The following article appears in the journal JOM, 49 (8) (1997), pp. 35-39. JOM is a publication of The Minerals, Metals & Materials Society

The Mechanical Properties of In-Situ Composites

INTRODUCTION

There are a number of processes and classes of in-situ composites, including processes where the precipitation of reinforcements occurs in the liquid state and processes where precipitation occurs in the solid state.¹ Processes in which precipitation occurs in the liquid state include those occurring with low undercoolings, such as traditional and directionally solidified eutectics; rapid solidification in which a supersaturation is created by rapidly cooling a liquid with the reinforcement constituents already in solution; and processes in which a large supersaturation is created by rapidly bringing the constituents of the reinforcement together to react. These latter processes include the XD^TM,²,⁴ Mixalloy,⁵ reactive-gas injection (Drexal process),^6,7 reactive sintering,⁸ reactive hot isostatic pressing (HIPing), and plasma conversion⁹ processes. Solid-state in-situ processes include internal oxidization,¹⁰ displacement reactions,¹¹ reactive milling, mechanical alloying, and cryomilling.¹² There are also traditional eutectoid reactions that can be considered solid-state in-situ reactions if an alloy of ferrite and cementite is considered a composite.

Generally, the reinforcements in discontinuously reinforced metallic- or intermetallic-matrix in-situ composites are on the order of 0.5-5 µm, and volume fractions range from 0-50 vol.%. The potential advantages of in-situ composites as compared to discontinuous metal-ceramic composites produced by ex-situ methods include:

• Smaller reinforcement particle size with higher strength (a contribution from composite-strengthening mechanisms) and improved fatigue resistance and creep
• Small, single-crystal reinforcements (lower propensity for particle fracture)
• Clean, unoxidized particle-matrix interfaces with higher interfacial strength (higher ductility and toughness) and improved wettability
• Thermodynamically stable particles that are weldable and castable, do not dissolve at high temperatures (vis-à-vis age-hardened alloys), and do not have a reaction layer (higher interfacial strength, improved corrosion, and long-term stability)
• Better particle-size distribution (improved mechanical properties)
• More conventional processing with the potential for lower cost and production with conventional equipment

STRENGTHENING MECHANISMS

Several possible explanations for increases in yield strength of discontinuously reinforced composites with equiaxed particles in the range 0.5-5 µm in diameter have been proposed in recent reviews.13,14 Continuum Models Continuum-mechanics models of composite behavior assume that load is transferred from the matrix to a nonplastically deforming reinforcement. One such model, shear-lag theory, assumes that load transfer occurs between a high-aspect-ratio reinforcement and the matrix by means of shear stresses at the particle-matrix interface.15 According to a modified shear-lag theory, which also accounts for load transfer at the particle ends,16,17 the yield stress of the composite y is y = ymf + ym (1-f) (A) where ym is the yield stress of the unreinforced matrix, L is the length of the particle perpendicular to the applied stress, t is the length of the particle parallel to the applied stress, A is the particle-aspect ratio, and f is the particle-volume fraction. For the equiaxed particles present in most in-situ composites, Equation A reduces to y = ym (B) It is important to note that this form of the equation is linear, with a very modest increase in y with increasing volume fraction, and there is no dependence of yield stress on particle size or microstructural scale. Finite-element solutions to the elastic-plastic continuum model for the problem of regularly spaced spheres have been obtained for two dimensions18 and three dimensions.19 These models yield similar results to the shear-lag theory at low volume fractions, but the elastic-plastic models predict a greater increase in strength above about 20 vol.%. Dislocation Density Effects A great deal of work has been performed on aluminum alloys reinforced ex-situ by SiC particles, plates, and whiskers. In many of these composites, dislocations are generated in the alloy matrix upon cooling or quenching from the processing or solutionizing temperature, due to a mismatch of the coefficient of thermal expansion (CTE) between the matrix and SiC reinforcement.20,21 The higher dislocation density increases the strength of the alloy.22,23 The amount of dislocation generation is affected by CTE, particle size, particle-volume fraction, and matrix strength.24 Single Dislocation-Particle Interaction Model: Orowan Strengthening Gliding dislocation can bow between small impenetrable particles and bypass them, leaving behind an Orowan loop. The effect of this process on the yield stress of the material is described by the Orowan relationship. For polycrystalline materials, this is y = (C) where M is the Taylor factor, µ is the shear modulus, b is the Burgers vector, v is the Poisson's ratio, r is the effective radius of the particles, ro is the cut-off radius, and is the mean interparticle separation on the slip plane.25 Multiple Dislocation-Particle Interaction Models Dislocation Pile-Up Models The yield stress of single-phase polycrystalline materials is phenomenologically observed to be proportional to d-1/2, where d is the grain diameter. This leads to the relation y = o + (D) known as the Hall-Petch relationship, which arose from experimental observations. Typically o is rationalized as either a frictional stress resisting the motion of gliding dislocations or as an internal back stress. ky is the Hall-Petch slope, which is considered to be a measure of the resistance of the grain boundary to slip transfer. Typical values of ky at room temperature are on the order of 0.05-0.1 MPa for face-centered cubic (f.c.c.) metals and 0.3-1.8 MPa for body-centered cubic (b.c.c.) metals. Although most studies have been carried out at ambient temperature, results indicate that the Hall-Petch relationship can also be used at elevated temperatures26 where ky decreases with increasing temperature. The same d-1/2 dependence of the yield stress has been observed in two-phase alloys where d is the characteristic microstructural dimension (i.e., the dimension that governs slip distance). Examples include drawn ferrite, where d is the dislocation cell size; drawn pearlite, where d is the interlamellar spacing of carbide; and polygonized aluminum, where d is the cell size. A very informative example is that of the competitive effects of cementite particles, grain boundaries, and subgrain boundaries on the behavior of spheroidized cementite in a ferrite matrix as analyzed by Anand and Gurland.27 Another example is the strengthening of deformation-processed composites such as drawn or rolled Cu-Nb, where the strength increase has been attributed to direct blocking of the dislocation motion by the interphase boundaries leading to a -1/2 relationship, where is the interphase separation.28-30 If the grain boundaries or subgrain boundaries are obstacles to dislocation motion, then the grain or subgrain size is the characteristic microstructural dimension. However, if particles are obstacles to dislocation motion, then the interparticle separation is the characteristic microstructural dimension. In either case, when dislocation pile-up is the operative strengthening mechanism, an inverse square root dependence on the characteristic microstructural dimension will be observed. When grain size and interparticle spacing have similar dimensions, dislocation pile-ups at both particle-matrix interfaces and grain boundaries may contribute. In these situations, the ease of dislocation annihilation or generation at a particular interface is at issue. Thus, the behavior of the interphase barrier may be as important as the spacing of these interfaces. Forest Hardening Models Another effect that dislocations may have is their role on work-hardening behavior at small strains. Kamat et al.31 have examined the tensile properties of 2024 and 2014 aluminum alloys reinforced with equiaxed alumina particles and suggest that dislocation tangles form around particles at small strains because of plastic incompatibility. These geometrically necessary dislocations link up to form dislocation cells with a diameter proportional to the interparticle spacing . They suggest that this should lead to a flow stress = (E) where is a proportionality constant. Thus, it is possible in some composites that the reinforcement of the matrix does not significantly alter the onset of plastic deformation. Instead, the observed increase in the 0.2% offset yield stress is due to an increase in the work-hardening rate at small strains.13 In these alloys, a -1 dependence of the 0.2% offset yield stress should be observed.

These benefits may not be realized for all systems, but tend to be very system and property specific.

In-situ composites also have disadvantages. The synthesis of an in-situ reinforcement is typically by a precipitation process, and, as such, the choice of in-situ reinforcements is limited to particles that are thermodynamically stable within a particular matrix. The size and shape of particles are controlled by nucleation and growth processes, so although significant flexibility may be available during processing and synthesis, the size and shape are ultimately determined by the kinetics of the system instead of the ex-situ synthesis and processing of the reinforcements. Although particle distribution may be more uniform, the potential also exists for the pushing of fine particles during solidification.

STRENGTHENING

Aluminum Matrix Systems

The proportional limit and 0.2% offset yield stress of Al-4Cu-1.5Mg composites containing 0, 2, 5, 10, and 15 vol.% TiB₂ at two average particle diameters are shown in Figure 2 for tensile tests conducted at room temperature. These alloys were extruded, solutionized at 493°C, water quenched, and then allowed to naturally age at room temperature. This heat treatment produces a dispersion of very fine Guinier-Preston (GP) zones that are sheared by the movement of dislocations such that the matrix deforms by planar slip. The yield stress increases with increasing particle-volume fraction. At any given particle-volume fraction, the composites with the small particles exhibit higher strength than the composites with larger particles. Note that the yield stress and the proportional limit increase at a similar rate with increasing volume fraction of TiB₂. Thus, the observed increase in strength is due to an increase in the initial resistance to the movement of dislocations and not a difference in the initial strain-hardening rate.

The applicability of the various characteristic slip-length models can be examined by plotting the yield stress against the various microstructural dimensions. The grain size for these alloys had been experimentally measured and is larger than the interparticle spacing by a factor of 3-5, depending on the alloy. In Figure 3, the yield stress is plotted against the inverse square root of the measured interparticle spacing, ^-1/2, with a least-square fit of the data.

The slope obtained from a similar fit against ^-1 yields a slope that differs from the Orowan slope predicted by Equation C in the sidebar by several orders of magnitude; thus, the observed dependence on particle spacing is not due to Orowan strengthening. Since particles of different size produce different amounts of strengthening, continuum models can be eliminated since they can not account for the observed particle-size effect. Further, transmission electron microscopy examination showed similar dislocation densities for reinforced and unreinforced alloys; therefore, differences in dislocation density are not responsible for the observed strengthening. Shear-lag theory for these equiaxed particles (Equation B) predicts an increase in the yield stress that is significantly smaller than what is observed. Hence, the data suggest that the strength increase in these composites is due to dislocation pile-ups through an interphase barrier-strengthening mechanism.

a

b

Figure 1. (a) Tensile yield stress and (b) elongation to fracture versus test temperature for composites produced by the XDTM process with various volume fractions of TiB2 with a Al-4Cu-1.5Mg matrix and 1.3 µm average particle size. Alloys are in peak-aged (T6) temper following exposure to an elevated temperature for 30 minutes and tested at a strain rate of 1.3 X 10-3s-1.

Figure 2. A variation of tensile yield stress and proportional limit with volume percent for Al-4Cu-1.5Mg composites reinforced with TiB2 with either 0.3 µm or 1.3 µm average diameter. The matrix is in the naturally aged (T4) condition.

Figure 3. Tensile yield stress as a function of the inverse square root of the interparticle separation for the Al-4Cu-1.5Mg composites shown in Figure 2.

Figure 4. Compressive yield stress versus test temperature for MoSi2 composites produced by the XDTM process with 30 vol.% of various reinforcement types. Specimens were tested in air at a strain rate of 7 X 10-5 s-1.

Figure 1, which shows these same Al-4Cu-1.5Mg/TiB₂ composites with the matrix in the T6 temper, reveals several important trends. The rapid drop in yield stress regardless of particle loading at about 200°C indicates that the mechanisms of strengthening in the reinforced alloys can be defeated by thermally activated processes. This is consistent with the mechanism of interphase-barrier strengthening. At high temperatures, dislocations could cross-slip, and the strength increase observed at low temperatures due to dislocation pile-up would be defeated. It is also possible that the grain boundaries and particle interfaces act as sinks for dislocations at high temperatures, further reducing the dislocation pile-ups. The annihilation of dislocation at grain and interphase boundaries would account for the apparent cross-over in strength as a function of loading that was observed at 350°C.

Sahoo and Koczak examined a series of Al-4.5Cu alloys reinforced with 0-11 vol.% TiC produced by reactive gas injection.³² They found increases in tensile yield and ultimate tensile strength on the order of 200 MPa, significantly greater than what would be predicted by the observed reduction in grain size. These results are consistent with the interphase-barrier strengthening.

Although the operative strengthening mechanisms have not been identified, other discontinuously reinforced aluminum-matrix in-situ composites for which strength and tensile data have been published include Al-5Mg and Al-5Cu reinforced with 5 vol.% or 10 vol.% TiC,³³ aluminum reinforced with 20 vol.% TiB_2, ³⁴ aluminum simultaneously reinforced with Al₂O₃ and TiB₂,³⁵ Al-5.8Cu-0.2Mg-0.2Mn (2519) reinforced with 4.4 vol.% or 6.4 vol.% TiC,³⁶ and Al-6.4Cu-1.4Li-0.3Ag-0.4Mg-0.15Zr (2195) reinforced with 4 vol.% TiB₂.³⁷ The compressive and flexural properties of aluminum reinforced with 30 vol.% TiB₂, TiC, or a mixture of TiB₂ and TiC have also been examined.³⁸

MoSi₂ Matrix Systems

In Figure 5, the compressive yield stresses of the 13 composites containing 15, 30, or 45 vol.% of the indicated particles are plotted versus the inverse square root of the measured planar interparticle spacing for test temperatures of 1,000°C and 1,200°C. For the unreinforced alloy, the grain size is taken as the effective slip distance and is used instead of in these plots. Values of k_y ranged from 2.8 MPa at 800° to 1.4 MPa at 1,200°C. These values are somewhat higher than the k_y values of 0.05-1.3 MPa determined for conventional metals at room temperature, but are consistent with particles acting as strong obstacles and with the difficulty of renucleating slip on the other side of a nondeforming particle. Such a high k_y value would also be expected in materials with a limited slip number of systems such as MoSi₂. Values of s_o decreased from 280 MPa at 800°C to 50 MPa at 1,400°C. Although a dislocation-strengthening model is applied to MoSi₂ below its ductile-to-brittle transition temperature of approximately 1,300°C,⁴¹ a dislocation-based strengthening argument is not inconsistent, as there is ample evidence of dislocation activity below 1,300°C.^42,43

The fact that the ZrB₂ and HfB₂ particles are plate shaped while the TiB₂ and SiC are nearly equiaxed particles is accounted for through the use of the experimentally measured . It should be noted that as the aspect ratio of the reinforcements becomes very large, load transfer would be expected to become a significant strengthening mechanism. The contribution of strengthening due to load transfer would overwhelm the role of dislocation slip in the matrix only when the aspect ratio becomes extremely large.

Flexural stress-strain curves for MoSi₂ reinforced with 30 vol.% SiC, 30 vol.% TiB₂, or 30 vol.% HfB₂ and an unreinforced MoSi₂ sample are shown in Figure 6.⁴⁴ The origins of neighboring curves have been displaced for clarity. Although not shown, the behavior of the alloys at 1,000°C is nearly identical to the behavior exhibited at 1,200°C. At 1,200°C, the stress-strain behavior is linearly elastic with fracture prior to yield. At 1,300°C, the TiB₂ and HfB₂ alloys undergo approximately 0.2% outer-fiber plastic strains prior to fracture, while the SiC-reinforced alloy exhibits about 1.0% outer-fiber plastic strains. The unreinforced alloy fractures prior to any plasticity at 1,300°C. At 1,400°C, all four alloys exhibit significant ductility and do not fracture prior to termination of the test. The apparent ductile-to-brittle transition temperature for the reinforced and unreinforced samples was approximately 1,300°C. To a first approximation, this transition temperature appears to be independent of particle type but is dependent on the MoSi₂ matrix grain size. These tests not only illustrate the effect of reinforcements on strength, but also their effect on the work-hardening rate of the composite.

The flexural strength of in-situ MoSi₂ composites reinforced with SiC produced by a solid-state displacement reaction have also been examined.¹¹,⁴⁵ For composites with between 30-67 vol.% SiC, larger SiC particles resulted in higher strength at elevated temperature, while a higher SiC volume fraction resulted in reduced strength. These trends were attributed to the SiC phase becoming continuous in these composites at large volume-particle sizes and high volume fractions.

TiAl and Other Aluminides

a

b

Figure 5. The elevated temperature compressive yield stress as a function of the inverse square root of the interparticle separation for MoSi2 composites produced by the XDTM process at (a) 1,000°C and (b) 1,200°C. Composites include a variety of particle types with reinforcement loadings of 15 vol.%, 30 vol.%, or 45 vol.%.

Near- Ti-Al alloys were also in-situ-reinforced with 40 vol.% and 50 vol.% TiB₂.⁴⁸ The strength of these alloys, with corrections to account for alloy chemistry, were found to lie on the same strength versus inverse square root of the interparticle spacing curve as the 7% and 12% alloys.

Other discontinuously reinforced intermetallic-matrix in-situ composites for which strength data have been published include Ni-Al reinforced with 20 vol.% TiC⁴⁹ and Nb-26Ti-48Al (at.%) reinforced with borides and carbides.⁵⁰

HIGH-TEMPERATURE STRENGTH AND STABILITY

In-situ copper-matrix composites reinforced with group IV borides and carbides have been produced by the XD^TM process. The tensile yield stress of copper reinforced in-situ with between 7.5 vol.% and 20 vol.% ZrB₂ is shown in Figure 7. These results are very similar to those shown for the Al-4Cu-1.5Mg/TiB₂ composites shown in Figure 1; in creases in reinforcement volume fraction provide increased strength. The elongation to fracture at room temperature of these composites range from 12% for the 7.5 vol.% composite down to 2% for the 20 vol.% composite, demonstrating the excellent thermal stability of this material. Fracture occurs by void nucleation at the particle/matrix interface, which is then followed by void growth and coalescence.

Samples of Cu-15ZrB₂ (vol.%) were heat treated in vacuum at temperatures 68-93% of the melting point for times up to 100 hours. These temperatures are in far excess of any likely use temperature. The room-temperature yield stress of this material as a function of time at temperature is shown in Figure 8. At the highest aging temperatures, there is a drop in the yield strength during the initial few hours, but after that the strength (and, by inference, the microstructure) remain relatively unchanged. The elongation to failure of these same materials showed relatively little change for the 650°C heat treatment and a modest increase from 2% up to 3% for the material heat treated at 980°C.

FRACTURE TOUGHNESS

Figure 6. The stress-strain curves for XDTM MoSi2 composites at elevated temperature in a four-point bend test. Samples were tested in air at a strain rate of 2 X 10-5 s-1.

K_IC = k       (1)

where E is the Young's modulus, _y is the yield stress, the interparticle spacing, and k depends on the nature of void initiation. Assuming a ductile void growth model, k is equal to the particle diameter divided by the interparticle spacing.

Kumar, Mannan, and Viswanadham examined the influence of TiB₂ on the toughness of NiAl.⁵⁴ They found that the addition of TiB₂ particles to stoichiometric NiAl did not significantly affect fracture toughness, while the addition of TiB₂ to nickel-rich NiAl degraded toughness. Another in-situ composite examined is MoSi₂ with SiC.¹¹

FATIGUE

Figure 7. Tensile yield stress versus test temperature for copper composites produced by the XDTM process with various volume fractions of ZrB2.

Figure 8. Room-temperature tensile yield stress of Cu-15ZrB2 (vol.%) composites heat treated at various temperatures as a function of time at temperature. The material shows very good thermal stability and a low coarsening rate of ZrB2.

CREEP

There have been a number of studies that have examined the creep behavior of in-situ-reinforced intermetallic-matrix composites. These include near- Ti-Al reinforced with TiB₂^58-60 or a mix of TiB₂ and Ti₂AlN,⁵⁹ NiAl with TiB₂,⁶¹ NiAl with AlN,^62-64 and MoSi₂ with SiC.^65,66 In general, reducing the grain size of the matrix contributes to higher creep rates, while increasing the volume fraction of the reinforcement contributes to lower creep rates. For most of these intermetallic-matrix composites, the presence of the reinforcing particles results in small grain sizes (on the order of the interparticle spacing), and creep is degraded when compared to an unreinforced material. Creep was improved by the presence of the particles in the aluminum composites that Krajewski et. al examined, which had a relatively large grain size. Thus, the key to good creep behavior for these intermetallic-matrix composites would appear to not only be to provide reinforcing phases, but to also process in order to obtain large grain sizes (something with which the reinforcing particles will interfere).

CONCLUSION

References

1. S.G. Fishman, In-Situ Composites: Science and Technology, ed. M. Singh and D. Lewis (Warrendale, PA: TMS, 1994), p. 1.
2. J.M. Brupbacher, L. Christodoulou, and D.C. Nagle, U.S. patent 4,710,348 (1987).
3. L. Christodoulou, D.C. Nagle, and J.M. Brupbacher, U.S. patent 4,774,052 (1988).
4. D.C. Nagle, J.M. Brupbacher, and L. Christodoulou, U.S. patent 4,916,029 (1990).
5. N.P. Suh, U.S. patent 4,278,622 (1981).
6. M.J. Koczak and K.S. Kumar, U.S. patent 4,808,372 (1989).
7. V. Shtessel, S. Sampath, and M. Koczak, In-Situ Composites: Science and Technology, ed. M. Singh and D. Lewis (Warrendale, PA: TMS, 1994), p. 37.
8. S.D. Dunmead et al., U.S. patent 4,909,842 (1990).
9. P.R. Taylor and M. Manrique, Processing and Fabrication of Advanced Materials IV, ed. T.S. Srivatsan and J.J. Moore (Warrendale, PA: TMS, 1996), p. 827.
10. C.R. Cupp, Progress in Metal Physics, 4 (1953), p. 151.
11. C.H. Henager, Jr., J.L. Brimhall, and L.N. Brush, Mater. Sci. Eng., A195 (1995), p. 65.
12. M.J. Luton et al., Multicomponent Ultrafine Microstructures, Mater. Res. Soc. Symp. Proc., 132, ed. L.E. McCandis et al. (Pittsburg, PA: MRS, 1989), p. 79.
13. R.M. Aikin, Jr., and L. Christodoulou, Scripta Metall., 25 (1991), p. 9.
14. M. Taya, Mater. Trans. JIM, 32 (1991), p. 1.
15. A. Kelly, Strong Solids (Oxford, U.K.: Clarendon Press, 1966), p. 123.
16. V.C. Nardone and K.M. Prewo, Scripta Metall., 20 (1986), p. 43.
17. V.C. Nardone, Scripta Metall., 21 (1987), p. 1313.
18. G. Bao, J.W. Hutchinson, and R.M. McMeeking, Acta Metall. Mater., 39 (1991), p. 1871.
19. C.L. Hom and R.M. McMeeking, Int. J. Plasticity, 7 (1991), p. 255.
20. R.J. Arsenault and R.M. Fisher, Scripta Metall., 17(1983), p. 67.
21. Y. Flom and R.J. Arsenault, Mater. Sci. Eng., 75 (1985), p. 151.
22. R.J. Arsenault and N. Shi, Mater. Sci. Eng, 81 (1986), p. 175.
23. M. Volgelsang, R.J. Arsenault, and R.M. Fisher, Metall. Trans. A, 17 (1986), p. 379.
24. C.T. Kim, J.K. Lee, and M.R. Plichta, Metall. Trans. A, 21 (1990), p. 673.
25. P.M. Kelly, Int. Metall. Rev., 18 (1973), p. 31.
26. M.E. Kassner and X. Li, Scripta Metall., 25 (1991), p. 2833.
27. L. Anand and J. Gurland, Metall. Trans. A, 7 (1976), p. 191.
28. G. Wassermann, Proc. 2nd Int. Conf. on the Strength of Metals and Alloys (Metals Park, OH: ASM, 1970), p. 1188.
29. P.D. Funkenbusch and T.H. Courtney, Scripta Metall., 15 (1981), p. 1349.
30. W.A. Spitzig, A.R. Pelton, and F.C. Laabs, Acta Metall., 35 (1987), p. 2427.
31. S.V. Kamat, J.P. Hirth, and R. Mehrabian, Acta Metall., 37 (1989), p. 2395.
32. P. Sahoo and M.J. Koczak, Mater. Sci. Eng., A131 (1991), p. 69.
33. H. Nakata, T. Choh, and N. Kanetake, J. Mater. Sci., 30 (1995), p. 1719.
34. A.K. Kuruvilla et al., Sripta Metall., 24 (1990), p. 873.
35. Z.Y. Ma et al., Scripta Metall., 31 (1994), p. 635.
36. M.K. Premkumar and M.G. Chu, Mater. Sci. Eng., A202 (1995), p. 172.
37. T.J. Langan and J.R. Pickens, Scripta Metall., 25 (1991), p. 1587.
38. I. Gotman, M.J. Koczak, and E. Shtessel, Mater. Sci. Eng., A187 (1994), p. 189.
39. R.M. Aikin, Jr., Ceram. Eng. Sci. Proc., 12 (1991), p. 1643.
40. R.M. Aikin, Jr., Mater. Sci. Eng., A155 (1992), p. 121.
41. R.M. Aikin, Jr., Scripta Metall., 26 (1992), p. 1025.
42. S.A. Maloy et al., Acta Metall. Mater., 40 (1992), p. 3159.
43. S.A. Maloy, T.E. Mitchell, and A.H. Heuer, Acta Metall. Mater., 43 (1995), p. 657.
44. R.M. Aikin, Jr., International Symp. on Structural Intermetallics, ed. R. Darolia et al. (Warrendale, PA: TMS, 1993), p. 791.
45. C.H. Henager, Jr., J.L. Brimhall, and J.P. Hirth, Mater. Sci. Eng., A155 (1992), p. 109.
46. S.L. Kampe et al., Metall. Trans. A, 25 (1994), p. 2181.
47. S.L. Kampe et al., Intermetallic Matrix Composites III, Mat. Res. Soc. Symp. Poc., 350, ed. J.A. Graves, R.R. Bowman, and J.J. Lewandowski (Pittsburgh, PA: MRS, 1994), p. 159.
48. M.L. VanMeter, S.L. Kampe, and L. Christodoulou, Scripta Met., 34 (1996), p. 1251.
49. Z.P. Xing et al., Metall. Trans. A, 28 (1997), p. 1079.
50. R.M. Aikin, Jr., P.E. McCubbin, and L. Christodoulou, Intermetallic Matrix Composites, MRS Symp. Proc., 194, ed. D.L. Anton et al. (Pittsburgh, PA: MRS, 1990), p. 307.
51. N.C. Beck Tan, R.M. Aikin, Jr., and R.M. Briber, Metall. Trans. A, 25 (1994), p. 2461.
52. J.R. Rice and M.A. Johnson, Inelastic Behavior of Solids, ed. M.F. Kannine et al. (New York: McGraw-Hill, 1980), p. 641.
53. G.T. Hahn and A.R. Rosenfield, Metall. Trans. A, 6 (1975), p. 653.
54. K.S. Kumar, S.K. Mannan, and R.K. Viswanadham, Acta Metall., 40 (1992), p. 1201.
55. G.M. Vyletel, D.C. Van Aken, and J.E. Allison, Scripta Metall., 25 (1991), p. 2405.
56. P.E. Krajewski, J.W. Jones, and J.E. Allison, Metall. Trans. A, 26 (1995), p. 3107.
57. P.E. Krajewski, J.E. Allison, and J.W. Jones, Metall. Trans. A, 24 (1993), p. 2731.
58. S.L. Kampe, J.K. Bryant, and L. Christodoulou, Metall. Trans. A, 22 (1991), p. 447.
59. P.L. Martin et al., Proc. Fourth Int. Conf. On Creep and Fracture of Engineering Materials and Structures (London: IOM, 1990), p. 265.
60. K. Sadananda et al., International Symp. on Structural Intermetallics, ed. R. Darolia et al. (Warrendale, PA: TMS, 1993), p. 809.
61. J.D. Whittenberger et al., J. Mater. Sci., 25 (1990), p. 35.
62. J.D. Whittenberger, E. Arzt, and M.J. Luton, J. Mater. Res., 5 (1990), p. 2819.
63. J.D. Whittenberger, E. Arzt, and M.J. Luton, Scripta Metall., 26 (1992), p. 1925.
64. T.R. Bieler, J.D. Whittenberger, and M.J. Luton, High-Temperature Ordered Intermetallic Alloys V, Mat. Res. Soc. Symp. Proc. Vol. 288 (Pittsburgh, PA: MRS, 1993), p. 1149.
65. M. Suzuki, S.R. Nutt, and R.M. Aikin, Jr., Intermetallic Matrix Composites II, MRS Symp. Proc., 273, ed. D.B. Miracle, D.L. Anton, and J.A. Graves (Pittsburgh, PA: MRS, 1992), p. 267.
66. M. Suzuki, S.R. Nutt, and R.M. Aikin, Jr., Mater. Sci. Eng., 162 (1993), p. 73.

For more information, contact R.M. Aikin, Jr., Department of Materials Science and Engineering, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, Ohio 44106; telephone (216) 368-4221; fax (216) 368-3209; e-mail: rxa19@po.cwru.edu.