THE B.C.C. TRANSITION METALS: W, V, Cr, Nb, Mo, Ta AND α-Fe
THE REFRACTORY b.c.c. metals have high melting points and moduli. Many of their applications are specialized ones which exploit these properties: tungsten lamp filaments (Chapter 19) operate at up to 2800°C, and molybdenum furnace windings to 2000°C, for example. They are extensively used as alloying elements in steels and in superalloys, raising not only the yield and creep strengths, but the moduli too. But above all, iron is the basis of all steels and cast irons, and it is this which has generated the enormous scientific interest in the b.c.c. transition metals.
����������� Maps for six pure b.c.c. metals (W, V, Cr, Nb, Mo and Ta) are shown in the Figures of this Chapter. Those for , β-Ti, α-iron and ferrous alloys are discussed separately in Chapters 6 and 8. The maps are based on data plotted on the figures and the parameters listed in Table S.1.
����������� Like the f.c.c. metals, maps for the b.c.c. metals show three principal fields: dislocation glide, power�law creep, and diffusional flow. The principal difference appears at low temperatures (below about 0.15 TM) where the b.c.c. metals exhibit a yield stress which rises rapidly with decreasing temperature, because of a lattice resistance (or Peierls' resistance). As a result, the flow stress of pure b.c.c. metals extrapolates, at 0 K, to a value close to 10-2 μ, independent of obstacle content for all but extreme states of work-hardening. On the other hand, the normalized strength of the b.c.c. metals at high temperatures (>0.5 TM) because diffusion in the more open b.c.c. lattice is faster than in the close�packed f.c.c. lattice.
����������� It is important to distinguish the transition metals from both the b.c.c. alkali metals (Li, K, Na, Cs) and from the b.c.c. rare earths and transuranic metals (γ-La, δ-Ce, γ-Yb, γ-U, ε-Pu, etc.). Each forms an isomechanical group (Chapter 18); members of each group have similar mechanical properties, but the groups differ significantly. The data and maps presented in this Chapter are for the b.c.c. transition metals only, and give no information about the other two groups.
Fig. 5.1. Tungsten of grain size 1 �m. The obstacle spacing is l =2 x 10-7m.
Fig. 5.2. Tungsten of grain size 10 �m.
Fig. 5.3. Tungsten of grain size 100 �m.
Fig. 5.4. Tungsten of grain size 1 mm.
����������� One map is presented for most of the metals, computed for a "typical" grain size of 100 �m. The influence of grain size is illustrated in detail for tungsten by maps for grain sizes of 1, 10, 102 and 103 �m (Figs. 5.1 to 5.4). The other metals follow the same pattern.
����������� Some of the b.c.c. transition metals show large elastic anistropy; then the various averages of the single crystal moduli differ significantly. We have taken the shear modulus at 300 K (μ0) as that appropriate for the anisotropic calculation of the energy of a <111> screw dislocation (Hirth and Lothe, 1968) , and have calculated a linear tempera�ture dependence either from single-crystal data or, when there was none, from polycrystal data. This is adequate for all but niobium which shows anomalous behaviour (Armstrong et al., 1966)  and iron (for which see Chapter 8).
����������� Lattice diffusion has been studied in all these metals and is shown for each as a data plot (below). For some there is evidence that the activation energy decreases at lower temperatures. We have used a dual expression for the lattice diffusion coefficient to describe this "anomalous diffusion" in vanadium; parameters are given in Table 5.1. Similar behaviour has been demonstrated for tanta�lum (Pawel and Lundy, 1965)  but can be well approximated by one simple Arrhenius relationship.
����������� Complete data for grain boundary and dis�location core self-diffusion are available only for tungsten. For chromium, an activation energy for core diffusion has been reported. All other core and boundary diffusion coefficients have been estimated using the approximation Qc ≈ QB ≈ (2Qυ/3) (see Brown and Ashby, 1980 , for an analysis of such correlations).
����������� The parameters describing lattice-resistance con�trolled glide were obtained, when possible, by fitting experimental data to eqn. (2.12). This rate equation gives a good match to experiment; but it must be realized that the values of the parameters ∆Fp �and �depend on the form chosen for ∆G: a different choice (one with the form of that in eqn. (2.9), for instance) gives different values. Further, the set of values for ∆Fp, �and �obtained by fitting eqn. (2.12) to a given batch of data are not, in practice, unique; changing any one slightly still leads to an acceptable fit provided the others are changed to compensate. The set listed here gives a good fit to the data shown in the data plots.
����������� The quantity �is determined approximately by extrapolating flow stress data to 0 K. It is more difficult to determine �experimentally unless the flow stress is known over a wide range of strain rates. The data of Briggs and Campbell (1972)  for molybdenum, and those of Raffo (1969)  for tung�sten, when fitted to eqn. (2.12), are well described by the value = 101l/s, and we have therefore used this for all the b.c.c. transition metals. Once �and are known, ∆Fp is found from the temperature� dependence of the flow stress. Finally, the rate�equation (eqn. (2.12)) is evaluated, and adjustments made to and ∆Fp to give the best fit to the data.
The yield stress of some b.c.c. metals decreases as the metal is made purer. There is debate as to whether the Peierls' stress results from an intrinsic lattice resistance or from small concentrations of interstitial impurities. The question need not concern us here, except that it must be recognized that the yield parameters refer to a particular level of purity.
����������� The critical resolved shear stress of b.c.c. single crystals is related to the polycrystalline shear strength by the Taylor factor, Ms = 1.67 (Chapter 2, Section 2.2). The data plots of this chapter record both; it can be seen that polycrystal data and single �crystal critical resolved shear stress data (for com�parable purities) differ by about this factor, at all temperatures. Table 5.1 records the polycrystal shear strength. As discussed in Chapter 3 (Section 3.2), the lattice resistance-controlled and obstacle�controlled glide are treated as alternative mechan�isms: that leading to the slowest strain rate is controlling. This results in a sharp corner in the strain rate contours. A more complete model of the mechanism interaction would smooth the transi�tion, as the data plots suggest. We have arbitrarily chosen an obstacle spacing of l = 2 x 10-7m (or a dislocation density of ρ = 2.5 x 10-13 /m2). This value is lower than that used for f.c.c. maps, and describes a lower state of work-hardening. It is not necessarily that of the samples recorded in the data plots, although it gives a fair description of most of the data.
����������� Figs. 5.1 to 5.4 show maps for tungsten of grain size 1 �m, 10 �m, 100 �m, and 1 mm, showing the influence of grain size. They are based on the data plotted in Figs. 5.5 to 5.7, and on the parameters listed in Table 5.1.
����������� Lattice diffusion data for tungsten are plotted in Fig. 5.7. There is good agreement between the data plotted as the lines labelled, 3, 5 and 6 (Andelin et al., 1965 ; Robinson and Sherby,1969 ; Kreider and Bruggeman, 1967 ). We have taken the co-efficients cited by Robinson and Sherby (1969)  because they typify these data. Boundary and core diffusion co�efficients are from Kreider and Bruggeman (1967) .
Fig. 5.5. Data for tungsten, divided into blocks. Each block is fitted to a rate-equation. The numbers are log10()
Fig. 5.6. Creep data for tungsten. Data are labelled with the temperature in °C.
Fig. 5.7. Lattice diffusion data for tungsten. The data are from (1) Vasilev and Chernomorchenko (1956); (2) Danneberg(1961) ; (3) Andelin et al. (1965) ; (4) Neumann and Hirschwald (1966); (5) Robinson and Sherby (1969) ; and (6) Kreider and Bruggeman (1967) .
����������� The high-temperature creep of tungsten has been reviewed by Robinson and Sherby (1969) , who demonstrated that most of the available data can be divided into high-temperature creep above 2000°C and low-temperature creep below. The high-temperature data are those of Flagella (1967)  �for wrought arc-cast tungsten � and of King and Sell (1965) .� These data show faster creep rates than those of Green (1959)  and Flagella (1967)  for powder-metallurgy tungsten.� The low temperature creep region is represented by data of Gilbert et al. (1965)� which show �between 1300°C and 1900°C with an apparent activation energy of about 376 kJ/mole (although they are nearly an order of magnitude slower than that of Flagella (1967)  in the overlapping temperature range). This general behaviour is also indicated by other papers.� The low-temperature yield parameters for tungsten are based on the polycrystalline yield data of Raffo (1969)  (see also Chapter 2 and Fig. 2.3).� These data are in general agreement with single-crystal critical resolved shear stress data of Koo (1963)  and argon and Maloof (1966) .
The field of dynamic recrystallization is based on the observations of Glasier et al. (1959)  and of Brodrick and Fritch (1964)  who tested tungsten up to 0.998 TM.
Fig. 5.8 show a map for vandium with a grain size of 0.1 mm. It is based on data shown in Fig. 5.9 and parameters listed in Table 5.1. There are so few creep data for vanadium that no creep data plot is shown.�
Fig. 5.8. Vanadium of grain size 100 �m
Fig. 5.9. Data for vanadium, divided into blocks. The numbers are log10 ().
Fig. 5.10. Lattice diffusion data for vanadium. The data are from (1) Lundy and Mehargue (1965) ; (2) Peart (1965) ; and Agarwala et al. (1968) .
����������� Lattice diffusion data are plotted in Fig. 5.10. It is found that the activation energy decreases from about 390 kJ/mole above 1350°C to about 310 kJ/mole below. We have used the coefficients determined by Peart (1965) , labelled 2 on the figure, as the most reliable.
����������� The dislocation creep parameters are based on Wheeler et al. (1971) , who found that the activation for creep, like that for diffusion, decreased (from 472 to 393 to 318 kJ/mole) with decreasing temperature, though the decrease occurred at a lower temperature. At their lowest temperatures the activation energy dropped to 226 kJ/mole which they ascribed to core diffusion. The stress exponent, n, increased from 5 at high temperatures to 8 at low temperatures in accordance with the expected low� temperature creep behaviour (Chapter 2, Section 2.4).
The lattice-resistance parameters are derived from the high-purity polycrystalline data of Wang and Bainbridge (1972) . They agree well with those derived from single-crystal data (Wang and Brain�bridge, 1972 ; Mitchell et al., 1970 ) when the Taylor factor conversion is included.
����������� There appear to be no observations of dynamic recrystallization in vanadium. The shaded field is that typical of the other b.c.c. metals.
����������� Fig.5.11 shows a map for chromium of grain size 0.1 mm. It is based on the data shown in Fig. 5.12 and the parameters listed in Table 5.1. There are so few measurements of creep of chromium that no creep plot is given.
����������� Lattice diffusion data are plotted in Fig. 5.13. We have chosen the coefficients of Hagel (1962)  (labelled 6) which come closest to describing all the available data.
Fig. 5.11. Chromium with a grain size of 100 �m.
Fig. 5.12. Data for chromium, divided into blocks. The numbers are log10 ()
Fig. 5.13. Lattice diffusion data for chromium. The data are from (1) Gruzin et al. (1959) ; (2) Paxton and Gondolf (1961)  (4) Bogdanov (1960) ; (5) Ivanov et al. (1962) ; (6) Hagel (1962) ; and (7) Askill and Tomlin (1965) .
The dislocation creep parameters are derived from the data of Stephens and Klopp (1972)  who tested high-purity iodide-chromium. Their data at 1316°C to 1149°C show a stress exponent of ��data at 816°C and 982°C show . The lower� temperature data, however, show no tendency toward a lower activation energy. There is, therefore, no conclusive evidence for (or against) a low-temperature creep field in chromium. The parameters describing the lattice resistance are based on the data of Marcinkowski and Lipsitt (1962)  for polycrystals. Twinning is observed in chromium at 88 K.
����������� There are no reports of dynamic recrystallization in chromium. The shaded field is that typical of other b.c.c. metals.
A map for niobium with a grain size of 100 �m, is shown in Fig. 5.14. It is based on data shown in Figs. 5.15 and 5.16 and on the parameters listed in Table 5.1.
The elastic constants of niobium have an anomalous temperature dependence (Armstrong et al., 1966) . Because of this, we have neglected the temperature-dependence of the modulus.
Fig. 5.14. Niobium (columbium) with a grain size of 100 �m.
Fig. 5.15. Data for niobium, divided into blocks. The numbers are log10()
Fig. 5.16. Creep data for niobium. Data are labelled with the temperature in °C.
Fig. 5.17. Lattice diffusion data for niobium. The data are from (1) Resnick and Castleman (1960) ; (2) Peart et al (1962) ; (3) Lundy et al. (1965) ; (4) Lyubinov et al. (1964)  single crystals; and (5) Lyubinov et al. (1964) , polycrystals
����������� Lattice diffusion data are plotted in Fig. 5.17. We have used coefficients derived from the measure�ments of Lundy et al. (1965) , labelled 3, which lie centrally through the remaining data, and cover an exceptionally wide range of temperature.
����������� Creep data for niobium are somewhat limited. The parameters listed in Table 5.1 are based on the measurements of Brunson and Argent (1962) ; these agree well with the data of Stoop and Shahinian (1966)  but not with those of Abramyan et al. (1969) , which show slower strain-rates.
����������� The low-temperature yield behaviour of niobium has been extensively studied for both polycrystals and single crystals. We have derived yield parameters from the data of Briggs and Campbell (1972) , which agree well with earlier studies.
����������� There appear to be no studies above 1400°C. The dynamic recrystallization field is arrived at by analogy with other b.c.c. metals.
Fig. 5.18 shows a map for molybdenum with a grain size of 100 �m. It is based on the data shown in Figs. 5.19 and 5.20 and on the parameters listed in Table 5.1.
Fig. 5.18. Molybdenum with a grain size of 100 �m.
Fig.5.19. Data for molybdenum, divided into blocks. The numbers are log10().
����������� Lattice diffusion data for molybdenum are plotted in Fig. 5.21. There is substantial agreement between all investigators. We have used the parameters derived by Askill and Tomlin (1963)  from studies of polycrystals (labelled 6).
����������� Creep of molybdenum has been well studied. The high-temperature creep parameters are derived from the data of Conway and Flagella (1968)  (which are more than an order of magnitude faster in creep rate than those of Green et al., 1959) , giving an exponent of n = 4.85. Several studies suggest the existence of a low-temperature creep field. Carvalhinos and Argent (1967) , Pugh (1955)  and Semchyshen and Barr (1955)  all found n = 6-8 for T = 0.4-0.53 TM, and an activation energy lower than that for volume diffusion.
Fig. 5.20. Creep data for molybdenum. Data are labelled with the temperature in °C.
Fig. 5.21. Lattice diffusion data for molybdenum. The data are from (1) Borisov et al. (1959) ; (2) Gruzin et al. (1959) ; (3) Brofin et al. (1960) ; (4) Danneburg and Krautz (1961) ; (5) Askill and Tomlin (1963) ; single crystals; (6) Askill and Tomlin (1963) , polycrystals; and (7) Pavlinov and Bikov (1964) .
����������� The low-temperature yield parameters are derived from data of Briggs and Campbell (1972) . These are not the lowest known yield strengths, and therefore do not describe molybdenum of the highest purity. Lawley et al. (1962) , found that the polycrystalline yield stress can be further lowered by nearly a factor of 2 by repeated zone refining, although the change in purity cannot be detected.
����������� Above 2000°C (0.8 TM) molybdenum of commer�cial purity shows dynamic recrystallization (Glasier et al., 1959 ; Hall and Sikora, 1959 ); pure molybdenum does so at a slightly lower temperature. The positioning of the dynamic recrystallization field is based on an average of these observations.
����������� Fig. 5.22 shows a map for tantalum with a grain size of 100 �m. It is based on data plotted in Figs. 5.23 and 5.24 and on the parameters listed in Table 5.1.
����������� Lattice diffusion data are plotted in Fig. 5.25. We have used the parameters of Pawel and Lundy (1965) , whose data are in good agreement with measurements of Eager and Langmuir (1953) , and which cover an exceptionally wide range of temperature.
Fig. 5.22. Tantalum with a grain size of 100 �m.
Fig. 5.23. Data for tantalum, divided into blocks. The numbers are log10 ().
Fig.5.24. Creep data for tantalum. Data are labelled with temperature in °C.
Fig. 5.25. Lattice diffusion data for tantalum. The data are from (1) Eager and Langmuir (1953) , (2) Gruzin and Meshkov (1955) ;� (3) Pawel and Lundy (1965) .
����������� The high-temperature creep parameters for tantalum are taken from Green (1965) . His steady�state data show an activation energy which increases with temperature (as pointed out by Flinn and Gilbert, 1966)� but this may be explained by the fact that the highest temperature tests lie in the field of dynamic recrystallization. The stress exponent is . There is some indication of low-temperature creep behaviour: the data of Schmidt et al. (1960)  at 1000°C and 1200°C show ��The yield para�meters for tantalum are based on data of Wessel as cited by Bechtold et,al. (1961) , and are in good agreement with the single-crystal data of Mitchell and Spitzig (1965) , adjusted by the appropriate Taylor factor (Chapter 2, Section 2.2).
����������� The existence of a field of dynamic recrystal�lization can be inferred from the data reported by Green (1965) , Preston et al. (1958)  and Glasier et al. (1959) . Its position was determined partly from these observations, and partly by analogy with other b.c.c. metals.
1.�������� Hirth, J.P. and J. Lothe, Theory of Dislocations. 1968: McGraw-Hill. 435.
2.�������� Armstrong, P.E., J.M. Dickinson, and H.L. Brown, Temperature dependence of the elastic stiffness coefficients of niobium (columbium). Trans. AIME, 1966. 236: p. 1404.
3.�������� Pawel, R.E. and T.S. Lundy, The diffusion of Nb95 and Ta182 in tantalum. J. Phys. Chem. Solids, 1965. 26: p. 927-42.
4.�������� Brown, A.M. and M.F. Ashby, Correlations for diffusion constants. Acta Metallurgica (pre 1990), 1980a. 28: p. 1085.
5.�������� Briggs, T.L. and J.D. Campbell, The effect of strain rate and temperature on the yield and flow of polycrystalline niobium and molybdenum. Acta Metallurgica (pre 1990), 1972. 20: p. 711.
6.�������� Raffo, P.L., Yielding and fracture in tungsten and tungsten-rhenium alloys. J. Less Common Metals, 1969. 17: p. 133-49.
7.�������� Andelin, R.L., J.D. Knight, and M. Kahn, Trans. AIME, 1965. 233: p. 19.
8.�������� Robinson, S.L. and O.D. Sherby, Mechanical behaviour of polycrystalline tungsten at elevated temperature. Acta Met., 1969. 17: p. 109.
9.�������� Kreider, K.G. and G. Bruggeman, Grain boundary diffusion in tungsten. Trans. AIME, 1967. 239: p. 1222.
10.������ Flagella, P.N. High Temperature Technology. in Third International Symposium- High Temperature Technology. 1967. Asilomar, Calif.
11.������ King, G.W. and H.G. Sell, The effect of thoria on the elevated-temperature tensile properties of recrystallized high-purity tungsten. Trans AIME, 1965. 233: p. 1104.
12.������ Green, W.V., Trans AIME, 1959. 215: p. 1057.
13.������ Koo, R.C., Acta Metallurgica, 1963. 11: p. 1083.
14.������ Argon, A.S. and S.R. Maloof, Acta. Met., 1966. 14: p. 1449.
15.������ Glasier, L.F., R.D. Allen, and I.L. Saldinger, Mechanical and Physical Properties of Refractory Metals. 1959, Aerojet General Corp.
16.������ Brodrick, R.F. and D.J. Fritch. in Proc. ASTM. 1964: ASTM.
17.������ Lundy, T.S. and C.J. Mehargue, Trans. AIME, 1965. 233: p. 243.
18.������ Peart, R.F., Diffusion V48 and Fe59 in vanadium. J. Phys. Chem. Solids, 1965. 26: p. 1853-61.
19.������ Agarwala, R.P., S.P. Murarka, and M.S. Anand, Diffusion of vanadium in niobium. zirconium and vanadium. Acta Metallurgica (pre 1990), 1968. 16: p. 61.
20.������ Wheeler, K.R., et al., Minimum-creep-rate behaviour of polycrystalline vanadium from 0.35 to o.87 Tm. Acta Metallurgica (pre 1990), 1971. 19: p. 21.
21.������ Wang, C.T. and D.W. Bainbridge, The deformation mechanism for high-purity vanadium at low temperatures. Met Trans., 1972. 3: p. 3161.
22.������ Mitchell, T.E., R.J. Fields, and R.L. Smialek, Three-stage hardening in vanadium single crystals. J.� Less-Common Metals, 1970. 20: p. 167-72.
23.������ Hagel, W.C., Self-diffusion in solid chromium. Trans. AIME, 1962. 22: p. 430.
24.������ Stephens, J.R. and W.D. Klopp, High-temperature creep of polycrystalline chromium. J. Less-Common Metals, 1972. 27: p. 87-94.
25.������ Marcinkowski, M.J. and P.A. Lipsitt, The plastic deformation of chromium at low temperatures. Acta Metallurgica (pre 1990), 1962. 10: p. 95.
26.������ Gruzin, P.L., L.V. Pavlinov, and A.D. Tyutyunnik, Izv. Akad Nauk., SSSR Ser. Fiz., l959? 5(5): p. 155.
27.������ Paxton, H.W. and E.G. Gondolf, Rate of self-diffusion in high purity chromium. Arch. Eisenhuttenw., 1959. 30: p. 55.
28.������ Bogdanov, N.A., Russ. Met. Fuels (English translation), 1960. 3: p. 95.
29.������ Ivanov, L.I., et al., Russ. Met. Fuels (English translation), 1962. 2: p. 63,V.
30.������ Askill, J. and D.H. Tomlin, Self-diffusion of chromium. Philosophical Magazine( before 1978), 1965. 11: p. 467.
31.������ Resnick, R. and L.S. Castleman, Self-diffusion of columbia. Trans. AIME, 1960. 218: p. 307.
32.������ Peart, R.F., D. Graharn, and D.H. Tomlin, Tracer diffusion in niobium and molybdenum. Acta� Met., 1962. 10: p. 519.
33.������ Lyubinov, V.D., P.V. Geld, and G.P. Shveykin, Izv. Akad. Nauk. S.S.S.R. Met. i Gorn Delo S, 137; Russ. Met. Mining . 1964?
34.������ Lundy, T.S., et al., Diffision of Nb-95 and Ta-182 in Niobium (Columbium). Trans. AIME, 1965. 233: p. 1533.
35.������ Brunson, G. and B.B. Argent, J. Inst. Met., 1962. 91: p. 293.
36.������ Stoop, J. and P. Shahinian. High Temperature Refractory metals. in AIME Symposium. 1966: Gordon & Breach Science Publishers, N.Y.
37.������ Abramyan, E.A., L.I. Ivanov, and V.A. Yanushkevich, Physical nature of the creep of niobium at elevated temperature. Fiz. Metal. Metalloved, 1969. 28(3): p. 496-500.
38.������ Askill, J. and D.H. Tomlin, Self-diffusion in molybdenum. Philosophical Magazine( before 1978), 1963. 8(90): p. 997.
39.������ Conway, J.B. and P.N. Flagella, Seventh Annual Report�AEC Fuels and Materials Development Program. 1968, GE-NMPO.
40.������ Green, W.V., M.C. Smith, and D.M. Olson, Trans.AIME, 1959. 215: p. 1061.
41.������ Carvalhinhos, H. and B.B. Argent, The creep of molybdenum. J. Inst. Met., 1967. 95: p. 364.
42.������ Pugh, J.W., Trans. ASM, 1955. 47: p. 984.
43.������ Semchyshen, M. and R.Q. Barr, Summary Report� to Office of Naval Research. 1955, Climax Molybdenum Co.
44.������ Borisov, Y.V., P.L. Gruzin, and L.V. Pavlinov, Met. i. Metalloved. Chistykh Metal., 1959. 1: p. 213.
45.������ Brofin, M.B., S.Z. Bokshtein, and A.A. Zhukhovitskii, Zavodsk. Lab., 1960. 26(7): p. 828.
46.������ Danneberg, W. and E. Krautz, Naturforsch., 1961. 16(a): p. 854.
47.������ Pavlinov, L.V. and V.N. Bikov, Fiz. Met. i Metalloved, 1964. 18: p. 459.
48.������ Lawley, A., J. Van den Sype, and R. Maddin, Tensile properties of zone-refined molybdenum in the temperature range 4.2-373o C. 1962. 91: p. 23.
49.������ Hall, R.W. and P.F. Sikora, Tensile Properties of Molybdenum and Tungsten from 2500 to 3700 F. 1959, NADA.
50.������ Eager, R.L. and D.B. Langmuir, Self-Diffusion of Tantalum. Physics Review, 1953. 89: p. 911.
51.������ Gruzin, P.L.a. and V.I. Meshkov, Fopr. Fiz. Met. i Metallolved., Sb. Nauchn. Rabot. Inst. Metallofiz. Akad. Nauk. Ukr., S.S.R. 570, AEC-tr-2926. 1955? Report.
52.������ Green, W.V., High temperature creep of tantalum. Trans. AIME, 1965. 233: p. 1818.
53.������ Schmidt, F.F., et al. 1960, Batelle Memorial Inst.
54.������ Wessel, E.T., L.L. France, and R.T. Begley. Columbium Metallurgy. in Proc. of AIME Symposium,. 1961. Boulton Landing. Interscience, N.Y.: AIME.
55.������ Mitchell, T.E. and W.A. Spitzig, Three-stage hardening in tantalum single crystals. Acta Metallurgica (pre 1990), 1965. 13: p. 1169.
56.������ Preston, J.B., W.P. Roo, and J.R. Kattus, Technical Report. 1958, Southern Research Inst.