2'447-23- P ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR Project 24~47 Preliminary Report on ci1uDIES OF THE STRUCTURE OF THE OMEGA PHASE IN Ti-8Cr AND Ti-I3Mo, ALLOYS by W. C.\\Bigelow 15 August 1957 METAIILLURGY RESEARCH BRANCH AERONAUTICAL RESEARCH LABORATORY (WCRRL)

X /-/ A

SUMMARY Studies have been rriade of Ti-8Cr and Ti-13Mo alloyrs by electron diffraction., electron microscopic., and singlecrystal x-ray dirfraction techniques to attempt to determine the structure of the transition omega phase of these alloys. The electron microscopic studies have shown the omega phase to occur in the f orm of platelets or rods 0 approximately 4QOA in thickness. The x-ray diffraction results obtained support those previously reported by 5 Silcock, Davies., and Hardy, and indicate a hexagonal 0 unit cell with a = 4.6bA and C = 2.82 is preferable to the cubic or orthorhombic cells proposed by otherA investigators. -

INTRODUCTION This report consists of a preliminary description Lwork which was undertaken at the request of the Metallurgy Research Branch of the Aeronautical Research Laboratory of the Wright Air Development Center, to investigate the crystallographic structure of the omega Phase of titanium-base alloys. Interest in this work arose from the fact that the structure determined for this phase by Parris, Schwartz, and Frost1'2'3 differed funda-'mentally from that determined independently by Spachner 4 zand Rostoker. In this connection it is interesting to note that a third structure has also been reported for the phase by Silcock, Davies, and Hardy.5 The purpose of the present research was to attempt to clarify the situation, and if possible to decide on a correct struture for the omega phase. Before proceeding with a description of the present work, it is desirable to review briefly the three previous investigations. REVIEW OF PREVIOUS INVESTIGATIONS The omega phase is a transition phase which forms in alloys of titanium and such elements as vanadium., chromium, iron, molybdenum, manganese, cobalt, and nickel,, as an intermediate step in the transformation from the body-centered cubic beta phase to the hexagonal alpha phase. The occurrence of the omega phase tends to supPress the martinsitic beta-to-alpha transition and increases the hardness and strength of the alloys. The studies of Parris, Schwartz, and Frost were carried out principally on a small single-crystal grain of Ti-8Cr alloy using rotation and Weissenberg single

426C (800~F). The solution and aging treatments were carried out on ar electrical resistance heater in a vacuum bell jar at 5 x 10 5 mm Hg pressure. From the data obtained it was concluded that the omega phase has a body centered cubic unit cell with ao 9.80o. The space group chosen was I13m with 54 atoms per unit cell. This structure is similar tO that of gamma brass in that its lattice parameter is approximately three times that of the parent matrix. It was also concluded that the (100) Planes of the omega phase were parallelto the (100) planes of the parent beta phase in the alloy. The studies by Spachner and Rostoker were carried out by powder x-ray diffraction techniques using 200mesh powders of Ti-8Cr, Ti-13Mo, and Ti-15V alloys. These powders were solution-treated for 30 min at 9000~C (1652), water-quenched, and then aged respectively, 8 hr at 5000 (932~F), 250 min at 4000C (7520F), and 100 hr at F (752). These heat treatments were carried out in argon-filled Vycor glass capillaries. The capillaries were water-atuenched after the solution and aging treatments to prevent transformations such as might occur during slow cooling. From the analysis of 30 lines in the Powder diffraction patterns an orthorhombic unit cell was chosen for t —he omega phase with the following lattice constants for. the three alloys: a b c Ti-8Cr 6.203 6.489R 3.3 Ti-13Mo 6.231 6.500 13.52 Ti.-15V 6.205 6.597 13.63 The studies by Silcock, Davies, and Hardy were carried out by single-crystal oscillation x-ray diffraction method'-s on a sige-rsaI gri of% a~r Ti0-15V alloy wIch.

obtained in this way showed diffraction spots which could be indexed on the basis of a cubic lattice of the size and orientation proposed by Parris, Schwartz, and Frost; however, it was also found that a simpler hexagonal unit cell with a = 4.60 and c ~ 2,82a would fit the dat equally well and provided better agreement between observed and calculated intensities. The space group chosen was P6/mmm with atom pobitions at, or near, 000, 2 2 For this hexagonal cell the precipitatio orientation would be with (0001) omega planes parallel t the (111) planes of the beta matrix, and it was proposed that coherency occurred on the (1210) omega planes. It is believed that the different results reported by these different groups of investigators result from experimental difficulties attending the study of the omega phase. Since it is a transition phase, it is intrinsically unstable, and probably cannot be prepared free of the beta phase; therefore pure single crystals of sufficient size for structural analysis by the usual x-ray diffraction methods cannot be obtained. Analysis of the omega structure from studies of a beta, singlecrystal grain containing the omega phase'is also fraught with difficulties. Due to the symmetry of the beta structure., more than one equivalent crystallographic orientation is possible for the omega crystals within a given beta crystal. Patterns obtained will thus be composites of patterns from these several orientations rather than the desired case of patterns from a single orientation. This greatly complicates the analysis of the patterns. In addition the omega reflections are weak compared with the beta reflect-ions in most cases, and in some cases coincide with beta reflections. This further increases the difficulty of accurately analyzing the patterns.

MATERIALS AND METHODS In the work to be described here, samples of Ti-8Cr and T-3Mo alloys were used. These were supplied by the Aeronautical Research Laboratory Qf the Wright Air Develo men t Center, and it is understood that they were prepared by Prof. Rostoker and his associates at the Armour Research Foundation of the Illinois InstitUte of Technology. The heat-treatment reported for the Ti-8Cr alloy is: solution treatment at 900~C (1652~F) and water-quenched, then aged hours at 400~C (752~F). The Ti-13Mo alloy was reportedly lsolution-treated at 900~C, water-quenched, then aged 250 m at 400~C. As received, the samples were in the form of rods 1/4" in diameter and 3-4" long. These were cut to shorter lengths more convenient for use, but were not given additional heat treatments. The different studies that have been made on these specimens include: (1) electron microscopic studies to determine the size, shape, and distribution of the omega particles in the alloys, (2) electron diffraction studies by reflection'I1echniques from polished and etched surfaces to attempt to obtain diffraction patterns of the omega phase from which information on its structure could be derived, t I) xZI-ray diffraction studies using a singlecrystal graiLn of-L t'he Ti-8Cr alloy and a Buerger precession camera. The electron m.Lcroscopic and electron diffraction studies involved examinations of surfaces of the specimens which were polished and etched so as to bring the particles of the omega phase into relief with respect to the beta matrix. For the electron microscopy, these surfaces were examined by means of surface replicas in the form of thin collodion ~ — _%_!1 _ -_ _ _ 31-_ fils, reare b evpoatio of a0 drop of_ a 0

palladium to increase their contrast, and then examined in a Philips Model EM-75 electron microscope. The electron diffraction studies were carried out by directing a collimated beam of electrons across the etched surfaces at a grazing angle so that it would strike the protruding omega particles and produce electron diffraction patterns from them. For this purpose an RCA Model EMD electron diffraction instrument was used. The x-ray diffraction studies were made on a singlecrystal grain of the Ti-8Cr alloy which was separated from the larger bar by careful etching. The grain was mounted in the goniometer head of the Buerger Precession Camera, and its crystallographic orientation relative to the coordinates of the instrument was determined. It was then re-oriented with selected crystallographic axes of the beta phase perpendicular to the x-ray beam and patterns were recorded for various levels of diffraction. RESULTS In the electron diffraction and electron microscopic studies, difficulties were encountered in finding procedures which would produce the high quality of polishing and etching required. Because of the high sensitivity of these techniques, the surfaces must be clean and free of all contaminants. Dirt or residues from the polishing or etching treatments may be picked up on the surface replicas and obscure the true structures when the replicas are observed in the electron microscope. Such contamination may also interfere in electron diffraction by producing extra diffraction rings or by preventing the electron beam from striking the true surface of the sample. For polishing, both mechanical and electrolytic methods were tried, including more than 20 polishing electrolytes of compositions given by Tagima7 and Osadchuk,8 and various modifications thereof. An even greater number of etching

reagents, both immersion ~nd electrolytic, were also tried. Again these included varliis reagents recommended for titanium and its alloys, and compositional modifications of these reagents. Due to the high resistance of the Ti-8Cr and Ti-13Mo alloys to electrochemical attack, unsatisfactory results were usually obtained with electrolytic polishing mwthods. When mechanical polishing methods were used, however, extreme difficulty was encountered in removing the abrasive polishing powders to obtain surfaces clean enough for electron diffraction. The most satisfactory overall results were obtained by mechanically polishing with 3/0 emery, and then with Linde "A" alumina on a wax lap, and following this with a light electrolytic polish in one of the reagents listed in Table I to remove worked metal and the polishing powders from the surfaces. Table I. Solutions used for Polishing Titanium Alloys (1) 77 ml glacial acetic acid 15 ml chromic acid (43%) 8 ml hydrofluoric acid (48%o) Should be used fresh for best results at a temperature of about 1000 with currents of 3-5 amp/cm2. (2) 100 ml solution (I) 6 ml perchloric acid (7cio) use as above (3) 37 ml nitric acid (tech. conc.) 37 ml lactic acid (85%) 18.5 ml1 hydrochloric acid (37%) 7.5 ml hydrofluoric Immersion, use at 60-80oc Most of the etching reagents tried produced pitting and staining of the surfaces; howeve-r, it.Tq tn Q wasposibl-.I= -et

Electron micrograprs have been obtained from both of the alloys which Show a fine dispersion of very small precipitate particles. Typical micrographs are reproduced in Figure 1. The particles generally appear to be in the form of rods or platelets, and their "diameter" or "thickness", as the case may be, appears to be of the order of o 400A. It is believed that these particles are the omega particles, for their size and distribution as indicated by the electron micrographs is in general agreement with the conclusions previously reached from the characteristics of their x-ray diffraction lines. 12,3 They are also of the proper size and dispersion to produce marked age-hardening of the alloys. The electron diffraction results were in general less satisfactory in that it was not possible to obtain patterns which could be definitely identified as arising from the omega phase. The principal difficulty apparently involved surface contamination during the polishing and etching operations, for in most cases diffuse halos or powder patterns of continuous rings were obtained rather than the spotty patterns expected from large-grained sample of the type being examined. Because of the small size of the omega particles and the general difficulties encountered in etching them into relief these results are not unexpected, for the slightest trace of surface contamination would be sufficient to prevent good electron diffraction patterns from being obtained. In the x-ray diffraction studies of the Ti-8Cr singlecrystal grain, patterns were obtained which showed the weak omega reflections among the stronger reflections from the beta phase. Unfortunately Dr. C. E. Nordman, who carried out most of these x-ray studies, is traveling abroad at the present time so that it is not possible to present the data in detail; however, the results will be summarized now, and a full report will be submitted upon his return in September. In general, the analyses of patterns from the

a. Ti-8Cr alloy. b. Ti-13Mo alloy. Fig. 1o Electron micrographs from the titanium-base alloys (X30,000). 8

hkl, hk2, hk3, and hk4 levels of the reciprocal lattice of the beta phase, plus patterns along 111, and 100 axes, showed very good agreement with the hexagonal structure 5 proposed by Silcock, Davies, and Hardy. Indices matching the triple-body-centered cubic cell proposed by Parris, 1 Schwartz, and Frost, could be assigned (as has been pointed out by Silcock, et al.); however, the patterns of the individual reciprocal lattice levels provided by the precession technique, showed too many reflections missing to make this structure plausible. It was not possible to show any reasonable agreement between the observed reflections and the orthorhombic structure proposed by Rostoker, et al. As a further check, a comparison was made of the "d" values reported by the various investigators. As shown in Table II, the "d" values and intensities reported by Parris Frost and associates, agree very well with those reported by Silcock, Davies, and Hardy. It is also interesting to note that almost the same set of t"dt" values appear in the data of Rostoker, et al. but that there are several extra lines. Since intensity data are not given, It is difficult to make a good comparison, but it is possible that these extra lines arise from some material other than the omega phase. For example., many of them agree very closely with'Values reported for the alpha phase of the titanium alloys. They could also re~sult from oxides or other surface contaminants, which would be more likely to contribute to powder patterns than to single-crystal patterns. DISCUSSION AND CONCLUSIONS From the results which have been obtained, and from an analysis of thoso- previously reported, it appears that the hexagonal structure pr oposed by Silcock, Davies, and Hrdy st epeerdt hecbcsrcuepooe

TABLE II. COMPARISON OF DIFFRACTION DATA POR OMEGA PHASE Ti-b-r- -Ti-7C'7 T7 ~Ti -bV Rostoker Parris,3 Silcock5 d d I hkl d I hkil 3.98 vvw 1 o*a 2.8o6 2.79 w 222 2.82 mw 0001 2. 560'8 2.2720 2.28 Vvvs 330 2.30 Vs ioTI1 1.8178 1.99 vvw 2Q2U0 1.7803 1.77 vvw 125 1.78 m 1121 1.7269 1.71 vvw 1.6244 1.62 Vs 600 1.62 s 2021F 1.5933 1.547 vvw 1.4741 1.50 vvw 12-30 1.4088 1.402 m 444 1.41 Vs 0002 1. 3613 1.3253 1.322 vvs 336 1.33 vvw 1231 1.3041 1.2494 1.2299 1.2013 1.200 rn 118 1.20 Vs 1 122 1.1507 1.15 Vs 66o 1.15 () 22-40 1. 1300 1. 0976 1.10 vvw 134 1.o0843 1.0692 i.o6 w 22-41 1.0290 1.026 VW 390 1.3 s 12-32 1.0089 0.9923 1.000 VW 448 0.99 vvw 4 o4o 0.9676 096 VW 277 0.97 S30-32 0.9426 0.949 Vs 666 0.94 m, 4 o-4

as to the- type of unit cell and the exact space group,, however, in spite of the fact that the hexagohlal cBe11 proposed g-ives reasonably good agreement with observed data. These uncertainties arise from difficulties in interpreting the patterns as a result of the fact that it has not been possible to separate the omega phase from the parent beta phase. The multiplicity of orientations of the omega particles in the beta matrix, and the fact that in some cases beta and omega reflections coincide makes unequivocal interpretation impossible. There is a possibility that some of these ambiguities could be resolved by electron diffraction studies, if the omeiga particles can be separated from the beta matrix by extraction replica techniques. Because of success which has bee~n achieved with these methods in studies of the phase of nickel-base alloys., it is proposed to attempt applying them to the omega phase before the final report on this work is submitted.

REFERENCES. W. M. Parris, C. M. Schwartz, and P. D. Frost:. Precipitation Hardening and Embrettement of High-Strength Titanium Alloys." WADC Technical Report 54-355, Wright Air t te- _ h- O Develomn''e r I-1tb 2. A E. Austin and J. R. Doig: "Structure of the Transition Phase Omega in Ti-Cr Alloys." Transactions AIME, Journal of Metals, January 1957, p. 27. 3. P. D. Frost, W. M. Parris, L. L. Hirsch, J. R. Doig, and C. M. Schwartz: "Isothermal Transformation of TitaniumChromium Alloys." Transactions ASM, 46, 231 (1954) 4. S. A. Spachner, R. F. Domagala; A. W. Goldenstein, and W. Rostoker: "Structural Changes in Commercial Titanium and Titanium-Base Alloys on Heat Treatment." WADC Technical Report 55-352, Wright Air Development Center, Ohio, 1956. 5. J. M. Silcock, M. H. Davies., H. K. Hardy: "The Structure of' the Omega Precipitate in Ti-16V Alloy." Inst itute of' Metals Monograph and Report Series No. 18.,'rTheI~i7-titute 6. L. 0. Brockway and W. C. Bigelow: "The Investigation of Minor Phases of Heat.-Resistant Alloys by Electron Diffraction and Electron Microscopy" WADC Technical Report 54-589, Wright Air Development CeFter77Oio,_ 9567 7. 5. Tajima and T. Non:. "Electropolishing Titanium." Products Finishing, 19, 26 (1954) 8. R. Osadchuk, W. P. Koster,, and S. F. Kahies: "Recommended Techniques for Polishing Titanium for Netallographic Examination." Metbal Progress, Oct. 1953, P. 129; "Metallographic&SruCtures in Commercial Titanium.," ibid.,. Nov. 1953, P. 913.

UNIVERSITY OF MICHIGAN 3 9015 02229 1085