THE UNIVERSITY OF ILLINOIS LIBRARY P54c ■ Return this book on or before the Latest Date stamped below. University of Illinois Library HAR09 2 i / 006 L161— H41 'ft THE COMPOSITION AND STRENGTH OF MORTARS REPORT ON THE RESULTS OF THE EXPERIMENTAL INVESTIGATION CONDUCTED FOR THE SCIENCE STANDING COMMITTEE OF THE ROYAL INSTITUTE OF BRITISH ARCHITECTS BY W. J. DIBDIN, F.I.C., F.C.S., etc. LONDON THE ROYAL INSTITUTE OF BRITISH ARCHITECTS 9 Conduit Street, Regent Street, W. 1911 Price Five SliiJIings PBINTED FOR THE UNIVERSITY OF LONDON PRESS, LTD., BY RICHARD CLAY AND SONS, LIMITED, LONDON AND BUNGAY CONTENTS SECTION • PAGE I Introduction . . 5 II Examination of Limes and Sands, etc., used . . . .14 III Diagrammatic Eesults 15 IV Eelation of Grading of Aggregate and the Presence of Clay to Strength of Mortar .21 ^ V Maximum Eesults obtained with Different Limes and Sands . 24 VI Eatios of Tensile to Crushing Strengths .... 26 VII Examination of Ancient Mortars ...... 27 3 VIII Comparison of Proportions of Lime to Aggregate in Ancient ^3 AND Experimental Mortars in Present Series ... 28 ^ IX Brick Tests for Adhesive Strength 29 X Trass 29 ^ XI Summarised Conclusions 36 ^ XII Series A. Tables and Diagrams. White Chalk Lime, etc. . 38 ^ XIII ,, B. ,, „ Dorking Greystone Lime, ETC. . . . . 41 t XIV „ C. ,, ,, Blue Lias Lime, etc. . 44 XV E). Effect of Clay in Aggre- gate .... 47 XVI Gr- i> 5» Comparison of Effect of X Clay in Aggregate . 50 ^ XVII „ E, F. „ Brick Adhesive Tests . 54 X iii 350539 SECTION I INTRODUCTION In accordance with the Scheme of Procedure adopted by the Science Standing Committee of the Royal Institute of British Architects on November 14, 1907, which I attach hereto, signed by the Chairman Mr. Lewis Solomon, the extensive series of experiments detailed therein has been completed, and I now have the honour of submitting for the considera- tion of the Committee the results as set out in the following tabulated statements, and observations on the conclusions deduced therefrom. For greater facility in comparison diagrams have been prepared of all the principal factors, and these accompany the tables. The work involved — 1. The collection and examination of representative samples of materials, viz. limes, sands and bricks. 2. The preparation from these of — 828 briquettes for tensile strength. 828 blocks for crushing „ 120 pairs of bricks for adhesive strength, making 1776 experiments in all. 3. The design and construction of special modifications to the compound lever cement testing machine for the purpose of effecting the crushing and adhesive tests. The method previously employed permitted only a pressure of 500 lb. per square iuch for the crushing strength, which was ample for ordinary mortars but insufficient for the present purpose, and by the kind- ness of Mr. G. E. Moore, M.Inst.Mech.E., who designed and constructed the exceedingly convenient and effective modifications in the machine described at length in the note attached, all difficulties on this point were overcome. The adhesive test was also facilitated by Mr. Moore's assistance — and his construction of the necessary adjunct to the same machine. At first, as reported to your Committee on July 27, 1908, Prof. J. D. Cormack, of University College, London, kindly granted the use of the large testing apparatus in his laboratory for the purpose of the experiments, but it was at once found on trial that the machine was unsuitable in con- sequence of its size and power, and it became necessary to design a more simple method, and I desire to acknowledge my indebtedness to Mr. Moore for his invaluable aid. 4. As an auxiliary to the foregoing a series of mortars from ancient buildings has been kindly forwarded to me for analysis by members of your Committee, who thereby have enabled me to place on record the results of the chemical and physical examination of a number of authenticated samples of representative mortar of variable qualities, which, having stood the test 6 COMPOSITION AND STRENGTH OF MORTARS of time, supply a valuable means of comparison with the series specially prepared and tested in connection with the present investigation. 5. The tabulation of all results, preparation of diagrams, and discussion of the experimental factors. 6. After the tests sanctioned by the Committee had been commenced it was contemplated to add a further series in connection with trass and pozzuolana; but it was soon realised that the work already in hand rendered it impossible, under the existing circumstances, to extend it at that time. Fortunately, however, this was a matter of minor importance in view of the excellent series of tests with trass that had been undertaken in Germany by government engineers, in connection with public works ; which tests, by the courtesy of Mr. William Challoner of Blackpool, I am enabled to add to this Report. The use of trass, a volcanic calcined clay, is an important industry on the Continent ; the output from the Tuffstein quarries at Krufft and at Kreutz, both in the Nette Valle near Andernach on the Rhine, is stated to be over 200,000 tons per annum, being used for German, Belgian, French, and Dutch engineering and building works. This large application is principally owing to the result of research work and experiments carried out by the Commission on Building Materials which has existed for the past twenty-five years under State subsidy in connection with the Government Experimental and Testing Institute at Charlottenburg. In consequence of the information thus placed at their disposal, the German State Railway and other departments have recently adopted standard specifications for mortars and concretes, which have been extracted from a paper on " The Use of Trass for Mortars and Concretes in Modern Building Construction," read at the sectional meeting of the National Association of Technical Engineers of Germany, by M. J. Mund, Cologne, in 1908. The samples of chalk lime, and of sand, were kindly supplied by Messrs. Hall & Co. Ltd., Croydon, and the blue lias lime was forwarded by Messrs. Chas. Nelson & Co. from their works at Stockton, Rugby. The stock bricks were supplied by Mr. W. Willett from the Brickfields, Langley, Bucks, and the Fletton Bricks by the Fletton Crown and Crowhurst Brick Co. Ltd. To all these firms my acknowledgment is due and hereby tendered for the facilities thus courteously afforded me. The ground grey Kent stock brick, as commonly used in London buildings, and red London clay were kindly arranged for by Mr. H. D. Searles-Wood, a member of your Committee. I have to acknowledge the assistance rendered in connection with this investigation by my son, Mr. F. J. A. Dibdin, who was largely engaged in con- nection with the former investigation, the results of which I had the honour to submit to the Royal Institute in December 1906, and who thus commenced the present work with no little experience. With the help of Mr. Leonard Cooper, F.C.S., etc., and Mr. Wilfred Corry,Mr. F. J. A. Dibdin prepared the whole of the testing briquettes, blocks, etc., and also assisted in connection with the alterations to the testing machine, as well as with the breaking and crushing tests at the respective periods. COMPOSITION AND STRENGTH OF MORTARS 7 8 COMPOSITION AND STRENGTH OF MORTARS CO O i o 1^ o o Xfl o CO o o o Sand in CO a> © 5S o I CD ^ ^ COMPOSITION AND STRENGTH OF MORTARS 9 10 COMPOSITION AND STRENGTH OF MORTARS to I ! o o O O 00 O C TT;^^ • >> j> -sV • 1 " 0-8 5-4 11-8 6-4 1 " 24-5 28-8 53-2 13-6 13-7 20-3 12-8 11-6 Passed yV" • 60-2 10-2 3-6 37-0 100 0 100-0 100-0 100-0 Passed* ^h" % 0-0 98*4 59-3 69-6 62-2 J/' % 0-0 73-9 30-5 16-4 48-6 tV" % 0-0 60-2 10-2 3-6 37-0 All samples free from clay. * See Section IV, p. 21. SECTION III DIAGRAMMATIC RESULTS A STUDY of the tabulated results as set out in the diagrams brings into view in a convenient form many points of interest well worth careful examination, a noticeable one being the decrease in strength of certain mortars, particularly those made with white chalk lime, after drying by storage on the shelves of the testing room in the basement of the building, which consequently was at a fairly constant temperature throughout the year. At first the fact that a mortar decreased in strength after three or twelve months seemed to throw some doubt upon the accuracy of the work, but a more careful comparison of the factors in view of the above con- sideration clearly points to the fact that a well-set mortar under certain conditions loses a percentage of its strength on drying. It will be noticed that this phenomenon applies almost entirely to those of an inferior quality. SERIES A. WHITE CHALK LIME Tensile Strengths Diagram Aa. — With Standard Sand the best results were obtained with proportions of 1 to 4, the two years' test giving 65 lb. per sq. in., all the tests above one month giving close agreement. Diagram Ah. — With Charlton Fine Sand the best two years' result, viz. 60 lb., was obtained with ratio of 1 to 4, proportions 1 to 5 falling to 32 lb. In this case the strength of the twelve months' samples was markedly higher, rising to 86 and even to 92 in the case of proportions 1 to 2, and it is only reasonable to assume that as the briquettes dried they lost the " grip " due to moisture. Diagram Ac. — With Pit Sand proportions 1 to 2 gave distinctly the best results, viz. 73 lb. in two years, the twelve months' test having been as high as 89. There was little practical difference between the other proportions. Diagram Ad. — With Thames Sand proportions 1 to 2 in two years gave 100, this being only 1 lb. less than the twelve months. As with pit sand, other proportions gave lower and fairly equal results. Diagram Ae. — With Ground Brick the best results were obtained with proportions 1 to 2 and 1 to 5, viz. in two years, 60 and 52 lb. respectively. 15 16 COMPOSITION AND STRENGTH OF MORTARS Crushing Strengths Diagram Aa. — With Standard Sand the maximum result in two years was obtained with 1 to 3, viz. 240 lb. per cu. in. ; proportions 1 to 2 and 1 to 4 gave nearly as good, but 1 to 5 gave only 45 lb. In the case of 1 to 3 the crushing strength increased up to 308 in twelve months and then fell to 240 in the dry mortar, i. e. in two years. Diagram Ah'. — With Charlton Fine Sand proportion 1 to 2 gave a strength of 300 lb. in two years, which steadily fell with increasing aggregate to 88 lb. with proportion 1 to 5. Diagram Ac'. — With Pit Sand proportions of 1 to 2 gave 247 lb., and 1 to 3, 253 lb. in two years, and then decreased with increased aggregate. In this diagram the effect of drying is noticeable, the twelve months' tests with 1 to 2 having been as high as 385. Diagram Ad'. — With Thames Sand proportions 1 to 2 gave 297 lb. in two years ; again the effect of drying is well shown, twelve months' tests with 1 to 2 having been as high as 406 lb. Diagram Ae . — With Ground Brick 312 lb. were obtained in two years with proportions 1 to 3, but this figure steadily fell to 152 with 1 to 5. SERIES B. DORKING GREYSTONE LIME Tensile Strengths Diagram Ba. — With Standard Sand far more consistent results are seen than in Series A in regard to the time factor. With proportions 1 to 3 the two years' test gave the maximum result, viz. 103 lb. per sq. in., a result, however, but little better than that obtained with 1 to 2, viz. 97 lb., proportions 1 to 4 and 1 to 5 fell off to 32 and 27 respectively, and it is of interest to note that with these latter " bad mortars," as they may be termed, the partially dried samples gave better results than the two years' tests. Diagram Bb. — With Charlton Fine Sand the results were all poor, only 50 lb. being obtained wjth proportions 1 to 2 in two years. Diagram Be. — With Pit Sand 75 lb. were obtained in two years with pro- portions 1 to 3. Here it is again very noticeable that the time sequence is in order with the best results, but erratic with the lower qualities. Diagram Bd. — With Thames Sand the best result (again in time sequence) was with proportions 1 to 2, viz. 88 lb. per sq. in. in two years. The effect of the moisture in this diagram is marked, with proportions 1 to 4 it might have been concluded from a three months' test that a strength of 95 lb. was attainable. Diagram Be. — With Ground Brick the results were all poor and erratic, the maximum result in two years being obtained with proportions 1 to 4, viz. 48 lb., although the moist briquettes tested at twelve months gave 60 lb. COMPOSITION AND STRENGTH OF MORTARS 17 Crushing Strengths Diagram Ba. — With Standard Sand 257 lb. per cu. in. were obtained in two years with proportions 1 to 2, the results steadily falling to 48 with 1 to 5, the " false result " due to the moisture in the twelve months' samples being very marked. Diagram Bh'. — With Charlton Fine Sand only 180 lb. per cu. in. were obtained in two years with proportions 1 to 5. Diagram Bc\ — With Pit Sand 333 lb. per cu. in. were obtained in two years with proportions 1 to 2, proportions 1 to 3, etc., giving distinctly inferior results. Diagram Bd', — With Thames Sand 243 lb. per cu. in. were obtained in two years with proportions 1 to 2. It is very interesting to note the " false result " given by the three months' test with proportions 1 to 3, viz. 430 lb. In this case evidently the effect of the capillarity of the water was acting under the most favourable conditions. Diagram Be'. — With Grottnd Brick the hest results, viz. 162 lb., were obtained with proportions 1 to 3, the time sequence being in order, but erratic in all other cases. SERIES C. BLUE LIAS LIME Tensile Strengths Diagram Ca. — With Standard Sand proportions 1 to 2 decidedly gave the best results, viz. 58 lb. per sq. in. in two years, falling to only 11 lb. with 1 to 4. Diagram Cb. — With Charlton Fine Sand, again, proportions 1 to 2 gave the best, viz. 40 lb. in two years, and the results fell off to about one half with proportions 1 to 3, etc. Diagram Cc. — With Fit Sand a marked improvement was effected, the time sequence being in order and 75 and 77 lb. being obtained with pro- portions 1 to 2 and 1 to 3 respectively in two years. With increase of aggregate the strength showed a marked diminution and the time sequence varied considerably. Diagram Cd. — With Thames Sand the proportions 1 to 3 gave 102 lb. per sq. in. in two years, the time sequence being in order. The deleterious effect of alteration of proportions either way is very marked, the strengths being lower and the time sequence erratic. Diagram Ce. — With Ground Brick the best tensile strengths were obtained, viz. 133 lb. in two years with proportions 1 to 2, the time sequence being perfect. With increased aggregate the results fell to 87 lb., with perfect time sequence, but after that proportions 1 to 4 and 1 to 5 gave variable results, the danger of accepting a one month test with 1 to 5 being distinctly indicated. 18 COMPOSITION AND STRENGTH OF MORTARS Crushing Strengths Diagram Ca. — With Standard Sand the two years' test gave 538 lb, per cu. in. with proportions 1 to 2. It is interesting to note that if the well- known proportions of 1 to 3 had been solely tested this mortar would have been dismissed as having a crushing strength of only 188. Here, clearly, the magic 1 to 3 is largely at fault. In the case of proportions 1 to 2 it will be noticed the time sequence is perfect. Diagram Ch\ — With Charlton Fine Sand the results were inferior, only 257 lb. being obtained in two years with proportions 1 to 2, and the twelve months' test gave higher results. Increased aggregate only made matters worse. Diagram Cc. — With Pit Sand a very different result was at once seen. All the proportions gave regular time sequences, and 1 to 2 gave 605 lb., 1 to 3 gave 650 lb., 1 to 4 gave 548 lb., but 1 to 5 gave only 303 lb. per cu. in. Diagram Gd'. — With Thames Sand the results were still better, viz. 785 lb. in two years for proportions 1 to 2, the time sequence being perfect. Increased aggregate only decreased strength. Diagram Ce. — With Ground Brick 910 lb. were obtained with proportions 1 to 2 in two years, the time sequence being excellent. As before, increased aggregate meant decreased strength. SERIES D. ADDITION of 5 PER CENT. CLAY to AGGREGATE all proportions 1 to 3 Tensile Strength Diagram Da. — White Chalk Lime with standard sand gave 50 lb. in two years and with pit sand 57 lb., but with Charlton only 35 lb. per sq. in. The twelve months' tests gave much higher results, as might be expected from the retention of ^noisture by the clay. Diagram Dh. — Dorking Grey stone Lime with pit sand gave the best of this set, viz. 67 lb. in two years, the Charlton fine sand giving only 20 lb. Diagram Dc. — Blue Lias Lime and standard sand gave 168 lb. in two years, and with pit sand 121 lb., whilst the best obtained with Charlton fine sand was only 34. Crushing Strength Diagram Da. — Wliite Chalk Lime and standard sand gave the best results, viz. 230 lb. per cu. in. in two years, and Charlton fine sand gave the lowest, viz. 70 lb. Diagram DV . — Dorking Greystone Lime also gave poor results ; the best being 177 lb. in two years with standard sand and only 82 with Charlton fine. COMPOSITION AND STRENGTH OF MORTARS 19 Diagram Dc\ — Blue Lias Lime and standard sand gave 876 lb. in two years. With pit sand, 550 lb. were obtained, but only 79 lb. with Charlton fine. SERIES G. COMPARISON OF RESULTS WITH AND WITHOUT CLAY white chalk lime, 1 to 3 Tensile Stkength Diagram Ga. — Without Clay the best result in two years was obtained with Charlton fine sand, viz. 55 lb. per sq. in. With Clay standard sand gave 50 lb. in two years, pit sand 57 lb. Charlton fine sand giving only 35 lb. in two years respectively, the time sequence being variable as in all low results, the twelve months' tests being much higher. DORKING GREYSTONE LIME Tensile Strength Diagram Gh. — Without Clay standard sand gave 103 lb. in two years, pit sand 75 lb., and Charlton fine sand only 35 lb. per sq. in. With Clay pit sand gave 67 lb., standard sand 48 lb., and Charlton fine only 20 lb. per sq. in. in two years, the twelve months' tests giving higher results in each case. blue lias lime Tensile Strength Diagram Gc. — Without Clay pit sand gave 77 lb. in two years, standard sand 38 lb., and Charlton fine only 26 lb. per sq. in. With Clay standard sand gave 168 lb., pit sand 121 lb., and Charlton fine sand 33 lb. per sq. in. white chalk lime Crushing Strength Diagram Ga. — Without Clay fine Charlton sand gave in two years 260 lb. per per cu. in. With Clay fine Charlton sand gave only 70 lb. per cu. in., whilst standard sand gave 230 lb. and pit sand 163 lb. per cu. in. Thus clay with white chalk lime in proportions of 1 to 3 is injurious. dorking greystone lime Crushing Strength Diagram GV . — Without Clay gave a maximum in two years of 228 lb. with standard sand, against Charlton fine sand 140 lb., and 200 lb. per cu. in. for pit sand. B 2 20 COMPOSITION AND STRENGTH OF MORTARS With Clay standard sand gave 177 lb. in two years, fine Charlton sand only 82 lb., and pit sand 140 lb. per cu. in. BLUE LIAS LIME Diagram Gc. — Without Clay pit sand gave 650 lb. in two years, the results being in sequence of time, Charlton fine sand only 156 lb., and standard sand 188 lb. per cu. in. With Clay standard sand gave 876 lb. in two years, Charlton fine sand 79 lb., and pit sand 550 lb. Brick Tests for Adhesion stock bricks Diagram E. All mortars 1 to DORKING GREYSTONE LIME Ground Brick gave an adhesive strength of 232 lb. in two years, being nearly double the strength obtained with Thames sand, four times that for pit sand, fine sand and standard sand. Fletton Bricks DORKING GREYSTONE LIME 1 to 3. Ground Brick gave the maximum, viz. 347 lb., the results steadily falling for Thames sand 293, pit sand 200, Charlton fine sand 87, and standard sand 122 lb. The Fletton bricks consistently gave a better bond than stock bricks. Note. — In the course of testing for adhesive strength it was noticed that in those cases where the greatest strength was obtained, the mortar, instead of leaving the brick, broke in itself, portions of mortar adhering to both bricks. 21 SECTION lY RELATION OF GRADING OF AGGREGATE AND PRESENCE OF CLAY TO STRENGTH OF MORTAR The effect of fine material in the aggregate is brought out very clearly in these experiments, especially in those in which 5 ^ of clay was added to the aggregate. Placing the aggregate in the order of increasing percentage of finer particles (see Table II, p. 14), we have — TABLE III Passed -^V in. mesh. Per cent. Standard sand .... O'O Pit sand .... 59-8 Ground brick . . . .62*2 Thames sand . . . . 69'6 Charlton fine sand . . . 98'4 Passed -gV in. mesh. Standard sand .... 0"0 Thames „ . . . .16-4 Pit „ . . . . 30-5 Ground brick . . . . 48*6 Charlton fine sand . . . 73*9 Passed yV in. mesh. Standard sand . . . . 0"0 Thames „ .... 3*6 Pit „ . . . . 10-2 Ground brick .... ST'O Charlton fine sand . . . 60 2 On tabulating the tests made with lime and aggregate in the ratio of 1 to 2 and 1 to 3, for example, we have the following series — 22 COMPOSITION AND STRENGTH OF MORTARS TABLE IV Results of Two Years' Tests. Proportion of Lime to Sand. 1 to 2. 1 to 3. Tensile. Crushing. Tensile. Crushing. Without Cloy. V> XliLc V^IldiK dllU O tclllLlcll U. OtlllU. 43 218 53 240 , , , , Fine Charlton Sand Oo OUU 00 Pit Sand .... 73 247 50 253 ,, Thames Sand 100 297 68 202 Ground Brick 60 213 40 312 97 257 103 228 Fine Charlton Sand . 50 143 1 35 140 Pit Sand . 62 333 1 75 200 Thames Sand 88 243 j 58 228 Ground Brick 38 123 1 27 162 Blue Lias and Standard Sand .... 58 538 j 38 188 ,, line Charlton Sand . 40 257 26 156 ,, ,, Pit Sand 75 605 i 77 650 Thames Sand .... 80 785 102 507 Ground Brick .... 133 910 1 87 657 With 5 % Clay. White Chalk and Standard Sand . i 50 230 Fine Charlton Sand ! 35 70 Pit Sand .... j 57 163; Dorking Greystone and Standard Sand . 48 177 ,, Fine Charlton . 1 20 82 Pit Sand . 1 67 140 Blue Lias and Standard Sand .... ; 168 876 ,, Fine Charlton Sand . 34 79 Pit Sand ..... 121 550 From this table it will be seen that — 1. The Fine Charlton Sand gave good results with white chalk lime, but with lias lime it made a very indifferent mortar ; and when clay was added it gave the lowest results. 2. Ground Brick, the next aggregate in order of fineness, gave good results with white chalk lime but inferior with Dorking greystone lime. With blue lias lime, however, it gave the second best results in tensile and the best results in crushing strength. Unfortunately, from the experimental stand- point, the test was not repeated with clay as, obviously, it is hardly likely that ground brick would contain unburnt clay. 3. Pit Sand, the third aggregate in order of fineness, gave good results with white chalk lime and Dorking greystone lime ; with blue lias lime it gave excellent results, viz. tensile 77, crushing 650. With the addition of clay in the white chalk lime tests, the crushing strength fell off, a similar result being obtained with Dorking greystone lime, the tensile strength in each case COMPOSITION AND STRENGTH OF MORTAES 23 being increased. When blue lias lime, however, was used, the tensile strength was increased from 77 to 121, but the crushing fell off from 650 to 550 lb. per cu. in. 4. Thames Sand, — As might be expected, the tests with this sand, although showing certain variations, on the whole ran fairly well with those made with pit sand, especially if we have regard to the experimental error unavoidable in work of this nature. 5. Standard Sand. — This, the coarsest of the aggregates, showed fair average results with white chalk lime, and twice the tensile strength with Dorking greystone lime. With blue lias lime the results were decidedly poor with proportions of 1 to 3, thus showing that evidently the voids were not com- pletely filled. With 1 to 2 far better results were obtained. Where a small quantity of clay was added to the sand the results were excellent, viz. 168 for tensile strength and 876 for crushing, against 38 and 188 respectively for the 1 to 2 tests, indicating that with such a coarse sand the addition of clay was distinctly advantageous as against its decidedly deteriorating effect when such a fine sand as that from Charlton was employed. Perhaps this point is one of the most interesting and valuable which has been brought out in the investigation. 24 COMPOSITION AND STRENGTH OF MORTARS SECTION V MAXIMUM RESULTS SHOWING VARIATIONS IN COMPOSITION OF MORTAR WHICH MAY BE EMPLOYED WITH DIFFERENT QUALITIES OF MATRIX (LIME) AND AGGREGATE (SAND) Results of Tests at End of Two Years TABLE V WHITE CHALK LIME with— Tensile. Crushing, standard Sand 1:3 53 lb. 240 lb. 1:4 65 „ 223 „ Average 1:31 59 lb. 231 lb. Fine Sand (Charlton) 1:2 53 lb. 300 lb. {Note. — 1 : 4 gave higher tensile, viz. 60, but lower crushing, viz. 217.) Pit Sand (ordinsirj) 1:2 "73 lb. 247 lb. 1:3 50 „ 253 „ Averao^e 1 : 21 61 lb. 250 lb. Thames Sand 1:2 100 lb. 297 lb. Ground Brich 1:2 60 „ 213 „ 1:3 40 312 1 : 4 38 „ 222 Average 1 : 3 46 lb. 249 lb. B. Tensile. Crushing, DORKING GREYSTONE LIME with- Standard Sand 1:2 97 lb. 257 lb. 1 : 3 103 „ 228 Average 1 : ^ 100 lb. 242 lb. Fine Sand (Charlton) 1:2 50 lb. 143 lb. 1:5 27 „ 180 „ {Note. — In this case the tensile and crushing strengths reversed, 1 : 2 giving the highest ten- sile and 1 : 5 the highest crushing strength. ) COMPOSITION AND STRENGTH OF MORTARS 25 Pit Sand (ordinary) 1 : 2 62 lb. 333 lb. 1 :3 75 „ 200 „ - Average 1 68 lb. 266 lb. Thames Sand 1 2 88 lb. 243 lb. {Note. — The crushing s trength 1 : 3 was 430 lb. ill three months, 323 lb in twelve months, but only 228 lb. in two years, thus showing an apparent remarkable deterioration due to drying.) Ground Brick 1 : 0 Z d8 lb. 123 lb. 1 3 27 „ 162 „ 1 4 48 „ 130 „ 1 : 5 38 142 „ Average 1 35 lb. 139 lb. C. BLUE LIAS LIME with— Tensile. Crushing. Standard Sand 1 : 2 58 lb. 538 lb. Fine Sand (Charlton) 1 :2 40 „ 257 „ Pit Sand (ordinary) 1 :3 77 „ 650 „ Thames Sand 1 :2 80 „ 785 „ 1 :3 102 „ 507 „ Average I 21 2 91 lb. 646 lb. Ground Brick I 2 133 lb. 910 lb. D. Tensile. Crushing. WHITE CHALK LIME + 5 Bed London Clay with — Standard Sand 1 : 3° 50 lb. 230 lb. DORKING GREYSTONE LIME + 5%i^gc^ London Clay with Standard Sand 1 : 3 48 lb. 177 „ Pit Sand I : 3 67 „ 140 „ BLUE LIAS LIME + 5 y Red London Clay with — Standard Sand 1 :3 168 „ 876 „ This table is of special interest as indicating at a glance the proportions of matrix and aggregate which should be used with any given materials of those experimented with, to yield the maximum results. For instance, if a Dorking greystone lime is employed, it will be seen that ordinary pit sand in the ratio of 1 to 2 will yield the highest crushing strength, viz. 333 lb. per cu. in., but if a blue lias lime is employed, ground brick, in the ratio of 1 to 2, will yield a crushing strength of no less than 910 lb. per cu. in., with a tensile strength of 133 lb. per sq. in.; but if a higher tensile strength is required, standard sand with 5 per cent, clay will give 168 lb. per sq. in. with a crushing strength of 876 ib. per cu. in. SECTION VI RATIOS OF TENSILE TO CRUSHING STRENGTHS TABLE VI Crushing ^ Tensile ~~ ^ Ratios of tests which gave maximum tensile strength thus, S'O. crushing strength thus, 7'8. ,, ,, ,, ,, in both tensile and crushing strengths thus, 6 '8. Average of Two Years' Tests. X = Standard Charlton Pit Thames Ground Sand. Sand. Sand. Sand. Brick. "White Chalk Lime 1 2 . 5-1 3-8 3-2 3-0 3-5 >) )> 1 3 . 4-5 4-7 5-1 2-9 7*8 5' >> 1 4 3-4 3-1 3-2 4-0 5-8 > J >> 1 5 . 1-7 2-8 3-5 3-9 2-9 Dorking Greystone Lime 1 2 . 2-6 2-9 5-4 2-8 3-2 5» M »5 1 3 . 2 '2 4-0 2-7 3-9 6-0 J5 J» 1 : 4 . 2-9 4-1 3 0 2-3 2-7 »> > > J ) 1 : 5 . 1-8 6-7 3-7 3-7 3-7 Blue Lias Lime 1 : 2 . 9-3 6-4 8-1 10-2 6 8 5 5 5 > 1 3 . 4-9 7-1 8-5 5-0 7-6 5) >> 1 4 - 13-6 6-7 9-1 5-7 6-0 5 5 5 5 1 5 . 10-4 5-8 7-2 6-1 6-6 With 5% C/a?/. White Chalk Lime 1 : 3 . 46 2-0 3-4 Dorking Greystone Lime 1 3 . 37 4-1 2-1 Blue Lias Lime 1 3 . 52 2-3 4-5 On abstracting those ratios which correspond to maximum results as shown in the tables we have the following : — TABLE VII Maximum Tensile Strengths White Chalk Lime to Thames Sand Dorking Greystone Lime to Standard Sand Blue Lias Lime to Ground Brick With 5 % Clay in aggregate. White Chalk Lime to Standard Sand . Dorking Greystone Lime to Pit Sand . Blue Lias Lime to Standard Sand Crushing Strength. White Chalk Lime to Ground Brick . Dorking Greystone Lime to Pit Sand . Blue Lias Lime to Ground Brick Witli 5^ Clay in Aggregate. White Chalk Lime to Standard Sand . Dorking Greystone Lime to Standard Sand . Blue Lias Lime to Standard Sand 26 Ratio X — 1 : 2 30 1 : 3 2-2 1 : 2 6-8 1 : 3 4-6 1 : 3 21 1 : 3 5-2 1 : 8 7'8 1 : 2 5-4 1 : 2 6-8 1 : 3 4-6 I : 3 3-7 1 : 3 5-2 27 SECTION VII THE EXAMINATION OF ANCIENT MORTARS In accordance with a suggestion made at a meeting of the Royal Institute of British Architects, some of the members have been good enough to collect a series of authenticated old mortars and to forward these for the purpose of analysis and physical examination. The results of these examinations are set out in the attached table (VIIL). The samples are twenty-nine in number and form a good representative set. The essential features brought out by the specially designed system of tests applied, in which particular attention has been paid to aggregate, may shortly be summarised as follows : — 1. That there is no evidence of the formation of soluble silica, the percentage of soluble silica in the best mortars being quite normal for modern mortars made from similar materials. Where larger quantities have been found, the character of the mortar leads to the belief that trass or pozzuolana, etc., had been employed. 2. That the presence of small quantities of clean ferruginous clay in the aggregate has certainly had no deleterious effect on the mortar, as such samples were amongst the best. 8. That the proportion of lime to sand and grit is usually much greater than 1 to 3. For instance, in the samples from Allington Castle, early thirteenth century, the relative volumes in three samples were 1 to 1'7, 1 to I'l and 1 to 1*9, the observations of Mr. W. D. Caroe, who kindly forwarded them, being respectively, " Very good indeed " — " Not so good as ' A,' but fair," and " In fair condition." In reference to Point 2 it may be remarked that the percentage of clean ferruginous clay in the sand in the above instances was respectively 8*6 per cent., 3*6G per cent, and 4*0 per cent. Mr. Caroe reported that the A sample is one of the best mediaeval mortars he had ever come across, second only to the thirteenth-century mortar at Corfe Castle which successfully resisted the gunpowder of the Cromwellian period. 4. Inferior mortars are at once detected by the examination. For instance, the sample from Painters' Hall, built under Sir Christopher Wren after the Fire of London, contained 15 "4 per cent, of earthy matter in the aggregate, which was largely composed of broken red brick, organic debris, carbon, etc., the proportion of lime to aggregate being 1 to 0*53. Under the circumstances it is surprising that a crushing strength of 110 lb. per cu. in. was obtained. 28 COMPOSITION AND STRENGTH OF MORTARS Another remarkable sample of bad 500-years-old mortar is one from the Guild Hall, Norwich, which contained coke and coal, the proportions of lime to aggregate being 3 to 1. It is not surprising that the city engineer reported that "the condition of the mortar was very bad, having slight adhesive powers, and much of the wall was in such weak condition as to enable it to be pulled down by the fingers without the use of tools." It may be noted that, whenever possible, samples of mortar should be collected in such a manner as to enable pieces of 1 cu. in. to be selected or cut out from larger pieces for the crushing test, which cannot be applied to the broken small stuff. SECTION VIII COMPARISON OF PROPORTIONS OF LIME TO AGGREGATE IN ANCIENT MORTARS AND THOSE WHICH HAVE GIVEN THE BEST RESULTS IN THE PRESENT INVESTIGATION Diagram illustrating Table VIII shows very clearly the relative proportions of lime to aggregate in various old mortars, some of them being of the highest quality, in comparison with the aggregate in those mortars which, under the two years' test, have given the best results in the course of the present investigation. From this it will be seen that the greater number of the old samples varied between one and two volumes of aggregate for one of lime, it being remembered always that the calculation is made on the assumption that greystone lime of 80 % of CaO and weighing 40 lb. to the cu. ft. was employed. The aggregate in the experimental mortars which gave the best results in two years is seen to vary from 2 to 3, the average of the whole being (excluding ground brick tests) 2*27. ■ From these facts it seems that there is good reason for reconsideration of the modern theoretical hard and fast requirement of proportions of 1 to 3 irrespective of the nature of the lime and aggregate. LIME = I TO /! ARS TARS IN PRESENT i IS. TWO YEARS TESTS! RY 28 COMPOSITION AND STRENGTH OF MORTARS Another remarkable sample of bad 500-years-old mortar is one from the Guild Hall, Norwich, which contained coke and coal, the proportions of lime to aggregate being 3 to 1. It is not surprising that the city engineer reported that "the condition of the mortar was very bad, having slight adhesive powers, and much of the wall was in such weak condition as to enable it to be pulled down by the fingers without the use of tools." It may be noted that, whenever possible, samples of mortar should be collected in such a manner as to enable pieces of 1 cu. in. to be selected or cut out from larger pieces for the crushing test, which cannot be applied to the broken small stuff. SECTION VIII COMPARISON OF PROPORTIONS OF LIME TO AGGREGATE IN ANCIENT MORTARS AND THOSE WHICH HAVE GIVEN THE BEST RESULTS IN THE PRESENT INVESTIGATION Diagram illustrating Table VIII shows very clearly the relative proportions of lime to aggregate in various old mortars, some of them being of the highest quality, in comparison with the aggregate in those mortars which, under the two years' test, have given the best results in the course of the present investigation. From this it will be seen that the greater number of the old samples varied between one and two volumes of aggregate for one of lime, it being remembered always that the calculation is made on the assumption that grej^stone lime of 80 % of CaO and weighing 40 lb. to the cu. ft. was employed. The aggregate in the experimental mortars which gave the best results in two years is seen to vary from 2 to 3, the average of the whole being (excluding ground brick tests) 2'27. From these facts it seems that there is good reason for reconsideration of the modern theoretical hard and fast requirement of proportions of 1 to 3 irrespective of the nature of the lime and aggregate. Table VIII. DIAGRAM SHOWING THE PROPORTION OF AGGREGATE IN VARIOUS MORTARS ANCIENT MORTARS BEST MORTARS IN PRESENT i INVESTIGATIONS. TWO YEARS TESTS I LIME = I TO AGGREGATE = X modern theory 7 1 ■■ i 1 V '! ' ! ted by H. D. Searles-Wood, Esq. T. H. Powell, Esq. 5 e, St., h, Bodiam Castle, ? 1402. (B. Dicksee.) Camber Castle, Henry VIII, 16th century, per Bernard Dicksee. Pharos, Dover. Roman. Studfall Castle. Roman. Nr. Hythe. Roman villa, Darenth. 52-04 0- 45 1- 64 26-10 0-38 16-60 0-19 52-02 2- 45 3- 60 23-16 0-62 13 02 0-38 44-10 0-30 2-70 27-10 5-30 7-80 0-00 70-20 0- 30 1- 40 11-30 0-20 9-20 0-00 55-10 0-90 2-10 19-10 2-90 15-90 0-00 2-32 4-50 12-70 7-40 4-00 99-72 99-75 100-00 100 00 100-00 Fine sand and coarse pebbles isricK TTlinf on /-I pebbles rSUCK 300 5 0 — 3-0 14-0 2-2 — 54-2 - - 4-5 5-7 — — — id 32-5 81-2: 52-0 = 1:0-6 Hard 29-0 72 : 52-0 = 1 :0-7 Very hard S3-9 44-9 : 44-1 = 1:1-0 14-1 35-2: 70-2 = 1:2-0 23-9 31-7 : 55-1 = 1:1-7 115 lb. + 700 lb. 100 lb. 53-0 lb. 100 lb. 53 lb. I \ \ h ■ \ J ' i TABLE VIII.— THE COMPOSITION OF ANCIENT MOKTARS DESOBIPTION CoUoctedbj'H. D. Searloa.Wood, Esq., F.ILI.B.A., etc. Collected by W. D. Cnna, Baq.. P.S.A., etc. Collected by Sydney PerkB, Colletted by H. D. SearloB-Wood. Esq. M.noi»ril Firet noor Park no,i.. Cl" l"" by 2-0 0-2 Slight trace fo Dodtam Castle (B.'Dickaee.) u I " t ■ 1.^',.' r 1 dM s IM A C-<«I,,rj,. 1W» anl„r„. 70-66 2- 96 9-06 0-82 3- 80 0 92 SludfaUCulle Insoluble in ili'ulo HCl ... SoluUo Sili™ Oxide of Iron and Alumina Lime (CaO) llasuosia Cmbonic Aci.l (CO.) Sulphuric Acid (So,) Oignnie Matter and Water of Hydration . 61 -50 0-60 11 25 00 Trace 6012 0-37 16-57 Troco 1300 2-92 1 62-50 38-95 i 0-65 1 -9.5 16-18 29-72 Trace Trace 12-60 22-50 0-20 1-47 71-50 0-30 12-70 10-00 61-50 5-30 9-66 Tr,ice 5-00 0-60 15-92 ^0-70 33-84 0-33 13-04 59-65 17-95 Trace 65-77 1- 20 2- 40 13-20 Trace 10-20 0-31 I 515 60-80 0-80 10-10 Trace 15-00 Trace 68-22 1-00 4-20 12-40 9-70 Trace 0 13 1 4 00 1 60-00 0-42 0-62 19-38 '0-6O 4-55 56 78 69-82 1-14 1-70 18-76 14-04 Nil Nil 12-60 7-00 0-33 0-45 10-00 16-04 "20 2-20 0-39 0-00 2-90 68-00 3-80 2-80 12-50 0-36 0-00 17-50 46-26 2-95 1-00 23-94 0-24 73-58 0-30 0-60 6-40 Trace 5 00 0 00 12-57 / Trace Trace 54-70 Trace Nil 0-35 40-23 400 2-00 16"80 1 s'f 1 51-08 1-14 5-60 18-30 0-4i 7^^ 52-04 26-10 0 38 52-02 'o-3S 0 00 70-20 '0-20 0-90 2-10 19-10 4-00 100 00 10000 100-25 00-24 99-05 100-00 100-00 100-00 99-97 100-45 100 00 99-9 99-68 100-13 100-00 1 99-51 99-75 100 00 Description of Earthy Matter . . ' ,, Grading uf Wmlud Satid, etc. Eetained ou \ inch Mesh . . . per cent Clay, etc. " " " . . '. 5 1 earthy, dirty with tibres of organic iiiat- sweepings Nil 20 4-0 20 K) Reddish hrown,earthy, fairly clean Fairly clean gravelly sand 6-0 120 40-0 Oloanochn Red gravellj 8-0 8-0 36-0 Red gravelly 120 V4 5-51 Black and '"" dobrif ! old rubbish .2-5'7 2-76 2-26 9-65 5-24 1-91 Cloy Brick will] 0-0 12-3 1- 16 Clay Flint, ivith and coal 5-2 2- 0 17-6 0-3 Fine clean Red clay. Clean saml 6-5 29-5 5-6 "clea'i^' 14-5 4- 6 3-5 5- 0 29-5 "clcan^' 7-8 9-8 11-3 Cl ^^1 Fine bnck- 0-0 0-0 0-3 4-6 62-5 2-7 01""' Fine dark- red 0-0 0 0 42-0 3-41 ""red pebbly, with line sand 14-1 10-7 Clean white sand with pebblra 45-5 2-6 1-9 19-9 3-20 Clay few pebbles '3-3 0-93 Clean clay Very fine and 0-0 45 43 The micro, seopieal exam organic mat- S d«bris°wit'i some fungoid nycelium and Clay Clean sand 19-9 5-8 36-4 0-86 Clay ^Clean grey 21 -0 - z Cl ^ ''l 1-84 3-35 17-42 40-23 _ Fine sand and coarse pebbles 30 0 6 0 3-0 14-0 Brick 'Jcbb? Brick - 69 1 61-9 69-4 69-36 62-3 27-2 69 4 65-1 70 7 60-0 j 56-8 60-8 57-9 46-36 75-4 71-1 - - - PercoutJige of Clay, etc. in Orit .... Liine corrected to Connnereial Liuie of 80%CaO Volume of Unslaked Livu to Sand and Grit Physical Character of Mortar .... Kcaetion Crushing Strength per cubic inch 20-5 •47-8 : 61 -5 = 1:1-3 Grey, friable Sliohtlv aKaline Too small z Light, Wable SUghtly Too small 20-71 "vi ^ Slightly alkaline 126 lb. 4-9 20-22 •47-2:62-5 ^^=1:1^-3^ SJightly Too small to test 7I-1V38-95 = 1:0-5 Uncpially nixed, luni|)s """Cd'""" Neutral Nono 110 lb. 76-6 lb. None 15-9 = 1 :l-8 Neutral None Too small 3-1 Neutral None 60 lb. 42-3 Fairly' hard, I'roTb. 22-44 :-,6-l :59-5 ll^d'''''d Neutral None Too small ! 100 lb. . 16-5 38-4 : 65-8 V"ety'ha,'d None 3-66 ss^e*:™ S 94 lb. 15-5 36-2 : 68-2 Neutral None Too small to test 94 lb. 4-5 24-2 60-6 : 60 Very 'hard Neutral 100 lb. 19-5 23-46 F:Urlyhanl Neutr.al 5-7 43^-9 :°69-8 )'ery hard Neutral 194 lb. 100 lb. 10-9 27-2:77-5 = 1:2-8 Fairly hard Neutral 5-6 15-6 39-0:68-0 = 1:1-6 Hard Neutral None Too small 2-0 29-93 = 1:0-6 " Very hai-d Slronply alkahiie Cousidei^ble quantity 100 lb. - Neutral 3-3 = '1:37'' Lird concrete Neutral 164 lb. 98-4 lb. 11-33 = l';2-5 Fairly hard Neutral Too small Weathered 25-6^ Very hard Neutral - grey 4-5 32-6 81-2:52-0 = 1:0-0 Hard 115 lb. 72': 52-0 V^' l" '! 53-0 lb. - 100 lb. •23-9 31-7:55-1 = 1:1-7 53 lb. • Factor used to convert weight of commercial lime to volume = 24. COMPOSITION AND STRENGTH OF MORTARS 29 SECTION IX BRICK TESTS FOR ADHESION Series E F was conducted with the view of ascertaining the adhesive power of certain mortars to two qualities of bricks, viz. Stock and Fletton. The mortars used were made with Dorking greystone lime in all cases, the aggregates used being, respectively, standard sand, fine Charlton sand, pit sand, Thames sand and ground brick, the proportions of lime to aggregate in all cases being 1 to 3. It was intended to obtain tests of the breaking strain at one and three months, as well as at twelve months and two years ; but the strength of the mortar at those periods was too low, in consequence of insufficient setting, to enable reliable results to be obtained, the bricks in many cases coming apart in the process of fixing them in the testing machine. The tests at twelve months and two years are set out in the Table and Diagram E, a, b, and c, and clearly indicate that in the case of the Fletton bricks the adhesive power is considerably greater than with the Stock bricks, the average results at the end of two years, in order of the aggregates given above, being — TABLE IX Standard Charlton Pit Thames Ground Sand. Fine Sand. Sand. Sand. Brick. Stock bricks 55 48 55 127 232 lb. Fletton „ 122 87 200 293 347 „ It is to be noted that the last three series, viz. pit and Thames sand and ground brick, in the case of the Fletton bricks particularly, the mortar adhered so firmly to the brick that it broke in itself instead of leaving the surface of the brick, as happened in all the tests which gave a lower result. It is noteworthy that the ground brick mortar gave the maximum results in both cases. SECTION X TRASS By the courtesy of Mr. William Challoner of Blackpool, I am enabled to add the following results of authoritative tests of mortars in which trass has been employed. It was intended to include a series of such tests, as well as others with pozzuolana, in the investigation, but the detailed work involved in connection with the large number of tests set out in the programme authorised by the committee rendered this impossible; 30 COMPOSITION AND STRENGTH OF MORTARS Fortunately, the work which has been done in Germany in connection with the use of trass, under stringent Government conditions, enables me to submit the following valuable information, and thus fill what would otherwise have been a serious blank. The abstract of the interesting and valuable paper by M. J. Mund of Cologne, in 1908, very fully discusses the question, and gives many details of German practice. The standard specification for mortars and concretes adopted by the Royal Board of Railway Administration, Germany — Herr Kirchner, Engineer and Technical Secretary to the Board — dated July 17, 1910, will be read with interest. It is to be noted that the lime used is a "fat" lime, similar to that obtained from Buxton limestone, and that the sand is Rhine sand, which is the normal sand used in Germany, graded to not more than 40 per cent, of voids. When slacking, the lime is broken into pieces not larger than will pass a 5-inch ring, and the water is added slowly, the lime being constantly agitated, and no part of it allowed to become chilled or lower than 70° F. Mr. Challoner informs me that by this method he has obtained 25 per cent, more " putty " from a ton of lime than by the ordinary English " slutching " process. TABLE X TESTS OF LIME, MORTAR WITH TRASS Tensile Tests, Pounds per Square Inch No. of Days. Made by Herr Helbing, Engineer, Essen, Germany, in connection with the Flood Dams, etc, of the Ems Waterworks, March 30, 1906. Made by Herr A. Unna, Engineer, Cologne. Lime 1 vol. Trass 1 Sand 1 ,, Lime 3 vols. Trass 1 vol. Sand 6 vols. Lime 1 vol. Trass 1 ,, Sand 1 Lime 1 vol. Trass IJ vols. Sand 1 vol. 7 days 14 „ . 28 „ . 3 months 6 „ . . 12 „ . 225 323 370 170 268 400 140 170 240 330 370 380 220 250 300 380 440 455 Compression Tests, pounds per square inch. Lime 1 vol. Trass vols. Saud If Lime 3 vols. Trass 1 vol. Sand 6 vols. 14 days 28 „ . . 100 „ 980 1400 2040 1300 1550 2250 Note. —The numbers in heavy type correspond, in time interval, to those in the committee's series. COMPOSITION AND STRENGTH OF MORTARS 31 TABLE XI TRASS IX LIME AND MORTAR Tests of Tensile Strength made by Mr. William Challoner. Pounds per Square Inch. Time Interval. Lime 1 vol. Lime 1 vol. Lime 1 vol. Lime 1 vol. Lime 2 vols. Trass 1| vols. Trass 1 ,, Trass 1 „ Sand 2 vols. Trass 1 „ Trass 1 vol. Sand 1 vol. Sand 1 Sand 3 vols. Sand 5 vols. 1 Day . 10 25 7 10 2 2 Days 18 45 16 20 8 3 „ ... 60 60 26 30 18 7 ... 223 108 80 95 70 14 „ ... 272 187 126 150 108 28 „ ... 313 350 182 187 138 3 Months . 349 324 225 212 166 6 „ ... 383 370 250 230 189 12 „ ... 2 Years 393 393 268 242 203 405 405 287 248 212 3 „ ... 417 415 292 250 217 Note. — The numbers in heavy type correspond, in time interval, to those in the committee's series. Cologne Bridge, 1847, under water 1 vol. sand, 1 vol. slaked lime, 7 J vols, trass, received from Mr. William Challoner. Per cent. Sand and grit . 48-20 Sol. silica .... . 9-70 Oxide of iron and alumina 7'55 CaO . 14-55 MgO . 0-46 CO2 . 8-06 SO3 ... . . . nil Water of hydration, etc. . . 11-44 99-96 Earthy matter (clay) . 2-45 Clay on sand . 3-1 Character of mortar very hard Too small in quantity for further tests. COMPOSITION AND STRENGTH OF MORTARS 0) O be 1 1 02 >1 >> -(J pH on «! c ^ '-r' cj c a fH O) ' T-i ,-1 CO 1^ t» a 'S eS C cS 03 S-1 COMPOSITION AND STRENGTH OF MORTARS 33 THE USE OF "TRASS" FOR MORTARS AND CONCRETES IN MODERN BUILDING CONSTRUCTION Abstract of Paper read at the Sectional Meeting of the National Association of Technical Engineers of Germany by M. J. Mund, Cologne, 1908. Translated and Edited by Mr. William Challoner. The great advantages to be derived from trass as an admixture for mortar is demonstrated in the only available practical manner by examples of ancient structures which still defy the ravages of time. These results we are able to show approximately by tests conducted by more or less artificial means in the laboratory, which by a careful selection of materials and accu- rate calculations of the proportions are very reliable when the experiments are carried out by men of trained experience in such work. It is further desirable that the peculiar necessities of each class of work should be carefully considered in order to ensure permanent results, and that the laboratory tests should be made with the same materials and under the same conditions as exist on the actual work. If these are not closely observed the information obtained from the tests will be of little practical value, or at any rate entirely misleading, as, for instance, if cement, lime, or trass be tested with aggregates dissimilar to those which will be ultimately employed in the actual work, or under conditions of humidity and temperature different from those which will then constitute their environment. It must also be borne in mind that although the binding materials, cement, lime, and trass may be of the highest quality, the resultant mixture of mortar may be inferior, or even useless, owing to an absence of requisite care in the selection, proportioning, or grading of the sand and stone which form the aggregates, and incidentally, perhaps, the largest proportion of the mixture. A great impetus has been given to the use of trass in building mortars by the appointment, some twenty-five years ago, of a commission which represented leading engineers and architects, cement manufacturers, lime and trass quarry owners. This commission was subsidised by the Government, and is in connec- tion with the Royal Experimental Institute of Charlottenburg. Amongst its principal objects were those of dealing in a practical manner with the important matter of building materials on the lines indicated above, and in addition to prevent the erroneous deductions which may be obtained from tests carried out in a manner less in accordance with actual conditions. The use of trass in mortar and concrete had been general for engineering construction work both under water and below ground for foundations, and the attention of the commission was now directed to the advantages which it offered over the existing practice by its use as an admixture for mortar in the superstructure. To this end the specifications, which lacked detail in other respects, were revised, and a series of careful tests made on new lines which gave interesting and successful results. 34 COMPOSITION AND STRENGTH OF MORTARS As a consequence, leading building engineers and architects of public works adopted these revised specifications for mortars, with such modifications as were suited to the peculiar necessities of condition and environment. The chief points for consideration in arranging the specifications for mortar are strength, density, permanency, and economy in cost ; at the same time due regard must be paid to the quality and form of the aggregates which are available in the locality. From amongst many public buildings the following are submitted, for which specifications for mortars were drawn up in accordance with the results of the tests referred to, trass with fat lime being substituted for the hydraulic and other more or less impure limes it had formerly been the practice to use : — New Post Office, Cologne, 1892. Law Courts, Cologne, 1893. Library and Record Office, 1894. For these collectively, some 3000 tons of trass were used. The Government architects of the above buildings, M. Brugger and Trimborn respectively, have kindly supplied copies of specifications and comparisons of cost for the different mortars used on the work, and they have expressed their unqualified approval of results obtained. The specification of mortar for the whole of the Library and Record Office, of which M. Brugger was architect, is as follows — 1 volume of trass, 2 volumes fat lime putty, 5 volumes sand. This mixture possessed the highest efficiency in point of strong density and cheapness as compared with the mortar composed of — 1 volume hydraulic lime, 2 J vols, sand, which was the specification formerly in general use for the same class of building. In the New Post Office and Law Courts, of which M. Trimborn was architect, the specification was varied in respect to proportion of sand, but no hydraulic lime was used for either foundations or superstructure ; for these works the following specification was adopted — 1 volume of trass, 1 volume of fat lime, 3 to 4 vols, of sand. In a communication the architect states : " The walls built with this mortar are excessively strong, and the admixture of trass in the mortar gave much better results than could have been obtained by the use of hydraulic lime or cement for the same purpose." COMPOSITION AND STRENGTH OF MOHTARS 35 For the building of the Gas Works for the town of Mulheim, the Govern- ment architect, M. Rathke, used similar specifications, the mortar being composed of equal volumes of fat lime and trass to 3 vols, of sand. M. Rathke reports that the mortar gave results in every way superior to mortar that contained hydraulic lime. In the foregoing illustrations there existed no abnormal difficulties due to wet ground, and therefore a mortar of a high degree of density was unnecessary. Where, however, wet ground is encountered, it is necessary to use a mortar of a higher degree of density, and the following proportions have been found the most suitable in order to secure an absolutely watertight mixture — 1 volume trass, 1 volume fat lime, 1 to IJ sand. It is well known to architects and builders that in cases where the mortar is not well flushed, percolation of water, either under slight pressure or by capillary attraction, usually takes place by way of the joints; also that the use of dense insoluble stone of a porous nature or badly burnt bricks would not be allowed. Owing to the natural conditions of low-water-bearing ground, the Dutch architects and builders have long realised the necessity for a careful selection of impervious building materials, and for the preparation of dense insoluble mortars ; for that reason hard burnt bricks and trass lime mortars have been in common use in Holland for centuries. Most of the buildings in Haarlem, Rotterdam, Amsterdam and The Hague are evidence of the success which has attended this practice, as although nearly all the buildings have their foundations below normal water level, they show few signs of dampness or depreciation due to this cause. The new Railway Station at Amsterdam may be taken as an example of a recent building on which mortar that contained trass has been used. The volumes of lime and trass are varied proportionately to meet the necessities of the varying degrees of moisture to which the foundations must be subjected, and the proportions of the two former ingredients are reduced for the superstructure, where a less dense mortar will fulfil the required conditions of strength to maintain the stability of the building while at the same time effecting an economy in cost. It may be suggested that some of the proved advantages of trass lime mortar may be obtained by the use of Portland cement mortar, but it is unnecessary to point out that the theoretical advantages of cement mortar are entirely discounted by the difficulties attending its use in actual practice. If the mortar is sufficiently rich in Portland cement, it cannot be spread unless water is continually added, and the stones or bricks must also be kept wetted to prevent the water from the mortar being absorbed. The result is that very little of the cement eventually remains in the mortar. This absorption does not take place in trass lime or even trass cement mortars. 36 COMPOSITION AND STRENGTH OF MORTARS SECTION XI SUMMARISED CONCLUSIONS Whilst there are many points of interest which a study of the tables and diagrams bring to light, it will be sufficient for the present purpose to indicate those of a more prominent character in a semi-tabulated form for convenience of reference. The diagrammatic expression of results shows clearly the absolute necessity for tests of this nature to be carried out over a fairly prolonged period. Probably the full period of two years adopted by the committee is ample for the purpose of comparative results, in order that the effect of moisture in the mortar may, as far as is practicable, be eliminated. In many cases, what I have termed the "time sequence" shows clearly that when a mortar is new or " green " it will have a greater strength than when dried, the effect of the moisture being to exert a capillary attraction, and thus give a false strength which is not maintained later on. The fact that the irregular "time sequence" appears practically only in cases of mortars of distinctly inferior quality is of no little interest. The effect of different sands is very varied. Fine Charlton sand gave its best results with white chalk lime, and its worst with blue lias lime, whilst the addition of clay was distinctly detrimental. On the other hand, the coarser sands, Thames and pit, gave far better results with the lias lime than with the chalk limes, and with these sands no advantage was obtained by the addition of clay ; but when this material was added in the proportion of 5 per cent, to standard sand used with blue lias lime in the proportions of 1 lime to 3 sand, the strength rose to the highest of any obtained in the series. It would thus seem that, given good materials, the essential point is to completely fill the voids, as it cannot be assumed that the addition of such a small quantity of clay could of itself raise the strength of the mortar over four and a half times. An interesting feature of the tests consists of the results set out in Section V, from which can be ascertained what matrix or aggregate should be preferred for use with either one or the other to secure the best results. If a Dorking greystone lime is employed, pit sand in the ratio of 1 to 2 will give the maximum crushing strength, equal to double that obtained with Charlton fine sand, and clay should not be present. If the lime is blue lias, ground brick is preferable in the ratio of 1 to 2, which gives nearly four times the strength available with fine sand. If a given lime and sand are to be employed, say Dorking greystone lime and ordinary pit sand, the usual proportion of 1 to 3 is not so good as 1 to 2, which gives a crushing strength of half as much again, with only a very little lower tensile strength. COMPOSITION AND STRENGTH OF MORTARS 37 It is unfortunate that the tests with clay added to the aggregate were confined to proportions of 1 to 3 in all cases, as there is ample reason for concluding that if these had been made with proportions of 1 to 2, very much better results Avould have been obtained. The ratios of tensile to crushing strength are distinctly variable, and do not appear to convey any marked indication of value. The examination of ancient mortars presents features of peculiar interest. In the first instance they distinctly point to the absence of the formation of soluble silica, which it has been assumed was formed in the course of time by the action of the lime in sand — a mistake undoubtedly arising from the presence of pozzuolana in ancient mortars. Another point is that the presence of clean ferruginous clay has certainly had no deleterious effect, this being present in quantities up to 8 per cent, on the sand in some of the finest samples examined. These examinations further give rise to the question as to the wisdom of the assumed propriety of the proportions of 1 to 3, so generally recommended at the present time, as nearly all the best old mortars contained proportions of 1 to 1, to 1 to 2, which agrees with the conclusions now experimentally demonstrated that larger proportions of matrix than 1 to 8 give better results. The tests for adhesion of bricks with various mortars clearly demonstrated the superiority of Fletton bricks over Stock bricks, and the distinct inferiority of standard sand, Charlton fine and pit sand, compared with Thames sand and ground brick as aggregates. The details of the respective tests are fully set out in the attached Tables XIII to XIX and the accompanying Diagrams. W. J. DiBDIN. February 1911. 38 COMPOSITION AND STRENGTH OF MORTARS ouno icoo o»OkO ursm t^fMCO CO'^Oi 00 OCO kCOO - CO t-^ ^ (O O 00 00 (>4 (M (M i-H i-i OOOOvOO OOlC ooo CO o> I— I cc -M 1^ CO eo t-, (Mr-((MCOCO(MCOOOO(M?D rH.-i,-i GM(N0O i-l(>Jr-l C^05 00-^CO OiOi— c (NCO-* C> Sand >> Sand Fine (M CO me ime ime ime ime o 1—1 I— 1 I— ( COMPOSITION AND STRENGTH OF MORTARS 39 CO .o> COCO.000 0Okrs«O CD«D0O ooooco ^COCO COC^C^ »-(.u:50 iooo oo«oo ooo CN CO 01>-ur5 0010>0-*00(>JiOTiiOOiCO C^i-lrHCOC.»O (Mt^CM i-trHOO r-lO-. «OI>.Ot^ CNl o-^o ocoo u;5i^-<*' 05^i— » (Mr-(rH i-lr-(i-lr-t(M(N i § G « ~ OS CO m m UIT COMPOSITION AND STRENGTH OF MORTARS <»t^O rHr-li— I ?OQOc-. GOOO C^i-HC<) Oi ao (MCJi— I COCOCO ?-H(NrH r-i r-{ OlOaC^ OOOOiOOOOkOkOkCO i-(CO(M C^(M(M »-H(Ni-( r-H,-.,-! ooo OOkO ooo OlOO CO(NCO (MrjfcO OOOO ocoo COCOtJI (N(>)C^ f-l(N(M C^tNCO OvOvrSOOO OOOOCX300 m o CN coQo-^ co(Moo COO(N ■<*<'*CCt^Ot-^ OOOC^ .— lOi— I ■n'->*iO 05005X^(>JC^ iO-^CO i-l(Nr-l ^i-Hi-H O rH .COo:)t--.C (N> Sand. Sand CO Lime 0) :: S rH l-H I— < '2 ■73 c C CO 22 o B I— 1 COMPOSITION AND STRENGTH OF MORTARS lOOi t-»t^(N f^-^O -^t^OO Cq r-H rl .-H r-( r-( CM 1-1 o 44 COMPOSITION AND STRENGTH OF MORTARS urJOOOiOOWiCQOOCOO OTJ1lO(Ml-Hi-l.-ll-ll-lr-lrH(M OOO^k^lOiCiCiOOOOiO (M(MCOi-lF-lr-l.-li-Hr-li-( O lO «0 CO iC CO iC'OOOiCOOOi^OOO COCOCOi-li-l(Mr-lr-l.-lrHi-l.-( vOOOOiOiCOOOvCiOO ^OOiOCOCOi— (OO'^OCOiCt^ CO'^ICOCMCMl-HtMr-lr-ICMr-li-l , (MC o 2 s . .a . g I— 1 I— 1 T-l J— 1 1-1 1-H COMPOSITION AND STRENGTH OF MORTARS ooloooovoounooo o o o oo oo > Sand CO ime 03 I— 1 T-H m § GO CO ^ „^ ^ ^ CO o \ Tha Ijime 03 03 ime T— 1 COMPOSITION AND STRENGTH OF MORTARS OOO OOO OOlO lOOlO (M»o?o co'to t^coo -^com o o 1— I lO 00 CD OOOiniOtOOO ooooioooooir:ico l^-^iO lOOkO -*XivO o-^-^ .-((MvO C 000 o iX> Tf< lO 10 00 o o O (M r-l l-H 00 VO O rtl (N CO .-t (M l-H Q -3 Co e Sand {Charltc 3 Sand Sand {ordinal 3 Sand • ^ 60 COMPOSITION AND STRENGTH OF MORTARS O I— i H O o 1/3 (M o CO oo (M OO (M (M lO O CO o O VO O W 1- CO ?o CO oo 00 05 mth. CO CO I— t CO CO CO O CO ^ O 00 o S CO (M O O VO (M (M CO o o (M i-H r-t Greystoi VO lO lO CO CO 'st< oo O (M (M (M (M in vo o CO CO (N CQ CO § CO §3 o i."^ I— 1 '^1 O CO COMPOSITION AND STRENGTH OF MORTARS OO o ^ (M OO rH rH „ g B o o Q ©4 ^ 5~ eS ■2 S S3 If g ft. s D 2 52 COMPOSITION AND STRENGTH OF MORTARS o o o CO (M (N !N (M o »o o C 05 O S 5» * " " S ' *" CO ^ 00 me . .'^^ s _ r-l COMPOSITION AND STRENGTH OF MORTARS lO CO CO O »0 CD oo Oi ao o o o o CO o o o VO CO CO CO CO «) o o o oo oo OS rJH CO Ot) o o o O CO O 00 o O Tji T-H (N T-H O (N O CO o rH T-H (M O lO o O 00 r-H l-t I-H (M O O O CD O Ttl O O o urs vrs CO CO rH (M (N r-l o o o inn (N CO (N Tt< lO o o oo CO (M i-t 3MiJ .JO:' ■■ ■■'■I Table Xill. D IAGRAM SERIES A.b. TENSILE STRENGTH. WHITE CHALK LIME. *FINE SAND CHARLTON) X 30 lO 12 MONTNS Js2 1 MONTH 3 MONTHS 2 „ 2 YEARS ■ \ \\ w \\ 60 \\ 3 MONTHS Z YEARS ^56 55 53 "-^ J- \ \ 44- 4-6 1 MONTH 30 37 32 25 'ze, I VOU. LIME I VOL. LIME I VOL. LIME I VOL. LIME •0^ Table XIII. DI AGRAM SERIES A b' CRUSHING STRENGTH. WHITE CHALK LIME . + F1NE SAND (CHARLTON) 5 650 Z O 500 1 MONTH 12 » .— 2 YEARS 12 MONTHS 3 MONTHS 32; 2 YEARS 3C 278 < ""s '\ ^^-'^"^^^ U \ \ \ ISO 1 MONTH 6 133 88 I 0^ iJfiM Table Xlll. DIAGRAM SERIES Ac. TENSILE STRENGTH. WHITE CHALK LIME. + PIT SAND(0RD1NARY; 140 130 10 f^MAfj ^kLlAHO 3TIHW u iHTWOW ....... ■'£^5'. Si Table XIII. DIAGRAM SERIES Ac! CRUSHING STRENGTH. WHITE CHALK LIME . + PIT SAND. (ORDINARY) I t i I 08 • ; W A- J 6 A > ' ; 1"'} "■ ■" • / •• 1 ■ 1 1 I I ! Table XIII. DIAGRAM SERIES Ad! CRUSHING STRENGTH. WHITE CHALK LIME + THAMES SAND. Table XIII. DIAGRAM SERIES A.e. TENSILE STRENGTH. WHITE CHALK LIME. + GROUND BRICK. i Table XIII. DIAGRAM SERIES Ae! CRUSHING STRENGTH. WHITE CHALK LIME + GROUND BRICK. s asm 1 ■ / / 'v . . , J 1'/ ■ " f c .a 5 f. 3_£jifl.J.i. I^^jiiifi. r j I Table XIV. DIAGRAM SERIES B a! DORKING GREY STONE LIME + STANDARD SAND 900 I— 1 ' I ! ! 3 3i fl ac__ ■: ITS r Table XIV. D I AGRAM SERIES B b! CRUSHING STRENGTH. DORKING GREY STONE LIME + F1NE SAND (CHARLTON) I o a e 3 jj^ 3 a m a s o a i a 3 rvi : J " . J.v^ pre , Y3B.p\.Omfi^0Ci V j^'a h } a f 0 ? 0 M A a t > ' p■^\ oov Table XIV. DIAGRAM SERIES B.C. 08 OS t y. 1 i i Table XIV. DIAGRAM SERIES Be! CRUSHING STRENGTH. DORKING GREY STONE LIME. + PIT SAND (ORDINARY) K 1 1 1 1 — - 1 ! i 1 MONTH 3 MONTHS 12 „ 2 YEARS ~ 1 1 ^ — 12 MONTHS ^\ \ i \ \ j 3M0NTHS 1 MONTH 232 '-^'"T 18* 137 I VOL. LIME 1 VOL. LIMt I VOL. LIME i. i; ( ■ ;'-(< I Table Xiv. DIAGRAM SERIES B.d! CRUSHING STRENGTH. DORKING GREY STONE LIME + THAIVIES SAND. Table XIV. DIAGRAM SERIES Be. TENSILE STRENGTH. DORKING GREY STONE LIME + GROUND BRICK O M 8 8 SI. ^ eai AMU .JOV Table XIV. DI AGRAM SERIES Be! CRUSHING STRENGTH. DORKING GREY STONE LIME. +GROUND BRICK. c o Table XV. D IAGRAM SERIES C.a. TENSILE STRENGTH. BLUE L i A S LIME. + STANDARD SAND. 1 . _ .. - 1 MONTH 3 MONTHS 2 7 YFARS ~ — , 3 MONTHS 12 MONTH.^ 63 2 YEARS t I MONTH 44 \ " V- \ IS^-'-'''''^ r- 1 1 1 vol.. LIME I VOL. LIME I VOL. LIME I VOL. LIME 2 „ SAND 3 „ SAND 4 „ SAND 5 „ SAND p I s.OJ Aid i I V*. . Table XV. DIAGRAM SERIES C.a! CRUSHING STRENGTH. blue lias lime. hStandard sand t MONTH 3 MONTHS — 2 YEARS 2 VEARS S36 IZ MONTHS \ \ 3 MONTHS \ \ '\ x \ \ \ \ \\ \ j 1 MONTH \ \\ ! \ '^'A \ \ \ Ids 167 ^~^~~-„.,^^^l20' ISO lit 95" 1 VOL. LIME 1 VOL. LIME 1 VOL. LIME 1 VOL. LIME 750 700 660 aoo 400 350 300 I Table XV. D IAGRAM SERIES C.b. TENSILE STRENGTH. BLUE LIAS LIME. + FINE SAND (CHARLTON) 1 MONTH 3 MONTHS O VITARR 1 3 & 12 MONTHS !> YCin.: ^ ! Table XV. D IAGRAM SERIES C.e. — TENSILE STRENGTH. BLUE LIAS LIME. ♦GROUND BRICK. 2 YEARS :I33 i MONTH 3 MONTHS 2 „ O VP&PS 12 MONTHS lis \ \ \ _VA 113 j j 3 MONTHS \ \ A \\ 95 / 93 90 87 78 i I VOL. LIME I VOL. LIME I VOL. LIME I „ BRICK 3 „ BRICK 4 „ BRICK I I.. / Table XV. DIAGRAM SERIES C.el CRUSHING STRENGTH. BLUE LIAS LIME. 300 2SO 200 ISO 100 SO o ^ I ooi-: j OO"' flB L. MA3l>Ai O OB8 0 Table XVI. DIAGRAM SERIES Da. TENSILE STRENGTH. WHITE CHALK LIME + AGGREGATE + 5% {.ondo^cl i F i — ; .-^ MTMOI^f : &HTIIOM f- r : -= — ■ — Table XVI. DIAGRAM SERIES D a! 9'nrmi>M c e< . y^ii-^- - \ V', "S, \ I Table XVI. D IAGRAM SERIES D b. TENSILE STRENGTH. DORKING GREY STONE LIME + AGGREGATE + 5% ]londo"cl*v I70 I 1 I60 ISO lO STANPARD SANO FINE SAND PIT SAND 1 1 3 (CHABLTONl lOBOlNARVl Table XVI. DIAGRAM SERIES D b! CRUSHING STRENGTH. DORKING GREY ST©WE LIME. + AGGREGATE + 5% j ^ondon^clav 800 750 7O0 650 6O0 O «° Z O 5O0 3 O 450 a 4O0 300 250 2O0 ISO lOO 50 O I r ^ / / STANDARD SAND ^'^•ft .^J^ (J p A . -■- ' i't 6 .-'i ■ ■■ • / Table XVI. DI AGRAM SERIES D c! CRUSHING STRENGTH. BLUE LIAS LIME + AGGREGATE + 5% {^o^^l%%^^y & 8 5 Table XVII. DI AGRAM SERIES G a. TENSILE STRENGTH. WHITE CHALK LIME WITHOUT CLAY IN AQGREQATE. WITH CLAY IN AGQREQATE. 1 MONTH 2 VEARR 8S 7 0 12 MONTHS / 61 12 MONTHS ^^^^^-—-^ 5 SO S7 50 4 4 42 2 YFABS 3 MONTHS §0 ; 5 3S 26 2 5 P MONTH 24 .16 z 1 STANDARD SAND FINE SAND r et- bEB idnvHc imch Table XVII. DIAGRAM SERIES G b. TENSILE STRENGTH. DORKING GREY STONE LIME. Table XV/I. DIAGRAM SERIES G.c. - — TENSILE STRENGTH. BLUE LIAS LIME. WITHOUT CLAV IN AGOREGATE. WITH OLAV IN AGGREGATE. 1 i Table XVIII. DIAGR AM SERIES G . a! CRUSHING STRENGTH. WHITE CHALK LIME: WITHOUT CLAY IN AGGREGATE. WITH CLAY IN AGGREGATE. I 700 1 MONTH IZrvlONTHS 308 12 MONTHS Z38 3 MONTHS 267 2 YEARS 240 0 - — — ,253 243 3 MONTHS 23f 2 rEARS 230 1 MONTH 122 1 MONTH Us ri STANDARD SAND FINE SAND PIT SAND STANDARD SAND FINE SAMO 1:3 (CHARLTON) (ORDINARY) |:3 iCHABLTONi Table XVIII. DI AGRAM SERIES G b! CRUSHING STRENGTH. — DORKING GREY STONE LIME WITHOUT CLAY IN AGGREGATE. WITH CLAY IN AGGREGATE. 900 800 lOO O 12 MONTHS 2 YEARS ■\. a 232 216 3 MONTHS 1 MOUTH 2 YEARS 3 MONTHS 1 MONTH 12 MONTHS lee 140 80 STANDARD SAND FINE SAND PIT SAND STANDARD SAND FINE SAND PIT SAND a. ■ .St 'S Z, op < ? ^ \_ '■ o 1 o .; t9 ■cc- 51^ ;0 0 Table XVIII. DIAGRAM SERIES G.c! — CRUSHING STRENGTH. BLUE LIAS LIME. , ; H Tiw d3C3 i/iba ^>\6\na a dot e : , ■ '■ ■ ' ''•*'< ;t . si?*"'" 3 Table XiX. DIAGRAM SERIES Ea.b.c.d.e. ADHESIVE STRENGTH. I I Table XIX. DI AGRAM SERIES F.a.b .c.d.e. ADHESIVE STRENGTH. FLETTON BRICKS BONDED WITH DORKING GREY STONE LIME + AGGREGATE 'I : 3