T. I. Tikhonov.

On Babbitts

Tomsk.

Typolithography of the Siberian Association of the Printed Word, corner of Dvoryanskaya and Yamskaya lanes, own building.

1915.

Contents

	Page
Composition of babbitts. Structure and melting temperature of binary alloys of lead and antimony, of tin and antimony, of zinc and antimony, of copper and tin, of copper and antimony, of zinc and lead, and of tin and zinc	1—14
Experimental investigations of babbitts, carried out by Engineer I. N. Valuev	15—21
Experimental investigations of babbitts of lead, tin, and antimony, and verification of the diagram of ternary alloys of the indicated composition	21—39
Experimental investigation of zinc and lead–tin babbitts. Experimental investigation of babbitts of the Siberian Railway. Manufacture of babbitts by pressing	39—76

On Babbitts

The composition of babbitts used at the present time in machine-building practice is very varied, but in the babbitts most frequently employed the constituent elements are Pb, Zn, Sb, Cu, and Sn.

Molten babbitt in the liquid state in most cases constitutes a close, uniform mass of the metals composing it. In view of the fact that liquid babbitt in the ordinary case is a rather complex solution, when such a solution is cooled it is entirely possible that it will decompose into its constituent components, depending on the cooling conditions. Each of the constituent components, when the temperature of the solution falls below its melting temperature, will begin to separate out in the form of individual solid particles, often having a clearly crystalline form. As the temperature falls, the number of individual solidified components increases, and finally, at the last moment before complete solidification, the solution is a mixture of solid alloy particles that have separated out and differ in composition, floating in a liquid uniform mixture, which solidifies in the alloy last and is the cementing link between the solid alloy particles that separated earlier. In babbitts this last-solidifying part, called the eutectic, constitutes the softest and most viscous part of the babbitt; the crystals and particles of the alloy components that separated earlier always possess greater hardness in comparison with the babbitt eutectic. When, owing to the composition of the alloy itself and the nature of its cooling, the quantity of separated crystals increases so much that too little eutectic mixture remains to bind the crystals properly to one another, the babbitt becomes hard, but also brittle. In good babbitts the number of separated crystals is not large, the crystals are small, and there is a sufficient quantity of soft, plastic eutectic binding them. The mechanical properties of any complex body, obviously, are strongly dependent on the mechanical properties of the elements composing it, and from this it follows that the study

the properties of babbitts that are complex in composition must begin with the study of the properties of the individual elements composing it. The structure of solidified babbitt is in direct dependence on the composition of the babbitt itself; the more complex the composition of the babbitt and the more varied the number of its constituent elements, the more complex is the structure of the solidified babbitt, and from this it follows directly that, before examining the complex structures of multicomponent babbitts, it is necessary to become acquainted, at least briefly, with the structure of the binary elements usually entering into the composition of babbitts. The cheapest element entering into the composition of babbitts is lead. Lead with antimony, as was established by Heycock and Neville in 1890—90¹, gives a eutectic at 13% antimony in lead. The solidification curve of the binary alloy of lead and antimony, according to the observations of the above-mentioned persons and later investigations, has the following form (Fig. 1). In constructing this curve, along the axis of abscissas

Fig. 1.

the percentage content of antimony in lead is plotted, and along the ordinate axis—the solidification temperatures of the alloys. As is evident from the curve, as antimony in lead increases from 0 to 13%, the temperature

the melting of the alloys gradually falls from the melting temperature of pure lead, 326.9°, to 248.5°, and at the latter temperature a eutectic consisting of 87% lead and 13% antimony separates from the liquid solution in solid form. With a further increase in the antimony content in lead, the melting temperature of the alloys gradually rises and approaches the melting temperature of pure antimony. After Heycock and Neville’s work on determining the solidification curve of alloys of lead and antimony, experimental investigations were carried out by Stead (1897) and Charpy, and the structure of the alloys obtained was also studied. From Charpy’s experimental observations ²) it follows that the eutectic of lead and antimony contains 12.5% antimony and corresponds to the compound with the formula Sb Pb₁, the melting temperature of the latter compound being 247° C. The eutectic of the indicated composition is observed in all alloys of lead and antimony from 0 to 95% antimony; but at the latter antimony content Stead observed still another arrest of temperature during cooling at 615° C and assumed the existence of a second eutectic substance. Later observations by Cambell (1902) and Goutermann ³) (1907) introduced certain additions to the observations of the above-named authors and made it possible to construct the solidification curve of lead–antimony alloys in the form given earlier. According to Goutermann’s investigations, the eutectic is observed in all alloys of lead and antimony at 248.5° C, provided that the antimony content in lead varies from 1 to 99%. In addition to the common arrest at 248.5° C in the fall of the thermometer temperature during cooling of lead–antimony alloys, Goutermann observed still a second point of temperature arrest at 244.8°, common to all lead alloys when the antimony content varies from 1 to 99%. The appearance of the second indicated common point is explained by Goutermann by the different solubility of crystals depending on their size, namely: larger crystals are less soluble in comparison with smaller ones, and hence it follows that, for larger crystals of the eutectic, the temperature of separation of the crystals lies somewhat higher than the temperature of separation of small crystals of the same eutectic; but this latter explanation by Goutermann of the appearance of the second point at 244.8° during cooling of the alloy is not accepted by Guertler, who assumes that the difference in the thermal effect owing to the different solubility of crystals depending on their size in the present case cannot be so noticeable. Further, according to Goutermann’s observations, the duration—

²) and ³) Guertler. Metallographie. p. 794.

the duration of the arrest of the thermometer at 244.8° C for different contents of antimony in lead was very different, and this arrest was most prolonged at a content of 50% antimony in lead; and this gave Guertler occasion to suppose that, at the indicated ratio of the constituent parts of the alloy, a new compound of antimony with lead separates out, in which the antimony content is close to 50% (conventionally designated on the diagram as “III”). The existence of the latter crystalline substance with a melting temperature of 244.8° C (at 50% Sb) has still not been proved by later observations on alloys of lead and antimony.

Thus the solidification curve of alloys of lead and antimony, on the basis of the general observations made in this field, has the form indicated earlier (Fig. 1). As for the mechanical properties of alloys of lead and antimony, then, according to Gontermann’s data, alloys with 13–40% antimony possess an extremely nonuniform structure and poor mechanical properties.

The second pair of metals entering into the composition of babbitts is tin and antimony. The study of alloys of tin and antimony begins with Seebeck’s works ⁴) in 1822. The fusibility curve of these two metals was given by Thomson in 1841. Heycock and Neville (1889–1890) were the first to draw precise conclusions about the state of the various separations at different contents of one metal in the other; they were the first to notice the fact—unexpected for them—of a rise in the fusibility curve of tin as a consequence of an insignificant addition of antimony. Observations on the study of the nature of alloys of tin and antimony have continued up to the most recent days, and the observations of the later period have been accompanied by metallographic investigations. On the basis of the observations of Zhemchuzhny (1901), Gallagher (1906), Williams (1907), Konstantinov and Smirnov (1910), and on the basis of the report by Portevin (1909), Le Gris (1911), and Loebe (1911), Dr. Guertler in his work Metallographie (p. 803) gives the following solidification curve of alloys of tin and antimony (see Fig. 2). The peculiarities of the diagram presented consist in the fact that alloys of tin and antimony, with an increase in the content of the latter in tin, do not have a common eutectic, but, depending on the content of tin and antimony, separate out three independent eutectics. Williams first noticed, on cooling alloys, a weak arrest of the thermometer at 430° C. Later Reinders, at 60% antimony, observed three heterogeneous forms: long needles of substance “1” (see the diagram, conventionally desig-

⁴) Guertler. Metallographie. p. 802.

marked “!”, crystals of substances III and IV (whereby Reinders did not distinguish the two latter substances), and all this is surrounded by the main mass of the first substance. Williams found three independent forms

[Figure: phase diagram with axes labeled “atomic % antimony” (top) and “weight % antimony” (bottom), with temperature scale at left; regions are labeled in Russian, including “liquid solution,” “liquid solution + crystals II,” “liquid solution + crystals IV,” “eutectic + crystals I and III,” “eutectic + crystals IV and II,” and marked fields I, II, III, IV.]

Fig. 2.

of precipitates at 50% antimony in tin. Finally, Reinders, on the cooling curve of one alloy at 55% antimony, found four successive thermal arrests. At the present time, as something firmly established, it is necessary to recognize the existence of four crystalline forms of precipitates in alloys of tin and antimony.

A whole series of observations showed that tin, up to 10% antimony, does not change its structure and changes its melting point only very slightly. Only Williams, in his experiments, observed that tin up to 8% antimony was homogeneous, without crystals, but at 9% antimony, in the tin field, rhombs of another substance, called “III” in the diagram, were distinctly noticeable. In turn, antimony, according to the observations of many authors, dissolves up to 9—10% tin without any noticeable disintegration of its crystals. Thus, according to Williams’s observations ⁵), tin with 91% antimony, after heating for 36 hours at a temperature of 400° C, clearly revea—

⁵) Guertler. Metallographie. S. 810.

the precipitation of a crystalline substance (designated in the diagram as “IV”), but the same alloy, with 90% antimony in tin, after heating for the same time and at the same temperature, was completely homogeneous. The latter alloy of antimony with 10% tin is designated in the diagram as “II.”

For all alloys of tin, beginning from 50 to 90% antimony in tin, under conditions of slow cooling, most experimenters have always observed a general halt of the thermometer at 430°, corresponding to the solidification of the common eutectic mass for all alloys of the above-mentioned composition. This latter substance (called “IV” in the diagram), which separates from the alloy at 430° C, has a crystalline form different from that of the substance cited earlier (“II”), which is a solid alloy of antimony with 10% tin. The crystalline substance “IV” is observed in alloys of tin with antimony in all those cases where the concentration of the alloy ranges from 50 to 57% antimony, which gave grounds for supposing that the composition of substance “IV” corresponds to the formula Sn Sb.

It must be noted that the point marked on the diagram (at 63% antimony) for the beginning of the separation of the crystalline substance “IV” on the eutectic horizontal at 430° C has not yet been firmly established.

The second halt of the thermometer during the cooling of tin alloys containing from 20 to 50% antimony is found, very weakly, at 319° C. Reinders explains the weak thermal effect of this latter halt of the thermometer by the fact that the substance separating at this temperature differs very little from the preceding separation “IV.” This latter substance, provisionally designated in the diagram as “III,” according to Stead’s data is most distinctly observed at a content of 40% antimony, and this gave grounds for supposing that this compound is a new crystalline form corresponding to the compound Sb₂ Sn₃. The metallographic compound Sb₂ Sn₃ differs from Sn Sb only by a slight difference in color upon etching, the base of the crystals being Sn Sb and the coating Sn₃ Sb₂. Some authors do not consider the compound Sn₃ Sb₂ at all.⁶) Thus, summarizing the previously cited results of examining the solidification curve of tin alloys with antimony, we have the following: all tin alloys containing from 0 to 10% antimony, after slow cooling after melting, constitute a uniform mass of solidified tin with 10%

⁶) Evangulov. Alloys. p. 311.

antimony. The same kind of homogeneous mass will occur in a solidified alloy of tin with antimony in the case where the antimony content in the tin is above 90%, but in the latter case the homogeneous mass will consist of antimony with 10% tin. All other tin alloys containing from 10 to 90% antimony, practically depending on the antimony content in the tin, in the solid state have clearly visible crystals of Sn Sb and Sn₃ Sb₂, immersed in a common homogeneous mass of tin with 10% antimony or antimony with 10% tin.

Next, the following pair of metals entering into the composition of babbitts and causing their hardness are zinc and antimony. Zinc with antimony gives a compound that is very hard, but also brittle. Of the existing compounds of zinc with antimony, most authors recognize two: Sb₂ Zn₃ and Sb Zn. The solidification curve of alloys of zinc and antimony, according to Guertler’s data, has approximately the following form⁸) (Fig. 3). Zinc is capable of dissolving antimony in itself in considerable quantity, and in doing so alloys are separated out whose melting temperature is higher than that for zinc and, from the addition of antimony

Fig. 3.

in itself antimony and, in doing so, alloys are separated out whose melting temperature is higher than that for zinc and, from the addition of antimony

⁸) Many ambiguities in the curve, arising from the different data of various authors in describing the curve, have been omitted; details on the curve may be found in Guertler’s book, Metallographie, pp. 777–785.

grow. An alloy containing 40% antimony ($\mathrm{Sb}_2\mathrm{Zn}_3$) has the highest melting temperature, 565° C, and on polished sections the alloy has a distinctly crystalline structure (conventionally denoted on the solidification curve as “IV”). Between the particles of zinc and of the zinc alloy with 40% antimony, upon gradual cooling, according to the data of Mönkemeyer and Arnemann, at a temperature very close to 412.5° C a common eutectic mass separates, in which the antimony content, according to Arnemann’s data, is about 2.5% by weight. With an increase of the antimony content in zinc from 40 to 100%, the melting temperature of the alloy begins gradually to decrease, and when the antimony content in zinc reaches 62% (atomic), the melting temperature of the alloy will be 482° C. For all alloys of zinc with antimony, beginning from 40 to 100% antimony in zinc, under conditions of slow cooling, a constant arrest is observed in the fall of the thermometer temperature at 482° C, which evidently corresponds to the solidification of a common eutectic mass for all alloys of zinc with antimony, provided that the antimony content in zinc is from 40 to 100%. One of the components of the latter eutectic mass is antimony (conventionally “III” in the diagram), and the other is a compound of zinc with antimony. This latter compound of zinc with antimony, as a whole series of observations by various authors has shown, in zinc alloys containing from 40 to 68% (atomic) antimony, under conditions of slow cooling after melting, itself decomposes. The most intense decomposition of the crystals of the indicated zinc compound with antimony, according to the data of Mönkemeyer and Zhemchuzhny, is observed at a temperature of 538° C, corresponding to 50% antimony in zinc, and it must be assumed that this temperature of 538° C corresponds to the separation of a new crystalline substance of the formula $\mathrm{SbZn}$. According to the data of the above-mentioned investigators, the thermometer-arrest point at 538° C is observed for all zinc alloys containing from 40 to 50% (atomic) antimony, which gave grounds to suppose that in the indicated interval of the cooling curve of the alloys, together with the previously indicated crystals of the alloy of formula $\mathrm{Sb}_2\mathrm{Zn}_3$ (crystals “IV”), other crystals of the zinc alloy with antimony of formula $\mathrm{SbZn}$ (crystals “VI”) also separate. Further, the same authors found that the melting point of the eutectic mixture of antimony and $\mathrm{SbZn}$ for all zinc alloys having more than 50% antimony lies on the horizontal at 506° C, and then also the percentage content of antimony in the latter eutectic of the alloy of zinc and antimony is very close to 68% (atomic). Finally, many investigators found arrests in the fall of the thermometer temperature

upon cooling alloys of zinc and antimony in all cases below 400° C, when the zinc alloys contained from 0 to 40% (atomic) antimony; moreover, the arrest point of the thermometer for all zinc alloys poor in antimony was about 321° C, and for alloys containing 40% antimony—about 360° C. The existence of these latter arrest points of the thermometer in the indicated interval of the solidification curve of zinc and antimony alloys gave many authors grounds to suppose the formation of a new crystalline form (crystals “V”), which, in all probability, corresponds to a crystalline transformation of the zinc itself, the temperature of whose crystalline transformation varies somewhat depending on the antimony content in the zinc. The different structural phases of zinc and antimony alloys, depending on the percentage content of antimony in the zinc and depending on the cooling conditions, are clearly shown by the designations on the solidification curve of the indicated alloys (Fig. 3).

Up to this point we have considered compounds with antimony of lead, tin, and zinc, which enter into the composition of babbitts; these compounds, as is known, have a crystalline structure and impart hardness to babbitts. In addition to antimony compounds that impart hardness to babbitts, in certain babbitts containing copper and tin there separate out very hard particles of compounds of the two latter elements. Without going into the details of the various compounds of copper and tin², it must be noted that in babbitts containing copper and tin, in the majority of cases one form of such compounds is encountered, namely: pale-pink, hexagonal stars on the polished section, having the formula of the compound $\mathrm{SnCu}_3$ (33.3% Sn). These star crystals are distinguished by great hardness and, with relief polishing of the section, always stand out clearly. It must be noted that the separation of hexagonal stars of the compound $\mathrm{SnCu}_3$ is observed in tin babbitts at 5% copper, and such an alloy is distinguished by the best mechanical properties. In babbitt alloys containing lead, antimony, and copper, in addition to hard crystals of lead with antimony or of pure antimony, pinkish separations of antimony with copper according to the formula $\mathrm{SbCu}_2$ are also noticeable on the polished section; moreover, it must be assumed that this compound increases the hardness of babbitts. A similar picture of the separation of crystals of the compound $\mathrm{SbCu}_2$ is also observed in tin babbitts with a high tin content. Bearing in mind that the brittleness of crystals of $\mathrm{SbCu}_2$ is greater than that of crystals of the compound $\mathrm{SnCu}_3$, it is necessary to conclude that tin—

an alloy with a lower copper content will be better, for the greater brittleness of babbitt contributes to its rapid wear. In antifriction alloys, in zinc babbitts, which sometimes contain a considerable amount of copper, deposits of copper and zinc in the form of brass are found in the polished sections, and these brass deposits, in all probability, increase the toughness of zinc babbitts.

Babbitts, by their purpose, cannot be especially hard, for hard babbitt linings can work well only on the condition of very accurate fitting of the working surfaces and of uniform loading. In new machines the latter conditions can never be observed, and the working surfaces of the liners must run in for a long time. But even in old babbitt liners, concentration of the load on a small area can be caused by accidental contamination, as a result of which local seizure of the working surfaces may appear at this area. For the reasons indicated above, it is highly desirable that the material of the liners possess a certain plasticity, under which it can easily deform and thereby promote a uniform distribution of the load over the rubbing surfaces. To satisfy the general requirements stated above for babbitts, it is necessary to choose babbitts of such a composition that, in structure, they consist of hard grains bound together by a plastic cement. During operation, the soft mass of the babbitt on the rubbing surface is quickly worn away, and only the hard particles of babbitt remain in contact with the working journal; moreover, these hard particles of babbitt must be well joined together by the plastic mass. In general, the entire mass of babbitt must be sufficiently plastic for uniform distribution of the load. With such a structure of the lining metal for liners, local contamination from dust and dirt is easily removed together with the lubricant, and hard grit that has accidentally entered remains pressed into the soft mass of the liner metal.

The soft mass of existing babbitts consists of tin, lead, and zinc in various proportions among themselves, depending on the composition of the babbitt. For completeness in considering the question of the structure of babbitts in general, it is necessary, if only briefly, to examine the possible compounds of the various metals with one another in the soft mass of babbitts. Tin and lead, when fused with each other, do not form chemical compounds. Both metals are soft; as the content of one in the other increases, the hardness of such an alloy in-

melts and reaches a maximum in the eutectic alloy consisting of 62% tin and 38% lead. The hardness of such a eutectic mixture will be less than the hardness of an alloy of 17% antimony and 83% lead³).

Next, the following pair of metals included in the soft part of babbitts are zinc and lead. The solidification curve of zinc–lead alloys, according to Guertler’s data, is shown in Fig. 4. Zinc, on the basis of a whole series of observations by Heycock, Neville, Arnemann, and others, can dissolve lead up to 4% without a noticeable change in its crystallization; the melting point of such a homogeneous mass will be about 417°, i.e., below the melting point of pure zinc (419.22° C). On the fusibility curve, the end of separation of a homogeneous mass (Fig. 4) from zinc and lead is to be taken at 3% Pb.

Fig. 4.

at \(418^\circ\) C, the melting temperature. The curve shown for the solidification of the alloy of zinc and lead within the limits from \(0.5\) to \(96.6\%\) lead in zinc was obtained experimentally by Arnemann (1910), and the general picture of the structure of the alloy obtained by him is represented on the curve. With a further increase in the lead content above \(96.6\%\) in zinc alloys, one must assume a further fall in the melting temperature of the alloys up to the time when a dense homogeneous mass of lead begins to separate from the alloy, with an insignificant content of zinc dissolved in it; moreover, the solidification temperature of the latter alloy of lead with an insignificant amount of zinc will be lower than the melting temperature of pure lead. According to the data of Matthiessen and Rose\(^ {11}\), the solubility of zinc in lead, without any noticeable change in crystallization, reaches \(1.6\%\).

Finally, the third pair of metals entering into the soft part of babbitts is tin and zinc. The solidification curve of this latter pair of metals (see Fig. 5) was constructed by Guertler from the data of Neycock, Neville, Wiedemann, Mazzotto, Arnemann, and others, and has, in general, the following significance: as is seen from the curve, all alloys of zinc and tin have one common eutectic mass with a melting temperature close to \(197^\circ\) C, while the composition of the common eutectic will be \(83.3\%\) (atomic) tin in zinc. Further, Arnemann observed at high temperatures (about \(300^\circ\)) a small residual—

Fig. 5.

\(^ {11}\) Guertler. Metallographie. S. 557.

of the thermometer during the cooling of the alloy, which he explained by the transformation of particles of zinc \(\beta\), containing a little tin, into particles of pure zinc \((\alpha)\). Finally, Mazzotto, in his observations, obtained slight arrests in the fall of the temperature of the alloy during cooling, already after the separation of the common eutectic; and this phenomenon was attributed by Mazzotto to a transformation of particles of tin. But further observations showed that the latter arrests of the thermometer are observed even in those cases when the zinc alloy contains very little tin. Thus the last-mentioned phenomenon in the solidification curve of the alloys of zinc and tin remains unexplained to the present time.

All existing babbitts, according to the predominating amount of one of the elements, may be divided into alloys of lead, tin, and zinc.

Binary alloys of lead and antimony, owing to their considerable cheapness, are often used in practice for casting bearings. From the previously considered solidification curve of such an alloy (see Fig. 1) it follows that the indicated alloy has a eutectic at \(13\%\) Sb, and with a higher content of antimony the alloy consists of solid crystals of antimony joined by the above-mentioned eutectic, which has sufficient plasticity. The brittleness of the grains of antimony does not harm the alloy, for these grains can easily shift in the plastic mass of the eutectic. With an antimony content above \(25\%\), the antimony crystals have a considerable size, touch one another closely, and the alloy, as a consequence, becomes very brittle. As a material for casting bearings, alloys of lead with \(13—25\%\) Sb are usually used, but these alloys are not suitable for high specific pressures. Dudley’s well-known alloy for bearings of slowly rotating shafts contains \(83.5\%\) Pb and \(16.5\%\) Sb.

A little copper is not infrequently added to alloys of lead with antimony, but in this case in the above-described structure of such alloys there appears a (slightly pinkish) compound of copper with antimony according to the formula \(\mathrm{SbCu}_2\). Thus the alloy used on the railways of the C-e d’ Est consists of \(65\%\) Pb, \(25\%\) Sb, and \(10\%\) Cu, while with a higher copper content the alloy becomes brittle.

Alloys of tin and antimony, commonly used for casting bearings, as follows from the previously considered solidification diagram of these alloys as a function of the antimony content (see Fig. 2), consist of rhombic crystals of the compound of tin and antimony according to the formula \(\mathrm{Sb}_2\mathrm{Sn}_3\) and \(\mathrm{SbSn}\), joined by a plastic mass of tin, the separated crystals being softer than pure

antimony. In practice, alloys of tin and antimony, owing to their considerable costliness, are used very rarely, and therefore the content of antimony in such alloys rarely exceeds 25%. To alloys of tin and antimony, in order to increase hardness, a little copper is often added; in this case, in the structure of the polished section, alongside the crystals of compounds of tin with antimony ($\mathrm{Sb}_2\mathrm{Sn}_3$ or $\mathrm{SbSn}$), there separate out yellow fir-tree crystals of the compound of copper with tin, according to the formula $\mathrm{SnCu}_3$, which in themselves possess great hardness. According to Charpy’s investigations, the alloy possessing the best properties is one having 83.33% Sn, 11.11% Sb, and 5.56% Cu, for at a higher copper content the hard fir-tree crystals of $\mathrm{SnCu}_3$ disappear.

As babbitts, in practice there are often used alloys of lead, tin, and antimony, with the antimony content in such alloys ranging from 10 to 18% and tin from 10 to 20%. The structure of such alloys consists of rhombic crystals of compounds of lead with antimony or tin with antimony, but the plastic eutectic binding them, under the influence of an admixture of tin in the lead, will be harder than the pure lead eutectic of a simple lead-antimony alloy, as a result of which the latter ternary alloys of lead, tin, and antimony are often used for casting liners subjected to a heavy load. Finally, among babbitts one must include an alloy of zinc with antimony with a slight admixture of tin; usually the content of tin and antimony in such alloys ranges between 10 and 15%. Zinc with antimony, as follows from the previously given solidification curve (see Fig. 3), gives a compound in crystalline form that is very hard and brittle, and, moreover, the mass cementing them is characterized by sufficient brittleness; therefore pure alloys of zinc and antimony are not used for casting bearings. A slight admixture of tin reduces the brittleness of the eutectic of the alloy, but on the other hand the tin content must not be greater than the indicated limit, for with a higher tin content in the alloy, alongside the compound of zinc with antimony, crystals of the compound of tin with antimony will begin to separate out, which will be softer than the crystalline compounds of zinc with antimony. In order to retain the tin in the eutectic of the alloy and at the same time to cause only crystals of zinc and antimony to separate out, it is necessary in the ternary alloy of zinc, tin, and antimony to keep the content of the latter within the above-mentioned limits. To increase the toughness of zinc alloys with antimony,

not infrequently a small quantity of copper is added to the alloy; in the latter case, on polishing, it is separated out in the form of individual rods of brass.

The composition of babbitts used in practice is very varied, and in most cases the adopted content of the babbitt used has not been subjected to critical evaluation. Only in recent times have several attempts been made to standardize the composition of the babbitts used on Russian railways, where, as is known, the consumption of babbitt is very considerable. Thus, at the XXV consultative congress of engineers of the traction service, the question was raised of the composition of the bronzes and babbitts used on Russian railways, and, by order of this congress, engineer I. N. Valuev addressed the administrations of the traction service of Russian railways with a request to report on the grades, compositions, and quantities of the materials consumed (bronzes and babbitts), on the methods of acquiring and manufacturing them, on the results of their service, on the technical conditions for their acceptance from factories and firms, and so forth. Together with the replies to the above-stated questions, samples of the materials used were also placed at Mr. Valuev’s disposal; these were in turn investigated by Mr. Valuev chemically and mechanically. The results of the latter investigations and the general conclusions drawn from them were reported by Mr. Valuev at the XXVIII consultative congress of engineers of the traction service⁴. Without dwelling on the various details in the investigations and results for the bronze samples that were at Mr. Valuev’s disposal, let us turn directly to the investigations carried out on babbitts. The chemical investigations of the babbitts were performed by the laboratory of the Moscow Electrolytic Works of I. K. Nikolaev, and their specific gravities were also determined there. Determinations of the hardness, ductility, and strength of the babbitt samples were made by Mr. Valuev, with the kind assistance of the administration of the Russo-French Shipbuilding Works, in the laboratory of that works. The general picture of the investigations of babbitts undertaken by Mr. Valuev was as follows:

Hardness was determined by pressing in a ball, according to Brinell’s method, with a Frémont machine. Ductility and strength were determined from graphical records on the same Frémont machine by crushing specially prepared bars of rectangular cross-section, \(8^{m}/_{m} \times 10^{m}/_{m}\), transverse to their length, with a steel cylin-

with a rod having a diameter of \(10^{m}/m\). The cylinder (with a length greater than \(10^{m}/m\)) was placed across the bar and gradually pressed into it until the bar split in half or produced a transverse crack. During this time, a pencil drew a curve on the diagram, the ordinates of which indicate at each moment the depth of penetration of the cylinder into the bar and the corresponding force of pressure on the cylinder at that moment. It is clear that the greater the viscosity possessed by the bar, the deeper one can press the cylinder into it without destroying it; and the bar may be considered the stronger, the greater the pressure on the cylinder at which the bar is destroyed. If we assume that we have a bar of absolute viscosity, i.e. one into which the cylinder can be pressed, without rupturing it, through its entire thickness, and call the viscosity of such a bar the number 100, then the viscosity of all other bars will be expressed as a percentage of the absolute viscosity: it will be represented by numbers ranging from 0 to 100. Such numbers are shown in the corresponding column of the tables given below. Strength in these same tables is shown by numbers proportional to the crushing force (for clarity), namely, 27 times smaller. ¹³) Microscopic investigations were also made of polished sections, likewise prepared by the factory of I. K. Nikolaev.

The coefficient of friction and the various degrees of wear of the babbitts under test were not determined. It is known that the coefficients of friction of different compositions of babbitts differ comparatively little among themselves, and that the quality of the lubricating oil used has a much greater influence on the frictional force. Tests for wear require a large expenditure of time and, nevertheless, cannot be placed under conditions such as occur in actual service.

Observations were also made of the temperatures of the melting and cooling points of the alloys. These data are not included in the tables, since suitable, well-equipped physical instruments were not available; direct observation with a thermometer, however, shows that for all the compositions of babbitts considered these data differ only slightly, namely: the beginning of melting is about \(230—235^\circ\), and the end of melting about \(240—260^\circ\).

In practice, in some cases, the length of the period of time between the moment when, during cooling, the thickening of babbitt begins (the end of melting, \(240—260^\circ\)) is of importance,

¹³) On the diagrams of the Frémont machine, \(1^{m}/m\) of abscissa was equivalent to 27 klg.

and the moment of complete solidification. This time is sometimes needed for production, the so-called “spreading” of the babbitt. The duration of the spreading period is the greater, the greater the heat capacity of the composition, which, in turn, is in direct relation to the heat capacities and the quantities of the metals entering into the composition. If the heat capacity of lead is taken as unity, then the heat capacity of the remaining constituent metals will be as follows: tin 1.75, copper 3, and antimony 1.7. Consequently, those compositions will spread the better which contain more tin and less lead; in other words, this capacity for “spreading” of a given babbitt is in direct dependence on its chemical composition.

Mr. Valuev first gives the results of his observations in one large general table, and then, in order to elucidate more clearly the composition of babbitts depending on their purpose, Mr. Valuev makes a selection from the general table, distributing the compositions of babbitts according to their purpose. The result of such a distribution of babbitts is the four tables below (I–IV), and it is necessary to note that the column “price by volume” is determined by Mr. Valuev according to the theoretical cost of the babbitts, with a correction for specific gravity. The volume of a babbitt having a specific gravity equal to 10 is taken as unity.

Then the cost of other babbitt compositions required to fill the same volume occupied by one cube of babbitt of specific gravity equal to 10 will be less than the cost of one cube of these compositions in direct proportion to their specific gravities.

Table 1.

Compositions of babbitts used only for fast-running locomotives.

No.	Roads	Firm or plant	Cu	Sn	Sb	Pb	Zn	Hardness	Viscosity	Strength	Price by volume
4	South-Western	Kiev workshops . . .	3,96	83,50	11,96	0,50	Trace	27,1	27	47,5	15,09
9	V.-Vensk.	Main workshops . .	4,26	80,00	13,00	2,60	Trace	28,4	30	47	15,00
20	Ryaz.-Ural	Leikin’s plant . .	4,40	70,00	13,04	12,00	0,13	27	7,5	22	13,45
24	Siberian	Own workshops . .	2,32	63,00	14,60	19,70	0,12	26,2	17	34	12,69
34	Siberian	Nikolaev plant . .	4,16	50,00	14,96	30,00	0,12	26,8	7,5	28	11,31
49	Siberian	R. A. M. O. . .	3,00	42,70	14,25	39,00	0,40	20,2	10	25	10,34
58	Chinese-Eastern	Nikolaev plant . .	3,20	40,00	16,20	40,00	0,12	22,5	4,5	20	9,96
76	Chinese-Eastern	R. Fr. (Verda) . .	1	26,30	9,20	63,26	Trace	18,3	23	32	8,16
102	Chinese-Eastern	T-vo A. S. L. . .	3,40	20,00	19,00	57,50	0,22	35,5	1,5	23	7,20

Table II.

Compositions of babbitts used for locomotives of all series indiscriminately.

No.	Railway	Firm or factory	Cu	Sn	Sb	Pb	Zn	Hardness	Viscosity	Strength	Price by volume
1	Northern	Own workshops	3,48	89,20	7,20	Tr.	—	23,1	38	47	15,84
2	Syzr.-Vyazem.	Leykin Factory	5,32	84,50	9,40	0,60	0,09	27,6	35,6	53	15,44
8	Mosc.-Brest	Own workshops	6,34	80,30	13,20	Tr.	—	30,3	17,3	42,5	14,79
10	Mosc.-Brest	Bryansk Factory	6,00	80,00	12,00	2,00	—	30	25	45	15,32
11	Samara-Zlat.	Own workshops	9,84	79,50	10,60	Tr.	—	33	18,1	43	17,23
18	Polish	Own workshops	7,40	72,00	18,28	2,02	—	36,1	5,3	38	13,78
19	Polish	Durans wheel	9,20	72,00	17,40	1,36	—	39,4	5,6	28,5	14,31
24	Siberian	Own workshops	2,32	63,00	14,60	19,70	0,12	26,2	17	34	12,69
25	Siberian	R. A. M. O.	2,00	62,60	15,00	20,00	0,12	25,5	17,5	33,5	12,49
30	Mosc.-Kursk	Own workshops	4,84	54,60	7,16	33,32	Tr.	18,7	26	34	12,27
42	M. V. R. (m.s.)	No. 1, own workshops	3,04	45,00	13,72	38,00	Tr.	21,9	24	27	10,19
44	Ekaterininsk.	Nikolayev Factory	3,40	44,82	11,24	40,00	0,21	26,4	12	27	10,67
50	South-Eastern	Lisovsk. Factory	3,36	42,50	13.60	40,00	0,16	20,8	11	24	10,31
54	Lib.-Romen	Own workshops	1.40	36,00	16,96	45,00	—	20	7	21	9,23
69	Lib.-Romen	R. A. M. O.	2,40	30,50	17,30	49,30	0,18	27,6	20,5	47	8,90
82	North-Western	St. Petersburg workshops	4,3	25,36	8,20	62,14	—	(recipe)	(recipe)	(recipe)	—
124	Lib.-Romen	Own workshops	0,20	0,60	18,50	80,50	Tr.	20	17,5	31	3,69

Table III.

Compositions of babbitts used for passenger cars.

No.	Railway	Firm or factory	Cu	Sn	Sb	Pb	Zn	Hardness	Viscosity	Strength	Price by volume
5	South-Western	Kiev workshops	5,34	83,00	11,28	0,29	Tr.	29,4	13,4	40,5	15,10
18	Polish	Own workshops	7,40	72,00	18,28	2,02	—	36,1	5,3	38	13,78
31	Lib.-Romen	Own workshops	4,60	54,64	10,20	30,00	0,13	23	20	28	11,85
35	Northern	Moscow workshops	3,92	49,20	12,80	34,00	Tr.	16,3	11,25	24,5	14,43
39	Northern	Schenker No. 1	4,20	46,00	12,28	37,31	0,12	22,4	12	27	11,40
40	Mosc.-Brest	Own workshops No. 2	1,80	46,00	9,60	42,00	0,18	17	20	26,5	10,80
46	Mosc.-Kursk	Own workshops No. 1	3,64	44,00	6,36	45,80	Tr.	19,8	15	25	10,99
54	Lib.-Romen	Own workshops	1,40	36,00	16,96	45,00	—	20	7	21	6,23
55	Lib.-Romen	Durans daifa	0,70	40,80	12,00	46,00	0,22	15,8	24	28	10,74

No.	Railways	Firm or works	Cu	Sn	Sb	Pb	Zn	Hardness	Viscosity	Strength	Price by volume
65	South-Western	South-Russian Works	1.80	34	8.88	55	0.22	20.2	6	21	9.35
77	Chinese-Eastern	Leikin Works	2.80	26.12	14.80	56.00	0.22	26.8	5.6	24	8.14
84	Northern-Vyazem.	same	3.00	24.90	13.40	58.50	0.12	25	6	23	7.99
96	M. V. R. (m.s.)	Own workshops	4.00	22.00	10.00	63.5	0.24	23.1	9.4	31	7.81
101	Ryaz.-Ural.	Lisovsk. Works	2.24	21.00	16.60	59.80	0.12	22	16.25	34	7.25
109	South-Eastern	same	1.00	15.00	13.90	69.70	0.20	23.5	5.6	29	6.28
116	same	R. Fr. (Berda)	1.40	6.00	16.94	75.00	0.27	22.7	5.17	25.38	4.82

Table IV.

Compositions of babbitts used for freight cars.

No.	Railways	Firm or works	Cu	Sn	Sb	Pb	Zn	Hardness	Viscosity	Strength	Price by volume
41	Tashkent	Main Workshops	1.70	48.00	7.20	42.80	0.17	17.1	33	30	10.89
54	Lib.-Romen.	Own workshops	1.40	36.00	16.96	45.00	—	20	7	21	9.23
62	Northern	Moscow Workshops	3.20	37.50	21.00	38.20	0.12	30.5	2.35	20.5	9.63
68	Ekaterinin.	Nikolaev Works	2	31.94	15.00	50.60	0.18	22.5	7.5	25	8.82
71	same	Leikin	2.20	22.80	18.20	50.00	0.24	27.2	3.75	29	8.41
79	Siberian	Nikolaev	2.36	26.00	13.36	57.50	0.25	24	6	23	8.16
81	Mosc.-Brest	Own worksh. No. 3	—	25.70	25.30	48.70	0.20	33.2	1.9	22	7.64
86	Polish	Pinsk worksh.	3.92	24.00	7.60	64.20	0.12	20.5	14	32	8.05
94	—	Doransk. Podsh.	0.60	22.50	7.40	69.00	0.20	13.3	20	26	6.57
95	—	Schwendler	4.20	22.50	23.64	49.44	0.17	33.6	1.6	23.5	7.57
99	South-Western	South-R. T-vo	2.60	21.50	18.40	57.30	0.24	29.7	2	25	7.31
101	Ryaz.-Ural.	Lisovskii Works	2.24	21.00	16.60	59.80	0.12	22	16.25	34	7.25
112	—	Gr. Kurovskie “Magnolia”	0.40	10.00	12.00	77.50	0.24	22.9	8.4	31	5.40
117	Central-Asian	Nikolaev Works	2	5.76	17.50	74.50	0.14	25.2	10	28	4.95
118	South-Eastern	Lisovskii Works	Tr.	5.00	15.50	79.40	Tr.	18.3	12	24	4.67
120	—	Henry Ratzel	0.10	3.05	16.20	80.40	0.20	20.3	15.6	30	4.01
121	North-Western	Balt. Worksh.	—	1.00	17.00	81.40	0.20	20.7	11.27	25.31	3.68
122	Central-Asian	Sormovo Works	0.35	0.49	16.20	82.50	Tr.	19	17	30	3.61

As is evident from the tables presented, the composition of babbitts used on Russian railways for the same purpose—

very varied, and, at the same time, it must be noted that the enormous majority of the traction-service administrations of the indicated railways, sending samples to Mr. Valuev, expressed complete satisfaction with their performance.

Further, in order to resolve the question of the best composition of babbitt depending on its purpose, Mr. Valuev, on the basis of general theoretical considerations, divides all babbitts into tin and lead babbitts, and at the same time considers the best to be all those babbitts which, in tests, revealed good mechanical properties; for the lower limits of good mechanical properties, the following were chosen: for hardness 20, for toughness 10, and for strength 25. Finally, in concluding his report, Mr. Valuev gives the following general Table V, in which he enters the best babbitts according to his tests.

Table V.

The best babbitts.

No.	Railways	Firm or works	Composition by analysis: Cu	Sn	Sb	Pb	Zn	Hardness	Toughness	Strength	Price by volume
		Of pure tin:
2	Syzr.-Vyazem.	Leikin Works . . .	5,32	84,50	9,40	0,60	0,09	27,6	35,6	53	15,14
9	V.-Vyatka	Main workshop . .	4,26	80,00	13,00	2,60	Tr.	28,4	30	47	15,00
14	—	Bryansk Works . .	4,92	78,90	12,40	3,80	—	28,3	28,75	55	14,70
4	South-Western	Kiev workshop .	3,96	83,50	11,96	0,50	Tr.	27,1	27	47,5	15,09
		Of tin:
24	Siberian	Own workshop . .	2,32	63,00	14,60	19,70	0,12	26,2	17	34	12,69
25	—	R. A. M. O. No. 1 .	2,00	62,60	15,00	20,00	0,12	25,5	17,5	33,5	12,47
26	South-Western	Kiev workshop .	3,20	59,62	11,68	25,00	0,12	23	18,1	34	12,52
36	North-Western	Balt. workshop . . .	2,00	48,80	14,00	35,00	0,14	21,4	36,5	31	11,20
		Of lead:
69	—	R. A. M. O. A. . .	2,40	30,50	17,30	49,30	0,18	27,6	20,5	47	8,90
116	—	R. Fr. (Berda) . .	1,40	6,00	16,94	75,00	0,27	22,7	5,17	25,38	4,82
101	Ryaz.-Uralsk.	Lisovsky Works	2,24	21,00	16,60	59,80	0,12	22	16,25	34	7,25
86	Polish	Pinsk workshop . . .	3,92	24,00	7,60	64,20	0,12	20,5	14	32	8,05
		Of pure lead:
121	North-Western	Balt. workshop . . .	—	1,00	17,00	81,40	0,20	20,7	11,27	25,31	3,68
124	Lib.-Romensk.	Own workshop . .	0,20	0,60	18,50	80,50	Tr.	20	17,5	31	3,69
120	—	Henry Ratzel . .	0,10	3,05	16,20	80,40	—	20,3	15,6	30	4,01
122	Central-Asian	Sormovo Works . . .	0,35	0,49	16,20	82,50	Tr.	19	17	30	3,61

From consideration of the last table, Mr. Valuev draws the conclusions that: 1) purely tin babbitts, when correctly formulated, possess high qualities, can satisfy the most stringent requirements imposed by railways on babbitts, and stand up to competition. 2) The qualities of tin babbitts have no advantages over lead babbitts; 3) as regards purely lead babbitts, in toughness they are equal to tin and lead babbitts, yielding to the latter only in hardness. At the same time, according to Mr. Valuev’s conclusions, purely lead babbitts surpass all the others in the stability of their properties, as a result of which the author expresses the supposition that, for the needs of railways, both by reason of its low cost and in its mechanical properties, a lead babbitt consisting of 83% lead and 17% antimony is quite suitable. In especially responsible cases of railway practice, Mr. Valuev recommends using a mixture made from the best tin babbitts, containing 5.55% Cu, 83.33% Sn, and 11.11% Sb, and from lead babbitts of the composition given above. To verify this last proposal of his, Mr. Valuev carried out a series of experiments, which in general gave good results. For preparing a mixture from the best tin and lead babbitts, Mr. Valuev advises first preparing separately the two above-mentioned babbitt compositions and then fusing and mixing them in the desired proportion. The point is that in tin babbitt, during cooling after melting, there should form the chemical compounds indicated above, which, evidently, in a concentrated solution will form more fully than in a solution diluted with lead.

Alongside the study of the mechanical properties of babbitts of various compositions, metallographic investigations of the latter were also carried out. Thus K. Loebe in 1911⁵ published his studies of babbitts made of lead, tin, and antimony. Campbell and Elder, on the basis of their observations of the above-mentioned composition of babbitts, gave an approximate diagram⁶ of the ternary alloy of tin, lead, and antimony. The indicated diagram is constructed on the general principle that the sum of the perpendiculars dropped from any point inside an equilateral triangle is equal to its height. The results of the experimental investigations, represented in the diagram (Fig. 6), are as follows: 1) the highest melting point of the ternary alloy (in the diagram

point O) of 80% Pb, 10% Sn, and 10% Sb is found at a temperature of 245° C, i.e. approximately 2 degrees below the solidification point of the eutectic of lead and antimony. 2) The lower melting point

Fig. 6.

(on the diagram, point P) was found in an alloy containing 40% Pb, 57.5% Sn, and 2.5% Sb, at a temperature of 189° C, consequently 7° above the melting temperature of the eutectic mass of the binary alloy of lead and tin. It is necessary to note that the positions of the lines Of, OP, and dP were plotted by W. Campbell approximately. The general significance of the diagram shown is as follows: at the vertices of the triangle A—there is antimony, at B—lead, and at C—tin. The eutectic mass of lead and antimony has 13% Sb and on the diagram corresponds to point r. The eutectic mass of lead and tin, at 62% Sn with a melting temperature of 182° C, lies at point q. On the line AC the equality of separations from tin and antimony (Sn Sb) is denoted by point g; f is the point of arrest of the thermometer at 422° C, d—the point of arrest of the thermometer at 240°, and e—a similar thermometer point for separations of tin in the form α (Snα) at 10% Sb.

The essence of the diagram presented, in general terms, is as follows: in the region Ar of—during the solidification of the alloy, antimony first begins to separate out in close combination with an insignificant-

amount of tin. Over the area of the diagram Br OPq, at first, during the solidification of the indicated ternary alloy, lead separates out; over the area fo Pd—first, from the solution as it solidifies, Sn Pb separates out, and, finally, over the area Cd Pq—tin first separates out in a dense compound with antimony (the tin is conventionally denoted “α”—Snα). Finally, if, for the sake of simplification, one assumes that, during the solidification of the above-mentioned ternary alloy, pure antimony separates out, then over the area of the diagram Aor, as the alloy solidifies, crystals of antimony will separate out until the remaining composition of the alloy reaches the boundary indicated on the diagram by the line ro. After this, lead and antimony will separate out simultaneously, while the composition of the liquid remainder of the alloy proceeds along the line ro and until it reaches the point O. At the concentration of the alloy corresponding to the point O,—particles of Sn Sb + Pb separate out from the alloy as it solidifies. If the concentration of the solution is greater than that corresponding to the point O, then particles of Sn Sb and Pb will continue to separate out from the solution along the line O p until the complete solidification of the alloy.

In the field Bro, during the solidification of the alloy, pure lead will first separate out until the concentration of the liquid remainder reaches the points of the line ro. The further picture of the solidification of the alloy will be the same as in the preceding case.

Over the area of the diagram Aof, during the solidification of the alloy, crystals of antimony first separate out until the liquid remainder of the alloy reaches the concentration of the points of the line fo; after this particles of Sn Sb will begin to separate out, and the concentration of the remaining liquid residue of the alloy approaches the point O. The picture of the solidification of the particles of the alloy at its concentration corresponding to the point O remains unchanged in comparison with the preceding cases. In the field f PO, the first product of solidification in the alloy is Sb Sn until the composition of the liquid remainder of the alloy reaches the boundary indicated by OP. In the latter case, particles of Sn Sb and Pb will begin to separate out simultaneously from the liquid remainder of the alloy, but, in doing so, the composition of the alloy residue will approach along the line OP toward the point p, at which, at the last moment of the solidification of the alloy, a dense mass of Snα + Pb separates out.

Over the area BPq, during the solidification of the alloy, particles of lead first separate out; these, we conventionally assume, do not contain tin dissolved in them, and such separation of particles of pure lead from the liquid remainder of the alloy will continue until

the concentration of the liquid residue does not reach the boundary Pq, at which particles of lead and tin \((\mathrm{Sn}\alpha)\) would simultaneously separate out from the liquid residue of the alloy, and the liquid residue of the alloy does not approach the eutectic (point q) of the binary alloy of tin and lead, which is the last to solidify. Over the area Cd Pq, during solidification of the alloy, particles of tin \((\mathrm{Sn}\alpha)\) will first separate out until the liquid residue approaches the boundary pq.

The line ro is the boundary of the antimony and lead particles of the alloy fields, separating during the solidification of the latter at the same temperatures. The line OP is the boundary between the lead fields and the Sb Sn fields and indicates to us where the simultaneous solidification of lead particles and Sb Sn occurs. Further, the line pq is the boundary and at the same time the line of simultaneous solidification of Pb and Sn. Finally, the line fo indicates to us the limit of transformation at which Sb Sn particles separate out at the expense of Sb, and the line dp is the limit of transformation of Sb Sn particles with the formation of \(\mathrm{Sn}\alpha\).

In conclusion, it must be noted that the ternary-alloy diagram of antimony, lead, and tin analyzed above is, in general, rather complex, but W. Campbell succeeded in confirming its main essential points by means of a series of metallographic photographs, for which he metallographically analyzed 49 alloys of different composition made of lead, tin, and antimony.

The investigation of babbitts of various chemical composition has been carried out by me repeatedly in the Metallographic Laboratory of the Tomsk Technological Institute; at the same time it is necessary to note that students of the Institute have taken and continue to take part in the indicated works, and, in the form of diploma projects, are assigned small investigations in the aforementioned field. Unfortunately, the equipment of the Institute’s laboratories does not make it possible to investigate babbitts for wear and heating in mechanical installations, since the Institute does not have electric power available around the clock, and the above-mentioned laboratory has no funds for the construction of the necessary apparatus.

We shall begin the description of the small experimental investigations of babbitts carried out in the metallographic laboratory of the Institute with experiments to verify the conclusions of Loebe and Campbell (see the diagram, Fig. 6). The first works aim to clarify the mechanism of solidification of a whole series of ternary alloys of lead, tin, and antimony. In order to introduce some system into the indicated problem and thereby facilitate the work itself, all experiments were divided into several series,

quantities by weight of Pb, Sb, and Sn in the ternary alloys were altered in such a way that the ratio of the weights \(\dfrac{\mathrm{Sn}}{\mathrm{Pb}}\) in each series was constant.

In all, in this way, 7 series of experiments were carried out:

I-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/1. \]

1) \(\mathrm{Sb} = 5\%\); \(\mathrm{Pb} = 47.5\%\); \(\mathrm{Sn} = 47.5\%\)
2) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 45\%\); \(\mathrm{Sn} = 45\%\)
3) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 40\%\); \(\mathrm{Sn} = 40\%\)
4) \(\mathrm{Sb} = 25\%\); \(\mathrm{Pb} = 37.5\%\); \(\mathrm{Sn} = 37.5\%\)
5) \(\mathrm{Sb} = 30\%\); \(\mathrm{Pb} = 35\%\); \(\mathrm{Sn} = 35\%\)

II-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/3. \]

1) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 67.5\%\); \(\mathrm{Sn} = 22.5\%\)
2) \(\mathrm{Sb} = 15\%\); \(\mathrm{Pb} = 63.75\%\); \(\mathrm{Sn} = 21.25\%\)
3) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 60\%\); \(\mathrm{Sn} = 20\%\)
4) \(\mathrm{Sb} = 25\%\); \(\mathrm{Pb} = 56.25\%\); \(\mathrm{Sn} = 18.75\%\)

III-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/4. \]

1) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 72\%\); \(\mathrm{Sn} = 18\%\)
2) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 64\%\); \(\mathrm{Sn} = 16\%\)
3) \(\mathrm{Sb} = 30\%\); \(\mathrm{Pb} = 56\%\); \(\mathrm{Sn} = 14\%\)

IV-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/5. \]

1) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 75\%\); \(\mathrm{Sn} = 15\%\)
2) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 66.6\%\); \(\mathrm{Sn} = 13.4\%\)
3) \(\mathrm{Sb} = 30\%\); \(\mathrm{Pb} = 58.3\%\); \(\mathrm{Sn} = 11.7\%\)

V-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/6. \]

1) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 77.1\%\); \(\mathrm{Sn} = 12.9\%\)
2) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 68.6\%\); \(\mathrm{Sn} = 11.4\%\)
3) \(\mathrm{Sb} = 30\%\); \(\mathrm{Pb} = 60\%\); \(\mathrm{Sn} = 10\%\)

VI-th series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = 1/7. \]

1) \(\mathrm{Sb} = 5\%\); \(\mathrm{Pb} = 83.1\%\); \(\mathrm{Sn} = 11.9\%\)
2) \(\mathrm{Sb} = 10\%\); \(\mathrm{Pb} = 78.75\%\); \(\mathrm{Sn} = 11.25\%\)
3) \(\mathrm{Sb} = 20\%\); \(\mathrm{Pb} = 70\%\); \(\mathrm{Sn} = 10\%\)
4) \(\mathrm{Sb} = 25\%\); \(\mathrm{Pb} = 65.6\%\); \(\mathrm{Sn} = 9.4\%\)
5) \(\mathrm{Sb} = 30\%\); \(\mathrm{Pb} = 61.25\%\); \(\mathrm{Sn} = 8.75\%\)

VI series.

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}}=\frac{1}{8}. \]

1) \(\mathrm{Sb}=5\%\); \(\mathrm{Pb}=84.45\%\); \(\mathrm{Sn}=10.55\%\)

2) \(\mathrm{Sb}=10\%\); \(\mathrm{Pb}=80\%\); \(\mathrm{Sn}=10\%\)

3) \(\mathrm{Sb}=15\%\); \(\mathrm{Pb}=75.5\%\); \(\mathrm{Sn}=9.5\%\)

4) \(\mathrm{Sb}=20\%\); \(\mathrm{Pb}=71.1\%\); \(\mathrm{Sn}=8.9\%\)

5) \(\mathrm{Sb}=25\%\); \(\mathrm{Pb}=66.6\%\); \(\mathrm{Sn}=8.4\%\)

The weight amount of the elements entering into each of the alloys totals 100 grams.

The positions taken for testing the above-mentioned alloys are shown, according to the Campbell and Elder diagram, in Fig. 6-a.

Fig. 6-a.

To obtain the alloys, an electric furnace of the firm “Kryptol Gesellschaft” was used.

To determine at every moment the temperature of the molten metal, a Le Chatelier pyrometer was used; the pyrometer was lowered through an opening in the furnace lid into the crucible, while the conductors running from the pyrometer were connected to copper clamps placed in a reservoir with ice; from the copper clamps thick copper conductors led to the galvanometer, which indicated the temperature of the molten metal.

Obviously, having the temperature of the given alloy at each moment, one can construct, in the usual way, the solidification curve.

The form of the solidification curve makes it possible to determine the critical points, establishing the moment at which a change in the structure of the alloy begins and characterized by the fact that at this moment a sudden slowing in the fall of temperature is observed, owing to the liberation of heat. In addition, the solidification curve of the alloy as a function of time was obtained with the aid of the recording pyrometer of N. S. Kurnakov, upon the detailed description of which we shall not dwell.

The experiment was carried out as follows: first the crucible, placed in the cup of the above-mentioned electric furnace, was heated to approximately \(400^\circ\) C; then a definite quantity of lead was lowered into it (melting point \(325^\circ\) C); when the temperature of the molten lead reached approximately \(500^\circ\) C, the corresponding quantity of antimony was lowered in, the mass being stirred with a wooden stick and sprinkled with charcoal dust, in order to protect the antimony from burning out.

Then, when the temperature of the liquid binary alloy reached about \(600^\circ\) C, tin was added.

Observations and readings on the solidification of each alloy were made from the moment when the temperature of the cooling alloy fell to \(650^\circ\), for which purpose the temperature was first brought up to \(700^\circ\) C.

Marking, by the galvanometer, every minute the temperature of the cooling alloy, the cooling curve was constructed from the points.

At the same time the recording pyrometer each time reproduced on photographic paper the solidification curve, and in order to obtain on the latter permanent points corresponding, for example, to \(600^\circ\) C, \(500^\circ\) C, etc., the small lamp of the apparatus was switched off with a plug, thereby obtaining on the curve temperature marks in the form of white intervals.

The cooling of each alloy was carried down to \(150^\circ\) C.

Before entering into the investigation of the material obtained, let us dwell somewhat on the method of constructing the fusibility diagram of ternary alloys. The fusibility curve of binary alloys can be represented on a plane—indeed: by laying off along the abscissa axis the various percentage contents of the element \(X\) of the binary alloy \(x + y\), and along the ordinate axis the corresponding temperatures at which arrests in the fall of temperature were observed, one obtains a fusibility curve lying in the plane.

This will not be so for ternary alloys: in this case the curves cannot be expressed on a plane, and here it is already necessary to resort to the construction of fusibility surfaces, plotting melting temperatures in a direction perpendicular to a certain coordinate plane that determines the composition of the ternary alloy.

For constructing the diagram one uses the “triangle” method. An equilateral triangle MNP is constructed (see Fig. 6-b), which, as indicated earlier, has the property that the sum

Fig. 6-b.

of the distances of any point \(x\), taken inside the triangle, from the sides \(MN\), \(NP\), and \(MP\) is a constant quantity, i.e.

\[ xm+xn+xp=\mathrm{conts}. \]

If we take \(xm+xn+xp=100\%\), then any ternary alloy can be expressed by some point \(A\).

By erecting at each of the points perpendiculars to the plane \(MNP\) and laying off on them the solidification temperatures of the corresponding ternary alloys, one may obtain, approximately, the surface shown in Fig. 6-b.

The perpendiculars \(\overline{1-N}\), \(\overline{2-M}\), and \(\overline{3-P}\) correspond to the melting temperatures of pure \(N\), \(M\), and \(P\).

The curves \(\overline{1-c-3}\), \(\overline{1-a-2}\), \(\overline{2-B-3}\), lying in planes passing through the sides of the triangle perpendicular to the plane \(MNP\), are the solidification curves of the binary alloys \(N\) and \(M\), \(N\) and \(P\), \(M\) and \(P\). These curves possess the eutectic points \(c\), \(a\), and \(B\).

The three surfaces \(\alpha\), \(\beta\), and \(\gamma\), intersecting one another, give the lines of separation from the alloy of binary eutectic mixtures; these lines in

in turn converge at one point \(A\), determining the temperature and composition of the ternary eutectic mixture, i.e., the eutectic alloy is determined by point \(A\).

On the basis of what has just been presented, it is possible to explain the mechanism of solidification of ternary alloys in general in the following manner.

First the component present in excess separates out (the 1st characteristic point), while the remaining mixture approaches more and more, in composition, a binary eutectic; and the separation, so to speak, of the excess component will continue until one of the lines of the binary eutectic is reached. Thereupon the second component, present in excess, begins to separate out simultaneously with the first (by a “component” of an alloy one must understand the free elements and those compounds which are not decomposed under the conditions considered), i.e., the separation of the binary eutectic will occur; such separation of the binary eutectic will continue until the remaining part corresponds to the composition of the ternary eutectic. This moment will be characterized by the appearance of the 3rd characteristic point on the solidification curve. Now the eutectic mixture too will begin to solidify, and this solidification will end at the temperature corresponding to the melting point of the ternary eutectic mixture.

It is evident that the construction of the solidification surfaces is a rather laborious task.

In order to simplify the matter considerably, it is sufficient only to project the curves of intersection of the surfaces onto the plane of the triangle \(MNP\), and the entire mechanism of solidification will be represented on the plane in the form of lines of one kind and another; these lines may give, in their intersection, one or several points, just as also occurs in binary alloys. These lines are found on the basis of cooling curves and the microscopic structure of the alloys.

The result of the latter construction is the diagram of Campbell and Elder considered earlier (see Fig. 6), but in order to verify the correctness of their conclusion, the ternary alloy of lead, tin, and antimony was taken; moreover, in the majority of the alloys lead is the predominant element. For the investigation an area of the diagram (see Fig. 6-a) close to point \(O\) was taken.

Point \(O\) can be established on the basis of cooling curves and microscopic structure. From a series of alloys with the ratio

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = \frac{1}{7}, \]

an alloy consisting approximately of 10% anti-

we, 79% lead and 11% tin, yielded a characteristic solidification curve. Only one halt in the fall of temperature was observed, at about 250° C.

Obviously, at this temperature, from the constituent parts of the alloy there formed a compound which then began to solidify without any subsequent transformations; otherwise another point should have appeared, determining the moment of a change in the structure of the alloy.

The polished section of such an alloy has a uniform structure of a dark gray color, and here and there small crystals of Sn Sb occur in the form of rhombi.

Consequently, the alloy under consideration differs somewhat from the eutectic alloy, as a result of which it was possible for a small quantity of Sn Sb crystals to appear, not even marked on the solidification curve. The observed alloy was somewhere near point O, but in any case in the region f O P (rhombi Sn Sb). As is seen, point O differs somewhat from Campbell’s O; namely, it seems to have shifted to the right of point O of the latter diagram. Let us next consider alloys from the region Br o Pq, i.e. from the region of excess lead, marked on the diagram by the numbers 1, 2, and 3.

The 1st alloy (Sb = 10%, Sn = 10%, Pb = 80%) lies near the line r o. The solidification curve gives two points, one at 252° C, and the other at 250° C. These two points make it possible to indicate that the alloy must lie on the line r o (in reality, slightly lower).

Thus the microscopic structure should show: a small quantity of the separated double eutectic of lead and antimony (temp. 252° C), surrounded by the products of transformation at O (the higher eutectic Sb Sn + Pb (see Fig. 7, Table I).

The solidification curve for alloys 2 and 3, situated in this same region and containing 5% antimony with

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = \frac{1}{8} \]

and

\[ \frac{\mathrm{Sn}}{\mathrm{Pb}} = \frac{1}{7}, \]

gives three characteristic points. For alloy 2, the first characteristic point is at 273° C; the second at 240° C, and the third at 190° C. On the basis of these data we can draw for ourselves a picture of the change in structure; the temperature of the last characteristic point (190° C) indicates that complete solidification occurred at point P, and consequently the change of the alloy proceeded along the line O P, but the alloy itself did not reach point O (temperature 240° C, and not 250°), but somewhere farther—along o P (below O). Alloy 3, judging by the solidification curve, passed through the same phases as the 2nd, i.e., first—

the separation of lead, then the separation of the double eutectic \(Sn\,Sb\) and \(Pb\), and finally complete solidification at point \(P\) at \(190^\circ C\).

The microscopic structure confirms what has been said: against the background of the polished section the crystals \(Sn\,Sb\) are clearly visible, as are dark places, or rather, separate dark inclusions (lead), and all this is surrounded by the eutectic at \(P\).

For alloys 4, 5, and 6 we have 3 characteristic points and, consequently, in the alloy we have separated antimony, a double eutectic \((Sb\) and \(Pb)\) along \(r-o\), and a eutectic mixture at \(O\). The microstructure of alloys 4 and 5 confirms this: the white crystals are antimony; the light mass is the double eutectic of antimony and lead; all this is surrounded by the higher eutectic at \(O\). It should be noted that the 2nd and 3rd characteristic points for alloys 4 and 5 almost coincide; from this one may conclude that the separation of the double eutectic began almost at point \(O\) and therefore continued for a very short time (see Fig. 7-a, Table I, alloy No. 5).

We shall not dwell in detail on the examination of the microstructure of all the alloys, but shall consider only the more characteristic of them, namely 15, 14, and 20—respectively \(30\%\), \(20\%\), and \(10\%\) antimony at

\[ \frac{Sn}{Pb}=\frac{1}{4}. \]

The fusibility curve of alloy 15 has three characteristic points, and in the microscopic structure of this alloy we see plates of antimony, in places rhombs \(Sn\,Sb\), and all this is surrounded by the products of transformation at \(O\). Alloy 14 contains a small quantity of antimony and a larger quantity of \(Sn\,Sb\) in the form of rhombs, and all this is surrounded by the eutectic at \(O\) (see Fig. 7-b, Table I). Alloy 20 lies in the region \(FOP\) and in its structure contains a few crystals of \(Sn\,Sb\), and a light mass—a eutectic mixture of \(Sn\,Sb\) and \(Pb\); all this is surrounded by the eutectic mixture at point \(P\), since the 3rd characteristic point has a temperature of \(190^\circ C\) (see Fig. 7 c, Table I).

Alloys 23, 24, 25, 26, 27, lying in the region \(FOP\), in their structure contain no separations of antimony at all (see Fig. 7d, Table I, alloy 25).

Alloy 27 has a uniform structure and evidently corresponds to an alloy on the line \(OP\), i.e., from the very beginning the separation of \(Sn\,Sb\) and \(Pb\) proceeded (see Fig. 7 e, Table I). The cooling curve clearly confirms this, since we have only 2 characteristic points: one at \(200^\circ C\)—the moment at which separation of \(Sn\,Sb\) and \(Pb\) from the alloy begins, and the other at \(189^\circ C\)—the solidification of the eutectic mixture at point \(P\).

Thus, on the basis of the experimental investigations just described of alloys of lead, tin, and antimony, one may come to the conclusion that the correctness of the fundamental propositions of Loebe and Campbell concerning the character of solidification and the structure of the above-mentioned alloys is confirmed by observations of the indicated alloys.

For further observations, in order to verify the conclusions from the diagrams of Loebe and Campbell, ternary alloys were taken, situated, in their composition, nearer to the other two vertices of Campbell’s triangle; for this purpose 24 alloys were taken, the percentage content of the constituent parts in these latter alloys being chosen not arbitrarily, but according to a definite system, given in the following Table V-a.

Table V-a.

Group	Ratio	Alloy	Pb	Sb	Sn
Group I	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/17$	Alloy No. 1	10	5	85
Group I	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/17$	Alloy No. 7	20	4.44	75.56
Group I	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/17$	Alloy No. 13	30	3.89	66.11
Group I	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/17$	Alloy No. 19	40	3.34	56.66
Group II	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/5$	Alloy No. 2	10	15	75
Group II	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/5$	Alloy No. 8	20	13.33	66.67
Group II	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/5$	Alloy No. 14	30	11.67	58.33
Group II	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/5$	Alloy No. 20	40	10	50
Group III	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/2$	Alloy No. 3	10	30	60
Group III	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/2$	Alloy No. 9	20	26.67	53.33
Group III	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/2$	Alloy No. 15	30	23.33	46.67
Group III	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1/2$	Alloy No. 21	40	20	40

Group	Ratio	Alloy	Pb	Sb	Sn
Group IV.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1$	Alloy No. 4	10	45	45
Group IV.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1$	Alloy No. 10	20	40	40
Group IV.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1$	Alloy No. 16	30	35	35
Group IV.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 1$	Alloy No. 22	40	30	30
Group V.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 2$	Alloy No. 5	10	60	30
Group V.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 2$	Alloy No. 11	20	53.33	26.67
Group V.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 2$	Alloy No. 17	30	46.67	26.33
Group V.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 2$	Alloy No. 23	40	40	20
Group VI.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 14$	Alloy No. 6	10	84	6
Group VI.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 14$	Alloy No. 12	20	74.66	5.34
Group VI.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 14$	Alloy No. 18	30	65.33	4.67
Group VI.	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}} = 14$	Alloy No. 24	40	56	4

The weight quantity of the elements entering into each of the alloys was taken in total as 200 gr. The alloys were melted in an electric furnace of the firm “Kryptol-Gesellschaft.” For measuring the temperature a “Le Chatelier” pyrometer was used.

To obtain the solidification curves, the apparatus of N. S. Kurnakov was used.

To obtain an alloy, a chamotte crucible was first heated in the above-mentioned electric furnace to 400° C; then a weighed portion of lead was lowered in, the lead was heated to 500° C, a weighed portion of tin was added, and at an alloy temperature of 600° C antimony was immersed. The mixture was stirred with a wooden stick. When the antimony had melted and the temperature of the bath had risen to 720–740°, the current was switched off, and at a temperature of 700° the apparatus of N. S. Kurnakov was switched on, the pyrometer readings being recorded every minute. At temperatures of 600°, 500°, 400°, 300°, and 200° the lamp in Kurnakov’s apparatus was switched off in order to fix constant points on the photographic paper. Cooling was continued down to 150°, and for some alloys—

to 130° C, after which the Kurnakov apparatus was switched off. Then, in order to free the thermometer from the alloy, the alloy was heated a second time.

In view of the fact that for the experiments it was necessary to have polished sections from slowly cooled alloy, and specimens for hardness testing from rapidly cooled alloy, one half of the alloy was accordingly cooled slowly, and the other rapidly.

This was done as follows: on a gas burner a fire-clay crucible was heated to 450°—500°, after which half of the obtained alloy, at a temperature of 620—750°, was poured into it; the crucible was covered with an asbestos plate, the gas was turned off, and in this way it was left to cool. The other half of the alloy was poured into a thick-walled cast-iron mold, whereby a quenched casting was obtained.

It was stated earlier that Campbell’s diagram was taken as the basis for the preliminary investigations; for greater clarity, Fig. 8 shows the position, over the area of the triangle, of the alloys taken for the subsequent experiments, and, in addition, indicates the observed critical points of the alloys investigated, with a conventional designation of the observed structure of the latter.

Turning to this diagram, we see that alloys 1 and 7 lie in the region cdpq, alloy 13 lies very close to the line pd, and alloy 19 close to the line qp.

For alloy No. 1 two critical points were obtained: 227° and 189° C. The polished section presents an even field of two colors (see Fig. 8-a, Table II).

Accepting the fundamental propositions of Campbell’s diagram, from the data obtained by us we must come to the conclusion that at first Sn α separated out (227°), then the process proceeded along the line pd and everything solidified at point p.

Alloy No. 7. The cooling curve gave two points: 207° and 181°. The polished section presents a field of two colors with separate white bodies, in which a tendency toward the formation of rhombs is noticeable. Such a character of the polished section indicates that Sn Sb first separated out; further, the process must have proceeded along the line pd—Sn α (207°), then along pq, and everything solidified at q.

With such an order of cooling the alloy must lie above the line pd. The temperature of separation of Sn Sb was not detected owing to the insignificant amount of Sn Sb.

Alloy No. 13. The temperature curve gave two points: 195° and 181°. The character of the polished section is the same as for alloy No. 7, but here white elongated bodies appeared; these bodies may be taken as forma-

formation of Sn Sb, since in the polished section their tendency to form a single whole is noticeable. The order of cooling is the same as in No. 7 (see Fig. 8-h, Table II).

Fig. 8.

Alloy No. 19. The temperature curve gave two points: 191° and 181°. The polished section presents a field of two shades, with a small number of rhombi. This indicates that Sn Sb was first precipitated; but how the further cooling proceeded is difficult to say: everything solidified at point q.

Alloys 2 and 8, according to the diagram, lie in the region fpd, and 14 and 20 in the region fop.

Alloy No. 2. The temperature curve gave three points: 290°, 220°, and 187°. The polished section presents a main field of two shades, in which

arranged white bodies of rhombic form. Comparing these data with Compbell’s diagram, we conclude that the white bodies are Sn Sb \((290^\circ)\), while the main field must be \(Sn\alpha\) \((220^\circ)\), and either the ternary eutectic of point \(p\), or \(Sn\alpha+Pb\), if everything cooled on the line \(pq\) (see Fig. 8-c, Table II).

Alloy No. 8. The temperature curve gave three points: \(280^\circ\), \(203^\circ\), and \(188^\circ\).

The character of the polished section is the same as No. 2, and everything that was said regarding the latter remains true for No. 8 as well.

Alloy No. 14. The temperature curve gave three points: \(260^\circ\), \(190^\circ\), and \(188^\circ\).

The character of the polished section is identical with Nos. 2 and 8. Everything said about them is applicable here as well. The temperature \(190^\circ\) is very close to the lower temperature \(188^\circ\); this indicates that the alloy reaches a state characterized by the line \(pd\), very close to point \(p\).

From all considerations concerning this alloy one must come to the conclusion that it lies to the right of the line \(fp\), i.e., in the region \(fpd\).

Alloy No. 20. The temperature curve gave two points, \(240^\circ\) and \(181^\circ\). The character of the polished section is as follows (see Fig. 8-d, Table II): in the upper part there are white bodies of rhombic form; as one approaches the lower part of the polished section these crystals acquire ragged edges and, in the very lowest part, pass into a network characteristic of \(Sn\ Sb+Pb\). Segregation is strongly pronounced here, but nevertheless the following conclusion may be drawn: first \(Sn\ Sb\) separated out \((240^\circ)\), then on the line \(op\)—\(Sn\ Sb+Pb\) (the temperature did not appear, evidently owing to the insignificance of the separation), and everything solidified at point \(q\).

Alloy No. 3. According to the diagram it lies in the region \(fpd\), while alloys 9, 15, and 21 lie in the region \(fop\).

Alloy No. 3. The temperature curve gave three points: \(360^\circ\), \(210^\circ\), and \(188^\circ\).

The polished section presents white bodies (see Fig. 8-e, Table II) with angles indicating a rhombic form, with inclusions from the main field, the main field being of two colors. The white bodies are \(Sn\ Sb\) \((360^\circ)\). According to Compbell’s diagram the main field must contain \(Sn\alpha\) \((210^\circ)\) and the eutectic of point \(p\) \((188^\circ)\).

Alloy No. 9. The temperature curve gave three points: \(340^\circ\), \(210^\circ\), and \(181^\circ\).

The character of the polished section is identical with the preceding one. The process is the same, only everything solidified at point \(q\).

Alloy No. 15. The temperature curve gave three points: \(320^\circ\), \(208^\circ\), and \(190^\circ\).

The character of the polished section is as follows: in the upper part of the section there are rhombi of Sn Sb; in the lower part these crystals have the structure of the Sn Sb+Pb net in the principal field. The order of cooling is the following: first Sn Sb separated out (320°), then along the line op—Sn Sb+Pb (208°), and everything solidified at p or near it (region fop).

Alloy No. 21. The temperature curve gave three points: 310°, 220°, and 189°.

The character of the polished section is the same as in No. 15, only the Sn Sb+Pb net is clearer (see Fig. 9, Table II).

Separation proceeded as follows: first Sn Sb (310°), then Sn Sb+Pb (220°), and everything solidified at p.

Alloy No. 4. In view of the fact that the temperature curve gave 5 points (the lower one 172°), and the polished section indicated the presence of a foreign element, we shall draw no conclusions concerning it.

Alloy No. 10. The temperature curve gave three points: 372°, 240°, and 181°.

The polished section has the following character: the greater part of the section is occupied by white bodies; in the dark field between these bodies a structure resembling the Sn Sb+Pb net is visible. Comparing the data of the polished section and of the temperature, we conclude that at 372° Sn Sb separated out; then along the line Op, Sn Sb+Pb separated out (240°), and everything solidified at point q.

Alloy No. 16. The temperature curve gave three points: 368°, 240°, and 181°.

The character of the polished section is the same as in No. 10, and everything said concerning it remains applicable here as well (see Fig. 9-a, Table II).

Alloy No. 22. The temperature curve gave three points: 340°, 240°, and 181°.

The rest is the same as in Nos. 10 and 16. On the ternary diagram these alloys should lie below line 10, i.e. in the region fop.

Alloy No. 5. The temperature curve gave four points: 485°, 401°, 247°, and a barely noticeable point at 187°. The character of the polished section is as follows: when etched with HNO₃, the white bodies are barely distinguishable by color; between the white bodies there is very little of the principal field, in which the Sn Sb+Pb net is visible (see Fig. 9 b, Table III); when etched with HCl+Fe₂Cl₆, the white bodies separated sharply.

Judging from the temperature and the character of the polished section, we arrive at the conclusion that up to 485° Sb separates out; then along the line fo, Sn Sb separates out (401°); further, along the line op, Sn Sb+Pb separates out (247°—here, apparently, the temperature was not determined accurately, since it should be less than 245°), and everything solidified at p or below.

Alloy No. 11. The temperature curve gave four points: 458°, 381°, 245° and 190°.

Alloy No. 17. The temperature curve gave four points: 430°, 369°, 249° and 181°.

Alloy No. 23. The temperature curve gave four points: 400°, 352°, 248° and 181°.

The character of the polished sections and the order of solidification are the same as for No. 5.

Alloy No. 6. The temperature curve gave three points: 585°, 343° and 241°. The character of the polished section is as follows: upon etching with HNO₃ the white bodies are faintly separated; there is little field; there is a network of Sn Sb+Pb. The order of cooling is as follows: first Sb (585°), then Sn Sb (343°) on the line fo and Sn Sb+Pb on the line op (343°) (see Fig. 9-c, Table III).

Alloy No. 12. The temperature curve gave three points: 560°, 310° and 241°. The order of cooling is the same as for No. 6. Both lie in the region Aof.

Alloy No. 18. The temperature curve gave two points: 539° and 248°. The polished section shows white elongated bodies and a field covered with a network. Comparing the data of the polished section with the temperature curve, we come to the conclusion that first Sb separated out (539°), then on the line or Sb+Pb separated out, and everything solidified, apparently, at O.

Alloy No. 24. The temperature curve gave two points: 490° and 248°. With regard to this alloy one may say the same as was said of the preceding one. Both lie in the region Aor.

In conclusion we may say that, in general, Loebe and Campell’s data concerning the ternary alloy Sn, Sb and Pb have been confirmed. If one assumes that the points f and d on the diagram were plotted accurately, then the latest investigations gave deviations on the lines fp and pd; but since there are no exact numerical data concerning the position of these lines, there is no need to speak of deviations.

Hardness tests of the alloys obtained were carried out in the mechanical laboratory, and, moreover, by the Brinell method (with a ball), for which it was necessary to give the specimens two parallel planes. When we began to give the specimens the required form, we noticed the impracticality of the form chosen by us (in the form of a cone). In order to obtain two parallel planes, it was necessary to machine the specimens on a lathe and a planing (shaping) machine. Owing to the high antimony content, some specimens on the planing machine...

crumbled on the lathe, and specimen No. 5 could not be turned at all. The pressure on the ball was taken as equal to 500 kilograms. The results of the test are as follows:

Table V-b.

	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=\dfrac{1}{17}$	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=\dfrac{1}{5}$	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=\dfrac{1}{2}$	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=1$	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=2$	$\dfrac{\mathrm{Sb}}{\mathrm{Sn}}=14$
Pb 10%	No. 1. $q=18{,}2$	No. 2. $q=21$	No. 3. $q=35$	No. 4. $q=38$	No. 5. $q=$	No. 6. $q=60$
Pb 20%	No. 7. $q=15{,}9$	No. 8. $q=20{,}1$	No. 9. $q=29{,}5$	No. 10. $q=44$	No. 11. $q=74$	No. 12. $q=54{,}5$
Pb 30%	No. 13. $q=15{,}19$	No. 14. $q=17{,}5$	No. 15. $q=27{,}8$	No. 16. $q=53$	No. 17. $q=66$	No. 18. $q=46$
Pb 40%	No. 19. $q=14{,}6$	No. 20. $q=12{,}6$	No. 21. $q=22{,}8$	No. 22. $q=42$	No. 23. $q=53$	No. 24. $q=34$

The table is arranged so that in all the alloys of each horizontal row the percentage content of lead is one and the same, while the ratio

$$ \frac{\mathrm{Sb}}{\mathrm{Sn}} $$

gradually increases; in the vertical rows, with one and the same ratio $\dfrac{\mathrm{Sb}}{\mathrm{Sn}}$, the percentage content of lead increases.

Even at a cursory glance at Table V-b one may draw the conclusion that, at one and the same percentage content of lead, the hardness increases with an increase in the ratio $\dfrac{\mathrm{Sb}}{\mathrm{Sn}}$, i.e., with an increase in the antimony content. This dependence, however, has a certain limit, since in group VI, despite the fact that the antimony content is greater than in group V, the hardness obtained was lower.

If we now turn to the vertical columns, we shall see that, at one and the same ratio $\dfrac{\mathrm{Sb}}{\mathrm{Sn}}$, with an increase in the percentage content of lead the hardness gradually decreases. The exception is group IV, where this dependence is violated.

In continuing the description of the work carried out in the metallographic laboratory of the institute on the investigation of babbitts, we shall begin with a description of the investigation of zinc babbitts, which in general have very limited application in practice, and then proceed to a description

of studies of tin, lead, and lead–tin babbitts, which in general have very considerable practical use.

Most zinc babbitts, called antifriction babbitts because of their low coefficient of friction, have the following chemical composition:

Zn	Sn	Cu	Sb	Pb
85	—	5	10	—
80	14.5	5.5	—	—
69	19	4	3	5

Noting that in the recommended compositions of zinc alloys the ratio

$$ \frac{\mathrm{Sn}}{\mathrm{Zn}} = \frac{14.5}{80} \quad \text{and} \quad \frac{19}{69}, $$

for the investigation there were first taken two pure alloys of zinc and tin, in which the above-mentioned ratio of tin to zinc was preserved; in addition, two more zinc alloys with 5 and 10% Sn were taken. Further investigations were conducted with the aim of determining the influence of antimony and copper on the alloy of zinc and tin; moreover, in preparing the test alloys Zn Sn and Zn Sn Cu, the weight ratio

$$ \frac{\mathrm{Sn}}{\mathrm{Zn}} $$

was kept constant. Since the maximum content of antimony in the zinc alloys cited above was 10%, the first alloy for investigation was composed in such a way that 10% of it was Sb, and 90% was divided between tin and zinc, according to their weight ratio

$$ \frac{14.5}{80}, $$

and thus an alloy of the following percentage content was obtained: Zn = 76.2%, Sn = 13.8%, and Sb = 10%. In the further preparation of babbitts from zinc and tin, the ratio

$$ \frac{\mathrm{Sn}}{\mathrm{Zn}} $$

was maintained in the latter, but the antimony content was successively 8%, 5.5%, and 3%. Consequently, the test alloys of zinc, tin, and antimony taken for the experiments were as follows:

Zn	Sn	Sb
76.2	13.8	10
77.9	14.1	8
80	14.5	5.5
82.1	14.9	3

The Zn Sn Cu alloy was prepared by replacing, in the last group of alloys, antimony with copper, but the highest percentage content

copper was to be taken as 5.5. Thus the last group of alloys taken from Zn, Sn, and Cu had the following composition:

Zn	Sn	Cu
80	14.5	5.5
82.1	14.9	3

The study of the babbitts taken consisted in determining the change in the position of the critical points, the hardness by the Brinell method, for which a ball of diameter \(10^{m}/m\) was used for indentation under a load of 500 klgr., and, finally, all the babbitts taken were examined metallographically.

The results of the observations obtained are presented in the following Table VI.

Table VI.

Zn.	Sn.	Sb.	Cu.	Hardness by Brinell.	Photograph of polished section at magnification 140 and etched with a 1% solution of \( \mathrm{NHO_3} \) in alcohol.
84.65	15.35	—	—	34.5	—
79.4	21.6	—	—	30	Fig. 10, pl. III.
90	10	—	—	40	” 11, ” III.
95	5	—	—	43	—
76.2	13.8	10	—	50	” 12, ” III.
77.9	14.1	8	—	48	” 13, ” IV.
80	14.5	5.5	—	43	” 14, ”
82.1	14.9	3	—	33	” 15, ”
80	14.5	—	5.5	52	—
82.1	14.9	—	3	51	” 16, ” IV.

Thus from the last table it is evident that the hardness of alloys of zinc and tin gradually falls with increasing tin content; the structure of zinc alloys with the indicated tin content, as the content of the latter gradually increases, passes from the structure of zinc with tin dissolved in it, through the eutectic of these two alloys close to the weight ratio 90 : 10 (Fig. 11), and further, in the fields of the latter eutectic, the number of separately precipitated tin particles increases with increasing content of the latter (Fig. 10).

For the following group of ternary alloys Zn Sn, Sb, an increase in hardness is noticeable with a gradual increase of antimony, while the struc-

the structure of these alloys is a eutectic of two metals, Zn and Sn, amid which are found nuclei of crystals—one must suppose, of compounds of tin and antimony⁷—and the number of crystals of the latter kind that separate out increases with increasing antimony content in the alloys (Figs. 14, 13, 12, and 15).

Finally, in the alloys of the last group, Zn Sn Cu, the increase in hardness with increasing tin content is noticeably weaker; this must be explained by the influence of the compounds of tin with copper that are formed, which, as was indicated earlier, possess considerable hardness. The structure of these last alloys is a general eutectic mass of tin and zinc, amid which are scattered particles of tin and of compounds of tin with copper (see Fig. 16).

All the above-mentioned zinc alloys were taken for testing in amounts of 200 g; they were prepared in a kryptol furnace, and the cast ingots were made in fireclay molds with cooling at room temperature. The burning-off of zinc during melting was reduced by a surface layer of ash and magnesia. The determination of the critical points for all the above-mentioned alloys was carried out with the aid of N. S. Kurnakov’s automatically recording apparatus; it was found that the melting temperature of the indicated Zn Sn Sb and Zn Sn Cu alloys increases with increasing antimony and copper content at a constant ratio of tin to zinc; the melting temperature of the Zn Sn Cu alloys increases especially noticeably with increasing copper content. An admixture of lead in ternary Zn Sn Cu and Zn Sn Sb alloys noticeably reduces the hardness of the alloys, but few experiments were made in this direction, since the indicated alloys are encountered very rarely in practice; in general it should be noted that zinc alloys as babbitts have only slight practical distribution, whereas tin and, in recent times, tin-lead babbitts have become more widely used.

The study of the latter group of alloys was begun with the group of compounds of tin with antimony; for this purpose alloys were prepared with a successive increase of the antimony content in tin, namely with 2.5%, 5%, 7.5%, 10, 15, 20, and then by fives up to 75% antimony in tin. The alloys were prepared in amounts of 40–50 g in a crucible in a kryptol furnace and, after melting, were rapidly cooled in a metallic mold. The results obtained in determining the—

the positions of the critical points and the general picture of the metallographic observations differ very little from the general picture of the solidification of two alloys of tin and antimony presented earlier in Fig. 2. Further, in the observations it was noticed that with an increase in the content of antimony in tin and with an increase in the cooling period, the size of the crystals of the compounds of tin and antimony increases; in addition, it was observed that the separated crystals of one and the same alloy grow noticeably during prolonged heating at temperatures from 248 to 316° C. The most distinctly expressed crystals were observed in tin alloys with 50% antimony, as a result of which it follows that these separated compounds of tin with antimony closely correspond to the compound of formula Sn Sb. In addition to the general investigations aimed at clarifying the question of the structure of alloys of tin and antimony, the latter were investigated for hardness by the Brinell method with the aid of a ball 10 m/m in diameter under a load of 500 klgr. The ball was pressed into the polished surface of the alloy and then the hardness of the latter was determined from the magnitude of the deformation. From the experiments it was found that the hardness of alloys of tin and antimony up to 10% of the latter increases rather weakly (from 17 to 30), then, with from 10 to 20% antimony, the hardness of the tin alloys increases very weakly (from 30 to 33), and, finally, after 20% antimony in tin alloys the hardness of the latter rises almost in a straight line, directly proportional to the content of antimony in tin, reaching 70 at 45% antimony. At 50% antimony in tin the alloy becomes completely brittle and does not withstand the ball test. The structure of tin alloys with an antimony content up to 10% is fine-grained, but with a further increase of antimony in the tin alloys, the size of the separated crystals increases.

In further investigations of babbitts in the metallographic laboratory of the Tomsk Technological Institute, the aim was pursued, as far as possible, of connecting these investigations with the practical application of babbitts. For the latter purpose, as the starting material were taken the babbitts used on the Siberian Railway. The composition of these babbitts was as follows:

1) Babbitt No. 1 for locomotives: 60% Sn, 21% Pb, 16% Sb, and 3% Cu.

2) Babbitt No. 2 for locomotives and passenger cars: 47% Sn, 38% Pb, 12% Sb and 3% Cu.

3) Babbitt No. 3 for freight cars: 23% Sn, 50% Pb, 24% Sb and 3% Cu.

The sum of all the indicated metals, according to the technical specifications of the Siberian Railway, must amount to not less than 99½% ...

For the investigations a series of alloys was prepared in which the ratio Sb : Sn and Cu : Sb : Sn, occurring in the babbitts of the Siberian Railway, was preserved, but the lead content was varied by three percent at a time up to 38%. In addition, tin alloys were prepared in which, at a constant ratio, the copper content was varied from 1 to 3%. Thus alloys Nos. 1–4 and 19–21 were composed with a constant Sb : Sn ratio, and in them only the percentage content of copper varied from 1 to 3%; then alloys Nos. 5–18, in which, at a constant Cu : Sb : Sn ratio, lead was added by 3% each time. For greater clarity, Table No. VII below gives the alloys composed from No. 1 to No. 21.

Table No. VII.

Nos. of photographs at magnification 100 : 1 and etched with a 1/2% solution of HNO₃ in water	Cu	Sb	Sn	Pb	Alloy No.
Fig. 17, plate IV	1,0	20,8	78,2	—	1
" 18	2,0	20,6	76,6	—	2
" 19, plate V	3,0	20,4	76,6	—	3
" 20 "	3,8	20,2	76,0	—	4
" 21 "	3,68	19,64	73,68	3,0	5
" 22 "	3,57	19,03	71,40	6,0	6
" 23 "	3,45	18,43	69,12	9,0	7
" 24 "	3,34	17,82	66,84	12,0	8
" 25, plate VI	3,23	17,21	64,56	15,0	9
" 26 "	3,11	16,61	62,28	18,0	10
" 27 "	3,0	16,0	60,0	21,0	11
" 28 "	2,89	15,39	57,72	24,0	12
" 29 "	2,77	14,78	55,45	27,0	13
" 30 "	2,66	14,18	53,16	30,0	14
" 31, plate VII	2,55	13,58	50,87	33,0	15
" 32 "	2,44	12,98	48,58	36,0	16
" 33 "	2,35	12,56	47,09	38,0	17
" 34 "	3,0	12,0	47,0	38,0	18
" 35 "	1,0	20,13	78,87	—	19
" 36 "	2,0	19,93	78,07	—	20
" 37, plate VIII	3,0	19,73	77,27	—	21

After all these preparations, observations were made of the melting temperatures of these alloys and of their microstructure, and then their hardness was determined with the aid of Brinell’s apparatus.

The determination of the melting temperatures was carried out by observing the cooling of the molten alloys. For this purpose a self-recording apparatus of Prof. Kurnakov was used, in which the readings of the pyrometer are automatically traced on photographic paper moving at a uniform speed. Thus, for each alloy a cooling curve was obtained. The experiment itself was performed as follows: an alloy in an amount of about 100 grams was placed in a crucible of refractory clay, inserted into a kryptol furnace, and heated to complete melting of the metal by means of an electric current. When all the metal had become liquid, the end of a Le Chatelier pyrometer was immersed in it. Since the thermocouple is acted upon by molten metals, especially antimony, which, according to Baikov’s indications⁸, forms with platinum a definite chemical compound, \(\mathrm{PtSb}_2\), and since this latter compound is obtained by direct contact of platinum with molten antimony and the reaction takes place with a tremendous evolution of heat, in working with antimony it is necessary to pay special attention to ensuring that the end of the pyrometer is completely protected from contact with the metal and its vapors. Usually, for this purpose, the thermocouple is placed in a quartz tube and the latter is immersed in the liquid alloy. In this way complete insulation is achieved between the pyrometer and the metal; at the same time the pyrometer retains its sensitivity and is protected from the action of the molten metals. The contacts of the ends of the thermocouple with the wires were placed in test tubes immersed in a mixture of melting ice. From the thermocouple, wires led to the pyrometer and to Prof. Kurnakov’s apparatus, which recorded the cooling curve.

After all the constituent elements of the alloy under investigation had passed into the liquid state, the kryptol furnace was switched off and a drum, driven in rotation by a clock mechanism, on which photosensitive paper had been fitted, was set in motion. The highest temperature to which the molten metal was heated ranged between \(750\)–\(790^\circ \mathrm{C}\). The pyrometer readings were recorded every \(100^\circ\) by switching off the small lamp in the lantern (“Zaychik”) for 10–15 sec. The cooling of the molten metal continued on average about \(1 \tfrac{1}{2}\) hours. The experiment ended when the pyrometer indicated that the temperature had fallen to \(180\)–\(150^\circ \mathrm{C}\).

The alloys were prepared before the experiment in such a way that first copper was melted in a fire-clay crucible with a small quantity of tin, in order to avoid oxidation; a small layer of powdered charcoal was poured on top; after this, tin was added in small portions, then lead, and finally antimony, while again the molten metal, to avoid oxidation, was covered with a layer of charcoal. The alloys prepared for the experiments contained no more than 3% copper. The results obtained by means of the cooling curves are presented in Table VIII. In this table, for each alloy all the pyrometer stoppages—the critical points—are given.

Table VIII.

Alloy No.	Stoppages 1	Stoppages 2	Stoppages 3	Stoppages 4	Notes
1	—	310	236	—
2	—	309	235	—
3	355	311	240	—
4	349	313	229	187
5	352	314	233	190
6	360	305	224	190
7	364	307	220	193
8	362	305	224	197
9	361	295	—	183
10	347	305	—	186
11	340	294	—	190
12	358	270	—	180
13	371	295	—	185
14	373	290	—	193
15	365	305	—	183
16	375	—	241	—	Stoppage No. 2 in alloy No. 16 is absent owing to a long interruption of the light beam.
17	370	298	—	193
18	377	293	228	—
19	—	317	241	—
20	—	311	243	—
21	326	315	235	—

If, from the position of the critical points of the alloys, one proceeds directly to consider the picture of the internal changes that is observed in the latter alloys in connection with the critical points, then the following must be noted: for alloys Nos. 1–3 the position of the 2nd critical point corresponds to the moment of separation of the compound of tin with antimony, which on the fields of the polished sections is noticeable in the form of separate crystals (see Figs. 17, 18, Table IV, and 19, Table V), while the third critical point corresponds to the solidification of the binding tin mass, containing up to 10% Sb. Further, the appearance of the higher 1st point in alloy No. 3 should be explained by the separation from the alloy of compounds of copper with tin. With a low copper content this point was not noticeable, but at 3% copper it appeared. On the fields of the polished sections of the indicated alloys the precipitates of the compounds of copper with tin gradually increase, and in the polished sections of alloy No. 3 (Fig. 19) these precipitates visibly cut through the crystals of the compounds of tin and antimony.

For all alloys, beginning with No. 5 through No. 18, there is a fourth arrest of the pyrometer at a temperature very close to 190°C, which must be explained by the formation in all the indicated alloys of a eutectic soft mass of 62% tin and 38% lead, with a melting temperature close to 182°C.^18) The second half of the soft cementing mass of babbitts, judging by the position of point No. 3 of the pyrometer arrest, for all babbitts containing tin and lead, represents a homogeneous mass of tin containing up to 10% antimony. The melting temperature of such a mass, as was indicated earlier (see Fig. 2), is very close to 243°C.

Thus from the preceding it is necessary to conclude that the soft cementing part of the babbitts under consideration consists, in alloys of tin not containing lead, of a tin mass containing up to 10% antimony in solution; in babbitts having tin and lead in their composition, the soft solidified part consists of a double mixture—of particles of tin with antimony of the preceding composition and of particles of the eutectic mass of tin and lead.

The arrest in the fall of temperature during cooling of the alloys after melting, conventionally designated by us in Table VIII as “2,” must be presumed to correspond to the separation from the molten alloy of crystalline compounds of tin with antimony or lead with antimony. It is possible to suppose that in alloys containing a considerable amount of tin and lead, both crystalline forms of compounds of the latter elements with antimony separate simultaneously.

^18) Guertler. Metallographie. S. 727.

Clearly expressed crystals of compounds of tin and lead with antimony, in the form of irregular rhombi, are distinctly visible in almost all polished sections of the alloys (see Fig. 17, Table IV—through 37, Table VIII).

The appearance of the highest arrest point in the fall of the thermometer temperature during cooling occurs in all those cases in which the copper content in the alloys is greater than 2%, and from this it follows that the point conventionally designated “I” corresponds, in the indicated alloys, to the separation of compounds of tin with copper; moreover, in the fields of the polished sections these compounds have the structure of weakly developed fir-trees or long rods cutting through the crystals of antimony compounds of tin and lead.

Further, all the above-mentioned alloys were tested for hardness by the Brinell method, the diameter of the ball used being 10 m/m and the load varying within two limits—200 and 500 kilograms. The diameters of the indentations obtained on the polished sections were measured with the aid of a microscope, and the hardness of the alloys was determined from a table (scale). Below is given Table IX of such measurements, the mean value from the two tests at 200 and 500 kilograms being taken for the hardness, according to which the curve was also constructed.

Table IX.

No. of alloy	Load 200 kg: d	Load 200 kg: Hardness	Load 500 kg: d	Load 500 kg: Hardness	Mean of two tests
1	3,3	23	4,4	31,2	27,1
2	2,6	37	4,2	34,5	35,75
3	2,7	34	4,1	36,0	35,0
4	2,6	37	4,1	36,0	36,5
5	2,65	36	4,2	34,5	35,25
6	2,7	34	4,3	32,6	33,3
7	2,75	33	4,25	33,6	33,3
8	2,8	32	4,3	32,6	32,3
9	2,9	30	4,45	30,4	30,2
10	3,0	28	4,5	29,7	28,85
11	3,15	25	4,65	27,8	26,4
12	3,1	26	4,8	25,9	26,0

Alloy No.	Load 200 kg	Load 200 kg	Load 500 kg	Load 500 kg	Mean of two experiments
	d	Hardness	d	Hardness
13	3.1	26	4.8	25.9	25.9
14	3.2	24	4.75	26.5	25.25
15	3.0	28	5.1	22.8	25.4
16	3.3	23	4.9	24.3	24.0
17	3.15	25	5.3	21.0	23.0
18	3.2	24	5.2	21.8	22.9
19	3.0	28	4.75	26.5	27.25
20	3.0	28	4.2	34.5	31.25
21	2.8	32	4.3	32.6	32.3

If these data are represented graphically, plotting along the X-axis the composition of the alloys, i.e., in our case the percentage content of lead in the alloy, and along the Y-axis the hardness, we obtain a curve (see Fig. 38). As is evident from these data, the hardness decreases up to 21% Pb, and from 21 to 33% remains almost unchanged; after this, a slow decrease again begins.

Fig. 38.

Further tests of babbitts of the lead-tin class were combined with practical tests of these babbitts under various service conditions on the Siberian Railway. In the summer of 1912, in the Krasnoyarsk workshops of the Siberian Railway, by order of the Chief of Traction of that railway, I. P. Arbuzov, experimental tests of babbitts of various compositions were organized. At my suggestion, two students of the Tomsk Tech—

of the Technological Institute, M. Mikhailov and A. Belyaev. The last two persons, at the end of the indicated practical exercises, presented to me a detailed report on the arrangement and results of the experimental investigations of babbitts in the Krasnoyarsk workshops of the Siberian Railway; subsequently, under my direction, they processed the samples of babbitts obtained by them in the metallographic laboratory of the Institute, which, in all, constituted their diploma work. I shall continue the further description of the arrangement of the experiments on the investigation of babbitts in the Krasnoyarsk workshops of the Siberian Railway according to the report presented to me by the two above-named students of the Institute; I must add that the said report agrees almost word for word with the published report of the Chief of the Krasnoyarsk workshops on the same question⁹.

For the experimental investigations babbitts of the following chemical composition were taken (see Table X).

Table X.

Name of babbitt	Sn	Sb	Cu	Pb	Specific gravity	Theoretical cost of 1 pood	Price by volume
No. 1 D . .	25	20	5	50	8.8	11 rub. 23 kop.	9 rub. 88 kop.
No. 1 N . .	30	16	4	50	—	—	—
No. 1 M . .	25	16	4	55	—	—	—
No. 1 K . .	20	16	4	60	9.15	9 rub. 71 kop.	8 rub. 88 kop.
No. 1 V . .	15	16	4	65	9.35	8 rub. 36 kop.	7 rub. 82 kop.
No. 3 K . .	—	17	—	83	10.13	4 rub. 14 kop.	4 rub. 19 kop.
No. 3 D . .	—	16.5	3	80.5	10.08	4 rub. 27 kop.	4 rub. 30 kop.
No. 3 E . .	—	16	4	80	10.10	4 rub. 31 kop.	4 rub. 35 kop.
No. 1 T . .	60	16	3	21	7.75	20 rub. 46 kop.	15 rub. 85 kop.
No. 2 T . .	47	12	3	38	8.32	16 rub. 02 kop.	13 rub. 32 kop.
No. 3 T . .	23	24	3	50	8.66	10 rub. 73 kop.	9 rub. 29 kop.

As regards wagon babbitt, during the experiments they endeavored: 1) to determine the suitability for actual service of No. 3 K and 2)

the effect of introducing into this same composition one or another quantity of Cu. The first, so to speak, touchstone for testing babbitts on the Siberian Railway was the hardness test. Since it was not possible in the Krasnoyarsk workshops to use the Brinell method, or any other generally accepted method, for determining hardness and to obtain results according to some definite scale, the following device was used to determine the relative hardness of the above-mentioned babbitts (see Fig. 39). The piece of babbitt to be tested, having two smoothly planed surfaces, is placed in the recess A of the apparatus constructed for this purpose.

Fig. 39.

The hardness of a given grade of babbitt is characterized by the diameter of the indentation obtained by pressing in a steel, conical (with an included cone angle at the apex of 60°), polished punch B. The load itself was produced with the aid of a simple lever press for testing springs, under the pressure roller of which was placed the head of the punch D, having a recess with the corresponding radius of curvature. The magnitude of the load in all tests remained constant and equal to one hundred and fifty poods. After several preliminary experiments, in which the duration of the load was varied within the limits from 15 seconds to 3 minutes, it was observed that after one minute of action of the punch there occurs a complete absence of further deformations that could have been taken into account with the measuring instruments available in the Krasnoyarsk workshops. Consequently, the duration of pressing was made equal to one minute in all experiments in testing the hardness of babbitt. The results and the mean values of the observations are given in Table XI below*).

“It is known that, in order to reduce friction and to avoid seizing, it is necessary that the contacting rubbing surfaces be made of materials as hard as possible. It would seem that the choice would be very simple; but, says Valuev, hardness alone is not enough, since with the slightest unevenness of the contacting rubbing surfaces, whether from poor fitting or from accidental causes, the pressure existing between them may be concentrated at several points, and then seizing of the surfaces may occur.

*) See the table on p. 52.

Table XI.

Name	d₁	d₂	d₃	d₄	mean d
Leykin No. 3 . .	9,8	9,8	9,8	10	9,85
No. 1 D . .	10,49	10,49	10,47	10,5	10,49
No. 3 T. Us. . .	10,6	10,4	10,5	11,6	10,77
No. 1 K . .	11,09	10,92	11,12	11,22	11,09
Leykin No. 1 . .	10,9	10,91	11,49	11,49	11,192
No. 1 V . .	11,39	11,20	11,20	11,5	11,32
No. 1 T. U. . .	11,42	11,42	11,39	11,89	11,53
No. 1 N . .	11,60	11,60	11,39	11,60	11,55
No. 1 M . .	11,60	11,40	11,52	11,80	11,58
Leykin No. 2 . .	→12,40	13,01	12,18	12,8	12,72
No. 2 T. U. . .	12,5	12,8	13	—	12,77
No. 3 E . .	13,3	13,2	13	13,4	13,3
No. 3 D . .	13,4	13,4	13,5	13,41	13,43
No. 3 K . .	14,2	14	13,8	13,67	13,92

To prevent this phenomenon, it is necessary that one of the rubbing surfaces belong to a plastic, viscous material capable, within certain limits, of undergoing changes in its shape without destruction. It has been observed that the hardness of a body is often accompanied by brittleness, i.e., by a low ability to resist impacts and shocks; in other words, it as it were possesses little strength. Thus alloys suitable for lining bearings must have sufficient hardness combined with viscosity and strength. The experiments undertaken to determine the strength and viscosity of the grades of babbitt under consideration, unfortunately, did not yield results worthy of confidence, owing to the imperfection of the instruments and machines available in the Krasnoyarsk workshops. These data, obtained by crushing bars of cross-section 1 sq. cm with the lateral surface of a cylinder of diameter 10 m/m, were too varied even for one and the same grade of babbitt, depending above all on the excessively primitive apparatus (a lever press), and then also on purely accidental causes that were difficult to eliminate: a somewhat less careful application of the load, the shaking of the forge floor under the blows of the steam hammer, etc.

As for placing trial babbitts on the bearings of locomotives, tenders, and cars, this was, properly speaking, to be the final and decisive test of the trial babbitt for determining the suitability of its use on the Siberian Railway.

The latter test is almost the only good test, apart from the hardness test. The first to be tested was babbitt No. 1 D.

In its physical properties (a detailed characterization is given below) it proved to be somewhat brittle, although tests in the axle-box and side-rod bearings of locomotive No. 1308 gave favorable results. Inspection after two trial runs (—96 versts) found the working surface of the side-rod bearing of the above-mentioned locomotive to be entirely sound. No damage to the lubricating channels, dents, galling, cracks, paraffin on the journal, etc., was observed.

Proceeding further along the path of a gradual reduction of the percentage content of tin in the composition of locomotive babbitt, the tests were concentrated on an alloy with a tin content of 20%—babbitt No. 1 K,—and in order to impart the greatest toughness, the contents of Cu and Sb were reduced from 5 and 20% to 4 and 16%. It must be noted that these percentage values of Cu and Sb, to facilitate subsequent observations, were adopted as constant in all compositions with one or another quantity of tin. In view of the fact that in the Krasnoyarsk main workshops almost exclusively major repairs of locomotives and cars were being carried out, the supervisors of the experiments found themselves in an extremely difficult position with regard to the possible productivity and increase in the number of experiments in producing babbitt of the desired composition. Consequently, the total number of experiments on locomotive babbitts carried out in the Krasnoyarsk main workshops during the summer of 1912 was far too insufficient to draw any definitive conclusion as to the advantages of one or another grade of the above-mentioned babbitts. The need for the speediest resolution of this question and for becoming acquainted with the conduct of investigations of lead-antimony babbitts previously carried out on all sections of the Siberian Railway prompted the supervisors of the experiments to apply with a request that investigations be conducted on the grades of babbitt they had produced and preliminarily tested in the Krasnoyarsk main workshops on the section of Krasnoyarsk Station. To this it must be added that locomotives repaired in the workshops, after two or three trial runs, are sent to

its own section and, consequently, escape further direct observation, whereas the locomotives of the Krasnoyarsk section of the traction service remained throughout under direct observation, with a very varied and ever-increasing mileage on the experimental babbitt. In mid-July 1912 permission was obtained, which made it possible to test babbitt No. 1 K more intensively and, moreover, under the usual actual conditions of its service on a locomotive. The total number of axle and drawbar bearings submitted for testing and poured with babbitt No. 1 K was 47. During the entire time of its service, both in the main workshops and on the section, no peculiarities or defects were observed. Upon inspection of some of the bearings, carried out by the directors of the experiments in the presence of the foremen of the locomotive-assembly shop, the working surfaces of the bearings were found to be sound; no running of the oil grooves, rolls, cracks, scoring of the journal, etc., was observed. To determine the degree of wear and durability it would be desirable to continue observations of the locomotives of the Krasnoyarsk section of the traction service further.

Babbitt No. 1 V, with a tin content of 15%, is the next step in the program of work for a gradual reduction in the percentage content of tin. Unfortunately, lack of time did not make it possible to carry out its tests fully in the scope and variety with which this was done for babbitt No. 1 K. Babbitt No. 1 V was fitted to 17 bearings.

With lead–antimony babbitts the matter stood somewhat differently. Parallel tests of the lead–antimony babbitts 3 K, 3 D, and 3 E gave quite favorable results. In all the rather numerous inspections of the bearings of wagons, tenders, and platforms, the directors of the experiments found the working surface to be quite sound. No rolls, cracks, chipping, scoring of the journal, or running of the oil grooves was observed. The total number of tests of lead–antimony babbitts was 285. It was impossible to note any substantial difference among the various grades of wagon babbitt, i.e., to take into account the influence of copper on the wear and durability of the babbitt with so insignificant a mileage on the experimental babbitt. The test alone on the stationary engine of the power station of the Krasnoyarsk main workshops would seem to speak in favor of the introduction of copper and of a certain superiority of babbitt No. 3 D over No. 3 K. As regards physical properties, the difference appears more clearly, whereas pure lead–antimony babbitt has a light-gray ...

in the color of the fracture; lead-antimony babbitt with an admixture of copper, with a darker fracture color, acquires a clearly expressed violet hue; hardness increases from the addition of copper (see in more detail below). To obtain the clearest possible idea of the technical properties of lead-antimony babbitts, as early as the beginning of July 1912 they were installed on the axle-box bearings of locomotives Nos. 1610 and 3163. Both locomotives endured the trial runs quite successfully; on the road the bearings did not heat up, but unfortunately the working surface was not inspected. This success opened a new path in the planned investigations; however, the low hardness, the extreme caution required in setting up experiments with pure lead-antimony babbitts on the line and, finally, lack of time prevented these tests from being developed to the desired extent. In all, the babbitt was installed on axle-box and connecting-rod bearings in a quantity of 285 pieces.

One of the important shortcomings of lead-antimony babbitts is their low hardness and heat capacity. Here it will not be superfluous to repeat a few words about the “spreading” properties of babbitt of one composition or another. The point is that when bearings are cast under the roller—and this occurs in application to locomotive axle-box bearings in general and to connecting-rod bearings in particular—on the line, the pasty state of the babbitt is of great importance. In this state the babbitt fills the bearing well and “spreads” well under the roller. The duration of the pasty state of babbitt is the greater, the greater the heat capacity possessed by its constituent parts. Taking, according to Valuev, the heat capacities¹⁰ Pb = 1, Sb = 1.7, Sn = 1.75, and Cu = 3, we may say that alloys with a high percentage content of Pb will have a short period of “spreading.” Below is Table XII of the heat capacities of all babbitts, from which it is seen that the heat capacity of lead-antimony babbitts is rather insignificant. Pure lead-antimony babbitt, on cooling, passes almost at once from the liquid state into the solid; the period of spreading of the pasty state is absent, which greatly complicates casting.

Table XII.

Grade of babbitt	No. 1 T. U.	No. 2 T. U.	No. 3 T. U.	No. 1 D.	No. 1 N.	No. 1 M.	No. 1 K.	No. 1 V	No. 3 K.	No. 3 D.	No. 3 E.	No. 3 M.
Heat capacity	1,62	1,496	1,4	1 4275	1,417	1,38	1,34	1,3025	1,139	1,1755	1,192	1,272

In addition to the above description of the experimental investigations of babbitts on the Siberian Railway, it is necessary to add a few words about the observations obtained concerning the qualities of the babbitts taken for testing.

Babbitt No. 1 K has a light silvery fracture and, with an insignificant tin content, possesses good technical qualities: in hardness, as is seen from the preceding Table XI, it is not inferior to the best grades of tin babbitts, while retaining sufficient toughness in service. Possessing good heat capacity, babbitt No. 1 K long preserves a dough-like structure, i.e. has a prolonged “smearing” period and fills the bearing under the journal very well. The latter property is especially valuable when pouring locomotive axlebox and connecting-rod bearings, especially if they are not to be bored. The poured surface comes out smooth, without shrinkage cavities, cracks, or blowholes.

Babbitt No. 1 B has a light-gray fracture and, with an insignificant tin content, less than in babbitt No. 1 K, possesses good properties. In hardness it differs little from tin babbitts, but is softer than babbitt No. 1 K. Its heat capacity is lower than that of the latter babbitt, but quite sufficient for the “smearing” period. Its toughness is moderate, but no chipping was observed when cutting the lubricating grooves.

Babbitt No. 3 K has a grayish fracture, belongs to the pure lead–antimony babbitts, possesses little hardness, but is distinguished by considerable toughness. Its heat capacity is small, but increases as the copper content in it increases. The “smearing” period is almost entirely absent, as a result of which pouring bearings under the journal is very difficult and cannot be performed without a soldering iron. When poured under a template it fills the bearing well and without defects.

Babbitt No. 3 D comes rather close to the preceding babbitt No. 3 K, but, in contrast to the latter, possesses somewhat greater heat capacity and hardness. The fracture of babbitt No. 3 D is gray, with a characteristic violet tint.

Babbitt No. 3 E, in contrast to the two preceding lead–antimony babbitts, has a larger quantity of copper, as a result of which it also possesses greater heat capacity.

As is seen from the foregoing, the principal aim of the tests of babbitts carried out on the Siberian Railway was to clarify the question of how far it is possible to replace expensive tin babbitts with more...

cheaper babbitts from the class of lead-tin-antimony alloys and from purely lead-antimony alloys. In addition, during the experiments described, an effort was made to elucidate the influence of an admixture of copper in lead-antimony babbitts.

The second half of the investigations of babbitts studied experimentally on the Siberian Railway was carried out in the metallographic laboratory of the Institute under my supervision, but with the active participation of the previously mentioned students (M. Mikhailov and A. Belyaev), who thus completed the second half of their diploma work. For a more thorough verification of the practical results obtained, all the babbitts tested in practice on the Siberian Railway were recast anew in the laboratory from pure metals and appear below under the, perhaps, not entirely convenient designation “chemically pure,” in distinction from samples of babbitts prepared in the Krasnoyarsk workshops of the Siberian Railway. The latter experimental observations, made in the metallographic laboratory of the Tomsk Technological Institute on the study of the above-mentioned babbitts (see Table X), may be divided into two groups: a) the study of the structure of the alloys and of their critical points, and b) the study of the physical properties of the babbitts—their hardness and specific gravity.

We know that any alloy in the liquid state represents a fairly homogeneous mixture of the metals composing it. This mixture may be simply a solution of one metal in another, or a solution of some chemical compound in the general mass of the alloy. Babbitts, too, are subject to this rule. When a liquid alloy begins to cool, then at a certain temperature, definite for each composition, one of the component pure metals or the corresponding chemical compound begins to separate out from the general mass of the liquid and solidifies in the form of separate crystals. The formation of crystals of this kind, accompanied by the liberation of a certain quantity of heat, slows the fall in temperature of the alloy and produces on the cooling curve a halt known as a “critical point.” Thereafter the separation of crystals of another metal or another chemical compound may begin, which again will cause a new halt, a new critical point. Finally, when the separation of all the metals and chemical compounds that were, as it were, in excess in the alloy has ended, there remains a still liquid mass, always identical in its composition for the given series of alloys. Upon solidifying, it forms, as it were,

a completely homogeneous body, so that even under a microscope it is impossible to distinguish the separate crystals of its constituent parts. This liquid homogeneous mass is called the eutectic mixture, and a greater or lesser amount of it in the entire mass of the alloy has a great influence on the general properties of the latter. The cooling curves, giving the relation between the fall of temperature and time, indicate the conditions and the moments of the beginning of the separation of crystals and of the formation of the eutectic mixture, and thereby make it possible, with the aid of the curves of critical points, to determine even the percentage composition of a given alloy, and conversely.

For binary alloys, which at the present time have been studied very thoroughly (Guertler, Borneman), this is quite correct and presents no difficulties; for ternary alloys, however, and still more for quaternary ones, the question becomes already more complicated. Polished sections may be of considerable use in this case, giving under the microscope a visual picture of the distribution of the forms and sizes of the crystals and the eutectic. The results obtained from the cooling curves may be represented by the following Table XIII.

Table XIII of critical points ²¹)

Babbit grades	Workshop samples I	Workshop samples II	Workshop samples III	Workshop samples IV	Samples from the Laboratory “chemically pure” I	Samples from the Laboratory “chemically pure” II	Samples from the Laboratory “chemically pure” III	Samples from the Laboratory “chemically pure” IV
No. 1 Tech. Cond. and Leikin	367°	295°	204° 202°	196°	356°	288°	196°	188°
No. 1 D	473°	306°	256°	194°	—	—	—	—
No. 1 K	482°	286°	256°	—	482°	276°	241°	—
No. 2 Tech. Cond.	432°	263°	213°	194°	432°	260°	216°	196°
No. 2 Leikin	402°	270°	211°	194°	—	—	—	—
No. 1 B	526°	276°	254°	—	497°	276°	247°	—
No. 3 Tech. Cond.	402°	315°	254°	—	406°	310°	256°	—
No. 3 Leikin	384°	315°	254°	—	—	—	—	—
No. 3 K	306°	256°	—	—	288°	256°	—	—
No. 3 D	530°	269°	256°	—	550°	280°	256°	—
No. 3 E	577°	270°	256°	—	—	—	—	—

²¹) The babbits of Leikin’s factory Nos. 1–3 have the same composition as the babbits of the same numbers of the Siberian Railway, according to the technical specifications (Nos. 1–3 Tech. Cond.).

The cooling curves of the alloys were obtained with the aid of the well-known apparatus of Prof. Kurnakov; the babbitt under investigation, in an amount of 100 gr., was melted in a kryptol electric furnace. Into a heated crucible of refractory clay with fireclay there was charged either a sample filing of babbitt cast in the Krasnoyarsk main workshops, or, in the corresponding order: copper with a small amount of tin and then, gradually, Sn, Pb, and Sb, in the case when the alloy was made up of pure metals in the Institute laboratory. Heating was carried out on the average to 700°–750° C; cooling was recorded by the apparatus down to 150° C, since it was assumed that below this temperature there are no critical points in the babbitt under investigation. The cooling time ranged from 1 h. 15 min. to 1 h. 25 min.

Returning to the table of critical points given above, upon closer examination one may draw the following conclusions:

A. The results of the investigation of specimens cast in the Krasnoyarsk main workshops and of “chemically pure” specimens give one and the same picture. The coincidence of the critical points for some babbitts is distinguished by great accuracy.

For example, for No. 2 of the technical specifications:

Workshop specimen	432°	263°	213°	194°
“chemically pure”	432°	262°	216°	196°
difference	0	−3	+3	+2

The greatest discrepancy for other grades can be explained by different melting conditions and by the resulting greater or lesser burning-out of one or another element of the given alloy.

B. Babbitts Nos. 1, 2, and 3 of the technical specifications of the Siberian Railway, and the corresponding Nos. of the Leikin works, have a common basic composition.

For example, for babbitts No. 1 tech. spec. and No. 1 of the Leikin works:

367°	295°	204°	196°
367°	295°	202°	196°
Difference 0	0	−2°	0

C. Tin babbitts with a tin content greater than 25% have four critical points.

D. A 25% tin content is the limit of existence of the fourth critical point (194–196° C).

At 25% Sn it is expressed very weakly—for example, for babbitt No. 1 D; and at 24% Sn—for babb. No. 2 tech. cond. and 2 of Leikin—it is not observed at all.

E. Babbitts with a tin content of less than 25% give only three critical points.

F. The upper critical point of lead babbitts is higher than that in tin babbitts, and rises still higher as the percentage content of lead increases.

Thus for lead babbitts,

\[ \begin{aligned} &\text{with a lead content of } 50\% && \text{critical point } 473^\circ \mathrm{C}.\\ &\text{” ” ” } 60\% && \text{” ” } 482^\circ \mathrm{C}.\\ &\text{” ” ” } 65\% && \text{” ” } 526^\circ \mathrm{C}. \end{aligned} \]

G. The lower critical point (254–256° C) for lead babbitts also appears at a higher temperature than that which corresponds to the eutectic mixture 87% Pb + 13% Sb.

Strictly speaking, the eutectic mixture 87% Pb + 13% Sb corresponds to a temperature of 248.5–244.8° C, and a certain elevation of it must, in all probability, be attributed to impurity of the constituent parts. This supposition is partly confirmed by the fact that for “chemically pure” babbitts prepared in the laboratory of the Institute, this point sometimes appears somewhat lower.

H. Lead-antimony babbitt No. 3 K is a binary alloy; the first critical point corresponds to the separation of crystals of pure antimony; the second—to the eutectic mixture 87% Pb + 13% Sb.

For the study of the structure of the babbitts, polished sections were prepared and photographed at a magnification of 100 times. The etchants used for the polished sections were: a 3% solution of nitric acid in water, bromine water, and HCl + Fl₂Cl₆. The general picture of the observed structural phenomena in the babbitts examined is as follows:

Babbitt No. 1 K (“chemically pure”), slow cooling—the edge of the casting (fig. 40, table VIII). The photograph represents the metallographic appearance of the edge of the lower part of a cone-shaped casting (100 gr.) obtained in the laboratory under slow cooling. The large crystals, which usually characterize slow cooling, are entirely absent here: they partly settled in the middle of the casting, and partly, under the influence of their lower specific gravity, floated upward. The main, fundamental, black mass is a mixture of two eutectic masses: 87% Pb + 13% Sb and 62% Sn + 38% Pb.

Babbitt No. 1 K (chemically pure) (fig. 41, table VIII). The same polished section, but the middle of the casting was chosen as the place for the photograph. Here we see characteristic white crystals of antimony compounds of very large size \((0.25 \times 0.2\ \mathrm{m}/\mathrm{m})\) and the principal dark soft mass of the composition indicated earlier. Between the crystals and, as it were, cutting through them, pass narrow, long rods of copper compounds. It is quite possible that these rods, scattered in all directions and joining together separate, harder grains of the antimony segregations, serve as a kind of framework in the softer principal mass and partly account for the considerable hardness of babbitt of this composition. In the next polished section these rods are expressed much more weakly.

Babbitt No. 1 K. (Krasnoyarsk workshops) (fig. 42, table VIII). The chief difference of this polished section, in comparison with the one just examined, consists in a considerable decrease in the size of each grain and of the copper rods, which, of course, follows from the conditions of more rapid cooling of this specimen, prepared in the Krasnoyarsk main workshops.

Babbitt No. 1 D (Krasn. workshops) (fig. 43, table VIII). This specimen has much in common with No. 1 K (Kr. worksh.), and only a considerable increase in the total number of white crystals of antimony and its compounds indicates an increase in the percentage content of the latter. The corresponding decrease in the eutectic indicates the greater hardness of this alloy and the increase in its brittleness, which was also observed in practice.

Babbitt No. 1 T. U. (Krasn. worksh.) (fig. 44, table VIII). Among the general, principal, dark-gray mass are scattered white rhombs of antimony and its compounds, and there are small, slightly grayish crystals in the fields of larger crystals of antimony compounds¹¹. The distribution of the crystals over the general surface of the polished section is fairly uniform. Copper compounds are observed in the form of separate inclusions both in the principal mass and among the crystals themselves.

Babbitt No. 1 T. U. (chemically pure) (fig. 45, table IX). It gives a vivid picture of the influence of slow cooling. The crystals of antimony compounds reach very large size. The copper compounds with tin have the appearance of narrow strokes, partly cutting through the crystals of antimony compounds, partly lying against the background of the principal mass.

Babbitt No. 1 V. (Red Workshop) (fig. 46, table IX). Amid the main dark-gray mass of the lead–antimony eutectic with an admixture of the tin-and-lead eutectic, white irregular rhombi of antimony segregations are scattered here and there, and, besides this, in very large quantity, very small crystals—evidently of the same antimony—and, finally, narrow rods of a milky color. When the polished section was etched with \( \mathrm{HCl} + \mathrm{Fe}_2\mathrm{Cl}_6 \), these milky rods took on a beautiful lilac color; consequently, we have here to do with copper compounds. These rods in places pierce the white crystals, and in places are grouped in the form of separate oases, surrounded and, in their turn, pierced by fine whitish zigzags and islands.

Babbitt No. 1 V. (Chem. pure) (fig. 47, table IX). The same picture as in the preceding polished section, but the size of the crystals is much greater (\(0.15 \times 0.175\ \mathrm{m}/\mathrm{m}\)), while their number is smaller.

Babbitt No. 2 T. U. (Red Workshop) (fig. 48, table IX). Here a particularly great abundance is noticeable of rods of copper segregations, scattered in all directions and reaching a very great length. The crystals of antimony segregations in places have a regularly expressed rhombic form and attain a size of \(0.1 \times 0.1\ \mathrm{m}/\mathrm{m}\).

Babbitt No. 2 T. U. (Chem. pure) (fig. 49 table IX). The dimensions of the crystals increased more than twofold (\(0.2 \times 0.275\ \mathrm{m}/\mathrm{m}\)); the rods of copper segregations disappeared; in their stead there appeared in large number, among the white grains and in the main mass, pinkish-violet inclusions.

Babbitt 3 K (Red Workshop) (fig. 50 table IX). This babbitt, being a binary alloy, is distinguished, as are also the subsequent ternary alloys of babbitts No. 3 D and 3 E, by great homogeneity. The main gray mass is a eutectic mixture (\(84\%\,\mathrm{Pb} + 13\%\,\mathrm{Sb}\)), while the white rhombi represent crystals of almost pure antimony, which was present in excess in the alloy before the formation of the eutectic.

Babbitt No. 3 D. (Red Workshop) (fig. 51 table X). The main black mass of the eutectic mixture \(87\%\,\mathrm{Pb} + 13\%\,\mathrm{Sb}\), with small white crystals of antimony and very thin, small compounds of copper of a pinkish-lilac color spread everywhere in large quantity, with a comparatively uniform distribution of all the constituent parts, indicate high technical qualities of the alloy as regards viscosity, with slight hardness. Practice fully justifies this supposition.

Babbitt No. 3 D. (Chem. pure) (fig. 52 table X). The cooling conditions in the present case also proved to have a very unfavorable influence. The crystals of antimony compounds became much larger in their dimensions. The segregation

copper in the form of pinkish-lilac inclusions is distributed rather abundantly not only among the white crystals of antimony, but also throughout the entire mass of the eutectic mixture.

Babbitt No. 3 E. (Krasn. workshops) (fig. 53, plate X). The principal dark-gray mass, as in the preceding lead-antimony babbitts, is a eutectic mixture of lead with antimony. Scattered within it are white crystals of antimony, of comparatively small size. There are considerably more copper segregations than in Babbitt No. 3 D.; they are all very small and are scattered in whole groups, both throughout the entire principal mass and among individual white antimony crystals. Some grains contain as many as 20 such small crystals.

Babbitt No. 3 T. U. (Krasn. workshops) (fig. 54, plate X). This polished section presents the following: there are many crystals of antimonial segregations, but little of the soft mass binding them, which characterizes the great hardness and low toughness of this alloy. The hardness of the soft lead mass of the alloy has been increased by the admixture of tin to it.

Babbitt No. 3 T. U. (Chem. pure) (fig. 55, plate X). On the polished section, large, long rods of copper compounds of very considerable dimensions are visible. The crystals of antimony compounds in the principal soft, black mass characterize the conditions of slow cooling of this alloy.

In conclusion, it remains only for us to emphasize the circumstance that slow cooling promotes the growth of individual crystals.

It was only necessary to examine the photographic images in order to be convinced of the correctness of the conclusion reached. All specimens of “chemically pure” babbitts obtained under slow cooling have crystals of considerably greater size than the specimens cast in the Krasnoyarsk workshops, where the cooling rate was considerably higher.

Thus, for example, in two mutually perpendicular measurements the crystals are:

Babbitt No. 1 V. Chemically pure: \(0.15 \times 0.18\ \mathrm{m/m}\).
» Workshops: \(0.09 \times 0.08\ \mathrm{m/m}\).

Babbitt No. 2 T. U. Chemically pure: \(0.25 \times 0.275\ \mathrm{m/m}\).
» Workshops: \(0.2 \times 0.1\ \mathrm{m/m}\).

With this we shall for the time being finish the general consideration of babbitts from the metallographic point of view and proceed to the determination of hardness and specific gravity.

All antifriction alloys, depending on their immediate purpose, must possess some degree of hardness. Under normal conditions a good babbitt consists, as is known from the foregoing, of a eutectic, soft mass, in which the harder crystals of pure metals and their chemical compounds, which separated out earlier, are distributed comparatively uniformly. These hard grains take up the load falling on the working surface of the bearing and transmit it (more or less uniformly) to the whole basic mass of the babbitt. The number and size of these crystals, on the one hand, and the external casting conditions, on the other, exert a great influence on the change in hardness of one or another grade of babbitt. Let us take, for example, a binary alloy corresponding to the eutectic mixture \(87\%\,\mathrm{Pb} + 13\%\,\mathrm{Sb}\) (see Fig. 1), and suppose that its hardness proves too low for us. Considering the curve of critical points of this pair, we at once see the necessity, in order to increase the hardness, of taking some excess of antimony, i.e. of increasing its percentage content and thereby promoting the separation of more hard crystals of pure antimony.

However, in this respect one must be extremely cautious. An excessive increase in the percentage content of antimony deprives the babbitt of one of its basic properties—toughness—and makes it completely unsuitable for actual service.

Let us now turn to the consideration of the influence of the external conditions of cooling. It must not be forgotten that the crystals appearing during the period of cooling of the alloy usually have a specific gravity different from that of the rest of the still-liquid mass, as a result of which they either rise to the surface or sink to the bottom. Such a phenomenon, as is known, is called “liquation” and appears especially sharply during the slow cooling of large masses. However, “at the present time,” says Wittorf (“Theory of Alloys as Applied to Metallic Systems”), “it may be regarded as fully established that, even with comparatively small quantities of metallic alloys (20–25 gr.), the structure of the layers adjacent to the walls of cooling is far from the same as in the inner regions of the casting.” The same thing was also observed during experiments with the slow cooling of babbitt cast in the laboratory in a quantity of 100 gr. Meanwhile it is clear that the more numerous the crystals and the smaller they are, the more homogeneous the total mass will be; and, conversely, the fewer the crystals and the larger they are, the more heterogeneous.

Until now only the cooling of babbitt has been discussed; however, the process of smelting itself hardly deserves less attention. An irrationally conducted smelting can ruin all calculations, and the product obtained will possess entirely different physical properties.

All these circumstances make it extremely difficult to obtain any absolute figures and compel serious attention to the choice of material for a given test. Thus, for example, when preparing polished sections during the experiments, an effort was made to follow the experiments of Neuskok and Neville, who, in studying bronze alloys, cut out samples in such a way that the side intended for preparation of the polished section was normal to the cooling surface and began at the latter. Unfortunately, it was not possible to carry out this condition completely, since the samples prepared earlier in the Krasnoyarsk main workshops were sent to the laboratory already sawn and in a considerably reduced form.

The first hardness tests were carried out in the summer of 1912 in the Krasnoyarsk main workshops of the Siberian Railway. The tests were then repeated again in the mechanical laboratory of the Tomsk Technological Institute on an Amsler machine. The load was 500 kg; the duration of its action was 3 minutes. The results obtained are presented in Table XIV.

Table XIV of hardness.

Sort of babbitt	\(d_1\)	\(d_2\)	\(d\) avg.	Surface	Hardness
No. 3 Leikin	4,30	4,30	4,30	15,2650	32,60
No. 1 D.	4,63	4,55	4,60	17,6087	28,40
No. 1 V.	4,60	4,80	4,70	18,4286	27,20
No. 1 K.	4,75	4,75	4,75	18,8527	26,50
No. 1 Leikin	5,10	5,10	5,10	21,9629	22,80
No. 1 T. U.	5,10	5,20	5,15	22,4357	22,30
No. 3 T. U.	5,25	5,25	5,25	23,3939	21,50
No. 2 Leikin	5,55	5,65	5,60	26,9392	18,60
No. 3 E.	5,60	5,60	5,60	26,9392	18,60
No. 2 T. U.	5,95	5,95	5,95	30,8316	16,20
No. 3 D.	6,15	6,05	6,10	32,6018	15,30
No. 3 K.	6,45	6,35	6,40	36,3828	13,80

Examination of the present table permits the following conclusions to be drawn:

A. Hardness increases as the percentage content of antimony increases, almost independently of the mutual ratio of the weight quantities of the remaining components (at least within the limits of the experiments cited above).

B. With one and the same percentage content of antimony, replacement of tin by lead causes a certain increase in hardness.

Tab. No. 1 T. U.—Sn—60%; Pb—21%; hard. 22.3.
” No. 1 K. —Sn—20%; Pb—60%; ” 26.5.

C. Pure lead–antimony babbitt is the softest of the babbitts investigated.

The addition to it of copper within the limits of 3–4% increases the hardness.

In the first part of the above-cited work an attempt was made to indicate the advantage obtained by replacing tin babbitts with lead ones, taking into account also the influence of specific gravity. To simplify the calculation of specific gravity, it was assumed that all the metals \(M_1, M_2, M_3\), etc., entering into the composition of the given babbitt do not combine chemically and do not form homogeneous solutions. Such an assumption is, of course, not correct, and the data obtained can be regarded only as a first approximation. The fact is that as soon as metals begin to dissolve in one another, or as soon as crystals of some chemical compound begin to separate out, the rectilinear dependence for specific gravity becomes curvilinear. In the metallographic laboratory the practical determination of specific gravity was carried out by double weighing of large and, in order to avoid adhesion of air bubbles, smooth masses of babbitt. We shall not dwell on the actual arrangement of the experiment, considering it too simple and generally known, but shall give only the final results. The babbitts are arranged in order of increasing hardness*).

The values of specific gravity are given at a water temperature of 20° C and atmospheric pressure \(p = 770\) mm. In addition, an attempt was made to determine the specific gravity by weighing babbitt filings in alcohol. However, the volatility of alcohol on the one hand, and the large quantity of air carried along by the filings on the other, greatly hinder both the performance of the experiments themselves and the obtaining of results deserving confidence.

*) For Table XV of specific gravity, see p. 67.

Table No. XV of specific gravity.

Babbitt grade	Workshop samples. $g_1$	Workshop samples. $g_2$	Workshop samples. $f=\dfrac{g_1}{g_2}$	“Chemically pure” samples $g_1$	“Chemically pure” samples $g_2$	“Chemically pure” samples $f=\dfrac{g_1}{g_2}$
No. 1 Leikin	454,350	60,000	7,570	—	—	—
No. 1 Tech. Cond.	590,155	76,310	7,734	50,600	6,350	7,990
No. 2 Leikin	832,200	100,150	8,310	—	—	—
No. 3 T. Cond.	694,600	82,600	8,410	66,130	8,010	8,258
No. 2 T. Cond.	706,500	83,770	8,434	52,330	5,950	8,795
No. 3 Leikin	1130,001	131,200	8,613	—	—	—
No. 1 D.	631,300	72,600	8,696	—	—	—
No. 1 V.	1051,000	110,880	9,070	31,310	3,330	9,402
No. 1 K.	1308,850	140,980	9,290	49,350	5,470	9,000
No. 3 E.	716,870	71,520	10,023	—	—	—
No. 3 K.	760,980	75,700	10,051	61,550	6,120	10,057
No. 3 D.	1039,640	102,100	10,182	63,270	6,240	10,140

In general, it must be noted that the specific gravity for any alloy is a relative quantity and may vary depending on 1) the specific gravity of its constituent parts and 2) the conditions of the casting itself, i.e. on the greater or lesser burning out of one or another metal. From the table given above only one conclusion can be drawn: that lead babbitts have a considerably higher specific gravity than tin babbitts.

In concluding the general discussion of babbitts, we consider it not superfluous to cite the highly interesting experiments of the director of the Kolomna Works, Mr. Manuilov, who communicated the following about them to Mr. Valuev:

“I,” he says, “investigated a very large number of grades, but only two grades of babbitt, those most often used in factories, are of real interest, namely: the more expensive tin babbitt, consisting of 11.11% Sb, 5.55% Cu, 83.34% Sn, and the lead babbitt, the cheapest, consisting of 85% Pb and 15% Sb. Both grades were tested under different loads and different speeds. Pressure was produced by means of levers with

with loads of various weights, pressing the bearing against a shaft 4″ in diameter rotating on rollers. From the mass of figures obtained, it turned out that, under normal operation, the two grades show no differences. The tin babbitt withstood a load of 114 klg. per sq. cm. at a speed of 3.5 mtr., the temperature of the bearing being 85–88°; the lead babbitt did the same. With an increase in pressure the temperature rose; at 130° C the load reached 133 klg. per cm.². The lead babbitt, at the same speed of 3.5 mtr. and a load of 133 klg., kept its temperature up to 120° C. The beginning of the seizing of the grooves in both alloys occurs shortly after the temperature rises above 135–140° C. A striking result is obtained: namely, the most expensive and the cheapest babbitt give almost the same result, which, in terms of temperature, is rather in favor of the lead alloy of this composition. One misunderstanding, caused by the value of the metals, compels people to seek the best alloy among tin babbitts. A psychological trait of people, making them consider the expensive to be good, has proved unsuitable for metallic alloys as well. Nature does not know human valuation and judges in its own way.

The experimental material obtained, on the one hand, and laboratory investigations, on the other, make it possible to draw the following practical conclusions:

I. The transition from tin babbitts to lead babbitts is entirely possible and is an urgent need of every factory enterprise.

II. Babbitt No. 1 K with 60% Pb is not inferior in its technical properties to tin babbitt with the same percentage content of tin (60% Sn).

III. Lead–red-lead babbitt can replace Nos. 2 and 3 of the Technical Specifications of the Siberian Railway.

IV. There are grounds to think of the possibility of a further reduction in the percentage content of tin, which, of course, requires further tests in this field.

In conclusion of the article on babbitts it is necessary to say a few words about the manufacture of babbitts. From the preceding it is not difficult to note that babbitts in the liquid state are very strongly subject to liquation, whence it follows directly that, when making babbitts by the usual method (melting), they should not be greatly overheated; in the liquid state they must be thoroughly stirred; and in making ingots it is necessary to cast the latter in metallic molds. Without going into the details of the manufacture of babb—

...by melting, it is necessary to note, if only briefly, those attempts that have been made recently to produce babbitts by pressing. Small experiments in the latter direction were also carried out in the metallographic laboratory of the Tomsk Technological Institute.

The question of obtaining alloys by the cold method, i.e., by compressing a powder-like mixture in a thick-walled vessel, belongs among the new and little-investigated ones in the technical literature.

It was first touched upon by Spring (1878–1882). This investigator carried out experiments at very high pressure: thus, for example, he obtained Wood’s alloy (8 parts Pb, 4 parts Sn, 15 parts Bi, 3 parts Cd) by compression at 50 tn/cm². The pressed alloy did not differ in any way from an alloy prepared by the usual method; it had the same specific gravity, color, hardness, and melted at 70° C.

In his experiments Spring observed that, with an increase in the melting temperature of the metals being compressed, the necessary pressure increases. To obtain, for example, brass, the pressure must be incomparably higher than 50 tn/cm², and a homogeneous alloy is obtained only after pressing 5–6 times, each time turning the metals being mixed into powder and thoroughly mixing them; the compressed alloy is somewhat darker than that obtained by the usual method.

For lack of literary data, it cannot be said whether any investigations on this question were conducted after Spring’s work. Only recently has interest in pressed alloys grown, and, as we shall see, attempts have appeared to apply them in technology.

In Metallurgie (1910, p. 430), R. Friedrich’s experiments are described. The latter took an alloy in a dough-like state during the solidification period, when, owing to the separation of its components, it becomes plastic, added to it some refractory metal, mixed, compressed, and cooled it. A mechanical mixture was obtained. To an alloy of 90% Sn and 10% Cu he added glass; to an alloy of 98% Pb and 2% Sb, at a temperature of about 250°, iron filings. The metal being added, in order to avoid cooling the mixture, was preheated to the same temperature as the mixture. The alloys obtained cannot be compared with melted alloys with respect to hardness, malleability, and thermal and electrical conductivity, since they are not homogeneous. However, by adding quartz, corundum, or carborundum, one can achieve great hardness; by adding metal oxides, one can reduce thermal and electrical conductivity.

In the same journal (1910, p. 553) there is a description of the experiments of W. Guertler. The latter attempted to obtain a cheap bearing metal; the soft material for the alloy was pure Pb or alloys of Sb, Zn, Sn with it, while the hard material, instead of the rather expensive Sn or Zn, was iron or cast iron. He obtained the Pb–Fe alloy by heating the powdered mixture to \(250—300^\circ\) (until the lead became plastic) and pressing it. A hard and compact mass was obtained, quite suitable for bearings. The author finds his method of pressing better than the above-described method of R. Friedrich, since it is possible to achieve a closer mixing of the constituent parts of the alloy. In contrast to Friedrich, Guertler regards the mixtures obtained as true alloys, only less homogeneous.

Finally, Masing investigated this question. He decided to test Spring, since during the period when the latter’s experiments were carried out (1878—1882) the structure of alloys had been little studied, and Spring’s conclusions about the identity of alloys that had been pressed and those obtained by the ordinary method gave rise to doubts.

On the basis of the general laws of physical chemistry, Masing assumed that in systems which, upon solidification, separate out pure components, the fused and compressed products would not differ essentially; but if both components of a binary system (Masing studied precisely this system) are capable of forming a compound or mixed crystals, then it is doubtful that the pressed and fused products would come out identical, since pressure does not accelerate a chemical reaction. Experiments with binary systems fully confirmed this proposition. Masing compressed a powdered mixture in a small cylinder (\(d = 14^{m}/m\)) with a pestle and brought the pressure up to 5000 atm., i.e., up to 50 klg. per \(m/m^2\). He did not obtain higher pressures, since the pestle became deformed. By means of a thermocouple he obtained heating and cooling curves.

As it turned out, metals which, on being fused, give neither compounds nor strong solutions do not enter into reaction under pressure either. When both components are capable of forming compounds, but do not form mixed crystals in considerable concentration, then under pressure a certain amount of compound is also formed. Owing to the low pressure, Masing obtained unsatisfactory sections, and he proved the formation of the compound only from the heating curve, namely: 1) by a halt of the thermometer during the crystallization of both eutectic mixtures formed by the compound and the two components, and 2) by large rises in the curve of temperature fall, corresponding to partial fusion of the components.

If both metals give mixed crystals, then, on compression, a certain quantity of them is also obtained. Finally, metals that give mixed crystals and compounds give these products also under pressure. When the compressed product is heated, diffusion takes place, and its structure becomes more homogeneous¹¹.

As we see, on the question of the identity of pressed alloys and alloys obtained by the ordinary method, only Spring has spoken out quite affirmatively. The other investigators either deny it (R. Friedrich), or else recognize it only in part, as, for example, Guertler and Masing. But Spring used colossal pressures, a minimum of 500 kg./m², and in some cases considerably greater ones. Unfortunately, we do not have at our disposal a detailed report on his works, while the brief notes in textbooks do not give a description of the experiments themselves; as a result, the question remains a mystery as to how the material of his apparatus could withstand such enormous stresses.

The apparatus constructed for the experiments (quite similar to Masing’s apparatus) and made in the Institute workshops from Böhler steel, certainly could not withstand them. Moreover, for the work only the Amsler press of the Institute’s mechanical laboratory was available, giving a maximum load of 60 tn.; to avoid damage to the machine, the laboratory recommends not bringing the load up to the maximum. The load was applied up to 55 tn. In order to obtain on this press a pressure of 50 tn. per cm.², it would have been necessary to make the cylinder diameter approximately 35 m/m². Hence it is clear that it was necessary to investigate the effect on the powdery mixture of comparatively small pressures.

The original apparatus for carrying out the experiments was constructed in the following form (Fig. 56). The diameter of the cylinder, and hence

Fig. 56.

and the pestle, \(=25^{m}/m\). Assuming the greatest load to be 55 tn., we obtain the compressive stress on the punch

\[ Kg=\frac{55000}{490}=112\ \mathrm{klgr.\ m}/m^{2}. \]

The apparatus was made of steel; the initially unhardened pestle became deformed and, after operation, was removed with difficulty, as a result of which the pestle had to be hardened and ground.

The required metals in pieces were filed with a locksmith’s file and sieved through a sieve; the grain size was \(0.3^{m}/m\); the thin but long shavings that passed through the sieve were removed. The quantity of mixture was prepared according to the capacity of the small cylinder (60–65 gr.). The powder-like mixture was thoroughly stirred, poured into the apparatus, and compressed under a press. The pressures were obtained at intervals of 5–10 tn. from 30 to 55 tn. Upon completion of pressing the load was maintained, and the mixture remained under pressure for 5–10 min. The alloy thus prepared was sawn in two: one half went to the preparation of a polished section, the other was again filed with a saw and compressed a second time. For this repeated pressing it was necessary to make another apparatus, since the capacity of the previous one was large. The new apparatus was a copy of the first; its diameter \(=15^{m}/m\). The greatest load was taken as 30 tn. The compressive stress on the pestle was

\[ Kg.=\frac{30000}{177}=170\ \mathrm{klgr.\ m}/m^{2}. \]

The first small cylinder became deformed under a load of 28 tn. It was necessary to make a new one and harden it.

The alloys obtained were sawn in two: one part went to the preparation of a polished section, the other to experiments for determining the melting temperature of the pressed alloys.

As a result of the experiments, only soft alloys—babbitts of the following composition—were obtained by pressure.

Group 1: Lead 77.25%,
Antimony 14%,
Tin 8.75%.
\(\left.\begin{array}{l} \\ \\ \end{array}\right\}\) Babbitts of the Russo-Baltic Works in Riga.

Sample No. 1, pressure 15 tn.—34 kg per \(m/m^{2}\).
" " 2, " 20 " 45 " " " Fig. 57, table X.
" " 3, " 30 " 61 " " "
" " 5, " 39 " 80 " " "
" " \(1/2\), " 28 " 159 " " "

Sample	Pressure	Load	Reference
Sample No. 1	15 tn.	34 kg per \(m/m^2\)
Sample No. \(2/2\)	15 tn.	85 kg per \(m/m^2\)
Sample No. \(3/2\)	20 tn.	114 kg per \(m/m^2\)
Sample No. \(5/2\)	25 tn.	142 kg per \(m/m^2\)	Fig. 58, plate XI.
Sample No. 4	30 tn.	170 kg per \(m/m^2\)	Fig. 59, plate XI.

Group 2.

Component	Content	Alloy
Tin	90%	British metal.
Antimony	8%	British metal.
Copper	2%	British metal.

Sample	Pressure	Load	Reference
Sample No. 6	25 tn.	56 kg \(m/m^2\)
Sample No. 7	30 tn.	61 kg \(m/m^2\)
Sample No. 8	35 tn.	79 kg \(m/m^2\)	Fig. 60, plate XI.
Sample No. \(6/2\)	15 tn.	85 kg \(m/m^2\)
Sample No. \(7/2\)	20 tn.	114 kg \(m/m^2\)	Fig. 61, plate XI.
Sample No. \(8/2\)	25 tn.	142 kg \(m/m^2\)
Sample No. 9	30 tn.	170 kg \(m/m^2\)

The sample was melted in a furnace (fig. 62, plate XI).

Group 3.

Component	Content	Alloy
Lead	60%	Alloy for bearings of railway cars.
Tin	20%	Alloy for bearings of railway cars.
Antimony	20%	Alloy for bearings of railway cars.

Sample	Pressure	Load	Reference
Sample No. 10	30 tn.	61 kg \(m/m^2\)
Sample No. 11	35 tn.	79 kg \(m/m^2\)
Sample No. 12	40 tn.	81.6 kg \(m/m^2\)	Fig. 63, plate XI.
Sample No. \(10/2\)	15 tn.	85 kg \(m/m^2\)
Sample No. \(11/2\)	20 tn.	114 kg \(m/m^2\)	Fig. 64, plate XII.
Sample No. \(12/2\)	25 tn.	142 kg \(m/m^2\)
Sample No. 13	30 tn.	170 kg \(m/m^2\)

The sample was melted in a furnace (fig. 65, plate XII).

Group IV.

Component	Content	Alloy
Lead	84%	Alloy for bearings of slowly rotating shafts.
Antimony	16%	Alloy for bearings of slowly rotating shafts.

Sample	Pressure	Load	Reference
Sample No. 14	30 tn.	61 kg \(m/m^2\)
Sample No. 15	35 tn.	79 kg \(m/m^2\)
Sample No. 16	45 tn.	92 kg \(m/m^2\)	Fig. 66, plate XII.
Sample No. \(14/2\)	15 tn.	85 kg \(m/m^2\)
Sample No. \(15/2\)	20 tn.	114 kg \(m/m^2\)	Fig. 67, plate XII.
Sample No. \(16/2\)	25 tn.	142 kg \(m/m^2\)
Sample No. 17	30 tn.	170 kg \(m/m^2\)	Fig. 68, plate XII.

Specimen melted in a furnace.—Fig. 69, Table XII.

Group	Component	Content	Alloy
V	Lead	78%	“Magnolia” alloy
V	Tin	16%	“Magnolia” alloy
V	Antimony	6%	“Magnolia” alloy

Specimen	Pressure	Result
No. 18	40 tn.	81.6 kg. m/m²
No. 19	45 tn.	92 kg. m/m²
No. 20	55 tn.	112 kg. m/m²
No. 18/2	15 tn.	85 kg. m/m²
No. 19/2	20 tn.	114 kg. m/m²
No. 20/2	25 tn.	142 kg. m/m²

Group	Component	Content	Alloy
VI	Zinc	69%	Alloy for bearings, “Babbitt”
VI	Tin	19%	Alloy for bearings, “Babbitt”
VI	Copper	4%	Alloy for bearings, “Babbitt”
VI	Antimony	3%	Alloy for bearings, “Babbitt”
VI	Lead	5%	Alloy for bearings, “Babbitt”

Specimen	Pressure	Result
No. 21	30 tn.	61 kg. m/m²
No. 22	50 tn.	102 kg. m/m²
No. 21/2	20 tn.	114 kg. m/m²
No. 22/2	25 tn.	142 kg. m/m²

Specimens bearing the index 1/2 were the result of repeated pressing; for this purpose the previously pressed material from the specimen whose number is indicated by the numerator was melted down again, poured through a sieve, and pressed again. It was not possible to carry out more than two repeated pressings, owing to the small capacity of the mold used for pressing, while an increase in the capacity of the mold was not permitted by the dimensions of the press employed. Polished sections were prepared from all the specimens and photographed, but below in the text (for economy) only a few photographs are given. Figs. 57, 58, and 62 were taken at a magnification of 300:1, and the rest at 140:1.

The pressed specimens of the babbitt of group I present on the polished sections a clear picture of a mechanical mixture (see Fig. 57, Table X; 58 and 59, Table XI), and only in the last stages of pressing are there, in the field of the polished section, slight signs of the separation of crystalline forms of antimony compounds.

A similar picture of a mechanical mixture is also noticeable in the pressed specimens of British metal (see Figs. 60 and 61, Table XI), and the distinct clarity of the indicated picture stands out especially prominently when the two above-mentioned photographs are compared with Fig. 62, Table XI,

giving a general picture of the structural constitution of the same British metal, but produced by melting. A quite distinct picture of a mechanical mixture of the constituent components was also found for babbitts of group III (see Fig. 63, Table XI, and Fig. 64, Table XII), which, in the solid state after melting (see Fig. 65, Table XII), have a structure noticeably different from the two presented above (Figs. 63 and 64). The general picture of a mechanical mixture was found in all pressed specimens of the remaining groups of soft alloys, and even lead-antimony babbitt, whose structure in the ordinary form is represented by Fig. 69, Table XII, in pressed specimens presents a vivid picture of a mechanical mixture (see Figs. 66, 67, and 68, Table XII).

In conclusion to the consideration of the general picture of the structural constitution of pressed alloys, it is necessary to add that the polished sections of specimens after repeated pressing had greater uniformity in their structure and in places revealed signs of the separation of crystals. For comparison of the homogeneity of alloys obtained in the usual way by melting with alloys obtained by pressing, small experiments were carried out to compare the melting temperature of the alloy.

For this purpose, a small cylinder of each alloy was taken, placed by weight in the crucible of a kryptol furnace; at the same time the test cylinder of alloy, together with the copper wires on which it had been suspended in the crucible, was connected into a common circuit with an electric bell. At the moment the cylinder of the observed alloy melted, the circuit was broken, as a result of which the bell ceased to emit sounds. In the crucible of the furnace, together with the test cylinder, a pyrometer was placed, which thus gave temperature readings at every moment of the experiment. As a result of the latter experiments the following was found: melted babbitt of group I had a melting temperature of 252°C; while the melting temperature of its constituent elements is as follows: 232°C for tin, 316° for lead, and 630.5° for antimony. A cylinder from pressed specimen 5/2 of the same babbitt, placed in the circuit of an electric bell under the conditions of the experiment described above, showed signs of melting at 237°C, for at this temperature the bell began to ring faintly and ceased entirely at 275°C. The temperature 237°C is very close to the previously indicated melting temperature of tin, and hence it should be assumed that at 237°C tin melted out of the pressed cylinder under observation, while there remained a skeleton of pressed lead and antimony; the remaining

the skeleton in the cylinder conducted current very poorly, as a result of which the bell worked very weakly; and then, under the action of the high temperature, the skeleton oxidized and melted at \(275^\circ\) C, thus breaking the common circuit of the bell. The remains of the oxidized cylinder were found at the bottom of the crucible.

Heating of cylinders made from pressed specimens, under conditions entirely identical with the experiment described above, was carried out for specimens of the remaining groups of alloys tested by pressing. The general results of the observations obtained fully resemble the previously described experimental picture for the babbitt of group I. Only small deviations were observed in the temperature at which melting of the test cylinder began.

Thus, the latter experiments undoubtedly prove that pressed alloys differ greatly from the same alloys obtained by melting, and constitute a mechanical mixture of their constituent components, which, when the pressed specimen is heated, melt independently of one another, beginning with the most fusible. Further, in the latter experiments with pressed cylinders of one group of alloy no common melting temperature was found, which indicates the complete absence of a eutectic in such alloys. But the general picture indicated earlier of structural changes in successively pressed alloys of one and the same composition undoubtedly reveals an increase in the homogeneity of the pressed alloy with increasing pressure; moreover, the homogeneity of the alloy increases when, as material for pressing, a crushed specimen of the same alloy made under the press is used.

Thus, summarizing the general conclusions of the few latter experiments on the manufacture of alloys (babbitts) by pressing, it must be noted that it is possible, by pressing, to obtain babbitts fairly close in their properties to babbitts obtained by the usual method of melting; for this it is necessary to increase the pressing pressure considerably, and then to carry out repeated pressings with preliminary thorough crushing of the material from the preceding experiment.

Tomsk,
July 28, 1913.

Noted Misprints

The editors ask that, when reading the article, the appended list of errors be kept in mind; through a misunderstanding during printing, they were omitted at the beginning of the article in an especially large number.

Page	Line	Printed	Should read
1	2 from top	inostroitelnyi	nostroitelnyi
—	9 from top	okhlazh-	okhlazhde-
—	10 from top	denii	nii
2	15 from top	abstsis	abstsisst
3	18 from top	Goutermam'a	Goutermann'a
—	20 from top	zatverdyvaniia	zatverdevaniia
4	12 from top	odelannykh	sdelannykh
—	14 from top	Goutermann'a	Goutermann'a
—	12 from bottom	dok-	do-
—	11 from bottom	lada	klada
—	— from bottom	Guertler	Guertler
—	1 from bottom	uslovno obez-	uslovnoe obo-
5	1 from top	pecheno	znachenie
—	3 from top	nashol	nashel
—	16 from bottom	turma	surma
—	15 from bottom	posledo-	posledo-
—	9 from bottom	nab-	na-
—	8 from bottom	liudat	bliudat
6	4 from top	diag-	dia-
—	5 from top	ramm	gramm.
7	7 from top	pogruzhennyia	pogruzhennye
—	12 from top	priz-	pri-
—	13 from top	naiutsia	znaiutsia
8	5 from top	po dannym Monkenmeyer'a i Arnemann'e'a	, po dannym Monkenmeyer'a i Arnemann'a,
—	8 from top	Arnemann'a	Arnemann'a
—	10 from top	dos-	dostig-
—	11 from top	tignet	net
—	16 from bottom	soosvetstvuiushchei	sootvetstvuiushchei
—	4 from bottom	protsen-	protsent-
—	3 from bottom	tnoe	noe
9	12 from top	pinka	tsinka

Page	Line	Printed	Should be
11	9	“, Krista-	Kristal-
12	7 from bottom	Arnemanu’a	Arnemann’a
13	5 from bottom	zat-	zatver-
14	4 from bottom	verděvaniia	děvaniia
—	20 from bottom	nogo	nago
—	15 from bottom	etoi	eto
—	11 from bottom	nez-	nezna-
—	10 from bottom	nachitelnaia	chitelnaia
—	6 from bottom	kristalliches-	kristalliche-
—	5 from bottom	kikh	skikh
15	6 from bottom	op-	opre-
—	5 from bottom	reděleny	děleny
16	22 from top	koeffitsiėtta	koeffitsienta
—	8 from bottom	nabliudeniia	nabliudenie
17	6 from top	os-	ostal’-
—	7 from top	tal’nykh	nykh
—	10 from top	inache govoria	inache govoria,
—	18 from top	Baluevym	Baluevym
19	Table IV	tsena po s”emu	tsena po ob”emu
20	1 from top	zamětit’	zaměnit’
— table V	1 from bottom	Aziattkaia	Aziatskaia
21	4 from bottom	izozrazhennye	izobrazhennye
22	2 from top	zat-	za-
—	3 from top	verděvaniia	tverděvaniia
—	2 from bottom	zatverděvanii	zatverděvanii
23	6 from top	uproshcheniia, predpolozhit’	uproshcheniia predpolozhit’
—	21 from bottom	svinets poka	svinets, poka
24	3 from top	prib-	pri-
—	4 from top	lizitsia	blizitsia
—	6 from top	(Sn α) poka	(Sn α), poka
—	12 from top	zatverděvaniia	zatverděvanie
26	8 from bottom	Gesellschaft	Gesellschaft
27	14 from bottom	i chtoby	i, chtoby
28	3 from top	plavnosti	plavkosti
—	16 from bottom	cons.	const.
29	5 from top	splav v”	splavov”
—	22 from bottom	zastěvaniia	zastyvaniia
30	19 from top	sdinitsa	svintsa
—	13 from bottom	siur’ma	sur’ma

Page	Line	From	Printed	Should read
30	10	ditto	syur'ma	sur'ma
31	14	ditto	Sn Sb,	Sn Sb
33	15	ditto	Geselschaft	Gesellschaft
36	4	ditto	trp	tri
—	1	from bottom	tenperatur	temperatur
37	15	from top	temtseratur	temperatur
—	ditto	ditto	(see fig. 9-a, table II)	(see fig. 9-a, table III)
38	21	from bottom	prodolgovztyia	prodolgovatyia
—	10	ditto	il'etsya	imetsya
—	7	ditto	i-ta	I-ta
40	5	from top	nkh	ikh
43	7	from bottom	parovozov	parovozov:
47	14	from top	rostut	rastut
—	22	ditto	tsementriruyushchei	tsementiruyushchei
—	25	ditto	sur'ma	sur'my
48	10	ditto	dlinnnkh	dlinnykh
51	1	from bottom	poverkhnostei.	poverkhnostei
52	16	ditto	tela,	tela
61	9	from top	korkasom	karkasom
63	14	ditto	po	no
69	9	from bottom	Cn	Cu
71	20	from top	izb'ezhanii	izb'ezhanie
75	19	ditto	nebol'shiya	nebol'shie
76	18	ditto	legkoplavnago	legkoplavkago

Table I.

Fig. 7. 100 : 1.

Fig. 7-a. 100 : 1.

Fig. 7-b. 100 : 1.

Fig. 7-c. 100 : 1.

Fig. 7-d. 100 : 1.

Fig. 7-e. 100 : 1.

Table II.

Fig. 8-a. 100 : 1.

Fig. 8-b. 100 : 1.

Fig. 8-c. 100 : 1.

Fig. 8-d. 100 : 1.

Table III.

Fig. 9-a. 100 : 1.

Fig. 9-b. 100 : 1.

Fig. 9-c. 100 : 1.

Fig. 10. 100 : 1.

Plate IV.

Fig. 13. 100 : 1.

Fig. 14. 100 : 1.

Fig. 15. 100 : 1.

Fig. 16. 100 : 1.

Fig. 17. 100 : 1;

Fig. 18. 100 : 1.

On Babbitts. T. I. Tikhonov.

Table V.

Fig. 19. 100 : 1.

Fig. 20. 100 : 1.

Fig. 21. 100 : 1.

Fig. 22. 100 : 1.

Fig. 23. 100 : 1.

Fig. 24. 100 : 1.

Table VI.

Fig. 25. 100 : 1.

Fig. 26. 100 : 1.

Fig. 27. 100 : 1.

Fig. 28. 100 : 1.

Fig. 29. 100 : 1.

Fig. 30. 100 : 1.

Table VII.

Fig. 31. 100 : 1.
Fig. 32. 100 : 1.

Fig. 33. 100 : 1.
Fig. 34. 100 : 1.

Fig. 35. 100 : 1.
Fig. 36. 100 : 1.

On Babbitts. T. I. Tikhonov.

Table VIII.

Fig. 37. 100 : 1.

Fig. 40. 100 : 1.

Fig. 41. 100 : 1.

Fig. 42. 100 : 1.

Fig. 43. 100 : 1.

Fig. 44. 100 : 1.

Table IX.

Fig. 45. 100 : 1.

Fig. 46. 100 : 1.

Fig. 47. 100 : 1.

Fig. 48. 100 : 1.

Table X.

Fig. 51. 100 : 1.

Fig. 52. 100 : 1.

Fig. 53. 100 : 1.

Fig. 54. 100 : 1.

Fig. 55. 100 : 1.

Fig. 57. 300 : 1.

On Babbitts. T. I. Tikhonov.

Table XI.

Fig. 58. 300 : 1.

Fig. 59. 140 : 1.

Fig. 60. 140 : 1.

Fig. 61. 140 : 1.

Table XII.

Fig. 64. 140 : 1.

Fig. 65. 140 : 1.

Fig. 66. 140 : 1.

Fig. 67. 140 : 1.

Fig. 68. 140 : 1.

Fig. 69. 140 : 1.

1. Dr. W. Guertler, Metallographie. S. 793. ↩

2. Guetler, Metallographie. S. 340. ↩

3. A more detailed solidification curve for tin–lead alloys may be found in Guertler, Metallographie, p. 722. ↩

4. Proceedings of the XXVIII consultative congress of engineers of rolling stock and traction in 1911, p. 265. ↩

5. Metallurgie, 1911, 8, 7 and 33. ↩

6. Metallurgie. 1912, S. 422. ↩

7. Separations of compounds of zinc with antimony, in all probability, are absent because in the alloys there is an excess of tin content over antimony content. ↩

8. Journal of the Russian Physico-Chemical Society, Vol. 32, No. 4. ↩

9. Report of the Chief of the Krasnoyarsk workshops, engineer M. S. Gorodetsky, on the investigation of babbitts carried out by the Krasnoyarsk workshops. Tomsk, 1912. ↩

10. Relative to Pb = 1. ↩

11. Descriptions of the works of Friedrich, Guertler, and Masing have been placed in various issues of the “Proceedings of the Russian Metallurgical Society” for 1910. ↩↩