Crack paths and hydrogen-Afinssisted crack growth response in AlSi 4340 steel

A study of the correlation between crack paths and crack growth response was undertaken to define better the elemental processes involved in gaseous hydrogen embrittlement. AISI 4340 steel fractured under sustained load in hydrogen and in hydrogen sulfide over a range of temperatures and pressures, whose crack growth kinetics have been well characterized previously, was chosen for study. Fractographic results showed that crack growth followed predominantly along prior-austenite grain boundaries, with a small amount of quasi-cleavage, at low temperatures. At high temperatures, crack growth occurred primarily by microvoid coalescence. The fracture surface morphology, which is indicative of the micromechanisms for crack growth, was essentially the same for hydrogen and hydrogen sulfide. Changes in fracture morphology,i.e., crack paths, corresponded to changes in crack growth kinetics, both of which depended on pressure and temperature. There was no evidence for crack nucleation in advance of the main crack, and this suggests that the fracture process zone is located within one prior-austenite grain diameter from the crack tip. The experimental results indicate that microstructure plays an important role in determining crack growth response. The prior-austenite grain boundaries are seen to be most susceptible to hydrogen embrittlement, followed by the (110)_α’ and (112)_α’ cleavage planes. The martensite matrix, on the other hand, is relatively immune. The observed changes in crack growth rate with temperature and pressure in the higher temperature region are explained in terms of the partitioning of hydrogen into the different microstructural elements and the consequent changes in the micromechanisms for fracture. Leave from the Department of Materials Science, Shanghai Jaio Tong University, Shanghai, People’s Republic of China. Formerly Research Associate, Department of Mechanical Engineering and Mechanics.     

A “Hydrogen partitioning” model for hydrogen assisted crack growth

A “hydrogen partitioning” model has been developed to account for the pressure and temperature dependence for hydrogen-assisted crack growth. The model gives explicit recognition to the role of hydr en-microstructure interactions in determining the distribution (or partitioning) of hydrogen among the various microstructural elements (principally between the prior-austenite grain boundaries and the matrix) and the rate of crack growth along the elements. It also takes into account the role of various rate controlling processes in determining the rate that hydrogen is being supplied to the fracture process (or embrittlement) zone. Quantitative assessment of the model indicates very good agreements between the model predictions and the observed crack growth responses for AISI 4340 and 4130 steels tested in hydrogen and for AISI 4340 steel tested in hydrogen sulfide. This model accurately characterizes the reduction in crack growth rate and the concomitant change in fracture mode at “high” temperatures. Through its integration with the earlier models, based on rate controlling processes, the model predicts the pressure and temperature dependence for K-independent crack growth over the entire range of environmental conditions.     

Gaseous hydrogen-induced cracking of Ti-5Al-2.5Sn

The kinetics of hydrogen-induced cracking have been studied in the Ti-5Al-2.5Sn titanium alloy having a structure of acicular α platelets in a β matrix. It was observed that the relationship between hydrogen-induced crack growth rate and applied stress intensity can be described by three separable regions of behavior. The crack-growth rate at low stress-intensity levels was found to be exponentially dependent on stress intensity but essentially independent of temperature. The crack-growth rate at intermediate stress-intensity levels was found to be independent of stress intensity but dependent on temperature in such a way that crack-growth rate was controlled by a thermally activated mechanism having an activation energy of 5500 cal per mole and varied as the square root of the hydrogen pressure. The crack-growth rate at stress-intensity levels very near the fracture toughness is presumed to be independent of environment. The results are interpreted to suggest that crack growth at high stress intensities is controlled by normal, bulk failure mechanisms such as void coalescence and the like. At intermediate stress-intensity levels the transport of hydrogen to some interaction site along the α-β boundary is the rate-controlling mechanism. The crack-growth behavior at low stress intensities suggests that the hydrogen interacts at this site to produce a strain-induced hydride which, in turn, induces crack growth by restricting plastic flow at the crack tip.     

The role of heat treating on the sour gas resistance of an X-80 steel for oil and gas transport

In this work, the role of the microstructure in the stress sulfide cracking (SSC) resistance of an API X-80 steel was investigated by exposure of as-received and heat-treated specimens to a H₂S-saturated aqueous National Association of Corrosion Engineers (NACE) solution. It was found that for similar corrosive environments and applied stress intensity factors of 30 to 46 MPa√m, crack growth in LEFM (linear elastic fracture mechanics) compact specimens is strongly influenced by heat treating. In the as-received alloy, crack growth in the direction normal to rolling was controlled by metal dissolution of the crack tip region in contact with the corrosive environment, with crack growth rates of the order of 1/W(da/dt)∼8.3×10⁻⁴ h⁻¹. Alternatively, crack growth in the direction parallel to the rolling direction did not show metal dissolution, but instead hydrogen embrittlement along segregation bands. In this case, crack growth rates of the order of 1.2×10⁻³ h⁻¹ were exhibited. In the martensitic condition, the rate of crack propagation was relatively fast (1/W(da/dt)∼4.5×10⁻² h⁻¹), indicating severe hydrogen embrittlement. Crack arrest events were found to occur in water-sprayed and quenched and tempered specimens, with threshold stress intensity values (K _ISSC) of 26 and 32 MPa√m, respectively. Apparently, in the water-sprayed condition, numerous microcracks developed in the crack tip plastic zone. Crack growth occurred by linking of microcracks, which were able to reach the main crack tip. In particular, preferential microcrack growth occurred across carbide regions, but their growth was severely limited in the ferritic matrix. Quenching and tempering (Q&T) resulted in a tempered martensite microstructure characterized by fine distribution carbides, most of which were cementite. In this case, the crack path continually shifted to follow the ferrite interlath boundaries, which contained mostly fine cementite precipitates. As a result, the crack was tortuous with numerous bifurcations along ferrite grain boundaries. Most of the tests were carried out in NaCl-free NACE solutions; the only exception was the as-received condition where 5 wt pct NaCl was added to the sour environment. In this case, crack growth did not occur after exposing the specimen to the salt-free NACE solution for 30 days, but addition of 5 pct NaCl promoted crack propagation.     

Mechanisms of hydrogen induced delayed cracking in hydride forming materials

Mechanisms which have been formulated to describe delayed hydrogen cracking in hydride-forming metals are reviewed and discussed. Particular emphasis is placed on the commercial alloy Zr-2.5 pct Nb (Cb) which is extensively used in nuclear reactor core components. A quantitative model for hydrogen cracking in this material is presented and compared with available experimental data. The kinetics of crack propagation are controlled by the growth of hydrides at the stressed crack tip by the diffusive ingress of hydrogen into this region. The driving force for the diffusion flux is provided by the local stress gradient which interacts with both hydrogen atoms in solution and hydrogen atoms being dissolved and reprecipitated at the crack tip. The model is developed using concepts of elastoplastic fracture mechanics. Stage I crack growth is controlled by hydrides growing in the elastic stress gradient, while Stage II is controlled by hydride growth in the plastic zone at the crack tip. Recent experimental observations are presented which indicate that the process occurs in an intermittent fashion; hydride clusters accumulate at the crack tip followed by unstable crack advance and subsequent crack arrest in repeated cycles.     

A short-time diffusion correlation for hydrogen-induced crack growth kinetics

Analysis of hydrogen-stress field interactions have led to kinetic criteria for slow crack growth. Using both elastic and plastic stress fields under opening-mode loading, criteria for stage I, II, III growth are developed in terms of the pressure tensor gradient at the crack tip. It is proposed that stage I (stress-intensity dependent) growth kinetics are predominantly controlled by the elastic stress field while stage II (nearly stress-intensity independent) kinetics are controlled by the plastic stress field. Measurements of slow crack growth in cathodically-charged AISI 4340 steel verify the overall aspects of the correlation. Detailed measurement and analysis of the increase in crack-tip radius with increasing applied stress intensity have led to a proposed decrease in crack growth rate during stage II growth. Some experimental evidence corroborates this later hypothesis and is consistent with long range diffusional flow of hydrogen as the controlling mechanism for crack growth kinetics. Partial fulfillment of the M.S. Degree at the University of Minnesota     

Embrittlement of a ferrous alloy in a partially dissociated hydrogen environment

Gaseous hydrogen embrittlement of quenched and tempered 4130 steel was studied as a function of temperature from −42° to 164°C in a partially dissociated hydrogen environment at low molecular hydrogen pressures (≈8 × 10⁻³ torr). Atomic hydrogen was created by dissociation of molecular hydrogen on a hot tungsten filament located near a crack opening. The presence of atomic hydrogen was found to increase the rate of hydrogen-induced, slow crack growth by several orders of magnitude and to significantly alter the temperature dependence of embrittlement from what is observed in the presence of molecular hydrogen alone. Based on a previous study, these observations are interpreted in terms of a difference between the hydrogen-transport reaction step controlling hydrogen-induced, slow crack growth in the molecular hydrogen and the atomic-molecular hydrogen environments. Finally, a comparison is made between the kinetics of hydrogen-induced, slow crack growth observed in the presence of atomic-molecular hydrogen and the kinetics of known, possible hydrogen-transport reactions in an effort to identify the reaction step controlling hydrogen embrittlement in the presence of atomic hydrogen.     

Hydrogen embrittlement studies of a trip steel

The conditions of cathodic charging, gaseous hydrogen environment, and loading for which a TRIP steel may or may not be susceptible to hydrogen embrittlement were investigated. In the austenitic state, the TRIP steel appeared to be relatively immune to hydrogen embrittlement. It was shown that it is the strain-induced martensitic phase, α^′, which is embrittled. In TRIP steel single-edge-notch specimens under fixed loads in gaseous hydrogen, slow crack growth occurs when the stress intensity level exceeds a threshold level of about 25 ksi-in.^1/2 and the growth rate varies approximately as the 2.5 power of the stress intensity level. The activation energy for this slow crack growth was found to be about 10,000 cal/g-atom, the approximate activation for hydrogen diffusion in martensite. Thus it was concluded that slow crack growth in TRIP steel loaded in gaseous hydrogen involves the diffusion of hydrogen through the α^′ phase. Formerly with the Lawrence Berkeley Laboratory, Berkeley, Calif.     

Fracture mechanics and surface chemistry studies of subcritical crack growth in AISI 4340 steel

Coordinated fracture mechanics and surface chemistry experiments were carried out to develop further understanding of environment enhanced subcritical crack growth in high strength steels. The kinetics of crack growth were determined for an AISI 4340 steel (tempered at 204°C) in hydrogen and in water, and the kinetics for the reactions of water with the same steel were also determined. A regime of rate limited (Stage II) crack growth was observed in each of the environments. Stage II crack growth was found to be thermally activated, with an apparent activation energy of 14.7 ±2.9 kJ/mole for crack growth in hydrogen, and 33.5 ± 5.0 kJ/mole in water. Fractographic evidence indicated that the fracture path through the microstructure was the same for these environments, and suggested hydrogen to be the embrittling species for environment enhanced crack growth in hydrogen and in water/water vapor. A slow step in the surface reaction of water vapor with steel was identified, and exhibited an activation energy of 36 ± 14 kJ/ mole. This reaction step was identified to be that for the nucleation and growth of oxide. The hydrogen responsible for embrittlement was presumed to be produced during this reaction. On the basis of a comparison of the activation energies, in conjunction with other supporting data, this slow step in the water/metal surface reaction was unambiguously identified as the rate controlling process for crack growth in water/water vapor. The inhibiting effect of oxygen and the influence of water vapor pressure on environment enhanced subcritical crack growth were considered. The influence of segregation of alloying and residual impurity elements on crack growth was also considered.     

Internal hydrogen-induced subcritical crack growth in austenitic stainless steels

The effects of small amounts of dissolved hydrogen on crack propagation were determined for two austenitic stainless steel alloys, AISI 301 and 310S. In order to have a uniform distribution of hydrogen in the alloys, they were cathodically charged at high temperature in a molten salt electrolyte. Sustained load tests were performed on fatigue precracked specimens in air at 0 ‡C, 25 ‡C, and 50 ‡C with hydrogen contents up to 41 wt ppm. The electrical potential drop method with optical calibration was used to continuously monitor the crack position. Log crack velocityvs stress intensity curves had definite thresholds for subcritical crack growth (SCG), but stage II was not always clearly delineated. In the unstable austenitic steel, AISI 301, the threshold stress intensity decreased with increasing hydrogen content or increasing temperature, but beyond about 10 wt ppm, it became insensitive to hydrogen concentration. At higher concentrations, stage II became less distinct. In the stable stainless steel, subcritical crack growth was observed only for a specimen containing 41 wt ppm hydrogen. Fractographic features were correlated with stress intensity, hydrogen content, and temperature. The fracture mode changed with temperature and hydrogen content. For unstable austenitic steel, low temperature and high hydrogen content favored intergranular fracture while microvoid coalescence dominated at a low hydrogen content. The interpretation of these phenomena is based on the tendency for stress-induced phase transformation, the different hydrogen diffusivity and solubility in ferrite and austenite, and outgassing from the crack tip. After comparing the embrittlement due to internal hydrogen with that in external hydrogen, it is concluded that the critical hydrogen distribution for the onset of subcritical crack growth is reached at a location that is very near the crack tip. Formerly Research Assistant, Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign.     

Evaluation of hydrogen-assisted cracking behavior of low-alloy steel in the range 95 °C to 350 °C

In this report, hydrogen-assisted cracking (HAC) behavior of low-alloy steel was evaluated using a constant elongation rate tensile test, and the temperature and crack tip strain rate effects were observed. It was found that temperature dependence of the threshold condition (C _σm ^c ) of HAC above about 100 °C followed the relation C _σm ^c = Kexp(−41,300/Rr) whereK is a constant andT is absolute temperature. The relationship between HAC growth rate and crack tip strain rate was established as almost linear, irrespective of temperature and hydrogen concentration at the crack tip. Hydrogen heat release tests were also performed. From these tests, formation and growth of microcracks which are trap sites of hydrogen were thought to be the mechanism of HAC in the steel. From this mechanism, HAC behavior of the low-alloy steel could be qualitatively explained.     

Chemical and metallurgical aspects of environmentally assisted fatigue crack growth in 7075-T651 aluminum alloy

A comprehensive study has been carried out on a 7075-T651 alloy to examine the influence of water vapor on fatigue crack growth. The kinetics of fatigue crack growth were determined as a function of water vapor pressure at room temperature and at 353 K. Detailed fractographic analyses and surface chemistry studies were carried out to identify the micromechanisms and to quantify the chemical interactions for corrosion fatigue crack growth in this alloy. Experiments were also carried out in ultra-high vacuum and in oxygen to provide for comparisons. Two regions of fatigue crack growth response were identified. In the low pressure region (below 67 Pa at 5 Hz), crack growth is controlled by the rate of transport of water vapor to the crack tip, and the response can be described by a model for transport controlled crack growth. At pressures above 67 Pa, additional increases in crack growth rate occurred, which are attributed to the further reactions of water vapor with segregated magnesium in this alloy. Different micromechanisms for crack growth have been identified for vacuum, oxygen, and water vapor. These micromechanisms are considered in relation to the environmental parameters through a modified superposition model for corrosion fatigue.     

Effects of frequency and temperature on short fatigue crack growth in aqueous environments

The growth of short fatigue cracks in a NiCrMoV steel forging was examined, under constant applied stress intensity range (ΔK = 31 MPa-m^1/2) in deaerated deionized water and 0.3 M Na₂SO₄ solution, as a function of frequency and temperature. Measurements were also made of the kinetics of electrochemical reactions of bare steel surfaces with the deaerated 0.3 M Na₂SO₄ solution, under free corrosion, to provide for comparison and correlation. Fatigue crack growth rate increased with reductions in frequency and with increases in temperature. The maximum amount of crack growth enhancement by the different environments appeared to be equal, although the crack growth response in deionized water appeared to be consistent with a faster reaction rate. The temperature and frequency dependence for corrosion fatigue crack growth corresponded directly with that for charge transfer between the “bare” and “filmed” metal surfaces under free corrosion. The results showed that shortcrack growth in the aqueous environments is controlled by the rate of electrochemical reactions, and is thermally activated with an apparent activation energy of about 40 kJ/M.     

Influence of impurity segregation on temper em brittlement and on slow fatigue crack growth and threshold behavior in 300-M high strength steel

Interactions between hydrogen embrittlement and temper embrittlement have been examined in a study of fracture and low growth rate (near-threshold) fatigue crack propagation in 300-M high strength steel, tested in humid air. The steel was investigated in an unembrittled condition (oil quenched after tempering at 650°C) and temper embrittled condition (step-cooled after tempering at 650°C). Step-cooling resulted in a severe loss of toughness (approximately 50 pct reduction), without loss in strength, concurrent with a change in fracture mode from micr ovoid coalescence to inter granular. Using Auger spectroscopy analysis, the embrittlement was attributed to the cosegregation of alloying elements (Ni and Mn) and impurity elements (P and Si) to prior austenite grain boundaries. Prior temper embrittlement gave rise to a substantial reduction in resistance to fatigue crack propagation, particularly at lower stress intensities approaching the threshold for crack growth(x0394;K _o). At intermediate growth rates (10^-5 to 10^-3 mmJcycle), propagation rates in both unembrittled and embrittled material were largely similar, and only weakly dependent on the load ratio, consistent with the striation mechanism of growth observed. At near-threshold growth rates (<10⁻⁵ to 10⁻⁶ mmJcycle), embrittled material exhibited significantly higher growth rates, 30 pct reduction in threshold ΔK_o values and intergranular facets on fatigue fracture surfaces. Near-threshold propagation rates (and ΔK_o values) were also found to be strongly dependent on the load ratio. The results are discussed in terms of the combined influence of segregated impurity atoms (temper embrittlement) and hydrogen atoms, evolved from crack tip surface reactions with water vapor in the moist air environment (hydrogen embrittlement). The significance of crack closure concepts on this model is briefly described. ntmis]formerly with the Lawrence Berkeley Laboratory, University of California in Berkeley. Formerly with the Lawrence Berkeley Laboratery, University of California in Berkeley.     

Hydrogen-enhanced fatigue crack growth in Ti-6Al-4V ELI weldments

The effects of a hydrogen environment on the fatigue crack growth rates in Ti-6A1-4V ELI (STA) and weld material were determined in the temperature range of ambient to -200°F. The hydrogen environment resulted in an acceleration of the crack growth rate and a change in the fracture mode for both materials in the temperature range of ambient to -100°F. At -200°F, there was no significant difference between the crack growth rates obtained in helium and hydrogen gas. The degree of hydrogén-enhanced crack growth was found to be dependent on the crack tip stress-intensity range, temperature, and microstructure of the material. The data is consistent with an embrittlement mechanism involving hydrogen diffusing ahead of the crack front.     

Initial hydrogen attack kinetics in a carbon steel

The kinetics of the initial stages of hydrogen attack in a commercial 0.3 pct C steel (grade A516) were investigated using anin situ dilatometer. The time, temperature and hydrogen pressure dependences of the rate of sample expansion were measured at hydrogen pressures from 1 to 20 MPa, and temperatures from 350 to 475 °C for sample strains of 10^-6 to 10^-3. Sample expansion began shortly after hydrogen exposure and proceeded at a nearly constant rate throughout the “incubation period” preceding rapid attack. At high temperatures and low pressures, this rate was proportional to P_{H 2} ^1.9±0.2 and had an apparent activation energy of 115 ± 9 kJ. At high pressures and low temperatures, the rate was proportional to PP_{H 2} ^1.0.62±0.07 and showed an apparent activation energy of 210 ± 13 kJ. This suggests that bubble growth during the incubation period occurs predominantly by grain boundary diffusion and is driven by near-equilibrium internal methane pressures. Sample expansion in the subsequent stages of accelerating growth probably is controlled by creep and methane generation. Formerly a Graduate Student at Ohio State University     

Predicting the kinetics of hydrogen generation at the tips of corrosion fatigue cracks

A model has been developed to predict the rate of generation of hydrogen atoms at the tips of fatigue cracks for steel cathodically protected in marine environments. The model incorporates crack-tip chemistry, scraping electrode measurements, and crack-tip deformation. The current density for generation of hydrogen atoms by reduction of water at the crack tip has been calculated for a range of electrochemical and mechanical variables (electrode potential, cyclic frequency, waveform, ΔK, and R value). The crack-tip current density is always greater than on adjacent crack walls and tends to increase with decreasing (more negative) potential. However, at potentials more negative than about-900 mV (SCE), at a cyclic frequency of 0.1 Hz, cathodic reduction of water on the external surface of the steel is predicted to be the dominant source of hydrogen atoms. Decreasing the frequency reduces the crack-tip current density and further emphasizes the dominance of bulk charging. There is little difference in hydrogen charging current densities at the crack tip for sinusoidal, triangular, or positive sawtooth waveforms, but square-wave loading provides a greater charging current. Increasing ΔK and R value have only a small effect on crack-tip current density but increase the area of the active surface and thus lead to more significant charging. Hydrogen-atom concentration profiles in fracture mechanics specimens and in tubular sections have been calculated for conditions in which bulk charging of the steel with hydrogen is dominant. To ensure that crack growth rates are “steady-state” values, test times have to be long enough to establish steady conditions of hydrogen charging. Crack growth data from fracture mechanics specimens may not be directly relevant to cracking in tubular sections because of hydrogen concentration gradients in the latter.     

Fracture mechanics and surface chemistry studies of fatigue crack growth in an aluminum alloy

Fracture mechanics and surface chemistry studies were carried out to develop further understanding of the influence of water vapor on fatigue crack growth in aluminum alloys. The room temperature fatigue crack growth response was determined for 2219-T851 aluminum alloy exposed to water vapor at pressures from 1 to 30 Pa over a range of stress intensity factors (K). Data were also obtained in vacuum (at < 0.50 μPa), and dehumidified argon. The test results showed that, at a frequency of 5 Hz, the rate of crack growth is essentially unaffected by water vapor until a threshold pressure is reached. Above this threshold, the rates increased, reaching a maximum within one order of magnitude increase in vapor pressure. This maximum crack growth rate is equal to that obtained in air (40 to 60 pct relative humidity), distilled water and 3.5 pct NaCl solution on the same material. Parallel studies of the reactions of water vapor with fresh alloy surfaces (produced either byin situ impact fracture or by ion etching) were made by Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS). The extent of surface reaction was monitored by changes in the oxygen AES and XPS signals. Correlation between the fatigue crack growth response and the surface reaction kinetics has been made, and is consistent with a transport-limited model for crack growth. The results also suggest that enhancement of fatigue crack growth by water vapor in the aluminum alloys occurs through a “hydrogen embrittle ment” mechanism.     

Hydrogen-induced slow crack growth in Ti-6Al-6V-2Sn

The effect of hydrogen and temperature on threshold stress intensity and crack growth kinetics was studied in Ti-6Al-6V-2Sn containing 38 ppm hydrogen. A slight decrease in threshold values occurred as temperature decreased from 300 K while they increased significantly above 300 K. For a given test temperature, crack growth rates exhibited an exponential dependence on stress intensity over a major portion of growth. At 300 K the rates reached a maximum. Slow crack growth occurred predominately by cleavage ofα grains which has been associated with hydride formation. The stress intensity required for hydride formation at a crack tip can be determined from hydrogen concentration and solubility considerations under stress. As these values differed from observed thresholds, a strong influence of microstructure was suggested and subsequently revealed by crack front examination. Quantification of this effect with a modified Dugdale-Barenblatt model relates the effective stress intensity at the crack tip to the applied stress intensity. Microstructure was also found to exert a strong influence on slow crack growth behavior when examined in terms of the effective stress intensity,K _eff. From Arrhenius plots of crack growth rates for variousK _eff, activation energies of 27.0 to 32.8 kJ/mol were obtained and related to the diffusion of hydrogen through theβ phase. The increase in crack growth rates with increasing temperatures up to 300 K is attributed to the temperature dependence of hydrogen diffusion. The decrease in crack growth rates above 300 K is related to a hydride nucleation problem.     

Hydrogen enhanced crack growth in 18 Ni maraging steels

The kinetics of sustained-load subcritical crack growth for 18 Ni maraging steels in high purity hydrogen are examined using crack-tip stress intensity,K, as a measure of crack driving force. Crack growth rate as a function of stress intensity exhibited a clearly definedK-independent stage (Stage II). Crack growth rates in an 18 Ni (250) maraging steel are examined for temperatures from -60°C to 100°C. A critical temperature was observed above which crack growth rates became diminishingly small. At lower temperatures the activation energy for Stage II crack growth was found to be 16.7 ± 3.3 kJ/mole. Temperature and hydrogen partial pressure are shown to interact in a complex manner to determine the apparentK _th and the crack growth behavior. Comparison of results on ‘250’ and ‘300’ grades of 18 Ni maraging steel indicate a significant influence of alloy composition and/or strength level on the crack growth behavior. These phenomenological observations are discussed in terms of possible underlying controlling processes. Formerly a Graduate Student and Research Assistant. Based on a thesis submitted in partial fulfillment of requirements for the M.S. degree in Metallurgy and Materials Science, Lehigh University, Bethlehem, PA.