Materials Week '97: Monday AM Session Abstracts

Focusing on physical metallurgy and materials, Materials Week '97, which incorporates the TMS Fall Meeting, features a wide array of technical symposia sponsored by The Minerals, Metals & Materials Society (TMS) and ASM International. The meeting will be held September 14-18 in Indianapolis, Indiana. The following sessions will be held Monday morning, September 15, during Materials Week 1997. To view other programming planned for the meeting, go to the technical program contents page.

ALLOY DESIGN AND SOLDERING TECHNOLOGIES FOR LEAD-FREE AND LEAD-BEARING SOLDERS: Session I

Session Chair: J.W. Morris, Jr., University of California, Berkeley, CA 94720

PHASE COARSENING AND CRACK GROWTH RATE DURING THERMO-MECHANICAL CYCLING OF 63Sn37Pb SOLDER JOINTS: P. Hacke, J. Fahmy, H. Conrad, Materials Science & Engineering Department, North Carolina State University, Raleigh, NC 27695-7907

The change in average phase size and in the crack area growth rate dA_c/dN during thermo-mechanical cycling (-30°C to 130°C) of 63Sn37Pb solder joints in shear were determined. The increase in phase size during the thermo-mechanical cycling was significantly greater than occurred in static isothermal annealing. Further, dA_c/dN increased during the thermo-mechanical cycling concurrent with the increase in phase size. The mechanisms for the phase coarsening and its effect on the crack area growth rate are discussed.

8:55 am INVITED

THE BEHAVIOR OF LOW AND INTERMEDIATE MELTING TEMPERATURE SOLDERS: D.R. Frear, Sandia National Laboratories, Albuquerque, NM 87185

A number of low and intermediate melting temperature solder alloys are used for hierarchical soldering in applications such as multichip modules and radar assemblies. Two alloys that have desirable properties include 50In-50Pb (MP=170°C) and 40Sn-40In-20Pb (MP=121°C). In addition to their relative melting temperatures, these alloys also have excellent stress relaxation properties that are necessary for assembling the multilevel ceramic layers required for radars. This paper will discuss experimental and computational results performed to determine the physical reactions, mechanical behavior, and lifetime reliability of 50In-50Pb and 40Sn-40In-20Pb. The reaction of these alloys with base metallizations such as Au-based thick films on alumina and copper/gold traces on Duroid-like substrates will be described. The time dependent constitutive relations were experimentally determined by compression creep tests. he derived constitutive relations were incorporated into a finite element tool that has been developed for near-eutectic Sn-Pb solders and lifetime predictions under used conditions for the In-Pb and In-Sn-Pb solders will be discussed. Experimental thermomechanical fatigue tests results for these alloys will be presented as validation for the computational simulations. This work was performed at Sandia National Labs which was supported by the United States Department of Energy under Contract DE-AC04-94AL85000. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy.

9:20 am INVITED

MICROSTRUCTURAL CONTROL FOR IMPROVED SOLDER PROPERTIES: S. Jin, H. Mavoori, Bell Laboratories, Lucent Technologies, Murray Hill, NJ 07974

Technical barriers to the high-density circuit and device interconnection and packaging include the absence of reliable solder joint materials that will withstand high strain/stress conditions, for example imposed by severe thermal expansion mismatch. When the solder joint size is reduced in high-density packaging, the shear strain on solder is substantially increased, which tends to cause accelerated microstructural coarsening and deterioration in creep and fatigue resistance. Microstructural control of solder alloys, e.g., in grain size or the morphology, size and distribution of precipitate phases can be effective in improving the mechanical properties and solder joint reliability. In this talk, various approaches for solder alloy processing and microstructural modifications for improved mechanical properties will be discussed.

9:45 am INVITED

SINTERING OF SOLDER JOINTS: M.A. Palmer, C.N. Alexander, R.W. Messler, Jr., W.Z. Misiolek, Rensselaer Polytechnic Institute, Troy, NY 12180-3590

The electronics industry has developed a manufacturing infrastructure based on the properties of eutectic lead-tin solder, which has a liquidus of 183°C. There is interest in alternatives to this work-horse alloy, based both on concerns about the toxicity of lead, and the need for more demanding applications, where temperatures may exceed 150°C or 93% of Tm. The high melting point of such alternative alloys, typically in excess of 220°C, raises manufacturing concerns. Forming solder joints, with the alternative sintering process at lower temperatures is investigated in this work. The objective is to apply the new proposed process using available equipment for today's technology. In this presentation we will examine the sinterability of eutectic tin-lead solder paste, through several proof-of-concept experiments. The results of these experiments, and their implications will be discussed.

10:10 am BREAK

10:30 am INVITED

VOIDING IN SOLDER JOINTS SUBJECTED TO HIGH TEMPERATURE AGING: L.E. Felton, Foster Miller, Inc., 195 Bear Hill Road, Walthem, MA 02154-1196; T.-Y. Pan, H.D. Blair, J.M. Nicholson, Ford Motor Company, P.O. Box 2053, Building R, M/D 3135, Dearborn, MI 48121-2053

Higher application temperatures have been the trend of automotive military electronic packages. A high temperature limit of 160°C has been proposed as the specification for the next generation printed-wiring board assemblies. However, voids have been shown to form in the intermetallic compound layer on electroplated copper when aged at such a temperature. A metallurgical study on samples with 63Sn-37Pb and 96.5Sn-3.5Ag solders aged at 160°C for 30 days has shown that the extent of voiding can be associated with different brighteners and/or surfactants in the plating bath. This voiding is usually confined to the Cu₃Sn layer and may ultimately cause mechanical failure of the solder joints. With time this void formation at elevated temperatures can be as deleterious to joint integrity as thermal cycle fatigue.

10:55 am INVITED

CREEP DEFORMATION OF NON-COMPOSITE Sn-Ag EUTECTIC SOLDERS UNDER ISOTHERMAL AND THERMAL CYCLING CONDITIONS: K.N. Subramanian, J.L. McDougall, A.W. Gibson, S. Choi, T. R. Bieler, Department of Materials Science and Mechanics, Michigan State University, E. Lansing, MI 48824-1226

Under normal operating conditions, soldered components in an automobile experience cyclic stress relaxation due to differential thermal expansion. The resulting strain and damage conditions are similar to creep deformation. Sn-Ag eutectic solder with and without intermetallic reinforcements were subjected to creep testing under isothermal and cyclic temperature conditions. Miniature single shear lap specimens used in this study are similar in size to soldered connections found in automotive circuit boards, and were fabricated using a melt reflow process similar to industrial practice. Strain measurements were made by comparing micrographs taken with a digital camera periodically throughout the experiment. Overall strain and strain heterogenities were quantified. The results for composite solder creep are compared with data for Sn-Ag solder from the literature.

11:20 am

MICROMECHANICAL CHARACTERIZATION OF SOLDER JOINTS: J.P. Lucas, A. Gibson, T. Bieler, K.N. Subramanian, Department of Materials Science, and Mechanics, Michigan State University, East Lansing, MI 48824-1226

Solder joints are in-situ composites consisting of highly distinctive phases. The microstructure of solders evolves almost continuously as solder exists at fairly high homologous temperatures. Moreover, microstructural evolution is enhanced by aging and thermo/mechanical history. The mechanical performance of solder in service depends on the microconstituent phases present in the solder joints. To control microstructural evolution, attempts are being made to stabilize the microstructure in-situ using thermally-stable reinforcement phases. To characterize the mechanical behavior and assess the stability of the solder, nanoindentation testing will be performed. Mechanical properties of the constituent phases, such as the stress, strain rate, creep behavior, and hardness will be determined spatially close to and far from interphase boundaries, and related to the microstructural stability of solder under simulated service conditions.

ALLOY MODELING AND DESIGN: Session I: Surfaces, Interfaces, and Grain Boundaries in Alloys

Session Chair: Prof. J.S. Faulkner, Alloy Research Center and Department of Physics, Florida Atlantic University, Boca Raton, FL 33431

SOME PROPERTIES OF Cu-Au ALLOY SURFACES: R.G. Jordan, Alloy Research Center and Department of Physics, Florida Atlantic University, Boca Raton, FL 33431; G.Y. Guo, CCLRC Daresbury Laboratory, Warrington WA4 4AD, UK

Experimental and theoretical studies of the low index surfaces of Cu-Au (ordered) alloys reveal a wealth of interesting details. In this talk we highlight some of our more recent results and discuss their implications with particular reference to the electronic structure of the alloys in the surface region. We focus in particular on surface states and the dependence of surface core level shifts on orientation and segregation. Supported by the NSF (#DMR-9500654) and NATO (#CRG.910981).

9:10 am INVITED

TIGHT-BINDING AND CVM CALCULATION OF PROFILES OF APB ENERGIES IN Ni₃Al: T. Shinoda, Y. Mishima, P. & I. Laboratory, Tokyo Institute of Technology, Nagatsuta 4259, Midori-ku, Yokohama 227, Japan; K. Masuda-Jindo, Department of Materials Science and Engineering, Tokyo Institute of Technology, Nagatsuta 4259, Midori-ku, Yokohama 227, Japan

A tight-binding (TB) recursion scheme, coupled to LMTO method, is used to calculate the APB energy of Ni₃Al and NiAl alloys. The total energy of the system is given by a sum of band structure energy and ionic repulsion. Since the electrostatic contributions cancel out, what is needed to calculate the effective pair interaction (EPI) energies is an average over all configurations, with fixed occupancy of sites i and j, of the one-electron band-structure term. The electronic structures of the alloys are calculated by tridiagonalizing a TB Hamiltonian using the continued fraction technique. We estimate the effective pair (cluster) interaction energies (EPI) using the direct configurational averaging (DCA) method [1] over a small number of randomly generated configurations. The present calculation of equilibrium profiles of (111) APB energy in Ni₃Al alloys is based on the CVM with the tetrahedron cluster approximation, using the empirical pair interaction energies as well as those estimated by a tight-binding recursion scheme coupled to a LMTO method. Calculated APB energies, that is the sums of each overall profile are compared with the experimentally obtained values in their magnitudes. Some arguments are tried with respect to the superiority to each other of the two pair interaction energies.

9:50 am

PHOTOEMISSION FROM THE (111) SURFACES OF Cu-Au ALLOYS: L.R. Masliah, R.G. Jordan, S.L. Qiu, Alloy Research Center and Department of Physics, Florida Atlantic University, Boca Raton, FL 33431; B. Ginatempo, Dipartimento di Fisica and Unità di Ficerca INFM di Messina, Università di Messina, Salita Sperone 31, 98166 Messina, Italy

We have calculated the electronic structure at the (111) surfaces of ordered Cu-Au alloys using the LMTO method in a supercell geometry and including the surface dipole terms. In order to make quantitative comparisons with our photoemission measurements we have used the potential functions in relativistic photocurrent calculations. In this talk we concentrate specifically on comparisons between experimental spectra and photocurrent calculations as functions of photon energy and distinguish details of the 'surface' and 'bulk' electronic structures. Supported by the NSF (#DMR-9500654) and NATO (#CRG.940120).

10:20 am BREAK

10:40 am

ANISOTROPY OF SEGREGATION AT ANTIPHASE BOUNDARIES IN A BCC-BASED BINARY ALLOY: Qiang Wang, Long-Qing Chen, Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802

A computer simulation on the kinetics and anisotropy of segregation at antiphase boundaries (APB's) in a model BCC-based binary alloy has been conducted using microscopic master equations in the first and second-neighbor pair approximations. Both stoichiometric and nonstoichiometric compositions are considered. It is shown that the degree of segregation at the APB's is highly anisotropic and the mobility of an APB at a given temperature is strongly composition-dependent. For the particular case of a cylindrical APB along the z direction, it is found that maximum segregation occurs along the []-direction and essentially zero along the [1 1 0]-direction. Despite the strong segregation anisotropy, however, it is demonstrated that the decrease in the radius of the cylindrical antiphase domain is linearly proportional to t^1/2 where t is the time.

11:10 am

SEGREGATION OF NIOBIUM SOLUTE IN NICKEL TOWARD GRAIN BOUNDARIES AND FREE SURFACES: Leonid Muratov, Bernard R. Cooper, Department of Physics, West Virginia University, Box 6315, Morgantown, WV 26506

The spatial redistribution of niobium atoms near the (100) and (111) free surfaces and selected grain boundaries (GB) of pure nickel has been considered in the low niobium concentration limit. Our simulations show a significant depletion in concentration of niobium immediately at a free surface. However, under the first two layers of pure nickel there exists a niobium enriched region with strongly temperature dependent concentration. This predicted non-monotonic distribution of niobium in the surface region may be very important for many applications and requires experimental confirmation. In contrast, at the grain boundaries, the concentration of niobium, pertinent to GB oxidation embrittlement, is predicted to be much higher than in the bulk. It monotonicaly decreases with the distance from GB, until reaching the bulk value. The calculation of the free energy uses atomistic potentials based on ab-initio quantum mechanical calculations, includes lattice relaxation around niobium atoms by using molecular dynamics (with 1440 or 2880 atoms in the modeling cell), and includes vibrational entropy phenomenologically within the local harmonic approximation.

11:40 am

EFFECTS OF OFF-STOICHIOMETRY ON ANTIPHASE BOUNDARIES IN MODEL Ll₂ ALLOYS AND -TiAl PHASE: Patrick Tepesch, Mark Asta, Andrew Quong, Sandia National Laboratories, Livermore, CA 94550

The effect of off-stoichiometry on the properties of conservative (100) and (111) antiphase boundaries (APB) in model L1₂ alloys are studied with the cluster variation method (CVM) in the Tetrahedron-Octahedron (T-O) approximation with first (V₁) and second (V₂) nearest neighbor pair interactions. Two model systems with = V₂/V₁ = æ0.1 and æ1.0 are considered. The effect of off-stoichiometry on APB properties, such as the equilibrium () and non-equilibrium (°) antiphase boundary free energies, depends quantitatively and qualitatively on temperature and . First-principles results on the -TiAl phase show that these effects can be important in real systems. This work was supported by the U.S. Department of Energy, Office of Basic Energy Science, Division of Materials Science.

BOUNDARIES AND INTERFACES IN MATERIALS: THE DAVID A. SMITH SYMPOSIUM: Session I

Session Chair: William A.T. Clark, The Ohio State University, Columbus, OH 43210

OPENING REMARKS: David B. Williams, Department of Materials Science and Engineering, Lehigh University, Bethlehem, PA 18015

8:35 am INVITED

MICROSTRUCTURAL DESIGN AND TAILORING OF ADVANCED MATERIALS: Gareth Thomas, Department of Materials Science, University of California, 561Evans Hall, Berkeley, CA 94720

We live and work in a world that depends on materials and their performance. The science of metallurgy and its extension to materials in general has arrived over the past 50 years at a position of great understanding, so much so that we can now tailor the desired microstructure for scientific properties using appropriate processing and controls. From a scientific approach to materials technology, the relationships between processing, structures and performance must continue to be established. This is the main task for materials scientists. Progress in experimentation and instrumentation has been truly remarkable these past 50 years - we can even image columns of atoms and do atom-by-atom spectroscopy! The electron microscope with all its modern capabilities for high resolution imaging, diffraction and spectroscopy remains central to our means of understanding the relationships between structure and composition. It will be the focus of this lecture to discuss present limitations of high resolution TEM and to draw upon examples of the role microscopy plays in materials engineering. In almost all cases, the problems relate to interfaces. Examples include development of cold formable, corrosion resistant low-carbon dual phase steels for concrete reinforcement; structural and electronic ceramics (including a new approach to joining high temperature ceramics); and a wide range of magnetic materials, involving nanostructures.

9:00 am INVITED

GRAIN BOUNDARY ARCHITECTURE FOR HIGH PERFORMANCE MATERIALS: T. Watanabe, Laboratory of Materials Design and Interface Engineering, Faculty of Engineering, Tohoku University, Sendai, Japan

Grain boundaries are important microstructural elements and a source of high-performance in polycrystalline materials. Because of their structural and configurational versatility and flexibility, grain boundaries can produce a large variety of effects on bulk properties of materials. The concept of "Grain Boundary Design and Control for High-Performance Materials" was proposed in 1984 on the basis of the relationship between grain boundary structure and properties. A new microstructural factor termed "Grain Boundary Character Distribution (GBCD)" has been introduced. The manipulation of GBCD and other factors enables us to design and develop high performance materials by controlling detrimental effects and by enhancing their beneficial grain boundaries. The concept has been proved to be applicable and powerful in designing and producing high-performance materials in connection with processing. An overview of recent achievements of Grain Boundary Architecture and prospect for the future will be given.

9:30 am INVITED

THERMAL PROPERTIES OF NANOSTRUCTURED MATERIALS: U. Erb, T. Turi¹, B. Szpunar⁺, G. Palumbo² and K.T. Aust, Department of Metallurgy and Materials Science, University of Toronto, Toronto, Canada M5S 3E4; ¹Department of Materials and Metallurgical Engineering, Queen's University, Kingston, Ontario, Canada K7L 3N6; ²Ontario Hydro Technologies, 800 Kipling Avenue, Toronto, Canada M8Z 5S4

The effect of grain boundary volume fraction on thermal expansion, specific heat, diffusion, and thermal stability of fully dense nanstructured materials produced by electrodeposition will be presented. This will be followed by a comparison with property measurements performed on materials produced by other synthesis methods. The variation in properties observed for materials produced by various synthesis methods will be discussed in terms of microstructural differences (i.e. grain boundary structure, porosity, impurity content, texture).

10:00 am BREAK

10:10 am INVITED

GRAIN BOUNDARY STRUCTURE IN NON-CUBIC CRYSTAL SYSTEMS: R.C. Pond, F. Sarrazit, M. Aindow*, Department of Materials Science and Engineering, University of Liverpool, Liverpool L69 3BX, UK; *School of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

Currently, the most widely accepted model of interfacial structure is based on the notion of reference configurations, which exhibit low free energy, and where any additional angular deviation and/or misfit present is accommodated by arrays of interfacial defects. This model has been extensively validated for the cubic crystal system, but further considerations may be necessary in other cases. This is illustrated in the present paper by experimental observations, using transmission electron microscopy, of interfaces in -Ti and ZnO. In the latter case, for example, a transition was observed from a facetted interfacial structure consistent with the above model to an alternative planar form based on a reference structure which was quasicrystalline in one dimension.

10:40 am

A GEOMETRIC CRITERION FOR "SPECIAL" CSL GRAIN BOUNDARY PROPERTIES: G. Palumbo^1,2 K.T. Aust², E.M. Lehockey¹, U. Erb², P. Lin¹, ¹Ontario Hydro Technologies, 800 Kipling Avenue, Toronto, Canada M8Z 5S4; ²Department of Metallurgy and Materials Science, University of Toronto, Toronto, Canada M5S 3E4

Although widely used for the characterization of grain boundaries within the coincidence site lattice (CSL) framework, the application of Brandon's criterion (for allowable angular deviation), has often resulted in grain boundaries which do not display "special" behavior being classified as low S CSL's; thus prompting some debate as to the overall validity of the CSL model. In this present work, the crystallographic basis for a previously formulated, more restrictive criterion for "special" boundaries is presented. The applicability of this criterion is shown to be supported by (1) a review of TEM evidence for the resolution of discrete intrinsic grain boundary dislocations, and (2) relative grain boundary performance in numerous experiments for corrosion, cracking, and creep cavitation susceptibility of several pure metals and commercial alloys.

11:00 am

RODRIGUES-FRANK MAPPING OF INTERFACE CRYSTALLOGRAPHY: Krishna Rajan, Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180-3590

The utility of Rodrigues-Frank vectors in crystallographically representing grain boundary misorientation will be described in this presentation. It will be shown that Rodrigues-Frank (R-F) mapping of microstructures is in fact well suited to experimental techniques which characterize grain specific orientation. R-F representations also permit the coupling of grain boundary misorientation associated with crystal symmetry to issues associated with sample symmetry. The value of such representations in problems ranging from deformation processing to thin film growth will be discussed.

11:20 am

ENGINEERING OF GRAIN BOUNDARIES IN LEAD ALLOYS FOR IMPROVED Pb-ACID BATTERY PERFORMANCE: E.M. Lehockey, G. Palumbo, A.M. Brennenstuhl, P. Lin, Ontario Hydro Technologies, 800 Kipling Avenue, Toronto, Canada M8Z 5S4

Lead-acid battery life is limited by intergranular degradation of the Pb alloy electrodes, including: corrosion (grain dropping), cracking, and grain boundary sliding (creep). Efforts to improve the resistance of the electrodes to these effects have focused on altering the microstructure to contain high frequencies of grain boundaries at/near low-S CSL misorientations. Elevating the frequency of these "special" interfaces from 12% in conventional material to 60% reduces weight loss (corrosion) and growth (creep) of PbCaSn electrodes, measured in industry standard tests, by 30% and 70% respectively. Similar processing increases the number of charge-discharge cycles to failure in PbSb alloy electrodes by 5-fold, arising from crack blunting at low-S CSL boundaries. These results clearly advocate a role for "grain boundary engineering" in developing advanced stationary and SLI batteries with superior reliability and energy density.

11:40 am

EVOLUTION OF THE FCC/L1₂ INTERFACE DURING DIFFUSION-BONDING: K. Fujiwara, Y. Ootoshi, Z. Horita, M. Nemoto, Department of Materials Science and Engineering, Faculty of Engineering 36, Kyushu University, Fukuoka 812-81, Japan

This study presents diffusion-couple experiments where Ni is bonded with Ni-based intermetallics having an L1₂-type crystal structure such as Ni₃X (X=Al, Si, Ga and Ge). Microstructural observations and selected area electron diffraction analyses have revealed that the Ni-rich solid solution phase (-phase) grows towards the intermetallic phase (-phase) during diffusion bonding. There exists an identical orientation relationship between the and phases when the annealing is conducted at higher temperatures but no such relationship is present when the couple is annealed at lower temperatures. It is then suggested that there is a critical temperature for the presence of the identical orientation relationship between the and phases. The critical temperature depends on the types of diffusion-couples and it appears that the lattice misfit is an important factor in determining the critical temperature.

12:00 pm INVITED

THE STRUCTURE OF NON-SPECIAL BOUNDARIES IN ALUMINUM: C.L. Briant, R. Bai, R. Phillips, V. Shinoy, A. Schwartzman, Division of Engineering, Brown University, Providence, RI 02912

Most studies of the structure of grain boundaries in metals have focused on special boundaries that have a high number of coincident lattice sites. Although these studies have yielded extremely valuable information, these boundaries are not the ones most often encountered in engineering practice. In this work we report on the structure of grain boundaries that are non-special in their orientation. Tilt boundaries of aluminum were prepared either by strain annealing or by growing bi-crystals from the melt. High resolution transmission electron microscopy images of these boundaries were obtained; all boundaries were of the <110> type. The structures of these boundaries were determined by high resolution TEM and were compared with structures simulated by the quasi-continuum method. The results of these studies will be discussed in terms of different grain boundary structure models. A discussion will also be presented on the use of the quasi-continuum method to simulate both the motion of grain boundaries and the response of grain boundaries to a mechanical stress. This work was sponsored by DOE contract DE-FG02-96ER45578.

DEFORMATION AND MICROSTRUCTURE: General Abstract Session I

Session Chair: Sriram Seshagiri, Wright Patterson Air Force Base, OH

CHARACTERIZING SUPERDISLOCATIONS IN Ll₂ STRUCTURES WITH TEM OBSERVATIONS AND IMAGE SIMULATIONS: EXAMPLES WITH APB- AND SISF DISSOCIATIONS: Mukul Kumar, Kevin J. Hemker, Dept. of Mechanical Engineering, John Hopkins University, Baltimore, MD 21218-2686

The nature of dissociated superlattice dislocation cores in Ll₂-modified Al₆₆Ti₂₅Cr₉, deformed at room temperature, has been characterized by weak-beam transmission electron microscopy and image matching by computer simulation techniques. The displacement fields associated with APB- and SISF- dissociated <110> superdislocations were calculated to account for the asymmetry in dislocation contrast, and to provide a better understanding of the formation of images from the cores of narrowly dissociated superdislocations. These techniques have been utilized in the present study to enhance the descriptive and quantitative information gathered from experimental weak-beam TEM observations of dissociated superdislocations prevalent on compressive deformation of cubic Al₃Ti at room temperature. The results indicate that suitable caution must be exercised in the interpretation of weak beam images from dislocations with large Burgers vectors, particularly when the spacing between the individual partial dislocations is on the order of a few nanometers, and small deviations from exact Bragg condition are employed.

8:50 am

DISLOCATION SUBSTRUCTURE IN NiAl SINGLE CRYSTALS: X. Shi, T.M. Pollock, S. Mahajan, V.S. Arunachalam, Department of Materials Science & Engineering, Carnegie Mellon University, Pittsburgh, PA 15213

NiAl single crystals, oriented for single slip, were deformed in compression over a range of temperatures (77K-673K). The resulting dislocation structures were examined using weak-beam transmission electron microscopy. Results indicate that the activated slip system is [001](110). The substructure consists mainly of near-edge dislocations and their density decreases with increasing temperature. These observations suggest that the edge dislocations have a low mobility in comparison to the screw dislocations. Another salient feature of the dislocation structure is the presence of a high density of jogs on both near edge and near screw zigzag dislocations. In this presentation, we will discuss the origins of the observed substructural features and comment on the relative mobilities of the edge and screw dislocations. The above work was supported by Dr. G. Yoder of ONR and the authors gratefully acknowledge the support.

9:10 am

USE OF THE ELECTRON CHANNELING CONTRAST IMAGING (ECCI) TECHNIQUE TO IMAGE DISLOCATIONS NEAR CRACK TIPS AND CRACK EDGES IN BULK SAMPLES: B-C. Ng, B.A. Simkin, M.A. Crimp, Department of Materials Science and Mechanics, Michigan State University, East Lansing, MI 48824-1226

The electron channeling contrast imaging (ECCI) technique has been developed in recent years to image near surface crystalline defects. This talk will review the basis of ECCI image formation and then discuss application of the technique to studies of fracture behavior. In our present studies, the technique is being used to image dislocations near crack tips and crack edges of bulk specimens. Examination of stable cracks shows that dislocations are generated ahead of the crack tips resulting in blunting of the cracks. Dislocations are also observed along the crack edges, left in the wake of the propagating cracks. Changes in dislocation density ahead of cracks have been correlated with material toughness. Interpretation of the images as a function of channeling conditions will be discussed. For example, the contrast of the dislocation images changes when the channeling conditions change from a positive (+) g vector to negative (-) g vector. This change of contrast channeling condition also serve to distinguish between dislocations and microcracks. This work was supported by the Office of Naval Research (Grant No. N00014-94-1-0204) and the National Science Foundation (Grant No. DMR#9257826).

9:30 am

STUDY OF DISLOCATION MOTION IN NiAl SINGLE CRYSTALS: B. Ghosh, M.A. Crimp, Department of Materials Science and Mechanics, College of Engineering, Michigan State University, East Lansing, MI 48824-1226

Dislocation mobility in commercial and high purity stoichiometric NiAl single crystals has been studied by in-situ straining transmission electron microscopy. Pre-existing dislocations, whether isolated or tangled, did not move at any time, even through the initiation and propagation of cracks. Dislocations generated upon straining glide past/through pre-existing dislocations. Images of mobile dislocations were captured dynamically by a wide angle camera coupled with a video cassette recorder. Straight slip traces left by the dislocations suggest easy glide and wavy slip traces present in the wake of dislocations indicate cross-slip of screw dislocations. Dislocation motion was found to be much slower in commercially pure crystals than in high purity crystals, suggesting solute drag. This retarded motion was characterized by small jumps of portions of individual dislocations, implying motion through kink migration. In hard oriented <001> crystals, <110> or <001> dislocations were found to be mobile on {110}planes. In the soft <110> orientation, <111> and <110> dislocations were observed to be mobile on {121}planes. This research has been supported by the Office of Naval Research (Grant # N 00014-94-1-0204).

9:50 am BREAK

10:00 am

AN ANALYSIS OF THERMALLY ACTIVATED DEFORMATION IN B₂ ALUMINIDES: T.M. Pollock, D. Lu, K. Eow, Carnegie Mellon University, Pittsburgh, PA 15213

Intermetallics with the B2 structure have recently been of interest for high temperature structural applications. While the limited ductility of many of these compounds is of concern, it is difficult to identify solutions to this problem, since ductility is a material property that is very sensitive to the details of the microstructure and processing. In this study, a variety of B2 aluminides have been investigated with the use of strain rate change experiments, in order to obtain phenomenological deformation parameters which may be capable of distinguishing intrinsic vs. extrinsic and short-range vs. long range barriers to deformation. Materials investigated to date include: RuAl, Ru₅₂Al₄₈, RuAl+0.5%B, Ru₅₃Al₄₇+0.5%B, Ru_50.5Al_44.5Cr₅, Fe₆₀Al₄₀+0.2%B, and single crystal and polycrystalline NiAl. The results of deformation experiments and accompanying electron microscopy studies will be presented.

10:20 am

CORRELATING YIELDING BEHAVIOR WITH DISLOCATION CORE STRUCTURES IN Ni₃GeFe₃Ge INTERMETALLIC ALLOYS: Mukul Kumar, T. John Balk, Kevin J. Hemker, Dept. of Mechanical Engineering, Johns Hopkins University, Baltimore, MD 21218-2686

The transition from positive to negative temperature dependence of 0.2% yield stress is investigated in the model pseudo-binary Ni₃Ge-Fe₃Ge system. Ni₃Ge and Fe₃Ge both L1₂ intermetallic alloys-show complete solid solubility as Fe is continuously substituted for Ni across the composition range. However, Ni₃Ge exhibits the yield stress anomaly, whereas the yield stress of Fe₃Ge shows a normal decline with temperature. Mechanical testing has verified this behavior, with the anomalous behavior gradually disappearing with increasing Fe content. It is proposed, and will be shown through weak-beam observations, that this transition is related to changes in the core structure of dissociated superdislocations. Results from mechanical testing will be presented and correlated with TEM observations of deformation structures. These results will be discussed in the light of planar fault energies calculated through computer simulations of images.

EXTRACTION AND PROCESSING OF TITANIUM: Session I: Titanium Recovery Processes

Session Chairs: Dr. B. Mishra, Department of Metallurgical & Materials Engineering, Colorado School of Mines, Golden, CO 80401; Mr. S. Daniel, Oremet Titanium, 530 W. 34th Avenue, P.O. Box 580, Albany, OR 97321

HIGH PURITY TITANIUM REFINING BY IODIDE PURIFICATION: R.K.F. Lam, Materials Research Corporation, Advanced Materials Division, 542 Route 303, Orangeburg, NY 10962

An improved iodide process has been developed to purify titanium metal to high purity. The process consists of (1) in-situ generation of titanium tetraiodide from crude titanium, (2) purification of titanium tetraiodide by distillation, and (3) decomposition of purified titanium tetraiodide to high purity titanium metal. Purity level of titanium metal is increased from 99.5% to 99.998%. Titanium tetraiodide is purified to 99.999% purity. Thermodynamic properties of titanium tetraiodide and impurity iodides are discussed. Properties of refined titanium metal and titanium tetraiodide are investigated. Automatic control was employed for maintaining reactor temperatures, chamber pressures, and electric power levels for heating up deposition surfaces. Experimental data and purity levels are compared with those reported by other researchers.

9:10 am

RECOVERY OF TITANIUM FROM BLAST FURNACE SLAGS: Z. Sui, N. Fu, Y. Zhang, School of Materials and Metallurgy, Northeastern University, Shenyang 110006, China

China bounds in mineral resources of iron, but some of that are composite mineral ores containing valuable nonferrous metals components such as Ti-V bearing magnetite ore in Southwest part China-Panzhihua. Due to complex mineralogy, very fine mineral dissemination and low grade, the Titanium Components were mostly concentrated in molten slags and separated from liquid pig -iron, when the ore was usually treated in Blast-furnace. So slags contained titanium became an important man-made resources, from that the titanium components could be recovered by the technique based on the precipitation selectivity. It was indicated by a series of experiments that the recovery efficiency of titanium components from the slags was directly related to the proper precipitating feature of Titanium component in the cooled slags, such as morphology, structure, phase composition, crystallized state, grain size and dispersity, which were obviously depended on the operation factors like the slags composition, temperature of heat-treatment, cooling rate and additive agent. Based on the optimum factors, the precipitating behavior of Titanium component from the slags could be artificially controlled by designing the operation conditions, which were investigated by a technique of TEM observation in situ of heating sample and then the thermodynamic analysis and the computer simulation combined with fractal description as auxiliary. The selective precipitation of Titanium bearing slags was studied as example, by which both available research method for studying on the precipitating selectivity of Boron and Rare Earth metals components from slags were also summarized.

9:35 am

SPECIATION OF TITANIUM DURING AND AFTER PROCESSING OF TUNGSTEN CARBIDE SCRAP FOR TUNGSTEN AND COBALT: R.P. Singh, M.J. Miller, Chemical Development Department, R&D Division, Chemical & Metallurgical Products, Osram Sylvania, Inc., Hawes Street, Towanda, PA 18848-0504

Tungsten carbide scarp contains 1 to 2 % titanium. At present during hydrometallurgical processing of tungsten carbide scrap, the recovery of titanium is not considered economically viable. However, strict environmental regulations and/or diminishing titanium sources may change all that in future. This paper would discuss the speciation of titanium during and after processing of tungsten carbide scarp for tungsten and cobalt recovery. Complete characterization of titanium-containing species, which is of fundamental importance in the development of extraction and processing methods of titanium recovery from tungsten carbide scrap, will be reported.

10:00 am BREAK

10:15 am

TITANIUM & TITANIUM ALLOYS RESEARCH AT THE INSTITUTE FOR NON-FERROUS AND RARE METALS, BUCHAREST, ROMANIA: T. Segaroeanu, Gh. Busila, V. Secanu, Gh. Morcan, L. Dumitru, St. Ivanescu, S. Dinescu, D. Capac, Institute for Non-Ferrous and Rare Metals, Bd. Biruintel 102, Sector 2, Bucharest, Romania

The research work performed within the Institute for Non-ferrous and Rare Metals of Bucharest, Romania was focused on the development of technologies for obtaining titanium from domestic raw materials, producing different titanium alloys and, also, on turning of the titanium and titanium alloys in to semi-finished products. For the titanium industry, we have produced complete technologies to process titanium slags through chlorination, purification and magnesiothermic reduction of the titanium chloride. Titanium and titanium alloys ingots were obtained by electron beam melting and by consumable electrode arc melting of the titanium sponge and pre-alloyed materials. Our research has continued with the processing of these ingots in rods, sheets, tubes, wires, and cast products. According to the Romanian industry needs, the research work in the field of titanium and titanium alloys was directed to the applications for medical use, chemical industry, naval applications and aircraft buildings. At the same time, we have developed titanium metal scrap processing technologies through molten salt electrolysis and we have obtained titanium metal powders. We have, also, obtained titanium metal powders via a hydrogenation-dehydrogenation process. Recently, our Institute has developed research programs in the areas of new corrosion resistant titanium alloys, intermetallic compounds and titanium-matrix composites.

10:40 am

THE CHEMICAL BASIS OF A NOVEL FLUORIDE ROUTE TO METALLIC TITANIUM: J. Besida, T.A. O'Donnell, T.K. Pong, D.G. Wood, Department of Chemical Engineering, University of Melbourne, Parkville, Victoria 3052, Australia

The conventional route to metallic titanium, the Kroll Process, involves reduction of TiCl₄ with magnesium and produces a sponge which is difficult to recover and is heavily contaminated with reaction products and excess reactant. After recovery, the sponge requires very costly vacuum arc refining to produce useable billets. Some earlier proposed fluoride routes also lead to sponge formation. In the present work K₂TiF₆, which unlike TiCl₄ is not air-or moisture-sensitive, is dissolved in molten cryolite, and has been shown at gram-scale experimental level to be reduced by metallic Al to produce a powder of metallic Ti which is about 99% pure and can be recovered as a free-flowing product. NaF is added in stoichiometric amount during the reaction to preserve the integrity of the liquid cryolite medium according to the overall equation: 3Na₂TiF₆ + 4A1 + 6NaF ~ 4Na₃AlF₆ + 3Ti.

11:05 am

CONTINUOUS PRODUCTION OF TITANIUM POWDER: S.J. Gerdemann, L.L. Oden, J.C. White, Department of Energy, Albany Research Center, 1450 Queen Avenue, S.W., Albany, OR 97321-2198

Although incremental improvements have been made to the Kroll process since its inception in 1948, the process in use today remains essentially the same batch process developed by Dr. Kroll and perfected by the U. S. Bureau of Mines. In this process, titanium tetrachloride (TiCl₄) is reduced by magnesium to produce titanium metal. There are two major limitations to the Kroll process: 1) it is a batch process and 2) the titanium sponge produced must undergo several purification steps before the metal is suitable for use. During the reduction of TiCl₄ with either magnesium (Kroll process) or sodium (Hunter process) the reaction proceeds so rapidly that the titanium metal sponge formed is an interlocking dendritic mass with inclusions of magnesium or sodium salts. The Albany Research Center (ALRC) is investigating a new, continuous titanium metal production process in which a titanium powder is produced in a bath of molten salt. In this process the rate of the reduction reaction is slowed and controlled by diluting the reactants with molten chloride salts. The diluted reactant streams are combined in a continuous stirred tank reactor, operated much like a crystallizer. New titanium metal forms on the already present small Ti particles and when the Ti particles become so large that they can no longer be suspended in solution, they fall to the bottom of the reactor and are removed. Initial experiments show considerable promise but problems remain in obtaining the purity and uniform particle size required.

FATIGUE AND CREEP OF COMPOSITE MATERIALS: Session I

Session Chairs: James M. Larsen, Wright Laboratory Materials Directorate, Wright-Patterson AFB, OH 45433; Paul Bowen, School of Metallurgy and Materials/IRC in Materials for High Performance Applications, The University of Birmingham, UK

MODELLING, CHARACTERISATION AND ASSESSMENT OF SELECTIVELY REINFORCED Ti MMCs: P. Bowen, N. Wang⁺, T.J.A. Doel, A.L. Dore and D.C. Cardona*, School of Metallurgy and Materials/IRC in Materials for High Performance Applications, The University of Birmingham, UK; ⁺British Steel plc., *Rolls-Royce plc

Selectively reinforced (clad) test pieces and model rings can show radically different failure modes compared with uniformly reinforced materials of similar volume fraction. This behaviour has been quantified and modelled as a function of cladding thickness and fibre volume fraction. Extensive transverse damage may often occur for fibre volume fractions of commercial interest at the first reinforced plane ahead of a mode I crack growing in the monolithic cladding layer. This damage may promote premature testpiece failure and prevents crack-tip shielding due to fibre bridging. Such observations have been made both for testpieces in bending and tension. The damage can be quantified by considering the magnitude and position of transverse stresses ahead of the growing crack-tip relative to those stresses that promote such damage in uniformly reinforced composites subjected to transverse loading. To characterise such modes of failure has necessitated extensive experimental and modelling studies. A dimensionless parameter has been developed in analytical form that predicts transverse stresses for a range of loading, testpiece geometry (cladding thickness and cladding: composite thickness ratios) and levels of selective reinforcement. When combined with experimental observations of damage initiation this parameter can define the propensity of a given geometry to transverse damage.

9:00 am INVITED

EFFECT OF IN-SITU MATERIAL PROPERTIES ON FATIGUE DAMAGE MODES IN TITANIUM MATRIX COMPOSITES: David Harmon¹, Kenneth L. Jerina², ¹McDonnell Douglas, St. Louis MO; ²Washington University, St. Louis MO

Titanium matrix composites (TMC) and their behavior under mechanical fatigue loads was the subject of this research. The primary objective was to explain fatigue damage modes in center notched TMC specimens. Two modes of damage have been observed in continuously reinforced, 0 degree unidirectional, SCS-6/Ti-15V-3Cr-3Al-3Sn (SCS-6/Ti-15-3) laminates. The fatigue specimens were destructively analyzed using optical microscopy to determine where cracks originated and how they grew throughout the specimen. A micromechanical model was developed to explain the fatigue crack patterns observed in the interface region surrounding the fibers of the woven and acrylic binder TMC material systems. A two dimensional model of a longitudinal lamina with a center hole was used to obtain a set of displacement boundary conditions for an element near the notch yet within the net section where the spiral crack patterns were observed. These boundary conditions were then used on a 3D unit cell model of the fiber, matrix, and interface.

9:20 am INVITED

ISOTHERMAL AND THERMO-MECHANICAL FATIGUE LIFE PREDICTION OF A TITANIUM MATRIX COMPOSITE, D.C. Slavik¹ and A.H. Rosenberger², ¹General Electric Aircraft Engines, Cincinnati, OH 45150, ²Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433

Higher costs associated with titanium matrix composites (TMCs) must be offset by significant advantages in material weight and/or performance prior to use. Life prediction models for TMCs are typically empirical and require a substantial testing program before the benefits of a new material can be fully assessed. Alternatively, models that rely on only the constituent stresses within the composite and available monolithic life prediction tools can be used for preliminary design without a substantial composite testing program. Fatigue life calculations using the local stresses and the constituent properties are evaluated and compared with composite fatigue data obtained for a continuous fiber SCS-6/Ti-6Al-4V system. Local stresses, calculated from the processing history and applied stress using a concentric cylinder model, are input into life prediction tools that include time-independent fatigue and time-dependent damage modes. These predictions are compared to experimentally measured fatigue lives as a function of test frequency (0.01 Hz to 10 Hz), temperature (23°C to 427°C), and thermo-mechanical temperature history (in-phase and out-of-phase). Life predictions, model limitations, and areas for model improvements are discussed.

9:40 am INVITED

A GOODMAN APPROACH FOR TITANIUM MATRIX COMPOSITES SUBJECTED TO HIGH CYCLE FATIGUE: D.P. Walls¹, T.E. Steyer², F.W. Zok², United Technologies, ¹Pratt & Whitney, West Palm Beach, FL; ²Materials Department, University of California, Santa Barbara, CA

A study has been conducted on high cycle fatigue in fiber reinforced titanium matrix composites. Under this type of loading, the composites generally exhibit the initiation and propagation of distributed matrix cracks and an associated stiffness loss and permanent residual strain. Additionally, the constitutive response of the composite exhibits stress/strain hysteresis. Continued cycling may eventually lead to fracture. Results of such observations for various load amplitudes and stress ratios are presented for a Ti-6Al-4V/SCS-6 system. A methodology for incorporating such information into a conventional Goodman diagram is proposed, as a technique to provide preliminary design guidelines.

10:00 am BREAK

10:20 am INVITED

DEGRADATION IN THE MECHANICAL PROPERTIES OF FIBER-REINFORCED TITANIUM MATRIX COMPOSITES UNDER FATIGUE LOADING, F.W. Zok, T.E. Steyer, S.J. Connell, Materials Department, University of California, Santa Barbara, CA; D.P Walls, United Technologies, Pratt & Whitney, West Palm Beach, FL

Fiber-reinforced titanium matrix composites (TMCs) exhibit a variety of damage processes under cyclic loading, including the formation and propagation of matrix cracks, bridging of the cracks by intact fibers, and debonding and sliding along the fiber-matrix interface. These processes are manifested in several ways, including the development of hysteresis in the stress-strain response, a loss in the elastic modulus, the development of permanent strain, and a reduction in fiber strength associated with fretting of the fiber surfaces. The present paper will address two aspects of this problem. The first involves the growth of a single dominant crack from a sharp notch. Experiments have been performed to directly measure both the interface sliding stress and the fiber strength within the bridged zones and the measurements used to simulate the fatigue crack growth characteristics and the fiber failure threshold. The second deals with the growth of multiple cracks in straight (unnotched) specimens and the effects of the cracks on the subsequent elastic and inelastic deformation characteristics. Micromechanical models for the behavior of bridged cracks of finite length will be presented and compared with the experimental data.

10:40 am INVITED

EFFECTS OF MATRIX MICROSTRUCTURE ON THE FATIGUE CRACK GROWTH RESISTANCE AND THE TOTAL LIFE OF SiC FIBRE REINFORCED Ti MMCS: P. McDonnell, S.V. Sweby, C. Barney, P. Bowen, School of Metallurgy and Materials/IRC in Materials for High Performance Applications, The University of Birmingham, UK

For the composite systems based on Ti-6Al-4V or IMI 834 matrix alloys, and reinforced with Sigma SM1140+ or SM1240 silicon carbide fibres, post processing heat-treatments have been carried out to promote increased matrix fatigue crack growth resistance. In this respect, near and fully matrix microstructures have been shown to improve crack arrest/catastrophic failure limits in the presence of an unbridged defect, and without any decrease in the total life (S-N) performance of such composites. However, because the transus temperature can be increased by approximately 50°C for the matrix present in the composite compared with that expected for the monolithic alloy alone, great care is required if these heat-treatments are not to result in a degradation of fibre strength. If the fibre strength is degraded, then severe reductions in both crack arrest/catastrophic failure limits and total life will result. Of interest, the severity of matrix-fibre interfacial reactions changes markedly for the different fibre-matrix systems considered, and hence different heat-treatment windows can be defined for specific fibre-matrix combinations. These observations will be quantified and discussed in detail and with particular reference to fibre fracture strength distributions and fibre-matrix interfacial strengths.

11:00 am INVITED

FIBER/MATRIX INTERFACIAL DEGRADATION IN SCS-6/TIMETAL^®21S COMPOSITE UNDER HIGH TEMPERATURE FATIGUE: J.L. Moran*, J. M. Larsen, J.R. Jira, M.L. Gambone, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433-7817; *Ball Aerospace & Technologies Corp., Fairborn, OH 45324

Fatigue crack growth in unidirectional titanium matrix composites containing the Textron carbon-coated SiC fiber, SCS-6, has been the subject of numerous investigations. It is now well known that bridging of Mode I matrix fatigue cracks by unbroken fibers has a major influence on crack growth rates at room temperature, particularly at low levels of applied stress. The relatively weak fiber/matrix interface in these materials promotes crack bridging by permitting interfacial sliding. At elevated temperatures, however, degradation of the fiber/matrix interfaces may be extensive, leading to major reductions in the contribution of crack bridging. The phenomenon of degradation of fiber/matrix interfaces in high-temperature fatigue crack growth was examined in tests of SCS-6/Timetal^®21S composite. Acoustic and optical microscopy were used to define the extent of interfacial degradation, and fiber push-out experiments were performed to quantify the effects of the degradation on fiber/matrix interfacial shear strength. Throughout the degraded region, the interfacial shear strength was reduced to near-zero values, while the shear strength beyond the degradation front was equivalent to that of the composite in its as-received condition. These findings are discussed with respect to material performance and life prediction under realistic usage conditions.

11:20 am INVITED

SURFACE-CRACK GROWTH IN A CONTINUOUSLY REINFORCED METAL MATRIX COMPOSITE: J.R. Jira, R. John, J.M. Larsen, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433; *University of Dayton Research Institute, Dayton, OH 45469

Turbine engine and aircraft components fabricated from continuous fiber reinforced titanium-alloy matrix composites (TMC) will experience cyclic loads during service and fatigue crack initiation and growth are expected to control component life. Hence, characterization of the fatigue crack growth behavior of a model TMC was initiated by the USAF under the Metal Matrix Composite Life Prediction Cooperative Program. The model TMC system chosen by the Cooperative was [0]24 SCS-6/Ti-6Al-4V. The results of the experimental investigation of propagation of part-through surface cracks in this material will be presented. Automated fatigue crack growth tests were conducted at a maximum stress level of 600 MPa at 23, 177 and 316°C with stress ratios of 0.1, 0.5 and 0.7. Fully bridged and partially bridged crack growth was observed. Crack extension was monitored using direct current electric potential and optical measurements. During some of the tests, the crack opening displacement profile was measured using the laser interferometric displacement gage system. Some of the tests were stopped prior to failure to determine crack shape. Results will be compared to through-crack behavior and implications for component design will be discussed.

11:40 am INVITED

FATIGUE CRACK GROWTH OF TI-MATRIX COMPOSITES WITH SPATIALLY VARIED INTERFACES: Benji Maruyama, Sunil Warrier*, NIST/Wright Laboratory, 2230 10th ST STE 1, Wright-Patterson Air Force Base, OH 45433; *Universal Energy Systems, 4401 Dayton-Xenia Rd, Dayton, OH 45432

Spatially Varied Interfaces is a design concept for composite synthesis whereby the interface mechanical response is tailored to the composite needs by varying the interface properties in patterns of weak and strong areas. In the SiC_f/Ti-alloy system, longitudinal fatigue crack growth experiments where longitudinal striped areas along the interface are systematically weakened, and the perturbation of the failure process is measured to gain a better understanding of the stress states and interface failure mechanisms.

GEORGE R. IRWIN SYMPOSIUM ON CLEAVAGE FRACTURE: Session I: Fracture Mechanics

Session Chairs: Kwai S. Chan, Southwest Research Institute, San Antonio, TX 78238; John D. Landes, Department of Mechanical and Aerospace Engineering, University of Tennessee, Knoxville, TN

OPENING REMARKS: Kwai S. Chan, Southwest Research Institute, San Antonio, TX 78238

8:30 am INVITED

BRITTLE CLEAVAGE FRACTURE: AN OVERVIEW OF SOME HISTORICAL ASPECTS: Paul C. Paris, Department of Mechanical Engineering, Washington University, St. Louis, MO 63130

Many aspects of cleavage fracture were not well understood by the mid-1950s when George R. Irwin coined the term "Fracture Mechanics." This discussion will attempt to relate some of the common concepts and misconceptions of that period. The leadership of Dr. Irwin in developing new concepts and methods for clarifying the understanding of "brittle fracture" will be emphasized here. An essential part of the progress made was to bring together the Engineering Mechanics, Metallurgy, and Physics of the subject into a reasoned balance in trying to further understand the nature of the phenomena. This is illustrated by recalling some of the "personalities" of that time period and their concepts and claims and how they interacted to make progress toward today's state of knowledge. Indeed today's closer ties between fields such as mechanics and metallurgy were in part caused by these requirements of analysis of "BRITTLE CLEAVAGE FRACTURE." We should all be thankful that Irwin was around to bring us together!

9:00 am

INTRODUCTION OF PROFESSOR GEORGE R. IRWIN: Professor Frank McClintock, Department of Mechanical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307

9:10 am INVITED

MICROSTRUCTURAL MECHANICS DESCRIPTION OF CLEAVAGE FRACTURING: R.W. Armstrong, G.R. Irwin, X.J. Zhang, University of Maryland, College Park, MD 20742

Grain size and particle size/distribution influences are considered in relation to the ductile-brittle transition and cleavage fracture toughness properties of steel materials. Lower ductile-brittle transition temperatures and greater fracture toughnesses are obtained at smaller grain sizes because of the greater microstructural stress intensities required for cleavage fracturing as compared with plastic yielding. Stereosection fractographic observations have revealed the spread of cleavage from single particle fractures and, more interestingly, from local hole-joining failures at particle clumps where rapid load transfer is reasoned to produce the necessary stress elevations for transition to cleavage.

9:40 am

INTERPRETATION OF RIVER LINE STEPS ASSOCIATED WITH THE GROWTH OF CRACKS: Derek Hull, Dept. of Materials Science & Engr., University of Liverpool, Liverpool L69 3BX, United Kingdom

The development of river line steps on fracture surfaces is common to cleavage in crystalline solids and to the fracture of amorphous and semi-crystalline solids. A brief review is given of the development of ideas about river lines, with particular reference to work on crystalline cleavage and the interpretation of river line patterns in terms of the intersection of cracks with arrays of screw dislocations. Topographical studies on fracture surfaces using optical microscopy, SEM, AFM and profilometry, are described which provide a more general understanding of the origin and development of river line steps. It is shown, following Sommer], that river line patterns are a consequence of local mixed mode I/III conditions and the geometrical constraints associated with the 'no twist' condition. These ideas about river line steps are extended to the interpretation of crack propagation under dynamic conditions and, in particular, to the influence of dynamic stress intensity and crack velocity on fracture surface topography.

10:00 am

SEM STEREO-SECTION FRACTOGRAPHY (SSF) OBSERVATIONS: X.J. Zhang, R.L. Tregoning, Naval Surface Warfare Center, Carderock, MD 20034; R.W. Armstrong and G.R. Irwin, University of Maryland, College Park, MD 20742

Cleavage initiation is a local microstructural event occurring within engineering materials that often are not macroscopically nor microscopically uniform. This increases the difficulty of quantifying the relationship between cleavage fracture initiation toughness measurements and microstructure. The SEM stereo-section fractography (SSF) technique allows simultaneous observations of both the fracture surface and the underlying microstructure. Longitudinal sectioning within one micrometer of the local cleavage initiation site is possible, which allows a direct correspondence to be established between fracture event and microstructure. The information obtained from SSF provides critical input for the micromechanical modeling of microstructural influences on fracturing behavior. A compendium is presented of such SSF results and their consequences for fracturing in A533B steel, HY-80 multi-pass weldment, and Ti₆Al₄V.

10:20 am BREAK

10:30 am INVITED

A THREE-DIMENSIONAL MODEL FOR POLYCRYSTALLINE CLEAVAGE AND PROBLEMS IN CLEAVAGE AFTER EXTENDED PLASTIC FLOW OR CRACKING: Frank A. McClintock, Department of Mechanical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307

Fractures in the Northridge and Kobe earthquakes, as well as in new laboratory configurations, show the need for predicting cleavage after plastic flow. Fitness-for-service rather than traditional design and maintenance necessitate quantitative predictions. Those from micromechanisms revealed by micro-fractography provide insight, are limited, and require so much microstructural data as to be impractical. (An addition to this armamentarium is a model presented here for the reduction, as the grain size decreases, in the lower shelf work due to twisting between adjacent cleavage facets.) Thus fracture mechanics is needed, which predicts cleavage in large structures from tests on small specimens. K(T) and J(Q) mechanics are valuable for elastic-plastic initiation. The limit of non-hardening fracture mechanics for fully plastic flow or crack growth before cleavage problems include perturbations due to cleavage and the effects of strain reversals, aging, and statistics. Stability and arrest require elastic-plastic analysis.

11:00 am

THE ROLE OF PLASTIC DEFORMATION IN CLEAVAGE CRACK PROPAGATION AND ARREST IN FERRITIC STEELS: E. Smith, Manchester University-UMIST Materials Science Centre, Grosvenor Street, Manchester M1 7HS, United Kingdom

The paper reviews and presents new perspectives, related to the role of plastic deformation in the propagation and arrest of cleavage cracks in ferritic steels. Thus, the paper addresses, amongst others, the following issues: (a) The interpretation of K_Ia, the crack arrest toughness, (b) Temperature versus mechanics related arrest criteria, (c) The role of plasticity in the context of cumulative versus non-cumulative modes of crack extension, (d) The role of shear lips in cleavage crack propagation, (e) The relation between the upper shelf for initiation toughness and the maximum temperature at which a crack is able to propagate by the cleavage mechanism, in the context of the ASME K_IC and K_Ia curves.

11:20 am

AN INSTABILITY AND ENERGY RATE MODEL FOR CLEAVAGE FRACTURE: C.E. Turner, Mech. Eng. Dept., Imperial College, London SW7 2BX, United Kingdom

The problem addressed is cleavage fracture in steel, with appreciable plasticity. A deterministic model is offered to relate different configurations. The view point is that although a criterion such as a critical stress over a small region or process zone ahead of the crack tip must be satisfied, a critical energy rate must also be available to move forward the crack-tip plastic zone to embrace the next small process zone, before a crack can run in an unstable manner. The driving force for this process in real-elastic-plastic material can only be the energy rate available for linear elastic unloading. In LEFM (such as on the lower shelf) this rate is G. In a partly plastic state (once off the lower shelf) it is I=G+G_p where G is the LEFM term and G_p is a transferral rate of elastic to plastic energy associated with crack growth but not a 'release rate' available for crack separation, per se. G_p is a function of both configuration and system compliance, as well of course, of the fracture load. Some estimates of I will be given from simple theory supported by two-dimensional elastic-plastic finite element studies. This energy rate, I, is then used to relate cleavage data at a given temperature, taken from the literature, for deep notch bending, compact tension, shallow notch bending and centre-cracked tension configurations.

11:40 am INVITED

THE INFLUENCE OF A GREAT AMERICAN SCIENTIST AND ENGINEER IN GERMANY: Erwin Sommer, Fraunhofer-Institut für Werkstoffmechanik, Freiberg, Germany

In March 1961 the German Society for Material Testing (DVM) organized the conference DVM-Tag 1961 in Würzburg. In one of the invited papers Dr. George R. Irwin co-authored by John E. Srawley discussed the basic principles of fracture mechanics. At this time, only a few scientists in Germany mainly physicists had studied his article on the fundamentals of fracture in the "Handbook of Physics." This first scientific visit to Germany in 1961 left deep traces because of a personal contact to another speaker of the conference: Frank Kerkhof, who had proven by his investigations that due to their extreme brittleness glasses were ideal materials for applying the methods of fracture mechanics. This contact lead to a second visit of Dr. Irwin to Germany in 1965 at Freiburg, followed up by a regular exchange of ideas and have stimulated many projects in the field of fracture mechanics some examples will be discussed.

MICROSTRUCTURE EVOLUTION, CHARACTERIZATION AND MODELING: Session I: Liquid-Solid Systems I

Session Chair: J.A. Dantzig, University of Illinois, Dept. of Mech & Industrial Eng., 1206 W. Green St., Urbana, IL 61801

THE MICROSTRUCTURE OF HIGH VOLUME FRACTION SOLID-LIQUID MIXTURES: T.L. Wolfsdorf, P.W. Voorhees, Dept. Materials Science and Engineering, Northwestern University, Evanston, IL 60208

The success of a variety of industrial materials production techniques, including liquid state sintering and semi-solid processing, is strongly influenced by the skeletal structures that form at a high volume fraction of solid in solid-liquid mixtures. In order to gain a fundamental understanding of the relation between processing, structure, and properties in these materials, we investigate the morphology and formation of the solid skeleton which forms in high volume fraction solid-liquid mixtures. Using microstructural tomography, a novel technique for imaging 3-D microstructures, we characterize the details of the skeletal topology and connectedness. Electron back-scattered diffraction analysis of the solid particles in the skeleton yields quantitative evidence for the mechanisms that are operative during skeletal formation. Our data suggests a model for skeletal formation and the origin of the skeletal stability. Based on this model, we recommend two specific methods to engineer the properties of these materials. Work supported by the Microgravity Sciences Div. of NASA.

9:05 am

DIFFUSION LIMITED COARSENING OF PRECIPITATES: Martin E. Glicksman, H. Mandyam, S.P. Marsh*, Materials Science and Engineering Department, Chemical Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590; *Naval Research Laboratory, Washington, DC 20375-5343

Interactions among precipitate particles undergoing late-stage diffusion-limited phase coarsening were modeled with several theoretical approaches: 1) time-dependent multiparticle simulations; 2) snap-shot simulations and perturbation methods; 3) nearest-neighbor distribution functions; 4) direct screening, with random field cells. These techniques will be described briefly, and their kinetic predictions for the evolution of a two-phase microstructure compared. Consistency is found at low volume fractions (V_V < 0.1) of the dispersed phase with those methods that permit, or are consistent with, diffusional Debye screening. Leading-order corrections to the coarsening rate constants always start as V_v^1/2 with contributions from dipolar interactions occurring next. At somewhat higher volume fractions (0.10 < V_V < 0.3) correlation effects become more significant, and direct (geometric) screening seems to work well. This work supported by the National Science Foundation, Division of Materials Research, under grant DMR-9633346.

9:40 am

TEM STUDY OF THE TWO-PHASE PARTITIONLESS MICROSTRUCTURE EVOLUTION IN RAPID SOLIDIFIED CO-18.5Al ALLOYS: H. Sieber, D.R. Allen, J. Perepezko, Dept. of Materials Science and Engineering, University of Wisconsin-Madison, Madison, WI 53706

Rapid solidification is known to produce solute trapping but in the limit of full partitionless solidification, only one phase typically forms for a given cooling rate or alloy composition. In rapidly quenched cobalt-aluminum (Co-Al) foils for 17.22 at.%Al XRD investigations show two partitionless phases are formed, a face-centered cubic (fcc) and an ordered body-centered cubic (B2) phase. The microstructure of the alloy changes with decreasing cooling rate from the edge towards the middle of the quenched foils. The different microstructure regions in Co-Al splat quenched (SQ) foils have been analyzed in detail by TEM investigations in plan view and cross section geometry. A comparison of the anti-phase domains size to the grain size indicates that the B2 phase solidifies as disordered bcc and orders during cooling in the solid state. The results are compared with partitionless phase evolution observed in other systems and consider in terms of the kinetic constraints imposed by this novel microstructure. The authors gratefully acknowledge the support of NASA (NAGW-2841 and NAG8-1278) and a NASA Graduate Student Researchers Program fellowship (for DRA).

10:15 am BREAK

10:30 am

PHASE FIELD SIMULATIONS OF COALESCENCE AND FRAGMENTATION IN A BINARY ALLOY: James A. Warren, William J. Boettinger, NIST, Metallurgy Division, Bldg. 223, B164, Gaithersburg, MD 20899

In the last few years, many advancements have been made in the use of the phase-field method for solidification modeling. This approach has allowed researchers to predict realistic microsegregation patterns produced by dendrite growth in binary alloys. One of the advantages of this approach, besides its computational simplicity, is that topological changes such as the coalescence of dendritic sidebranches during growth and fragmentation of dendrites during melting are handled without any modification of the model. Recent efforts into three dimensional alloy simulations, as well as directional dendritic solidification will be discussed.

11:05 am INVITED

SIMULATION OF MICROSTRUCTURAL EVOLUTION IN TWO-PHASE SYSTEMS: Veena Tikare, Sandia National Laboratories, P.O. Box 5800, Albuquerque, NM 87110

Modeling of microstructural evolution presents unique challenges because of the large number thermodynamic, mechanistic and spatial variables which must be considered simultaneously. Microstructural evolution in two-phase systems is further complicated by characteristics of the two individual phases plus their interactive characteristics. These are thermodynamic characteristics such as mutual solubility of each in the other, wetting or non-wetting of each by the others, etc. and mechanistic characteristic such as transport mechanism of each of the two phases. The spatial arrangement of the two phases must also be considered. In this paper, the incorporation of characteristics of two separate phases into the Potts and the phase-field model will be presented. The advantages and disadvantages of each of the models will be discussed. The appropriate application of each of the models will also be discussed. This work was performed at Sandia National Laboratories, supported by the U.S. Department of Energy under contract number DE-AC04-95AL85000.

11:40 am

PHASE FIELD COMPUTATIONS USING AN ADAPTIVE GRID TECHNIQUE: N. Provatas, N. Goldenfeld, J. Dantzig, University of Illinois, Dept. of Mech & Industrial Eng., 1206 W. Green St., Urbana, IL 61801

In recent years, the phase field technique has been developed to model evolution of solidification microstructures. In the method, the liquid-solid interface is modeled as a diffuse region whose thickness is characterized by an order parameter, known as the phase field. One of the difficulties in applying the phase field method is the conflicting requirements of high resolution needed to successfully capture the physical phenomena at the interface, and the simultaneous need to fully resolve the diffusion field ahead of the advancing front. This had led to the use of very dense grids, with concomitantly large computation times, and also to the study of high growth regimes, where the diffusion field is relatively small. In this work, we describe an adaptive gridding procedure for solving the phase field equations, where high resolution is available near the interface, with more appropriate grid dimensions to resolve the diffusion field.

MODELING CARBURIZING, NITRIDING AND RELATED DIFFUSIONAL SURFACE TREATMENTS: Session I: Process Control I

Session Chair: J.E. Morral, University of Connecticut, Storrs, CT

MODELING THERMODYNAMICS AND KINETICS OF CARBURIZING AND NITRIDING - A BASIS FOR SOUND PROCESS CONTROLS: J. Ågren, Royal Institute of Technology, Dept. of Materials Science and Engineering, Division of Physical Metallurgy, Stockholm S-100 44, Sweden

Abstract not available.

9:00 am

MONITORING CARBURIZATION IN ROTARY RETORT FURNACES: C.A. Stickels, ERIM, 4683 Freedom Dr., Ann Arbor, MI

Small parts, such as links for steel chain and rollers for roller bearings, are often carburized in rotary retort furnaces. Parts are continuously fed into one end of a tubular retort, and move through the slowly revolving retort, driven by a helical rib on the inside of the retort. The carburizing gas is introduced at the discharge end and is vented at the charge end of the retort. Thus, there are counter-current flows of parts and the carburizing atmosphere. The carburizing process can be described as one in which the carbon demand of the workload and the carburizing potential of the furnace atmosphere vary along the length of the retort. For purposes of computation, the furnace was divided into a series of zones, within each of which the carbon potential and the carbon demand were constant. For a given set of input variables (temperature, production, rate, atmosphere flow rate, inlet atmosphere composition, etc.) the variations in carbon demand and atmosphere composition from zone to zone can be found. Calculations show how the amount of carburization in each part varies as a function of a) the inlet atmosphere composition, b) the flow rate of the atmosphere and c) the production rate. The "answers" provided by the model are sensitive to values estimated for rates of reaction within the gas phase and at the surface of the parts.

9:30 am

ON LINE DIFFUSION CALCULATIONS: T. Guler, Thermovac Pty., Ltd., 27 Biby St., Tugun 42224, Queensland, Australia

This paper will present several examples of "Simulation and Process Registrations" supported by metallurgical results on sealed quench furnaces. It will show that the presently used oxyprobes and simple controllers are more or less an off-line monitoring system and that on-line diffusion calculation has rendered the current methods as obsolete and superceded.

10:00 am BREAK

MODELING CARBURIZING, NITRIDING AND RELATED DIFFUSIONAL SURFACE TREATMENTS: Session II: Process Control II

Session Chair: M.J. Fischer, Surface Combustion, Inc., Maumee, OH

NITRIDING PROCESS CONTROL TECHNIQUES: R. Fincken, Marathon Monitors, Inc., 3100 E. Kemper Rd., Cincinnati, OH 45241

Abstract not available.

11:00 am

A DISCUSSION OF UNIQUE APPLICATION USING CAR-CALC MODELING SOFTWARE: R.L. Houghton, Hayes Heat Treating Inc., 800 Wellington Ave., Cranston, RI 02910

CARB-CALC software can accurately predict case depths in low alloy steel. The carbon profile data created from CARB-CALC can be automatically loaded into AC3 software, which can accurately predict hardness results of hardened and tempered low alloy steels with uniform or non-uniform carbon distribution. Both software packages are available from Marathon Monitors, Inc. This paper describes several examples of predictive models as compared to the heat treatment results. Both CARB-CALC and AC3 were used in the process development stage. The heat treatments were performed in a vacuum carburizing furnace with integral oil quench. The applications chosen for discussion are unique, since either the material was a non-standard carburizing grade or the process had to yield an unusual case profile requirement. The modeling software proved to be a valuable tool in the process development of these heat treat applications.

11:30 am

ON LINE USE OF INTERACTIVE DIFFUSION MODELS FOR CONTROL OF CARBURIZING AND NITRIDING: J.G. Conybear, I. Liddell, Ispen Industries International, 3260 Tillman Drive, Bensalem, PA 19020

Abstract not available.

PHASE TRANSFORMATIONS: General Abstract Session

Session Chair: Mukul Kumar, John Hopkins University, Laurel, MD 20723

CONTINUOUS COOLING OF CAST NICKEL ALUMINUM BRONZE (NAB): T.A. Marsico¹, P.R. Howell², ¹The Applied Research Laboratory; ²The Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802

Cast NAB was continuously cooled from 1029°C to 500°C. The first phase to form during continuous cooling of NAB from 1029°C was proeutectoid a as grain boundary allotriomorphs at 989°C. By 823°C, planar interfaces has developed between the and the proeutectoid a phases. The proeutectoid a was found to grow via the ledge mechanism. The k_ii and k_iv phases were first observed at 832°C. The k_iii phase was first observed at 823°C which marked the start of the eutectoid reaction during continuous cooling. In contrast to what is observed during isothermal transformation, the k_iii phase is present on the is side of the a/ interphase interface. This was the first time the very early stages of the eutectoid reaction is continuously cooled NAB was clearly documented. At 823°C, both discrete and lamellar k_iii were observed. A model describing the genesis and evolution of the eutectoid reaction in continuously cooled NAB will be presented. EPMA x-ray maps and concentration profiles indicated that a certain level of solute, primarily in terms of Ni and Al contents, must be achieved before cooperation can occur. The NAB eutectoid reaction product was found to grow via ledge mechanism and found to multiply via branching. Also clearly evidenced at 823°C was the encapsulation of k_ii by k_iii in contrast to the previous belief that k_ii provided a nucleation site for k_iii. At 768°C, precipitate free zones (PFZ's) were found to form around the peripheries of the proeutectoid a "grains". Also observed at 768°C was the impingement of proeutectoid a "grains" and the sympathetic nucleation of a onto a phase. From 768°C to 500°C, the proeutectoid a phase and the eutectoid reaction product, a+k_iii, continued to grow at the expense of the phase. At 500°C, martensite was still observed in room temperature microstructures indicating the eutectoid reaction did not have time to reach completion.

8:50 am

ISOTHERMAL TRANSFORMATION OF CAST NICKEL ALUMINUM BRONZE (NAB): T.A. Marsico¹, P.R. Howell², ¹The Applied Research Laboratory; ²The Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802

Cast NAB was isothermally transformed at temperatures from 1000°C to 500°C. A semiquantitative TTT curve for NAB was generated. NAB was fond to solidify as 100% phase. However, the 100% phase field was found to be very narrow. The proeutectoid a reaction was impossible to suppress even during rapid quenches from the melt. The proeutectoid a phase was found to form on grain boundaries, as allotriomorphs, at 1000°C. The morphology of the a phase was found the follow the DubÇ classification. The k_ii and k_iv phases were first observed at 880°C. The k_ii phase was always found to nucleate in the phase while the k_iv phase was always found to nucleate in the a phase. At 880°C, the encapsulation of the k_ii phase by the growing a phase was observed. This was the first time this process was clearly documented. A model describing this process was presented. The k_iii phase was first observed at 815°C which marked the start of the eutectoid reaction in NAB. For the first time the early stages of the eutectoid reaction in NAB were documented. The k_iii phase was directly viewed forming on the side of the a/ interphase interface. A model describing the very early stages of the eutectoid reaction in NAB was presented. At 815°C, the eutectoid reaction was mostly discrete but became lamellar as the transformation temperature was lowered. Also observed at 815°C was the formation of the "k_iii films". This was the first time "k_iii film" formation was observed and documented. A model describing the formation and growth of these films was presented. The sympathetic nucleation of a onto a was observed and discussed. This was the first time sympathetic nucleation of proeutectoid a in NAB was clearly observed. At 500°C, no martensite was observed in the room temperature NAB microstructures indicating the eutectoid reaction had reached completion.

9:10 am

CHARACTERIZATION OF CONCENTRATION PROFILES WITH AVERAGE EFFECTIVE INTERDIFFUSION COEFFICIENTS AND CHARACTERISTIC DEPTHS: Y.H. Sohn, M.A. Dayananda, Purdue University, School of Materials Engineering, 1289 MSEE Building, West Lafayette, IN 47907-1289

Concentration profiles of isothermal multicomponent diffusion couples can be asymptotic or exhibit various features including relative maxima and minima in concentrations, zero- flux planes and flux reversals for individual components. Selected isothermal diffusion couples in the Cu-Ni-Zn and Ni-Cr-A1 systems that develop these features are examined on the basis of average effective interdiffusion coefficients D_i^eff and characteristic depth parameters d_i calculated directly from the concentration profile of component i on either side of the Matano plane. These coefficients and parameters are related to the composition at the Matano plane. The ratio of D_i^eff to d_i determined on one side of the Matano plane is compared with the corresponding ratio determined on the other side in order to characterize the various types of concentration profiles. For asymptotic concentration profiles the ratio of D_i^eff to d_i evaluated on the higher concentration side of the Matano plane is greater than that for the lower- concentration side. For a component developing a zeroflux plane (ZFP), the D_i^eff to d_i ratio determined on the ZFP side of the couple is lower than that for the other side. The D_i^eff to d_i ratios are observed to be negative for both sides of the Matano plane for concentration profiles exhibiting a minimum on the higher concentration side and a maximum on the lower concentration side of the diffusion couple.

9:30 am

QUALITATIVE BOND VALENCE STUDY OF TRANSITION METAL HYDRIDES WITH THE FLUORITE STRUCTURE: Bryan Molloseau, M. De Graef, Department of Materials Science and Engineering, Carnegie Mellon University, 5000 Forbes Ave., Pittsburgh, PA 15213-3890

Transition metal intermetallics have potential applications in hydrogen-based propulsion systems for aircraft and future hypersonic vehicles. Hydrogen-induced degradation of structural materials is a major concern in the design of these systems. To obtain an in-depth understanding of the effect of hydrogen on the stability and properties of transition metals and their hydrides we employed at Monte Carlo implementation of the Bond Valence method (MCBVM), based on Pauling's crystal chemistry rules. The value of the MCBVM is to quantitatively predict bond lengths and unit cell dimensions, to test the validity of proposed crystal structures, locate sites for substitutionally disordered materials, and to locate weakly scattering atoms such as H and Li. Because the fluorite structure is the most prevalent dihydride crystal structure for transition metals, it was selected for this study. We will show that the MCBVM can be used to qualitatively predict charge transfer in metal hydrides. First-principals work on MgH₂, Pd-H and Ni-H compounds has shown a charge transfer from the metal to the hydrogen atoms; the MCBVM is in good agreement with these results, indicating that it may be used as a less-complicated first approach to the study of metal hydrides.

9:50 am

IN-SITU REACTION CONTROL IN THE Ti-Al-Si SYSTEM: J.S. Park, J.H. Perepezko, Department of Materials Science and Engineering, University of Wisconsin-Madison, 1509 University Ave., Madison, WI 53706

In the ternary Ti-Al-Si system, TiSi₂ and TiA1 phases were used as reactants in order to produce the Ti₅Si₃ phase as a reinforcement in a TiAl matnx. Following reaction at 1373K for 100 hours, several intermediate phase layers were produced such as TiSi, porous Ti₅Si₄, TiAl₃, Ti₂Al₅ and TiAl₂, but the desired Ti₅Si₃ phase was not produced. An analysis of the diffusion path indicated that Ti₅Si₃ formation required an enhanced Ti flux which was provided by a kinetic bias layer of Ti. In the Ti biased reaction, the Ti₅Si₃ phase was produced along with several intermediate phases. The reaction products of the biased reaction showed stability under long term annealing at 1373K. The diffusion pathway control is discussed in terms of chemical potential changes during in-situ synthesis. The support of ONR (N00014-92-J1554) is gratefully acknowledged.

10:10 am BREAK

10:20 am

HYDRIDING OF GRADE II AND GRADE III TITANIUM: Z. Wang, C.L. Briant, K.S. Kumar, K.J. Van Vliet, Division of Engineering, Brown University, Providence, RI 02912

In this paper we report a study of hydriding of grades II and III titanium exposed to sea water environments. These environments included aqueous solutions of NaCl of various concentrations as well as simulated sea water. The pH values ranged between 1 and 8 and the temperatures of the exposures varied between room temperature and 90°C. The results show that hydrides form in the material if the electrochemical potential is below approximately -700m V(SCE). The corrosion potential of titanium is well above this value. However, this potential can be achieved by impressing a cathodic current on the sample or by galvanically coupling the titanium to zinc or aluminum. Galvanic coupling to a low alloy steel will also produce this potential at elevated temperatures. Mechanical tests show that the elongation to failure of grade II titanium is not significantly affected when a thick hydride layer is formed on its surface. However, the less pure grade III titanium shows a decrease in elongation when tested in an environment that produces hydrides on the surface. This work was sponsored by ONR Contract N00014-96-1-0272.

10:40 am

TEM CHARACTERIZATION OF LAYERED MAGNETIC STRUCTURES: H. Geng, M.A. Crimp, Department of Materials Science and Mechanics, Michigan State University, East Lansing, MI 48824-1226

Thin films and multilayers of magnetic materials are being studied for sensor and information storage applications. However, a clear understanding of the magnetic properties, including giant magnetoresistance (GMR) requires complete characterization of the structure. In the present study, conventional and high resolution TEM are being used to examine the micro and atomic structure of GMR multilayers and spin valves. Specimens comprised of (Nb//Cu/FeMn/Py/Cu/Py//Nb) and (Nb//Ag/Py/Ag/Py/FeMn//Nb) multilayers grown on Si (001) substrates have been characterized. TEM images reveal excellent contrast between Nb, having the BCC structure with a {011}growth plane and the spin valves. However, contrast between the individual layers of the spin valves is poor. HREM analysis indicates non-equilibrium structures may exist in some of the spin valve layers. The specifics of these structures have been found to be a function of layer thickness and composition. This research has been supported by the NSF through the Materials Research Science and Engineering Center (Grant No. 61-2268).

11:00 am

THE EFFECT OF COMPOSITION ON THE STRUCTURAL MAGNETIC AND ELECTRICAL PROPERTIES OF THE SYSTEM La_1-xCa_xMnO₃: Th. Leventouri, J.J. Neumeir, D. H. Goodwin, Alloy Research Center, Physics Department, Florida Atlantic University, Boca Raton, FL 33431

An unusually large magnetoresistance occurs in La_1-xCa_xMnO₃ compounds for 0.2<x<0.5. The effects of processing and composition on the evolution of the structure in the system La_1-xCa_xMnO₃ (0<x<1) will be presented through Rietveld analysis of x-ray powder diffraction data. Magnetic as well as electrical properties will be discussed.

PROCESS AND MATERIAL BEHAVIOR MODELING SYMPOSIUM: Session I

Session Chair: S. Lee Semiatan, WL/MLLN, Wright Patterson AFB, OH

THERMAL STRESS DEVELOPMENT DURING THE VACUUM ARC REMELTING (VAR) PROCESS: M.K. Alam, Dept. of Mechanical Engineering, Stocker Engineering Center, Ohio University, Athens, OH 45701; S.L. Semiatin, Materials Directorate, Wright Laboratory, WL/ML, WPAFB, OH 45433

The development of temperature nonuniformities and the resulting thermal stresses during the VAR process were modeled using the commercial finite element method (FEM) code ABAQUS. Solutions were obtained for different values of ingot diameter, crucible-ingot interface heat transfer coefficients, and lengths of the solidified ingot. The stress solutions were obtained by using an elasto-plastic model with temperature dependent thermo-mechanical properties. Model predictions revealed that the maximum tensile thermal stresses are developed at the bottom of the ingot. The magnitude of the stresses increased with increase in ingot diameter and the interface heat transfer coefficients. The predicted development of large tensile stresses correlates well with observations of thermal cracking during VAR of near-gamma titanium aluminide alloy ingots.

8:55 am

PREDICTION OF DISTORTION IN SUPERALLOY COMPONENTS AFTER HEAT TREATMENT: T.C. Tszeng, W.T. Wu, Scientific Forming Technologies Corporation, Columbus, OH 43202

Unacceptable distortion or residual stresses are common problems in heat treating high-temperature structural materials such as superalloys. The ability to predict component distortion relies upon an accurate calculation of complex thermal, mechanical, and metallurgical changes during the heating and quenching cycles in the processes. Among other considerations in such a mathematical model, creep deformation is an important factor. In this presentation, progress in the incorporation of creep deformation in the modeling of several heat treating processes for superalloy components is summarized. Specifically, modifications to the computer program DEFORM to model solution treatment (heating, soaking, and quenching) and stress relief of Inconel 718 components are described. The calculated results are analyzed with reference to the prediction of component distortion after heat treatment. This study is partially funded by a US Air Force/Navy SBIR Award (Contract No. F33615-95-C-5238).

9:20 am

MICROSTRUCTURE/PROPERTY CONSIDERATIONS IN HIGH TEMPERATURE FORGING OF AEROENGINE DISK COMPONENTS: P.S. Follansbee, M.F. Henry, J. Nic, Physical Metallurgy Laboratory, General Electric Corporate R&D Center, Schenectady, NY 12301

The manufacture of modern aeroengine Ti and Ni-base disks places unique demands on cost and quality. Of all disk manufacturing operations, forging is the most amenable to simulation, which has become an essential tool in minimizing cost and assuring quality. In this presentation, we will review several important features of high temperature forging of disk components. We discuss some of the metallurgical features affecting forging of high gamma prime Ni-base components. We will emphasize the difficulty in prescribing constitutive behavior of these materials under high temperature conditions due to evolving microstructural conditions - particularly grain growth and grain refinement via dynamic recrystallization. The suitability of state variable approaches for these applications will be discussed.

9:45 am

CONSTITUTIVE RELATIONSHIP FOR SUPERPLASTICITY IN -TiAl ALLOYS: R.S. Mishra, A.K. Mukherjee, Department of Chemical Engineering and Materials Science, University of California, Davis, CA 95616

Several exciting developments are taking place in processing of -TiAl alloys for enhanced superplasticity. Multi-axial isothermal forging can produce 0.1-0.3 mm grain sizes in -TiAl alloys. These ultra-fine grained -TiAl alloys have exhibited low temperature superplasticity and provide the possibility of high strain rate superplasticity. Such possibilities are discussed with the help of constitutive equations for optimal superplasticity in -TiAl alloys. A three-dimensional 'superplasticity map' has been developed for -TiAl alloys to depict the optimum strain rate-temperature grain size domain for superplasticity. The use of these constitutive equations to develop a 'processing map' to produce ultra-fine grained -TiAl will be presented.

10:10 am

SUPERPLASTIC DEFORMATION AND CAVITATION MODEL BASED ON GRAIN BOUNDARY SLIDING: A.K. Ghosh, Department of Materials Science and Engineering, The University of Michigan, Ann Arbor, MI 48109-2136

The importance of grain boundary sliding (GBS) in superplastic deformation is well known. It is also known that the extent of sliding is a function of grain boundary character and varies from one region to another within a test specimen, leading to local regions of stress concentration and grain rotation. Recently, an attempt was made to model the response to the stress concentration by considering glide-climb creep in the grain mantle region. Grain boundary migration occurs naturally as a consequence of this process, as long as accommodation processes are adequate. At low stresses, where triple point accommodation problems are minimal, cavity nucleation is a result of satisfying a normal stress criterion on grain boundary particles. A simplified polycrystal model is developed to track the opening of voids and their growth as a function of strain rate, which is consistent with the deformation model itself. Results based on such an initial model will be presented.

10:35 am

MODELING OF MECHANICAL AND MICROSTRUCTURAL BEHAVIOR OF SUPERPLASTIC 7475 ALUMINUM: W.L. Moore, M. Zelin, D.E. Ebersole, Concurrent Technologies Corp., 1450 Scalp Ave., Johnstown, PA 15904

Increased use of superplastic aluminum alloys for component fabrication depends on the increased ability of design engineers to predict post-formed mechanical and identify areas where inspection is critical. To this end, software enhancements to the MARC^TM analysis code have been developed as a key part of a program to optimize the superplastic forming process for aluminum aircraft assemblies. The program focuses on blowforming of 7475 aluminum in the superplastic condition, with particular attention on prediction of forming process parameters such as optimized forming schedules (including backpressure), thickness, stress and strain distributions, and microstructural changes (including cavitation and grain growth). The predictions of the process simulation, and their subsequent verification by comparison with the characteristics of formed components, are summarized.

11:00 am

ANALYSIS OF CAVITATION BEHAVIOR OF A SUPERPLASTICALLY DEFORMED NEAR-GAMMA TITANIUM ALUMINIDE ALLOY: C.M. Lombard, Materials Directorate, Wright Laboratory, WL/MLLM, WPAFB, OH 45433; A.K. Ghosh, Dept. of Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109; S.L. Semiatin, Materials Directorate, Wright Laboratory, WL/ML, WPAFB, OH 45433

The uniaxial hot tension behavior of a near-gamma titanium aluminide alloy sheet (Ti-45.5Al-2Cr-2Nb) was determined in the as-rolled condition (initial grain size 3 to 5 µm) and rolled-and-heat treated (1177°C/4 hours or 1238°C/2 hours) conditions (initial grain size 10 to 12 µm). Microstructure evolution, cavitation rates, and failure modes were established via constant strain rate tests at 10^-4 sec^-1 to 10^-2 sec^-1 and test temperatures between 900 and 1200°C. Interrupted tests were also conducted in order to study cavity nucleation and growth. Cavity shapes and volume fractions were analyzed as a function of total strain. For all initial microstructural conditions, the failure mode was established as predominantly cavitation/fracture controlled. The cavitation behavior was interpreted using a theoretical analysis of the isothermal hot tension test under cavitating conditions.

11:25 am

STRIP CAST SIMULATION TO INVESTIGATE PHASE SELECTION IN 3xxx SERIES Al ALLOYS: L. Carroll¹, K.A.Q. Reilly¹, B. Cantor¹, P.V. Evans², ¹Oxford Centre for Advanced Materials and Composites, Dept. of Materials, University of Oxford, Parks Road, Oxford OX13PH, UK; ²Alcan International Limited, Banbury Laboratory, Southam Road, Banbury, Oxon OX167SP, UK

Electron beam surface processing is an experimental technique capable of generating steady-state growth velocities in the strip casting regime. Growth velocity is an important factor in determining intermetallic phase selection during solidification. In this work the influences of solidification rate and alloy content, including the addition of grain refiner, have been examined with respect to phase selection. Phase identification has been accomplished through a combination of phase extraction, x-ray diffraction, transmission electron microscopy and energy dispersive x-ray microanalysis. The aim of this work is to optimize alloy design in 3xxx series Al alloys in order to take full advantage of strip casting for production of beverage can body sheet.

TESTING AND INSPECTION TECHNIQUES FOR STRUCTURAL INTEGRITY EVALUATION AND POWER PLANT LIFE EXTENSION: Session I: Small Specimen Test Techniques

Session Chairs: Prof. James C. Li, Materials Science Program, Department of Mechanical Engineering, University of Rochester, Rochester, NY 14627; Dr. R. Viswanathan, Electric Power Research Institute, 3412 Hillview Ave, Palo Alto, CA 94303

OPENING REMARKS: Fahmy M. Haggag, Advanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak Ridge, TN 37830

8:40 am

USE OF SMALL SPECIMENS FOR EVALUATING STRUCTURAL INTEGRITY OF NUCLEAR REACTOR PRESSURE VESSELS: R.K. Nanstad, D.E. McCabe, M.A. Sokolov, D.J. Alexander, S.K. Iskander, Metals and Ceramic Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6151

The Oak Ridge National Laboratory is engaged in the evaluation of small mechanical test specimens for potential applicability to assessments of structural integrity of nuclear reactor pressure vessels. The Heavy-Section Steel Irradiation Program, funded by the U.S. Nuclear Regulatory Commission, is performing experiments with various types of small specimens for fracture toughness, Charpy impact toughness, crack-arrest toughness, and tensile tests. The use of so-called subside specimens could provide the capability to obtain mechanical property measurements from materials for which the available material is extremely limited. Materials which have been thoroughly characterized in previous studies with relatively large specimens are used for the small specimen evaluations. Both metallurgical and mechanical factors are considered in the evaluation of the test results with a view towards application of the techniques to operating reactor vessels. Research sponsored by the Office of Nuclear Regulatory Research, U.S. Nuclear Regulatory Commission, under Interagency Agreement DOE 1886-8109-8L with the U.S. Department of Energy under Contract DE-AC05-960R22464 with Lockheed Martin Energy Research Corp.

9:00 am

CONFOCAL MICROSCOPY METHOD TO CHARACTERIZE THE MICROHARDNESS INDENTATION TOPOLOGIES AND CONSTITUTIVE BEHAVIOR: T. Santos, G.R. Odette, G.E. Lucas, Department of Mechanical Engineering and Materials, University of California, Santa Barbara, CA 93106

Abstract not available.

9:20 am

APPLICATIONS OF THE STRESS-STRAIN-MICROPROBE TECHNOLOGY TO EVALUATE THE STRUCTURAL INTEGRITY OF METALLIC STRUCTURES: Fahmy M. Haggag, Advanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak Ridge, TN 37830; K. Linga Murty, North Carolina State University, Raleigh, NC 27695-7909

The Stress-Strain Microprobe (SSM) uses an automated ball indentation technique to obtain flow data from a localized region of a test specimen or component. This technique is being used to rapidly determine the yield strength and microstructural condition of a variety of materials including pressure vessel steels, stainless steels and nickel-base alloys. The SSM provides an essentially nondestructive technique for the measurement of local yield strength data. This technique is especially suitable for the study of complex or high variable microstructures such as weldments and weld heat affected zones. In this study yield strength data measured by the SSM are discussed and where possible compared to data obtained by conventional tensile tests. The sensitivity of the SSM to the presence of residual stresses is also discussed.

9:40 am

APPLICATIONS OF THE STRESS-STRAIN MICROPROBE TECHNOLOGY IN FIELD TESTING OF PRESSURE VESSLES PIPING, AND AIRCRAFT ENGINE DISKS-VIDEO PRESENTATION: Fahmy M. Haggag, ADvanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak RIdge, TN 37830

10:00 am BREAK

10:10 am

EVALUATION OF IN-SERVICE MECHANICAL DEGRADING OF COATINGS IN GAS TURBINE BLADES USING A SMALL PUNCH TESTING METHOD: J. Kameda, T.E. Bloomer, Ames Laboratory and CATD, Iowa State University, Ames, IO 50011; Y. Sugita, A. Ito, Electric Power R & D Center, Chubu Electric Power Co., Inc., Nagoya 458, Japan; S. Sakurai, MERL, Hitachi Ltd., Hitachi 317, Japan

The application of the standard mechanical testing method is not suitable for evaluating in-service degradation of gas turbine blade coatings because of the complex shape and localized degradation near surfaces. This paper examines how in-service and thermal environmental attack influence the mechanical properties (22-950'c) of CoNiCrAlY coatings over Rene' 80 substrates in gas turbine blades using a small punch (SP) testing technique in conjunction with scanning Auger microprobe analysis. SP tests have clearly shown strong dependence of environmental condition. In-service operation under combined fuels of kerosene and liquefied natural gas (LNG) led to a two-fold increase in the ductile-brittle transition temperature (DBTT) over coatings observed under mainly LNG due to more extensive oxidation and grain boundary sulfidation. However, the DBTT of coatings did not change during thermal aging at 870°C in air that produced only oxidation. These findings imply that the grain boundary sulfidation would exert a stronger embrittling effect on the CoNiCrAlY coatings than the oxidation. The integrity of blade coatings is discussed based on the SP test result.

10:30 am

LOCAL MECHANICAL PROPERTIES MEASURED BY IMPRESSION TESTS: James C. M. Li, Materials Science Program, Department of Mechanical Engineering, University of Rochester, Rochester, NY 14627

Micro and nano indentations have been used to probe local mechanical properties such as hardness and elastic modulus. However, since most indenters are of conical or pyramidal shape, it is impossible to produce steady state deformation. To solve this problem, cylindrical indenters have been used and it has been shown that a steady impression velocity can be achieved at a constant load and the load and temperature dependences of such velocity are comparable to those of the steady state deformation of tensile or compressive specimens of uniform cross section. Stress relaxation measurements can be performed also using a cylindrical indenter and the stress-strain rate relation obtained is similar to that from steady state measurements. In addition, the impression recovery experiments give indication of the dimensional stability of the material. Work supported by NSF through DMR9623808 monitored by Dr. Bruce MacDonald.

10:50 am

USING INTEGRATED ULTRASONIC TECHNIQUE TO ASSESS RADIATION DAMAGE AND DAMAGE DUE TO SERVICE ENVIRONMENTS: J.A. Wang, Computational Physics and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6370; Fahmy M. Haggag, Advanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak Ridge, TN 37830

Abstract not available.

11:10 am

AN IMPRESSION TEST METHOD FOR CHARACTERIZATION OF MECHANICAL BEHAVIOR: A REVIEW: M.A. Imam, H.Y. Yu, B.B. Rath, Material Science and Technology Directorate, Naval Research Laboratory, Washington, DC 20375-5320

Mechanical behavior of materials is measured using a simple localized impression testing technique, consisting of a flat-ended cylindrical punch. This method has significant advantages over the conventional approach, such as the need of small specimen volume, constant stress tests at constant load, and temperature and stress dependence of creep rate obtained from a single sample. Over the years, its application has been extended to study elastic modulus, yield strength, ductility, and superplasticity of materials. This approach has also been used to study fatigue, stress relaxation, and mechanical anisotropy of materials. The status of the impression test methodology and its applications are reviewed. The importance of the impression test method as an experimental tool in designing new materials, both bulk as well as thin films, will be discussed.

11:30 am

MICROCRACK METHODOLOGY FOR MORE EFFECTIVE RELIABILITY INSPECTION SCHEDULING OF METALLIC MATERIALS: E.Y. Chen, Physical Metallurgy Laboratory, GE Corporate Research & Development, P. O. Box 8, Schenectady, NY 12301; L. Lawson, 226 Interstate Parkway, Bradford, PA 16701; M. Meshii, Department of Materials Science and Engineering, Northwestern University, Evanston, IL 60208

In many metallic materials, the formation of microcracks is a continuous and often accelerating damage process. In such cases, calculations using conventional reliability theory on the rise of the hazard function (i.e., the probability of failure) show that the idea of equispaced inspections may not be sufficient for avoiding failure prior to the next inspection interval. To better understand the impact of microcracks in structural integrity evaluation, fatigue microcracking will be characterized for a 304 stainless steel from primarily replicas of smooth specimens. Since microcracks form in clusters, they will be analyzed as such in terms of distributions. This has been done by modeling their growth as a semi-Markov process represented by an embedded Markov chain. The simulation takes into account factors such as the stress intensity applied to the crack, the distribution of microstructural obstacles which may retard growth, and the probability that two cracks will coalesce. From these computations, it is seen that the long crack Paris equation does an effective job of predicting microcrack growth when microstructural variables are considered. The derived data in turn allow engineers to schedule inspections based on when detectable cracks would most likely appear, and then space them more closely late in life when the risk of failure increases with the density of cracks is higher.