Materials Week '97: Tuesday 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 Tuesday morning, September 16, 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 III

Session Chair: D.R. Frear, Sandia National Laboratories, Albuquerque, NM 87185

RIPENING-LIMITED INTERMETALLIC COMPOUND FORMATION IN SOLDERING REACTIONS: K.N. Tu, Dept. of Materials Science & Engineering, UCLA, Los Angeles, CA 90095; H.K. Kim, Samsung Corporation, Seoul, Korea; P.A. Totta, IBM East Fishkill Facility, Hopewell Junction, NY 12533

In interfacial reactions, not all intermetallic compounds grow in layers as it is assumed in the diffusion-limited or interfacial-reaction-limited mode of growth. During chip-joint in electronic packaging, the compound formed between molten SnPb solder and the Cu thin film on a Si chip is not layer-like, rather it is scallop-like. The growth of these scallops is limited by ripening. The ripening can lead to spalling of the compounds from the chip surface; it weakens the solder joint and is a serious device reliability issue. In this talk, the ripening morphology and kinetics will be presented.

8:55 am INVITED

MECHANICAL BEHAVIOR OF SOLDER MATERIALS IN FIBER-OPTICS STRUCTURES: E. Suhir, Bell Laboratories, Lucent Technologies, Murray Hill, NJ 07974

Mechanical behavior of solder materials and joints, used in fiber-optics engineering, is discussed, as well as some general considerations underlying the reliability requirements for such materials and joints. Several analytical stress models for the evaluation of thermal and mechanical stresses in optical fibers soldered into ferrules are suggested: 1) evaluation of the thermally induced stresses in an optical glass fiber soldered into a ferrule; 2) evaluation of the "global" and "local" thermal mismatch stresses in a fiber whose ends are soldered onto a ferrule (capillary); 3) evaluation of the "mechanical" interfacial shearing stress in a fiber soldered into a ferrule and subjected to tension; 4) predicted curvatures and stresses in a fiber-optic interconnect subjected to bending; 5) evaluation of the role of the bending stress in, and the appropriate length of, a glass fiber specimen soldered into a ferrule and subjected to a pull-test force; and 6) predicted "mechanical" and thermal stresses in, and optimization of the configuration of, an optical fiber interconnect in a laser package design. The developed analytical stress models, calculation procedures, and recommendations can be helpful in the materials selection, and in the analysis and design of solder joints in fiber optics.

9:20 am INVITED

PROPERTIES OF SEVERAL DIE ATTACH SOLDERS AS A FUNCTION OF TEMPERATURE: M.W. Weiser, Johnson Matthey Electronics, Spokane, WA 99216

The coefficient of thermal expansion, elastic, and shear properties of several solder alloys that are used for die attach in power devices were measured as a function of temperature ranging from as low as -65C to just below melting point. The solders included high Pb alloys like Pb₅Sn, Pb₅In_2.5Ag, and Pb_2.5Ag₂Sn; semi-hard alloys like Sn₂₅Ag₁₀Sb (alloy J) and Sn_8.5Sb; and a some lower melting alloys such as Sn_3.5Ag and Sn₃₇Pb for comparison to literature data. The elastic properties were evaluated by measuring the longitudinal and shear wave velocities in the ultrasonic region. The shear properties were evaluated in a TO-247 package with a bare Cu lead frame where the Si die was replaced with a Ni plated Mo tab.

9:45 am INVITED

STUDY OF DIE BOND RELIABILITY IN MICROELECTRONICS POWER DEVICES USING SURFACE ANALYSIS TOOLS: A. Scandurra, Laboratorio Superficied Interfaci-Consorzio Catania Ricerche, c/o SGS Thomson Microelectronics, 95100 Catania, Italy

Die bond reliability is one of the most important quality parameter in microelectronics power devices. A variety of phenomena occurring during thermomechanical fatigue in Pb-Sn2-Ag2.5 soft die bond in microelectronics power devices are discussed. In addition reliability of the die bond is correlated with surface chemical composition of the base metals, soft solder microstructure evolution and composition of fracture surfaces developed under fatigue. The usefulness of surface analysis tools like Secondary Ion Mass Spectrometry (SIMS) and X-ray Photoelectron Spectroscopy (XPS) in these applications is discussed.

10:10 am BREAK

10:30 am INVITED

CHARACTERIZATION OF INTERFACIAL MICROSTRUCTURES OF LEAD-TIN SOLDER JOINTS: G. Ghosh, Dept. of Materials Science and Engineering, Northwestern University, Evanston, IL 60208-3108

Various metallization schemes, such as Pd/Ni/Cu, Pd/Ag/Cu etc., are used in electronic packaging. During soldering metallized layer(s) react with the liquid solder and form various intermetallics both at the solder/substrate interface and in the solder. Diffusion of atomic species in the solid-state also lead to the formation and growth of intermetallic(s). In order to improve the reliability of electronic packaging, it is necessary to understand and control the diffusion path(s) and reaction mechanisms between the metallization layers and lead-tin solders. In this study we have used a variety of electron microscopy techniques, such as SEM, TEM and AEM to characterize the interfacial microstructures. The complex interfacial microstructures formed, on various substrates, due to interdiffusion will be discussed in detail.

10:55 am INVITED

MICROSTRUCTURE AND PROPERTIES OF SOME Sn-Zn-(X,Y) ALLOYS: R. Hwang, N. Jiang, J. Clum, E. Cotts^*, Mechanical Engineering, ^*Physics Department, State University of New York at Binghamton, Binghamton, NY 13902

Multicomponent (ternary and quaternary) alloys based on the Sn-Zn eutectic (198C) system have been examined to test for the interaction of composition with processing conditions in controlling microstructure and properties (microhardness, load relaxation, spreading reaction, etching reaction). Alloying additions have been tested with the objective of maintaining the eutectic melting behavior of the base alloy. The process conditions include rate of solidification, percent compressive strain, aging temperature and time. A simple factorially designed experimental plan was used to conduct and interpret the tests. In addition to the effect of alloying composition, a major effect of solidification rate and a secondary effect of the interaction between deformation and annealing was observed. Alloying additions have been seen to refine the structure of the base Sn-Zn eutectic as well as alter the stability of the microstructure.

11:20 am

THE DEVELOPMENT OF A LEAD FREE, IN-SITU COMPOSITE SOLDER ALLOY CONTAINING Ni OR Fe-BASED COMPOSITE PHASES: S. Choi, A.W. Gibson, J.L. McDougall, T.R. Bieler, K.N. Subramanian, Department of Materials Science and Mechanics, Michigan State University, East Lansing, MI 48824

The implementation of lead-free solder is impeded due to the lack of experience, material property data base, and reliability criteria for lead-free solder joints. Microstructural instability from thermomechanical cycling of solder is often detrimental to the performance of solder joints. Elemental Ni or Fe and additional Sn were added to eutectic Sn/Ag alloy to obtain about 20 Vol % of small in-situ composite intermetallic particles. Since coarsening of microstructural features during deformation is known to affect creep-fatigue resistance, the effect of the composite phase on aging behavior was investigated. Small single shear lap joints with a size similar to joints in microelectronic applications were fabricated using a melt reflow process similar to industrial practice. The effects of composite phase on microstructural evolution were monitored with aging temperature and time to determine the kinetics of aging processes. The effect of the composite phase on mechanical properties of the solder is also discussed.

11:40 am

MECHANICAL FATIGUE AND FATIGUE/CREEP OF SOLDERS: A.W. Gibson, K.N. Subramanian, T.R. Bieler, J.P. Lucas, Department of Materials Science and Mechanics, Michigan State University, East Lansing, MI 48824-1226

Automotive electronics and solders are exposed to relatively harsh environments, especially in under-the-hood applications. The solders are exposed to high and low frequency mechanical fatigue, as well as thermal fatigue/creep caused by temperatures fluctuating from -40 to +150C. To combat the detrimental effects of fatigue and/or creep, composite strengthening strategies are considered. Single shear lap solder joints, approximating the size used in microelectronics, are fatigue tested in cantilever bending mode utilizing a shaker table. Specimens are tested at room temperature and elevated temperatures to simulate under-the-hood temperature variations. Effects on microstructural coarsening, and crack initiation and growth are compared in similarly tested composite and non-composite solders.

ALLOY MODELING AND DESIGN: Session III: Physical, Chemical, and Mechanical Properties of Alloys I

Session Chair: Dr. D.D. Johnson, Sandia National Laboratories, Livermore, CA 94551

EFFECT OF COMPOSITION AND STRUCTURE ON THE VOLTAGE OF Li-INTERCALATED COMPUNDS: G. Ceder, M.K. Aydinol, A.F. Kohan, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139

Lithium-intercalation compounds are an important part of rechargeable lithium batteries. Using the ab-initio pseudopotential method it is possible to predict the average voltage at which Li-capability occurs in metal oxides. We use this newly developed capability to systematically investigate the effects of metal ion, chalcogenide ion, and crystal structure on the intercalation voltage. For existing compounds, such as LiCoO₂ and LiMn₂O₄, the agreement with experiments is extremely good. In the lithium-metal-oxides, lithium is completely ionized with its valence charge transferred partly to the metal and partly to the oxygen ions. Charge transfer to oxygen seems to correlate well with high output voltage. We demonstrate how this can be used to design compounds with very high energy density.

9:10 am

VIBRATIONAL ENTROPY DIFFERENCE BETWEEN ORDERED AND DISORDERED ALLOYS: Axel Van de Walle, Gerbrand Ceder, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139

Previous experimental measurements seem to indicate that a significant portion of the entropy difference between ordered and disordered Ni₃A1 is due to vibrational effects. In order to assess this usually large importance of vibration, ab-initio computations of the phonon density of states were performed. The Full Potential LMTO method was used to compute the forces acting on each atom as function of their displacement. These results were used to fit spring constants in a Born-von Karman model. While the vibrational entropy of an ordered phase can easily be obtained from such a model, the entropy of a disordered phase requires a few more steps. The vibrational entropy obtained for different ordered structure were represented by a cluster expansion. The vibrational entropy of the disordered phase was then extrapolated using this expansion. These results were also compared to the ones obtained by linear response theory.

9:40 am

VIBRATIONAL ENTROPY DIFFERENCES BETWEEN hP24 AND FCC Co₃V BY HIGH TEMPERATURE INELASTIC NEUTRON SCATTERING: P.D. Bogdanoff , B. Fultz, Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA 91125; J.L. Robertson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831

We report our measurements for the difference in vibrational entropy between the hP24 and fcc high temperature phases of Co₃V, as measured by high temperature inelastic neutron scattering. Previous work has shown that near the transition temperature the vibrational entropy of the hP24 phase is a very strong function of temperature, undergoing a 0.10 k_B/atom increase over the 200K below the transition. The anharmonic effects in the single phase regions mask the entropy difference between hP24 and fcc at the transition temperature. Earlier work done at Caltech measured the vibrational entropy difference between hP24 and fcc at 0.11+/-0.02 k_B/atom using low temperature calorimetry. Work supported by U.S. DOE DE-FG03-96ER45572.

10:10 am BREAK

10:30 am

NON-COLLINEAR MAGNETISM OF Fe-RICH FeNi ALLOYS: Y. Wang, Pittsburgh Supercomputing Center, Pittsburgh, PA 15213; G.M. Stocks, D.M.C. Nicholson, W.A. Shelton, Oak Ridge National Laboratory, Oak Ridge, TN 37831

The magnetic structure of Fe-rich FeNi alloys has long been a subject of great scientific interest and controversy. In this study, we attempt to understand an interesting phenomenon that the average magnetic moment of the alloys in the -phase (fee) decreases dramatically in the composition range near 70% Fe. Although the observation was made more than thirty years ago, the mechanism for this moment collapse is still controversial. In our approach, the non-collinear locally self-consistent multiple scattering (LSMS) method is applied to the magnetic structure calculation of large unit cell samples consisting of Fe and Ni atoms. The Fe and Ni atoms are randomly distributed on a fcc lattice. The moment directions are initialized to be randomly oriented, and then, as the self-consistent iterations proceed, are allowed to rotate to minimize the total energy. A stable magnetic structure of the alloy is determined by the final moment configuration. We compare our results with experiments and discuss the implication of our results for the INVAR mechanism.

11:00 am

EMBEDDED ATOM METHOD CALCULATIONS OF VIBRATIONAL THERMODYNAMIC PROPERTIES OF ORDERED AND DISORDERED Cu₃Au AND Ni₃Al: Dane Morgan, Jeffrey Althoff, Didier de Fontaine, University of California, Berkeley, CA 94720; Mark Asta, Stephen Foiles, D.D. Johnson, Sandia National Laboratories, Livermore, CA 94551

Recent work had suggested that vibrational effects play a significant role in determining some phase diagrams. In order to better understand these effects, we investigate the vibrational properties of disordered and ordered Cu₃Au and Ni₃Al using the Embedded Atom Method (EAM). We calculate vibrational thermodynamic quantities within the quasi-harmonic approximation. The vibrational entropy is found to be strongly dependent on volume and cell-internal relaxations. For fully relaxed structures the dependence on lattice decoration of the vibrational entropy is found to be significantly smaller than that suggested by recent experimental results. Research supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences.

11:30 am

COMPOSITION DEPENDENCE OF THE MAGNETIC MOMENT DISTRIBUTION OF Ni_cCu_1-c ALLOYS: G. Malcolm Stocks, D.M.C. Nicholson, W.A. Shelton, Oak Ridge National Laboratory, Oak Ridge, TN 37831; Yang Wang, Pittsburgh Supercomputing Center, Pittsburgh, PA 15213; I.A. Abrikosov, Physics Department, Uppsala University, S-75121 Uppsala, Sweden

We have studied the composition dependence of the magnetic structure of Ni_cCu_1-c alloys throughout the composition range for which these alloys are magnetic. The disordered solid solution phase is modeled using large (256-atom) super-cells and the calculations are performed using the first-principles locally self-consistent multiple scattering (LSMS) method utilizing the coherent potential approximation (CPA) boundary condition as proposed by Abrikosov et al. We present results for the composition dependence of the magnetic moments and the local environment dependent fluctuations in magnetic moments. We will focus on the extent of these fluctuations in the region of the critical concentration for the onset of magnetism. Research sponsored by the Division of Materials Science, Office of Basic Energy Research Sciences, and by the Division of Mathematical, Information, and Computation Sciences, Office of Computational and Technology Research, US DOE, under contract DE-AC05-96OR22464 with Lockheed Martin Energy Research Corp.

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

Session Chair: Kevin M. Knowles, The University of Cambridge, Cambridge, UK

DISLOCATIONS IN STRAINED LAYER INTERFACES: Peter J. Goodhew, Department of Materials Science & Engineering, The University of Liverpool, Liverpool L69 3BX, UK

Two dislocation configurations are particularly important in semiconductors. Threading dislocation (TDs) intersect the interfaces between layers and can have any Burgers vector. Misfit dislocations (MDs) have a component of their Burgers vector which acts to relieve the strain in the system, and they often lie in or parallel to interfaces between layers. In this paper we consider the origin of both types of dislocation, and their scope for multiplication. The device and growth requirements of strained layers and fully-relaxed layers with predictable lattice parameters but long term stability are different and sometimes conflict. We will report experimental work on III-V systems, particularly InGaAs grown on GaAs. We have studied the distribution of spacings of MDs and the implications these have for the dislocation sources which must be operating, their effect on the surface topography of a growing layer, the effect of vicinal plane substrates on TD formation, and the use of etching techniques to reveal TD densities even in thin heterolayers.

9:00 am INVITED

INTERFACE PROPERTIES AND PHASE STABILITIES IN METALLIC MULTILAYERS: Hamish L. Fraser, Rajarshi Banerjee, Suliman Dregia, Mark Asta^*, Andrew Quong^*, Department of Materials Science and Engineering, The Ohio State University, Columbus, OH; ^*Sandia National Laboratory, Livermore, CA

In a recent study of phase stability in thin multilayered samples of Al and Ti, it has been noted that as the scale of the microstructure is decreased, the Ti layers undergo a transition from hcp to fcc and then revert back to hcp in even thinner layers, whereas the Al layers undergo a transition from fcc to hcp. We have developed a thermodynamic model to account for these observations, which is based on the influence of the relative energies of the interfaces between the various stable and metastable configurations. Because of the apparent importance of these various interfaces on phase stabilities in these new types of materials, their structures have been characterized using high resolution and analytical electron microscopy, and the degree of compositional intermixing is being evaluated by high resolution SIMS. The information gained from these experiments is being used in theoretical computations (using a density functional approach) of bulk and interfacial energies so that a detailed basis to the thermodynamic model may be developed.

9:30 am INVITED

SHEAR BOUNDARIES IN LAMELLAR TiAl: P.M. Hazzledine, Materials Directorate, Wright Laboratory, WL/MLLM, Wright-Patterson AFB, OH 45433; UES Inc, 4401 Dayton-Xenia Road, Dayton, OH 45432

In lamellar TiAl six orientation variants of the tetragonal phase form a multilayer in which the lamellar boundaries are {111} planes. The tetragonality of the unit cell causes in-plane misfits in the lattice parameters of the order of 1%. In very thin lamellae the misfits are taken up elastically whereas in thicker lamellae some of the misfit is taken up by van der Merwe dislocations. The elastic strains are shears in the plane of the lamellae and the dislocations are either single sets of screw dislocations or cross grids of screw dislocations. Both the elastic coherency stresses and the mismatch dislocations affect the strength of the multilayer; the coherency stresses by direct interaction with glissile dislocations and the mismatch dislocations by acting as barriers to glissile dislocations crossing the lamellar boundaries. Neither has any effect on dislocations which are glissile in the plane of the lamellae. In this paper the strength of TiAl is discussed with reference to the structure of the lamellar boundaries.

10:00 am BREAK

10:10 am INVITED

GRAIN BOUNDARIES IN ORDERED ALLOYS: Ian Baker and Easo George, Thayer School of Engineering, Dartmouth College, Hanover, New Hampshire, NH 03755

Many strongly-ordered intermetallic compounds fail through brittle intergranular fracture. In this paper research on the structure and chemistry of grain boundaries in ordered alloys is reviewed. The observations are discussed in terms of the technique used for grain boundary examination with emphasis on the limitations of each technique. Most work on grain boundary structure/chemistry in ordered alloys has been on L1₂ compounds, in particular Ni₃Al. In this material, several researchers have noted that the grain boundaries in Ni₃Al can be enriched in nickel. This is usually observed when the material is nickel-rich and doped with boron. Grain boundaries in Ni₃Si seem to be similarly enriched in nickel. There have been few studies of grain boundaries in B2 compounds. However, grain boundary compositions in the B2 compounds NiAl and FeAl were usually found to be the same as the nearby matrix. The implications of these observations are discussed in terms of the mechanisms which have been presented for the brittle nature of the grain boundaries in these materials. Research supported by the US Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences, contract #DE-FG02-87ER45311.

10:40 am

GRAIN BOUNDARY STRUCTURE AND SEGREGATION IN NiAl: David E. Luzzi, Richard W. Fonda¹, Min Yan², Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA; ¹Naval Research Laboratory,Washington, DC; ²Los Alamos National Laboratory, Los Alamos, NM

A combined experimental and theoretical modeling approach utilizing HREM and atomistic structure calculations has been applied to the structure of a =5 (310) [001] grain boundary in Ni-rich NiAl. The resultant structure model, containing nickel antisite defects adjacent to the boundary plane, is consistent with HREM data, is the lowest energy structure at 0K via molecular statics calculations, and is stable at higher temperatures as determined from Monte Carlo calculations. The calculations were then extended to the study of stoichiometric and Al-rich grain boundaries. Segregation of point defects to the boundary is favored in all off stoichiometric NiAl. In Al-rich NiAl, either Ni vacancies or Al antisite defects may be found at the grain boundary, whereas only Ni constitutional vacancies are favored within the bulk.

11:00 am

EFFECTS OF INTERLAYERS ON THE ADHESION OF THIN COPPER FILMS: Michael D. Kreise, Neville R. Moody^*, William W. Gerberich, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN; ^*Sandia National Laboratories, Livermore, CA

Nanoindentation techniques are used to evaluate the effect of very thin "adhesion promotion" layers of refractory elements such as chromium and tungsten on the adhesion of thin copper films of 100nm to 1000nm thickness on SiO₂ substrates. Such interlayers produce changes in the work of adhesion, W_ad, which has previously been identified as leveraging all other contributions to the adhesion toughness, G_l. The present study examines in particular the effect of interlayers on the plastic dissipation contribution. Other variables of interest are film thickness, residual stress, and the effect of thick refractory overlayers.

11:20 am

STRUCTURE AND MECHANICAL PROPERTIES OF Cu/Nb AND Cu/Ni NANOSTRUCTURED MULTILAYERS: T.E. Mitchell, Y.C. Lu, J.D. Embury, M. Nastasi, H. Kung, Center for Materials Science, Mail Stop K765, Los Alamos National Laboratory, Los Alamos, NM 87545

Cu/Nb multilayers prepared by sputtering onto Si substrates with layer thicknesses ranging from 11Å to 5000Å have been characterized by transmission electron microscopy and nanoindentation. The films are strongly textured with fcc and bcc close-packed planes and directions tending to be parallel. For the 11Å layers the Cu is found to grow pseudomorphically on Nb in the bcc structure. It is thought that, for thicker layers, the bcc Cu loses coherency and transforms martensitically to the fcc phase, resulting in the observed Kurdjumov-Sachs orientation relationship. The hardness follows a Hall-Petch relationship for the larger thicknesses, but is controlled by Orowan dislocation bowing at the smallest thicknesses. In addition, Cu/Ni multilayers have been grown epitaxially on NaCl and Cu single crystal substrates. The multilayers are single crystal but contain a high density of dislocations. They are being deformed in order to understand slip propagation across the interfaces.

11:40 am INVITED

COMBINED ATOMISTIC AND HREM STUDIES OF INTERFACES IN AB TYPE INTERMETALLICS: NiAl vs TiAl: V. Vitek, Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA 19104

The atomic structure of grain boundaries and other interfaces in intermetallic compounds have been studied very extensively since they are often paths of brittle fracture or affect critically the plastic properties. Most of the atomic level calculations have been performed using the central-force potentials of the embedded-atom type to describe atomic interactions. In the case of NiAl such calculations revealed structures that are in an excellent agreement with HREM observations while in TiAl calculations employing such central force potentials exhibit some significant discrepancies. On the other hand ab initio electronic structure calculations lead to a very good agreement with HREM observations. It will be shown that the reason for this difference between NiAl and TiAl must be sought in a notably different nature of bonding in these two alloys. In particular, the bonding is practically metallic in NiAl while it has significantly covalent character in TiAl. Hence, combined HREM observations and atomistic calculations of interfacial structures can contribute significantly to our understanding of the nature of bonding in compounds and alloys.

EXTRACTION AND PROCESSING OF TITANIUM: Session III: Physical Metallurgy

Session Chair: Dr. G.J. Kipouros, Dept. of Mining & Metall. Engg., Technical Univ. of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4; J. Monsees, International Titanium Association, 1871 Folsom St., Suite #100, Boulder, CO 80302

REACTION PATHWAYS DURING METALLOTHERMIC REDUCTION OF TiCl₄ BY UTILIZING CHEMICAL POTENTIAL DIAGRAMS: T.H. Okabe, T. Uda, E. Kasai and Y. Waseda, Research Center for Metallurgical Process Engineering, Institute of Advanced Materials Processing, Tohoku University, 2-1-1 Karahira, Aobaku, Sendai 980-77, Japan

Phase equilibria and reaction pathways during metallothermic reduction of TiCl₄ has been considered by utilizing isothermal chemical potential diagram for Ti-M-Cl (M = Mg, Na, Al) at 1073°K. Importance of electronically mediated reaction (EMR) is discussed for each metallothermic reduction by explaining mechanism of titanium deposition in the Kroll or Hunter process. In the Kroll process, long-range electronically mediated reaction (LR-EMR), in which titanium deposit and reaction container act as electron conductor, was found to be dominant, and this may result in formation of sponge titanium. Both electron and ion transfer in NaCl solutions are quite likely to occur in the reaction pathway during sodiothermic reduction of TiCl₄. The formation of titanium powder in the Hunter process may be attributed to short-range electronically mediated reaction (SR-EMR). The possibility of pure titanium production by an aluminothermic reduction are also discussed.

9:10 am

OXIDATION OF Ti₃Al AND Ti₃Al-Nb INTERMETALLICS: R.G. Reddy, X. Wen and Y. Li, Department of Metallurgical & Materials Engineering, University of Alabama, Tuscaloosa, AL 35487

The oxidation behavior of Ti-Al-Nb intermetallic compounds were investigated. The experiments were carried out using TGA in the temperatures between 750 to 1100°C in the oxygen atmosphere. The samples were analyzed using X-ray, SEM and EDS. From the experimental results, oxidation rate constants, and diffusion coefficients were calculated. The oxidation products formed were identified. The mechanism of oxidation of intermetallic compounds was discussed.

9:35 am

SURFACE HARDENING BEHAVIOR OF TITANIUM ALLOYS IN CARBURIZATION: A.S.M.A. Haseeb, M.F. Islam, O. Alam, S.A.M. Tofail, Department of Metallurgical Engineering, Bangladesh University of Engineering & Technology, Dhaka-1000, Bangladesh

Surface hardening of titanium alloys has been attracting a lot of interest in recent years because of their increasing use in machine components requiring improved tribological properties. The present papers describes a comparison of the surface hardening behaviour of different commercial titanium alloys in pack carburization. Carburization of Ti-3Al-2.5 V, Ti-6Al-4V, and Super Alpa-2 was carried out in a packed bed containing 50% Charcoal, 38% coke,10% Barium Carbonte, 2% Sodium Carbonate. Carburization was done in the temperature range of 850°C to 900°C, for 1 to 3 hour. The carburized samples were investigated by optical microscopy, microhardness measurements and X-ray diffraction (XRD) technique. Surface hardness as high as 800 VHN was achieved in the carburized samples. The hardness profile formed on each of the alloys are discussed in terms of phases formed and microstructure developed during the carburization process.

10:00 am BREAK

10:15 am

ISOSTATIC DIFFUSION WELDING OF DISSIMILAR TITANIUM ALLOYS: M.F. Islam, M.O. Alam, Department of Metallurgical Engineering, Bangladesh University of Engineering & Technology, Dhaka- 1000, Bangladesh: N. Ridley, Materials Science Centre, UMIST and University of Manchester, Grosvenor Street, Manchester, M1 7HS, UK

Microduplex Ti-3Al-2.5V (IMI-325), Ti-6Al-4V (IMI-318) and Ti-25Al-10Nb-3V-1Mo (Titanium Aluminide) sheet materials have a considerable potential for superplastic deformation at temperatures in the range of 850-870°C, 860-910°C and 940- 980°C, respectively. Isostatic diffusion welding of dissimilar couples of titanium alloys (Ti-3Al-2.5V to Ti-6Al-4V and Ti-6Al-4V to Ti-25Al-10Nb-3V-1Mo) was carried out at temperatures of 850 - 920°C and pressures of 2.1 MPa to 6 MPa in order to form a sound weld between dissimilar materials. An assessment of the quality of welds produced was made on the basis of metallographic examination, lap shear test and scanning electron microscopy of the lap shear fracture surfaces. EDAX (energy dispersive X-ray detector) microanalysis using a scanning electron microscope on metallographic samples across the weld line was also performed to characterize the diffusion affected interface region produced between the dissimilar alloys.

10:40 am

EFFECT OF THE TYPE OF CARBON ON CHLORINATION OF REFRACTORY OXIDES: Carbochlorination of Tantalum Oxide: J. Gonzalez, M.C. Ruiz, INTEQUI, Universidad Nacional de San Luis; A.E. Bohe, D.M. Pasquevich, Centro Atomico Bariloche, CNEA, CONICET, 8400 Rio Negro, Argentina

The effect of the type of carbon on the reaction of carbochlorination of a refractory oxide, such as tantalum oxide, was investigated by thermogravimetry. Reactants and residues were characterizad by XRD, SEM and BET. The carbons utilized were graphite, carbon black and carbon obtained from sucrose. In order to associate the reduction capacity of carbon with their reactivity in the carbochlorination, they were burning of with oxygen and a mixture of oxygen-chlorine. It was observed that due to the chlorine presence the oxidation rate is delayed and carbon black and sucrose carbon increment their mass due to the chemisorption of chlorine on the active sites. From experimental results it is viewed that the reduction capacity of carbons, is associated to the reactivation in the carbochlorination of tantalum oxide; this reactivity order is: carbon of sucrose > black carbon > graphite. The same order was observed for the mass increment in chlorine, reactivity with oxygen and mixture oxygen-chlorine.

11:05 am

PHASE EQUILIBRIA IN THE TITANIUM-IRON-OXYGEN SYSTEM: S. Itoh, A. Kikuchi, Department of Metallurgy, Tohoku University, Aza Aoba, Aramaki, Aoba-Ku, Sendai 980-77, Japan; T. Azakami, Department of Machinery, Saitama Institute of Technology, Saitama, Japan; K. Itagaki, Institute for Advanced Materials Processing, Tohoku University, Sendai, Japan

Phase relations and the equilibrium partial pressures of oxygen in the titanium-iron-oxygen ternary system have been studied at 1173-1373 K by means of a thermogravimetric method in a CO-CO₂ gas mixtures and X-ray diffraction technique for the quenched samples after the equilibrium experiments. The objective of this study is both to clarify the relation between the phase relations in the Ti-Fe-O ternary system and the equilibrium partial pressures of oxygen, and to discuss the possibility of upgrading natural ilmenite ore to a rutile substitute for extractive metallurgy of titanium. The results are summarized as follows: The phase diagram and isobars of oxygen in the titanium-iron-oxygen ternary system at 1173 to 1373 K have been determined. From the present results upgrading ilmenite (FeTiO₃) to a rutile (TiO₂) substitute is possible thermodynamically.

FATIGUE AND CREEP OF COMPOSITE MATERIALS: Session III

Session Chairs: Peter K. Liaw, Dept. of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; D.B. Miracle, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433

CALCULATION OF THE SHEAR DECOHESION STRENGTH OF TI-MMC INTERFACE: D. Osborne, H. Ghonem, Mechanics of Materials Laboratory, Department of Mechanical Engineering, University of Rhode Island, Kingston, RI 02881

An experimental / numerical procedure has been carried out in order to calculate the shear decohesion strength of the fiber/matrix interface in a continuous fiber reinforced titanium metal matrix composite. Transverse loading experiments are used to determine the externally applied stress at which the interface decohesion under Mode I loading conditions occurs. The applied stress necessary to cause initiation of the Mode I interface decohesion was verified by measuring the displacement of the material near the interface during the loading process through a high magnification long range microscope. Using Finite Element Analysis, the corresponding local stress is calculated. This local stress is used to determine the minimum valid thickness for a pushout sample. Fiber pushout tests on samples which exceed this minimal thickness are then carried out in order to determine the applied load required to cause debonding under shear loading conditions. The stress state in a sample of this thickness is numerically calculated in order to identify the shear stress in the interface at the point when the interface begins to debond. This shear stress is taken in this study to be the shear decohesion strength of the fiber/matrix interface. This shear strength was found to decrease with increasing temperature due to the relaxation of residual stresses with increasing temperature. Results of this study were compared to the interface shear strength calculated from pushout tests based on the assumption of constant shear stress along the thickness of the sample.

9:00 am INVITED

ISOTHERMAL & THERMOMECHANICAL FATIGUE OF TITANIUM MATRIX COMPOSITE SUBJECT TO TRANSVERSE LOADING: A.H. Rosenberger, R. John*, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433-7817, *University of Dayton Research Institute, Dayton, OH 45469-0128

Unidirectional titanium matrix composites (TMCs) have a significant strength to weight benefit over monolithic materials when loaded in the fiber direction. The benefit of TMCs, however, is severely diminished when off-axis loading becomes appreciable such as the centrifugal loading in hoop wound rotating ring structures. This study examined the isothermal and thermomechanical fatigue (TMF) behavior of a transversely loaded unidirectional SCS-6/Ti-6Al-4V composite. Isothermal fatigue experiments were conducted at temperatures of 23, 371°C, and 427°C. Generally the isothermal fatigue life was short at applied stresses above the fiber debond stress and considerably longer at lower stresses. In-phase and out-of-phase TMF were conducted with a temperature range of 23 to 427°C. A simple model was used to conduct life predictions of these composites using fatigue properties of the matrix, taking into consideration the geometry of the transverse specimens, and temperature phasing of the TMF. Experimental results, life prediction modeling, and areas for model improvements are discussed.

9:20 am INVITED

LOW-CYCLE FATIGUE BEHAVIOR OF TRIMARC-1/Ti-6Al-2Sn-4Zr-2Mo: D.J. Buchanan^*, R. John^*, J.M. Larsen, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433-7817; ^*University of Dayton Research Institute, Dayton, OH 45469-0128

Continuous fiber reinforced titanium matrix composites (TMC) are planned as replacement materials for conventional steel components in elevated temperature aerospace applications. The results of an experimental and analytical investigation of isothermal low-cycle fatigue behavior of Ti-6Al-2Sn-4Zr-2Mo monolithic and [0]₁₀ metal matrix composite reinforced with silicon-carbide (Trimarc-1) fibers are presented. The fatigue tests were conducted at temperatures of 23, 163 and 371°C. The fatigue tests were load-controlled with stress ratios -1.3 and 0.1 for longitudinal tests and 0.1 for transverse tests. Fractographic studies are being conducted to document the influence of fiber spacing and fiber touches on the fatigue lives. This paper will also describe the application of a life fraction model based on micromechanics for longitudinal and transverse fatigue loading.

9:40 am INVITED

A COMPARISON OF DEFORMATION AND FATIGUE BEHAVIORS OF HIPPED FOIL AND SHEET TITANIUM ALLOYS: M.G. Castelli^*, B.A. Lerch, *NYMA, Inc., Brook Park, OH 44142, *NASA-LeRC, Cleveland, OH 44135

Micromechanics modeling of fiber reinforced MMCs requires a good and accurate knowledge of constituent properties. Much emphasis has been placed upon fabricating unreinforced matrix materials which have processing histories, and ideally mechanical properties, which are similar to the matrix materials in the foil/fiber/foil composites (in-situ matrix). This has required the processing of matrix plates which have been fabricated via foil consolidation rather than by more conventional wrought methods. However, the fabrication of such plates is complicated by the fact that the matrix material must be rolled into foils, a process which is not always easily accomplished. In addition, a consolidated foil plate, while giving comparable properties to those of the matrix in the composite (ideally), is expensive when compared to the standard wrought product. To alleviate these problems and yet maintain appropriately representative properties, a plate can be consolidated out of matrix sheets. These sheets do not require the extensive working required to produce the foils, and fewer sheets are needed to obtain the desired plate thickness. The question is, does the consolidated sheet material exhibit the same properties as that of the consolidated foils? This study investigates the mechanical properties of HIPped foil vs. HIPped sheet of Ti-matrix materials commonly employed in SiC reinforced Ti-composites. The Ti-materials investigated are: Ti-15-3, TIMETAL 21S, and Ti-6-4. Tensile, LCF, creep, and stress relaxation properties were examined and compared for both HIPped foil and sheet materials. Microstructural examinations were conducted to provide input to interpreting the macroscopic behaviors.

10:00 am BREAK

10:20 am INVITED

DEFORMATION AND RUPTURE MODEL OF [0] METAL MATRIX COMPOSITE UNDER SUSTAINED LOAD: N.E. Ashbaugh*, J. Metzcar*, A.H. Rosenberger, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433-7817; *University of Dayton Research Institute, Dayton, OH 45469-0128

A one-dimensional model for the deformation and rupture of a [0] titanium alloy matrix composite (TMC) has been developed for creep (sustained load) conditions. The model was based on the statistical strength distribution of the fibers and time dependent characterization of the matrix. Acoustic emissions from broken fibers have been monitored to assess the progression of fiber damage. The evaluation of the number and location of fiber breaks in interrupted test specimens has provided relevant information for assessment of the fiber/matrix load interactions and for correlation with acoustic emission events. Deformation and rupture predictions will be compared to experimental data from continuous fiber reinforced [0]8 SCS-6/Ti-6Al-4V tested at 427, 482 and 538°C and at various stress levels. Application of the model to forward predict the test results or to infer characteristic fiber strength distributions will be discussed.

10:40 am INVITED

CREEP OF SiC/Ti-1100 COMPOSITE IN AN AIR AND A VACUUM ENVIRONMENT: M.L. Gambone, A.H. Rosenberger, Wright Laboratory Materials Directorate, Wright-Patterson Air Force Base, OH 45433-7817

The titanium-alloy matrix composites in turbine engine components are required to withstand prolonged exposures at elevated stress and temperature; thus, understanding and predicting creep behavior is important. The near-alpha titanium alloy, Ti-1100, was developed for greater strength and creep resistance at temperatures as high as 593°C and has been proposed as a composite matrix to exploit these improved properties. This study examines the creep behavior of a unidirectionally-reinforced SiC fiber/Ti-1100 composite at 540°C in both laboratory air and vacuum. The dramatically reduced creep life of the composite tested in air compared to vacuum is attributed to environmentally assisted intergranular cracking in the Ti-1100 matrix. Modification of the matrix microstructure through heat treatments of the composite is also shown to increase creep life. The McLean/Curtin creep model is demonstrated to describe creep behavior of the composite in vacuum where matrix cracking does not occur.

11:00 am INVITED

ESTIMATIONS OF MMC TRANSVERSE RUPTURE AND RESIDUAL STRENGTH AFTER CREEP - A MICROMECHANICS APPROACH: E. Wung, Allison Engine Company, Rolls-Royce Aerospace Group, Indianapolis, IN 46206-0420

The high temperature creep behavior of continuous fiber reinforced SCS-6/Ti64 materials subjected to loading in the transverse direction is investigated with a micromechanics approach. The objective is to identify a suitable damage parameter which can be used to predict the rupture time or the residual strength if the material does not rupture during the hold time. Due to the weakly bonded fiber/matrix interface and the ductile nature of the Ti-6-4 material at elevated temperature, a large deformation theory approach is utilized in the stress analysis procedure which allows for the calculation of the true stresses and the logarithmic strains at the macro level as well as in the constituents. Owing to the evaluation of the true stress and strain quantities, the model prediction can accurately reproduce even the earlier stages of the traditional tertiary creep regime. A damage parameter based on the reduction of the cross-sectional area of the matrix material is proposed. This parameter is shown to correlate well with creep rupture and residual strength data. The incorporation of this damage parameter into an elaborated FEM computational scheme for the creep life prediction and the residual strength estimation of gas turbine engine components will be discussed.

11:20 am INVITED

TRANSVERSE CREEP OF SiC/Ti-6Al-4V FIBER-REINFORCED MMC's: D.B. Miracle¹, B.S. Majumdar², and S.G. Warrier²; ¹AF Wright Laboratory/ Materials Directorate, Dayton, OH; ²UES, Inc., 4401 Dayton-Xenia Rd., Dayton, OH 45432

The creep response of an 8-ply SiC (SCS-6)/Ti-6Al-4V composite has been measured at 427°C at several levels of stress. In addition to a conventional transverse creep sample geometry, two additional sample geometries were considered, with the intent of assessing the influence of the stress concentration at the sample free edge. These two geometries include a cruciform sample and a straight-sided sample with a layer of matrix material diffusion-bonded over the cut fiber ends. The transverse creep response of each sample geometry will be presented and discussed. Microstructural observations will be reported to describe the mechanisms of creep damage and failure. The contribution of the interfacial region to the transverse properties will be discussed.

11:40 am INVITED

DEGRADATION OF RESIDUAL STRENGTH IN SCS-6/Ti-15-3 DUE TO FULLY REVERSED FATIGUE: J.R. Calcaterra, S. Mall, Department of the Air Force, AFIT/ENY Building 640, 2950 P Street, Wright-Patterson AFB, OH 45433

Little attention has been given to residual strength degradation in Titanium Matrix Composites (TMCs) after cycling. To address this concern, fatigue tests on SCS-6/Ti-15-3 have recently been carried out at the Air Force Institute of Technology (AFIT). These tests were conducted in strain control at 427°C with R=-1. The main goal of the experiments was to determine their effect on the residual strength behavior of TMCs with fiber volume fractions (Vf) of 15, 25 and 42%. After test completion, fracture surfaces of each specimen were examined in a scanning electron microscope. Results indicate that fiber volume fraction seems to have an effect on both strain-controlled fatigue life and residual strength degradation. Lower fiber percentages result in material where the properties of the matrix, such as hardening or cracking, play a much larger role in the composite response. Despite these distinctions, all specimens tested retained the majority of their strength prior to failure.

FERROUS METALLURGY: General Abstract Session

Session Chair: Long-Qing Chen, University of Pennsylvania, Philadelphia, PA

ON LINE MEASUREMENT OF STRAND DISPLACEMENT AND ITS EFFECT ON INTERNAL QUALITY OF STEEL SLAB DURING CONTINUOUS CASTING: J.D. Lee, C.H. Yim, Technical Research Laboratories, Iron & Steel Making Research Team, POSCO, Pohang P.O. Box 36, 1, Koedong-dong, Nam-ku, Pohang-ski, Kyungbuk, 790-785, South Korea

Strand bulging is presumed to play a major role in the formation of internal cracks and centerline segregation in continuous casting. Meanwhile, the other factors such as roll displacement, which are due to ferrostatic force, roll bending and wear in the mechanical system, have not yet been investigated in detail. In the present paper, strand behavior at various positions of caster has been monitored during casting and its deformation has been discussed. Slab and roll displacement, which affects internal quality of slab, has been studied in relation to such casting conditions as pinch roll pressure, casting speed etc. Variation of slab displacement with pinch roll pressure at inlet of straightening may leads to quality defects and has been optimized by on-line monitoring of strand. It was also concluded that roll displacement control, which is based on on-line measurement during casting, specially around crater end should be emphasized for internal quality and effective maintenance of machine. In addition, dynamic bulging analysis considering strain accumulation has been performed to predict the critical strain for internal cracks and compared with experimental data.

8:50 am

HOT WORKING AND MICROSTRUCTURE DEVELOPMENT IN 409 FERRITIC STAINLESS STEEL: H.J. McQueen, E.V. Konopleva, N.D. Ryan, R. Zaripova*, Mech. Eng., Concordia University, 1455 de Maisonneuve Blvd. W., H-549-34, Montreal, Quebec, Canada

As continuously cast and homogenized 409 steels were deformed by torsion in the T range of 600-1200°C with strain rates of 0.01-5s^-1 to different strains (0.5 1, 2, 4 ) and to fracture. Flow curves exhibited peaks at 800-1200°C and gradual softening which was higher at higher stresses, thus due to deformation heating. At 600 and 700°C with =0.01 s¹, failure occurred at or just after the peak. The activation energy and strength were lower than those of 430 and 434 steels. In the as-cast segregated austenite, appearing at grain boundaries and within grains, raised the strength and diminished the nearby substructure recovery, causing some recrystallized nuclei. After testing at 600°C, deformed grains show little trace of recovery. At 700-1000°C, elongated grains contain substructure which is more developed at higher T and and lower . While the homogenization has eliminated the stringers, microprobe examination confirmed that there were small differences in composition between elongated grains which affect subgrain sizes and etching. There is no evidence of dynamic recrystallization in the range studied.

9:10 am

MÖSSBAUER EFFECT STUDY OF PHASE TRANSFORMATION & TRANSITION STATE IN MECHANICALLY ALLOYED Fe-Zn CUBIC MATERIALS: O.N.C. Uwakweh*, A. Jordan, Z. Liu, University of Cincinnati, Department of Materials Science and Engineering, Cincinnati, OH 45221-0012

The Mössbauer Effect measurement of Fe₂₅Zn₇₅ alloy corresponding to the mixed phase +₁ cubic phase reveals the presence of four distinct Fe-sites with quadrupole splitting (QS) of 1.10, 0.241, 0.0773 and 0.0772 mm/s. After aging at 130°C for 30 min., a new site with QS of 1.5 mm/s, and relative abundance of 5% is observed. At stable equilibrium following aging at 400°C for 1 hr., Fe-sites corresponding to the single phases of and ₁ are observed, with their total abundance in line with equilibrium compositions as determined by the lever rule. Comparison with the single phases show that this transient stage at 130°C aging is unique in the mixed phase composition and is detected for the first time in this system. X-ray diffraction (XRD) measurements confirm the existence of this stage.

9:30 am

KINETICS & TRANSFORMATION OF BALL-MILLED Fe-Zn-Al MATERIALS: Oswald N.C. Uwakweh, Zhentong Liu, Department of Materials Science and Engineering, 578 Engineering Research Center (ERC) Building, University of Cincinnati, Cincinnati, OH 45221-0012

Fixed ratios of Fe and Zn corresponding to -(Fe₃Zn₁₀), _l-(Fe₅Zn₂₁), -(FeZn₇) and -(FeZn₁₃) with the addition of 5% A1 (wt.) were ball milled in an argon gas atmosphere Non-isothermal kinetic analyses of the mechanically alloyed materials, based on DSC measurements revealed two diffusion controlled processes during the evolution of the + 5% Al and + 5% Al compositions with activation energies of 2272 kJ/mole, and 1591 kJ/mole respectively. Also endothermic and exothermic reactions detected are consistent with respect to the formation of the Fe₃Al, Fe₂Al₅, and -FeZn₇ phases. Based on FeA1₂ formation at 440°C for the + 5% Al, the revision/re-evaluation of the Fe-ZnAl equilibrium phase diagrams is proposed. The + 5% Al, and _l + 5% Al materials evolved similarly, except at 400°C where the former consisted of -Fe+ +, while the later was without the phase.

9:50 am

PHASE IDENTIFICATION IN ULTRA-LOW CARBON STEEL WELDMENTS: Richard W. Fonda, Naval Research Laboratory, Code 6324, Washington, DC 20375

While identification of phases is well established for medium and high carbon steels, the proper identification of phases in low and ultra-low carbon steels is still a problem which has not been adequately addressed. In this study, the decomposition of austenite in ultra-low carbon steels is examined by a combination of dilatometry, optical microscopy, and transmission electron microscopy. Samples cored out of an advanced ultra-low carbon steel weldment were thermally cycled at different cooling rates to simulate a wide variety of wading conditions. Dilatometric analysis is used to independently determine the volume fraction of each microconstituent present. These various samples are then examined by optical microscopy to determine the efficacy of different etchants in revealing and differentiating the various phases. Characteristics of these phases, both at the optical and the transmission electron microscopy levels of resolution will be discussed.

10:10 am BREAK

10:20 am

THE STUDY OF METASTABLE TO STABLE EQUILIBRIUM TRANSFORMATION OF MECHANICALLY ALLOYED Fe-Zn-Si MATERIALS: O. Uwakweh, Z. Jordan, P. Maziasz*, University of Cincinnati, Materials Science and Engineering Department, Cincinnati, OH 45221-0012; ^*Metals & Ceramics Division, Oak Ridge Natinal Lab (ORNL), Oak Ridge, TN

The ball-milling of elemental powders corresponding to (Fe₃Zn₁₀) + 0.12 Si; ₁ (Fe₅Zn₂₁) + 0.12 Si; (FeZn₇) + 0.12 Si; and (FeZn_l3) + 0.12 Si mechanically alloy to crystalline states. Differential scanning calorimetry (DSC) measurements show that they evolve through characteristic stages. The activation energies for these stages are 120 ± 0.32 kJ/mole, and 130 ± 0.99 kJ/mole ( + 0.12 Si); 161 ± 0.16 kJ/mole (₁ + 0.12 Si) respectively. Three stages with activation energies of 131 kJ/mole, 167 ± 0.2 kJ/mole and 244 ± 3.21 kJ/mole are observed for the + 0.12 Si alloy. Evolution of the + 0.12 Si alloy, exhibits two stages with activation energies of 96 ± 2.08 kJ/mole and 641 ± 0.45 kJ/mole, while the peak at 420 ± 3°C depicts the eutectic reaction in Zn-Si, and the melting of Zn in Fe-Zn binary systems.

10:40 am

METASTABLE TRANSFORMATION OF BALL-MILLED Fe-Zn-Si ALLOYS IN EQUILIBRIUM PHASE EVALUATION: O. Uwakweh, A. Jordan, 578 Engineering Research Center (ERC) Building, Materials Science & Engineering Department, University of Cincinnati, Cincinnati, OH 45221-0012

Fe-Zn intermetallics with the addition of 0.12 wt % Si corresponding to an isocomposition line in the ternary system is investigated with respect to FeSi formation. The mechanically alloyed -(Fe₃Zn₁₀) + 0.12 % Si; ₁(Fe₅Zn₂₁) + 0.12 % Si; (FeZn₇) + 0.12 % Si; and (FeZn₁₃) + 0.12 % Si, and the intermediate compositions (i.e., mixed phase alloys) yield metastable crystalline materials through ball-milling of pure elemental powders. DSC measurements in the 200-600°C temperature range indicate the presence of characteristic stages during their structural change. An invariant reaction occurring at 420-423°C is identified in the (FeZn₇) + 0.12 % Si and + + 0.12 % Si composition range. This reaction explains the phenomenon associated with the Zn coating of Si bearing steels, otherwise known as the Sandelin Effect. 2-D representation of the domain of FeSi boundaries is constructed from DSC and XRD measurements.

11:00 am

ANALYSIS ON THE INTERLAMELLA SPACING OF HYPEREUTECTOID PEARLITIC STEELS: Kyung-Tae Park, Sam-Kyu Cho, Jong-Kyo Choi, POSCO Technical Research Lab., P.O. Box 36, Pohang, Korea

Unless proeutectoid cementite exists, pearlitic cementite thickness would be a primary factor influencing the brittleness of hypereutectoid steels. In the present study, the variation of pearlitic cementite thickness with carbon content was analyzed by using a simple carbon mass conservation relation during austenite/pearlite transformation. The analysis revealed that pearlitic ferrite thickness always decreased with increasing carbon content while cementite thickness either increased or decreased with increasing carbon content. The critical condition for the variation of cementite thickness with carbon content was examined and the result was compared to the interlamellar spacing measurement data reported previously.

11:20 am

TOUGHNESS OF THIN COATED STEELS WITH HARDENED SUBSTRATE: T. Arai, Arvin TD Center, 2020 15th St., Columbus, IN 47201

Impact test and static bending tests made clear the fracture behavior of hardened steels coated mainly by TRD, as well as PVD and CVD. Impact strength and transverse rupture stress are more susceptible to the fracture behaviors of the hardened substrates rather than that of the coatings. In the case of brittle substrates, high-alloyed tool steels tempered at the low temperature, cracking will precede in the substrates, leading to failure. Cracking in the coatings preceding cracking in the tough steel substrates, tempered at the high temperature, usually does not induce the fractures of specimens. In some steels tempered at the low temperature cracking in the coatings induces cracking in the substrate simultaneously or at higher stress level. As a result, the coated steels are inferior in toughness to uncoated hardened steels.

11:40 am

DEFORMATION AND AGING BEHAVIORS OF Fe-Mn-Al-Cr-C DUPLEX ALLOY: M. Hadji, M. Temmar, T. Sahraoui, S. Zidelmel, Institute of Mechanical Engineering, University of BLIDA, P.O. Box 270, route de soumaa BLIDA 09000, ALGERIA

The microstructural changes which occur during aging of cold rolled Fe-Mn-Al-Cr-C duplex alloy have been studied. Typical laminate microstructures were developed during the first and second cycle of deformation. It was observed that at room temperature deformation the austenite deformed by the formation of deformation twins. The age hardening behaviors is observed to occur in four stages: a) very fine precipitation in ferrite with formation of precipitation free-zones (PFZ); b) significant homogeneous and heterogeneous precipitation in the deformation twins within the ferrite matrix; c) coursening of precipitation in ferrite and at grain boundaries, d) precipitation of secondary equiaxed ferrite within the austenite. A correlation between the mechanical properties and the microstructures is discussed.

GEORGE R. IRWIN SYMPOSIUM ON CLEAVAGE FRACTURE: Session III: Micromechanical and Microstructural Modeling

Session Chairs: R.O. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 94720; W.W. Gerberich, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455

8:30 am INVITED

RECENT ADVANCES IN THE MICRO-MODELLING OF CLEAVAGE FRACTURE IN STEEL: J.F. Knott, School of Metallurgy and Materials, The University of Birmingham, Birmingham B15 2TT, UK

The paper covers two aspects of the modelling of cleavage fracture in steels. The first relates to events at the microstructural scale, where fracture initiators are smaller than 10 µm (usually, smaller than 5µm) in size. Here, the discussion focuses on the ways in which the local cleavage fracture stress depends on microstructure. Attention is drawn to behaviour in weld metals, where initiation sites can be clearly identified on fracture surfaces. The relation of local fracture stress to fracture toughness relies on the identification of a "critical distance" and the uniqueness of this "distance" when dealing with cleavage following warm prestressing, or cleavage ahead of a growing fibrous crack, is questioned. Comparisons are made with the so-called "local approach" which is statistically based. The second aspect covers events at the meso-scale: specifically, the ways in which fracture toughness results should be analysed, when the microstructure comprises a two-phase mixture of tough and brittle areas, of order 50-200 µm in size. The traditional Weibull approach is challenged, because it does not represent physical reality and because it produces ultra-pessimistic lower-bound values. The re-examination of the Weibull approach has implications with respect to micro-scale modelling.

9:00 am

CLEAVAGE-QUASI CLEAVAGE IN FERRITIC AND MARTENSITIC STEELS: G.R. Odette, K.V. Edsinger, G.E. Lucas, Department of Mechanical Engineering, University of California Santa Barbara, Santa Barbara, CA 93106-5070

Confocal microscopy-fracture reconstruction and SEM were used to characterize the sequence-of-events leading to cleavage in a low alloy pressure vessel steel and two 8-12Cr martensitic steels as a function of temperature. While differences between the steels were observed, they shared some common characteristics that differ from the conventional view of cleavage. Most notably cleavage does not occur as a single weakest link event; rather it is the consequence of a critical condition when a previously nucleated dispersion of microcracks suddenly coalesce to form a large, rapidly propagating macroscopic crack. It is argued that the critical event can be treated as a bridging instability. The stabilizing effect of the ductile ligaments separating the cleavage facets increases with increasing temperature. Indeed, even in the ductile tearing regime cleavage facets form a significant fraction of nuclei for larger microvoids.

9:20 am

STATISTICAL AND CONSTRAINT FACTORS IN CLEAVAGE INITIATION: G.R. Odette, K.V. Edsinger, G.E. Lucas, Department of Mechanical Engineering, University of California Santa Barbara, Santa Barbara, CA 93106-5070

The size dependence of effective cleavage initiation toughness K_ee(T) (defined by the load-displacement conditions at initiation) of steels are mediated by both statistical and constraint factors. Statistical effects are controlled by the total high stress volume even under plane strain, small scale yielding, e.g., K_Ic1/B^-1/4. Constraint loss and reductions in the stress fields occurs for shallow cracks, large scale yielding and deviations from plane strain. The interplay between these factors is examined by analyzing the observed K_ee(T) behavior for specimens with different W, B and a/W using FEM simulations of the crack tip fields and confocal microscopy, fracture reconstruction and SEM characterization of the sequence-of-fracture-events. Observed versus actual sequences and complications such as crack tip strain, the transition to ductile tearing and ultimate loss of specimen capacity are discussed.

9:40 am

THE FRACTOGRAPHY-MODELING LINK IN CLEAVAGE FRACTURE: Anthony W. Thompson, Materials Sciences Div., Lawrence Berkeley National Laboratory, University of California, Berkeley, CA 94720

Cleavage fracture has historically been modelled, out of necessity, in rather idealized terms. In real materials, however, there are a number of difficulties in linking such models with metallographic and fractographic observations. Some of the most vivid examples occur for ₂ titanium aluminide alloys, in which, when the microstructure contains primary ₂ particles, the primary particles crack first. When "basketweave" or Widmanstätten structures of ₂ laths comprise the microstructure, it appears that individual laths crack first. And in colony structures, cracking occurs first across the ₂ lath colonies. Both detailed fractographic observations, and also a statistical model for brittle fracture by failure of weakest links, have been developed. The extent to which this can be interpreted in classical cleavage terms will be discussed. This work has been supported by the U.S. Air Force Office of Scientific Research.

10:10 am BREAK

10:20 am INVITED

MICROMECHANISMS OF CLEAVAGE FRACTURE: A.R. Rosenfield, Rosenfield and Rosenfield, 1650 Ridgway Pl., Columbus, OH 43212

This paper will review the history of crack-arrest technology, i.e. defining the conditions necessary to arrest a fast-moving cleavage crack in steel. Prof. Irwin's major contributions to the topic will be summarized and our research at Battelle stemming from these contributions will be discussed. The history of the concept of crack-arrest toughness, the development of a standard test method, and unresolved issues will be reviewed. Major emphasis will be placed on the observation and modelling of the microstructural contributions to toughness stemming from the unbroken ductile ligaments left in the wake of the advancing crack.

10:50 am

AN APPLICATION OF THE J-Q MODEL FOR ESTIMATING CLEAVAGE STRESS IN THE BRITTLE TO DUCTILE TRANSITION: John D. Landes, University of Tennessee, 310 Perkins Hall, Knoxville, TN 97996-2030; Carlos A.J. Miranda, IPEN, San Paulo, Brazil

A recent model has been proposed by the authors to predict cleavage failure in the transition for steels based on a weak link mechanism and a crack tip stress field modified for planar constraint by the J-Q theory. The model uses the distribution of toughness results at a single temperature to predict the same at a different temperature or for a different geometry. In this model a material cleavage stress is needed to predict when the weak link fracture is triggered. This cleavage stress is a key input for the application of the model but is not a property that is routinely measured and is hence not available for most steels. Using a characteristic of the model this cleavage stress can be estimated from the result of two distributions of toughness values tested at two different temperatures in the transition. In this paper the method to estimate a value of cleavage stress is presented and the result is used to predict the toughness distributions for structural components. Examples are given for several steels.

11:10 am

MODELING A CLEAVAGE-CHARACTERISTIC STRESS (S_co) OF FERRITIC STEELS: D.M. Li, School of Materials Science & Engineering, Harbin Institute of Technology, Box 433, Harbin 150001, China; M.Yao, Materials Science Department, Yanshan University, Qinhuangdau 066004, China

The cleavage fracture of ferritic steels has been studied under different specimen geometry (smooth, notched and cracked bars/sheets), loading mode (plain tension, three-/four-point notch/crack bending and center-cracked sheet tension) and strain rate (quasi-static to impact) conditions. The data of mechanical properties testing, fractographic observation and FEM simulation are incorporated to model and substantiate a "cleavage-characteristic stress," S_co, as representing a material's resistance to cleavage fracture. A concept of "effective plastic zone" has been emphasized in constructing the proposed probabilistic fracture model. Based on the parameter S_co, a unified cleavage fracture criterion is established accounting for the different conditions involved.

11:30 am

AN INVESTIGATION INTO THE MECHANISM OF CLEAVAGE FRACTURE IN A DUAL PHASE STEEL: Wei-Di Cao, Allvac, An Allegheny Teledyne Company, Monroe, NC 28110-0531

The cleavage fracture behaviors in notched and precracked specimens of a dual phase steel were studied as a function of grain size and test temperature. It was found that the relationship between critical fracture stress and bainite packet size can be described by _FF=_F^o + K_FF (d_pp)^-1/2, and _F is relatively temperature independent within the test temperature range. The cleavage fracture toughness decreased with increasing grain size, but rose again after passing a critical grain size, leading to a minimum in K_ICI-grain size dependence. A strong temperature dependence of K_ICIwas observed at higher test temperature near transition temperature of cleavage-dimple fracture. To explain the observed behaviors, a detailed fracture process observation was conducted by using acoustic emission monitor, cross section observation and fractography on interrupted test specimens. A comparison was made with major available theories and a modeling methodology is suggested to better explain the experimental phenomena.

HIGH CYCLE OF FATIGUE OF STRUCTURAL MATERIALS: Session I: Overviews and Perspectives

Session Chairs: Prof. Richard Hertzberg, Lehigh University; Materials Department, Whitaker Lab #5, Bethlehem, PA 18015; Prof. Wol Soboyejo, The Ohio State University, Dept. of Materials Science and Engineering, Columbus, OH 43210

OPENING REMARKS: Prof. Wol Soboyejo, The Ohio State University, Dept. of Materials Science and Engineering, Columbus, OH 43210

8:00 am INVITED

THE PARIS LAW FOR FATIGUE CRACK GROWTH IN TERMS OF THE CRACK TIP OPENING DISPLACEMENT: F.A. McClintock, Prof. Emeritus of Mechanical Engr., Massachusetts Institute of Technology, Cambridge, MA 02139

A convenient normalization of the Paris law is to divide the growth rate by the Burgers vector b and the range of stress intensity factor by not only E (Barsom, Speidel), but also by ÷(2b) where 2 is an empirical factor to make the threshold knee occur at DK_th/E÷(2b)=1, da/(bdn)=1. For non-normalized data, the reference point E÷(2b), b is compared with actual threshold knees for different alloys and conditions. Microstructurally short fatigue cracks, with ranges of stress intensity below the long crack threshold, grow faster than expected form the Paris law. This faster growth per cycle is compared to the crack tip displacement from several different idealized crack tip deformation mechanisms: pure slip on one plane, alternating slip on two symmetrical planes, and continuum sliding off from six fans at a non-hardening plastic crack tip. The fortunate deficit in fatigue crack growth rate from the CTOD is due to a combination of roughening, closure, and obstacles requiring polyslip. Apparently the smaller deficit in harder alloys, due to relatively less resistance contributed by polyslip, exceeds the benefit of smaller CTOD for a given DK.

8:25 am INVITED

COMMENTARY ON THE PARIS RELATION: M.E. Fine, Dept. of Materials Science and Engineering, Northwestern Univ., Evanston, IL 60628

The empirical establishment of the Paris Relation relating fatigue crack propagation rate to stress intensity amplitude is the basis for much experimental, theoretical, and modeling of fatigue failure in metals that continues to this day. A number of issues concerning the relation will be discussed. In the history of a fatigue failure when does the Paris regime begin? Can the Paris relation be extended to small cracks? What is the relation between flaws and cracks? How do flaws become cracks? When can the Paris Relation be integrated to give a useful estimation of fatigue life? What are the fundamental parameters that determine the crack propagation rate in the Paris regime? What determines the hysteretic plastic work during a loading cycle?

8:50 am INVITED

THE PARIS EXPONENT AND DISLOCATION CRACK TIP SHIELDING: J. Weertman, Dept. of Materials Science and Engineering and Dept. of Geological Sciences, Northwestern Univ., Evanston, IL 60208

A crack blunting dislocation emission mechanism likely leads to a Paris fatigue crack growth rate law of exponent 2 if dislocation shielding is not important. This talk discusses how dislocation crack tip shielding may affect an increase in the Paris exponent with a simultaneous reduction in the fatigue crack growth rate.

9:15 am INVITED

PARIS LAW EXTENSIONS FOR LOW (K) AND HIGH (K) REGIMES: A.B.O. Soboyejo, Dept of Aerospace Engr., W.O. Soboyejo, Dept. of Materials Science and Engr., The Ohio State University, Columbus, OH 43210

Fatigue crack growth and fracture in structural metallic materials is a stochastic process. The applicability of Paris law will be demonstrated to cover the entire zones of low stress intensity factor range (K) and high intensity factor range (DK) in structural metallic materials, provided appropriate modifications are made to account for the "death rate" in the stochastic process model for the characterization of the low (K) regime, and the "birth rate" in the stochastic process model for the characterization of the high (K) regime. Stochastic process models which incorporate the paris law, including the essential effects of multi-parameter variables and their possible statistical variabilities, which can contribute to the effective driving force which can cause fatigue crack growth and fracture, are presented in this technical paper. From the stochastic process model proposed, appropriate reliability functions are developed, in order to quantify the probability of survival, or structural metallic materials, under fatigue conditions. Possible applications of the principles developed here, in the area of engineering design, in order to minimize and control crack initiation, propagation and failure, in metallic materials and structures, will be discussed.

9:40 am INVITED

PROGRESS IN UNDERSTANDING CORROSION FATIGUE CRACK GROWTH: R.P. Wei, Dept. of Mechanical Engineering and Mechanics, in conjunction with AFOSR, Lehigh University, Bethlehem, PA 18015

The introduction and promotion of the use of the linear fracture mechanics parameter DK by Paul C. Paris and his associates in the early 1960s to characterize the driving force for fatigue crack growth profoundly affected fatigue research and design. The impact of this contribution is highlighted through a perspective overview of the progress in understanding corrosion fatigue crack growth in metallic alloys and its application to design over the past 30 years. Directions for future research are discussed.

10:05 am BREAK

10:20 am INVITED

PERSONAL REFLECTIONS OF PAUL PARIS' FIRST GRADUATE STUDENT: R.W. Hertzberg, New Jersey Zinc, Dept. of Materials Science and Engineering, Materials Department, Whitaker Lab #5, Bethlehem, PA 18015

The introduction and promotion of the use of the linear fracture mechanics parameter DK by Paul C. Paris and his associates in the early 1960s to characterize the driving force for fatigue crack growth profoundly affected fatigue research and design. The impact of this contribution is highlighted through a perspective overview of the progress in understanding corrosion fatigue crack growth in metallic alloys and its application to design over the past 30 years. Directions for future research are discussed. The early post-graduate years were highlighted by collaborative failure analyses of assorted structural components, including the famous 1969 failure analysis of the F-111 wing box. Simultaneous participation in fracture mechanics short courses broadened my technical training and prepared me for my teaching and research career at Lehigh University. The focus of my textbooks and my research interests will be examined in the context of my training with Paul Paris. Finally, a secret involving Paul Paris and one of my earlier failure analyses will be revealed.

10:45 am INVITED

HISTORY OF CONSTANT LIFE DIAGRAMS: G.P. Sendeckyj, Wright Laboratory, Materials Directorate, WL/MLLN, Wright-Patterson AFB, Dayton, OH 45433

A historical review of the development of constant life diagrams (variously referred to as Goodman, Smith, Haigh, etc. diagrams) is presented. It shows that neither Gerber nor Goodman published the first constant life diagram. Goodman never drew what is now called the Goodman diagram and the so-called Goodman law was in general engineering use before Goodman's book first appeared. Similar comments are shown to hold for the various other recent constant life diagrams.

11:10 am INVITED

CRACK GROWTH UNDER VARIABLE-AMPLITUDE AND SPECTRUM LOADING IN 2024-T3 ALUMINUM ALLOYS: J.C. Newman Jr., Mechanics of Materials Branch, NASA Langley Research Center, Hampton, VA

The damage-tolerance approach used today began when Professor Paul Paris proposed that fatigue-crack growth could be correlated with the "stress-intensity factor range." This concept has revolutionized the treatment of crack growth in aircraft structures. When Elber discovered "crack closure", the effective stress intensity factor range began to explain many crack growth load-interaction effects. This paper is dedicated to these achievements. The present paper is concerned with the application of a "plasticity-induced" crack closure model to study fatigue crack growth under various load histories. The model was based on the Dugdale model but modified to leave plastically deformed material in the wake of the advancing crack. The model was used to correlate crack growth rates under constant-amplitude loading and then used to predict crack growth under variable-amplitude and spectrum loading on thin sheet 2024-T3 aluminum alloys. Predicted crack-opening stresses agreed well with test data from the literature. The crack growth lives agreed within a factor of two for single and repeated spike overloads/underloads and within 20 percent for spectrum loading. Differences were attributed to fretting-product-debris-induced closure and three-dimensional affects not included in the model.

11:40 am

INTRODUCING THE K_MAX SENSITIVITY CONCEPT FOR CORRELATING FATIGUE CRACK GROWTH DATA: J.K. Donald, Fracture Technology and Associates, 2001 Stonesthrow Road, Bethlehem, PA 18015; Gary H. Bray and Ralph W. Bush, Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069

MECHANICAL BEHAVIOR OF BULK NANO-MATERIALS: Session II

Session Chair: Fernand Marquis, Dept. of Metallurgical Engineering, South Dakota School of Mines & Technology, Rapid City, SD 57701

SINTERING OF NANOPARTICLES: R.S. Averback, Huilong Zhu, M. Ghaly, M. Yeadon, J.M. Gibson, Department of Materials Science and Engineering, University of Illinois, 1304 W. Green St., Urbana, IL 61801

The sintering of nanoparticles has been investigated by a combination of molecular dynamics computer simulations and in situ observations in a UHV TEM. Because their small radii of curvature, nanoparticles begin to sinter when they contact by plastic deformation. Different mechanisms have been identified in metals: dislocation glide in pure metals, viscous flow along grain boundaries of intermetallics and bulk viscous flow in amorphous alloys. Similar mechanisms occur when nanoparticles contact substrates. The time scale for sintering is some tens of picoseconds. Direct observations of particle-particle sintering and particle substrate sintering by TEM will also be reported.

9:00 am

NANO-DUPLEX STRUCTURED (Mo,Ti)-DISILICIDES REACTIVELY SINTERED FROM PRETREATED POWDERS: T. Aizawa, M. Suzuki, B.K. Yen, Univ. of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan

Spex-type milling was used to yield nano-grained powder mixture both in Mo-Si and Ti-Si systems. These pretreated powders were mixed with the specific chemical composition, and reactively densified by using the hot-pressing, resulting in nearly-full dense, nano-grained (Mo-Ti)- disilicide duplex structure. Variation of hardening at room temperature and flexure strength at elevated temperatures with increasing the mole content of TiSi₂ into MoSi₂ was investigated to correlate the nano-structured duplex microstructure with the mechanical properties.

9:25 am

MICROSTRUCTURE AND PROPERTIES OF NANOCOMPOSITES OBTAINED THROUGH SPD-CONSOLIDATION OF COMPOSITE POWDERS: Y.T. Zhu¹, I.V. Alexandrov ², V.A. Shundalov ², R.Z. Valiev², T.C. Lowe¹, ¹Mail Stop G755, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545; ²Institute of Physics of Advanced Materials, Ufa State Aviation Technical University, Ufa 450000, Russia

Nanocomposites have great potential for structural applications because they exhibit not only excellent mechanical properties such as high yield strength and hardness, but also much better thermal stability than single-phase nanocrystalline materials. In this study, fully dense Cu-SiO₂ and Al-A1₂O₃ nanocomposites were obtained through consolidation of nanocrystalline composite powders using Severe Plastic Deformation (SPD). The microstructure and texture were analyzed to understand how microstructure development during deformation and subsequent annealing correlates with microhardness. Theoretical analysis using a dislocation model is employed to interpret the results.

9:50 am

INFLUENCE OF DYNAMIC DENSIFICATION ON NANOCRYSTALLINE STRUCTURE IN Ti-Si ALLOY: P. Counihan, N.N. Thadhani, School of Materials Science and Engineering, Georgia Tech, Atlanta, GA 30332-0245

Dynamic densification was used for consolidation of mechanically alloyed, amorphous Ti-Si alloy powders, employing a 3-capsule, plate-impact, gas-gun loading assembly. The powders densified under two-dimensional shock loading conditions at impact velocities of 300 and 500 m/s were observed to retain the amorphous structure as evidenced by XRD and TEM analysis. The densified material was subsequently annealed at various temperatures (600-1200°C) and times to produce a crystallized (Ti₅Si₃) nanoscale microstructure (>40 nm). TEM and XRD analysis along with microhardness measurements were used to characterize the microstructure and properties of the resulting material. In this presentation characteristics of the compacts and the influence of dynamic loading on the nanocrystalline structure in the crystallized compacts will be discussed. Funded by ARO and Georgia Tech.

10:15 am BREAK

10:30 am INVITED

SUPERPLASTICITY IN NANOCRYSTALLINE MATERIALS: R.S. Mishra, A.K. Mukherjee, Dept. of Chemical Eng. and Matls. Sci., Univ. of California Davis, CA 95616

Superplasticity is a grain size dependent phenomenon, because of which, it is expected that nanocrystalline materials would exhibit the effect at lower temperatures or higher strain rates. Both these features are attractive for technological applications. Lower forming temperature for ceramics can result in significant cost saving on power and tooling. The optimum strain rate for conventional superplasticity is in the range of 10^-5-10^-3 s^-1, which is quite slow for large scale forming of components. A shift in optimum superplastic strain rates to 10^-1-1 s^-1 is desirable. Our initial results on tensile deformation in a number of metallic and intermetallic materials show that the fundamental mechanism of superplasticity might be different. The results are analyzed to provide a mechanistic insight on superplasticity in nanocrystalline materials. The role of grain size on diffusional and dislocation accommodation during superplasticity is discussed.

11:00 am

LOW-CYCLE FATIGUE AND LOW TEMPERATURE CREEP BEHAVIOR OF ULTRAFINE-GRAIN (UFG) COPPER: S.R. Agnew¹, J.R. Weertman¹, R.Z. Valiev ², ¹Northwestern University, Evanston, IL 60208; ²Institute for Metals Superplasticity Problems, Ufa, Russia

Studies of low-cycle fatigue and low temperature (1/3T_m) creep behavior of UFG Cu made by severe plastic deformation have been designed to reveal information about the dominant deformation mechanisms, in addition to determining the microstructure's stability. During fully reversed low-cycle fatigue a considerable degree of cyclic softening has been observed, e.g., at _p =1.5% the stress amplitude decreased from 380 to 230 MPa (40%) prior to saturation. Changes in the microstructure resulting from the deformation were determined via TEM analysis before and after testing. A marked change in creep behavior has been observed at 160°C, where a long transient is followed by a reduced creep rate. TEM shows that the grain size has increased from 0.25 to 0.5 µm, with a high density of twins observed in the post-creep case. Research supported by Los Alamos National Laboratory Contract 7764Q0016-35.

11:25 am

MICROSTRUCTURAL CHARACTERISTICS OF ULTRAFINE-GRAINED Al AND Al ALLOYS PRODUCED USING EQUAL-CHANNEL ANGULAR PRESSING: Y. Iwahashi¹, Z. Horita¹, M. Furukawa², M. Nemoto¹, T.G. Langdon³, ¹Department of Materials Science and Engineering, Faculty of Engineering, Kyushu University, Fukuoka 812-81, Japan; ²Department of Technology, Fukuoka University of Education, Munakata, Fukuoka 811-41 Japan; ³University of Southern California, Los Angeles, CA 90089-1453

Ultrafine-grained structures of pure Al (99.9%) and A1-Mg solid solution alloys were produced using equal-channel angular extrusion whereby an intense plastic strain is introduced into the material. The structural evolution with respect to the imposed strain was examined using transmission electron microscopy. It is demonstrated that homogeneous equiaxed grain structures, with grain sizes of the order of ~1 µm and ~0.5 µm, may be attained in A1 and an Al-1% Mg alloy by subjecting the materials to equivalent strains of ~400% and ~600%, respectively. The results confirm that the addition of 1% Mg to the A1 matrix is effective in promoting grain refinement.

MICROSTRUCTURE EVOLUTION, CHARACTERIZATION, AND MODELING: Session III: Novel Systems & Methods

Session Chair: S.P. Marsh, Naval Research Laboratory, Code 6325, 4555 Overlook Ave. SW., Washington, DC 20375-5343

GRAIN AREA DISTRIBUTIONS IN EVOLVING THIN FILMS: M.A. Palmer, M.E. Glicksman, K. Rajan. Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY

The properties of a thin film for example, mechanical strength, electrical conductivity, and creep resistance depend upon the size of the individual grains which make up the thin film. The distribution of grain sizes as well as the average grain size will therefore be important. In this paper the grain area distribution in an evolving thin film is examined. It is found that two dimensional theories accurately describe the behavior of larger grains. However these models do not take into account the finite thickness or the finite width of the thin film. It has been found that these characteristics cause deviations from what would be predicted by the two dimensional models. Most notably the grain size distribution appears to broaden with time. Several tabulated distributions, as well as those predicted by numerous computer simulations, will be compared with the data, and the implications of the broadening will be discussed.

9:00 am

SIMULATION OF MICROSTRUCTURAL EVOLUTION IN ELECTRONIC GLASS-CERAMICS: Indrajit Sinha, Rajiv Kumar Mandal, School of Materials Science and Technology, Institute of Technology, Banaras Hindu University, Varanasi-221 005, India

It has been recognized that the presence of a connected network of the crystalline phase improves the performance of electronic glass-ceramics. Keeping this in view the theoretical simulation of microstructural evolution of the ceramic phase in a glassy matrix using the static Monte Carlo technique has been done in two and three dimensions. The problem has been approached from the perspective of the percolation theory. A normal distribution of probable sites of the ceramic phase in question has been done and the condition in commensurate with the physical realization that nucleation sites have a low probability of occurrence has been used to identify them.

9:30 am

MICROSTRUCTURE EVOLUTION DURING SPRAY FORMING OF LIQUID IMMISCIBLE ALLOYS: Rajiv Kumar Mandal, S.N. Ohja, School of Materials Science and Technology, Institute of Technology, Banaras Hindu University, Varanasi-221 005, India

Liquid immiscible alloys based on Al-Pb and Cu-Pb Systems were spray deposited using different processing conditions. The scanning electron microscopy of spray deposits invariably revealed uniform dispersion of submicron size lead particles in fine equiaxed grains of the matrix phase. In contrast, the atomized and overspray powder particles of these alloys indicated a bimodal size distribution of lead particles in the intercellular or interdendritic regions. The X-ray diffraction study and EPMA of the preform and atomized powders indicated presence of a new metastable phase in Cu-Pb alloys. The microstructural evolution during spray forming of liquid immiscible alloys will be discussed in light of the heat flow at the gas-droplet interface and consequent rapid solidification of droplets as well as that of the preform during spray deposition process.

10:00 am

CHARACTERIZATION OF THE BETA TO AMORPHOUS PHASE TRANSFORMATION IN A BULK TITANIUM-BASED ALLOY: K.J. Doherty¹, D.J. Li², G.J. Shiflet¹, S.J. Poon² , ¹University of Virginia, Dept. of Materials Science and Eng., Charlottesville, VA 22903; ²University of Virginia, Dept. of Physics, Charlottesville, VA 22901

Partial amorphization of bulk titanium-based alloys was obtained via annealing of a metastable crystalline material. A BCC solid solution (b) was formed in the Ti-Cr-TM (TM = one or more transition metals) system after arc-melting or air-cooling of a solutionized ingot. Initial examinations using x-ray diffraction patterns show a broadening of the major crystalline peaks along with an emergence of an amorphous halo with increasing aging time. Further evaluation using electron microscopy reveals complex phase transformations which vary significantly by modifying the aging temperature and time. Compositional fluctuations also change the amorphization characteristics. The premise for amorphization is that the free energy difference between an amorphous phase and the metastable crystalline phase is lowered by the appropriate additions of the proper transition metals to the binary Ti-Cr system.

10:30 am

ATOMIC-SCALE MEASUREMENT OF COMPOSITION PROFILES NEAR GROWING PRECIPITATES IN THE Cu-Co AND Ni-Al SYSTEMS: Ian Rozdilsky, A. Cerezo, G.D.W. Smith, Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, U.K.

The three-dimensional atom probe can reconstruct the positions of the majority of atoms within a small volume of material to sub-nanometer resolution. This technique has been used to measure atomic scale composition profiles in the immediate vicinity of nanometer scale precipitates during diffusional growth/coarsening in the Cu-Co and Ni-Al systems. Compositional data obtained from the interface of individual precipitates is used to examine the atomic attachment/detachment process.

11:00 am

A MACROSCOPIC MODEL FOR MULTIPHASE PRECIPITATION AS BASED ON THE CLASSIC NUCLEATION AND GROWTH THEORY: Juan C. Márquez, Ney Luiggi A., GFCES, Dpto. de F'sica, Escuela de Ciencias, Universidad de Oriente. Cumaná. Apdo. Postal 299, Sucre, Venezuela

The study of the multiphase precipitation process of a binary alloy system, with the nucleation and growth theory as its starting point, has led us to determine how the concentration of clusters of different sizes varies in a system where the decomposition of the solid solution is simultaneously shifted into a metastable phase and a stable phase. Both the metastable and the stable clusters evolve independently until the onset of the process of dissolution of the metastable phase. This premise leads us to a system of differential equations with as many equations as clusters participate in the process. As all the metastable clusters can be integrated into a single metastable phase and all the stable clusters can be integrated into one single stable phase, this system may well reduced to a macroscopic scheme for each phase, with one single differential equation defining all metastable cluster and one single differential equation defining all stable clusters. The insertion or non-insertion of the nucleation stage in this model depends on how a critical sizes of the stable and metastable clusters are defined and fixed. If the starting clusters are dimmers only the growth stage is characterized, three differential equations being sufficient to typify the process. The solutions obtained are mathematically different from the empirical Johnson-Melh-Avrami model. Our model is applied to the Pb-Ca binary and commercial aluminum alloy, our results being in qualitative agreement with the experiment. The not quantitative agreement is related to how the model is assumed relative to the reaction constant for each type of solute in the system examined.

11:30 am

ON THE TIME FOR COMPLETE RECRYSTALLIZATION: C.H. Wörner, Instituto de Física, Universidad Católica de Valparaíso, Casilla 4059, Valparaíso 02, Chile

Usually, Johnson-Mehl-Avrami-Kolmorogov (JMAK) kinetics is employed to follow the recrystallization process. In this paper, by using simple models of nucleation and growth, it is shown that the time for completion is finite (as opposed to the JMAK theory). It was found that in one, two, and three-dimensional systems, with instantaneous nucleation conditions, this time scales as 1/vn, 1/vn^1/2, and 1/vn^1/3, respectively; n being the nucleus density per unit length (area, volume) and v the growing interface isotropic speed.

12:00 noon

CONTAINERLESS DIRECTIONAL SOLIDIFICATION OF TEXTURALLY ALIGNED LAMELLAR TiAl Bimal Kad, Mike Scott,¹ and Dennis Dimiduk,² AIMES-0411, University of California San DIego, LaJolla, CA 92093, ¹UES Inc., 4401 Dayton-Xenia Road, Dayton, Ohio, ²WL/MLLM Materials Laboratory, WP-AFB, Ohio 45443.

The plastically anisotropic response of PolySysnthetically Twinned (PST) colonies of lamellar TiAl is such that the best combination of strength and ductility is obtained when the laminates are aligned parallel to the loading axis. This characteristic mechanical response has led to intensive efforts to texturally align lamellar TiAl microstructures by solidification, or deformation, processing based methodologies. Solidification schemes in the (Ti-48-50at%Al) composition range of interest are particularly complicated on account of closely spaced L+ and L+ peritectic phase fields, where the texture of the primary solidification product dictates the final laminate texture, based on epitaxial solid state transformation schemes. Additionally, the details of the L-> S portion of the phase diagram, in this composition range, are sketchy at best, and any phase boundary variabilities introduced by interstitial impurities or minor ternary additions are unknown. This presentation will report on our progress in producing aligned are unknown. This presentation will report on our progress in producing aligned Lamellar TiAl, within the framework of above mentioned phase diagram uncertainties. In particular efforts are directed at skirting the L+ phase field and extending the L+ phase field, with a view to sustaining a desirable texture in the primary solid.

NON-LINEAR EFFECTS IN MATERIALS SCIENCE: Session I: Diffusion

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

LINEAR AND NONLINEAR MODIFICATIONS OF THE DIFFUSION EQUATION: John W. Cahn, Materials Science and Engineering Laboratory, NIST, Gaithersburg, MD 20899

There has been increased interest in the formulation of the diffusion equations, modified to reflect specific complications of various applications. Many of these applications will be described later in this symposium. The modified equations can be nonlinear and may contain source or sink reaction terms, higher gradients, nonlocal terms, surface diffusion, etc. The study of these modified equations has often yielded valuable insights into the phenomena being studied, as well as results that sometimes deviate from what is expected from the linear diffusion equation in surprising ways. Various physical problems will be discussed to illustrate how modifications to the equations originate, and what qualitative effects these have on solutions.

9:00 am INVITED

RELAXING THE ASSUMPTION OF LOCAL THERMODYNAMIC EQUILIBRIUM AT INTERFACES IN STRESSED DIFFUSION COUPLES: William C. Johnson, Department of Materials Science and Engineering, Thornton Hall, University of Virginia, Charlottesville, VA 22903-2442

Impediments to establishing local thermodynamic equilibrium at an interface during a diffusional phase transformation will affect the motion of the interface, the interfacial compositions, and the composition profiles in each phase. A continuum theory on interfacial motion will be presented which accounts for deviations in chemical equilibrium at the interface and impediments to interface motion. Results are applied to the motion of a planar interface in a stressed, binary diffusion couple and show that large, time-dependent deviations in the interfacial compositions with respect to the stress-free equilibrium compositions are possible. This work is supported by NSF under Grant DMR-9496133.

9:30 am INVITED

DIFFUSION AT LATTICE STEPS-AN ATOMIC VIEW: Gert Ehrlich, Materials Research Laboratory and Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801

Steps on a crystal surface dramatically alter the ordinary diffusion of atoms over the surface and play a vital role in the growth of crystals. The atomic events taking place at steps are now accessible to direct examination by taking advantage of the unique ability of the field ion microscope to image individual metal adatoms without significant perturbation. Recent studies of jump processes which take place when atoms attempt to incorporate at ascending and descending steps on closepacked fcc surfaces have revealed a variety of unexpected phenomena, such as atom interchange and funneling, which will be briefly reviewed. Supported by the Department of Energy under Grant No.DEFG02-9lER-45439.

10:00 am INVITED

THE EFFECTS OF COARSENING ON CASCADES OF SPINODAL DECOMPOSITION IN MULTICOMPONENT ALLOYS: David J. Eyre, Department of Mathematics, University of Utah, Salt Lake City, UT 84112

Cahn-Hilliard equations suitable for multicomponent alloys are employed to study spinodal decomposition. Numerical simulations of spinodal decomposition show that the phase separation into multiple phases proceeds via a sequence or cascade of separations, and that coarsening of the solution is required before subsequent phase separation can take place. Numerical simulations suggest that this coarsening can delay phase separation for long times. In this talk, results of an investigation into the role of coarsening in secondary phase separation will be presented.

10:30 am BREAK

10:40 am

DIFFUSIONAL EVOLUTION OF ARBITRARILY SHAPED PRECIPITATES UNDER ELASTIC FIELDS: P.H. Leo, Department of Aerospace Engineering and Mechanics, University of Minnesota, Minneapolis, MN 55455

We study the diffusional evolution of arbitrarily shaped precipitates embedded coherently in an infinite elastic matax. We develop two distinct approaches to this problem. The first is a sharp interface method, where the composition and elastic fields are calculated by using a boundary integral method, and-the precipitate-matrix interface is tracked through its normal velocity. The second is a diffuse interface (Cahn-Hilliard) model, where microstructure is tracked by the evolution of smooth composition and displacement fields. In both formulations, nonlinearities enter through the curvature of the precipitate-matrix interface. We compare the two formulations. We also give results for precipitate shapes and motions as a function of elastic inhomogeneity, interfacial energy anisotropies and elastic anisotropies.

11:00 am

SIMULATION OF MICROSTRUCTURE IN 422 STAINLESS STEEL DURING NITRIDING: Carlene Hannigan, Advanced Technology Center, The Torrington Company, Torrington, CT 06790; Caian Qiu, J.E. Morral, Department of Metallurgy and Materials Engineering, Institute of Materials Science, University of Connecticut, Storrs, CT 06268

Microstructure evolution in commercial 422 stainless steel (Fe-11.3Cr-0.8Mn-1.0Mo-0.8NI-1.0W-0.2C wt %) has been investigated during nitriding by computer modeling. The nitriding process is simulated by means of the finite difference program DICTRA, which assumes diffusion control and local equilibrium. Composition profiles and diffusion paths have been calculated for various nitrogen potentials at the surface.

11:20 am

LINEAR RESULTS FROM THE NON-LINEAR FORM OF THE DIFFUSION EQUATION WHEN ANALYZING DIFFUSION COUPLE DATA: Yoon-Ho Son, New-Tech Co., Ltd, Daegu, Korea; J.E. Morral, University of Connecticut, Storrs, CT

When the diffusivity of a material is independent of concentration, the so-called "diffusion equation" is linear. Several methods of analyzing diffusion couple data to measure the diffusivity have been developed based on this linear form. However if the diffusivity is proportional to concentration, then the diffusion equation is nonlinear via a term proportional to the concentration gradient squared. Despite significant changes in concentration profiles that can result from the non-linear term, it has been found numerically that no detectable error is introduced when using the constant diffusivity, "square root diffusivity analysis" on diffusion data. The analysis always gives a diffusivity which is the average of initial alloy diffusivities. This observation has been made for all possible variations of the diffusivity in binary systems and for selected variations of the diffusivity matrix for ternary systems. The conclusion is that the "square root diffusivity analysis" can be applied to diffusion couples with significant concentration differences without necessarily introducing error.

PROCESS MODELING IN THE HIGH TEMPERATURE MATERIALS INDUSTRY: Session I

Session Chair: Marvin McKimpson, Institute of Materials Processing, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931

OPTIMIZATION OF WORKABILITY FOR HOT EXTRUSION OF CI2200 ALLOY: Steve C. Medeiros, S. Venugopal, W.M. Mullins, and James C. Malas, Materials Process Design, Materials Directorate, WL/MLIM, Bldg. 653, Wright-Patterson Air Force Base, Ohio, 45433-7746

A systematic study to determine the cause of problems such as severe tooling wear, eccentricity, and poor dimensional accuracy associated with hot-tube extrusion of deoxidized high phosphorus copper has been performed by optimizing the intrinsic and extrinsic workabilities. The intrinsic workability was optimized by studying the deformation characteristics in the temperature range from 750 to 950°C and a strain rate range from 0.01/s to 10/s using the constant strain rate, isothermal compression test method. Optimization of the extrinsic workability was performed by analyzing the flow of metal through the system using the upper bound technique. The results of both these analyses are i) current processing conditions are within a stable region and ii) the introduction of a flow control plate into the container will cause a decrease in the lateral loading on the mandrel resulting in decreased wear on the dies.

9:00 am

INDUSTRIAL APPLICATIONS OF PROCESS MODELING: B.A. Mueller, A. Hines, D. Hirvo, T. Simon, Howmet Corporation, 1500 South Warner St., Whitehall, MI 49461

Process modeling is being applied in industry to reduce development cycle times and improve processes. This application is the result of reductions in model construction time and improvements in model accuracy and predictive capability. Process modeling is able to accurately locate macroshrinkage, and provide trends in microporosity location and severity in equiaxed castings. Model size becomes an issue for large structural castings. For single crystal castings where defects such as spurious grains, freckles and boundaries form, process modeling is capable of predicting thermal profiles, the extent and location of the mushy zone, and fundamental parameters such as the thermal gradients and growth rates. However, the technology needs to be developed further to account for the effect of orientation on growth kinetics and to predict yield. In this presentation we highlight the above with examples of applications to production components.

9:30 am

THERMAL-MECHANICAL AND MICROSTRUCTURAL MODELING OF HEAT TREATING PROCESSES FOR SUPERALLOY COMPONENTS: T.C. Tszeng, W.T. Wu, Scientific Forming Technologies Corporation, 700 Ackerman Road, Suite 255, Columbus, OH 43202-1559

The stringent requirements in mechanical properties of superalloy components continue to demand better heat treating processes. In addition, unacceptable distortion or residual stresses are common difficulties in heat treating of superalloy components. A process design engineer is often facing the dilemma of needing to meet all of the mechanical and metallurgical requirements in the heat treated components. To better understand the thermal, mechanical and microstructural changes in the heat treating and subsequent machining processes for superalloy components, a process modeling system was developed based on the existing computer code DEFORM (design Environment for Forging). We will give an overview in the issues of general heat treating conditions, material constitutive models, metallurgical models, implementation and computational results. This study is partially funded by a US Air Force/Navy SBIR Award (Contract # F33615-95-C-5238).

10:00 am

MICROSTRUCTURE DRIVEN DESIGN FOR HOT DEFORMATION PROCESSES: J.C. Malas, S. Venugopal, W.G. Frazier, E.A. Medina, S. Medeiros, W.M. Mullins, N.U. Deshpande, A. Chaudhary, Materials Process Design, Materials Directorate, WL/MLIM, Bldg. 653, Wright-Patterson Air Force Base, OH 45433-7746

A new design approach based on the application of systems engineering principles to optimize microstructure development during hot working processes has been developed. Two stages of analysis and optimization form the basis of this microstructure driven design strategy. In the first stage, the optimal strain, strain rate and temperature trajectories for the `safe' processing of the material have been calculated. The optimum trajectories have been arrived at based on the kinetics of certain dynamic microstructural behaviors, thermo-physical characteristics of the material and the intrinsic hot workability of the material, along with a chosen optimality criterion. In the second stage, a process simulation model is used to calculate process control parameters, (e.g. ram velocity, die shape and billet temperature) needed to insure that the material follows the trajectories calculated in the first stage. This approach has been validated with an example of hot extrusion of steel, nickel and titanium alloys. Extrusion experiments were performed by using the optimized process parameters. The observed microstructural features in the extruded products were in close agreement with the desired ones.

10:30 am

APPLICATION OF PROCESS MODELING FOR HEAT TREAT OPTIMIZATION OF A HIGH STRENGTH CAST PLUS WROUGHT SUPERALLOY FOR ADVANCED SHAFT APPLICATIONS: Paul D. Genereux, Sue Goetschius, Jack Shirra, Pratt & Whitney Aircraft, MS 114-43, 400 Main Street, East Hartford, CT 06108

The current generation of high thrust turbofans require high strength shaft materials to withstand the torque loads applied due to high fan bypass ratios. In this paper, the development of a model to optimize the processing of a high strength superalloy for shaft applications will be discussed. The critical property requirement for shaft applications is elevated temperature yield strength. The modeling approach consisted of first developing a relationship between yield strength and microstructure / processing for the alloy. This was followed by the development of a process model predicting the microstructure (g' size) resulting from the heat treatment of the shaft. The model was then utilized to identify the effect of key process parameters such as solution temperature, transfer time and quench media on yield strength. It was determined that the greatest strength benefit resulted from reducing the transfer time and a heat treat facility was identified to minimize transfer time delays. Additional areas for process model development will be highlighted, particularly simulating the effect of billet conversion processes on billet/product grain size.

11:00 am

A UNIFIED APPROACH TO THE MODELING OF PLASMA ARC COLD HEARTH MELTING: Y. Pang, K.O. Yu, Concurrent Technologies Corporation, 1450 Scalp Ave., Johnstown, PA 15904

Titanium alloys are primarily melted, refined and cast into ingots via plasma arc melting (PAM), electron beam melting (EBM), and vacuum arc remelting (VAR) for jet engine applications. The PAM process offers superior ability to remove harmful hard alpha and high density inclusions and to potentially improve chemical segregation and product yield. To date, computer models are unavailable to establish the relationships between refining efficiency, ingot structure, and process conditions for the process. The NCEMT is developing a state-of-the-art PAM process simulation system, which includes three process models: plasma torch, refining hearth, and ingot solidification, to enhance the level of understanding and to optimize the process windows for consistent elimination of inclusions and imperfections in cast ingots. The objective of this presentation is to illustrate what capabilities these individual models entail and how they are integrated to yield a useful tool for process optimization. This work was conducted by the National Center for Excellence in Metalworking Technology, operated by Concurrent Technologies Corporation under contract No. N00140-92-C-BC49 to the U.S. Navy as part of the U.S. Navy Manufacturing Technology Program.

11:30 am

KNOWLEDGE-INTEGRATED SOLUTION HEAT TREATMENT PROCESS FOR TURBINE AIRFOILS: J.S. Chou, K.O. Yu, Concurrent Technologies Corporation, 1450 Scalp Ave., Johnstown, PA 15904

An optimization methodology for the solution heat treatment of directionally solidified (DS) and single crystal (SX) superalloy turbine airfoils has been developed. This methodology includes modeling the dissolution kinetics, predicting the alloy incipient melting temperatures, and preventing the formation of recrystallized grain defects. It provides a way to solution heat treatment processes for alloys René N4 and René N5 have been developed and implemented in production at PCC Airfoils and Howmet. These two new processes halved the total solution heat treatment time for René N4 and René N5 turbine airfoils used in F404 and F414 engines. This work was conducted by the National Center for Excellence in Metalworking Technology, operated by Concurrent Technologies Corporation under contract No. N00140-92-C-BC49 to the U.S. Navy as part of the U.S. Navy Manufacturing Technology Program.

TESTING AND INSPECTION TECHNIQUES FOR STRUCTURAL INTEGRITY EVALUATION AND POWER PLANT LIFE EXTENSION: Session III: Pressure Vessel Steels--Mechanical and Fracture Properties & Radiation Embrittlement

Session Chair: Prof. K. Linga Murty, Department of Nuclear Engineering and Department of Materials Science & Engineering, North Carolina State University, Raleigh, NC 27695-7909; Mr. Fahmy M. Haggag, Advanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak Ridge, TN 37830

CHARACTERIZATION OF PRESSURE VESSEL STEELS AND THEIR WELD GRADIENT PROPERTIES UTILIZING THE ABI TECHNIQUE: EFFECTS OF TEMPERATURE: Peter Q. Miraglia, K. Linga Murty, North Carolina State University, Nuclear Engineering Department, Raleigh, NC 27695-7909; Fahmy M. Haggag, Advanced Technology Corporation, 661 Emory Valley Road, Suite A, Oak Ridge TN 37830; Vikram N. Shah, Idaho National Engineering Laboratories, P.O. Box 1652, Idaho Falls, ID 83415-3870

The mechanical properties of nuclear pressure vessel weld material were characterized using the stress-strain-microprobe (SSM) based on the automated ball indentation (ABI) technique with the express purpose of evaluating gradient properties varying from the base, heat-affected-zone (HAZ), and the weld nugget. ABI-derived stress vs strain curves correlated with the conventional destructive tensile tests which were also performed on SSM using sub-size specimens. Mechanical properties derived from the ABI tests include yield strength, strength coefficient, strain hardening exponent (uniform ductility), and Brinell hardness. In addition, the tensile strengths are estimated from these data using the strain-hardening relation. In this particular weld, the strength properties of the weld nugget were bracketed by the base and HAZs. ABI tests clearly demonstrated the viability of SSM in characterizing localized properties in systems with gradient properties with least amount of specimen preparation, and in much shorter time period when compared to the conventional destructive tests which involve tedious and extensive specimen preparation. In addition, since ABI tests leave a very small impression with compressive residual stresses and no stress concentrations, the technique is regarded essentially non-destructive. The effect of temperature on the ABI-derived mechanical properties is being studied from -150°C to 250°C with the main aim being the development of correlations with fracture energies and DBTT. Experimental results, data analyses, and fracture correlations to-date will be presented. This project has been funded by the INEL University Research Consortium. The INEL is managed by Lockheed Martin Idaho Technologies Company for the U.S. Department of Energy, Idaho Operations Office, under Contract No. DE-ACO7-94ID13223.

8:50 am

RADIATION EMBRITTLEMENT PREDICTION MODELS, AND IMPACT OF HEAT-AFFECTED ZONE MATERIALS TO RADIATION EMBRITTLEMENT AND ANNEALING RECOVERY: 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; K. Linga Murty, North Carolina State University, Raleigh, NC 27695-7909

Abstract not available.

9:10 am

THE EFFECTS OF SIDE GROOVES ON THE DUCTILE-TO-BRITTLE TRANSITION BEHAVIOR OF SUBSIZE CHARPY SPECIMENS: David J. Alexander, Metals and Ceramic Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6151

It has been suggested that by machining side grooves of appropriate dimensions on subsize Charpy specimens the same ductile-to-brittle transition temperature (DBTT) can be measured as would be obtained with standard full-sized specimens. Therefore, small half-size Charpy specimens have been fabricated from A 533 grade B class 1 pressure vessel steel. Side grooves of 0, 10, and 20% of the specimen thickness were machined on each side of the specimens. These specimens were tested and compared to previous data generated with full-size specimens. The measured DBTTs will be compared, and the effects of the side grooves on the DBTT will be discussed. Also, the effects of the side grooves on the upper-shelf energy levels will be considered.

9:30 am

NOVEL METHODS TO DETERMINE FRACTURE TOUGHNESS AND CHARPY IMPACT PROPERTIES FROM AUTOMATED BALL INDENTATION TESTS: T. Byun, Korea Atomic Energy Research Institute, Taejon, S. Korea; 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

Abstract not available.

9:45 am

A UNIFIED TOUGHNESS MASTER CURVE-SHIFT METHOD FOR USING SMALL SPECIMENS: G.R. Odette, Department of Mechanical Engineering and Materials, University of California, Santa Barbara, CA 93106

Abstract not available.

10:00 am BREAK

10:10 am

DYNAMIC STRAIN AGING IN BASE AND WELD METAL OF SA106 GR.C PIPING STEEL: J.W. Kim, I.S. Kim, Department of Nuclear Engineering, Korea Advanced Institute of Science and Technology, 373-9, Mabuk-ri, Kusong-gu, Taejon 305-701, Korea; I.S. Ra, K.K. Jee, Korea Power Engineering Company, Inc., 360-9, Mabuk-ri, Kusong-myon, Yongin, 44-910, Korea

The characteristics of dynamic strain aging (DSA) in base and weld metal of SA106 Gr.C piping steel were examined through tensile and J-R testing carried out under various temperatures and loading rates. Results indicated that both metals were susceptible to DSA at a certain range of temperature, including nuclear power plant operating temperature, which depended on strain rate. DSA temperature range of weld metal was broader and appeared higher temperatures than that of base metal. This might be caused by smaller grain size and higher Mn content in weld metal. Crack initiation resistance, J_i, and crack growth resistance, dJ/da, in DSA region were about 40% lower than those at RT for both metals.

10:30 am

A CONFOCAL MICROSCOPY-FRACTURE RECONSTRUCTION METHOD TO MEASURE DYNAMIC FRACTURE TOUGHNESS IN SUBSIZED FRACTURE SPECIMENS: G.R. Odette, E. Donahue, G.E. Lucas, Department of Mechanical Engineering and Materials, University of California, Santa Barbara, CA 93106

Abstract not available.

10:50 am

USE OF A STRESS-STRAIN MICROPROBE FOR MATERIAL TESTING: N.F. Panayotou, D.G. Baldrey, Lockheed Martin, Schenectady, NY 12301

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 highly 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.