METALLURGICAL AND MATERIALS TRANSACTIONS A
	ABSTRACTS Volume 29A, No. 3A, March 1998 This Month Featuring: Symposium on Fundamentals of Gamma Titanium Aluminides: Part II; Alloy Phases; Transformations; Transport Phenomena; Mechanical Behavior; Solidification; Materials Processing; Composite Materials. View March 1998 Table of Contents.

SYMPOSIUM ON FUNDAMENTALS OF GAMMA TITANIUM ALUMINIDES: PART II

Autogenous Gas Tungsten Arc Weldability of Cast Alloy Ti-48Al-2Cr-2Nb (Atomic Percent) versus Extruded Alloy Ti-46Al-2Cr-2Nb-0.9Mo (Atomic Percent)
D.J. BHARANI and V.L. ACOFF
This study examines procedures for consistently producing sound (crack and void free) welds using the autogenous (without filler metal) gas tungsten arc (GTA) welding process. Cast alloy Ti-48Al- 2Cr-2Nb (at. pct) and extruded alloy Ti-46Al-2Cr-2Nb-0.9Mo (at. pct) have been examined to determine if sound welds can be produced using autogenous GTA welding without any preheat. Experimentation consisted of GTA spot welding samples of gamma titanium aluminide at weld current levels of 45, 55, 65, and 75 A for a duration of 3 seconds. For the cast alloy, current levels of 45, 55, and 65 A for 3 seconds produced similar fusion zone microstructures, which consisted of a dendritic solidification structure. The fusion zone microstructure of the 75 A for 3 seconds current level differed significantly from the lower current levels. It also consisted of a dendritic solidification structure; however, the morphology was quite different. For the extruded alloy, current levels of 45 and 55 A for 3 seconds produced fusion zone microstructures similar to the lower current level samples of the cast -TiAl, which consisted of a dendritic solidification structure. The fusion zone microstructures of the 65 and 75 A samples were similar to each other, but they had a dendritic solidification structure of a different morphology than that of the 45 and 55 A samples. For both alloys at all current levels, microhardness profiles showed an increase in hardness from the base metal to the fusion zone. There were no significant differences in the average fusion zone hardness as a function of increasing current level. However, nanoindentation testing did show that certain phases and microconstituents in the fusion zone did have significant variations in hardness in relation to the enrichment and depletion of chromium.

Coherency Stresses in Lamellar Ti-Al
M.A. GRINFELD, P.M. HAZZLEDINE, B. SHOYKHET, and D.M. DIMIDUK
General formulas are given for the coherency strains and stresses in a multilayer far from the free surfaces. The multilayer is assumed to be a periodic stack of different elastically isotropic materials, but there may be any number of layers in the stack and they may each have any thickness and any elastic constants. The results are applied to lamellar Ti-Al alloys, in which there are shear misfits between different layers and both shear and biaxial misfits between the and ₂ layers. In a fully coherent multilayer, the stresses would be large, in the GPa range, and in high strength, thin lamella alloys, the coherency stresses are a substantial fraction of a GPa. The shear stresses act principally on hard mode deformation systems, and the biaxial stresses place every ₂ lamella in biaxial compression. This biaxial compression, which, for dislocation glide, is equivalent to a uniaxial tension normal to the lamella, is particularly large when the ₂ volume fraction is small.

Effect of Deformation Temperature on Fatigue and Fracture Behavior in TiAl Polysynthetically Twinned Crystals
Y. UMAKOSHI, H.Y. YASUDA, T. NAKANO, and K. IKEDA
The temperature and orientation dependence of cyclic deformation, fatigue life, and fracture behavior in TiAl polysynthetically twinned (PST) crystals were investigated, focusing on the change of plastic strain energy and deformation mode in the domains. Stress-controlled fatigue tests were performed at 1 or 10 Hz using the same stress amplitude in tension and compression (R = -1) over a temperature range from -196°C to 700°C. The fatigue strength at = 45 deg ( being the angle between the loading axis and lamellar planes) decreased monotonically with increasing temperature. At = 0 deg, the fatigue strength was high up to 500°C, but the fatigue life decreased rapidly above 600°C because of dynamic recovery and interlamellar separation. The plastic strain energy- stress amplitude curves in specimens fatigued with = 45 deg increased monotonically with stress amplitude for all temperatures and for higher temperatures with = 0 deg. At 25°C and -196°C with = 0 deg, three regions in the plastic strain energy-stress amplitude curves were observed. This anomalous change in the plastic strain energy at lower temperatures was due to a transition in primary deformation mode between twinning and slip by ordinary dislocations in some domain orientations.

Atomistic Simulation of Fracture in TiAl
JULIA PANOVA and DIANA FARKAS
Atomistic simulations of fracture in L1₀ TiAl were carried out using embedded atom method (EAM) interatomic potentials and molecular statics. We studied the behavior of semi-infinite cracks under mode I loading in different orientations of the crack front and plane. For the [](111) orientation, we observed dislocation emission involving the formation of superlattice intrinsic stacking fault (SISF). For the [001](110) orientation, we observed the emission of ordinary 1/2[110] edge dislocations that were highly mobile and had a compact core. We found that cracks with [001](100) orientation cleaved near the Griffith value of loading in a purely brittle manner. Similar behavior was observed for cracks with [](100) orientation.

Finite Element Analysis of Cavitating Facet Interaction in a Fully Lamellar Titanium Aluminide Alloy under Creep Conditions
ANIRBAN CHAKRABORTY and JAMES C. EARTHMAN
Creep constrained grain boundary cavitation in a fully lamellar (FL) form of a titanium aluminide intermetallic alloy has been studied using finite element (FE) techniques. Two different forms of FL models were considered. Cavitation was modeled in the presence of grain boundary sliding (GBS) for the case of straight former grain boundaries. Models of cavitation without GBS were also performed for a FL microstructure with serrated former grain boundaries. The effect of cavitating facet interaction on rupture life has been studied. A comparison between the FL forms and a dual- phase equiaxed microstructure having the same phase ratio (₂/) was also made to examine the relative susceptibility of these microstructures to high-temperature damage. It has been observed that the overall effect of interaction between cavitating facets increases the rupture time significantly when these facets are on adjacent grains. However, in the presence of GBS, cavitation on the facet with narrower separation effectively reduces the cavity growth rate on the facet with wider separation.

Microstructural Development and Creep Deformation in Equiaxed , + ₂, and + ₂ + B2 Titanium Aluminides
ERIC A. OTT and TRESA M. POLLOCK
The development of microstructure and its influence on creep properties have been studied for structures including equiaxed , duplex, and other structures of varying ₂ morphology in two Ti-48Al- 2Cr-2Nb alloys. Heat treatments at 1125°C have been utilized to produce equiaxed microstructures in alloys with or without Mo additions. The transformation produces ₂ plates with several orientation variants within grains during subsequent annealing of the equiaxed microstructures below the transus. Formation of this ₂ morphology results from rapid up-quenching (UQ), and this structure persists through annealing, cooling, and creep testing. Differences in minimum creep rates for several microstructures containing varying amounts of multi- or single variant /₂ grains are shown to be minimal. The presence of Mo has also resulted in improved creep resistance in equiaxed and + ₂ + B2 structures, as compared to similar microstructures in the Ti-48Al-2Cr- 2Nb alloy. Deformation during creep at 760°C at stresses between 200 and 400 MPa occurs by a combination of twinning and dislocation glide without recrystallization, resulting in power-law stress exponents in the range of 6 to 9. Only minimal strain path dependence of the minimum creep rate is detected in a comparison of creep rates in stress jump, stress drop, and single stress tests.

ALLOY PHASES

TRANSFORMATIONS

Upper Acicular Ferrite Formation in a Medium-Carbon Microalloyed Steel by Isothermal Transformation: Nucleation Enhancement by CuS
I. MADARIAGA, J.L. ROMERO, and I. GUTIÉRREZ
The isothermal transformation vs time of a medium-carbon microalloyed steel at 450°C, following austenitization at 1250°C for 45 minutes, has been investigated using optical microscopy, scanning electron microscopy, and transmission electron microscopy (TEM). At short times, the fine microstructure of acicular ferrite is nucleated at MnS inclusions, which are covered by a shell of a hexagonal CuS phase. The special orientation between MnS and the CuS crystals of this shell enables the formation of a low-energy interface between the ferrite and the inclusion with, at the same time, the ferrite satisfying one of the 24 variants of the orientation relationship into the Bain region with austenite. As the treatment times are increased, the increase in the volume fraction of acicular ferrite being formed raises the carbon concentration of the austenite, such that some retained austenite instead of martensite is observed for these intermediate treatment times. This retained austenite transforms to ferrite plus carbides at long treatment times, resulting in a final microstructure of acicular ferrite, very similar in nature to those encountered in the case of upper bainite formation.

TRANSPORT PHENOMENA

H_B + H_L + H = k_BT_P

in a temperature region T_P > T₀ + T, below which deviation from linearity of the relationship takes place; T mainly depends on the value of H_L. In the preceding expression, T₂₀ is the initial temperature of heating; excess energy H is found to be a constant, 0.09 eV, for a fixed value of H_L, 0.082 eV, for lattice diffusion of hydrogen in body-centered-cubic iron. The influence of diffusion and trap parameters, i.e., pre-exponential factor of hydrogen diffusivity, trap site density, parameter of trapping, heating rate, and specimen size, enters into the single parameter in the preceding expression.

Hydrogen and Deuterium in Pd-25 Pct Ag Alloy: Permeation, Diffusion, Solubilization, and Surface Reaction
E. SERRA, M. KEMALI, A. PERUJO, and D.K. ROSS
In this work, the hydrogen and deuterium transport parameters such as permeability, diffusivity, and solubility in a Pd-25 pct Ag alloy are studied, using a permeation gas phase technique and, additionally, a computer-controlled microbalance system. In the region in which Sieverts' law is valid, the hydrogen solubilities found using both techniques are in very good agreement. From permeation measurements, the surface constants for the adsorption (k₁) and release (k₂) of hydrogen and deuterium in Pd-25 pct Ag alloy are also determined. The measurements cover the temperature range from 323 to 773 K and a pressure range from 1 to 10⁵ Pa.

A Mathematical Model for the Solute Drag Effect on Recrystallization
MASAYOSHI SUEHIRO, ZI-KUI LIU, and JOHN ÅGREN
The model for the solute drag effect in phase transformations has been applied to recrystallization, i.e., moving grain boundaries. In this model, the total driving force is dissipated by the interfacial energy, the finite interfacial mobility, the solute drag in boundaries, and diffusion in the matrix ahead of the interface, of which all are taken into account consistently. The effects of the Gibbs energy of segregation and the diffusivity of impurity atoms in boundaries were investigated. The results show that the Gibbs energy of segregation mainly affects the critical composition at which the drastic change in the boundary velocity appears, and the diffusivity of impurity atoms in boundaries mainly affects the velocity reduced by the solute drag effect. In other words, the Gibbs energy of segregation and the diffusivity of impurity atoms in boundaries can be evaluated from experimental data by means of the present model. This model was applied to the Al-Mg system, and the Gibbs energy of segregation and the diffusivity of Mg in boundaries were evaluated from experimental data. The evaluated Gibbs energy of segregation agrees with the estimate based on elastic energy considerations. The diffusivity estimated from this model is smaller than that measured along the grain boundary.

MECHANICAL BEHAVIOR

Gaseous Hydrogen Embrittlement of a Hydrided Zirconium Alloy
J.-H. HUANG and M.-S. YEH
ZIRCALOY-4 plate specimens were gaseously hydrided up to 340 ppm H and then tested in a hydrogen gas environment of various pressures up to 2020 kPa at 25°C, 100°C, and 200°C. Notched tensile specimens were chosen to better understand the ``ductile-brittle transition'' associated with hydrogen content and hydrogen pressure. The purpose of the present investigation is to understand the synergistic effect of hydrogen gas and internal hydrides on the mechanical properties of ZIRCALOY-4. The results showed that for both uncharged and hydrided specimens, the notch tensile strength decreased with increasing hydrogen pressure as well as increasing temperature. Compared with uncharged specimens, the specimens with hydrides had lower values of notch tensile strength. A ductile-brittle transition was found on specimens tested at 25°C and at hydrogen pressures between 0 and 1010 kPa. For the specimen containing 220 ppm H, the reduction of area (RA) at 25°C and at hydrogen pressures of 1010 kPa and above was drastically reduced, resulting in almost completely brittle behavior. This hydrogen and hydride-induced cracking was found to be an autocatalytic process. From the fractographic finding, the ductile-brittle transition was closely related to the precipitation and distribution of brittle hydrides. The ductile-brittle transition disappeared as the temperature increased to 100°C and above. This can be attributed to the improved ductility of the zirconium matrix with increasing temperature.

Effect of Tungsten Particle Shape on Dynamic Deformation and Fracture Behavior of Tungsten Heavy Alloys
DONG-KUK KIM, SUNGHAK LEE, and HEUNG-SUB SONG
The effect of the tungsten particle shape on the dynamic deformation and fracture behavior of tungsten heavy alloys was investigated. Dynamic torsional tests were conducted using a torsional Kolsky bar for five alloys, one of which was fabricated by the double-cycled sintering process, and then the test data were compared via microstructures, mechanical properties, adiabatic shear banding, and fracture mode. The dynamic torsional test results indicated that in the double-sintered tungsten alloy whose tungsten particles were very coarse and irregularly shaped, cleavage fracture occurred in the central area of the gage section with little shear deformation, whereas shear deformation was concentrated in the central area of the gage section in the other alloys. The deformation and fracture behavior of the double-sintered alloy correlated well with the observation of the impacted penetrator specimen and the in situ fracture test results, i.e., microcrack initiation at coarse tungsten particles and cleavage crack propagation through tungsten particles. These findings suggested that the cleavage fracture mode would be beneficial for the self-sharpening effect, and, thus, the improvement of the penetration performance of the double-sintered tungsten heavy alloy would be expected.

Solid Particle Erosion of an Fe-Fe₃C Metal Matrix Composite
B.A. LINDSLEY and A.R. MARDER
The erosion resistance and morphology of spheroidized Fe-C alloys containing 0.2 to 1.4 wt pct carbon was investigated. The Fe-C alloy system was chosen as a model metal-matrix composite for the study of the effect on erosion of a hard second phase in a ductile matrix. Alloys were austenitized and water quenched to form martensite, then tempered at 690°C for different times to produce carbide sizes of 0.4, 0.8, 1.6, and 2.4 µm. Utilizing these materials, it was found that the erosion resistance increased as the microstructural features decreased in size, with the important microstructural variables being carbide spacing and ferrite grain size. These variables control dislocation motion in the ferrite and, in turn, affect the plastic deformation and the erosion resistance of the spheroidized alloys. For the 0.4 to 1.4 pct C alloys, the carbide spacing was sufficient to determine erosion rate, whereas, for the 0.2 pct C alloys, ferrite grain size became the controlling structure. Microstructural spacing, which is a measure of the mean free path between both the grain boundaries and the carbides, was found to describe all of the erosion data. A Hall-Petch-type relationship was found between microstructural spacing and both erosion rate and hardness.

Warm-Temperature Tensile Ductility in Al-Mg Alloys
ERIC M. TALEFF, GREGORY A. HENSHALL, T.G. NIEH, DONALD R. LESUER, and JEFFREY WADSWORTH
Several binary and ternary Al alloys containing from 2.8 to 5.5 wt pct Mg were tested in tension at elevated temperatures (200°C to 500°C) over a range of strain rates (10^-4 to 2.0 s^-1). Tensile ductilities of up to 325 pct were obtained in binary Al-Mg alloys with coarse grains deformed in the solute-drag creep regime. Under test conditions in which solute-drag creep controls deformation, Mg in concentrations from 2.8 to 5.5 wt pct neither affects tensile ductility nor influences strain-rate sensitivity or flow stress significantly. Strength is shown to increase with increasing Mg concentration, however, in the power-law-breakdown regime. The solute-drag creep process, which leads to superplastic-like elongations, is shown to have no observable grain-size dependence in a binary Al-Mg material. Ternary alloying additions of Mn and Zr are shown to decrease the strain-rate sensitivity during solute-drag creep, negatively influencing ductility. An important cause of reduced ductility in the ternary alloys during creep deformation is found to be a transition from necking- controlled failure in the binary alloys to cavitation-controlled failure in the ternary alloys investigated. An increase in ternary element concentration, which can increase the relative volume percentage of proeutectic products, increases cavitation.

The Influence of Crystallographic Orientation and Strain Rate on the High-Temperature Low-Cyclic Fatigue Property of a Nickel-Base Single-Crystal Superalloy
Z.F. YUE and Z.Z. LU
Fully reversed low-cyclic fatigue (LCF) tests were conducted on [001], [012], [], [011], and [] oriented single crystals of nickel-based superalloy DD3 with different cyclic strain rates at 950°C. The cyclic strain rates were chosen as 1.0 X 10^-2, 1.33 X 10^-3, and 0.33 X 10^-3 s^-1. The octahedral slip systems were confirmed to be activated on all the specimens. The experimental result shows that the fatigue behavior depends on the crystallographic orientation and cyclic strain rate. Except [001] orientation specimens, it is found from the scanning electron microscopy (SEM) examination that there are typical fatigue striations on the fracture surfaces. These fatigue striations are made up of cracks. The width of the fatigue striations depends on the crystallographic orientation and varies with the total strain range. A simple linear relationship exists between the width and total shear strain range modified by an orientation and strain rate parameter. The nonconformity to the Schmid law of tensile/compressive flow stress and plastic behavior existed at 950°C, and an orientation and strain rate modified Lall-Chin-Pope (LCP) model was derived for the nonconformity. The influence of crystallographic orientation and cyclic strain rate on the LCF behavior can be predicted satisfactorily by the model. In terms of an orientation and strain rate modified total strain range, a model for fatigue life was proposed and used successfully to correlate the fatigue lives studied in this article.

SOLIDIFICATION

Primary Spacing in Directional Solidification
DEXIN MA and PETER R. SAHM
A new analytical model is developed to explain the variation in primary spacing with growth velocity V. In this model, dendrite growth is resolved into two parts: the growth of the center core and that of the side arms, which are separately treated. In contrast to the assumption in the current models, it is only the dendrite core, not the entire dendrite, whose curvature radius at the tip is directly related to dendrite tip radius R. The primary spacing is considered to be the sum of core diameter and twice the sidearm length. As long as the growth of side arms is suppressed, it becomes cellular growth. As a result, this model gives a reasonable dependence of cell and dendrite spacing on the process parameters. The proposed model has been applied to several alloys to compare its predictions both with experimental data and with the analytical expression of the Hunt-Lu model.

Communication: The Influence of Gravity-Related Convection on Secondary Arm Evolution in NH₄Cl-H₂O
Mary Helen McCay, John A. Hopkins, and T. Dwayne McCay

MATERIALS PROCESSING

COMPOSITE MATERIALS