METALLURGICAL AND MATERIALS TRANSACTIONS A
	ABSTRACTS Volume 27A, No. 12, December 1996 This Month Featuring: Transformations; Transport Phenomena; Mechanical Behavior; Environment; Welding & Joining; Surface Treatment; Solidification; Materials Processing; Composite Materials. View December 1996 Contents.

TRANSFORMATIONS

The Deposition of Aluminide and Silicide Coatings on -TiAl Using the Halide-Activated Pack Cementation Method
T.C. MUNRO and B. GLEESON
The halide-activated pack cementation method (HAPC) was utilized to deposit aluminide and silicide coatings on nominally stoichiometric -TiAl. The deposition temperature was 1000°C and deposition times ranged from 2 to 12 hours. The growth rates of the coatings were diffusion controlled, with the rate of aluminide growth being about a factor of 2 greater than that of silicide growth. The aluminide coating was inward growing and consisted of a thick, uniform outer layer of TiAl₃ and a thin inner layer of TiAl₂, with the rate-controlling step being the diffusion of aluminum from the pack into the substrate. Annealing experiments at 1100°C showed that the interdiffusion between the aluminide coating and the -TiAl substrate was rapid. In contrast to the aluminide coating, the silicide coating was nonuniform and porous, consisting primarily of TiSi₂, TiSi, and Ti₅Si₄, with the rate-controlling step for the coating growth believed to be the diffusion of aluminum into the -TiAl ahead of the silicide/-TiAl interface. The microstructural evolution of the aluminide and silicide coating structures is discussed qualitatively.

TRANSPORT PHENOMENA

MECHANICAL BEHAVIOR

Plastic Zone and Pileup around Large Indentations
D.F. BAHR and W.W. GERBERICH
Mechanical properties of cold-worked molybdenum, grade 4 titanium, and an - titanium alloy are measured with tensile tests and by indentations using conical indenters with 105, 120, and 137 deg included angles. The extent of plastic deformation and pileup around an indentation is measured using profilometry. Various models predicting the extent of plastic deformation and pileup are compared to the actual measured values. As inferred from indentation, the calculated yield strength of the material from the mean pressure does not correlate well to the yield strength measured by tensile testing. The plastic zone size surrounding an indentation can also be used to determine the yield strength of the material, and this does correlate to the yield strength measured by tensile tests. Furthermore, the extent of plastic deformation is relatively independent of the included angle of the indenter for the range of materials used in this system. Models predicting the amount of pileup at the edges of the indentation appear to approach but overestimate the actual amount of pileup in the materials tested.

The Balance of Mechanical and Environmental Properties of a Multielement Niobium-Niobium Silicide-Based In Situ Composite
B.P. BEWLAY, M.R. JACKSON, and H.A. LIPSITT
This article describes room-temperature and high-temperature mechanical properties, as well as oxidation behavior, of a niobium-niobium silicide based in situ composite directionally solidified from a Nb-Ti-Hf-Cr-Al-Si alloy. Room-temperature fracture toughness, high-temperature tensile strength (up to 1200°C), and tensile creep rupture (1100°C) data are described. The composite shows an excellent balance of high- and low-temperature mechanical properties with promising environmental resistance at temperatures above 1000°C. The composite microstructures and phase chemistries are also described. Samples were prepared using directional solidification in order to generate an aligned composite of a Nb-based solid solution with Nb₃Si- and Nb₅Si₃-type silicides. The high-temperature mechanical properties and oxidation behavior are also compared with the most recent Ni-based superalloys. This composite represents an excellent basis for the development of advanced Nb-based intermetallic matrix composites that offer improved properties over Ni-based superalloys at temperatures in excess of 1000°C.

Effects of the Alumina Scale on the Room-Temperature Tensile Behavior of Preoxidized MA956
J. CHAO, J.L. GONZÁLEZ-CARRASCO, J. IBÁÑEZ , M.L. ESCUDERO, and G. GONZÁLEZ-DONCEL
This article deals with the effects of the -Al₂O₃ scale (~5 µm) developed during preoxidation (1100°C/100 hours) of MA 956 on its room-temperature tensile behavior. The tensile tests were made in the strain-rate range of 10^-5 to 10^-1s^-1. It is shown that the scale, fine and tightly adherent to the substrate, affects the tensile behavior in two relevant ways. First, the yield strength and the tensile strength are lowered with respect to those of the scale-free material. This is explained in terms of the residual stresses generated in the scale during preoxidation. From the analysis of the differences in the yield strength of preoxidized MA956 with respect to the scale-free material, residual compression stresses in the scale of about 5500 MPa were obtained. These high stresses account for the surprisingly high tensile strain achieved (1.4 pct) before scale spallation occurs. Second, a ductile to brittle transition (DBT), which is not observed in the scale-free samples, occurs at intermediate strain rates (10^-3s^-1). The brittle fracture is related to the increase of the triaxiality state in the substrate near the scale/metal interface.

Shear Ligament Phenomena in Fe₃Al Intermetallics and Micromechanics of Shear Ligament Toughening
H. CHIU and X. MAO
The environment-assisted cracking behavior of a Fe₃Al intermetallic in an air moisture environment was studied. At room temperature, tensile ductility was found to be increased with strain rate, from 10.1 pct at 1x10^-6 s^-1 to 14.3 pct at 2x10^-3 s^-1. When tensile tests were done in heat-treated mineral oil on specimens that have been heated in the oil for 4 hours at 200°C, ductility was found to be recovered. These results suggest the existence of hydrogen embrittlement. Shear ligaments, which are ligament-like structures connected between microcracks, were observed on the tensile specimens. They undergo ductile fracture by shearing and enhance fracture toughness. This toughness enhancement (represented by J_l) was estimated by a micromechanical model. The values of the unknown parameters, which are the average ligament length l, the area fraction V, and the work-to-fracture ₁₁, were obtained from scanning electron microscopy (SEM) observation. The total fracture toughness K_c and J_l were reduced toward a slower strain rate. The experimental fracture toughness, K_Q, was found to be increased with strain rate, from 35 MPa at 2.54x10^-5 mm·s^-1 to 47 MPa at 2.54x10^-2 mm·s^-1. The fact that strain rate has a similar effect on K_Q and K_c verifies the importance of shear ligament in determining fracture toughness of the alloy. With the presence of hydrogen, length and work-to-fracture of the shear ligament were reduced. The toughening effect caused by shear ligament was reduced, and the alloy would behave in a brittle manner.

Microstructural Evolution and Superplastic Deformation Behavior of Fine Grain 5083Al
P.A. FRIEDMAN and A.K. GHOSH
The microstructural evolution during superplastic deformation of a fine grain Al-4.7 pct Mg alloy (5083Al) has been studied quantitatively. Starting from an average grain size of 7µm, grain growth was monitored in this alloy both under static annealing and with concurrent superplastic deformation at a high test temperature of 550°C. Grain size was averaged from measurements taken in longitudinal, transverse, and thickness directions and was found to grow faster during concurrent superplastic deformation than for static annealing. A grain growth law based on an additive nature between time-based and strain-based growth behavior was used to quantify the dynamics of concurrent grain growth. The extent of void formation during deformation was quantified as the area fraction of voids on L-S planes. This void fraction, referred to as the cavity area percent, was recorded at several levels of strain for specimens deformed at two different strain rates. A constitutive equation incorporating this grain growth data into the stress-strain rate data, determined during the early part of deformation, was generated and utilized to model the superplastic tensile behavior. This model was used in an effort to predict the stress-strain curves in uniaxial tension under constant and variable strain rate conditions. Particular attention was paid to the effects of a rapid prestrain rate on the overall superplastic response and hardening characteristics of this alloy.

Conditioning Monitoring by Microstructural Evaluation of Cumulative Fatigue Damage
C. FUKUOKA, Y.G. NAKAGAWA, J.J. LANCE, and R.N. PANGBORN
The objective of this work is to evaluate the damage induced below and above the fatigue limit (_t = 360MPa) in pressure vessel steels, such as SA508. Fatigue damage was induced in samples taken from an SA508 steel plate by various loading histories in order to examine the influence of prior cyclic loading below the fatigue limit. Cell-to-cell misorientation differences were measured by the selected area diffraction (SAD) method. Surface cracking was also studied by the replication method. Small cracks were observed after precycling both below and above the fatigue limit. It was, however, found that fatigue test bars had a longer lifetime after precycling below the fatigue limit, while precycling above the fatigue limit caused other specimens to fail even when subsequently cycled below the fatigue limit. Cell-to-cell misorientation usually increases with accumulation of fatigue damage, but it was found that the misorientations measured after precycling below the fatigue limit decreased again at the beginning of the subsequent cycling above the fatigue limit. It should be noted that the misorientation at failure was always about 4 to 5 deg, regardless of loading histories. Misorientation showed good correlation with the fatigue lifetime of the samples.

Evidence of Fracture Surface Interference for Cracks Loaded in Shear Detected by Phase-Shifted Speckle Interferometry
R.U. GOULET, T.S. GROSS, and D.A. MENDELSOHN
Interference of the asperities on a crack loaded in pure, remote shear wedges the crack faces open, thereby inducing a mode I stress intensity factor (SIF). The interference also shields the crack tip from the applied mode II SIF. Three-illumination beam, phase-shifted speckle interferometry was used to measure the three-dimensional incremental displacement fields in a 7 x 11 mm area around a mode I fatigue precrack in a Al 7075 specimen loaded in 94 increments of increasing shear. The displacement fields were accumulated relative to the unloaded state by sampling at appropriate lo cations in the incremental fields to optimize spatial resolution and compensate for large rigid body motions. The induced mode I SIF and the effective mode II SIF were estimated from the crack tip shear displacement (CTSD) and crack tip opening displacement (CTOD). A digitized fracture surface profile was numerically shifted according to the experimentally measured crack face displacements to determine the locations of fracture surface interference as a function of applied load.

Creep Lifetime Prediction of Oxide-Dispersion-Strengthened Nickel-Base Superalloys: A Micromechanically Based Approach
M. HEILMAIER and B. REPPICH
The high-temperature creep behavior of the oxide-dispersion-strengthened (ODS) nickel-base superalloys MA 754 and MA 6000 has been investigated at temperatures up to 1273 K and lifetimes of approximately 4000 hours using monotonic creep tests at constant true stress , as well as true constant extension rate tests (CERTs) at . The derivation of creep rupture-lifetime diagrams is usually performed with conventional engineering parametric methods, according to Sherby and Dorn or Larson and Miller. In contrast, an alternative method is presented that is based on a more microstructural approach. In order to describe creep, the effective stress model takes into account the hardening contribution caused by the presence of second-phase particles, as well as the classical Taylor back-stress caused by dislocations. The modeled strain rate-stress dependence can be transferred directly into creep-rupture stress-lifetime diagrams using a modified Monkman-Grant (MG) relationship, which adequately describes the interrelation between representing dislocation motion, and lifetime t_f representing creep failure. The comparison with measured creep-rupture data proves the validity of the proposed micromechanical concept.

An Evaluation of the Creep Properties of Two Al-Si Alloys Produced by Rapid Solidification Processing
LUBOS KLOC, STEFANO SPIGARELLI, EMANUELA CERRI, ENRICO EVANGELISTA, and TERENCE G. LANGDON
Tensile creep tests were conducted on two Al-Si alloys produced by rapid solidification: an Al-Si-Ni-Cr alloy and an Al-Si-Cu-Fe alloy, designated alloys A and B, respectively. The creep curves of these two alloys in the temperature range from 493 to 573 K were markedly different, with alloy A exhibiting a normal creep curve with a very short tertiary region and alloy B exhibiting an extended tertiary stage associated with strain localization. The minimum creep rates varied, with the applied stress raised to exponents of ~9.0 and ~8.5 for the two alloys, respectively. The hardness of alloy B decreased with time during the creep testing, but there was little or no change in the hardness of alloy A. These differences in the creep and hardness characteristics are attributed to the evolution of precipitates within the two alloys during creep testing. A detailed analysis shows that, over the temperature range examined experimentally, alloy A exhibits a creep strength that is superior to conventional Al-based alloys and comparable to, or even higher than, some SiC-reinforced Al composites.

Correlation of Microstructure and Fracture Toughness in High-Chromium White Iron Hardfacing Alloys
SUNGHAK LEE, SEONG-HUN CHOO, EUNG-RYUL BAEK, SANGHO AHN, and NACK J. KIM
A correlation is made of microstructure and fracture toughness in hypereutectic high-chromium white iron hardfacing alloys. In order to investigate the matrix effect of these alloys, in particular, four different matrices such as pearlite, austenite, and a mixture of pearlite and austenite were employed by changing the ratio of Mn/Si, while the total volume fraction of carbides was fixed. The hardfacing alloys were deposited twice on a mild steel plate by the self-shielding flux-cored arc-welding method. Fracture toughness was increased by increasing the volume fraction of austenite in the matrix, whereas hardness and abrasion resistance were nearly constant. In situ observation of the fracture process showed that cracks initiated at large primary carbides tended to be blocked at the austenitic matrix. This suggested that fracture toughness was controlled mainly by the amount of austenite in the matrix, thereby yielding the better toughness in the hardfacing alloy having the austenitic matrix. Considering both abrasion resistance and fracture toughness, therefore, the austenitic matrix was preferred for the high-chromium white iron hardfacing alloys.

Microstructure and Fracture of SiC-Particulate-Reinforced Cast A356 Aluminum Alloy Composites
SUNGHAK LEE, DONGIL KWON, and DONGWOO SUH
A microstructural analysis of local microfracture of cast A356 Al-SiC_p composites fabricated by permanent mold re-casting and squeeze-casting methods was made. Notch fracture toughness tests were conducted on these composites to identify critical fracture parameters using a stress-modified critical-strain criterion. The composite microstructure shows continuous networks of densely populated SiC and eutectic Si particles along the intercellular regions. Squeeze casting produces a more homogeneous structure and larger spacing of brittle particles and increases the tensile ductility and fracture toughness, while strength levels are almost identical to the re-casting case. The calculated values of the microstructurally characteristic distance l* for the re-cast and squeeze-cast composites are about 40 µm, which is comparable to the average sizes of the intercellular network. However, the reference critical strain *₀ for squeeze casting is larger than that for re-casting, showing a trend to higher ductility and fracture toughness.

Notch Fracture in -Titanium Aluminides
MADAN G. MENDIRATTA, ROBERT L. GOETZ, and DENNIS M. DIMIDUK
The notch fracture behavior of two -titanium aluminide alloys, having duplex and fully lamellar microstructures, has been investigated as a function of notch geometry and test temperature. The un-notched tensile properties and notch fracture loads are used to perform finite element analysis (FEA) to determine triaxial tensile stresses and effective plastic strains in the vicinity of notch roots. These results, together with fractographic examinations of notch failures, indicate that a crack nucleates in the triaxial tensile field when the effective von Mises stress just exceeds the uniaxial tensile yield stress. The high tensile stress component then propagates the nucleated microcrack to failure with local stress intensity reaching the toughness of the material. Thus, both plasticity and high tensile stress are required to cause notch failure.

Elevated Temperature Deformation Behavior of a Dispersion-Strengthened Al-Fe,V,Si Alloy
SHANTANU MITRA
The deformation behavior of a rapidly solidified, dispersion-strengthened Al alloy containing 11.7 pct Fe, 1.2 pct V, and 2.4 pct Si was studied at test temperatures up to 450°C using constant-stress creep and constant strain-rate tensile tests. Apparent stress exponents (n) up to ~24 and an activation energy of 360 kJ/mol were obtained with the standard Arrhenius type power-law creep equation, which also suggested a change in behavior at ~300°C. Substructure-invariant and dislocation/dispersoid interaction models were found to be inadequate for explaining the behavior. When the data were replotted as ^l/n vs , two regimes were found between 350°C and 450°C. A model with a pseudothreshold stress (_Th' for the higher stress regime resulted in n ~3, indicating solute drag in this regime. Transmission electron microscopy (TEM) showed departure-side pinning of dislocations at higher stresses. In the lower stress regime, TEM showed dislocation subgrain structures. Here, the model resulted in a stress exponent of ~4.5 indicating the dislocation climb mechanism. At temperatures below ~300°C, a single regime was found along with lower activation energies and a stress dependence of ~3. Dislocation pipe diffusion is proposed to explain the lower activation energy. The origin of _Th' has been tied to dislocation generation at the grain boundaries.

Fracture Characteristics, Microstructure, and Tissue Reaction of Ti-5Al-2.5Fe for Orthopedic Surgery
MITSUO NIINOMI, TOSHIRO KOBAYASHI, OSAMU TORIYAMA, NORIAKI KAWAKAMI, YOSHIHITO ISHIDA, and YUKIHIRO MATSUYAMA
The microstructure of Ti-5Al-2.5Fe, which is expected to be used widely as an implant material not only for artificial hip joints but also for instrumentations of scoliosis surgery, was variously changed by heat treatments. The effect of the microstructure on mechanical properties, fracture toughness, and rotating-bending fatigue strength in the air and simulated body environment, that is, Ringer's solution, was then investigated. Furthermore, the effect of the living body environment on mechanical properties and fracture toughness in Ti-5Al-2.5Fe were investigated on the specimens implanted into rabbit for about 11 months. The data of Ti-5Al-2.5Fe were compared with those of Ti-6Al-4V ELI, which has been used as an implant material mainly for artificial hip joints, and SUS316L, which has been used as an implant material for many parts, including the instrumentation of scoliosis surgery. The equiaxed structure, which is formed by annealing at a temperature below transus, gives the best balance of strength and ductility in Ti-5Al-2.5Fe. The coarse Widmanstatten structure, which is formed by solutionizing over transus followed by air cooling and aging, gives the greatest fracture toughness in Ti-5Al-2.5Fe. This trend is similar to that reported in Ti-6Al-4V ELI. The rotating-bending fatigue strength is the greatest in the equiaxed structure, which is formed by solutionizing below transus followed by air cooling and aging in Ti-5Al-2.5Fe. Ti-5Al-2.5Fe exhibits much greater rotating-bending fatigue strength compared with SUS 316L, and equivalent rotating-bending fatigue strength to that of Ti-6Al-4V ELI in both the air and simulated body environments. The rotating-bending fatigue strength of SUS316L is degraded in the simulated body environment. The corrosion fatigue, therefore, occurs in SUS 316L in the simulated body environment. Fatigue strength of Ti-5Al-2.5Fe in the simulated body environment is degraded by lowering oxygen content in the simulated body environment because the formability of oxide on the specimen surface is considered to be lowered comparing with that in air. The mechanical property and fracture toughness of Ti-5Al-2.5Fe and Ti-6Al-4V ELI are not changed in the living body environment. The hard-surface corrosion layer is, however, formed on the surface of SUS316L in the living body environment. The Cl peak is detected from the hard-surface corrosion layer by energy-dispersive X-ray (EDX) analysis. These facts suggests a possibility for corrosion fatigue to occur in the living body environment when SUS316L is used. The fibrous connective tissue and new bone formation are formed beside all metals. There is, however, no big difference between tissue morphology around each implant material.

Flow and Fracture of Bimaterial Systems Based on Aluminum Alloys
TODD M. OSMAN and JOHN J. LEWANDOWSKI
The effect of property mismatches on constrained plastic flow in aluminum alloys was investigated via both finite element modeling (FEM) and experimentation. Double-notched tension tests on monolithic aluminum alloys and notched trilayer laminates, consisting of the aluminum alloy and a discontinuously reinforced aluminum material, were used to experimentally study the degree of constraint developed in aluminum alloys for use in bimaterial systems. Constraint levels in bimaterial systems were found to be affected by mismatches in elastic modulus and strength. The trends observed in the development of constrained plastic flow in these studies were rationalized based upon the effects of stress triaxiality on the flow and fracture behavior of the various aluminum alloys investigated.

Hydrogen-Induced Cleavage Fracture of Fe₃Al-Based Intermetallics
L. QIAO and X. MAO
Hydrogen-induced fracture of ductile Fe₃Al-based intermetallics was studied through mechanical testing, fracture surface observation, and in situ transmission electron microscopy (TEM) tests of tensile specimens. Mechanical properties of ordinary ductile X-80 pipeline steel (low-alloy steel) were tested and compared with Fe₃Al intermetallics. Elongations of the Fe₃Al alloy decreased from 14 to 10 pct, with increases in the strain rate from 10^-6 to 10^-3/s. The elongation reduction of Fe₃Al was caused by the hydrogen-induced fracture. There was no elongation reduction when the testing was done in mineral oil. Non-necking occurred near the fracture section, and the fracture surfaces mainly consist of cleavage and partial intergranular morphologies. Elongation near the fracture surface of the Fe₃Al intermetallics was about 14 pct, which is the same as the total elongation. For the pipeline steel, however, an elongation near the fracture cross section was greater than 130 pct, which was much higher than its total elongation of 17 pct. In situ TEM observation on a tensile test sample showed crack propagation accompanied by dislocation plasticity. When the Fe₃Al was precharged cathodically, the crack tip was sharp. Its radius was much less than that obtained without hydrogen charging. The crack propagated along the grain boundary for the charged specimens, but penetrated the grain boundary for the specimen without hydrogen charging. Effects of hydrogen on plastic deformation and grain-boundary cracking are discussed in this article.

An Analysis of the Flow Stress of a Two-Phase Alloy System, Ti-6Al-4V
R.E. REED-HILL, C.V. ISWARAN, and M.J. KAUFMAN
An analysis of the tensile deformation behavior of a two-phase body-centered cubic (bcc)-hexagonal close-packed (hcp) alloy, Ti-6Al-4V, has been made. This has shown that the temperature dependence of the flow stress, the logarithm of the effective stress, and the strain-rate sensitivities can be described by simple analytical equations if the thermally activated strain-rate equation contains the Yokobori activation enthalpy H=H^o ln (*₀/*), where H^o is a constant, * the effective stress, and *₀ its 0 K value. The flow stress-temperature plateau region (500 to 600K) also can be rationalized analytically in terms of oxygen dynamic strain aging in the alpha phase.

Elastic Moduli of Titanium-Hydrogen Alloys in the Temperature Range 20°C to 1100°C
O.N. SENKOV, M. DUBOIS, and J.J. JONAS
The elastic properties of a series of polycrystalline titanium-hydrogen alloys (containing up to 25 at. pct H) were measured over the temperature range 20°C to 1100°C. The latter limits permitted investigation of adjacent parts of the +, , and phase fields. A laser ultrasonic technique was employed to measure the temperature and hydrogen-concentration dependencies of the elastic constants. The room-temperature elastic properties of the alloys depended only slightly on hydrogen concentration, remaining almost independent of the volume fraction of the hydride phase. In the phase field, the addition of hydrogen decreased the shear and Young's moduli and increased the bulk modulus, Lamé constant, and Poisson's ratio. The Young's and shear moduli decreased more rapidly with increasing temperature than in cubic phases. By contrast, Poisson's ratio increased with temperature. In the phase field, the temperature dependence of the elastic constants was weak. However, alloying with hydrogen increased the shear and Young's moduli, decreased Poisson's ratio, but did not appreciably affect the bulk modulus and Lam;aae constant. The different effects of hydrogen on the elastic constants of alpha and beta titanium are interpreted in terms of the influence of dissolved hydrogen on the stability of the hexagonal close-packed (hcp) and body-centered cubic (bcc) lattices in the vicinity of the -to- transformation. The present results are also used to help account for the effect of hydrogen concentration on the mechanical properties of Ti-H alloys.

Measurement of Friction under Sheet Forming Conditions
W. WANG, R.H. WAGONER, and X.-J. WANG
Among the tests used to evaluate friction in sheet forming processes, bending-under-tension (BUT) tests have received considerable attention. A new test was developed incorporating a smooth increase of wrap angle during deformation, even at high deformation rates, thus replicating typical conditions at die radii. A corresponding data analysis procedure was introduced to separate the effects of bending and friction and to account for the strain-rate sensitivity of the test material. The new test and analytical procedure were used to investigate the frictional behavior for four sheet alloys: interstitial-free (IF) steel, high-strength galvanized (HSG) steel, 2008-T4 aluminum, and 70/30 brass. A range of testing rates, pin radii, and lubrication conditions were employed, and the variation of friction coefficient (µ) as a function of contact radius, contact angle, and punch speed was measured. Complex variations of µ with respect to tool radius and wrap angle were observed, consistent with previous studies. No significant, systematic variation of µ was found for punch speeds ranging from 1.7 to 170 mm/s, contrary to reports in the literature. Failures in 2008-T4 aluminum shifted from tensile localization to bending as the tool radius was reduced. The other materials consistently failed by tensile localization.

The Quench Sensitivity of Cast Al-7 Wt Pct Si-0.4 Wt Pct Mg Alloy
D.L. ZHANG and L. ZHENG
The effect of quenching condition on the mechanical properties of an A356 (Al-7 wt pct Si-0.4 wt pct Mg) casting alloy has been studied using a combination of mechanical testing, differential scan ning calorimetry (DSC), and transmission electron microscopy (TEM). As the quench rate decreases from 250°C/s to 0.5°C/s, the ultimate tensile strength (UTS) and yield strength decrease by approximately 27 and 33 pct, respectively. The ductility also decreases with decreasing quench rate. It appears that with the peak-aged condition, both the UTS and yield strength are a logarithmic function of the quench rate, i.e., UTS or _y=A log R+B. The term A is a measure of quench sensitivity. For both UTS and yield strength of the peak-aged A356 alloy, A is approximately 32 to 33 MPa/log (°C/s). The peak-aged A356 alloy is more quench sensitive than the aluminum alloy 6063. For 6063, A is approximately 10 MPa/log (7°C/s). The higher quench sensitivity of A356 is probably due to the high level of excess Si. A lower quench rate results in a lower level of solute supersaturation in the -Al matrix and a decreased amount of excess Si in the matrix after quenching. Both of these mechanisms play important roles in causing the decrease in the strength of the peak-aged A356 with decreasing the quench rate.

ENVIRONMENT

Effect of Nitrogen on the Oxidation Behavior of Ti₃Al-Based Intermetallic Alloys
T.K. ROY, R. BALASUBRAMANIAM, and A. GHOSH
The effect of nitrogen on the oxidation behavior of Ti-25Al, T-24Al-15Nb, and Ti-25Al-11Nb (at. pct) has been examined in this study. The gases employed were nitrogen and oxygen-nitrogen and oxygen-argon mixtures in various proportions at a total pressure of 1 atmosphere. The experiments were carried out in the temperature range of 1100 to 1300K by thermogravimetry. The suitability of employing the parabolic rate law as the basis of interpretation of weight gain vs time data has been discussed. Oxidation resistance of Nb-containing alloys was superior to that of binary Ti-25Al, irrespective of the gas composition employed. The nitridation rates of alloys, with or without Nb, were more than an order of magnitude lower than those for oxidation. The scales were characterized by X-ray diffraction (XRD) and scanning electron microscopy with energy dispersive X-ray (EDX) analysis. The scales formed in all the conditions mostly consisted of TiO₂ and Al₂O₃. However, TiN was observed in scales of Nb-containing alloys in all nitrogen bearing gases and seemed to primarily account for improved oxidation resistance of the preceding in comparison to alloys without Nb. Nitrogen pretreatment was provided for some samples before oxidation for further elucidation of the role of nitrogen.

WELDING & JOINING

Infrared Transient-Liquid-Phase Joining of SCS-6/21S Titanium Matrix Composite

Microstructural Features of Friction Welded MA 956 Superalloy Material
C.Y. KANG, T.H. NORTH, and D.D. PEROVIC
The microstructural features of MA 956 friction welds were examined using a combination of optical microscopy and transmission electron microscopy (TEM). The MA 956 base material contained a uniform distribution of small diameter (<20 nm) Y₂O₃ particles. Limited numbers of larger diameter (>100 nm), spherically shaped Al₂O₃, Ti(C,N), Y₂O₃-Al₂O₃, and Al-Ti-Y containing particles were also observed in the MA 956 base material. In the recrystallized region, the grain size was largest at the bondline and increased markedly in the radial direction of the welded joint. Increasing the forging pressure from 50 to 150 MPa during the friction welding operation markedly increased the strain rate and decreased the grain size at the joint centerline. The friction welding operation substantially altered the particle chemistry, dimensions, and shape in the joint region. The number of aluminum-rich or titanium-rich particles was substantially decreased and large irregularly shaped particles were formed.

Nonuniform Distribution of Carbonitride Particles and Its Effect on Prior Austenite Grain Size in the Simulated Coarse-Grained Heat-Affected Zone of Thermomechanical Control-Processed Steels
D.W. TIAN, L.P. KARJALAINEN, BAINIAN QIAN, and XIAOFENG CHEN
The spatial distribution of carbonitride particles in the simulated coarse-grained heat-affected zone (HAZ) of Nb-Ti microalloyed thermomechanical control-processed (TMCP) steels was investigated using a scanning transmission electron microscope (STEM). It was found that the particles in quenched coarse-grained HAZ were frequently distributed in a nonuniform way, forming clusters and arrays of particles. This nonhomogeneity is defined by the grouping tendency of particles and described by the closeness of the average number density (the mean particle number per unit area) to the average local number density (the mean particle number per unit area, excluding the examined areas without particles). A high concentration of Nb (0.04 mass pct in this article) promoted the formation of carbonitride particle arrays and clusters because of its high segregation tendency at grain and subgrain boundaries during the cooling of a slab. Some of these particles remain undissolved at the peak temperature of a welding thermocycle and may result in sympathetic nucleation of new particles on them. The effectiveness of the particle groups to restrict grain growth is discussed.

SURFACE TREATMENT

SOLIDIFICATION

Macrotransport-Solidification Kinetics Modeling of Equiaxed Dendritic Growth: Part I. Model Development and Discussion
L. NASTAC and D.M. STEFANESCU
An analytical model that describes solidification of equiaxed dendrites has been developed for use in solidification kinetics-macrotransport modeling. It relaxes some of the assumptions made in previous models, such as the Dustin-Kurz, Rappaz-Thevoz, and Kanetkar-Stefanescu models. It is assumed that nuclei grow as unperturbed spheres until the radius of the sphere becomes larger than the minimum radius of instability. Then, growth of the dendrites is related to morphological instability and is calculated as a function of melt undercooling around the dendrite tips, which is controlled by the bulk temperature and the intrinsic volume average concentration of the liquid phase. When the general morphology of equiaxed dendrites is considered, the evolution of the fraction of solid is related to the interdendritic branching and dynamic coarsening (through the evolution of the specific interfacial areas) and to the topology and movement of the dendrite envelope (through the tip growth velocity and dendrite shape factor). The particular case of this model is the model for globulitic dendrite. The intrinsic volume average liquid concentration and bulk temperature are obtained from an overall solute and thermal balance around a growing equiaxed dendritic grain within a spherical closed system. Overall solute balance in the integral form is obtained by a complete analytical solution of the diffusion field in both liquid and solid phases. The bulk temperature is obtained from the solution of the macrotransport-solidification kinetics problem.

Macrotransport-Solidification Kinetics Modeling of Equiaxed Dendritic Growth: Part II. Computation Problems and Validation on INCONEL 718 Superalloy Castings
L. NASTAC and D.M. STEFANESCU
In Part I of the article, a new analytical model that describes solidification of equiaxed dendrites was presented. In this part of the article, the model is used to simulate the solidification of INCONEL 718 superalloy castings. The model was incorporated into a commercial finite-element code, PROCAST. A special procedure called microlatent heat method (MLHM) was used for coupling between macroscopic heat flow and microscopic growth kinetics. A criterion for time-stepping selection in microscopic modeling has been derived in conjunction with MLHM. Reductions in computational (CPU) time up to 90 pct over the classic latent heat method were found by adopting this coupling. Validation of the model was performed against experimental data for an INCONEL 718 superalloy casting. In the present calculations, the model for globulitic dendrite was used. The evolution of fraction of solid calculated with the present model was compared with Scheil's model and experiments. An important feature in solidification of INCONEL 718 is the detrimental Laves phase. Laves phase content is directly related to the intensity of microsegregation of niobium, which is very sensitive to the evolution of the fraction of solid. It was found that there is a critical cooling rate at which the amount of Laves phase is maximum. The critical cooling rate is not a function of material parameters (diffusivity, partition coefficient, etc.). It depends only on the grain size and solidification time. The predictions generated with the present model are shown to agree very well with experiments.

Modeling of Primary and Secondary Dendrites in a Cu-6 Wt Pct Sn Alloy
N. TIEDJE, P.N. HANSEN, and A.S. PEDERSEN
Microstructure of gas-atomized CuSn6 particles has been investigated using scanning electron microscopy (SEM), and it is shown that the dendrite arm spacing (DAS) is related to the particle diameter (d) so that DAS=0.19d^0.72. Formation of microstructures in the particles are modeled using a numerical solidification model. This model concerns tips of cells and dendrites, but in the present investigation, it is, in a simple manner, extended to comprise whole cells and dendrites. Furthermore, ripening of dendrite arms is taken into consideration. It is found that for increasing growth rates there is a transition from dendrites to cells when the growth velocity approaches the diffusional velocity in the melt, i.e., when the Peclet number is equal to one. It is also shown that both primary stem spacing and dendrite spacing are related to the ratio between the volume in the liquid where there is solute diffusion and to the surface area of the cells and dendrites (D/A). The relation between spacing and D/A is the same for cells and dendrites, indicating that the spacing selection is controlled purely by solute diffusion in the melt.

Communication: Lamellar Growth of Eutectic Equiaxed Grains
ADRIAN V. CATALINA and DORU M. STEFANESCU

MATERIALS PROCESSING

Modeling Particle Fracture During the Extrusion of Aluminum/Alumina Composites
C.H.J. DAVIES, W.-C. CHEN, D.J. LLOYD, E.B. HAWBOLT, I.V. SAMARASEKERA, and J.K. BRIMACOMBE
Particle fracture during the extrusion of a 6061/Al₂O₃/20p composite has been modeled using a modified comminution formulation. It has been assumed that the particles contain a Poisson distribution of flaws, and that the distribution is specific to the alumina and the method of production; the particle distribution in the extrudate was characterized by the Rosin-Rammler (RR) distribution. The model relates macroscopic deformation variables to fracture and, starting with the distribution in the as-cast material in each case, is able to predict with reasonable accuracy changes in size distribution for three extrusion ratios. Some discrepancy between prediction and experiment occurs at small sizes. This is believed to result mainly from inaccuracies in the measured data and effects of the continuous size distribution in representing a set of discontinuous data. The model is potentially generalizable to any particulate-reinforced metal matrix composite (MMC).

The Squeeze Casting of Hypoeutectic Binary Al-Cu
M. GALLERNEAULT, G. DURRANT, and B. CANTOR
A combination of thermocouple measurements, optical microscopy, and image analysis has been used to investigate the effects of applied pressure, melt temperature, mould insulation, and addition of grain refiner on the cooling/solidification behavior and resulting macro- and microstructure in squeeze cast Al-4.5 wt pct Cu ingots. Channel macrosegregates are formed in Al-4.5 wt pct Cu squeeze castings because of an increased rate of heat removal due to the application of pressure. The increased density of solute-rich liquid in the Al-Cu system causes channel segregates to form with a characteristic V pattern. Pressure and melt superheat increase the temperature gradients and cooling rate during squeeze casting but have only a minor effect upon the formation of channel segregates. The addition of grain refiner disperses but does not eliminate channel segregates. The application of pressure during squeeze casting changes the solidification behavior from mushy to near-plane front, leading to normal rather than inverse solute concentration profiles.

Modeling Recrystallization Kinetics, Grain Sizes, and Textures During Multipass Hot Rolling
HANS ERIK VATNE, KNUT MARTHINSEN, ROAR ØRSUND, and ERIK NES
A physically based model for the evolution of recrystallization microstructures and textures during hot rolling of aluminum is presented. The approach taken differs from similar models developed for steels. The present model is based on recent experimental investigations directed toward identifying the nature of the nucleation sites for recrystallized grains of different crystallographic orientations. Particle stimulated nucleation (PSN) and nucleation from cube bands and grain boundary regions have been incorporated in the model. The multipass aspect complicates the modeling due to partial recrystallization between the rolling passes. Two different approaches have been suggested to handle this. The model has been applied to predictions of recrystallization kinetics, recrystallized grain sizes, and recrystallization textures during multipass hot rolling of aluminum. The predictions are reasonable compared to experimental results.

Communication: Synthesis of Full-Density Nanocrystalline Tungsten Carbide by Reduction of Tungstic Oxide at Room Temperature
M. SHERIF EL-ESKANDARANY, M. OMORI, M. ISHIKURO, T.J. KONNO, K. TAKADA, K. SUMIYAMA, T. HIRAI, AND K. SUZUKI

Communication: Synthesis of Nanocrystalline Ni₃Cu by Sol-Gel Route
S.K. PRADHAN, A. DATTA, M. PAL, AND D. CHAKRAVORTY

COMPOSITE MATERIALS

Martensitic Transformations in NiMnAl Phase Alloys
R. KAINUMA, H. NAKANO, and K. ISHIDA
Martensitic transformation characteristics in a wide range of alloy compositions in the NiMnAl system have been investigated by differential scanning calorimetry (DSC), X-ray diffraction (XRD), and transmission electron microscopy (TEM). The results indicate that while the NiMnAl alloys with relatively high Al and Mn contents transform martensitically to either the 14M(7R) or the 10M phase with a long-period stacking-order structure, those with low Al or Mn contents transform to the L1₀(3R:2M) structure. The Ms temperature and the c/a ratio of the L1₀ martensite increase with decreasing Al or Mn content. In the Ni-30Mn-20Al alloy, typical premartensite microstructures, such as needlelike, modulated, and tweed structures, are observed and it is also found that a distorted B2 structure is formed as an intermediate phase prior to transformation to the 10M martensite.

Infiltration of Fibrous Preform by Molten Aluminum in a Centrifugal Force Field
YOSHINORI NISHIDA, ITARU SHIRAYANAGI, and YOSHIBUMI SAKAI
The pressure to infiltrate molten aluminum into alumina short-fiber preforms was generated by centrifugal force, and it was proved that centrifugal force is effective as a motive force for the infiltration and sound composites that are produced. The initiation pressure of the infiltration was also calculated and obtained clearly. The start pressure agreed well with the threshold pressure obtained theoretically. The characteristics of the infiltration in the preform was discussed using a one-dimensional solution of D'Arcy's law in a centrifugal force field. It was made clear that when preform surface pressure is low, the region having lower pressure than the threshold pressure appears in the infiltrated region.

Influence of Reinforcement Volume Fraction and Size on the Microstructure and Abrasion Wear Resistance of Hot Isostatic Pressed White Iron Matrix Composites
E. PAGOUNIS, M. TALVITIE, and V.K. LINDROOS
The changes in the microstructure and wear resistance of a powder metallurgical high-Cr white iron after the incorporation of TiC particles were studied in the present work. Various reinforcement volume fractions and sizes were used in order to examine their influence on the three-body abrasion wear resistance. The experiments were carried out at three different austenitizing temperatures. The most important observation after a microstructural examination was the increased amount of martensite in the composites subjected to identical heat treatment procedures with the unreinforced alloy. The austenite-to-martensite transformation in the composites increased with the TiC volume fraction and with the austenitizing temperature. This indicates that the two parameters have a key role in the transformation mechanism, which seems to be mechanically induced. The increasing of martensitic transformation with the TiC content in the composites enhanced continuously the supporting ability of the iron alloy matrix to the TiC particles, which in turn increased the wear resistance of the composites. The abrasion wear resistance increased with the TiC volume fraction until the onset of spalling. However, in composites containing coarse reinforcements, spalling occurred earlier in the wear process. This decreased wear resistance significantly because spalled TiC particles contributed additionally to wear.

Influence of Matrix Structure on the Abrasion Wear Resistance and Toughness of a Hot Isostatic Pressed White Iron Matrix Composite
E. PAGOUNIS, M. TALVITIE, and V.K. LINDROOS
The influence of the matrix structure on the mechanical properties of a hot isostatic pressed (hipped) white iron matrix composite containing 10 vol pct TiC is investigated. The matrix structure was systematically varied by heat treating at different austenitizing temperatures. Various subsequent treatments were also employed. It was found that an austenitizing treatment at higher temperatures increases the hardness, wear resistance, and impact toughness of the composite. Although after every different heat treatment procedure the matrix structure of the composite was predominantly marten sitic, with very low contents of retained austenite, some other microstructural features affected the mechanical properties to a great extent. Abrasion resistance and hardness increased with the austen itizing temperature because of the higher carbon content in martensite in the structure of the com posite. Optimum impact energy values were obtained with structures containing a low amount of M (M₇C₃+M₂₃C₆) carbides in combination with a decreased carbon content martensite. Structure austenitized at higher temperatures showed the best tempering response. A refrigerating treatment was proven beneficial after austenitizing the composite at the lower temperature. The greatest portion in the increased martensitic transformation in comparison to the unreinforced alloy, which was ob served particularly after austenitizing the composite at higher temperatures, was confirmed to be mechanically induced. The tempering cycle might have caused some additional chemically induced transformation. The newly examined iron-based composite was found to have higher wear resistance than the most abrasion-resistant ferroalloy material (white cast iron).

Creep Deformation and Damage in a Continuous Fiber-Reinforced Ti-6Al-4V Composite
S.W. SCHWENKER and D. EYLON
Mechanisms of longitudinal creep deformation and damage were studied in an eight-ply unidirectional-reinforced SCS-6/Ti-6Al-4V composite. The composite was creep tested in air under constant tensile load at temperatures from 427°C to 650°C and stresses from 621 to 1380 MPa. In situ acoustic emission (AE) monitoring and post-test metallographic evaluation were used to study fiber fracture and damage during creep. At low creep stresses, creep rates continuously decreased to near-zero values. This was attributed to a mechanism of matrix relaxation and the time-dependent redis tribution of load from the ductile matrix to the elastic fibers. At higher stresses, progressive fiber overload occurred during creep loading. In this case, the composite exhibited a stage of decreasing creep rate (due primarily to matrix relaxation), followed by a secondary stage of nearly constant creep rate due to fiber fracture. The results indicate that interfacial oxidation damage and inefficient load transfer at elevated temperatures significantly decreased the capability of broken fibers to carry load. As a result, additional time-dependent stress redistribution occurred in the composite, which was responsible for the secondary creep stage.