1997 TMS Annual Meeting: Wednesday Abstracts

RECENT ADVANCES IN FRACTURE--A Symposium Dedicated to Professor Emeritus Frank A. McClintock: Session VI: Fatigue Fracture

Session Chairpersons: Professor Robert O. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 97420-1760 and, Center for Advanced Materials, Lawrence Berkeley National Laboratory, Berkeley, CA 94720; Dr. Michael F. Henry, General Electric CR&D, P.O. Box 8, K1-MB229, Schenectady, NY 12309

2:00 pm INVITED

MECHANISMS OF FATIGUE FRACTURE IN METALLIC, CERAMIC, AND INTERMETALLIC MATERIALS: Robert O. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 97420-1760 and, Center for Advanced Materials, Lawrence Berkeley National Laboratory, Berkeley, CA 94720

From his early micro-mechanical modeling of fatigue-crack propagation in anti-plane shear to his later hypothesis for the existence of a fatigue threshold, Frank McClintock has made many seminal contributions to the understanding of fatigue fracture in metallic materials. In this presentation, current thinking on the mechanics and mechanisms of fatigue is examined, with particular emphasis on the similarities and differences between metal fatigue and the failure of monolithic and composite intermetallics and ceramics under cyclic loading. This is achieved by considering the process of fatigue-crack growth as a mutual competition between intrinsic mechanisms of crack advance ahead of the crack tip (e.g., alternating shear), which promote crack growth, and extrinsic mechanisms of crack-tip shielding behind the tip (e.g., crack closure and bridging), which impede it. The widely differing nature of these mechanisms and their specific dependence upon the alternating and maximum driven forces (e.g., K and Kmax) provide a useful distinction of the process of fatigue-crack propagation in the different classes of materials.

2:25 pm INVITED

OBSERVATIONS ON THE NUCLEATION AND MICROSTRUCTURALLY DEPENDENT CRACK PROPAGATION IN STRUCTURAL METALLIC MATERIALS: David W. Hoeppner, Department of Mechanical Engineering and Quality and Integrity Design Engineering Center, University of Utah, 3209 MEB, Salt Lake City, UT 84112

In the mid 1800's Sorby used the optical microscope to study fatigue deformation of materials and started an activity that continues with some vigor today. The early investigators of fatigue deformation wanted to learn the mechanism(s) of fatigue crack nucleation and propagation. This progress was aided by many investigators including Frank McClintock. McClintock and other investigators inspired many investigators to pursue the observation of fatigue deformation so that improved understanding of fatigue mechanisms could be attained. However, many studies that were made were "static". That is, materials were exposed to cyclic loading and then viewed with the observational technique. While this allowed some progress to be made, this did not deal with the actual deformation(s) that were occurring and thus progress in understanding fatigue was very limited. However, some investigators did mount microscopes on fatigue machines and make what became known as "in-situ" observations. Some of the contributions made will be reviewed as background of this paper. The development of the scanning electron microscope has changed the understanding of cyclic deformation response in engineering and model materials. Starting in 1970 numerous investigators around the world developed fatigue machines that were either placed in the chamber of an SEM or were attached to the SEM. Contributions of Kromp, Weiss, Strickler, Kikukawa, Jono, Aachi, Davidson, Hoeppner and others will be reviewed. In addition the past 25 years has seen the development of numerous in-situ systems. This has led to much greater understanding of fatigue crack nucleation and microstructurally dependent propagation. These studies and observations will be reviewed in the paper.

2:50 pm INVITED

DISLOCATION-CRACK INTERACTIONS DURING FATIGUE CRACK GROWTH: K. Sadananda1, N. Louat2 A. K. Vasudevan3, 1Naval Research Laboratory, Code 6323, 4555 Overlook Drive SW, Washington D. C. 20375; 2Fairfax Materials Research Inc., 5613 Marble Arch Way, Alexandria, VA 22315; 3Code 332, Office Naval Research, N. Quincy Road, Arlington, VA 22217

Fatigue crack growth occurs due to irreversibility associated with the plastic flow under cyclic load. Although crack growth occurs by plastic flow, the use of elastic fracture mechanics parameter to quantify crack growth may still be justified since material for the most part is under elastic loading, normally termed as small scale loading. In the past, concepts such as plasticity induced crack closure have crept into the literature and led to massive confusion in the literature in understanding and quantifying fatigue crack growth. We present new unifying concepts using dislocation models that accounts, load-ratio effects, so-called anomalous effects of short cracks, over-load and under-load effects, as well as concepts that connect the crack nucleation through Kitagaawa diagram.

3:25 pm INVITED

FATIGUE FRACTURE STRIATION FORMATION AND ITS RELATION TO CRACK TIP BEHAVIOR: Campbell Laird and Pedro Peralta, Department of Materials Science and Engineering, University of Pennsylvania, 3231 Wanut Street, Philadelphia, PA 19104

The mechanisms of fatigue striation formation are reviewed in terms of ductile and brittle fracture processes. Depending on these processes, fatigue striations can show considerable morphological variations, which are explained in terms of detailed slip processes and dislocation structures as well as fracture mechanisms. The problem of striation formation during intergranular crack propagation is also treated with respect to new results obtained on bicrystals having misorientation. These results are helpful for understanding how striation formation is related to the geometry and crystallography of slip at the crack tip.

3:50 pm BREAK

4:00 pm INVITED

INFLUENCE OF MICROSTRUCTURE ON LOW CYCLE FATIGUE FRACTURE: S. L. Mannan, K. Bhanu Sankara Rao, Materials Development Division, Indira Gandhi Center for Atomic Research, Kalpakkam 603 102, India

High temperature low cycle fatigue (LCF) has been an area of great interest in the last decades because of its relevance in nuclear and aerospace industrial applications. In general, the cyclic deformation and fracture of alloys engineered for high performance applications depend critically on the stability of initial microstructure during cyclic loading and on the slip mode, both of which in turn depending on temperature, govern the cyclic stress response and the mode of crack initiation and propagation. In order to attain adequate fatigue resistance in high temperature materials, a thorough understanding of quantitative relationship between initial microstructure and the fracture mode is required. Grain size, degree of prior cold work, thermal aging that occurs prior to and during service are the major structural parameters that influence LCF of austenitic stainless steels while the size and distribution of ' and carbides control the cyclic deformation and fracture in nickel base superalloys. Low cycle fatigue studies have been conducted on 304SS to examine the effects of grain size (75, 310 and 700 µm), cold work (0, 10, 20 and 30%) and thermal aging (923K: 1000h, 3000h, 5000h) on LCF fracture over a very wide temperature range (300-1023K). Fine grained material exhibited better endurance in terms of total and plastic strain amplitudes at all temperatures. The temperature dependence of fatigue life showed a complex behaviour. Medium and coarse grained alloy displayed a continuous reduction in life with increase in temperature with recovery in life at elevated temperatures. The effects of grain size on life were rationalised on the basis of martensite transformation, dynamic strain aging, oxidation and creep in the temperature range of their operation. The superior fatigue resistance of fine grained material has been attributed to the occurrence of trangranular fracture at all the temperatures. Cold work has been found to be both beneficial and harmful for fatigue life depending on the temperature. LCF life decreased with increasing PCW at and below 823K, whereas at 923K, PCW levels greater than 10% cold work exhibited recovery in life due to the reduced incidence of intergranular cracking. Thermal aging prior to LCF testing promoted transgranular cracking and improved fatigue resistance. The influence of microstructure (A: free from carbides and ', B: peak aged spherical ' of 18 nm diameter and C: overaged spherical ' of 35 nm diameter) on strain controlled LCF behaviour of Nimonic PE-16 superalloy has been studied at 923 K. Coffin-Manson plots describing the plastic strain amplitude versus life showed that Microstructure A had maximum fatigue resistance while C displayed the least. Microstructure B showed a two slope behaviour in the Coffin-Manson plot. For this condition samples cycled at low strain amplitudes exhibited much shorter lives than would be expected by extrapolation from high strain portion of the plot. Microstructure A exhibited crack initiation in planar slip bands, followed by transgranular propagation marked by fatigue striations. In B, the fracture mode at high strains was transgranular, while at low strains propagation by both trans and intergranular (mixed mode) was observed. Alloy C displayed cleavage facets very frequently on the fracture surface, and crack propagation was mixed mode at all strain amplitudes. The microstructural dependence of LCF behaviour and fracture modes have been rationalised on the basis of operative deformation mechanisms, degree of slip homogenization and the evolving microstructure during cyclic deformation. The degree of homogeneity was assessed by slip band spacing measurements on tested samples. These studies on 304SS and Nimonic PE-16 emphasise the need for optimisation of microstructure for maximizing fatigue resistance.

4:25 pm INVITED

CORROSION FATIGUE BEHAVIOR OF METALS AND ALLOYS: David J. Duquette, Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180

The qualitative aspects of the effects of corrosion on the fatigue behavior of metals and alloys have been recognized for more than 60 years. However, for much of that time, corrosion was assumed to lower fatigue resistance, simply by either inducing stress raisers in the form of pitting, or by reducing the cross-sectional area of a component. Modern research has shown that there may be a complex synergism between corrosion and the cyclic deformation response of metallic materials. This synergism may affect precrack deformation, crack nucleation and/or crack propagation. This presentation will address this synergism, and attempt to demonstrate that neither simple corrosion models for crack initiation, nor superposition models where corrosion rates and fatigue crack propagation rates are numerically added, are adequate to represent corrosion fatigue damage.

4:50 pm INVITED

THE EVOLUTION OF A CONSTITUTIVE RELATION FOR FATIGUE CRACK GROWTH: Arthur J. McEvily, Department of Metallurgy and Institute of Materials Science, U-136, University of Connecticut, Storrs, CT 06268

Load sequence effect can exert a strong influence on the rate of fatigue crack growth. For example, a large spike overload can result in a period of crack growth retardation after the overload. In order to calculate the number of delay cycles resulting from such an overload, a constitutive equation is needed, and it has been proposed that the following equation provides a generally valid constitutive relationship for the rate of fatigue crack growth: da/dN = A (Keff - Keffth)2 where a is the crack length, N is the number of cycles, A is a material constant, Keff is the effective range of the stress intensity factor, Keffth is the effective range of stress intensity factor at threshold. This equation can only be used if crack closure effects are accounted for. Examples of the use of the equation to calculate the number or retardation cycles following an overload as a function of specimen thickness will be given. In addition, the application of the equation to the study of short crack growth involving high-low and low-high load sequences will also be demonstrated.

5:15 pm INVITED

MICROMECHANICS OF FATIGUE AND FRACTURE IN LAMELLAR TiAl: Bimal K. Kad, Robert J. Asaro, Department of Applied Mechanics & Engineering Sciences, University of California at San Diego, La Jolla, CA 92093

While phenomological correlations, or simple rules of mixtures, are an important initial step in predicting the composite bulk response, the complexities of inhomogeneous deformation in layered morphologies, strain partitioning and incompatibilities between the respective micro-constituents, and discrete interface slip characteristics present formidable challenges for microstructural design. Thus, finite element based numerical procedures, incorporating phyically based crystal plasticity models, are employed to study the evolution of non-uniform deformation, under monotonic and fully reversed cyclic loadings, in lamellar TiAl microstructures. The impetus for such efforts is to gather fundamental insight into microstructure sensitive deformation mechanisms, and to extract additional information, not obtainable from traditional mechanical property data measurements. Such an effort is particularly desirable to help track various aspects of plastic anisotropy of specific layers, and micro-constituents are implicit in polycrystalline aggregates. We will present several examples of experimentally observed, and numerically computed results, to identify hot spots for strain localization in fully reversed loadings, and prescribe microstructural remedies to alleviate such effects.

5:40 pm

STEADY-STATE FATIGUE CRACK GROWTH IN COMPOSITE LAMINATES: J. Tong, Department of Mechanical and Manufacturing Engineering, University of Portmouth, Anglesea Building, Anglesea Road, Portmouth PO1 3DJ, UK

Matrix crack growth behaviour under mechanical fatigue loading has been studied in a quasi-isotropic GFRP laminate. Detailed experimental observations have been made on the growth of individual cracks and the accumulation of cracks in ±45°C as well as 90° plies. The results show that when the crack length is sufficiently long compared with the layer thickness, the growth of fatigue cracks in off-axis layers is essentially independent of crack length, a phenomenon termed as steady-state cracking. A generalised plane strain finite element model has been constructed and used to relate the crack growth rate to the associated strain energy release rate. A good correlation has been achieved which indicates that strain energy release rate may be the appropriate parameter to characterise stable matrix growth behaviour in composite laminates.