Search Documents

By E. J. KENNEDY

METHODICAL AND EFFECTIVE: thus may be characterized the fall meeting of the Iron and Steel and Institute of Metals Divisions at the Hotel Statler, Buffalo, N. Y., on Oct. 4 and 5. Approximately 200 registered. Interest centered 1111 the technical papers which were keenly followed, with substantial discussion: During the entire week of the National Metal Congress (Oct. 3 to 'i), there were also other sessions of interest' to metallurgists, directed by the America11 Society for Steel Treating, American Welding Society, Production Division of the Society of Automotive Engineers, the Iron arid Steel and Machine Shop Practice Divisions of the American Society of Mechanical En,' the Wire Association and the American Drop Forging Institute. The National Metals Exposition at the 174th Regiment Armory was well attended.

Jan 1, 1932

By T. A. Wright

THE-Institute of Metals Division opened on Tuesday afternoon with Wheeler P. Davey as chairman and G. E. Edmunds as vice-chairman. Four papers were on the program, two being of a fundamental character. A paper on "Internal Surface in Minerals and Metals," by R. S. Dean and J. Roster, was presented by the latter; C. H. Gottschalk discussed coercive forces in powders. In answer to a question by the chairman. The author stated that he had not investigated [he temperature coefficient of resistance. A paper on "Recrystallization and Grain Growth in Cold Worked Polycrystalline Metals," by L. W. Eastwood, Arthur E. Bousu, 'and C. T. Eddy, was presented by Mr. Eastwood. A. great many quantitative data were given, with little reference to the theory. Theoretical considerations were discussed by Dr. Gottschalk, and C. A. Edwards of the University College of Wales, contributed very interesting comment on the relationship of size and critical strain. Dr. Edwards discussed the necessity of considering strain with time and temperature.

Jan 1, 1935

By A. J. Lena, M. F. Hawkes

Changes in hardness, tensile properties, microstructure, electrical resistance, and X-ray diffraction effects indicate that lattice strains are necessary for the embrittlement of ferritic stainless steels when heated for relatively short times at 475°C (885°F). It is suggested that 475°C (885°F) embrittlement is due to the accelerated formation of an intermediate stage in the formation of s under the influence of these strains. FERRITIC stainless steels (low carbon alloys of iron with more than 15 pct Cr) are subject to two forms of embrittlement when heated in the temperature range of 375° to 750°C. The embrittlement which occurs after long time heating between 565" and 750°C is well understood; it is caused by the precipitation of the hard, brittle s phase. Sigma is an intermetallic compound of approximate equi-atomic composition with an extended range of formation in Fe-Cr alloys. The maximum temperature at which this form of embrittlement can occur is dependent upon chromium content; and is approximately 620°C for a 17 pct Cr steel and 730°C for a 27 pct Cr steel. The other form of embrittlement occurs after relatively short heating periods in the range of 375" to 565°C; in the higher chromium steels, hours may be sufficient as compared to months for s embrittlement. This phenomenon is not at all well understood and several controversial theories have been proposed. The rate and intensity of embrittlement increase with increasing chromium content but the maximum rate occurs at 475°C re-gardless of chromium content. As a result of this, the phenomenon has been termed 475°C (885°F) embrittlement. The effect of 475°C embrittlement on the properties of ferritic stainless steels has been thoroughly reviewed by Heger.1 The embrittlement causes a pronounced decrease in room temperature impact strength and ductility, a large increase in hardness and tensile strength, and a decrease in electrical resistivity and corrosion resistance. Microstructural changes accompanying embrittlement are minor and difficult to interpret with a general grain darkening, appearance of a lamellar precipitate, grain boundary widening, and precipitation along ferrite veins having been reported at various times. With the exception of reported line broadening, X-ray diffraction studies by conventional Debye analysis of solid samples have been of little value. BY making use of electron diffraction methods, Fisher, Dulis, and Car-roll' have recently shown the existence of a chromi-um-rich, body-centered cubic phase in 27 pct Cr steels which had been aged at 482°C (900°F) for as long as four years. Two types of theories have been advanced to account for the embrittlement. The first of these requires the precipitation of a phase not inherent in the Fe-Cr system with various investigators suggesting a carbide,3 nitride,3 phosphide,4 or oxide." Theories of this type have difficulty accounting for the influence of alloying elements on the embrittlement and for the facts that a minimum chromium content is necessary for embrittlement and the intensity of embrittlement increases with increasing chromium content. The second type of theory that has been proposed relates 475°C embrittlement to s phase formation which is inherent in the Fe-Cr system. An assumption of this kind can adequately explain the influence of alloying elements, for they exert an effect on 475°C embrittlement similar to that on s phase for-mation as can be seen in Table I. The minimum chromium content is essentially the same for both phenomena and it has been shown12,13 that s is a stable phase in the embrittling temperature range. In addition, it has been reported14,15 that pure alloys embrittle to the same extent as commercial type alloys. There are, however, several factors which have prevented complete acceptance of a s phase theory. Foremost of these is that the embrittlement can be removed by reheating for short time periods above 600°C, which in the higher chromium steels is within the stable s region. No s has ever been observed after one of these curing treatments, nor has any s been found as a result of embrittlement at 475°C. In addition, the simple precipitation of s cannot explain the time-temperature relationships for reactions between 350°and 750°C. This behavior is shown schematically in Fig. 1. Newell 16 and Ried-rich and Loib4 have shown that 475°C embrittlement follows a C-type curve as illustrated, while Short-

Jan 1, 1955

By P. Shahinian, M. R. Achter

In a comparison of the creep -rupture properties of nickel in air arid in vacuum there is a reversal in relative strengths with variations in stress. At low stresses the properties are better in air while at high stresses they are better in vacuum. In the early stages of creep in a high stress test the specimen in air nzay have the lower creep rate but the one in vacuum is creeping at a lower rate in the later stages. A mechanism, involving two competing processes in terms of oxidation strengthening and weakening by surface-energy reduction, is proposed to explain these reversals. It had generally been assumed that, relative to their properties in air, the high-temperature strengths of metals would be increased in an inert atmosphere as a result of the elimination of the deteriorating effects of oxidation. However, several recent investigations have shown that the creep strength may be greater in air than in nonoxidizing gases. Pickus and Parker1 found that the creep rate of nickel was lower in air than in hydrogen or nitrogen. Internal oxidation was mentioned by Shepard and Schalliol as a possible cause for the longer creep life of Hastelloy C in air than in vacuum or helium. Sweetland and Parker, who observed lower creep rates for aluminum and copper in air than in helium, discuss the possible applicability of the dislocation-barrier mechanism to explain their results. Surface oxide layers may interfere with the escape of dislocations through the surface or with their generation there. In addition to showing the strengthening in air of low-alloy and stainless steel, and nickel and cobalt-base alloys, shahinian4 presented notch-effect data which indicate the role of stress concentrations; alloys of relatively low ductility were more sensitive to atmosphere strengthening when notched specimens were used. Although Bleakney's copper wires had a lower reduction of area in air than in vacuum, it is difficult to interpret his results in terms of strength changes since the stresses employed were not measured. In contrast to these data showing the strengthening in air, a number of investigations, most at relatively low temperatures, have demonstrated that the environment may cause a decrease in strength. These results are generally explained in terms of the effect of environment on surface energy. In a review of the subject, Benedicks and Harden use the concept of atomic bonding to explain how the adsorption of impurities would decrease the energy of a clean surface. A decrease in surface energy would facilitate the propagation of cracks or extension of the surface. Use of this mechanism is made by sato7 to explain the weakening of steel specimens bent in various liquids, by Potek and Shcheglakovs for the reduction of fracture strength of steel in molten metal, and by Benedicks and Harden for the reduction of strength of a variety of materials in a number of liquids. A similar explanation might be used for the results of Forestier and Clauss. The fracture stress of fine wires pulled in a variety of gases, including air, is lower than in vacuum. They attributed their results to the "tendency" for the gases to condense on the metal surface. That the strength of metals is affected by the type of gas in which the test is conducted has been demonstrated convincingly; the mechanism of the effect, however, is still uncertain. This investigation represents one in a series to explore the influence of variations in alloy composition and test conditions on the atmosphere effect in the expectation that the results would afford an insight into the processes taking place. To determine the effect of specimen composition, the present investigation was conducted

Jan 1, 1960

By N. Pessall, A. D. McQuillan

The phase relalionsllips in an isothernzal section of the Ti-Zr-H system at 750°Chave been investigated by measurement of hydrogen equilibrium pressures. The resulting diagram indicates that the hydride phase has an anomalously wide range of homogencity extending from 66.6 to 30 at. pct H in alloys containing about 15 al. pet Zr. Measurement of the heat of solution of hydrogen as a function of alloy composition in ß-phase alloys indicates no anomalous behavior in the region of 15 at. pet Zr, and suggests that the extensive range of homogeneity in the hydride phase at about this composition must he attributed to an enhanced stability of the, hydride phase. THE manner in which titanium and zirconium react with hydrogen and the constitutional diagrams of the two metal-hydrogen systems are qualitatively very similar,'-6 as can be seen from Fig. 1 which shows the constitutional diagrams of the two systems. Quantitatively, however, there are some differences. For example, the exothermic heat of solution of hydrogen in ß titanium is smaller than that in ß zirconium, and consequently the equilibrium hydrogen pressure of a Ti-H alloy is higher than a Zr-H alloy at the same temperature and hydrogen concentration. Other differences manifest themselves in the hydride phases which exist at high hydrogen concentrations. In the Ti-H system, above room temperature, there exists only a single hydride phase in which the metal atoms are arranged on a fee lattice, but in the zirconium system there are two hydrides, and whereas the hydride of lower hydrogen concentration has a fee structure, the higher hydride has a face-centered tetragonal structure. The tetragonal hydride retains its stability up to temperatures in excess of 450oC.7 From the foregoing, it would be expected that the constitutional diagram of the ternary Ti-Zr-H system would have a very simple form, especially as titanium and zirconium are completely miscible in both the a and ß forms.8 The present work, which consists of a study of an isothermal section of the ternary system, was carried out in order to check the assumption frequently made that the electronic structure of the two metals and the mechanism by which hydrogen is held in solution in the metals is the same for each. Although this proved largely to be true, certain inter esting and unexpected features of the ternary system emerged which must be explained before we can really say that the reactions of titanium and zirconium with hydrogen are fully understood. EXPERIMENTAL PROCEDURE The most convenient method of studying hydrogen-metal systems is by the measurement of the hydrogen pressure in equilibrium with the hydrogen dissolved in the metal as a function of hydrogen concentration and temperature. In binary systems, the results are easily interpreted in terms of alloy constitution, since two-phase regions can be recognized by the fact that the hydrogen pressure at constant temperature does not vary with the hydrogen concentration in the alloy when it is in the two-phase condition. This criterion does not apply in ternary systems, but there is usually a marked change of slope in the isothermal pressure-composition curves when a phase boundary is crossed, and with care, and a knowledge of the limiting binary systems, the phase structure of the alloys can usually be deduced. In the

Jan 1, 1962

By P. Pietrokowsky

lntermediate phases in the binary metal alloy systems Ti-Zn, Ti-Hg, Zr-Zn, Zr-Cd, and Zr-Hg have been investigated by X-ray powder diffraction methods. Gamma-TiBHg and Zr3Hg have a beta tungsten structure; TiHg and ZrHg are analogous to ordered AuCu (L l0 type); delta-TixHg and ZrHg, are isomorphous with ordered AuCu,, (L l2 type); ZrZnz is face-centered cubic (C 15 type); TiZna crystallizes in a C 14 type structure; and Zrll-, occurs in cubic and tetragonal modifications in which a random distribution of atoms exists. Additional information for the intermediate phase TiZn, is presented. THE physical metallurgy of titanium and zir-conium when alloyed with elements in subgroup 11-B of the periodic table has received comparatively little attention in the literature. Laves and Wallbaum' reported the existence of several intermediate phases in the system Ti-Zn. They identified TiZn as a CsCl (B 2) structure and described TiZn, as an ordered Cu,Au (L 1, type) crystal structure. Gebhardt' investigated very limited zinc-rich regions in the systems Ti-Zn and Zr-Zn. Phase relationships in these partial constitution diagrams were deduced from the results of thermal analysis and metallography, X-ray results being inconclusive. The first zinc-rich intermediate phase in either system was not identified. More recently, Anderson, Boyle, and Ramsey" have had occasion to refer to a partial Ti-Zn phase diagram (unpublished data) covering the region 0 to 10 atomic pct Ti, wherein two intermediate phases corresponding approximately to TiZn,, and TiZn,, were indicated. The metals used in this investigation and typical analyses as determined by the suppliers are given in Table I. The iodide-process zirconium which was used was not hafnium free. Experimental Methods Alloy Preparation: The relative ease with which titanium and zirconium combine with other elements places extensive restrictions on the methods by which alloying can be achieved. The direct current electric arc melting procedure was found to be unsatisfactory for preparing these binary alloys and for this reason solid state diffusion methods were utilized. The iodide process titanium and zirconium rods were filed by hand. Fines were subjected to a magnetic separation and classified as to size as either finer than 200 mesh or coarser than 200 but finer than 80 mesh. Powder mixtures of various binary combinations of titanium and zirconium with zinc and cadmium were compacted in a 12 mm diameter die and then placed in Vycor or quartz vials which were evacuated and sealed. The mercury alloys were prepared in a similar manner; the proper weight of liquid mercury being placed in a vial with minus 200 mesh titanium or zirconium. These ampoules were then heat treated at various temperatures and times depending on the characteristics of the particular alloy system. Upon completion of the heat treatment the ampoules were cooled rapidly and the specimens weighed. The loss in weight varied from 1 to 3 pct (except as noted in the text) and for this reason the calculated compositions were accepted as a fair measure of the actual alloy compositions. X-Ray Dif raction Techniques: Powder diffraction experiments were conducted at room temperature with a 143.2 mm Debye-Scherrer X-ray camera in which the film was placed in the Straumanis arrangement. When copper radiation was used, a nickel filter was employed, and for the cobalt target a filter of iron oxide was used. In both cases, the /i? filter was placed between the X-ray target and the powder specimen which was mounted on a glass fiber. The lack of information based on optical metallography placed reliance on the results obtained from the X-ray experiments for determination of

Jan 1, 1955

By J. M. Roberts, D. M. Barnett

The activation energy required to break a pinned dislocation line away from its condensed atmosphere of impurity atoms is calculated as a .function of applied stress, without neglecting line tension. Reasows are presented for not assuming the effect of line tension to be negligible when considering the problem of freeing a small segment of dislocation from an impurity atmosphere. The dislocation is considered to he pinned at each atomic size along its length and the pinning points are assumzed immobile. The development is a refinement of the origznal cottrell-bilby1 theory. The solution is compared to recent approximations developed by cottrell3 and Haasen5 in which line tension con- IN 1949 Cottrell and bilby1 considered the problem of determining the force required to pull a dislocation away from its condensed atmosphere of impurity atoms. They also examined the effect of thermal fluctuations on this force. The dislocation line was pictured as bowing out in a triangular loop under the action of an applied stress, and the activation energy necessary to free the dislocation from its atmosphere was calculated as a function of stress, line energy, and dislocation-solute atom interaction. siderations were neglected. A qualitative explanation of the yielding process in terms of the activation of Frank-Read sources is presented, but the lack of a realistic solution to the dynamic dislocation problem involved prevents an extension of the model at present. A self-consistent correlation between the present calculations and experimental data for delay time associated with yielding and the temperature dependence of the upper yield stress was made. Favorable agreement was noted. It is concluded that extension of the original Cottrell-~ilby' theory which includes line tension effects can just as well describe the yielding process as other approximations3,5 which neglect line tension. Cottrell and Bilby found the form of the activation energy vs stress curves to be almost independent of the line energy and the interaction energy (i.e., these parameters only affected the activation energy U by a scale factor). A check of the theory was accomplished by comparison with the experimental data of McAdam and Mebs2 for the lower yield stress as a function of temperature. cottrell,3 in 1957, attempted to obtain a simple closed-form solution for the activation energy as a function of stress by linearizing the quartic equation relating displacement of the dislocation line and applied stress. It was argued that since the contribution to the line energy/atomic plane of a dislocation is made by the long-range stress field of the dislocation through a term of the type (Gb3/2p)ln A/ro, where

Jan 1, 1963

By J. W. Downey, M. V. Nevitt

MUELLER and Knott' have reported that Ti2Cu has the Si2Mo-type (Cllb) structure. The lattice parameters of Ti2Cu and the positions of the titanium and copper atoms are given by them as follows: ao = 2.944A; co = 10.786A 6 atoms per unit cell: 2 Cu (a) 000; 1/2: 1/2, 1/2 4 Ti (e) 002; 1/2, 1/2, 1/2 + z 00 -2 ; 1/2. 1/2. 1/2- z where z = 0.341 We have found eight other binary compounds which appear to be isostructural with Ti2Cu. All of them involve the combination of a titanium group element with a copper group element or a titanium group element with palladium. The compounds and their lattice parameters are given in Table I. The alloy samples were prepared by arc melting metals having a purity of 99.9 pct or better. They were homogenized for 72 hr at 900°C in evacuated capsules and quenched in tap water. Filings were prepared, and these were reannealed for 4 hr at 700°C, also in evacuated tubes. The lattice parameters were determined by a least squares Bnalysis of the back reflection lines of patterns made with cobalt radiation. It was not possible to make a precise determination of the lattice parameters of Zr2Au because the filings underwent rapid oxidation at room temperature. The evidence supporting the large tetragonal cell containing 6 atoms involves the presence of the reflections 002, 101, 004, 112, 114, 202, and other reflections with higher indices for which I = 3n. Three or more reflections corresponding to these indices, which cannot be indexed on the basis of a small body-centered tetragonal cell containing 2 atoms, are observed in the patterns of all of the compounds listed in Table I. They are strongest and most numerous when the components have large differences in their atomic scattering factors, as is the case in Ti2Pd, Zr2Cu, and Hf2Cu. The values of the z parameter should be regarded as approximate They were selected on the basis of the best agreement between the observed intensities of two pairs of closely spaced powder pattern lines (004): (112) and (114): (202) and the corresponding intensities calculated for several values of z. An ordering scheme was assumed in which titanium, zirconium, or hafnium atoms occupy the titanium positions of Ti2Cu, and copper, silver, palladium, or gold atoms occupy the copper positions. Calculated and observed intensities of the aforementioned lines and the other lines of the patterns are in satisfactory agreement, but a refinement of the z parameters and a final confirmation of the ordering arrangement must await a more detailed study. It is interesting to note that in transition metal systems at A2B stoichiometry the Si2Mo-type structure competes for stability with two other fre-

Jan 1, 1962

By J. W. Downey, R. A. Morris, M. V. Nevitt

An experimental survey has determined the occurrence of Ti2 Ni-type phases in binary and ternary alloys of Ti, Zr, or Hf with first, second, or third long-period transition metals and also in ternary systems in which 0 is the third component. The existence of eleven previously unreported isomorphs is cited and some approximate phase boundaries are given. Lattice parameters are reported for the new phases. The periodic table relationships involved in the occurrence of the phase are discussed. A previous paper' reported the occurrence of a group of Ti2Ni-type phases (structure-type E93) containing Zr in combination with Rh, Ir, and Pt, and Hf in combination with Rh, Pd, Ir, and Pt, and cited evidence to show that all of the Zr phases and the Hf-Pd phase are not binary phases but occur only when 0 is present. More recently one of the present authors found that in the ternary systems of Ti and 0 with Cr, Mn, Fe, Co, and Ni the occurrence of the Ti2Ni-type phase could be correlated qualitatively with valency electron concentration, and that the role of 0 appeared to be that of an electron acceptor.' This paper describes the results of a broader investigation of the occurrence of Ti2Ni-type phases in binary and ternary systems containing Ti, Zr, or Hf. EXPERIMENTAL PROCEDURE Alloys were prepared by arc melting. The purity of the metals Zr, Rh, Hf, Ir, Pd, and Pt was given in the previous paper.' The additional materials used for this work were metals with a minimum purity of 99.9 pct and reagent grade ZrO2, TiO2, Cr2O3, and NiO. Chemical analysis of several alloys in each system indicated good agreement between intended compositions and cast-button compositions. For the heat treatment precautions were taken to avoid oxygen contamination. The specimens were wrapped in Mo foil and sealed in fused quartz tubes at a room temperature pressure of 1 x mm of Hg. These capsules were then annealed in a vacuum furnace at an external pressure of 5 X mm of Hg. Since the major constituent in most of the alloys studied was the brittle Ti2Ni-type phase, Debye-Scherrer patterns with sharp lines could be obtained with powder crushed from the annealed solid samples without being reannealed. Metallographic, X-ray diffraction, and density determination techniques previously described1, 2 were used. RESULTS AND DISCUSSION Zr-Rh-O, Zr-Ir-Oand Zr-Pt-O Systems-Figs. 1 through 3 show the approximate boundaries of the Ti2Ni-type phase fields at the temperatures indicated in the ternary systems Zr-Rh-O, Zr-Ir-O and Zr-Pt-0. In each isothermal section the Ti2Ni-type phase field is labeled E and the unidentified binary phase found in the previous work' is labeled ?. The dotted lines, which are not phase boundaries, correspond to the composition (Zr2 B)1-xOx. In each system the Ti2Ni-type phase field is displaced to the high-Zr side of this composition line. However, the departures from Zr2B stoichiometry for the metal components are not as high as Zr3B, as was inferred from the previous work. Lattice parameters of the Ti2Ni-type phases in the alloys shown as triangular data points in Figs. 1 through 3 are given in Table IA. The probable compositions of the Ti2Ni-type phases have been estimated from the isothermal sections. Although there are a few exceptions that probably result from uncertainties in phase boundary location and from the effect of variation of oxygen content, the generalization can be

Jan 1, 1961

By G. Thomas, O. Johari

MEIERAN and Richman' have proposed a method for the determination of electron-diffraction patterns from twinned bcc crystals. It is the purpose of this note to present a general method capable of predicting the positions of all reflections from both matrix and twin regions for all possible orientations of cubic crystals. The method has been verified for twinning in explosively deformed copper, and is found to be a very useful tool when used in conjunction with stereographic projection, and dark-field electron microscopy. Let an isometric crystal be twinned in any (hkl) plane. The twinning can be considered either as a reflection in the twin plane or as a 180-deg rotation about the twin axis. In reciprocal space, the 180-deg rotation about the twin axis would place a given point p, q, Y to a new position P, Q,R, such that both will be equally inclined to the twin axis. This is represented in Fig. 1. If the coordinates of three such pairs of points are determined, the matrix of the transformation can be obtained.' It is also seen that such pairs would satisfy another condition, i.e., the vector [p,q,r] plus the vector [P,Q,R] will yield a multiple of hkl, i.e., the indices of the twin plane, because in isometric systems the axis [hkl] is perpendicular to the plane (hkl). This can be simplified by taking the points along the three axes. If a point h2 + k2 + 12, 0, 0 is taken along the x axis, it is seen that its new indices should be so as to yield a sum 2h2, 2hk, 2hl,i.e., 2h(h,k,1). (Also the points h2 - k2 - 12, 2hk, 2h1, and h2 + k2 + 12, 0, 0 are equidistant from the origin and both points make equal angles with the twin axis, i.e., cos-' h/dh2 + k2 + l2.) Similar operations for the points yield respectively, after the 180-deg rotation about the twin axis. Following the method due to Buerger,' the matrix of the transformation is obtained as follows: Using this matrix the indices of any other point in reciprocal space can be determined. This matrix is much simplified for (111) twinning commonly encountered in fcc metals, since are each equal to —1; hence

Jan 1, 1964

By G. D. Oxx, L. F. Coffin

The concept of using a phase that is liquid at service temperatures as a component of coatings for refractory metals has been described. The liquid, an alloy of gold and silicon, is retained on a molybdemum surface by a capillary system made of molybdenum disilicide. The coating has the advantage of good thermal shock and has a self-henling chracteristic. In order for highly stressed structures to exceed a service temperature of 2000°F, it has become apparent that a development that does not depend on the traditional iron, cobalt, or nickel-base alloy is needed. Alloys of the refractory metals, i.e., molybdenum, tungsten, tantalum. niobium (colum-bium), and rhenium are potentially useful at extremely high temperatures that more conventional alloys would never be expected to achieve. A new alloy of molybdenum has become available that has a 100-hr rupture strength of 35,000 psi at 2200°F.1 As a result of this work, it may be assumed that a material with adequate mechanical properties at 2200?F is now available. Unfortunately, this alloy and all other known alloys of the refractory metals suffer from not being; serviceable in an oxidizing atmosphere for a very long time. In order to permit general use of refractory metals at high temperatures, it is necessary to prevent destructive oxidation by appropriate alloying or by protective coatings. The liquid phase coating reported here is representative of a concept for the protection of metals that permits higher service temperatures and introduces a new group of materials for selection as coatings. Coating Requirements—Service conditions that represent potential applications for refractory metals vary considerably: however, it is possible to consider two conditions that are usually present. The refractory metal1 component must be heated to the service temperature at least once and usually frequently. When a solid coating is used, usually the coating and basis metal do not have the same coefficient of expansion. This difference causes thermal stress in the coating that is aggravated by rapid heating and cooling and eventually causes coating failure. The second consideration is the probability that there will be damage to the coating by some environmental condition. In jet engines, for example, it is expected that large particles, stones, metal parts, and so forth, will strike a bucket at high speed and cut a hole in the coating. In addition to the above considerations, the coating must also have other fundamental properties. Obviously, it must be oxidation resistant. In addi- tion, it must prevent permeation of oxygen and subsequent oxidation of the basis metal at the interface. Also, diffusion of the basis metal to the outside surface and subsequent oxidation there must be prevented. Finally, the coating must not react with the basis metal to form a weak bond at the interface. It should be noted that the coating need not support a load: it must only remain intact. Liquid Phase Coating Design Factors—In the use of a liquid as a principal coating constituent, it would be expected that the two service conditions mentioned above would be satisfied. A liquid would flow under the influence of the thermal strain developed by coating-basis metal expansion mismatch such that failure from this source would not be expected to occur. It is also probable that the ability of a liquid to flow would provide a self-healing effect. Thus, damage caused by particles in the atmosphere would be repaired. A third advantage is evident in that vapor pressure rather than melting point limits the service temperature. This latter advantage permits use of low melting but oxidation resistant metals such as gold or copper that would be quite useless if the solid state was required. A low viscosity liquid alone. however, is a totally unsuitable coating because it will simply flow off a component, particularly under the influence of a high acceleration field. A method of preventing liquid loss must be devised. In this case, it is desired that the liquid have a low viscosity so that it may easily flow into flaws; thus, raising the viscosity is not a satisfactory solution. One other inherent difficulty is caused by the high mobility of atoms in the liquid state as compared to the solid state. Because of this mobility, it would be expected that gases would diffuse more rapidly through a liquid. It is also probable that the solubility of oxygen in the liquid would be comparatively high. These factors would tend toward rapid permeation of oxygen and oxidation of the substrate. It is most reasonable that this problem could be solved by incorporating a solid phase, which is impermeable to gases, as a component of the coating. To overcome the tendency of a liquid to flow easily, it is possible to use surface tension to advantage. If the surface were made up of a large number of capillary tubes of sufficiently small diameter, then surface tension would hold the liquid even against the acceleration of a centrifugal field.

Jan 1, 1961

By C. H. Ma, J. S. Makris

This paper describes a modified X-ray diffraction microscope wzethod based on the general principles of Berg-Bawett's Method, Lang's Method, and Woosters1 Technique. This new method may be used for stationary or oscillatory photomacrographs, with both reflection and transmission techniques. In the stationary technique, a small point-source of a divergent X-ray beam of white radiation is used; in the oscillatory technique, a monochromatic divergent beam is employed. The chief advantages of the method described herein are that only approximate alignment of the Bragg angle -of-interest is needed during exposure, and that a large portion of a crystal-specimen surface can he irradiated. In recent years, the study of dislocations through X-ray diffraction microscope techniques has received considerable attention as concerns the per,fection of single crystals. Very promising results have appeared in the literature.'-' These techniques are very useful when the dislocation density of a crystal is near 105 per sq cm or lower. The principle of X-ray diffraction microscopy is based on the differential effect of X-ray primary extinction between the distorted parts and the perfect parts of a crystal. When an X-ray beam impinges on a set of planes of a crystal at the correct Bragg angle, diffraction of X-rays will occur with a 90-deg phase shift of the diffracted beam. This once-diffracted beam still forms the correct Bragg angle with the planes ahead of it, if the planes are not distorted. Thus, part of the strength of the once-diffracted beam is rediffracted and the beam continues to travel in the direction of the primary X-ray beam. This twice-diffracted beam not only takes away part of the strength of the once-diffracted beam, but also causes destructive interference with the primary beam because it now is exactly 180 deg out of phase with the primary beam. As the primary beam goes farther and farther into the crystal, its intensity is continuously weakened. Asa result, the ultimate X-ray diffraction power of a set of undistorted planes is largely extinguished. This phenomenon is called primary extinction.1,2'4 If the set of planes ahead of it is distorted, the once-diffracted beam will have less chance to be rediffracted, because it could not meet the exactly correct Bragg angle for rediffraction. If a part of the beam does rediffract, this twice-diffracted beam is not exactly 180 deg out of phase with the primary beam, and thus it causes less destructive interference. As a result, the registered X-ray diffraction power of the distorted part of a crystal appears stronger than that of the perfect part. Thus, point-by-point contrast of a diffraction image is theoretically possible from a crystal containing dislocations. Two most popular X-ray diffraction microscope techniques today are Berg-Barrett's Method and Lang's Method. Both of these techniques require a

Jan 1, 1964

By Y. L. Yao

A scale based on existing data for the mutual solid solubilities of metals in binary systems has been compiled. This scale is useful for estimating the relative magnitude of two solubilities, provided that these two metals do not form a continuous series of solid solutions. In his classic work on the theory of metallic solid solutions, Hume-Rothery proposed that, other things being equal, a metal of lower valency should be regarded as more likely to dissolve one of higher valency than vice versa. He called this the "relative valency effect". Later he found that the application of this general principle was limited to the three univalent metals: copper, silver, and gold. In dealing with alloys of these metals with metals of higher valency, Hume-Rothery stated, it is usually found that the solid solution of the univalent metal is of the greater extent.&apos; The question of whether there is a systematic trend in comparing the relative magnitude of mutual solubilities, if valency is not considered a criterion, naturally arises. The present paper proposes a "mutual solid solubility" scale on an empirical basis. The order of its arrangement is determined by the relative magnitude of mutual solid solubilities. Some applications of and inferences from this scale are given. ORIGIN AND SCOPE OF THE SCALE An extensive survey of over 400 binary phase diagrams shows that, with certain exceptions and within certain limitations, metals can be arranged in the following order: Na Zn Cd Ga Bi Ge Sb As Hg Sn Si Mg Li Ba In Tl Pb Al Ca Be Mn Cr Re W Mo Cu V Ta Fe Ru Cb Co Ti Hf Ni Os Zr Ir Pt Au U Rh Ce Pd Ag Th so that a metal higher on the scale tends to have a higher solubility in a metal lower on the scale than vice versa. some experimental solubility data are given in Table I. These data, unless otherwise specified, are taken from Hansen and Anderko.6 Throughout this paper, the term "solubility of A in B" is taken to mean the maximum solid solubility of A in B in atomic percent, while the maximum solid solubility of B in A in atomic percent is called "reciprocal solubility". The exceptions found in the above-mentioned survey are: Al-Ag system, 20.34 at. pct Al in Ag, 23.8 Ag in Al; Mg-Pb system, 5.9 Mg in Pb, 7.75 Pb in Mg; As-Sn system, vanishingly small amount of As in Sn, a large As primary solid solution; Cd-Hg system, Cd in Hg <Hg in Cd; Mg-T1 system, 5(?) Mg in ßT1, 15.4 T1 in Mg; W-Ti system, -21 W in ßTi, -25 Ti in W; Zr-Ag system, probably small amount of Zr in Ag, 20 Ag in ßZr.7 Considering the large number of examples, the exceptions are few. It is remarkable that metals may be so arranged without considering other factors such as size, crystal structure, electronic structure, and so forth. A limitation to the use of the scale is the fact that it cannot be tested for two metals forming a continuous series of solid solutions. It is interesting to point out that when the continuous solid solutions have a miscibility gap the critical composition is

Jan 1, 1961

By N. A. Tiner, Saara Asunmaa

The early stages of nucleation of vacuunz-deposited gold coatings on electropolished surfaces of pure nickel and binary alloys of Ni-A1 and Ni-Ti have been studied, employing carefully controlled temperature and vacuum conditions. The preferential nucleation of gold on slip-plane edges on the crystal surfaces gave rise to a very pronounced decoration effect which was used in studying the morphology of the pain boundaries, etch pits, and slip planes on crystal faces. Characteristic surface patterns were observed with local dimensional variations. Most pronounced geometric patterns were ohserved in the pain boundaries of nickel and on the crystal faces of Ni-A1 alloy. The nucleation behauior of gold indicates contamination on the Ni-T1 alloy and the getter properties of titanium are presented as a tentative interpretation for the detected contamination. The markings and imperfections on cleavage surfaces of crystals have long been observed by optical microscopy.1,2 Surface-replica techniques are also used to detect the smallest steps and structural details on crystal surfaces by electron microscopy. Recently, Bassett3-5 and Sella, Conjeaud, and Trillat 6 have shown that, when gold is deposited onto a rocksalt cleavage surface, there is preferential nucleation at steps and slip-plane edges on the crystal surface. This gives rise to a very pronounced decoration effect which is invaluable for revealing the lattice steps on a cleavage surface. According to Bassett,3 gold atoms arriving during vacuum evaporation at the crystal surface have a certain mobility on the surface after arrival, particularly at elevated temperature. If a significant

Jan 1, 1964

By D. Kuhlmann-Wilsdorf

A new theory of work hardening is developed which rests on only a few simple principles and is applicable to a wide variety of materials and dislocation structures. It explains, qualitatively, the general characteristics of the three-stage shear stress-shear strain curve of fcc metals, as well as the most important features connected with easy glide. The value of, the work-hardening coefficient in stage 11, is derived quantitatively, from first principles, and is shown to be insensitive to many changes in detailed dislocation behavior. The low rate at which energy is stored during mechanical working in stage 11, the dependence of slip line length on stress, and the so-called Cottrell-Stokes law are explained. Although the theory is primarily developed for fcc metals and alloys, it is largely applicable also to some other materiuls, and in particular apparently to poly crystalline simple steels in theh linear range of hardening. FOR pure fcc metals and alloys, after many different pretreatments and under a great variety of testing conditions, the well-known three-stage work-hardening curve is commonly observed. Moreover, the work-hardening coefficient in stage 11, 011, bears an almost constant relationship to G, the modulus of rigidity, such that K = G/Brr" = 300, within a factor of about 2 either way. The very fact that the three-stage curve and the value of K are so very persistent, can hardly be understood except on the assumption that some quite simple phenomena are responsible. In this light all available theories of work hardening must be judged unsatisfactory: Each of them is based on some quite specific dislocation model, while it is known that metals with similar values of K may have widely different dislocation structures, for example, dislocation tangles in the case of pure fcc metals, and piled-up groups of dislocations in a-brasstype alloys. Thus, while a particular theory might conceivably represent a special case with a reasonable degree of accuracy, it cannot possibly have illuminated the underlying principle. Another severe criticism applying to the presently most widely discussed theories is that they employ empirical or even quite unexplained parameters in order to arrive at a numerical value of K. This indicates that some vital aspect of the work-hardening process has remained unexplained. In the present paper, a new theory of work-hardening is developed which does not suffer from the above shortcomings. It is derived from first principles, and is founded on the realization that the underlying causes for the three-stage work-harden- ing curve and for the persistence of the numerical value of O,, must be few and simple. As a first step it is shown herein that the experimental oklservations on "easy glide" can be explained on the assumption that in this stage dislocations multiply, beginning in a number of restricted area:;, and from there penetrate into crystal regions still substantially free of glide dislocations until a quasi-uniform dislocation distribution has been established. During "easy glide", the resistance to the movement of the foremost dislocations, advancing into still largely a dislocation-free regions, is believed to control the flow stress. This resistance is determined by various factors, among them the reaction of the dislocations against bowing-out into loops, which is due to their line tension. Without such bowing-out no dislocation multiplication would occur, but the corresponding contribution to the flow stress stays practically constant all through stage I. According to the present theory, stage TI begins as soon as there are no more areas left into which newly formed dislocations have not yet penetrated, i.e., as soon :is a quasi-uniform dislocation density has been attaned. Regardless of how, in detail, the dislocations should happen to be arranged at the end of "easy glide", whether in pileups or tangles, on one glide system or on several, the stress necessary to overcome the line tension of the segments must now increase, because the dislocation density increiises while the average free dislocation length decreases. It is argued that all other contributions to the flow stress stay at about the same level as in stage I, and that the said increase in the stress necessary to bow out dislocation segments into loops is responsible for the major part of work hardening in stage 11. Stage TII of the stress-strain curve is explained by largely follouing the ideas of Seeger and his co-workers; however, it is pointed out that the start of stage m may on occasion indicate the onset of "conservative climb" instead of cross slip. Indirect evidence indicates that linear work hardening persists even under the profuse action of cross slip, but at a lower rate. Only when, in addition to cross slip, climb operates extensively, does the work-hardening rat,. drop sharply, ultimately down to zero. BRIEF SURVEY OF EXPERIMENTAL EVIDENCE The stress-strain curve of crystals of many different substances are qualitatively similar. No significant permanent plastic deformation takes place below the so-called "critical resolved shear stress". As the applied stress is raised beyond this level. yielding occurs with little or no work hardening and

Jan 1, 1962

By H. H. Hausner, H. S. Kalish

IN recent years zirconium and beryllium have become of great interest because of their special properties. Zirconium is known for its remarkable ability to absorb the gases oxygen, nitrogen and hydrogen, has extensive use as a getter in electron tubes and has indeed become the primary metal in the manufacture of photoflash bulbs. Beryllium is of interest because it is light metal and as such could supplement aluminum and magnesium in light metal applications. The late development of zirconium ductile at room temperature produced by both the iodide and calcium reduction methods has stimulated frequent articles on this metal in the recent literature. Beryllium is so closely akin to zirconium crystall-ographically that some investigators are of the opinion that beryllium should be ductile at room temperature if it were pure enough. The crystallographic relationship of beryllium and zirconium, the interest created in Be-Zr alloys where the zirconium is believed to function as a remover of solid solution gas in beryllium and thus increase the ductility of the beryllium rich alloy, and the amenability of these metals to powder metallurgy methods stimulated this approach to an investigation of the system Be-Zr. Preliminary work showed that a zirconium compact melted when pressed in contact with a beryllium compact and sintered at 1250°C, without any apparent final bond between the two. A search of the literature revealed nothing pertinent about the Zr-Be system or about any zirconium - beryllium alloys. In 1936 Misch³ vaguely identified the compound ZrBe2 and reported that it has a deformed cubic structure. Two patents," , were issued to Mach-lett Laboratories in 1943 and in 1946 regarding the addition of small amounts of zirconium to beryllium in order to improve the ductility of beryllium so that it may better be used for X ray windows. In considering the alloy formation of beryllium with other metals Raynor thought it unlikely that zirconium would have any solid solubility in beryllium. From this indefinite information nothing could be ascertained about the possible use of Zr-Be alloys or the use of these metals intimately. A brief preliminary study of the system Zr-Be by powder metallurgy methods was undertaken to provide regions for more intense studies of Zr-Be alloys and as a guide to a formal phase diagram determination. The investigation described in the following pages was completed in less than three weeks. The usefulness of this rapid method to facilitate phase diagram studies and new alloy investigations will be shown. Materials and Procedure Zirconium powder was produced from crystal bar stock of high purity supplied by Foote Mineral Co. The crystal bar was converted to hydride at 730°C, crushed, and the hydride decomposed in purified argon at 850°C. 60 pct of the powder passed through a 325 mesh screen, 40 pct of the powder was —140 mesh, +325 mesh and the average particle size of the total was 8 to 12 microns. The beryllium powder was used as re-

Jan 1, 1951

By N. Brown

From the theories of flow and fracture it is shown that the difference in reciprocals of the transition temperatures (OK) is a quantitative measure of temper ernbrittlement. Experimental data are given which support this conclusion. STUDIES of temper embrittlement have been made with various viewpoints; some involved a study of the kinetics of the reaction,&apos;,2&apos;" others were concerned with changes in the mechanical properties," . and still others investigated the structural differences between the embrittled and unembrittled states by means of the microscope," -ray ,&apos; measurements of electrode potential," and electrical resistivity." In all these studies, the difference in impact properties was the criterion for distinguishing between the embrittled and unembrittled states. By common agreement the difference in transition temperatures is now taken as a measure of the degree of embrittlement, and this measure is used as the quantitative measure upon which studies of the phenomenon are based. Vidal and Jolivet"&apos; suggested that the degree of embrittlement was the same whether measured by impact or by a slow bend test. Their conclusion was based on tests at constant temperature. Jaffe and Buffum" using the impact and slow bend test showed that, for a given degree of embrittlement, the difference in transition temperatures varied with the method of testing. Hultgren and Chang" showed that the difference in transition temperature produced by the V-notched Charpy impact test was not the same as the value produced by the keyhole-notched impact test. This is a very fundamental point because there can be no true measure of temper embrittlement unless it is invariant with respect to the test method. Theory The purpose of this analysis is to show how the transition temperature depends on the stress distribution and the state of the material. The criterion for fracture may be stated as follws: where iF is some function of the principle stresses and F is a critical condition depending on the state of the metal. Similarly yielding occurs when Where is another function of the prin-ciple stresses and Y is a critical value depending on the state of the material. There is some controversy as to the exact form of the function f,.(u,, u,, us) but most experimental data give the following criterion:= where 8, and 6, = 21 or 0. Combinations of 6, and 8, give the various criteria such as the normal stress, the hydrostatic tension, and the maximum shear stress laws. Agreement is general that the octahedral shear stress law adequately describes the yield condition. where the bracketed factor will be called the design factor and depends on the stress distribution. Whether the metal decides to yield or fracture is given by the condition that The effect of temperature and strain rate upon Y has been studied by various investigators, It has been shown by MacGregor and Fisher" that

Jan 1, 1955

By R. E. Ogilvie, J. I. Goldstein

The a and y solubility limits in the Fe-Ni phase diagram have been redetermined at temperatures above 500°C. Both a diffusion-couple and a quench and anneal technique were used. The solubility limits were measured with an electron-probe micro-analyzer. The nickel concentration at the phase boundary is increased below 700°C and the a solid-solubility range is much larger than had been previously measured. The solubility limits are also extrapolated to 300°C. It is suggested that the solubility of nickel in a FeNi reaches a maximum in the temperature interval 400o to 500°C. THE equilibrium diagram has been of great use in the field of meteoritics, where the phase relations between kamacite (a) and taenite (?) in metallic meteorites can be described by means of the Fe-Ni diagram. The study of metallic meteorites by electron-probe microanalysis has cast some doubt on the accuracy of the presently available Fe-Ni diagrams.1, 2 Recent thermodynamic studies of the Fe-Ni system also suggest that the diagram may be in error.3 For these reasons the high-temperature (800o to 500°C) part of the diagram was redetermined. INTRODUCTION The currently accepted Fe-Ni diagram is that of Owen and Liu.4 Above 910°C, there is a region of complete solid solubility, ? (fee). Below 910°C, the a (bee) phase is stable in pure iron. The effect of increasing amounts of nickel is to stabilize the y phase. The phases that form when Fe-Ni alloys are heated or cooled bear little relation to the equilibrium diagram. If an alloy is cooled from the y state and held at a temperature within the a + y field, no evidence has been found for the occurrence of the y - a transformation.5 If the alloy is cooled to low enough temperatures the y phase breaks down into a supersaturated bee phase called a2. In fact, in alloys over 27 pet Ni the y phase is retained at room temperature.6 The a, phase has the same composition as the original y. The temperature at which a2 forms, the M, temperature, has been determined experimentally.&apos; A state of equilibrium can be approached by cooling below M, to form a2 and then reheating the alloy into the two-phase region of the diagram. The y phase will then begin to precipitate out of the a2 phase and grow. The growth of the phase, however, is quite slow. Using the interdiffusion coefficients -Da of Goldstein et al.,8 it is estimated that it takes about 1 year to grow a 10-µ-wide region of ? at 700°C in a 5 pet Ni alloy. Owen and Liu4 used the technique just described to form the equilibrium phases. The phases present were determined by means of X-ray analysis. The accuracy of their diagram depends on the number of alloys available near the phase boundary at a given temperature. In this study two different techniques were used to determine the a/a + ? and ?/a + ? solubility limits. The inherent accuracy of both techniques was greatly improved over that used by previous investigators because the phase-boundary compositions were measured with an electron-probe microanalyzer. PROCEDURE The two methods used to determine the Fe-Ni diagram are the diffusion couple (D.C.) and the quench-and-anneal (Q.+A.) techniques. In the first method, diffusion couples whose diffusion path goes through a two-phase region of the phase diagram were used. A description of the technique used for making the diffusion couples has been described in a previous paper.&apos; After the diffusion treatment a discontinuity in the resultant concentration vs distance profile was measured. The nickel concentrations in the a and ? phases at the interface of the discontinuity are the solubility limits of the a and ? phases in the phase diagram at the diffusion temperature. If the interface com-

Jan 1, 1965

By Stanley L. Lopata, Eric B. Kula

SEVERAL methods are available for determining pole figures by X-ray means.&apos; The earlier film methods have been replaced by techniques in which the intensities are measured by Geiger counters on an X-ray diffractometerZm7. These methods utilize either flat transmission or reflection samples,214°8 cylindrical specimens,3 or spherical specimens.7 A single transmission or reflection sample will not yield information over the complete pole figure. The cylindrical specimens suggested by Nortod and the spherical specimen of Jetter and Borie7 have the advantage of allowing the whole pole figure to be obtained without any corrections to the intensity for absorption, There is a lower limit to the size of sheet which can be conveniently studied, however, and sample preparation can be rather tedious. Probably the most common method today for determining the complete pole figure is that developed by Schulz. A flat reflection sample is used for determining the pole figure from the center out to about 70". Because of the geometry of the system used, little or no correction for absorption or irradiated volume is necessary.4&apos;6 A separate transmission sample is used for the region near the edge of the pole figure. This procedure requires two separate samples, one of which must be a thin carefully-prepared transmission sample. Furthermore, since the reflection and transmission data are in different arbitrary intensity units, a region of overlap must be obtained, and the intensity data from one set of measurements converted to units of the other. These are serious disadvantages of this method, and they point out the need for a simplified procedure. Since the reflection technique can be used for planes whose normals lie up to about 70° from the sample surface normal, it is apparent that a complete pole figure can be obtained by reflection alone if sample surfaces are cut at oblique angles to the rolling plane Specifically, if a rolled sample is cut so that the normal to the surface formed lies equidistant (54" 44&apos;) from the rolling plane normal, rolling direction, and transverse direction, then complete information for one quadrant of the pole figure can be obtained by reflection from the surface. Fig. 1 shows this oblique surface, as well as the position of the pole of this surface in the pole figure. When a surface has been cut oblique to the rolling plane, the standard polar stereographic net is inconvenient to use, and it would be more desirable to have the center of the net coincide with the pole of the oblique plane. Fig. 2 shows such a net, where the center has been offset 54&apos; 44&apos; to correspond to a specimen cut as in Fig. 1. This net was in effect obtained by rotating a standard polar stereographic net 54" 44&apos; with the help of a Wulff net. With the experimental setup used. a sample can be cut from plate of 1/2-in. thickness or br Gen- erally thinner sheet is under investigation, and a composite specimen must be used. A convenient procedure has been to bond together, using epoxy resin, sufficient sheets to form a cube. These are clamped in a vise, and when dry, the whole vise is rotated and a flat surface ground at the required angle. The sample is then mounted inside a larger steel ring using Koldmount, the backside ground flat and parallel to the approximate desired thickness. The sample is then polished and etched to remove any effects of cold working during grinding. With this method only one quadrant of the pole figure is obtained. Often pole figures show symmetry around the rolling and transverse directions, and any slight asymmetry is due to scatter. Should the pole figure not be symmetrical, four oblique surfaces, corresponding to the four quadrants of the pole figure, would have to be examined. In most cases there is symmetry at least around the rolling direction, which reduces to two the number of quadrants to be investigated. For many purposes all that is required is an average polefigure for the four quadrants. This canreadilybe obtained with composite specimens of sheet material, where sheets Corresponding to each of the four quadrants can be intermixed. The four quadrants can be obtained by considering sheet in the normal position, and by rotatiolls of 180" around the sheet normal, the rolling direction, and the transverse direction. If desired, the whole thickness of the sheet can be used, yielding an average of the surface and interior textures; or the surface material can be removed from each sheet, resulting in a pole figure for the center alone. Eugene S. Meieran of the Massachusetts Institute of Technology has independently developed the same method. His results, adapted to a pole figure goniometer with a specimen spiraler, will be described in another publication.

Jan 1, 1962

By Wilhelm in der Schmitten

A necessary condition for a uniaxial stress distribution in compression testing is that the specimen end surfaces make full contact with the apparatus compression plates. In addition all compression surfaces must be machined perpendicularly to the loading axis. In general this condition is not sufficiently fulfilled in a simple apparatus. Therefore on application of the load the compression plate may first touch the specimen at one point only. This may cause crumbling of the specimen if the material under investigation is brittle (e.g., ionic crystals). It also precludes a precise study of the beginning of deformation as the effective cross sectional area is unknown. In addition a complex stress state is set up which renders quantitative work impossible, e.g., on dislocation density and arrangement as a function of stress. To overcome these difficulties a "self-adjusting" upper compression plate is required. A simple and reliable method of achieving this is shown in Fig. 1: A steel hemisphere is placed on the specimen. As soon as the compression plate (which no longer needs precise machining) makes contact with this hemisphere the load is uniformly distributed over the whole specimen surface, and a homogeneous uniaxial state of stress results. An advantage of the method is that the loading is always central, even when the specimen surface is not parallel to the compression plate (Fig. 1, dashed line). Therefore to a first approximation a uniaxial state of stress is maintained if the misorientation of the specimen end surface is small. Two typical stress strain curves of identically prepared sodium chloride crystals are shown in Fig. 2; curve 1 was taken without and curve II with the hemisphere under otherwise identical conditions. The improvement in the accuracy of defining the beginning of elastic and plastic deformation is clearly indicated. The writer wishes to thank the Deutsche Forschungsgemeinschaft for financial support, and Professor P. Haasen for providing facilities for this work.

Jan 1, 1962