UNIVERSITY OF CALIFORNIA LOS ANGELES PRINCIPLES OF METALLOGRAPHY 10124 PUBLISHERS OF BOOKS F O R^, Coal Age v Electric Railway Journal Electrical World v Engineering News-Record American Machinist ^ Ingenieria Internacional Engineering S Mining Journal v Power Chemical & Metallurgical Engineering Electrical Merchandising iiiiiiiiiiiiiiiiTfliiiiiiTiliiiiiflliiiiiiriliiiiiilfliiiiiriTiiiiiiTITiiiiiflniiiiiriliiiiiiiFiiiiiifiTiniiiliiiiiiii PRINCIPLES OF METALLOGRAPHY BY ROBERT S. WILLIAMS, S.B., PH.D. ASSOCIATE PROFESSOR OP ANALYTICAL CHEMISTRY AND INSTRUCTOR IN METALLOG- RAPHY, MASSACHUSETTS INSTITUTE OP TECHNOLOGY McGRAW-HILL BOOK COMPANY, INC. 239 WEST 39TH STREET. NEW YORK LONDON: HILL PUBLISHING CO., LTD. 6 & 8 BOUVERIE ST., E. C. 1920 COPYRIGHT, 1919, BY THE MCGRAW-HILL BOOK COMPANY, INC. TO WHOSE INSPIRATION MY INTEREST IN METALLOGRAPHY IS DUE PREFACE This little book has been written to meet the needs of those students of General Science on Engineering who do not specialize in Metallography but who will use it to a limited extent in connection with their professional work. It is hoped that it will be of service, also, to the general reader as an introduction to an increasingly important branch of service and as an aid to the better under- standing of the more highly specialized books. Greater emphasis has been laid on the applications of Metallography than on the physico-chemical principles involved but it is believed that the fundamental ideas on which metallography is based have not been neglected. In the appendix will be found a few of the tables most commonly used by the metallographist, a suggested out- line of a brief laboratory course and a descriptive list of the more important books and journals dealing with the subject. Thanks are due to the authors of many of the stand- ard books on metallography which have been freely used in the preparation of this little volume and grateful acknowledgment is made for the use of a few drawings which have been copied with minor changes from other books. Special thanks are due to Messrs. Bauer and Deiss from whose book on "The Sampling and Chemical Analysis of Iron and Steel" most of the microphotographs of steel and iron have been taken. It is a pleasure to express my appreciation for the services of Professor L. F. Hamilton who has helped greatly by his kindly criticism of the proof. CONTENTS PAGE PREFACE v " CHAPTER I. THE SIMPLE ALLOY DIAGRAM 1 II. LABORATORY METHODS OF METALLOGRAPHY 18 III. THE ALLOY DIAGRAM AND ITS MEANING 39 IV. THE NON-FERROTJS ALLOYS OF TECHNICAL IMPORTANCE. ... 69 V. IRON AND STEEL 9 ^ VI. DEFECTIVE MATERIAL 123 APPENDIX 136 INDEX . 151 PRINCIPLES OF METALLOGRAPHY CHAPTER I THE SIMPLE ALLOY DIAGRAM An alloy may be defined as the solid which results when two or more metals harden from the molten state. In a few rare cases alloys are prepared in other ways as, for instance, by compression of the solid metals or by electrolytic deposition, but the amount of material alloyed in this way is negligible. Metallography, or Physical Metallurgy as it is some- times called, is the general study of metals and alloys. In its more restricted sense it deals with finished alloys, their physical and*chemical properties, their internal structure, the^methods of investigation and, perhaps most important of all, the study of the mechanical prop- erties and defects of the commercial alloys. When a pure metal is melted under conditions which make it possible to determine the changes of tempera- ture during the cooling and these changes are indicated in graphical form, a curve of the following form is ob- tained (Fig. 1). The temperature readings are taken at definite time intervals and in plotting the curve these temperatures are used as ordinates while the correspond- ing time intervals are abscissae. The solidification of lead has been selected as an example, with time inter- vals of ten seconds. It will be noticed that the curve 1 'METALLOGRAPHY falls smoothly until the temperature of 327C. is reached when it breaks sharply and remains horizontal indicating a constant temperature during an appreci- able period. The curve again falls gradually to ordi- nary temperatures without further abrupt changes in direction. The horizontal line represents the transition of the lead from the liquid to the solid state. The heat which is necessary to maintain the mass at constant Time in .Seconds FIG. 1. Cooling curve of pure lead. temperature is the latent heat of solidification. Since the amount of heat liberated at this time is proportional to the mass of metal solidifying, it is evident that, other conditions being equal, the length of the horizontal is a measure of the amount of material present. This does not hold absolutely in practice as it is impossible to get ideal cooling conditions but it is true to such a degree of approximation that the fact is of great value in the thermal study of alloys. The horizontal has the addi- tional physical significance that it indicates the only temperature at which solid and molten lead will stay in THE SIMPLE ALLOY DIAGRAM contact with each other indefinitely. At any higher temperature solid lead will disappear and at a lower one there will be no liquid. Several other methods of drawing these " cooling curves" will be described later in connection with the laboratory study of the alloys. The Two Layer Alloy. The simplest possible alloy is one which results from the solidification of two metals 200- M WSAl 15* Al 50*Al 30 * 10* Pb 25* Pb 50* P6 70 * Time in Seconds FIG. 2. Cooling curves of aluminum, lead and four alloys. which do not mix even when they are in the molten condition. In the strictest sense this is probably not a true alloy but the subject deserves consideration because of the practical application of certain of these non- miscible metal pairs in the form of solid emulsions. If aluminum and lead are mixed in varying proportions, the mixtures melted, allowed to cool under conditions as 4 PRINCIPLES OF METALLOGRAPHY nearly identical as possible and temperature readings are taken at definite time intervals during the cooling, a series of curves like that shown in Fig. 2 will be obtained. Curve No. 1 is the cooling curve of pure aluminum and each successive curve represents the cooling of an alloy containing more lead than the one before it until pure lead solidifies as indicated in Curve No. 6. This method of representing the temperature changes in a series of alloys is not satisfactory even in so simple a case and it becomes impossible in more complicated cases. This has lead to the use of a graphical form of representation known sometimes as the "freezing point diagram" but, more accurately, as the equilibrium dia- gram of the alloys. Disregarding for a moment the time taken for each alloy to solidify, as shown by the horizontal lines in the series of curves, it is evident that the temperature changes may all be indicated on a chart in which the ordinates represent temperatures and the abscissae percentage composition of the different alloys. Since, in the case under consideration, the melting point of neither metal is affected by the presence of the other, the freezing point diagram consists of two horizontal lines only, one at the melting point of aluminum, the other at the melting point of lead. In his various papers on Thermic Analysis, Tammann 1 has shown that, in addition to the simple freezing point diagram given above, the introduction of a curve, or curves, showing the time interval during which the tem- perature stays constant for each alloy of a series is exceedingly valuable both in the construction and in the interpretation of these equilibrium diagrams. The time curves are obtained and applied to the diagram as follows: a straight edge is applied to the sloping parts of each cooling curve as shown in Curve No. 2, Fig. 2 *Zt. Anorg. Chem., 37 (1903), 303; 45 (1905), 205; 47 (1906), 289. THE SIMPLE ALLOY DIAGRAM and the resulting sloping lines are connected by hori- zontal lines drawn through the horizontal parts of the cooling curves. The distance between the points of intersection of each horizontal line with the two oblique lines which it crosses represents the time either of solidi- fication or of some other change taking place at constant temperature. In Tammann's earlier work these time lines were drawn from the horizontal temperature lines to which they referred as shown in Fig. 3. This led to Al - 658.7 P6-327 Composition FIG. 3. Aluminum-lead diagram. a certain amount of confusion as one set of ordinates had to be used to indicate both time and temperature. In the more recent work of other metallographists the time curves are made a separate part of the diagram as shown in Fig. 4. Since, as was stated on p. 2, the time taken for a metal to solidify is proportional to the amount of material present when the cooling conditions are substantially identical, it is obvious that the time of solidification of the aluminum will be greatest where its amount is greatest and will decrease to zero where only lead is present. The triangle which is formed by drawing a line through the ends of the time perpendiculars is 6 PRINCIPLES OF METALLOGRAPHY useful in several ways. From it may be determined not only how much time will be required for any alloy in the series to solidify, no matter what its composition, but also, what is far more important, the percentage composition of the alloy. It is only necessary to cool an equal weight of the unknown specimen under standard conditions, determine the time interval on its cooling curve and locate the same time interval on the equi- 00/U90 SO 70 GO 50 4 U 3U ; o i|o 2)0 a|o 4|o s|o c|o T|O Composition FIG. 4. Aluminum-lead alloys (GROYER). The ordinates of the triangles marked "time in seconds" correspond to the horizontal lines in the curves of Fig. 2. librium diagram. The point at which the ordinate of the time triangle has the same length as the unknown time interval indicates the percentage composition of the alloy. This is the fundamental idea of Thermic Analysis and, while the results do not compare in accuracy with ordinary quantitative methods nor can the method be used at all in many cases, it is quite possible, in those instances in which it may be used, to get results very rapidly indeed and with a degree of accuracy which is often sufficiently great for much commercial alloy work. The complete equilibrium diagram makes it possible THE SIMPLE ALLOY DIAGRAM 7 to predict the physical condition of any alloy in the series at any temperature included in the diagram. This question of the interpretation of diagrams is one of the greatest importance to the metallographist. Consider as an example the behavior of an unknown alloy x, Fig. 4, during its cooling from a temperature 1 to the tem- perature 4. Since the diagram shows that in all mix- tures of this series the first heat evolution occurs at the temperature of the line A B it follows that all alloys above that line and, therefore, alloy x will be in the liquid condition at the temperature 1. When the tempera- ture has fallen to that indicated by A B, or point 2 on the dotted line, a temperature effect is noticed on the diagram. Since this evolution of heat takes place at the same temperature for all alloys of the series and since it corresponds to the melting point of aluminum, the assumption is justified that at and below the point 2 the mass consists of crystals of aluminum suspended in a mass of fused material. That this assumption is correct may be shown by removing some of the solid material at any temperature lower than that of point 2 and higher than point 3. It will be found to be pure aluminum. At point 3 on the line CD a second heat effect is shown and since this point corresponds to the melting point of lead the -same line of reasoning as before shows that this effect represents the solidification of the lead which the mixture contained. As no further heat changes are shown on the diagram it may be assumed that alloy 4 is a solid mixture of aluminum and lead. The diagram as drawn gives the following general information with regard to alloys of aluminum and lead. At any temperature above that represented by AB, any alloy of aluminum and lead, no matter what its percent- age composition, will consist of a mixture of molten 8 PRINCIPLES OF METALLOGRAPHY aluminum and molten lead and in this particular case owing to the great difference in the specific gravity of the two substances it will consist of two distinct liquid layers. In the area between AB and CD all alloys con- sist of a mixture of solid aluminum and molten lead, the relative amounts of the two metals varying with the percentage composition. In the temperature range below CD the alloy is made up of solid lead and solid aluminum. Unless the alloy is vigorously stirred during the solidification, the solid will be found to consist of two distinctly separated layers. After the construction of the diagram, or frequently in practice simultaneously with its construction, a micro- scopic study of the solid alloys is made. A highly pol- ished surface free from scratches is obtained on which the internal structure of the alloy is brought out by treatment with suitable etching reagents which attack one constituent more than the other and produce in this way depressions in the surface, color changes of the constituent attacked or other differences which are noticeable under the microscope. The method of polish- ing the specimens and preparing them for microscopic examination will be considered in the section on Labora- tory Methods, p. 18. The following microphotograph (Fig. 5) of a lead-aluminum alloy shows the way in which the microscope is used to confirm the results obtained by temperature measurements. As was to be expected from the diagram the microscope shows two distinct layers, the dark lead at the bottom and the lighter colored aluminum above it. Plastic Bronze. The most important example of the alloys formed by non-miscible metals is the class of copper-lead alloys known as plastic bronzes, much used as bearing metals. Melted lead forms an emulsion with melted copper and if the cooling of the alloy takes place THE SIMPLE ALLOY DIAGRAM 9 rapidly enough the lead will be found more or less uni- formly distributed throughout the mass of copper in the form of spherical drops. By the addition of nickel or other high melting material in small amounts it is pos- sible to prepare an alloy containing 50 per cent, lead though the usual plastic bronzes contain only from 15 to 30 per cent. lead. In alloys of this type the copper gives the necessary strength while the lead increases the plasticity of the alloy and acts as a lubricant. FIG. 5. Aluminum-lead alloy (75 X)- While it is probably true that no two metals are absolutely insoluble in each other, either in the liquid or in the solid state, the solubility in the case just de- scribed is so slight that for practical purposes it may be disregarded. There are many alloys of this class but those of copper and lead are the only ones of technical interest. The Eutectic Alloy. The next type of alloy to be considered is that in which the two metals are completely miscible in the liquid state and completely non-miscible, or insoluble in each other, in the solid state. It is a 10 PRINCIPLES OF METALLOGRAPHY well-known fact that in most cases in which one sub- stance is added to another, in which it will dissolve, the freezing point of the solvent is lowered. An illus- tration of this phenomenon is the preparation of the ice-salt freezing mixture which is able to produce a tem- perature twenty-one degrees below that of ice alone. Numerous alloys behave in the same way. One of the best known of these is the lead-antimony alloy which has many important commercial uses. If a small amount of lead is added to molten antimony the freezing point FIG. 6. Lead-antimony cooling curves. of the latter is lowered to a considerable extent and increasing quantities of lead still further lower the freez- ing point. If, on the other hand, a small amount of antimony is added to pure lead the melting point of the lead is also lowered and, as in the case of the antimony, is progressively lowered by the addition of greater quantities of antimony. The effect of the addition of each metal to the other is shown in the series of curves of Fig. 6. It is obvious that since each metal lowers the freezing point of the other, the lines connecting these freezing points must intersect at some point as shown by the dotted lines in the drawing Fig. 6. This point THE SIMPLE ALLOY DIAGRAM 11 of intersection is one of great interest and importance and has been called the eutectic point. The alloy cor- responding to the composition at which the two lines intersect is the eutectic alloy and the temperature is the eutectic temperature. If the data given by the cooling curves is assembled in the form of an equilibrium diagram, as before, the diagram takes the form shown in Fig. 7. The significance of this type of diagram can be under- stood most readily by considering the physical changes which take place in a few special cases as, for instance, 50 30 Composition FIG. 7. Lead-antimony diagram. 10 Sb during the cooling of alloys 1, 2, and 3 in Fig. 7. Since the V-shaped curve was obtained by connecting the freezing points of the separate alloys, it is evident that the area above the V represents a temperature range in which everything is in the molten condition. This is fre- quently called the liquidus. As the temperature of alloy 1 falls no change takes place until the line PbB is reached, at which temperature pure lead begins to separate. The result of the separation is to leave a solution richer in antimony than the original solution and, therefore, one which has a lower freezing point. Pure lead continues 12 PRINCIPLES OF METALLOGRAPHY to separate with the consequent formation of solutions increasingly rich in antimony and therefore with lower melting points. The fact that the solution from which the lead is crystallizing is of constantly changing compo- sition is the reason for the shape of the cooling curves (see Fig. 6) in alloys of this type. The curve is not a horizontal line as in the case of a pure metal but, as it represents an infinite number of freezing points, it ap- pears in the form of a change in the direction of the normal curve, or as a " hold " as it is often called. As the liquid from which the lead is separating becomes richer in antimony it approaches the eutectic composition in- dicated by B. Since this represents the lowest possible temperature at which lead and antimony alloys can solidify, it is evident that when the residual liquid finally reaches the eutectic composition the metal will solidify at this constant temperature. The same reasoning applies to alloy 3 except that in this case the antimony crystals separate first. The primary separation of antimony is followed by an enrichment of the remaining liquid with lead until the eutectic composition is reached again. At the composition 2 no change takes place until the eutectic temperature is reached when lead and anti- mony solidify together in the form of the eutectic mix- ture. The line DBE (the eutectic line) represents that temperature below which the alloy is solid and is there- fore called the solidus line. To summarize the statements made above, it may be said that any alloy having a composition between D and B shows two heat evolutions on cooling, one which corresponds to the primary separation of lead and a second which is due to the solidification of the eutectic mixture. The amount of residual material having the eutectic composition is greater the nearer the composi- tion approaches that of the point B and is zero at THE SIMPLE ALLOY DIAGRAM l point D. Along SbB antimony is the primary separa- tion and the eutectic mixture of antimony and lead is the secondary. Because of the constant temperature at which the eutectic separates it was formerly believed that the eutectic was a compound. The microscope shows that this is not the case but that, on the contrary, the eutectic alloy is an extremely intimate mixture of the two com- ponent metals. Since the eutectic is a mixture of the two metals and since, as shown in the diagram, the eutec- tic line extends from one side of the diagram to the other, it follows that while lead and antimony are wholly mis- cible and soluble in each other in the liquid state they are wholly non-miscible, or insoluble in each other, in the solid state. That the metals are insoluble in each other in the solid state must be true as the diagram shows that, however small an amount of either metal is added to the other, there is always the secondary heat effect at the eutectic temperature. The practical application of the time curves is much more evident from this diagram than from the preceding one. As the time taken for the eutectic to solidify is greatest where there is most eutectic, it follows that the time curve has its maximum at the eutectic composi- tion and drops to zero at the pure metals. This fact is most useful in the construction of the diagram of two unknown metals. Formerly the location of the eutec- tic point was a matter of repeated trials with no advance information as to the probable location of the point. With the introduction of the time-line idea the question is much simplified. It is only necessary to determine the time-lines for a few alloys. After plotting these lines on a horizontal base the ends are connected. The in- tersection of the two oblique lines resulting gives a close approximation to the eutectic composition so that its 14 PRINCIPLES OF METALLOGRAPHY exact determination, if that is desired, is a matter of a very few additional experiments. FIG. 8. Lead-antimony alloy with excess lead. Corresponds to @, Fig. 7 (75 X) (HOMERBERG). FIG. 9. Lead-antimony eutectic. Corresponds to @. Fig. 7 (75 X) (HOMERBERG) . The microscopic structure of these alloys is exactly what would be predicted from the diagram. All alloys THE SIMPLE ALLOY DIAGRAM 15 from pure lead to the eutectic composition show pri- mary lead crystals surrounded by more or less of the FIQ. 10. Lead-antimony alloy with excess antimony. Coi responds to , FlG. 7 (HOMEKBEBG). FIG. 106. Bismuth-tin. Eutectic with slight excess tin (SAWYER). eutectic B, depending on the composition (Fig. 8). Alloys from the antimony side to the eutectic, show 16 PRINCIPLES OF METALLOGRAPHY primary antimony crystals imbedded in the eutectic (Fig. 10). The alloy having the composition B shows simply the fine-grained eutectic structure without pri- mary crystals of either metal (Fig. 9). A better defined eutectic is shown in Fig. 106, the Bi-Sn eutectic. The lead-antimony alloys are commercially of much importance. Antimony is too brittle to be of use alone but because of its hardness it gives to lead properties which are very desirable for certain purposes. The alloys are used for acid proof coatings, for type and for light bearings. The eutectic alloys of greatest commercial importance Composition FIG. 11. Lead-tin diagram. are in the tin-lead series of which solders are made. The incomplete diagram, Fig. 11, shows those changes which are of technical interest but omits certain facts which will be considered later. Of the many mixtures used, two characteristic types are tin solder, containing about 37 per cent, lead (corresponding to the eutectic composition) and plumber's solder with approximately THE SIMPLE ALLOY DIAGRAM 17 67 per cent. lead. The diagram shows that between the points A and B, corresponding to a temperature drop of about 70, the alloy consists of lead crystals supported in a molten metal. This produces the pasty consistency which makes the plumbers " wiped joint" possible. The two examples just given illustrate in a general way the factors which are commonly determined in the study of alloys. Many other types of alloy- diagrams have been worked out which deal with the formation of intermetallic compounds, solutions of one ' metal in another, transitions of one compound into another, either during the solidification of the alloy or after it has completely solidified, and other possible changes which may take place. Before considering these more complex alloy diagrams and their applications in prac- tical work it will, perhaps, be simpler to study some of the methods and forms of apparatus that are actually used in the laboratory preparation and microscopic study of the alloys. CHAPTER II LABORATORY METHODS OF METALLOGRAPHY The preceding chapter indicates that in the laboratory study of the alloys and in the construction of their diagrams, one of the most important factors to consider is the succession of heat changes which take place when the molten mixture of metals passes into the completely solid state. In a few cases, notably in the cooling of steel, changes of vital importance occur far below the point of solidification of the alloy so that it is necessary in special instances to follow the cooling to very low temperatures. A study of these thermal changes nec- essarily involves three factors: (1) A method of melting the mixed metals, (2) a container in which the metals can be melted, and (3) an apparatus for measuring tem- peratures and, especially, temperature changes. 1. Furnaces. The furnaces used in the melting of metals vary so much with the amount of material to be melted and with the melting points of the metals in- volved that but one furnace will be described in detail. There is no difficulty, however, in building or buying small furnaces for any sort of alloy work. For the study of the heat changes which take place in alloys of low melting point (90QC. or lower) a con- venient form of apparatus is shown in the photograph Fig. 12 and given in detail in the sketch Fig. 13. A is an iron tube about 5 inches long and from 2^ to 4 inches in diameter. It can be made quite readily by threading one end of a short piece of steel pipe and screwing on a cap. The open end is fastened to a triangle of heavy 18 LABORATORY METHODS OF METALLOGRAPHY 19 FIG. 12. Photograph of a simple type of cooling curve apparatus. Thermo-Couple opper Leads to MilU : Voltme)e Milli-voltmeter Fio. 13. Diagram sketch of melting system and cooling curve apparatus. 20 PRINCIPLES OF METALLOGRAPHY iron or chromel wire so that the tube may be suspended over a burner by means of a ring and stand. The source of heat may consist of one or more Tirrell or Meker burners or, for very high temperatures, a blast lamp. The flame is protected from draught by means of a concentric cylinder of asbestos cloth separated from the outer wall of the iron tube by a space of ^ to 1 inch through which the gas flame can pass. This asbestos collar also retains the heat and makes the melting proc- ess easier and quicker. This simple method of heating may be replaced by any of the more elaborate types of furnaces. The more expensive wire wound electric re- sistance furnaces of the vertical type (dental furnace type) are much easier to control if provided with a vari- able resistance. For the high melting elements like nickel, cobalt, chromium and the metals of the platinum group a carbon resistance furnace is necessary, the details of construction of which may be found in any book on applied electricity. A recently developed electric fur- nace of the induction type promises to be of great value in the study of the high melting alloys. 2. Container for the Melted Metals. The material and shape of the container depend on the temperature at which the alloys melt, the properties of the melted metal and the shape of the furnace. For use with the furnace just described a tube of the shape shown in B, Fig. 13, is most convenient. This tube is embedded in sand in the iron case A so that heat may be distributed to it as uniformly as possible. For alloys with a low melting point (less than 700C.) hard glass tubes are most useful as the filling of the tube and adjusting of the tem- perature measuring instrument can be watched. For alloys which melt above the softening point of glass, tubes of porcelain or fused quartz are required. Thin- walled tubes of unglazed porcelain may be used for LABORATORY METHODS OF METALLOGRAPHY 21 almost all metals. In a few instances where the metal forms an oxide which is highly reactive chemically (e.g., chromium oxide or manganese oxide), it may be necessary to use the much more fragile magnesia tubes. This expedient, fortunately, is seldom needed. Tubes of this size limit the amount of alloy to be studied to about 40 grams. When larger weights of metal are to be used larger furnaces and crucibles must be substituted. Filling the Tube or Crucible. The metallic elements of which the alloy is to be composed are weighed in the desired proportions in the form of small chips, clippings or drillings and, in most cases in laboratory practice, are mixed before they are introduced into the tube. In exceptional cases it may be necessary to melt the less volatile component first, adding the more volatile ele- ment in successive small portions. The amount of material to be used is determined by the accuracy de- sired in the final results. As a laboratory experiment or as a preliminary survey of the field to determine the general shape of the equilibrium diagram, 20 to 30 grams of the mixture is enough. It must be clearly emphasized, however, that where great accuracy is re- quired and slight heat changes are to be looked for, the amount of material must be very greatly increased, often up to 400 to 500 grams. As the metals used should be of great purity and are therefore expensive it is sel- dom desirable to experiment with such large amounts of material. Stirring. There is often a marked tendency for the molten metals to separate into layers, especially if they differ considerably in specific gravities and mix with diffi- culty or not at all. In such cases the liquid mixture must be stirred during the course of the experiment. This can usually be done by means of a glass or porce- lain rod having a circular bend at the bottom through 22 PRINCIPLES OF METALLOGRAPHY which the thermometer or other temperature measur- ing device can pass (see C, Fig. 13). This rod is moved slowly up and down during the cooling of the melted alloy. Oxidation in the Tube. It is necessary in all cases to protect the metals during the melting and during the solidification from the oxidizing effect of the air. In a few instances this may be done by covering the surface with powdered charcoal but it is usually more convenient and more effective to melt the metals in an inert gas. Hydrogen, carbon dioxide and nitrogen have all been used for this purpose. The gas passes from the generator into the melting tube through the bent glass or porcelain tube, Fig. 13. As a safety measure, if hydro- gen is used, it is best to pass the gas through a small drying tube containing a number of disks of wire gauze. The gauze will cool the gas so that there is no danger of setting fire to the hydrogen in the generator or storage tank. The oxidation effects are slight with the low melting metals protected in this w r ay so that it may be assumed that the composition of the resulting alloy is the same as the composition of the mixture from which it was made. This is not the case with high melting, easily oxidized metals, even under the most favorable condi- tions, so that the final composition of the alloy should be determined by chemical analysis. Weight and Atomic Per Cent. In all industrial alloy work, and for most laboratory purposes, the metals are mixed according to their percentages by weight of the total amount of material used. It will be shown later that some metals form intermetallic compounds. In this case a system based on the atomic relationships is more convenient. For example, a compound of tin and magnesium containing 70.95 per cent, tin by weight LABORATORY METHODS OF METALLOGRAPHY 23 is known. This percentage composition gives no indica- tion of the relation of the atoms in the compound. If now the composition is indicated in atomic per cent., it will be found to be 33.33 atomic per cent, tin and 66.66 atomic per cent, magnesium, showing at once that the formula is SnMg 2 . The following expression shows the method of converting weight per cent, into atomic per cent. : A = atomic weight of first metal; B = atomic weight of second metal; p = weight per cent, of A; q = weight per cent, of B. Then: Atomic per cent, of A = - and: 100 q~ Atomic per cent. x)f B = r 3. Measurement of Temperature Changes. 1 A mer- cury thermometer will serve for the measurement of the temperature changes with the very low melting alloys as those of sodium, potassium or the amalgams. Almost all alloy work, however, requires higher tempera- tures than can be determined in this way and some form of pyrometer must be used. As in the case of the fur- naces, many excellent pyrometers of different sorts are obtainable. One of these pyrometers will be described as it illustrates the general method of use which may be applied with slight variations to any of the other instru- ments. Protected from direct contact with the molten 1 For a detailed description of temperature measurements see "The Measurement of High Temperatures" by BUKGESS and LE CHATLIEB. 24 PRINCIPLES OF METALLOGRAPHY metal by a glass, quartz or porcelain tube E (Fig. 13), are two wires of different metals joined at the bottom to form the thermal junction or thermoelement. For the measurement of low temperatures one of these wires may be copper, the other constantan (a copper-nickel alloy). For somewhat higher temperatures, a thermal couple made of chromel-iron may be used. For tem- peratures up to about 1650C., a rare metal couple, one wire of which is platinum, the other an alloy either of platinum and rhodium or platinum and iridium, is required. The two wires forming the thermal couple must be insulated from each other inside the protecting tube. This may be done by wrapping one wire with asbestos thread or by incasing it in short lengths of capillary tubing of quartz or porcelain. The loose ends of the wire pass into a jar F where they are kept either at constant temperature or at some temperature which may be determined from time to time and will not vary more than two or three degrees during the experiment. A convenient arrangement is to have these wires pass through a cork stopper into a Thermos or other vacuum walled bottle. If ice is kept in the container during the run, no correction for the temperature of the "cold junc- tion" is needed. Otherwise, a correction for the tem- perature of the cold junction must be made. The cold ends of the thermocouple are connected inside the jar F to insulated copper wires leading to the instrument G. The instrument on which the temperatures are read is commonly a millivoltmeter. Most of the millivoltmeters recommended for temperature work of this character are provided with two scales, one of which reads millivolts and the other temperatures directly, either in Centigrade or Fahrenheit degrees. The temperature scale may be used with only a single pair of metals and even then the reading varies somewhat with continued use of the couple, LABORATORY METHODS OF METALLOGRAPHY 25 so that, except for commercial work, the millivolt scale is almost always used in spite of the obvious advantages of a direct temperature reading. In order, then, to establish the connection between the millivoltage as read and the temperature to which it corresponds, calibration is necessary. This can be most simply done by determining the melting points of a series of pure substances and constructing a plot, using the known temperatures as ordinates and the instrument readings as abscissae. The following list of pure substances and melting or boiling points is given for convenience but any materials having definite melting points at suitable intervals may be used. Degrees C. Water b. p 100 Tinm. p '... 231.9 Sulphur b. p 444.5 Antimony m. p 630 Silver m. p 960. 2 Copper m. p ' 1082.8 In locating the freezing points of the substances used in the calibration, it is convenient to follow the same procedure that is followed later in studying the alloys themselves. The pure substance, tin for example, is heated until it is completely melted. The supply of heat is then cut off and the melted mass allowed to cool slowly. Temperature readings are now taken at defi- nite time intervals. If the temperature is high and is falling rapidly, intervals of five seconds are allowed be- tween each consecutive reading of the millivoltmeter scale. When the cooling rate is normal (about 5 to 10 degrees per minute) an interval of ten seconds between readings will show any material changes in the cooling rate. At low temperatures where the difference in temperature between the alloy and its surroundings is small, the cooling will be so slow that a much less frequent reading of the instrument is required. The frequency of 26 PRINCIPLES OF METALLOGRAPHY readings is a question of judgment but ten seconds be- tween readings may be taken as a reasonable interval. If the millivoltages are now plotted as ordinates and the time intervals as abscissae, a curve will be obtained show- ing a horizontal break at that millivoltage corresponding to the freezing point of the material in question. This horizontal line which indicates, not only the temperature of solidification but the time taken for the material in question to solidify, varies in length with the amount of material used and with its latent heat of fusion. The constancy of this time interval under the same condi- tions make its determination of great value in making the alloy diagrams. Various types of millivoltmeters are in use, some direct reading, some of the mirror type in which the deflection is magnified by reflecting a beam of light on a scale at some distance from the instrument and still others of the recording type. Whichever instrument is used, the general method of study consists in establishing a connection between the temperature changes and the intervals of time during which these changes occur. Instead of using a millivoltmeter as a temperature meas- uring instrument, a potentiometer may be used. This instrument requires somewhat more care and experi- ence in operation than the millivoltmeter but the results are far more accurate. Where exact temperatures, rather than relative temperatures or rapid temperature changes are required, the potentiometer should always be used. The modern forms of potentiometer can be operated so much more rapidly than the early types that their use by metallographists is constantly increasing. Having considered the various factors involved in the study of the temperature changes taking place when an alloy is cooled it may be helpful to summarize these factors in a brief description of an actual melting opera- LABORATORY METHODS OF METALLOGRAPHY 27 tion. Referring once more to Fig. 13 (p. 19), the proc- ess is as follows. Metals X and Y are weighed in a finely divided condition, such weights of the metals being taken as will produce an alloy of the required com- position and in such amounts that with the apparatus described the total weight will be between 20 and 30 grams. These are mixed in the tube B, imbedded in sand in the furnace tube A. The stirring rod C, the gas intake tube and the thermoelement protector E pass through a cork stopper or a suitably perforated brass cap. The insulated wires pass to a constant temperature bottle F and then, by copper leads, to a millivolt-meter or potentiometer G. Heat is applied to the sand bath and the metals melted, allowing the introduction of the thermocouple tube and the stirring rod. The heat is then shut off and the temperature instrument is read at definite time intervals until allj the heat changes have taken place and the hard alloy is cooling at a uni- form rate. The readings are carefully recorded. In the study of a binary (two component) alloy this operation is repeated with a series of metal mixtures of varying composition. If the nature of the equilibrium curve is wholly unknown, mixtures are generally taken which vary in composition by 10 per cent, intervals from one pure metal to the other. These eleven points (this includes the melting points of the constituent metals) will usually indicate the general shape of the diagram, after which the necessary number of additional mixtures can be selected for study in the vicinity of the more essential points such as eutectics, intermetallic com- pounds or other characteristic features suggested by the preliminary survey. In simple cases a very few addi- tional mixtures will give all the information necessary, while in the more complex alloys 40 or 50 points are sometimes needed to establish the diagram. 28 PRINCIPLES OF METALLOGRAPHY Plotting the Cooling Curves. Various methods of plotting the experimental results are in use. 1 Of these methods the simple time-temperature curve is used more frequently than any other although the " inverse rate" curve has decided advantages when the heat effects dT ~3T FIG. 14. Inverse rate cooling curve. (T = temperature, t = time.) are small. With this latter method the ordinate rep- resents temperature as in the simple curve but the abscissa is the reciprocal of the rate of cooling, that is, the time necessary for the temperature to fall through a definite small interval (5 to 10 for example). The resulting curve has the form shown in Fig. 14 in which t = time in seconds and T = temperature in degrees. Constructing the Diagram. After plotting the in- dividual cooling curves, a sheet of coordinate paper of suitable size is selected and, using percentage composi- tions as abscissae and temperature holds on the cooling curves as ordinates, a diagram is constructed as indi- cated in Chapter I. The relative length of any hori- 1 See GULLIVER, "Metallic Alloys," Ed. 2, p. 175. LABORATORY METHODS OF METALLOGRAPHY 29 zontal lines that may be found in the set of cooling curves is next determined and these " time-lines" are also plotted as an independant but closely associated part of the diagram. Preparation and Microscopical Examination of the Polished Alloys. For the examination of the ordinary alloy, a specimen having a surface about ^ square inch is large enough. In special cases of defective material, as for instance, broken rails or similar articles, much larger pieces are needed. The small specimen for polish- ing is first obtained as a cubical or cylindrical piece by cutting it from the larger piece with a hack saw, or a power saw if available. Where the material is too hard to cut, as in the case of Duriron and similar high silicon alloys, the specimen is broken with a hammer and a fragment of suitable size selected for polishing. If only a very small piece of metal is available, it can be handled by placing it in a molten, readily fusible alloy and allow- ing the whole mass to solidify. The tiny fragment can then be polished with the larger mass of metal in which it is imbedded. The next step in the preparation of the specimen con- sists in filing the surface to be examined or grinding on a carborundum wheel. If the surface is originally very rough, two or more wheels may be used to advantage, each finer than the one preceding. The subsequent opera- tion of polishing consists in rubbing the specimen on abrasive materials of increasing fineness until a scratch- free, mirror-like surface is obtained. In polishing the specimen it should be turned through ninety degrees each time the change is made from one polishing surface to the next. The rubbing on each abrasive surface should be continued until the scratches produced by the next coarser surface (perpendicular to the direction of polish- ing) have been removed. The number of polishing 30 PRINCIPLES OF METALLOGRAPHY surfaces needed in the preparation of a specimen de- pends on the hardness of the material. With hard steel specimens a number of grades of abrasive material are needed each but little finer than the one preceding it, while with soft alloys like those of lead or tin very few intermediate grades of polishing material are required. For the average specimen the following sequence of abrasives will serve: coarse file, fine file, very fine emery wheel, emery papers (French emery "Marke Hubert" is best) 1C, IF, 0, 00, 000 and 0000. In using the finer grades of emery paper, from 00 down, a drop of oil should be applied to the surface of the specimen. From this point the abrasion is carried out by finely divided pow- ders used in the form of aqueous suspensions which are sprinkled from time to time on the finest quality of broadcloth or chamois skin. An excellent series of polishing powders has been prepared especially for metal- lographic purposes by the Norton Alundum Company of Worcester. Three powders that will answer for most purposes are 60 minute emery or carborundum, followed by "alundum F" and finally by "levigated alumina." Many metallographists prefer jeweler's' rouge as a final polishing agent rather than alumina. Rouge gives a brilliant final polish but has the disadvantages: (1) that it tends to cause a slight flowing of the surface metal, especially with the softer alloys; and (2) that it is a dirty and unpleasant abrasive with which to work. Polishing may be done by hand if necessary but the operation is a long and tiring one so that machine polish- ing is to be recommended if an apparatus is available. For hand polishing, the various grades of paper and the cloths which are to be used as bases for the polishing liquids are tacked to smooth boards (3 inches X 8 inches is a convenient size). Many excellent polishing machines are now obtain- LABORATORY METHODS OF METALLOGRAPHY 31 able. They differ somewhat in detail but all consist in general of rotating disks of wood or metal over which the abrasive paper or cloth is stretched. Machines hav- ing disks revolving in a horizontal plane are more con- venient to use than those of the vertical type. l The liquid suspension of abrasive powder maybe applied to the cloth by means of a camels hair brush, by a wash bottle or by the simple shaking bottle suggested in the sketch, Fig. 15. FIG. 15. Flask for suspended abrasives. Scrupulous care in the use of polishing papers and powders must be taken to avoid the transfer- of a coarse abrasive to a finer surface. A single particle of grit on the final polishing wheel may injure an otherwise perfect specimen so that repolishing is necessary. Etching is a selective chemical action which will effect one constituent of the alloy more than the other and its purpose is to develop the internal structure of the polished alloy. The number of etching solutions which have been used is very large and varies from simple reagents like dilute acids to extremely complex mixtures for special purposes. The special etching agents used in the examination of brass and bronze and of steels will be considered later in connection with the study of these important technical alloys. For general 1 A detailed description of mechanical polishing devices will be found in SATTVEUR, "Metallography and Heat Treatment of Iron and Steel," Ed. 2, p. 54 et seq. 32 PRINCIPLES OF METALLOGRAPHY purposes dilute acids (2 per cent, nitric or hydrochloric) or dilute alkalis (sodium hydroxide or ammonium hy- droxide) are used. One of the most successful general reagents is made by dissolving 10 grams of ferric chloride (FeCh) in 100 cc. of alcohol. It will be realized that etching is a chemical problem so that the particular reagent used, as well as its strength, depends on the chemical solubility of the components of the alloy. For this reason specific directions cannot be given. In most cases, however, the problem is one which is readily solved by a few trials if the effect of the reagent is watched under the microscope. Etching is most uni- form if the specimen is immersed face upward in a shallow dish containing the reagent. The formation of gas bubbles on the surface must be prevented by keep- ing the dish in constant rocking motion or by swabbing the exposed surface with a bit of absorbent cotton at fre- quent intervals. After etching, the surface is washed with water and with alcohol and is dried either with cotton or by a warm blast of air. Perfectly dried speci- mens may be kept in a desiccator for a long period with- out change. Specimens may be preserved indefinitely by coating them with cellulose acetate. For etching spots on a large surface or as a rapid method for determining the most suitable etching agent for a given alloy, a small swab of cotton moistened with the reagent may be used. The Microscope. 1 The polished and etched surface of the specimen must be examined by reflected light. If a low power objective is used (magnifying less than 25 diameters), a beam of light can be reflected on the surface in an oblique direction, passing below the objec- tive. As the magnification increases the distance between the objective and the surface becomes so small that 1 See SAUVEUR, " Metallography and Heat Treatment of Iron and Steel, Ed. 2, p. 67. LABORATORY METHODS OF METALLOGRAPHY 33 this method of illumination is impossible. In such cases the microscope must be provided with a " vertical illuminator, " the principle of which is shown in Fig. 16. The beam of light, coming from a small arc lamp, a nitrogen filled tungsten or other powerful light source, passes into the side arm of the illuminator and is reflected by a prism or plate down to the surface of the speci- FIQ. 16. Sketch of vertical illuminator (sheet glass type) dotted lines show the direction of rays reflected from the specimen. men, illuminating a spot on the surface so intensely that it can not only be examined but photographed. Such an illuminator can be fitted to any microscope but, where much metallographic work is to be done, one of the regular metal microscopes will be found much more convenient and effective. The magnifications most commonly used are 5x, lOx, 50x, 75x, lOOx and 200x. In rare cases, with extremely fine-grained structure, a magnification of 500x or even lOOOx may be necessary. As a matter of permanent record and for purposes of comparison of samples from different sources, photo- graphs of the etched surfaces are often made and all 34 PRINCIPLES OF METALLOGRAPHY modern metal microscopes are provided with a photo- graphic attachment. Photographing Metal Specimens. The beam of light A. The complete metallographic camera. B. Enlarged section showing the illuminator and microscope. FIG. 17. The Leitz metal microscope. from the polished specimens can be reflected to the eye- piece for visual examination or it may pass through the LABORATORY METHODS OF METALLOGRAPHY 35 bellows of a camera to a sensitive plate. The preceding cuts, Fig. 17 a and b, show the arrangement of the Leitz metallographic microscope. It is very similar to other metal microscopes and illustrates the general principles of them all. Plates. For making microphotographs the chief req- uisites of good . plates are color sensitiveness and fine grain. With steel samples in which the surface of the specimen is light, with black or gray markings, the in- expensive " Stanley" plates will be found satisfactory. For colored specimens, like brasses and bronzes for example, a more color-sensitive plate has to be used. The "Wratten M" plate gives excellent results. It has the disadvantage that it cannot be handled near the ordinary dark room light but must be manipulated either in absolute darkness or by a special "safe light." " Standard Orthonon" plates, also made by the Eastman Company, are effective both for steel and for brass photo- graphs. The author has found the "Wellington Ortho Process" plates very satisfactory for all purposes and they can be handled without danger near a fairly bright ruby light. In photographing colored metal surfaces a color screen of some sort and a piece of ground glass are generally placed between the illuminating lamp and the specimen to be photographed. The color screen in- creases the contrast between the components while the ground glass is to diffuse the light and prevent the reflec- tion of the hot carbons of the arc on the photographic plate. A yellow-green screen gives excellent results in photographing the yellow alloys of copper. Exposure of the Photographic Plate. The time of exposure may vary from a few seconds to ten or fifteen minutes depending on the source of light, the nature of the specimen, the color of the screen and the kindx>f plate used. The operator soon recognizes these factors so 36 PRINCIPLES OF METALLOGRAPHY that he can estimate the length of exposure with con- siderable accuracy. When the conditions are entirely unknown the following test method will save time and plates. Expose as usual for a few seconds. Instead of shutting off the light, as would ordinarily be done, push in the opaque screen, which is used to cover the plate in the holder, about one-half inch (^ in.), in order to shut off a portion of the exposed plate. After another short interval push in the screen another half inch. Repeat this operation until the opaque screen has been pushed into place in the holder. When the plate is developed it will show a series of bands each of which represents an exposure for a somewhat longer time than the one preceding it. Select the correctly exposed strip from the banded negative, record its time of exposure and use substantially the same length of exposure for photographs taken under similar conditions. Development of the Exposed Plates. Directions for the preparation of developing solutions will be found in each box of plates. Tray development requires much experience for complete success so that the use of a developing tank is strongly recommended. The opera- tion of the tank is simple and the results are positive and uniform. Of the many kinds of developer on the market it is probable that pyro-soda will give the best results for metallographic work. After the development is com- pleted, the plates are placed in saturated "hypo" (sodium thiosulphate) solution until the white opaque coating has dissolved and the negatives have become trans- parent. The clear negatives are then washed in run- ning water for an hour, after which they are placed in a drying rack and allowed to become perfectly dry and hard. Printing, Finishing and Mounting. Photomicrographs require the greatest possible detail in printing so that a glossy printing paper is commonly used. Glossy Velox, LABORATORY METHODS OF METALLOGRAPHY 37 Glossy Cyko or other papers of the same sort give excel- lent results. Printing directions are always inclosed with the paper. After the prints are washed a brilliant finish is obtained by placing them face down on a ferro- type plate. To keep them from sticking to the surface of the ferrotype plate it must be cleaned before use with a solution of beeswax in benzol or by a prepared cleaner such as the Ingento Polishing Compound. After the prints are in position on the plate the surplus water is removed by means of a print roller (squeegee) and they are then allowed to dry. The dry prints are then mounted on suitable cards for examination and filing. Dry mounting tissue is excellent for this purpose. Plates, paper, mounting tissue, etc., can be purchased of any dealer in photographic supplies. If many photo- graphs are to be taken printed cards of the general style shown in Fig. 18 will be found almost indispensable. In addition to the thermic and microscopic methods of alloy study, other properties of alloys are occasion- ally used in determining their constitution. Among these less important properties are electrical and heat conductivity, heat expansion, and magnetic effects. An excellent discussion of the connection between these different properties and the equilibrium diagram will be found in Desch, " Metallography," Ed. 2, p. 230. In closing this chapter on the methods of alloy study it must be stated that the mechanical testing of the metals for tensile strength, hardness, elongation and other physical properties is of great and constantly in- creasing importance to the practical metallographist. A discussion of this phase of the subject is beyond the scope of this book and the reader is referred to Rosen- hain, " Physical Metallurgy," in which the methods of mechanical testing are considered in detail. 463286 38 PRINCIPLES OF METALLOGRAPHY 3 u o CO -P 0> /AA AAA AA A/\AAAAAA/f /v/ Vvv\AAAAAAAA?v AAAAAA/ . AAAAAA/^oi%\AAAAAA7 , / W N A i AAA A h0( %Bi.lQ 20 y 40 50 60 70 80 Z FIG. 32. Method of showing the composition of ternary alloys. mixture. This is done most easily by the use of plotting paper with triangular coordinates. The classic example of the simple ternary is the series composed of lead, tin and bismuth. In the diagram, Fig. 32, the three corners represent the pure metals. The base of the triangle of which each metal is the apex represents therefore zero concentration of that element. For example, the line 1 An excellent discussion of Ternary Alloys will be found in GULLIVER, "Metallic Alloys," Ed. 2, p. 340, from which much of the following material is taken. THE ALLOY DIAGRAM AND ITS MEANING 01 Pb-Sn represents zero per cent. Bi, the line Pb-Bi zero per cent. Sn and Bi-Sn is zero per cent. Pb. Starting with these lines as bases and approaching the element whose percentage is desired, each line parallel to the base indicates one per cent, assuming, as is usually the case, that the triangle is divided by 100 parallel lines in each of the three directions. Point X in the figure there- fore represents 20 per cent. Pb, 50 per cent. Sn and 30 per cent. Bi. A study of the geometry of this figure shows two other facts which are of importance in the actual construction FIG. 33. Combination of the binary surfaces Pb-Sn, Sn-Bi and Bi-Pb. of these diagrams. A line drawn from any corner to any point on the opposite side represents an alloy in which two of the metals have a constant relationship to each other while the percentage of the third metal varies. Pb-Y, for example, is a line on which the relation of tin to bismuth is always 7 to 3. A line parallel to one side of the triangle represents an alloy in which one metal has a constant percentage while the other two vary. On WZ the per cent, of Sn is always 15 per cent, while the lead increases from zero per cent, at Z and the bismuth from zero per cent, at W. \ The next step in the construction of the model would 62 PRINCIPLES OF METALLOGRAPHY be to erect three plane figures on the three edges of the triangle corresponding to the three binary alloys. The sketch (Fig. 33) shows the appearance of such a space model. If lead is then added to the binary eutectic of lead and tin the melting point of the mixture is lowered. The same effect is noticed when tin is added to the lead-tin eutectic and when bismuth is added to lead-tin. Since these three ?t 250 225 200 175 150 FIG. 34. Ternary diagram with contour lines (CHAHPT). lines all slope downward their intersection must lie at a point lower than the melting point of any of the binary eutectics. This point is shown at E, Fig. 31 (p. 59) and is the ternary eutectic. The shape of the space model for an alloy of this type will be apparent from the consider- ations j ust outlined. The three lines connecting the single ternary eutectic with the three binaries form the lower edges of the three valleys made by the intersections of the three curved surfaces, composing the liquidus surface. If this surface is projected on its triangular base and THE ALLOY DIAGRAM AND ITS MEANING 63 contour lines drawn, representing the intersections of a series of parallel horizontal planes with the space model a figure of the shape shown in Fig. 34 results. These contour lines, which are naturally isothermals (lines of constant temperature), show that the surfaces above the binary eutectic valleys are convex. Consider, next, the changes which take place on cooling such an alloy as is represented by X in the diagram. As the temperature falls, pure lead separates along the line X- Xi, the relationship of bismuth and tin staying constant. At Xi enough lead has separated so that lead and tin are in the eutectic ratio and the two metals crystallize to- gether along the bottom of the binary eutectic valley until the composition E is reached, at which point the re- maining liquid solidifies as the ternary eutectic . Although , accurately defined, a eutectic should be of constant melt- ing point, the binary mixtures of eutectic composition vary in melting point with the amount of the third element. As a matter of easy statement, however, the mixtures of binary eutectic composition are universally referred to as binary eutectics. If a number of ternary alloys of Pb, Sn and Bi are studied and the points deter- mined at which the binary eutectics begin to form, planes drawn through these points will give the binary eutectic surface. This is found to consist of six twisted surfaces, each intersecting its neighbor in such a way that there will be three ridges, the binary eutectic lines, and three valleys, the projections of which connect the ternary point E with the three corners of the triangle. Since all alloys of the series become solid at the temperature of the ternary eutectic, the solidus surface is a horizontal plane through the ternary eutectic point E. In the actual construction of these ternary diagrams, the common practice is to study a number of vertical sections from which the space model or its projection can 64 PRINCIPLES OF METALLOGRAPHY be assembled. In the lead-tin-bismuth series, for example, a fairly complete study of the alloys of the compositions represented by the lines Pb-A, Sn-B and Bi-C, Fig. 34, gives a general idea of the shape of the model, and the necessity for further study in the area Bi-O-B is apparent. Referring again to Fig. 32, it will be seen that a series of alloys starting with the composition represented by X and with gradually increasing percentages of bismuth (sec- tion X-Bi) will give much additional information with re- gard to conditions in the eutectic area. This series might be followed by another section in which the percentage of bismuth is kept constant at 50 per cent., while lead and tin are varied. By studying several sections in this way it is soon possible to construct the space model accurately. Microscopic Appearance of Ternary Alloys. The microscopic study of these alloys is not satisfactory. The primary crystals are perfectly normal but the binary eutectic separations occur so slowly and over so consid- erable a temperature range that segregation generally takes place and the normal eutectic structure is lost. The ternary eutectic is, also, so finely divided and so intimate a mixture that the component elements can be found only by careful double etching and then with much difficulty. That three distinct structure elements are present, however, is shown in the photograph, Fig. 35. In the class of technically important ternary alloys are included Babbitt metal and nickel silver (formerly called German silver). The composition of Babbitt metal varies over a considerable range but it is usually an alloy of tin, antimony and copper. A common com- position is tin, 90 per cent., antimony, 7 per cent, and copper, 3 per cent. Microscopic examination shows the alloy to consist of crystals of SbSn and Cu 3 Sn imbedded in a tin matrix. Babbitt metal is an antifriction alloy and because of its comparatively low THE ALLOY DIAGRAM AND ITS MEANING 65 tensile strength is commonly used to line bronze bushings, the bronze giving the needed strength and the Babbitt the low frictional resistance. The microscope is of great use in the making of Babbitt lined bearings as by its aid it is possible to detect segregation in the metal and also to determine the size of the SbSn crystals. Large crystals, due to very slow cooling when the Babbitt is cast, and excessively fine crystals, due to a sudden chilling of the molten metal, both produce unsatisfactory bearing FIG. 35. Ternary diagram of lead, tin and bismuth. 75 X (HOMEHBERG). surfaces. A temperature of approximately 100C. for the mould has been found to give satisfactory results. Nickel silver is composed of copper, nickel and zinc. It has a wide use in the production of non-corrodible articles, table ware (with or without silver plating) , and the like. Alloys with four or more constituents are not un- common. Some of them find important application in the manufacture of easily melted fuse plugs for automatic sprinkler systems. Wood's metal corresponds to the 66 PRINCIPLES OF METALLOGRAPHY ternary eutectic of Bi, Pb and Sn to which a small amount of cadmium is added. It melts at 70C. The Phase Rule. 1 This chapter on equilibrium diagrams would not be complete without reference to the Phase Rule which, while it is of little use to the technical metallographist, has been a most valuable tool in the study of equilibrium diagrams, making it possible to state in any given case what the equilibrium conditions actually are. Knowing the conditions which should exist at equilibrium, the microscope makes it possible to decide at once whether or not equilibrium has been reached. The general statement of the phase rule is as follows : F = C + 2 -P in which F represents the number of degrees of freedom' C the number of components and P the number of phases. In the general case these factors are often difficult to define but in the application to metallic alloys no such difficulty is met. The components, C, are obviously the metals. The degrees of freedom are the changes which the alloy can undergo, namely, changes of temperature, con- centration and pressure. Since vapor can be neglected with most alloys and since the pressures commonly met in alloy practice are too small to have any appreciable effect, changes in pressure can be omitted, reducing the variables (degrees of freedom) to temperature and con- centration. A phase is denned as a homogeneous, physically distinct substance. In dealing with alloys, it may be a pure metal, a metallic compound or a solid solution. In addition, each physical state of the sub- stance, whether solid, liquid or gas constitutes a separate phase. Neglecting the vapor phase and the effect of pressure, 1 FINDLEY, "The Phase Rule and Its Applications." THE ALLOY DIAGRAM AND ITS MEANING 67 the Phase Rule for alloys may be reduced to the simple form F = C + 1 -P The number of components in a binary alloy is 2, so the expression is still further simplified and takes the form F = 3 -P A concrete illustration of the use of the Phase Rule is given in the following diagram, Fig. 36. A point at r : Divariant (Temp, and cone, may change ) Monov'ariant (Temp, or cone) may change but not both) Non-variant .(Neither temp. nor. cone. may change) I Triple point (Liquid and two Solids may exist here) Concentration FIG. 36. Phase rule diagram for binary alloys. X lies in the liquid phase. Substituting 1 in the sim- plified expression it becomes, F = 3 1 = 2. In words, the alloy now possesses two degrees of freedom. Both temperature and composition can be varied within the area bounded by AEB and the alloy will stay molten. This field then represents an area of divariant equilibrium. At the point Xi on the line AE, the crystal is beginning to separate but is in contact with the liquid. Two phases, crystal and liquid, are present and the expression becomes, 68 PRINCIPLES OF METALLOGRAPHY F = 3 2 = 1. The alloy, now, has only one degree of freedom, for any change in temperature is accompanied by a change in concentration along the line AE, and a change in concentration necessitates a change in tempera- ture. The lines AE and BE are therefore lines of mon- ovariant equilibrium. At the point E, the eutectic point, or at any other point on the eutectic line, CED, two crystal phases, A and B, are in contact with a liquid phase of composition E. (Solid E contains A and B in the form of very fine crystals.) Under these con- ditions, the expression becomes F = 3 3 and the system becomes non-variant. Neither the temperature of the mixture nor the composition of the three phases can change until one of the three phases has disappeared. The temperature cannot fall until all the liquid phase, E has solidified nor can it rise without the disappearance of either A or B. It must not be understood that the mixture can have only the composition represented by the point E. It may have any composition along the line CD but, in this case, the composition of each phase remains the same, the difference in original composition producing changes in the relative amounts of the three. The line CD, therefore, is a line of non-variant equilibrium. One of the principle uses of the Phase Rule is to deter- mine whether or not true equilibrium has been reached. It is evident that, since an alloy with less than zero degrees of freedom is an impossibility, there can never be more than two crystal phases in contact with, or separat- ing from, a two component liquid metal. Therefore, in a case like that indicated in Fig. 26, p. 53, and illustrated in Fig. 28, p. 55, the presence of three phases, in the micro- scopic section of the solid alloy, is a positive indication of incomplete equilibrium. CHAPTER IV THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE No attempt will be made to describe or even to name the large and constantly increasing number of non- ferrous alloys used in practice. Certain of them are of such great technical interest and importance, however, that their properties must be considered. Many have been referred to in connection with the equilibrium dia- grams and others are of such special character that their consideration is out of place here. The majority of the important non-ferrous alloys not yet discussed fall into one of two groups; the smaller, containing aluminum and its alloys, the larger, the alloys of copper, particularly the brasses and bronzes. Aluminum Alloys. While the commercial develop- ment of the alloys of aluminum has produced many alloys of light specific gravity coupled with valuable mechanical properties, it has also very great possibilities for investigation, notably along the lines of alloying the aluminum base with two or more other metals. Pure aluminum, as cast, has a tensile strength of about 14,000 pounds per square inch, but by cold work, as in wire drawing, this may be increased to nearly 50,000 pounds. The commonest casting alloy is that with a com- position of 92 per cent, aluminum and 8 per cent, copper which has a tensile strength of about 20,000 pounds but is more readily corroded than aluminum itself. The most valuable alloy is that known as Duralumin, which is aluminum containing from 3.5 per cent, to 5.5 per cent, copper, 0.5 per cent, to 0.8 per cent, manganese and 70 PRINCIPLES OF METALLOGRAPHY about 0.5 per cent, magnesium. It has several remarkable properties and is the only aluminum alloy which has been successfully heat treated. As cast, the alloy has a tensile strength of about 35,000 pounds per square inch and an elongation of 17 per cent, in two inches. If it is heated to 400-500C. and then quenched, there is no great change in the physical properties. If, how- ever, the quenched alloy is allowed to "age" for a few days, the tensile strength will increase to about 58,000 pounds and the elongation to 23 per cent. By hard rolling, duralumin may have its tensile strength increased to 85,- 000 pounds per square inch. In sheets, tubes and similar articles it breaks at about 50,000 pounds. Weight for weight, duralumin is as strong as the best steel and for the same strength has greater rigidity. Unlike other aluminum alloys containing copper it is markedly re- sistant to corrosion and compares favorably with copper under similar conditions. The properties of duralumin make it a most useful alloy for the construction of various articles especially in airplane and motor parts. 1 The preparation of aluminum alloys for metallographic examination presents some difficulties. The preliminary polishing should always be done on emery papers mois- tened with oil and the final polishing cloths must never be allowed to dry. Liquid abrasives are absolutely neces- sary for successful results. The structure of the polished alloys is developed by immersion in dilute hydrofluoric acid followed by treatment with nitric acid. For detail work a 0.10 per cent, solution of sodium hydroxide in 50 per cent, alcohol is recommended. Copper and Its Alloys. Enormous quantities of pure copper are used in the electrical industry because of its 1 An excellent discussion of aluminum alloys is given by MERICA, Chem. and Met. Eng., 19 (1918), 135 and Bull. 76, Bureau of Standards, April, 1919. THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 71 very high conductivity. Copper forms solid solutions with many of the common elements, including practically all of those with which it is associated in its production from the ore. A fact of great technical importance is the remarkable effect of the existence of solid solutions on the conductivity of an alloy. If the alloy is of the eutectic type, the elements are completely insoluble in each other in the solid state and the conductivity of the solid alloy is practically the sum of the conductivities FIG. 37. Relation between solid solution and electrical conductivity. A. is the equilibrium curve and B the curve of conductivity. of the component metals. When solid solutions occur, the conductivity drops off very sharply, the correspond- ing curve taking the form of a steep sided U as in Fig. 37, in which A is the equilibrium curve and B the curve of conductivity corresponding to it. The rapid decrease in conduction due to slight addition of the dissolving element makes evident the harmful effects of even small percentages of dissolved impurities on the conductivity of the copper and the necessity of accurate analysis if the metal is to be used for electrical work. Oxygen dissolves in copper to an appreciable extent and also unites with it to form cuprous oxide, Cu 2 0. The com- 72 PRINCIPLES OF METALLOGRAPHY pound then reacts with the copper to form a eutectic series of alloys with a eutectic at 3.5 per cent. Cu 2 (0.39 per cent, oxygen). Since the presence of this eu- tectic has a harmful effect both on the electrical and mechanical properties of copper, all high grade copper is deoxidized in the process of manufacture. This may be done by the use of phosphorus, silicon, boron and prob- ably, other readily oxidized elements. Deoxidizing by means of silicon in the form of copper silicide gives copper with high conductivity. Boron is highly effective as a deoxidizer and boronized copper is not infrequently specified for electrical work. For general industrial purposes, copper is used in the form of rolled sheets, tubes, bars and drawn wires. These mechanical operations 1 have such marked effects not only on the physical but on the metallographic proper- ties of copper that they will be considered in some detail. Copper as cast has a tensile strength of from 17,000 to 20,000 pounds per square inch and its ductility is indi- cated by an elongation of from 40 to 50 per cent, in two inches. It is possible by simple mechanical work, such as rolling or drawing, to increase the tensile strength to almost 50,000 pounds per square inch. This increase in tensile strength is accompanied by a great increase in hardness and a marked decrease in elongation. Hard brass wire may have an elongation of only 1 or 2 per cent. Several theories have been proposed to account for this phenomenon of hardening by means of mechanical work, a phenomenon which is by no means confined to pure copper but is a general property of metals and alloys. The most logical theory and the one which gives the most adequate explanation of the known facts is known 1 For a description of the mechanical testing of alloys and of the effects of work see ROSENHAIN, "Physical Metallurgy," Chapter XI. THE NON-FERRO US ALLO YS OF TECHNICAL IMPORTANCE 73 as the amorphous cement theory which was proposed by Beilby and has been carefully studied by Rosenhain FIG. 38A. Moderately worked Muntz metal (Cu (O'Daly.) -Zn 40%). 75 X. 1 m FIG. 38B. Spe ;r hard drawing. and his associates. This theory, now generally accepted, assumes that when a metal is subjected to such an amount of strain, either tensile or compressive, that its 74 PRINCIPLES OF METALLOGRAPHY elastic limit is exceeded, the crystals become elongated not by a simple stretching of the material but by a slip- ping taking place along certain of the crystal planes. FIG. 40. That the elongation of the crystals is very marked under certain conditions is shown in the photomicrographs, Figs. 38, A and B. The changes taking place within the 7 HE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 75 crystal are shown in Figs. 39 and 40 sketched from Rosenhain's photographs which represent a piece of soft iron before and after straining. The dark lines crossing the crystals in Fig. 40 are due to the fact that the surface is no longer plane but is covered with a large number of 8 S * S , D A Before straining B After straining FIG. 41. Sketch showing the way in which slip bands are produced in strained metal. (Rosenhain.) microscopic ridges formed by the displacement of many crystal layers. These dark lines, or bands, are called slip bands and are characteristic of overstrained metal (see also Fig. 48, p. 87). The sketches, Figs. 41 and 42, illus- FIG. 42. Sketch showing the optical reason for the appearance of slip bands. (After Rosenhain.) trate the probable nature of slipping and the optical reason for the dark slip bands. Rays "A" (Fig. 42) striking the horizontal surfaces are reflected back into the eyepiece and produce light bands. Rays "B" strike the oblique surfaces, are reflected out of the field and cause the black 76 PRINCIPLES OF METALLOGRAPHY lines or bands. Granting that slipping does occur along crystal planes, it is easy to believe that the rubbing of the surfaces wholly destroys the crystalline character of an extremely thin layer of the metal, producing what has been called an amorphous cement or amorphous binding material between the displaced layers. Since it may be assumed that the slipping would first occur along those planes where the crystalline cohesion was least, it would naturally follow that a greater tension would need to be applied to cause additional elongation. This fact, together with the belief that the amorphous material is both harder and stronger than the crystalline form, would account for the increased hardness and ten- sile strength of material, which has been subjected to mechanical work. As the amorphous material has no planes along which slipping can take place to relieve an imposed strain, it follows that a sudden shock is apt to cause it to fracture. The same fact limits the amount of mechanical work which can be done on a metal before it becomes necessary to cause a partial recrystallization of the amorphous material by annealing. The annealing of metals having an excessive amount of amorphous material has the function of decreasing the brittleness and reducing the strain hardness. Etching of Copper Alloys. The most effective reagent for alloys rich in copper is ammoniacal hydrogen peroxide. For ordinary brasses containing 70 per cent, copper and 30 per cent, zinc the proportions should be approximately one part of 3 per cent, hydrogen peroxide to five parts of strong ammonia (sp. gr. 0.90). The amount of hydrogen peroxide must be increased with higher percentages of cop- per and decreased as the copper decreases. The ratio of peroxide to ammonia should be about 1 to 10 for use with alloys of the Muntz metal type (Cu 60 per cent., Zn 40 per cent.). The reagent does not keep and it is THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 77 essential for success that it should be prepared immediately before use. The mixture is applied to the polished sur- face by swabbing with cotton soaked in the liquid, after which the surface is washed in running water and again treated for a few seconds with the etching reagent. Alternate treatments with alkaline peroxide and water will soon develop the crystal structure in such a way that the finest details become visible. This method of etching requires [a little experience but the results will be found to repay any time spent in practice with it. Dark, overetched surfaces are commonly due to excess of ammonium hydroxide, while surfaces lacking in detail are usually caused by too much peroxide although this lack of contrast may be due to severe overstrain, a-brass (p. 82) is colored buff or brown by the mixture while 0-brass (p. 82) is generally yellow. Ammonium persulphate, (NH^SaOg, in strong am- monia (1 gr. per 20 c.c. ) is often used to identify /3-brass which it attacks more readily than it does a-brass. Nitric acid (sp. gr. 1.20) is used for rapid development of the crystal structure. The resulting etched surface shows strong contrast but the details are not so sharply denned as with the peroxide and the alloy is not so satisfac- tory to photograph. Ferric chloride solution (p. 32) is effective for use with arsenical brass. Bronze. Copper-tin. The equilibrium diagram of the copper-tin alloys is very complex as Fig. 43 shows, con- sisting of one definitely established compound Cu 3 Sn and five solid solutions in which it is probable that the compounds CuSn, Cu 5 Sn 2 , Cu 4 Sn and Cu 5 Sn exist, al- though this fact has not been fully established. The important tin bronzes are practically all included in that section of the diagram in which the percentage of tin is less than 30 per cent, or, in other words, the desirable 78 PRINCIPLES OF METALLOGRAPHY properties are associated with the a and /3 crystals. Four classes of bronzes are of importance. 1. Coinage bronze containing 96 to 92 per cent, copper is used largely in the production of "copper" coins and medals, the small amount of tin present increasing the hardness and wearing qualities of the copper. 100% Sn Composition FIG. 43. Copper-tin diagram. 2. Gun metal and gear bronze vary in composition from 92 to 88 per cent, copper. Gun metal is no longer used in the manufacture of ordnance but is often used where strong, heavy castings are to be made. In order to increase the fluidity of the metal and make the casting operation simpler, a small amount of zinc is frequently THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 79 added. A standard alloy of this class has the composi- tion 88 per cent, copper, 10 per cent, tin and 2 per cent, zinc. This is usually known as Government Bronze, " G " Metal or simply "88, 10 and 2." It has a fairly high tensile strength, 3238,000 pounds per square inch, and is often used where pressures, steam or hydraulic, are to be met, or for bearings subject to heavy loads. Another bronze of the same class consists of 89 per cent. Cu and 11 per cent. Sn and is very generally used, under the name of English gear bronze in the manu- facture of heavy gears. A brilliant, mirror surface is developed at the contact between the teeth of the gear and the driving mechanism and an excellent bearing and wearing surface results. 3. Bearing Bronzes. These alloys vary from 87 to 81 per cent, copper and usually contain one or more elements in addition to the tin. The best example of this class is the bearing bronze to which phosphorus (in the form of phosphor copper) and lead have been added. These alloys, called phosphor bronzes, are of two classes, those to which phosphorus is added only as a deoxidizer and those in which an excess is present to act as a hardener. In the first class the function of the phos- phorus is simply to increase the strength and ductility of the alloy by removing the Cu 2 eutectic and other oxides. In many cases, in spite of the marked improvement in physical properties, the actual amount of phosphorus is negligible. In the second class, the phosphorus, even though present in small quantities, usually less than 1 per cent., forms extremely hard particles of Cu 3 P, too brittle in themselves to be used in a bearing but forming and excellent non-abrasive skeleton in the strong, tough bronze matrix. Lead is frequently added to bronze in small amounts and, as it is an insoluble constituent, is found fairly uniformly distributed throughout the metal 80 PRINCIPLES OF METALLOGRAPHY in the form of drops, Fig. 44, A. Lead gives to the metal two valuable characteristics. It makes it more easily A. Unetched bronze showing lead drops. B. Etched to show dendritic structure. FIG. 44. Phosphor bronze with 4 per cent, of lead. 75 X. machined and, to a certain extent, self lubricating because of the soft, greasy character of the suspended THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 81 lead. The lead drops are sources of weakness in an otherwise strong metal so that the amount x>f lead must be carefully adjusted to fit the conditions under which the bearing is to be used. For most purposes the lead con- tent is less than 2 per cent., although in the plastic bronze, previously mentioned (p. 9), it may reach 50 per cent. 4. Bell Metals. These alloys contain from 80 to 75 per cent, copper and are of especial interest because they must be worked either at a temperature above dull redness or the hot metal must be suddenly chilled (quenched) and then worked cold. Reference to the diagram (Fig. 43) will show that work is done in both cases on the jS-solid solution. In the first case, the work is done while the alloy is in the /3-temperature range and, in the second case, the sudden chill retains the bronze in the condition in which it existed at the higher temperature (seep. 57). With the increase in tin to more than 25 per cent., the brittleness becomes so great that the alloys are handled only with difficulty and are used exclusively for decorative purposes where the material is not subjected to strain or shock. Brass. The copper-zinc alloys are the most important of the copper alloys because they are relatively inexpen- sive as compared to tin bronze. The diagram for the brasses is, like that of the bronzes, very complex, consisting of six series of solid solutions which probably contain, as in the other case, definite compounds (Fig. 45). The 7-solid solution, which begins to 'be formed when the percentage of zinc is increased above 50 per cent., probably contains the compound Cu 2 Zn 3 and is so brittle that alloys in which it occurs are practically valueless except for decorative purposes where strength and duc- tility are not required. This limits the technically im- portant brasses to three classes: a-brass, from to 36 82 PRINCIPLES OF METALLOGRAPHY per cent, zinc; a + |8-brass, from 36 to about 42 per cent, zinc ; and /3-brass, from 42 to about 50 per cent, zinc, at which point the tensile strength and ductility both drop to nearly zero. The following curves show the relation- ships between tensile strength properties and metallo- graphic constitution of cast copper-zinc alloys (Fig. 46). It will be seen that the strength of cast brass increases from about 28,000 pounds per square inch with pure copper to more than 60,000 pounds per square inch with Composition FIG. 45. Copper-zinc diagram. (Shepard.) 45 per cent, zinc (pure 0), so that by choosing the com- position, any desired strength within these limits may be obtained. It must be clearly understood, at this point, that the figures just given are not absolute, but relative, as the physical properties vary within fairly wide limits, even with cast material, depending on various factors such as the shape of the cast piece, the material of which the mould is made and the rate at which the metal cools. THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 83 The actual number of commercial brasses is very great but, for convenience, they may be grouped in a few classes. The cost decreases with the increase in the per- centage of zinc so that very high copper alloys are not much used. 1. Gilding Metal or Jewelry Brass. This contains from 1 to 20 per cent, zinc and is used under various trade names in the manufacture of cheap jewelry. The color of some of the alloys in the group is not unlike that of standard gold. 100 j{ Cu 95 86 55 50 45 75 70 65 Composition FIG. 46. Relation between chemical composition of copper-zinc alloys and tensile strength. (After Johnson J. Inst. Metals, xx, 233.) 2. Dutch Metal. These alloys contain between 20 and 25 per cent, zinc, are very malleable and are used largely in the hammered form as substitutes for gold leaf. 3. Brass for Cold Working. The group in which the zinc varies from 27 to 35 per cent, includes by far the larger number of the technically important brasses. This range of compositions lies at the zinc rich end of the a-brass field and, therefore, includes the alloys of high tensile strength coupled with maximum ductility. In 84 PRINCIPLES OF METALLOGRAPHY this class are found the alloys used for sheet metal, tubes, wire, cartridge cases and other articles which are to be subjected to severe mechanical work. 4. Muntz metal and similar alloys contain from 37 to 45 per cent, zinc and include the a and /3- and the pure ^-brasses. Pure /3-brass does not exist in copper-zinc alloys which have been cooled slowly from a high tem- perature but may be obtained as a perfectly homogene- ous solid solution by quenching an alloy with 60 per cent, copper from a temperature of 800C. Owing to the fact that normally cooled alloys of the Muntz metal type are composed of two components, a and ft, they are usually worked hot so that the metal may be in its homo- geneous condition. Pure /3-brass shows little or no tendency to twin even after it has been worked and annealed. It is colored yellow by NH 4 OH and H 2 O 2 and may be distinguished in this way from a-brass which under similar treatment becomes distinctly brown. A mixture of NH 4 OH and ammonium persulphate is often successfully used in the examination of Muntz metal as the /3-brass is colored yellow while the a-brass is practically unaffected by a short treatment with this reagent. The characteristic appearance of a-brass in a ground-mass of untwinned /3-brass is shown in the follow- ing photographs, Fig. 47, A, B, C, and D, the differ- ences in the size of the a-masses being due to differences in heat treatment and chemical composition. Alloys of the Muntz metal type are fairly resistant to corrosion by salt water if the a and ft crystals are small and intimately mixed. They are used to a con- siderable extent in the sheathing of wooden ships, for condenser tubes and for other purposes where the lesser ductility is not a serious objection. Because of the cheapness of zinc the Muntz metal alloys are sometimes substituted for the more expensive a-brass. These THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 85 A. Cast Muntz metal (150 X). B. Muntz metal annealed at 750 and quenched. Chiefly /3-brass. 75 X. (Johnson and Jermain.) C. Muntz metal annealed at 750 and D. Muntz metal annealed at 750 and cooled ooled in air. 75 X . (Johnson and Jermain.) in furnace. Shows brown islands of a-brass in a matrix of /3-brass. 75 X- (Johnson and Jermain.) FIG. 47. Muntz metal. 86 PRINCIPLES OF METALLOGRAPHY alloys may be made much less readily corroded by the addition of about 1 per cent, tin as in Naval Brass. 5. Brass Solder. For brazing iron, the solder con- tains 35 per cent, of zinc while, for soldering brass, the alloy of 50 per cent, is most commonly used. 6. White Brass. When the percentage of zinc is more than 50, the resulting alloys become increasingly light in color and are very brittle. These alloys are known as the white brasses and are used only for orna- mental castings. a-Brass. The most important single alloy is that whose composition is very near 70 per cent, copper and 30 per cent. zinc. It possesses great ductility, about 56 per cent, elongation and a tensile strength when cast of 30,000 to 35,000 pounds per square inch, a strength which is greatly increased by mechanical work. One of the most important uses of this " 70-30" brass is in the production of shell or cartridge cases which, in the process of manufacture, are subjected to severe mechanical work. Practically all brass shell cases from those used for the small revolver to the large cases used in naval guns are made by'a series of punching and draw- ing operations. As was stated on p. 72, this working produces an elongation of the crystal grains and a marked hardening and increase in brittleness of the metal. The following photograph, Fig. 48, is of interest in showing the value of metallographic as well as chemical control of cold-worked brass. The chemical analysis is excellent but the strain has been enough to produce great distor- tion of the crystal grains and the production of numerous slip bands. This strained condition can be wholly relieved by suitable annealing. The striking feature of cold-worked a-brass which has been annealed is the pro- duction of twin crystals characterized by alternate dark and light bands (Fig. 50, C) . These twins may be so small THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 87 as to be scarcely visible at a magnification of seventy-five diameters or they may be so large that with the same FIG. 48. Badly strained a-brass showing distorted crystals and many slip bands. 75 X. i-< i . .j I-?* ^ -**, *- * : '*V: '-; FIG. 49. Cartridge brass as cast. 75 X. n agnification a single crystal may cover the entire field of vision of the microscope. These differences in crystal 88 PRINCIPLES OF METALLOGRAPHY size are shown in Figs. 49 and 50 A, B, C, D, which are all a-brasses of the same chemical composition (70 per cent, copper and 30 per cent, zinc), but differ in the amount of mechanical and heat treatment. Although it may be said in a general way that the higher the anneal- ing temperature the larger the crystal grains, other con- ditions being equal, it is also true that the crystal size depends on the amount of cold work to which the brass has been subjected. 1 This means that for each sample of cold- worked brass there is an annealing temperature which will produce crystals of the desired size, this tem- berature depending on the extent to which the brass has been deformed by mechanical work. Fine crystal grains indicate increased tensile strength and hardness with de- creased ductility while large crystals are always accom- panied by softness, lower tensile strength and greater ductility. This connection between crystal size and physical properties has led to the introduction of definite grain size requirements in many specifications, not only for brass and bronze but also for steel. A convenient means of studying grain size has been proposed by Jeffries 2 and is recommended by the American Association for Testing Materials. The method consists in projecting the mag- nified image of the specimen onto a ground glass plate on which has been inscribed a circle 79.8 millimeters in diameter (area = 5000 sq. mm.). The ground glass is placed with its ground surface toward the specimen and on the outer, smooth surf ace the number of whote crys- tals included in the circle is counted. This may be done conveniently by checking each crystal with a soft (glass) pencil. The number of grains intersecting the circum- 1 MATHEWSON and PHILLIPS, Am. Inst. Min. Eng., Feb., 1916. 2 ZAY JEFFRIES, "Grain Size Measurements," Met. and. Chem. Engr., vol. xviii, p. 185. THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 89 A. Severely cold worked cartridge metal be- ginning to recrystallize. 75 X. B. Cold worked and annealed. Small twin crystals are visible. 75 X. C. Moderately cold worked cartridge metal D. Moderately cold worked cartridge brass fter annealing. 75 X- (The characteristic annealed at 700 C. 75 X. (Large twin cry s- winned structure is very marked.) tals due to overheating.) FIG. 50. Brass 70 % copper, 30 % zinc. 90 PRINCIPLES OF METALLOGRAPHY ference of the circle is then counted and 0.5 of this number, added to the number completely included in the circle, gives a close approximation to the total number of crystals present. To obtain the number of grains per square millimeter, the crystal count is multi- plied by a factor which depends on the magnification used. The standard magnifications, as recommended by the American Society for Testing Materials, are, for steels, 50-100-250 and 500 diameters and for non-ferrous alloys, 25-75-150 and 250 diameters. The multiplying factors are given in the following table. Diameter of circle in milli- meters Magnification used Multiplying factor to obtain grains per square millimeter 79.8 10 0.020 79.8 25 0.125 79.8 50 0.500 79.8 75 1.125 79.8 100 2.000 79.8 150 4.500 79.8 250 12.500 79.8 500 50.000 If the grain size is to be expressed as the average diameter of the crystal in millimeters, or its area in ju 2 , the following formulas from Jeffries paper will be found useful : z = completely included grains; w = boundary grains; x = equivalent number of whole grains in 5000 sq. mm. (circle 79.8 mm. in diameter or rectangle with area of 5000 sq. mm.) ; x = ~ w + z. m = magnification; / = multiplying factor used to obtain grains per THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 91 square millimeter (see table) ; / = n = number of grains per sq. mm.; n = fx : d = 1,000,000 d = diameter of average grain in mm. : d = ~F Vn a = area of average gram in M 2 - a One of the most serious defects in worked brasses and bronzes is the strained condition leading to the formation of what are known as season cracks. Various articles of cold worked brass may be so badly strained and so imperfectly annealed, that storage for a period of from several weeks to a number of months, particularly in a moist climate, leads to a spontaneous break down of the strained metal and the production of large or small season cracks. It happens, frequently, that this dan- gerous condition is not at all apparent even on careful inspection. Because of its comparative frequency, it has become the custom in many instances to insist on a test for season cracking with a specified number of samples from an entire lot. This can be done effectively by immersion of the suspected sample for 4 hours, in a \Y^ per cent, solution of mercuric chloride or mercuric nitrate. The season cracking phenomenon is greatly accelerated by this treatment and the tendency to crack at once disclosed. It is essential that a-brass which is to be exposed to severe mechanical treatment should be free from bis- muth, antimony, iron and lead as all are sources of weakness. For brass which receives only a moderate treatment, small percentages of iron or lead will not be dangerous. Iron gives to the alloy increased strength coupled with increased hardness and decreased ductility, while lead acts, as it does with bronze, to reduce the 92 PRINCIPLES OF METALLOGRAPHY tensile strength but to make the brass far more readily machined. Bismuth and antimony tend to form brittle envelopes around the a-crystals and to destroy the duc- tility that makes cold work possible. For special purposes, small percentages of other ele- ments are added to the brass alloys. The use of lead has been mentioned as improving the machining quali- ties though reducing the tensile strength, but only when the percentage of lead is very low (less than 0.5 per cent.) can the brasses be worked hot. Tin, when added in small quantities increases the hardness of brass but causes a marked decrease in ductility. When added to 70-30 brass in amounts from 1 to 1.5 per cent., an alloy which is very resistant to sea water, Admiralty metal, is formed. The addition of manganese in amounts less than 4 per cent., gives to the brasses very desirable properties. It is usually added to brasses of the Muntz metal type, together with small amounts of tin, iron and aluminum. An alloy of this general type, called, unfor- tunately, manganese bronze when it should be man- ganese brass, is much used in making propeller blades, rudders, ship fittings exposed to sea water, and for other engineering purposes requiring a strong, non-corrodible alloy. Aluminum, when added in very small amounts, increases the fluidity of molten brass to a marked degree, rendering the casting operation easier and producing cleaner castings. It materially increases the strength of the brass but rapidly reduces its ductility so that the amount added should never exceed 3 per cent. Aluminum Bronze. Another technical copper alloy of great importance is the alloy with aluminum known as aluminum bronze. The diagram, Fig. 51, is not unlike the tin-bronze and brass diagrams in its general character and complexity. The only alloys of technical importance, however, lie in the a-field (copper solid THE NON-FERROUS ALLOYS OF TECHNICAL IMPORTANCE 93 solution varying from to about 11 percent, aluminum) and, at the opposite side, in the T? field, which is a series of solid solutions of copper in aluminum, saturated at about 10 per cent, copper. The light alloys of aluminum with copper have already been mentioned, p. 69. The alloys in the a-field have remarkable physical properties, the addition of aluminum causing a striking increase in tensile strength. In castings, for example, while 30 per s,w SspQZ L2J3B32 FIG. 51. Aluminum bronze diagram. (Carpenter and Edwards, Gwyer, Curry.) cent, 'zinc gives a brass with a tensile strength of about 30,000 pounds per square inch and 10 per cent, tin will give a bronze with about 40, 000 pounds tensile strength, the addition of 10 per cent, aluminum to copper gives an aluminum bronze with a strength of about 70, 000 pounds per square inch. Its elongation is about 20 per cent. , nearly as great as that of brass and twice as much as that of tin-bronze and the alloy is considerably harder than either. Aluminum bronze is used in the manufacture of castings requiring strength and toughness, and is espe- 94 PRINCIPLES OF METALLOGRAPHY cially resistant to shock or to alternating stresses. It has the added advantage that it is from 10 to 15 per cent, lighter than the corresponding brasses and tin bronzes. This alloy would be more extensively used if it were not for certain difficulties in its manufacture which are apt to cause lack of uniformity in the finished product. Properly made aluminum bronze is a very valuable alloy. l 1 Extended discussion of the technical non-ferrous alloys will be found in LAW, "Alloys and Their Industrial Applications;" GULLIVER, "Metal- lic Alloys;" DESCH, "Metallography." References to current practice will be found in the "Institute of Metals" (British) and in the "American Institute of Metals" which has recently become affiliated with the "Institute of Mining Engineers." CHAPTER V IRON AND STEEL The most important applications of metallography, as well as the most difficult, lie in the uses, defects and methods of heat treatment of iron and steel. The diffi- culties of the study of this series of iron-carbon alloys may be traced to various causes, among them the fact that in the technical study we are dealing, even in what are known as the " plain carbon-" steels, not with a simple alloy of iron and carbon but with an exceedingly complex mixture of iron, carbon, phosphorus, manganese, sulphur and silicon. While the effects of the last four elements, when they are present in small quantities, as is usually the case, are not comparable with the effects produced by comparatively slight changes in carbon content, no one of the constituents can be wholly neg- lected. Increase in any one of them, above a certain well- recognized maximum, causes far-reaching changes in the physical and metallographic properties of the metal. When elements like chromium, nickel, vanadium or tungs- ten are added in making the alloy steels (self hardening, high speed tool steel, etc.), the situation becomes so complex that little is known from an equilibrium standpoint. An enormous amount of work is yet to be done in sys- tematizing the present knowledge of the properties of the alloy steels. A second factor which complicates the exact study of the iron-carbon diagram is that the iron exists in various allotropic forms, each one of which has different physical properties, notably magnetic properties and each one of 95 96 PRINCIPLES OF METALLOGRAPHY which varies in its ability to dissolve carbon. During the cooling of chemically pure iron (electro-deposited) five or six holds in the curve have been noted by various investigators but, because of uncertainties as to the possible effects of dissolved gases, it is generally assumed that iron exists in three allotropic forms; (1) 7-iron, stable above 900 and practically nonmagnetic; (2) /3-iron, existing between 780 and 900; and (3) -iron, stable below 780 and strongly magnetic. 1 The changes just indicated occur on cooling the iron and, from the initial of the French word, "refroidessement," are often referred to as the Ar 3 , Ar 2 and ATI points, respectively. The changes take place at slightly higher temperatures on heating and, in abbreviation of the word, "chauffage, " are often called the Ac 3 , Ac 2 and Aci points. They are also called critical points or transformation points. As the percentage of carbon is increased the Ar 3 , Ac 3 , Ar 2 and Ac 2 are progressively lowered until at the tempera- ture of 780 to 790C. and with the carbon content in- creased to 0.85 per cent, the three critical points coincide. The temperature range which includes all the critical points is called the critical range and is of great impor- tance in the annealing of steel. This temperature range extends from about 800 to 900 with low carbon steels, decreases to an interval of 10 (from 780-790C.) at the composition 0.85 per cent, carbon and then increases from this point as the carbon content increases. A third complication in the iron-carbon system is due to the fact that many of the reactions involved occur in the solid state and therefore, unless the alloys cool with extreme slowness, the changes take place X A group of English metallographists question the existence of /8-iron and consider only the a and y forms. The great majority of metallog- raphists, however, still agree that thermal evidence warrants belief in the /8-form. IRON AND STEEL 97 incompletely or not at all. It is also true that in the higher carbon range, from 4.3 per cent. C upward, the equilibrium relationships have never been satisfactorily settled. The enumeration of these various difficulties might lead to the impression that a study of the metallography of iron and steel is a hopeless task. This is far from true as many of the difficulties have been overcome and the study has been carried on along so many lines that information of the utmost value to the makers and users of iron and steel has been obtained. It is also true that so much remains to be done that there is an unlimited field of investigation for those who have the opportunity and the inclination to carry on the work. It is beyond the scope of this book, nor is it its purpose, to consider except in a general way the various branches of the metallography of iron and steel. For detailed information the reader is referred to one of the larger books dealing solely with this side of the subject. 1 Classification of the Iron Carbon Alloys. Dealing first with that group of alloys consisting chiefly of iron with varying amounts of carbon, the metals are divided into three groups, depending on the carbon content and method of manufacture; (1) wrought iron, (2) steel and (3) pig or cast iron. Wrought iron contains normally less than 0.3 per cent, carbon and is prepared by melting the crude pig iron, as it conies from the smelting furnace, with hematite (iron oxide). The resulting pasty mass is first hammered and then rolled to remove from it most of its impurities. The chief foreign substance is an iron silicate slag which is never wholly removed by the squeezing of the rolls but becomes elongated, giving to wrought iron its characteristic fibrous structure, Fig. 52. 1 SAUVEUR, "Metallography and Heat Treatment of Iron and Steel." 7 98 PRINCIPLES OF METALLOGRAPHY This material has been largely replaced by the cheaper mild steel which has many of the same physical properties. Wrought iron is sometimes specified, however, because of its easy welding and its resistance to shock. Steel includes the alloys having less than 1.7 per cent, carbon and the properties of this group of alloys is sub- FIG. 52. Wrought iron. ject to the widest variation, depending on the method of production, the rate of cooling, the subsequent heat treatment and other factors. The term iron includes the alloys from 1.7 per cent, carbon upward, usually not in excess of 4 or 5 per cent. The Equilibrium Diagrams. Many metallographists have studied the alloys of iron and carbon in the greatest detail and have proposed new equilibrium diagrams or IRON AND STEEL 99 have suggested modifications of the first one. Though differing somewhat from it in special points, all of them, with the exception of the Upton diagram, 1 resemble in general outline the early diagram of Roberts- Austen. 2 1 UPTON, /. Phys. Chem., 12 (1908), 506. 2 ROBERTS- AUSTEN, Proc. Inst. Mech. Eng., 1899, 35. 100 PRINCIPLES OF METALLOGRAPHY This is especially true of the alloys in the steel range (less than 1.7 per cent. C). Much more uncertainty exists in the range of the irons, largely, no doubt, because the lesser importance of this group has not warranted the immense amount of study that has been given to steel. A simplified and probably incomplete iron-carbon diagram is given in Fig. 53 from which most of the im- portant general relationships can be studied. The diagram is a combination of eutectic (p. 11), solid solu- tion (p. 39), compound (p. 49) and eutectoid (p. 57). The valuable properties of steel depend to a large ex- tent on the fact that, as is always true of reactions in the solid state, the decomposition of the solid solution 7 into its components requires time and can be prevented almost wholly by a sufficiently sudden cooling. It is evident from the diagram that pure iron never separates from a liquid solution of carbon in iron but that the solid which first separates along the line FeE is a solid solution of carbon in iron becoming a saturated solution when 1.7 per cent, carbon has been added. Referring to the diagram, it will be seen that all steels are originally solid solutions of carbon in iron, the carbon content varying from almost zero up to the saturation point, 1.7 per cent. The eutectic E is a mixture of the solid solu- tion B and the solid which separates along the line EC. Whether the solid separating on this line is actually the definite compound (Fe 3 C) is open to question and will be considered later. Decomposition of the solid solution 7 takes place along the lines FP and BP which intersect at the eutectoid point P. FP is a line along which pure iron separates and BP represents the separation of the definite compound Fe 3 C. Consider, now, the changes which take place when a, IRON: AN STSSP;'* : ; : : : /. 101 steel containing 0.5 per cent, carbon cools from the molten state to ordinary temperatures as indicated by the line xx' in Fig. 53. At x the metal is liquid. When the line FeE is reached a solid solution begins to separate and at the temperature represented by the intersection of xx'wiih FeB the steel has wholly solidified. Through- out the area FeFPB, the alloy is a solid solution, possibly of carbon in iron, but, more probably, of the compound Fe 3 C in iron. Along the line FP a change in the solid state takes place with a separation of pure iron and, as a result, a change in the concentration of remaining solid solution until it contains about 0.85 per cent, carbon and has reached a temperature of from 680 to 700. At this temperature and composition the final change in the solid solution takes place with the formation of the eutectoid P, an intimate mixture of the pure iron and the compound Fe 3 C. Steels containing less carbon than that corresponding to the eutectoid (0.85 per cent. C) are known as hypo-eutectoid steels while those from 0.85 to 1.7 per cent. C are the hyper-eutectoid steels. In this latter range the solid solution decomposes along the line BP with the separation of the compound Fe 3 C, reducing the carbon content until the eutectoid point P is again reached, when the same eutectoid mixture as before is formed. When complete equilibrium has been established, hypo-eutectoid steels consist of varying amounts of pure iron imbedded in the eutectoid P, while the hyper-eutectoid steels are mixtures of Fe 3 C with the same eutectoid. At P, only the eutectoid will be found. Incomplete Transformations. If the behavior of steel on rapid cooling was as simple as has just been indi- cated there would be only three classes of steels, (1) mixtures of iron and the eutectoid, (2) the eutectoid itself and (3) mixtures of Fe 3 C and the eutectoid. Thus 102 .'. the physical properties of the steel could be readily determined by a knowledge of the physical properties of the three substances involved. It is, however, rarely the case in practice that complete equilibrium is established, so that by far the larger number of technical steels rep- resent imperfect equilibria due to incomplete trans- formations along the lines indicated in the diagram. Extremely rapid cooling, such as quenching from a high temperature in liquid air, produces the unchanged solid solution; while either very slow cooling or annealing for a considerable time at a temperature just below the eutectoid temperature, will give the eutectoid mixture. Between these extremes are various intermediate transi- tion forms with varying physical properties. Several of these intermediate substances have such characteristic microscopic structure and definite physical properties that they have been named as a means of distinguishing them from each other. These names are not at all de- scriptive but have a certain historical interest as they are based on the names of men who have been leaders in the development of the science of metallography. The un- decomposed 7-solid solution is called Austenite, after Roberts- Austen, one of the pioneers in the metallography of steel. Following this is the more usual component found in quenched steel, Martensite, after the German metallographist Martens; next Troostite, from the French chemist Troost; Sorbite, after Sorby, and finally Pearl - ite, resembling mother of pearl, the only product of which the name is at all descriptive. The transition briefly stated is from Austenite > Martensite Troostite Sor- bite >Pearlite. To complete the naming of the common constituents of steels, the pure iron separating along the line FP has been called Ferrite and the compound Fe 3 C, Cementite, as it is the important constituent of those steels which have been hardened by the cementation IRON AND STEEL 103 process. It is hardly necessary to say that the changes are not as abrupt as the limited series of names might indicate but that there are still other intermediate prod- ucts to which such combination names as troosto- sorbite and sorbitic-pearlite have been given. The name Osmondite, from Osmond the French metallographist who was a pioneer in the microscopic study of steel, is sometimes given to the product at the exact boundary between troostite and sorbite but is not frequently used Etching of Steel and the Microscopic Appearance of Its Constituents. Many etching reagents have been suggested, some of them extremely complex mixtures, but for most purposes three or four different solutions will be found sufficient. Nitric Acid and Alcohol. The most commonly used reagent is a solution containing 4 c.c. of concentrated nitric acid (sp. gr. 1.42) in 96 centimeters of ethyl alcohol. This is used by immersing the specimen for about 10 seconds and agitating the liquid constantly to prevent the retention of gas bubbles on the polished surface. After treatment, the specimen is washed thoroughly in running water and dried either by patting gently with soft linen or cotton, or by means of a blast of air. Picric Acid. For low carbon steels a solution of 5 grams of picric acid in 95 c.c. of alcohol is frequently used. Alcoholic Hydrochloric Acid. Martens and Heyn recommend the use of a solution containing 1 part of hydrochloric acid (sp. gr. 1.19) in each 100 parts of alco- hol. The reagent acts much more slowly than alcoholic nitric acid, requiring about one minute, but the results are excellent. Special Reagents. Copper Ammonium Chloride. For examination of the specimen without the microscope (macroscopic examination), and, especially, for the pur- 104 PRINCIPLES OF METALLOGRAPHY pose of locating segregated areas in large specimens, a solution of 8 parts of copper ammonium chloride in 100 parts of water is used. This reagent is suitable only for plain carbon steels and must be applied by immersing the specimen. Iodine Solution. A solution containing 6 parts of iodine in 100 parts of alcohol is used for macroscopic examination and is applied by swabbing the polished surface with cotton soaked in the solution. The opera- tion is continued for five minutes, a fresh portion of reagent being added as soon as the iodine color produced by the preceding treatment has disappeared. The iodine reagent must be prepared immediately before use. It is effective in showing excessive slag, phosphorus segre- gations and other irregularities. Kourbatoff's Reagent for Cementite. Cementite is not affected by the usual reagents but is colored black or brown by immersion from three to five minutes in a boiling solution of sodium pier ate in sodium hydroxide. The reagent contains 2 parts of sodium picrate in 98 parts of a 25 per cent, solution of sodium hydroxide. Ferrite is unaffected by this reagent so that Kourbatoff's mixture furnishes a sure means of distinguishing between these two components. Time Required for Etching. With all these reagents the time required varies somewhat with the nature of the steel under examination. If pearlite requires 10 seconds immersion, as it will with alcoholic nitric acid, sorbite will need 7 or 8 seconds, martensite about 5 seconds and troostite, which is most sensitive to chemical reaction, will take only 2 or 3 seconds for the develop- ment of its structure. Occurrence and Physical Properties of the Constitu- ents of Steel. General. In dealing with slowly cooled steels in complete equilibrium it is only necessary IRON AND STEEL 105 to consider the physical properties of the three possible constituents, (1) ferrite, (2) pearlite, and (3) cementite. These are shown in the following table (Sauveur). Constituents Tensile strength in >ounds per square _ incn Elongation, per cent, in 2 inches Hardness Ferrite About 50 000 40 + Soft Pearlite About 125 000 10 + Hard Cementite 5000 (?) Very hard and brittle 180 90 100 80 M 5 WO 70 1 s 1 s EFFE STOFC j A. ARBON ON PHYSICAL PROPER nnealed just above - Ac 3 Tensile Strength Elastic Limit Yield Point Elongation Contraction of Area TIES B- C- D- D \ \ I 3 1 1 I 100 ! 50 a | SO-40 H 60^30 I * 40 20 20 10 \ _ A ^^^ ~y( X ^ \ / S^fr \ A ^ >^ -\ J- x^ ^ ^ \ / B ^^. ^/ x S 1.00 1.20 FIG. 54. Diagram showing effect of carbon. (Ajter J. H. Nead, Am. Ins*. Min. Eng. (1916), 2341.) As every annealed steel is a mixture of two of these three constituents, it is to be expected that the physical prop- erties can be fairly closely predicted from the chemical 106 PRINCIPLES OF METALLOGRAPHY composition. The tensile strength, for example, in- creases from about 50,000 pounds per square inch with pure iron, to 125,000 pounds per square inch with pear- lite (0.85 per cent. C), and then decreases as the amount of cementite increases. In the same way the ductility, as represented by the percentage elongation, decreases from that of pure iron and becomes very low indeed A B FIG. 55. Pearlite and ferrite. A = 123 X. B = 1650 X. when the carbon content is more than 1.20 per cent owing to the increase in brittle cementite. These general relations of pearlitic (fully annealed) steels are shown in the sketch, Fig. 54, and the microscopic appearance is given in Figs. 55 and 56. The former illustrates a hypo-eutectoid steel, the latter is typical of the hyper-eutectoid class. The thumb print struc- ture in each case is the eutectoid pearlite. By heat- IRON AND STEEL 107 ing pearlitic steels for a number of hours at 600- 700 the laminated structure gradually disappears, due to the coagulation of the layers of cementite in the eutectoid, with the formation of spherical masses. This operation is known as spheroidizing and the cemen- tite as spheroidized cementite. FIG. 56. Cementite and pearlite 1650 X- (The white mass is cementite and the dark eutectoid pearlite.) The slowly cooled (pearlitic) steels have many indus- trial uses. Very low carbon steel, or extra mild steel, containing less than 0.1 per cent. C is used for articles which must be readily worked like rivets or horseshoe nails and is also used for material which is to be subjected to cementation (see p. 117). Low carbon, or mild steel, with carbon up to 0.25 per cent, may be 108 PRINCIPLES OF METALLOGRAPHY used for screws, bolts, agricultural implements, sheets, wire and structural steels, though much of this material, due to treatment in the process of manu- facture, is not strictly pearlitic. Steels containing from 0.25 per cent, to about 0.60 per cent, carbon form the class usually known as medium high carbon or half hard steel. As the carbon content increases the use of completely pearlitic steel is decreased. Wholly or partially pearlitic steels in this range are used for castings, shafting, piston rods and cylinders for compressed gas. When the carbon is from 0.6 to 0.85 per cent., the steels are classed as high carbon or hard steels and among the many uses may be mentioned tires, springs, cheap cutlery, wire, certain agricultural tools and wood work- ing tools. Practically no steels in the range from 0.85 to 1.25 per cent, carbon are used without some heat treatment which will modify 'the pearlitic character to a greater or less degree. Rapidly Cooled and Tempered Steels. The changes which take place when a steel is quickly cooled (quenched) and subjected to a later heating (tempering or annealing) are so numerous that only a few can be considered as illustrating the general character of the resulting proper- ties. The first metallographic constituent of chilled steel which might be expected is austenite (Fig. 57) , but it decomposes so quickly on cooling that it is never formed in the commercial hardening of plain carbon steel. It is the chief constituent of certain alloy steels, however, and will be considered later (p. 113). The constituent commonly produced when steel is quenched is mar- tensite (Fig. 58), the first transition product of the de- composition of austenite. This substance is extremely hard, brittle and unworkable so that pure martensitic steel is rarely found in practice. Associated with other IRON AND STEEL 109 constituents such as troostite or free ferrite, in low carbon steels, and with cementite, in the high carbon steels, it is always found in hardened tool steels in amounts which vary with the method of its production. The quenching of an edged tool may be taken as a single example of the production of martensite. The tip only of the tool is heated to redness and dipped in water. The brittleness of the resulting martensite is reduced, with a consequent sacrifice of hardness, by allowing heat to flow from the unquenched portion of the tool FIG. 57. Austenite and martensite. (The dark masses are martensite and the light ground mass austenite.) 350 X- to the tip, until the desired softening results, when the entire tool is quenched and the final product has the combined properties of chilled and tempered; steel. The decomposition which takes place in the tool just described, after it has been subjected to partial quench- ing, leads to the formation of troostite (Fig. 59). As troostite is a decomposition product of martensite, it is to be expected that in its physical properties it will differ from martensite in decreased hardness and increased ductility. Its hardness is about halfway between that of 110 PRINCIPLES OF METALLOGRAPHY martensitic and pearlitic steel of the same carbon content. The tempering of troostite leads to a rapid increase in duc- FIG. 58. Martensite. 900 X. FIG. 59. Martensite and troostite. 117X. tility with a decrease in hardness. It may be formed by cooling slowly through the transformation interval, as for example when small pieces of steel are quenched IRON AND STEEL 111 in oil, but it is much more frequently produced by re- heating (tempering) a chilled steel below 400. It is a constituent of practically all plain carbon steels, which have been hardened (tools for example) and is associated in varying amounts with martensite, depending on the temperature of tempering. When great hardness is required and brittleness is of less importance, as in the production of razor blades for instance, a temperature of about 200 is used, producing relatively small amounts of troostite. When toughness as well as hardness is re- FIQ. 60. Sorbite. 350 X. quired, the amount of troostite is increased by tempering at 300 to 400. Most tools are tempered between 200 and 300. While it is almost never the custom in practice to temper above 400 it is, perhaps, easier to correlate the physical properties of sorbite (Fig. 60) with those of troostite by considering the former as produced by the tempering of troostitic steels in the range from 400 to 600. Sorbitic steels are softer and more ductile than troostitic and not as soft as pearlitic steels. Sorbite, like troostite, is one of the decomposition products of austenite and is, in fact, imperfectly formed pearlite. 112 PRINCIPLES OF METALLOGRAPHY It may be produced by heating a chilled steel in the range between 400 and 650 but it is usually made by regulating the cooling rate in such a way that, while the chilling action is not great enough to produce martensite, it is too rapid to allow the complete formation of pearlite. It may be formed (1) by cool- ing small pieces in air, (2) by quenching larger pieces in oil from a temperature just above the critical range or (3) by quenching small pieces in water from a point near the bottom of the critical range. Though slightly less ductile than pearlitic, sorbitic steel has so high a tensile strength and elastic limit that it is used for the highest grade of structural work. It is not possible to give absolute values to the physical properties of sorbite as its character varies so much with the method of pro- duction. It may be stated for purposes of comparison that while sorbite sometimes reaches a tensile strength of 140,000 pounds per square inch, ordinary pearlite will have a strength of about 110,000 pounds per square inch while the coarsely laminated form of pearlite will show a tensile strength of only about 70,000 pounds per square inch. The following table from H. C. Boynton gives the relative hardness of the different constituents of steel as compared with ferrite (pure iron) as a standard. It must be remembered that these values, particularly for the intermediate forms like sorbite and troostite are only approximations. Ferrite = 1 Pearlite = 43 Sorbite = 52 Troostite = 88 Austenite =104 Martensite = 239 Cementite = 272 IRON AND STEEL 113 Alloy Steels. In addition to the elements silicon, phos- phorus, manganese and sulphur, always found in small quantities in carbon steels, other elements are often added to make the ternary or quaternary alloy steels, many of which have remarkable physical properties. The subject of alloy steels is so large and the informa- tion concerning them changing so rapidly that only a very general discussion can be given here in spite of the great and constantly increasing importance of these alloys. A few general principles seem to hold, though even these must not be accepted as absolutely established but rather as suggestive. 1. If the carbon content is kept constant, the addition of the alloying element in increasing amounts causes the steel to be first pearlitic, then martensitic and finally, with a sufficiently high percentage of the third element, to become austenitic. 2. By holding the percentage of the alloying element constant and increasing the carbon content the changes under similar cooling conditions are^ as before, from pearl- itic, to martensitic, to austenitic steel. 3. It follows, almost as a corollary to (1) and (2), that the higher the carbon content the less the amount of alloying element needed to complete the structural change and, conversely, the higher the percentage of alloying element the less carbon is needed to change the structure and properties. These changes are shown graphically in the diagram (Fig. 61). The number of known alloy steels is great and con- stantly increasing. Among the commoner ternary steels are those containing either nickel, manganese, tungsten, chromium, vanadium, molybdenum or silicon, in addi- tion to the carbon. In the quaternary class may be found chrome-nickel, chrome-tungsten, chrome-vana- dium, nickel- vanadium and others. Other and still more 114 PRINCIPLES OF METALLOGRAPHY complex series are the chrome-nickel-vanadium, chrome- tungsten-vanadium and the like. A detailed discussion of these alloys is impossible and only a few will be con- sidered to illustrate in a general way the properties of each group. The strictly metallic elements like manganese, nickel and chromium lower the critical points of steel very decidedly so that austenite and martensite can be formed much more easily than is the case with carbon steel. A steel containing from 1 to 1.5 per cent, carbon and 0.2 0.4 0.0 0.8 1.0 Per Cent Carbon FIG. 61. Constitutional diagram of alloy steels. (Sauveur after Guillet.) from 10 to 15 per cent, manganese can be obtained easily in the austenitic condition by reheating the cast steel to about 1000C. and quenching in water or oil. The resulting steel is hard and resistant to wear but, at the same time, possesses much ductility. It has been used in making rails subjected to excessive wear, as on sharp curves. The less metallic alloying elements, like tungsten, vana- dium and molybdenum have little or no effect on the critical points (p. 96) but, due to the formation of double carbides, tend strongly toward the production of cemen- titic steels. IRON AND STEEL 115 If manganese is added to a tungsten steel, the alloy first formed on cooling is of the cementite class. If, however, this steel is reheated to a high temperature and cooled in the air, the carbide which is dissolved at the high temperature is retained in the martensitic con- dition. Such a steel is said to be " self hardening." One of the most important of the alloy steels is the chrome-tungsten or high speed tool steel. Such an alloy, with from 10 to 20 per cent, tungsten and 2 to 10 per cent, chromium, has the characteristics of cementite when slowly cooled. On reheating to a very high tem- perature, often almost to the melting point of the steel, the carbide dissolves and, if the cooling is fairly rapid, the steel retains an austenitic or martensitic structure with corresponding physical properties, notably great hardness. The striking fact in this case, however, is that the martensite formed in this way shows no tendency to soften even at relatively high temperatures, approach- ing 600C. This makes it possible to run a cutting tool at such speed that, while its edge will become visibly red, it will still retain its hardness, a property absolutely impossible with common carbon steel. These illustra- tions will serve to show some of the possibilities of alloy steels. Heat Treatment of Steel. 1 The most important property of steel is the power which it has of changing its physical condition under the influence of heat. Many of the changes have been considered in connec- tion with the discussion of the various metallographic constituents: sudden cooling or quenching, for example, produces the hard martensite; tempering hardened steel softens it, producing troostite or the still softer con- stituent sorbite. Annealing may be carried out for one of three reasons; (1) to increase the softness and 1 BTJLLENS, "Steel and Its Heat Treatment," Ed. 2. 116 PRINCIPLES OF METALLOGRAPHY ductility, (2) to relieve the strains produced by chilling or by mechanical work, or (3) to reduce the grain size. The second object has been considered in the case of worked or strained brass (p. 76) to which strained steel is wholly analogous. Severely worked steel shows the same elongated structure illustrated in Fig. 38 and the restoration of its normal structure by annealing is of the same character as with brass. Improvement of the physical properties may be brought about by .annealing whereby the size of the crystal grains is reduced. Large crystal grains are almost always an indication of weakness while the production of small grains invariably leads to greatly improved mechanical properties. Temperature of Annealing. The steel must be heated to a temperature slightly above its critical range in order to have the crystalline structure affected, and it has been found that the higher the temperature above the critical range the more coarsely crystalline the result- ing steel will be. The most suitable temperature varies, of course, with the carbon content but should be ap- proximately as follows for plain carbon steels. (American Society for Testing Materials.) Carbon content Annealing range Less than 0. 12 per cent 875 to 925C. 0. 12 to 0.25 per cent 840 to 870C. 0.25 to 0.49 per cent 815 to 840C. 0.49 to 1.00 per cent 790 to 815C. After the desired temperature has been reached the object must be kept at that temperature until it is heated throughout its mass. It is then cooled either (1) in the annealing furnace, producing the softest, weakest and most ductile metal; (2) in air, giving a somewhat harder and less ductile material; or (3) by quenching in oil, giving the hardest, strongest and least ductile steel of IRON AND STEEL 117 the three. The finest possible structure would be ob- tained by quenching from a point as near the critical range as possible, but, except with very low carbon con- tent, the resulting steel would be hard and lacking in ductility. The double anneal overcomes this difficulty. The operation consists in reheating the hardened steel to about 650 (close to, but below its critical range), which serves to relieve the hardness and at the same time to retain the fine structure. Case Hardening. An operation closely allied to heat treatment is case hardening, a process which is carried on for the purpose of adding a hard, non-abrasive sur- face to a strong, ductile steel. The steel used for this purpose has a low carbon content (0.2 per cent, or less) and is carburized by heating in contact with a carbon furnishing substance, either in the solid, liquid or gase- ous form. Numerous materials have been used of which charred leather, potassium ferrocyanide and barium car- bonate may be considered types. As carbon dissolves very slightly in a-iron, if at all, the case hardening tem- perature must be above the critical range and is usually between 850 and 1000C. The depth to which the carbon penetrates depends on the length of time during which the steel is in contact with the carbonaceous material and varies from 0.5 millimeter to 5 millimeters with an average depth of from 2 to 3 millimeters. Metallographic examination of a case-hardened specimen shows a surface coat of hard cementite over a band of eutectoid, free from cementite and ferrite, and below this band the soft interior mass or core in which ferrite largely predominates. The coarse structure of the core produced by the long heating during the case hardening process, must then be refined by suitable heat treatment. Cementation is a similar operation applied to wrought iron for the purpose of changing it into steel. It differs 118 PRINCIPLES OF METALLOGRAPHY from case hardening in that the steel produced in this way is afterward melted and the carbon becomes uniformly distributed in the ingot, forming cement steel. Cast Iron. Reference to the iron-carbon diagram (p. 99) shows that the cast iron field extends from 1.7 to 4 or 5 per cent. C. The liquidus curve has two branches, along one of which (FeE) austenite should separate and along the other (CE) cementite. The cementite separating along CE is so readily decomposed on cooling that three classes of cast irons are commonly recognized, depending on the cooling rate and on the presence of constituents other than carbon. 1. White iron results from rapid cooling (chill cast- ing) of the metal and has the white fracture and FIG. 62. Chilled casting. (Actual size.) a layer is white iron. hard, brittle qualities of cementite. Chill castings are too hard to machine and are seldom used in the production of small articles. It is often necessary to produce soft, fairly strong cores with extremely hard surfaces as, for example, in car wheel treads or the sur- faces of rolls. In such a case the white iron (cementite) IRON AND STEEL 119 surface may be produced by casting against a highly heat-conducting material like an iron plate. The pro- FIG. 63. Section made from white border, a, in Fig. 62. 350 X. FIG. 64. Graphic temper carbon in malleable iron. 350 X. duction of white iron is also favored by the absence, or low percentage, of silicon and the presence of high percentages of manganese and sulphur (Figs. 62 and 63). 120 PRINCIPLES OF METALLOGRAPHY An important decomposition product of white iron is produced by annealing for several days at a temperature of about 730C. Under these conditions the hard, white cementite decomposes into graphite and ferrite, the graphite separating, however, not in the massive form but as an amorphous black powder (temper carbon) (Fig. 64). This malleabilizing process produces an iron which is much stronger than gray iron of the same com- FIG. 65A. Taken at the part y, in Fig. 62. 350 X shows gray iron. position. The tensile strength of malleable iron is more than 40,000 pounds per square inch as against an approximate 20,000 pounds for gray iron with the same amount of free carbon in the fibrous instead of the powdery form. Malleable iron is used for castings which are to be subjected to shock, especially in the manufac- ture of small castings which would be made of steel if it were not for the technical difficulties involved in steel casting. 2. Gray cast iron is produced when the cementite first formed is allowed, by decreasing the rate of cooling, to decompose into ferrite and graphite. This gives a dull IRON AND STEEL 121 FIG. 65B. Graphite in iron. 115X- The halves represent two specimens of * iron with different carbon content. FIG. 66. Section taken at the transition zone, /3 in Fig. 62 mottled iron 350 X. 122 PRINCIPLES OF METALLOGRAPHY gray appearance to the fractured metal. Under these conditions, the graphite separates in the form of long fibrous masses unaffected by etching reagents (Figs. 65a and 656). The separation of graphite is increased by the addition of silicon which, in gray castings, is usually present in amounts varying from 2 to 4 per cent. The tensile strength of gray iron varies from 18,000 to about 23,000 pounds per square inch. 3. Mottled Iron. By a suitable regulation of the cooling rate, an iron containing both free cementite and graphite is produced with properties depending on the relative amounts of the soft and hard constituent (Fig. 66). CHAPTER VI DEFECTIVE MATERIAL Not the least important of the many uses of metallog- raphy is its application to the study of defective or unsuitable material. A distinction must be made be- tween the terms "defective" and " unsuitable" as a perfect alloy may be wholly unsuited to the purpose for which it has been, or is to be, used. The causes of defective metal are numerous but they may generally be classified in one of four groups: 1. Incorrect chemical composition; 2. Improper mixing, melting or casting; 3. Unsuitable mechanical treatment; 4. Improper heat treatment. 1. Incorrect Chemical Composition. Large errors in chemical composition produce effects which are too ob- vious to need more than a reference. If an alloy which was supposed to be brass with 70 per cent, copper, actually contained but 50 per cent, copper and was sub- jected to severe cold work the results would be disas- trous. Such errors, however, are found by the analyst rather than by the metallographist. Slight errors in chemical composition are generally far more serious because they are so much less readily detected. Not infrequently these differences in chem- ical composition are so localized as to escape detection by analytical processes and yet they are of far-reaching effect on the physical properties of the material. Some of these defects have been referred to, as for instance, 123 124 PRINCIPLES OF METALLOGRAPHY the formation of brittle envelopes around brass crystals when bismuth or antimony is present (p. 92). The commonest chemical defect is the presence of segregated material which is without question one of the most common causes of failure in metals. Segregated im- purities occur in many technical alloys but to a far greater extent in steel and iron in which their presence may do serious harm. Until recently but little infor- FIG. 67. Slag inclusions in wrought iron. 350 X. mation has been available on defects of this sort in non- ferrous metals. During the past year, however, two papers 1 have been published showing by microphoto- graphs, inclusions of the oxides of tin and zinc as well as of casting sand and other non-metallic substances in brass and bronze. In the case of iron and steel, chemical inclusions usu- 1 CoMSTocK, "Non-metallic Inclusions in Brass and Bronze," /. Am. Inst. Metals, March, 1918. CARPENTER & ELAM, "An Investigation on Unsound Castings of Admiralty Bronze," J. Inst. Metals (British), xix, 155. DEFECT I VE MA TERIAL 125 ally consist of slag (commonly silicates of iron or other elements), sulphides of iron or manganese, phosphides of iron or manganese or metallic oxides. Slag is a normal constituent of wrought iron but may at times segregate in such a way as to become harmful (Fig. 67). Small amounts of oxide and silicate are always present in steel and if the quantity is not excessive and is uni- FIG. 68. Islands of manganese sulphid formly distributed will not seriously effect the mechan- ical properties of the metal. Irregular distribution of the non-metallic material in segregated areas is always a source of danger and can be detected more readily by the metal microscope than in any other way. Sulphides, either of manganese or iron, are sources of danger, especially in material which is to be worked hot (rolled, hammered or drawn). Manganese sul- phide is sometimes found in steel rails in a form which 126 PRINCIPLES OF METALLOGRAPHY under the microscope resembles long, narrow islands, dove gray in color (Fig. 68) . Sulphide inclusions are easily detected by the production of "sulphur prints." In the older method of Heyn this is effected by hold- FIG. 69. Sulphur print on (Actual size.) ing a piece of white silk moistened with mercuric chlo- ride and hydrochloric acid in contact with the polished surface of the metal (Fig. 69). Black stains of mercuric sulphide are produced at the points of contact with the FIG. 70. Sulphur print of a defective boiler plate showing sulphide streaks. sulphide spots. The same effect is obtained more simply by soaking photographic printing paper (Velox or Cyko) in 2 per cent, sulphuric acid and applying the paper to the polished sample. Black spots or streaks DEFECTIVE MATERIAL 127 of silver sulphide are produced. The print may be made permanent by fixing it in "hypo" (sodium thio- sulphate) solution in the usual way. The presence of sulphide streaks is shown in the following photograph, A. Rolled from the top of the B. Rolled from the bottom of the ingot. ingot. FIG. 71. Steel I-beams showing sulphide and phosphide segregations. Fig. 70. That the lines are actually due to sulphide inclusions is shown by chemical analysis of the separate layers (Heyn). Layer I shows 0.067 per cent, sulphur, layer II, 0.201 per cent, and layer III, 0.240 per cent. 128 PRINCIPLES OF METALLOGRAPHY Phosphorus in the form of phosphide is one of the most harmful of the non-metallic inclusions in mild steel. It occurs in the ingot and if too small an amount of the ingot top is removed to eliminate the phosphide segre- gations, these will go through the mechanical operations of rolling, forging, etc., without coming to the surface. The result is shown in Fig. 71 and illustrates in a striking fashion the advantages of metallographic over chemical examination in cases of this sort. The streaks are due in this case to sulphide as well as to phosphide inclu- sions. Phosphide inclusions are the usual cause of "cold shortness." This inclusion of non-metallic impurities in alloys has been estimated to cause more than 75 per cent, of the failures found in technical practice. 2. Improper Melting, Mixing or Casting. Another, though less frequent, reason for defective material is imperfect mixing of the molten alloy, producing layer formation in the solid. Plastic bronzes (p. 9) some- times fail because of imperfect mixing which causes a much greater concentration of lead in one part of the casting than in another. In the light bearing-alloys of the Babbitt metal class and in type metals containing antimony, irregular distribution of the cubical crystals of SbSn is of common occurrence and is always a cause of unsatisfactory material (Fig. 72 a and b). Segregation in brasses of the Muntz metal type is not at all infrequent and in rolled or drawn material gives rise to distinct layer formation. Fig. 73 shows a- and 0-brass in adjacent streaks in a condenser tube. The a-brass may be found also in the form of islands or spots in the (3-field as shown in Fig. 74 a and 6. These few cases illustrate the possibilities of layer formation in alloys which are duplex in structure and have components differing in specific gravities. DEFECTIVE MATERIAL 129 3. Unsuitable Mechanical Treatment. Work, either cold or hot, may produce bad material if improperly done. FIG. 72. Variations in size and distribution of SbSn crystals in babbitt metal. The effects of cold work on brass have been considered on p. 72 and illustrated in Figs. 18 and 47. The char- 130 PRINCIPLES OF METALLOGRAPHY acteristic fibrous structure of overworked brass is also shown in Fig. 75. Steel is more commonly worked hot and defective metal may be produced when the rolling, hammering or other operation is commenced at too high a temperature. If, for example, a steel containing ferrous sulphide (FeS) is rolled at a temperature much above 900C. it is probable that the iron sulphide is actually molten and rolling would naturally cause serious FIG. 73. Layer formation in defective condenser tube showing a- and a + /3-layers. 75 X. cracks in the finished material. If, on the other hand, mechanical operations are carried on (or are finished) at too low a temperature, severe strain hardening may result. Both these conditions, incipient cracking from too hot work and overstrain from cold work may be detected by the microscope. Unequal amounts of work on different parts of the same article is a frequent cause for metal failure. This produces internal strains which may ultimately lead to " season cracks" (p. 91) or may DEFECTIVE MATERIAL 131 cause immediate cracking on reannealing. Fig. 18, p. 37, shows a crack which was probably produced in this way. A. Etched ,-ith NH 4 OH and (NH^tStOs. a-brass very slightly attached. 75 X. , B. Part of the same tube etched with NH 4 OH and H 2 O 2 . 75 X- FIG. 74. Formation of islands of a-brass in |8 field. Cold work may be done unintentionally in certain mechanical operations and may cause serious trouble. 132 PRINCIPLES OF METALLOGRAPHY In fitting large pieces of structural steel such as are used in bridges, building frames and the like, it is sometimes necessary to hammer the parts into position. This has been known to cause bad strain hardness in a limited area with a resulting failure of the metal. In a gas tank recently constructed a large number of rivets broke because of the overstrain produced by hammering when they were too cold. Defects produced in this way can- not, of course, be controlled by metallographic exami- FIG. 75. or-brass strained and not annealed. 75 X. nation but the causes of failure can often be detected by the microscope and the responsibility placed. One other type of defective material is due to a combi- nation of mechanical and chemical difficulties and is caused by the rolling or hammering of surface oxides or " scale" into the metal under treatment. This is one of the reasons for hard spots in rolled brass and, as it is apt to be a surface condition, it frequently leads to rapid corrosion of the material. 4. Improper Heat Treatment. The term "heat treat- ment," is most commonly used in connection with heating DEFECTIVE MATERIAL 133 or cooling operations (quenching, tempering, etc.) ap- plied to the solid alloys but, if the term is applied in a broader sense to include the rate at which a molten alloy cools, it explains many cases pf defective material. The casting of Babbitt metal will serve to illustrate the way in which defective, or at least unsuitable, material may be produced by variations in the cooling rate of the liquid metal. Too rapid cooling produces undeveloped crystals of the tin-antimony compound SnSb, while very slow cooling causes the formation of large cubes of the compound and greatly increases the tendency of these crystals to rise to the top of the cast metal. Both conditions are responsible for unsatisfactory bearing surfaces. If the metal against which the Babbitt is poured has a temperature of approximately 100C. the SnSb crystals will be found to be moderate in size and regularly distributed. Heat treatment, as applied to solid alloys, consists of annealing, tempering or quenching and any one of these operations if badly executed will lead to defective or un- suitable material. The microscope is an invaluable tool in studying defects of this kind. Annealing carried on at too high a temperature may cause " burning" or " overheating." Burning is nothing but oxidation and is visible under the microscope as a marked thickening of the grain boundaries due to oxide formation. It is a cause of brittleness and produces fracture between the grains because of the oxide envelope with which the crystals are surrounded. Overheating is distinguished from burning in that it is not accompanied by oxidation but is a cause of excessive crystal growth. The very large crystals of overheated a-brass are shown in Fig. 50 (p. 89). In the case of steel the best physical properties are associated with fine structure so that annealing at temperatures too far above the critical range 134 PRINCIPLES OF METALLOGRAPHY will produce material which, if it is not actually defective, is at least inferior to the steel which would be obtained by suitable treatment. Annealing in an unsuitable atmosphere may also pro- duce defective material. Steel, for instance, may be decarbonized and therefore materially weakened by long heating in an oxidizing atmosphere. Tempering is carried out for the purpose of relieving strain hardness or increasing the amount of softer ma- terial at the expense of the hard material as, for example, in the change of martensite to troostite or sorbite. Tem- pering at too high a temperature may unduly decrease the hardness and at too low a temperature may leave an excessive amount of the harder constituent. In either case the resulting material is defective in the sense that it is not in the proper mechanical condition for the pur- pose to which it is to be applied. Quenching incorrectly performed is another cause of defective or unsuitable material. Quenching from too high a temperature generally forms excessively large crystals which produce correspondingly weak metal and it may also cause a form of strain hardness which not infrequently leads to cracks. Quenching is done for the purpose of retaining the metal in the condition (usually a hard state) in which it exists at the higher temperature. If the temperature from which the metal is quenched is too low the desired hardness will not be obtained and the structure of the soft form will be seen under the micro- scope. Finally, if the specimen is large, the cooling rate may be unequal in different parts and local differ- ences in hardness and strain will be produced. Warp- ing or cracking of the quenched piece not infrequently results. This brief discussion of some of the causes of defective material will serve to illustrate the applications^ of DEFECTIVE MATERIAL 135 metallography to an important engineering problem and also emphasize the value of metallographic control both to the producer and to the consumer of alloys. This new branch of physical science is only one of the factors in the study of metals and, if possible, should always be used in connection with ordinary chemical analysis and mechanical testing rather than indepen- dently of them. There are times, however, when it is the only means of study that may be used without damage to the material under investigation as, for instance, in the inspection of a finished article. In conclusion it must be said that, while metallog- raphy has become an almost indispensable science in the metal industries, the interpretation of microscopic ap- pearances in complex cases requires a highly skilled and experienced observer. The inexperienced metallog- raphist should be extremely careful not to overestimate the value of his observations. APPENDIX TABLE 1 Outline of a Brief Course in Experimental Metallography 1. Assemble a melting point apparatus and calibrate a thermo- element (p. 18). 2. Prepare a series of alloys, study the cooling curves and construct the equilibrium diagram* (p. 4). Lead-antimony, bismuth-tin and lead-tin are suggested. 3. Sectionalize, polish, etch and examine the alloys prepared in (2) (p. 29). 4. Photograph at a magnification of seventy-five diameters (75 X), develop the plates, print, trim and mount the photographs (p. 34). 5. If time permits study as in (2), (3) and (4) the following alloys: (a) Tin-magnesium. (Open maximum.) (6) Copper-antimony. (Open maximum and concealed maxi- mum.) (c) Tin-antimony. (Several series of solid solutions, probably with a compound.) 6. Experiments with Brass. Polish, etch and examine (p. 29). (a) Cast brass. (6) Cold-worked a-brass. (Examples of this material occur in spring brass, sheets, condenser tubes, cartridge cases and similar articles.) (c) Examine severely cold-worked a-brass. Anneal for ten minutes at 550C. and reexamine. (d) Anneal samples of cold-worked a-brass for ten minutes each at the temperatures 450, 550, 650, 750 and 850 and com- pare the resulting crystals with reference to size. (e) Examine cast Muntz metal (60 per cent. Cu-40 per cent. Zn). (/) Examine extruded or hot-worked Muntz metal. (g) Anneal several specimens of hot-worked Muntz metal at 800 and * If but a limited time is available, the number of mixtures may be divided among the members of the class so that each will make but one or two melts. The results are then collected and arranged in the form of the complete diagram. 136 APPENDIX 137 (1) Quench in cold water (/3-brass). (2) Cool in air. (3) Cool slowly in the annealing furnace. Anneal a quenched specimen one-half hour at 400 and compare with (1). 7. Examination of Bronze. (a) Examine cast bronze and compare its structure with manganese bronze and aluminum bronze. (6) Examine cold-worked bronze and anneal to restore normal structure, (c) Examine leaded phosphor-bronze after polishing and before and after etching. (The unetched specimen will show the distribution of lead.) 8. Experiments with Steel. (a) Examine cast steel with about 0.3 per cent, carbon. Anneal for one hour at 875 and compare grain size with that of the original sample. (Etch with HNOs, p. 103.) (&) Heat specimens of steel containing approximately 0.1 per cent., 0.3 per cent., 0.6 per cent, and 0.9 per cent. C and cool very slowly through the critical range or heat for 30 minutes just be- low the critical range. (Note the relative quantities of ferrite and pearlite.) Examine at 100 X and 500 X. (c) Quench a small specimen with about 1.2 per cent. C from 1200 (Martensite). (d) Quench a small specimen with about 0.8 per cent. C from 1000 and reheat to 600 (Sorbite). (e) Quench steel used in (d} as before and reheat to 350 (Troostite) . Examine at 100 X and 500 X. (/) Polish two specimens of normally cooled high carbon steel (about 1.4 per cent. C). Etch one piece with nitric acid, the other with sodium picrate (p. 104). Examine at 100 X and 500 X (Cementite). (g) Quench a specimen with at least 1 per cent, manganese arid about 1.5 per cent, carbon from 1400 in ice water (Austenite) . (h) Examine high-speed tool steel. Heat for twenty minutes between 500 and 600 and reexamine after slow cooling. (i) Heat a sample of steel containing less than 0.2 per cent, car- bon in a mixture of 60 per cent, charcoal and 40 per cent, barium carbonate for at least two hours at 1000. Cool and examine the entire cross section (case hardening). Note the thickness of the cementite layer. 0') Select two pieces of steel with 0.5 per cent. C. Heat one to 138 PRINCIPLES OF METALLOGRAPHY 900, the other to 1200. Cool both in air and examine with special reference to size of the crystal grains. Experiments with Iron. (a) Wrought Iron. Polish, etch with alcoholic HN0 3 and examine wrought iron using one longitudinal and one transverse section. Note especially the color and distribu- tion of, the slag. (6) Gray Cast Iron. Polish a high-silicon, high-carbon iron and examine without etching (Graphite). (c) White Cast Iron. Examine a high -manganese, low-silicon iron which has been chill cast. (d) Malkable Iron. Examine malleable iron etched and unetched. Note the small rounded masses of graphite due to the decomposition of cementite and compare these with the graphite plates or needles in (6). (e) Mottled Iron. Examine the transition stage which will be found between the outer case and inner core of a chilled casting. Mechanical Testing. -If testing machines are available a study of the various mechanical properties as hardness, tensile strength, elastic limit, elongation, etc. should be made with some of the com- moner industrial alloys, brass, the different bronzes, steel, iron and the like and the results obtained studied in connection with the metallo- graphic examination. The outline just given is merely suggestive and the number of experiments may well be increased by the study of defective material as, for example, broken rails or axles, corroded boiler or condenser tubes, overheated steel or any of the cases of the same sort that may especially interest the experimenter. The field for investigation and research is unlimited. APPENDIX 139 TABLE 2 Books and Journals. (This list is not complete but includes the publications most readily obtainable). The asterisk * indicates the books especially recommended for the general reader. General Alliages Me"talliques Cavalier 1909. General metallography dealing extensively with physical and mechanical properties. Die binaren Metallegierungen Bornemann 1909. Brief theoretical discussion and many equilibrium diagrams. Einfiihrung in die Metallographie Goerens 1906, * Translation Introduction to Metallography Ibbotson 1908. A small general metallography but a classic in that it is the first book to deal with equilibrium diagrams as well as microphotography. Etude Industrielle des Alliages Me"talliques Guillet 1906 Gen- eral study of industrial alloys with many microphotographs. Etude Theorique des Alliages "Me'talliques Guillet 1904 Dis- cussion of the methods of alloy study; hardness, conductivity, magnetism, density, etc. Handbuch der Metallographie Guertler 1912-1914. The most complete and detailed treatment of metallography yet published. Much emphasis is laid on theoretical considerations. * Introduction to Physical Metallurgy Rosenhain 1917. Deals especially with mechanical testing, physical properties and indus- trial applications of alloys. Summary of defective material. Lehrbuch der Metallographie Tammann 1914. Deals with the subject almost wholly from the theoretical viewpoint. Contains an extended discussion of the valence theory as related to intermetallic compounds. * Metallic Alloys Gulliver 1913. One of the standard books on general metallography. Considers crystallography and the mechanism of crystallization of metals. Excellent treatment of ternary alloys. Metallographie Heyn and Bauer 1909. Small general metallo- graphy largely theoretical. Metallographie Ruer 1907 (Translated by Mathewson) .Ele- ments of metallography considered chiefly from the theoretical side. * Metallography Desch 1913. General metallography with spe- cial emphasis on physico-chemical principles and an extended 140 PRINCIPLES OF METALLOGRAPHY discussion of the connection between constitution and physical properties. Metallography Hiorns 1902 An introduction to the microscopic study of alloys. Physikalische Chemie der Metalle Schenk 1909. Six lectures on applications of physical chemistry to metallography. Traite" de Metallographie Robin 1912. A large work on general metallography including many subjects not usually considered. A complete set of diagrams and many microphotographs. Traite Complet D'Analyse Chimique Goerens and Robin 1911 Metallographie Microscopique. General metallography with many diagrams and microphotographs. IntermetaJic Compounds Chemical Combinations Among Metals Giua 1918 (Translated by Robinson). Theoretical treatment of the nature and properties of intermetallic compounds. Intermetallic Compounds Desch 1914. A monograph dealing with the constitution and physical properties. (hardness, conduc- tivity, etc.) of intermetallic compounds. Industrial Alloys (Non-ferrous) Alloys Sexton 1909. Deals with the physical and mechanical properties of the non-ferrous alloys and the methods of preparation. Alloys and Their Industrial Applications Law 1914. A book designed primarily for the engineer with a brief theoretical discussion and a detailed description of the mechanical proper- ties of technical alloys. Mixed Metals Hiorns 1912. A descriptive text-book on the composition, methods of manufacture and technical uses of non-ferrous alloys with a brief theoretical treatment. Practical Alloying Buchanan 1910. A description of foundry practice with alloy formulas. Essentially a foundryman's handbook. Iron and Steel *Cast Iron in the Light of Recent Research Hatfield 1912. An extended discussion of the properties of cast iron with many microphotographs. Deals with shrinkage, casting temperature, growth, etc. APPENDIX 141 Crystallization of Iron and Steel Mellor 1905. Six lectures on the applications of metallography in engineering. Die praktische Nutzanwendung der Priifung des Eisens mit Hilfe des Mikroskopes Preuss 1913. A brief discussion of the use of the microscope in the study of good and of defective steel and iron. Iron, Steel and Other Alloys Howe 1903. One of the earlier classics on the metallography of iron and steel by the dean of American metallographists. Metallographie und Warmebehandlung Hanemann 1915. A dis- cussion of applied metallography and heat treatment of steel, designed for engineers. * The Metallography and Heat Treatment of Iron and Steel Sauveur 1916. A detailed treatment of the metallography of iron and steel with many excellent microphotographs. One of the modern standard works on iron and stee? The Metallography of Steel and Cast Iron Howe 1916. A large book dealing with many theories of the behavior of iron and steel and especially with the mechanism of plastic deformation. Highly speculative and of interest to the advanced student. * Microscopic Examination of Steel Fay 1917. Intended for inspectors of steel ordnance materials. A very brief discussion of methods and many microphotographs of good and of de- fective material. The Microscopic Analysis of Metals Osmond 1913 (Translated by Stead). An English edition of Osmond's classic in the field of the microphotography of steel and iron. Includes many of Osmond's remarkable photographs of good and of defective material. * Physico-chemical Properties of Steel Edwards 1916. A discus- sion of the chemical and structural constitution of steel and the effects of heat treatment based on the equilibrium diagram. The Sampling and Analysis of Iron and Steel Bauer and Deiss 1912 (Translated by Hall and Williams) .Part I deals with the applications of metallography to the selection of suitable samples for analysis. Heat Treatment of Steel Heat Treatment of Tool Steel Brearly 1916. Essentially a book for the tool maker but gives many applications of metallography. Steel and Its Heat Treatment Bullens 1917. A detailed discus- 142 PRINCIPLES OF METALLOGRAPHY sion of the heat treatment of steel and the uses of the microscope in following the changes involved. Traitements Thermiques des Produits Me"tallurgiques Guillefr 1909 A discussion of the methods and results of quenching, tempering and annealing of alloys, chiefly iron and steel. Journals American Institute of Metals 1906-1918 (Incorporated with Am. Inst. of Min. Eng., 1919). Deals with the production and metallography of non-ferrous alloys. American Institute of Mining Engineers. Includes the metallo- graphy of iron and steel and since Jan., 1919 has published the non-ferrous metallography formerly printed in Am. Inst. of Metals. American Society for Testing Materials 1899 to date. In the metals section has published much information on the physical properties and metallography of alloys. Bureau of Standards, United States. Publishes frequent valuable monographs on various branches of metallography. Journal of the Iron and Steel Institute (British) 1896 to date. Deals with the preparation and metallography of iron and steel. Published semi-annually. Journal of the Institute of Metals (British) 1909 to date. A semi- annual publication of the Society which covers for non-ferrous alloys the field occupied by the Iron and Steel Inst. in the steel industry. Mitteilungen aus dem Eisenhiittenschen Instituts Aachen Wust 1906 to date. A series of valuable contributions to theoretical and industrial metallography. Zeitschrift fur Metallographie. Published original articles in English, French, Italian and German and included abstracts of all metallographic articles. In addition to the journals devoted primarily to the properties of metals the following journals frequently publish important articles on metallography. Chemical and Metallurgical Engineering, Metallurgie, Revue de M6tallurgie, Stahl und Eisen, Zeitschrift fur anorganische Chemie. APPENDIX TABLE 3 143 The Common Industrial Alloys. The composition of these alloys varies between fairly wide limits with a corresponding variation in physical properties and applications. The information given under the heading "Uses" is intended merely to indicate the general nature of the alloy. * Name Per cent, composition Uses Admiralty metal .... Aluminum brass .... Bronze, Aluminum Bearing Cu 70 Zn 29 Sn 1 Cu 70-68 Zn 27-31 Al 1-3 Cu 90, Al 10 Cu 70-90 Condenser tubes for use with salt water. Marine fittings. Propeller blades, rudder frames, sea valves. Hard, non-corrodible. Used in parts exposed to tanning, sulphite and simi- lar corrosive liquors. Bearings of various sorts. Gear Sn 1-10 Pb 0-15 Zn 0-27 Cu 89 Sn 11 Used for heavy gears usu- Gun Cu 88. Zn 10-8, ally against steel. Strong valves and fittings. Phosphor Plastic Sn 2, Pb 2 Cu 80-77 Sn 8-10 Pb 9.5-15 P 0-1.0 Cu 70-50 Bearing metal, wire, rods, steam fittings. Bearing metal. Silicon Pb 30-50 Ni trace Cu 96, Sn 4, Si tr. Telegraph wires, electrical Babbitt metal Sn 70-90 work. Bearings and antifriction "Genuine" Sb 7-24 Cu 2-22 Sn 88.9 Sb 7.4 Cu 3.5 lining for bronze bushings. 144 PRINCIPLES OF METALLOGRAPHY TABLE 3. Continued Name Per cent, composition Uses Bell metal Cu 80-75 Bells, gongs, etc. Sn 20-25 (Sometimes Ag, Ni or other metals) Brass, Gilding metal . . Cu 99-80 Cheap jewellery, gold paint. Zn 1-20 Dutch metal Cu 80-76 Thin sheets as substitute Zn 20-24 for gold leaf. Standard Cu 73-66 Brass for cold working. Zn 27-34 Sheets, tubes, cartridges. White Cu less than 45 Ornamental castings not requiring strength. Brazing metal Cu 85-Zn 15 Britannia metal Sn 95-90 Cheap table ware. Sb 5-10 Cu 1-3 Chromel Ni 60 Resistance wire for heating (Nichrome) Cr 40 units, crucibles, triangles. (approximate) tongs. (Patented.) Cupro-nickel Cu 98-52 Projectile driving bands, Ni 2-48 rifle bullet caps, electrical resistances. Constantan Cu 60. Used with copper or iron to Ni 40 ' make thermocouples. Delta metal Cu 60 See Sterro metal. Zn 40 Mn 0.5-2 Duralumin Al 95.5 Strongest and best of alumi- Cu 3.0 num alloys. Used in air- Mn 1.0 plane and automobile Mg 0.5 parts. Fusible metals Lipowitz Bi 50 These and other ternary and Pb 27 quaternary alloys are Sn 13 used for fuse plugs for Cd 10 automatic sprinklers. Woods Bi 38 Pb 31 Sn 15 Cd 16 APPENDIX 145 TABLE 3. Continued Name Per cent, composition Uses German silver See nickel silver. Gun metal Cu 92-88 Gears, heavy hydraulic cast- Sn 8-12 ings. Hercules metal Aluminum brass with Same as aluminum brass Fe with added toughness. Invar Fe 64 Low coefficient of expan- Ni 36 sion. Used in clocks, pre- cision instruments. Magnalium Al 90-94 Scientific instruments. Bal- Mg 10-6 ance beams. Magnolia metal Pb 78 (Lead base babbitt) Sb 16 Antifriction, bearing alloy. Sn 6 Manganese bronze-. . Cu 8S Propeller blades. Non- Sn 10 corrodible and great wear- Mn2 ing qualities. Manganin Cu 82 High electrical resistance ! Mn 15 and low temperature co- Ni 2.3 efficient. Fe 0.6 Monel metal Ni 72 Almost n o n-c orrodible. Cu 26.5 Used for propeller blades, Fe 1.5 wire, sheets, etc. Muntz metal < Cu 60 Sheathing for ships, bolts, Zn 40 nuts, condenser tubes. Naval brass Cu 62 Properties like Muntz Zn 37 metal. Less easily cor- Sn 1 roded by sea water. Nickeliu Cu 74.5 Resistance wire. Ni 25 Fe 0.5 Nickel silver Ni 18-25 Table ware, cheap jewellery, (German silver) . . . Zn 20-30 base for silver plating. Cu (Remainder) Palau Pd Substitute for platinum in Au chemical crucibles, dishes, etc. (Patented.) Pewter . . Sn 85-90 Platters, bowls, cups, etc. Sb 15-10 Little used at present. Platinoid Cu 60 High resistance wire but Zn 24 not suitable for heating Ni 14 coils. W 1-2 146 PRINCIPLES OF METALLOGRAPHY TABLE 3. Continued Name Per cent, composition Uses Platinite Fe54 Same coefficient of expan- Ni46 sion as glass. Used as C 0.15 substitute for platinum in equipping incandescent lamps. Platinum Iridium . . . Pt 90 Standard meter and other Ir 10 standards. Thermo- couple with platinum. Rheotan Cu52 High resistance but not Zn 18 suitable for heating coils. Ni25 Fe5 Shot metal Pb99 Casting bullets and small As 1 shot. Solder Soft Pb 67, Sn 33 Plumbers solder. Medium Pb 50, Sn 50 Hard Pb 33, Sn 67 Speculum metal. . . . Cu 70-65 Takes a high polish. Sn 30-35 Formerly used in reflectors for telescopes. Steel Plain Carbon. . . C 0.05-0.15 Boiler plate, rivets, sheet steel, case hardening stock. C 0.15-0.25 Structural work bridges, shafting. C 0.25-0.40 Axles, connecting rods, pis- ton rods. C 0.4-0.75 Rails, steel castings. C 0.6-0.8 Cutlery, wood working tools, drills. C 0.8-1.0 Springs, lathe tools, drills. C 1.0-1.2 Large lathe tools, axes, knives. C 1.2-1.5 Saws, files, balls for bear- ings, razors. Chrome Cr less than 3 Projectiles, files. Chrome-tungsten C 0.25-1 High speed tools. May be W 5-25 run at 500-600C. without Cr 2-10 losing their edge. Vd 0.25-1 APPENDIX TABLE 3. Continued 147 Name Per cent, composition Uses Steel (continued) "Chrome- C 0.25-1 Gears and springs. vanadiurn Cr 0.8-1.1 Vd 0.15 Manganese Mn 6-15 Used on sharp railroad curves, frogs, switches, etc., where wear is hard. Nickel Ni 3-4 Drive shafts crank shafts, gears and other automo- bile parts. Nickel- Ni 1-4 Armor plate chromium . . . Cr 0.45-2 Silicon Si, less than 5 Has high permeability and low hysteresis. Used in dynamo construction. Stellite Co 80-50 Non-corrodible. Used in Cr 20-50 cutlery, surgical instru- ments. (Patented.) Sterro metal Cu 60 Strong as mild steel and not (Aich's metal Delta Zn 38 easily corroded. Used in metal) : Fe 2 hydraulic cylinders, sea water valves. Type metal Pb 60-85 Sb 8-20 Sn 5-35' . 148 PRINCIPLES OF METALLOGRAPHY TABLE 4. Temperature Conversion Table (Condensed) Degrees Centigrade to Degrees Fahrenheit. Degrees Fahrenheit = % Degrees Centigrade + 32. Degrees Centigrade = % (Degrees Fahrenheit - 32). Decrees* Centigrade 10 20 30 40 50 60 70 80 90 Degrees Fahrenheit 32 50 68 86 104 122 140 158 176 194 100 212 230 248 266 284 302 320 338 356 374 200 392 410 428 446 464 482 500 518 536 554 300 572 590 608 626 644 662 680 698 716 734 400 752 770 788 806 824 842 860 878 896 914 500 932 950 968 986 1004 1022 1040 1058 1076 1094 600 1112 1130 1148 1166 1184 1202 1220 1238 I 1256 1274 700 1292 1310 1328 1346 1364 1382 1400 1418 1436 1454 800 1472 1490 1508 1526 1544 1562 1580 1598 1616] 1634 900 1652 1670 1688 1706 1724 1742 1760 1778 17961 1814 1000 1832 1850 1868 1886 1904 ! 1922 1940 j 1958 1976 1994 1100 2012 2030 2048 2066 2084 2102 ! 2120 2138 2156 2174 1200 2192 2210 2228 2246 2264 2282 2300 2318 2336 2354 1300 2372 2390 2408 2426 2444 2462 2480 2498 2516 2534 1400 2552 2570 2588 2606 2624 2642 ! 2660 | 2678 2696 2714 1500 2732 2750 2768 2786 2804 2822 1 2840 2858 2876 2894 1600 2912 2930 2948 2966 2984 3002 j 3020 3038 3056 3074 1700 3092 3110 3128 3146 3164 3182 3200 3218 3236 3254 1800 3272 3290 3308 3326 3344 3362 I 3380 ; 3398 3416 3434 1900 3452 3470 3488 3506 3524 3542 3560 3578 3596 3614 2000 3632 3650 3668 3686 3704 3722 3740 3758 3776 3794 APPENDIX 149 TABLE 5. Melting Points and Atomic Weights of the More Important Metals and Metalloids Element Symbol | weight j Melting point, deg. C. Aluminum Antimony Arsenic Barium Al Sb As Ba 27.1 120.2 75.0 137.4 658.7 630.5 850.0 (?) 850.0 Beryllium Be 9 1 1278 Bismuth Boron Cadmium Bi B Cd 208.0 11.0 112.4 271.0 2000.0-2500.0 (?) 320.9 Calcium Ca 40.1 800. Oca. Carbon C (Diamond) 12 > 3600.0 Cerium Ce 140 25 > 800.0 Caesium Cobalt Chromium Copper Gallium Gold Indium Iridium Cs Co Cr Cu Ga Au In Ir 132.9 59.0 52.1 63.6 70.0 197.2 115.0 193.0 26.0 1480.0 1520.0 1084.1 30.0 1063.5 155.0 2350.0 (?) Iron Fe 55.9 1530.0 Lanthanum La 138 9 810.0 (?) Lead Pb 206.9 327.4 Lithium Magnesium Manganese Mercury Li Mg Mn Hg 7.03 24.36 55.0 200 186.0 635.0 1260.0 -38.9 Molybdenum Nickel Osmium Mo Ni Os 96.0 58.7 191.0 2500.0 (?) 1451.0 2700.0 (?) Palladium Pd 106.5 1549.0 Phosphorus P 31.0 I. 44-11. 930 Platinum Pt 194.8 1780.0 K 39 15 62.5 Rubidium Rb 85.5 38.0 Ruthenium Selenium Ru Se Si 101.7 79.2 28.4 2450.0(7) 217.0 1420.0 Silver Ag 107.93 961.5 Sodium Na 23.5 97.5 150 PRINCIPLES OF METALLOGRAPHY TABLE 5. Continued Element Symbol Atomic weight Melting point, deg. C. Sulphur S Strontium Sr 32.6 87.6 f I. 112.8-II. 119.2 \ III. 106.2 >Ca,< Ba(?) Thallium Tl 204.1 302.0 Tellurium Te 127 6 450 Tin Sn Titanium Ti Tungsten W Uranium . Ur Vanadium V 119.0 48.1 184.0 238.5 51.2 232.0 1800.0 >3000.0 < 1850.0 1720.0 Zinc Zn 65.4 1 419 INDEX Abrasives 30 methods of use of 31 Admiralty metal 92 Alloy steels 113 diagram for 114 Alloys, burning of 133 definition of 1 eutectic 9 mechanical hardening of 72 mechanical testing of 138 microscopical examination of 29 overheating of 133 oxidation of 22 preparation of, for examination 29 stirring of, during melting 21 ternary 58 two-layer 3 Aluminum, alloys of 69 effect of, on brass . . . .' 92 preparation of, for examination 70 Aluminum bronze 92 Aluminum-lead alloys 6 Amorphous binding material 76 Amorphous cement theory 73 Annealing, defective 133 reasons for . . y, . . . 116 temperature of 116 Antimony-copper alloys 58 Antimony-lead alloys 10 structure of 14 uses of 16 Atomic per cent., calculation of 23 use of 22 Atomic weights, table of 149 Austenite 108 151 152 INDEX Babbitt metal 64 casting of 65, 133 segregation in 128 Bell metal 81 Bismuth-tin alloys 15 Books on metallography, list of 138 Brass 82 a-form 86 annealing of 88 0-form 84 cold-working of 83 experiments with 136 jewelry 83 solder 86 twinning of 86 white \ 86 Bronze 77 aluminum 92 bearing 79 coinage 78 diagram for ' - - 78 experiments with 137 gear 78,79 Government 79 manganese 92 phosphor 79 plastic & C Carbon, effect of, on steel 105 Case hardening 117 materials used for 117 temperatures for. . '. . ..." Cementation 117 Cementite 102 Kourbatoff's reagent for 104 spheroidized 107 Changes in solid state 56 Chemical composition, effect of 123 Chromel 49 Compounds, intermetallic Concealed maximum 52 diagram of 53 INDEX 153 Conductivity, effect of solid solutions on 71 Constantan 49 use of, in thermocouple 24 Containers for melted metals . . . 20 Cooling, rate of . . 25 Cooling curve, form of , 1, 2 plotting of 28 Copper 70 deoxidizing of 72 mechanical hardening of 72 properties of 72 Copper alloys, etching of 76 Copper-manganese alloys 43 diagram of . . 44 uses of 48 Copper-nickel alloys 48 diagram of 48 uses of 49 Copper-silver alloys 40 diagram of .' . 40 uses of . 42 Critical points 96 Critical range .... 96 D Diagrams, equilibrium .... 4 experimental, construction of . . 27, 28 freezing point . 4 interpretation of ... 7, 39 Duralumin 69 heat treatment of 70 Dutch metal . 83 E Etching 31 of copper alloys 76 of steel . .... . . ... ... . ......... 103 reagents for . ... . .,...-....-.- 32 Eutectic .......... 11 alloy 11 location of, by time lines 13 point 11 temperature 11 Eutectoid . 57 154 INDEX F Ferrite 102 Furnaces 18, 19 Fuse plugs 65 G German silver 64 Gilding metal 83 Gold, white 49 Grain size '. 88 factors for 90 formulas for 90 measurement of 88 Gun metal 78 H Heat treatment of steel 115 books on 141 High speed tools 115 "Hold, "definition of 12 Hydrogen, use of in melting alloys 22 I Industrial alloys. 143 books on 140 composition of 143 uses of 143 Intermetallic compounds 49 books on 140 hardness of 55 uses of 55 Iron, allotropic forms of 96 cast 118 critical points in 96 experiments with 138 gray 120 malleable 120 mottled 122 white 118 wrought 97 Iron and steel, books on . 140 INDEX 155 J Journals, list of 141 K Kourbatoff 's reagent for cementite 104 L Laboratory course in metallography 136 Laboratory methods of metallography 18 Lead, effect of, on brass 92 Lead-tin alloys 16 diagram of 16 uses of 16 Liquidus 11 curve 45 M Magnesium-tin alloys 51 Malleabilizing 120 Manganese bronze 92 Manganese steel . 114 Manganin 48 Martensite 108 Maximum, concealed : 52 open 50 Melting, methods of 27 Melting points, table of 149 Metallography, books on 138 definition of 1 laboratory course in 136 laboratory methods of 18 Metals, atomic weights of 149 melting points of 149 Microscope, metallographic 34 vertical illuminator for 33 Millivoltmeter, calibration of 25 types of 26 use of 24 Monel metal 49 Mounting photographic prints 36 cards for 37 Muntz metal 84 segregation in 128 156 INDEX N Nichrome '.' Nickel silver Non-ferrous alloys .... 69 O Open maximum diagram of with solid solutions Overheating, definition of I 33 P Palau 49 Pearlite Phase rule definition of terms diagram illustrating .... 67 uses of 67 Phosphide in steel . . . Plastic bronze segregation in . . . . - Plates, photographic development of exposure of . kinds of partial exposure of Polishing . by hand 30 mechanical Potentiometer, use of Printing, photographic Prints, mounting of .... 37 Pyrometers Q Quenching 115 incorrect * 3 S Season cracks 91,130 test for 91 INDEX 157 Segregation 124 in brass and bronze 124 in Muntz metal 128 in plastic bronze 128 in steel 125 Slag in iron 125 Slip bands ~. . . . 75 Solder, brass 86 plumber's 16 tin 16 Solid solution 39 curve of 46 development of 42, 43 effect of, on conductivity 71 freezing of 43 microscopic appearance of '. 47 Solidus 12 curve 45 Sorbite Ill production of 112 Specimens, preserving of 32 Spheroidizing 107 Steel 95 alloy 113 annealing of 115 case hardening of 117 chrome-tungsten : . . . 115 cold working of 130 composition of 98 diagram for 99 etching of 103 experiments with 137 heat treatment of 115 high speed tool 115 hot working of 130 hyper-eutectoid '.,.;.... .... .... 101 hypo-eutectoid 101 incomplete transformations in 101 manganese '. 114 metallographic constituents of . . . . V . . ... 102, 112 phosphide in 128 quenching of 108, 115 sulphide in tempering of 108, 115 uses of 107 Stellite. 49 158 INDEX Strain hardness 76 Sulphides in steel 125 detection of, by sulphur prints 126 Surrounding, appearance of 55 cause of . .54 Temperature, conversion table for . 148 measurement of 23 Tempering 115 incorrect temperature for 134 Ternary alloys 58 binary eutectics in 63 binary surfaces of 61 changes occurring in 63 composition of 60 diagram of, with contour lines 62 microscopic appearance of 64 space model of 59 solidus surface of 63 uses of 64 Testing, mechanical ........... 138 Thermal junction 24 Thermic analysis 4 Thermoelement 24 Thermos bottle, use of, for cold junction 24 Time curve 4 construction of 4 uses of 6,13 Tin, effect of, on brass . 92 Transition 54 Troostite 109 production of 110 W Weight per cent., conversion to atomic ...... t 23 Wood's metal . 65 1 24 UNIVERSITY OF CALIFORNIA, LOS ANGELES THE UNIVERSITY LIBRARY This book is DUE on the last date stamped below 19511 S NOV 4195 2 MAR 3 - DEC 1 1953 AUG 6 1959 AUG 2 7 1959 2 1976 P 519 THE L UNWERStl , O.. Uf LOS ANGELES 3 1158 OC ; SOUTHERN REGIONAL UBRARY FACILITY A 001 187889 9