 
 
 Case-hardened steel, 15. 
 
 Castings, 8. 
 
 Cellular double bottom, 37. 
 
 5, ,, ,, tanks, 161. 
 
 Cement, disuse of, 146, 152, 280. 
 Cementation steel, 15. 
 Cementing, 280. 
 Centre keelson, 39, 41. 
 
 ,, ,, standing upon floors, 209. 
 
 ,, of pressure, 58. 
 
 ., through plate, 41, 
 
 ,, ,, ,, keelson, 210. 
 
 Channel bar frames, 193. 
 
 ,, pillars, 208. 
 Class, grades of, 29. 
 
 ,, maintenance of, 30, 
 Classification, 26-32. 
 
 ,, societies, purpose of, 26. 
 
 Club-footed, 72. 
 Cofferdams, 155. 
 Collision bulkheads, 214. 
 Companion-way, framing of, 199. 
 Comparison of butt laps and straps, 171. 
 ,, of scantlings of types, 120. 
 
 ,, of vessels for strength, 96. 
 
 Compensation for hold beams, 38. 
 Composite deck houses, 250-253. 
 Construf>tion, details of, 167-269. 
 Continuity of strength, 246-247. 
 Corrosion, steel versus iron, 24. 
 Crucible steel, 15. 
 Curve of bending moments, 94. 
 
 ,, ,, buoyanc}', 93. 
 
 „ ,, loads, 84, 94. 
 
 ,, ,, shearing forces, 94. 
 
 ,, ,, ,, stresses, 85. 
 
 ,, ,, weights, 91. 
 
 Deck cargoes, stresses due to, 74. 
 ,, erections, 126. 
 ,, ,, and dead weiglit, 107. 
 
 ,, houses, composite, 250-253. 
 ,, stringers, 41. 
 
INDEX. 
 
 Deck weights, 63. 
 Decks, 220-225, 
 Declivity of keel, 178, 266. 
 
 ,, ,, launcliing ways, 177, 267. 
 Deep framing iu lieu of hold beams, 47, 97, 
 193. 
 ,, tanks, 164. 
 Deformation, tendency to longitudinal, 65, 
 69. 
 ,, ,, ,, transverse, 61. 
 
 Depth of hold, 50. 
 
 ,, moulded, 50. 
 Derrick tables, 248-250. 
 Derricks and masts, 248, 250. 
 Details of construction, 167-269. 
 Diaphragm plates, 124. 
 Disposition of material, 96. 
 
 ,, ,, weiglit, 68. 
 
 Double bottom framing, 36, 191. 
 
 ,, ,, tanks, 160. 
 
 Doublings to decks, 223. 
 
 ,, ,, outside plating, 226, 228-30. 
 
 Doxford's turret steamer, 138-42. 
 Drift punch, 153, 167. 
 Ductility of steel, 19, 77, 
 
 Erection of transverse framing, 191. 
 Erections on deck, strength afforded by, 
 104. 
 ,, ,, ,, strengthening at ends 
 
 of, 104. 
 Elasticity of steel, 19, 76, 
 Elastic limit of steel, 77, 19, 20, 76. 
 Engine casings, 255. 
 Excessive strength, 108. 
 Extreme breadth, 50. 
 
 ,, proportions, 50. 
 " Equivalent girder," 87. 
 
 Faving surfaces, 185. 
 Fatigue, 21. 
 
 Ferromauganese in iron and steel, 14. 
 Flat plate keel, 184. 
 Floor plates, 35, 189. 
 ,, ,, bending, shearing, and punch- 
 ing of, 189-190. 
 Forces exerted upon ships, 52. 
 Forgings, steel and iron, 16. 
 Foundation plates, 210. 
 Frame bending, 186, 
 ,, butts, 36. 
 ,, bevelling, 186. 
 Frames, 34, 185, 192, 193. 
 
 ,, bulb angles, 192. 
 
 ,, channels, 193, 
 
 ,, punching of, 185, 
 
 ,, Z, 192, 
 Framing, 185-194. 
 Freeboard tables, 126. 
 Fundamental types of vessels, 107. 
 
 Grades of class, 29. 
 
 'Great Eastern,' system of construction, 
 
 70, 136. 
 Gudgeons, stern frame, 234, 245-6. 
 
 Hanging keel, 191. 
 
 Hatch coamings, 197, 250. 
 
 Hatches, 250. 
 
 Heel piece, 177, 190. 
 
 Hinged top mast, 250. 
 
 Hold beams, compensation for, 38. 
 
 Hold pillars, widely spaced, 146, 149. 
 
 Hollow pillars, 203. 
 
 Intercostal keelsons, 41, 210. 
 
 Insulation, 255. 
 
 Iron, classification of, 11. 
 
 ,, sections, 18. 
 
 ,, tests for, 23, 
 
 ,, thickness of plates, 23. 
 
 Keel, declivity of, 178, 266. 
 
 Keel blocks and launching ways, 177. 
 
 ,, ,, foundation for, 177. 
 Keels, 182-184, 191, 
 Keelsons, 39, 209, 
 
 Launching, 265-9. 
 
 ,, stresses in, 75, 
 
 ,, ways, declivity of, 177-8, 267. 
 
 Laying off, 178, 
 Length over all, 49. 
 
 ,, between perpendiculars, 49, 
 Limit of elasticity, 77, 
 " Lloyd's depth," 50, 
 " ,, length," 50, 
 
 ,, Register, function of, 27, 28. 
 Load Line Act, 1890, 28, 126. 
 ., lines, 29. 
 
 ,, ,, societies which assign, 28. 
 ,, ,, standard of strength for, 28. 
 Loading aground, stresses due to, 74. 
 Local stresses, 52, 72. 
 Locking pintle, 246. 
 Longitudinal bulkheads, value of. 213. 
 ,, framing, 33, 39. 
 
 ,, stresses, 52, 65. 
 
 ' Lucania,' structural features of, 132. 
 
 M'Glashan's side water ballast tank, 165. 
 M'Intyre tanks, 36, 160, 
 Machinery space, structure of, 247-248, 
 Maintenance, 270-286. 
 
 ,, of class, 30. 
 
 ,, ,, top side of steel decks, 
 
 275-277. 
 Malleable iron, 3, 10. 
 
 ,, cast iron, 10. 
 Manganese in iron and steel, 5. 
 Margin plate, 41. 
 Masts and Derricks, 248-250. 
 Maximum stress, 125. 
 Mild steel, 3. 
 Modes of construction, 42. 
 Modification in depth of transverse fram- 
 ing, 97. 
 Moment of inertia, 79, 96. 
 
 ,, ,, ,, calculation of, 80. 
 ,, ,, ,, integrational formula 
 of, 81. 
 
INDEX. 
 
 289 
 
 Moment of inertia, illustration of, 83. 
 ,, ,, ,, midship section, 100. 
 ,, ,, ,, of girder, 88, 90. 
 
 New features in modern shipbuilding, 
 
 146. 
 Nickel in steel, 5, 21. 
 Numerals for scantlings, 47, 112, 113. 
 Neutral axis, 78. 
 
 ,, ,, calculation for position of, 
 
 99. 
 ,, surface, 78. 
 
 Ocean Steamship Company's steamer, 
 
 146-152. 
 Oil carried in bulk, first steamer, 152. 
 , steamers, 152-60. 
 
 , ,, entrance to shaft tunnel, 159. 
 
 , ,, riveting, 153, 158. 
 
 ,, stability of, 153. 
 ,, stiffening of bulkheads, 157. 
 ,, stresses of, 154. 
 ,, ventilation of, 160. 
 ,, workmanship in, 153. 
 Outside plating, compensation for openings 
 
 in, 228. 
 Oxide, removal of from stesl, 272. 
 
 Packing iron, 227. 
 Panhead rivets, 168. 
 Panting stresses, 72. 
 
 ,, ,, arrangements to resist, 
 
 250. 
 Partial a\vning deck vessel, 126. 
 Partially waterborne vessels, 74. 
 Passenger steamer, strength of light type, 
 
 99. 
 Peak tanks, 163. 
 Petroleum, expansion of, 152, 154. 
 
 ,, gas from, 152, 160. 
 
 Phosphorus in steel and iron , 4. 
 Pig iron, 3, 6. 
 Pillars, 35, 202. 
 
 ,, channel and tee bar, 208. 
 heads and feet, 204-208. 
 ,, large round plate, 149, 
 ,, of exceptional strength widely 
 
 spaced, 149. 
 ,, portable, 208. 
 ,, round hollow. 203. 
 Pintles, rudder, 234, 245-246. 
 Plughead rivets, 169. 
 
 Poop, bridge, and forecastle, vessel with, 
 129. 
 ,, front bulkhead, 253. 255. 
 Pressure per frame space, 58. 
 ,, upon bulkhead, 58. 
 ,, ,, tank top, 60. 
 
 Preventives against decay, 271. 
 Priestman's self-trimming steamer, 144. 
 Propeller shaft, cause of fracture, 69. 
 Propulsion by steam, stresses due to, 72. 
 ,, ,, sail, ,, ,, 73. 
 
 Puddling furnace, 10. 
 Pumping, 259-265. 
 Punching of frames, 185, 
 
 QiTAUTER deck, raised, 122. 
 Quarter pillars, 203. 
 
 Racking stresses, 64, 97. 
 Raised quarter deck vessels, 122. 
 Registered breadth, 50. 
 ,, depth, 50. 
 ,, length, 50. 
 Registration societies, value of, to ship- 
 owner, 101. 
 ,, ,, (See also Classifica- 
 
 tion), 26. 
 Reverse frames, 35, 188. 
 
 ,, frames, bending, bevelling and 
 punching of, 188. 
 Rider plate, 39. 
 Rivet tests, 24, 
 
 Rivets and riveting, 167-171, 183. 
 ,, ,, ,, in bar keel, 183. 
 
 ,, ,, ,, rules for, 175. 
 
 ,, ,, ,, spacing of 175-176, 
 
 183-184, 185, 188, 
 234, 245. 
 ,, ,, ,, table of, 175, 
 
 Ropner's trunk steamer, 142-144. 
 Rose or strum box, 261-262. 
 Round iron pillars, 203. 
 Rudders, 74, 238-246. 
 
 ,, pintles, 234, 245-246. 
 
 ,, stoppers, 246. 
 
 ,, unusual design of, 150. 
 
 Sailing ships, 131. 
 
 Scantlings for awning deck vessel, 120. 
 ,, ,, .spar deck vessel,. 120. 
 
 ,, ,, three deck vessel, 114, 120. 
 
 ,, ,, two deck vessel. 111. 
 
 Scarphs in stern frames, 233. 
 Scrieve board, 180. 
 Scuppers, 225. 
 
 Self-trimming steamers, 138, 142, 144. 
 Semi-box beams, 201. 
 Shade deck vessel, 131. 
 Shearing stresses, curve of, 85, 94. 
 Sheer, vessels without, 146. 
 Shell plating, 225-231. 
 
 ,, ,, length of plates in, 227. 
 
 ,, ,, localities of increased thick- 
 
 ness, 226-227. 
 ,, ,, strength aHbrded by, 226. 
 
 ,, ,, structural value of, 41. 
 
 ,, riveting, 226. (See Rivets and 
 riveting). 
 Shelter deck vessels, 127. 
 
 ,, ,, ,, scantlings of, 115. 
 Side bar keel, 184. 
 ,, keelson, 40, 210. 
 ,, stringer, 41. 
 ,, tanks, 164, 165. 
 Siemens steel, 2, 13, 15. 
 
 ,, ,, strength of, 14. 
 
 Silico-spiegel, 16. 
 Silicon in steel, 5. 
 Single deck vessels, 109, 110. 
 Sluice valves, 260. 
 Snap head rivets, 170. 
 Spar deck vessel, 49, 109, 115. 
 
 19 
 
290 
 
 INDEX. 
 
 Spar deck vessel, of excessive strength, 
 117. 
 ,, ,, ,, scantlings of, 120. 
 Spectacles for twin screws, 238. 
 Spiegeleisen, 14. 
 
 Steamers carrying oil in bulk, 152-160. 
 Standard of strength for load lines, 28. 
 Steel and iron, durability of, 270. 
 ,, castings, 16. 
 , , , , tests for, 24. 
 
 ,, decks, 220. 
 ,, ,, doublings to, 223. 
 ,, ,, maintenance of, 223, 275. 
 ,, ,, wood sheathed, 223. 
 ,, for ships' use, composition of, 2, 3, 
 
 13. 
 ,, plates, thickness of, 23. 
 ,, sections, 18. 
 ,, ships, percentage of, 1. 
 , , tests of, 21 . 
 Stern arrangement for twin screws, 236- 
 238. 
 
 ,, frames, 231-8. 
 ,, ,, after post dispensed with, 
 
 234-5. 
 ,, ,, connection to hull, 283-4. 
 
 ,, ,, method of dispensing with, 
 
 146, 150. 
 ,, ,, scarphs in, 233. 
 
 "Strain," 75. 
 
 ,, in dry dock, 64. 
 
 Strength, continuity of, 246-7. 
 
 ,, of vessels. Board of Trade in- 
 
 structions for com- 
 parison of, 105, 125. 
 ,, in vessels of special proportions, 
 
 105. 
 Strengthening under bow, 72. 
 Stress, 19, 75, 78. 
 
 ,, calculation, value of, 97, 103. 
 ,, comparison of, 101, 103. 
 ,, per square inch, 87, 90, 99. 
 Stresses, illustration of tensile and com- 
 pressive, 80. 
 local, 72, 124. 
 Stringers, 39. 
 Stoppers, rudder, 246. 
 Suction pipes, size of, by Lloyd's rules, 
 
 265. 
 Sulphur in iron and steel, 4. 
 
 Table of riveting in beam knees, 202. 
 
 „ 175. 
 Tanks, cellular double bottom, 161. 
 
 ,, deep midship, 164. 
 
 ,, fore and after peaks, 163. 
 
 ,, M'Intyre double bottom, 160. 
 
 ,, side (margin plate), 161. 
 Tap rivets, 170. 
 Tee bar pillars, 208. 
 Telescopic mast, 250. 
 Templates, 231. 
 
 Tempering steel, 16. 
 
 Three deck vessels, 49, 108, 113. 
 
 ,, ,, ,, scantlings of, 115, 120. 
 
 Top mast, lowering of, 250. 
 Tensile strength, 19. 
 Tests for rivets, 24. 
 
 ,, for iron and steel, 21. 
 ,, for steel castings, 24. 
 Transverse bulkheads, function of, 213. 
 ,, framing, erection of, 191. 
 
 33, 34, 185-194. 
 ,, ,, modification in depth, 
 
 97. 
 ,, ,, unusually wide, spac- 
 
 ing of, 146, 148. 
 ,, stresses, 52. 
 
 Treatment of plates and bars in shipyard, 
 
 25. 
 "Trunk" steamers, 142. 
 
 ,, ., for coal cargoes, 144. 
 
 ,, ,, ,, rice ,, 144. 
 
 Tunnel, construction of, 255. 
 "Turret" steamers, rules for freeboard 
 of, 142. 
 ,, ,, reserve buoyancy of, 
 
 142. 
 ,, ,, structural features 
 
 of, 138. 
 Twin screws, stern arrangement for, 236-8. 
 Two deck vessels, 109, 110. 
 
 ,, ,, ,, scantlings of, 111. 
 
 'Tween deck bulkheads, 64. 
 Types of vessels, 107, 131, 
 
 ,, ,, modifications of, 107. 
 
 ,, ,, determination of, 108. 
 
 Unclassed vessels, 31. 
 Ultimate strength, 77. 
 
 Valve chest, 264. 
 Ventilators, 257-9. 
 Ventilation, 256-9, 284. 
 Vibration, 73, 247. 
 
 Vessels with poop bridge and forecastle, 
 129. 
 
 Wash plates, 40, 164, 211. 
 Water ballast arrangements, 160. 
 
 ,, pressures, 52. 
 Watertight doors, 220. 
 Web frames in lieu of hold beams, 45, 193. 
 Well deck vessels, 128. 
 Whaleback steamer, 138. 
 Weight, disposition of, 68. 
 Working stress, 78, 104. 
 Wrought iron, 3, 10. 
 Wood decks. 224. 
 
 ,, durability of, 270. 
 
 Z frames, 140, 192. 
 
 FEINTED BY NEILL AND COMPANY, LIMITED, EDINBURGH. 
 
 n::>