The effect of residual stress on the fatigue crack growth behavior of Al-Si-Mg cast alloys—Mechanisms and corrective mathematical models

The fatigue crack growth (FCG) behavior of various types of alloys is significantly affected by the presence of residual stress induced by manufacturing and post-manufacturing processes. There is a qualitative understanding of the effects of residual stress on fatigue behavior, but the effects are not comprehensively quantified or accounted for. The difficulty in quantifying these effects is largely due to the complexity of residual-stress measurements (especially considering that parts produced in similar conditions can have different residual-stress levels) and the lack of mathematical models able to convert experimental data with residual stress into residual-stress-free data. This article provides experimental, testing, and mathematical techniques to account for residual-stress effects on crack growth rate data, together with two methods for eliminating residual stresses in crack growth test specimens. Fracture-mechanics concepts are used to calculate, in simple and convenient ways, stress-intensity factors caused by residual stresses. The method is advantageous, considering that stress-intensity factors are determined before the actual test is conducted. Further on, residual-stress-intensity factors are used to predict the residual-stress distribution in compact tension (CT) specimens prior to testing. Five cast Al-Si-Mg alloys with three Si levels (in unmodified (UM) as well as Sr-modified (M) conditions) were analyzed both with and without residual stress. Fatigue cracks are grown under both constant stress ratio, R=0.1, and constant maximum stress-intensity factor, K _max = const., conditions. The mechanisms involved in crack growth through residual-stress fields are presented.     

Fatigue crack growth behavior in niobium-hydrogen alloys

Near-threshold fatigue crack growth behavior has been investigated in niobium-hydrogen alloys. Compact tension specimens (CTS) with three hydrogen conditions are used: hydrogen-free, hydrogen in solid solution, and hydride alloy. The specimens are fatigued at a temperature of 296 K and load ratios of 0.05, 0.4, and 0.75. The results at load ratios of 0.05 and 0.4 show that the threshold stress intensity range (ΔK _th ) decreases as hydrogen is added to niobium. It reaches a minimum at the critical hydrogen concentration (C _cr ), where maximum embrittlement occurs. The critical hydrogen concentration is approximately equal to the solubility limit of hydrogen in niobium. As the hydrogen concentration exceeds C _cr , ΔK _th increases slowly as more hydrogen is added to the specimen. At load ratio 0.75, ΔK _th decreases continuously as the hydrogen concentration is increased. The results provide evidence that two mechanisms are responsible for fatigue crack growth behavior in niobium-hydrogen alloys. First, embrittlement is retarded by hydride transformation-induced and plasticity-induced crack closures. Second, embrittlement is enhanced by the presence of hydrogen and hydride.     

Fatigue crack growth and stress redistribution at interfaces

The role of the interface in redistributing stress around cracks in multilayered ceramic/metal composites is investigated. The emphasis is on the different effects of interfacial debonding or of plastic slip in the metal phase adjacent to strongly bonded interfaces. The experiments are conducted on alumina/aluminum multilayered composites. Monotonic loading precracked test pieces causes plastic shear deformation within the aluminum layer at the tip of the notch without debonding. However, interfacial debonding can be induced by cyclic loading, in accordance with a classical fatigue mechanism. Measurements of the stress around the crack demonstrate that debonding is much more effective than slip at reducing the stress ahead of the crack.     

Fatigue damage and crack nucleation mechanisms at intermediate strain amplitudes

Polycrystalline copper was fatigued in rotary bending at constant intermediate surface strain amplitudes at 26 Hz under ambient conditions. The specimens were interrupted at various life fractions, their surfaces prepared metallographically and scrutinised to ascertain the types of fatigue damages, namely, short cracks which are confined to individual grains or isolated grain boundary facets, and their role in fatal crack formation. The results show that, at intermediate strain amplitudes, slip band and twin boundary crack damages predominate during early stages of cycling, while grain boundary crack damages remain relatively insignificant even at the stage when fatal cracks have developed. However, depending on the strain amplitude level, the transgranular crack damages may or may not be instrumental in fatal crack formation. At the lower amplitude end of the transition region, fatal cracks are formed by interlinkage of slip band and twin boundary damages. At the higher amplitude end, even though grain boundary damages are negligible initially, they degenerate rapidly on further cycling and eventually evolve into fatal cracks. The present findings show that some 0.05% plastic strain amplitude is required to propagate intergranular cracks. Once the above condition is met, cracks would propagate rapidly along the interface and the crack nucleation mode would change from transgranular to intergranular.     

The effects of microstructure on elevated temperature crack growth in nickel-base alloys

The effects of microstructure on the fatigue and creep crack growth of Waspaloy and P/M Astroloy were evaluated at 650°C. In Waspaloy, changes in γ′ size and distribution did not markedly affect fatigue crack growth. An increase in fatigue crack growth rate occurred at low test frequencies and was associated with a transition to intergranular crack propagation. In P/M Astroloy, a coarser grain size lowered the fatigue crack growth rate. Serrated grain boundaries, though beneficial under creep loading, have no effect in fatigue.     

Solid solution decomposition mechanisms in cast and microcrystalline Al-Sc alloys: III. Analysis of experimental data

The results of experimental studies of the decomposition of the solid solution in cast and microcrystalline (MC) Al-X wt % Mg-0.22 wt % Sc-0.15 wt % Zr (X = 0, 1.5, 4.5) aluminum alloys that were described in part I of this work are analyzed. The data on the electrical resistivity of the alloys are used to determine the volume fraction of Al₃Sc(Zr) particles precipitating in the temperature range 240–400°C. The model developed in part II of this work is used to reveal the mechanism of particle precipitation in the cast and MC alloys. It is shown that the particles in the cast and MC alloys precipitate mainly on dislocations. The differences in the decomposition kinetics of the solid solution in the cast and MC alloys are explained by different mechanisms of dislocation structure recovery mechanisms occurring in them.     

The effect of Mg on the microstructure and mechanical behavior of Al-Si-Mg casting alloys

The microstructure and tensile behavior of two Al-7 pct Si-Mg casting alloys, with magnesium contents of 0.4 and 0.7 pct, have been studied. Different microstructures were produced by varying the solidification rate and by modification with strontium. An extraction technique was used to determine the maximum size of the eutectic silicon flakes and particles. The eutectic Si particles in the unmodified alloys and, to a lesser extent, in the Sr-modified alloys are larger in the alloys with higher Mg content. Large Fe-rich π-phase (Al₉FeMg₃Si₅) particles are formed in the 0.7 pct Mg alloys together with some smaller β-phase (Al₅FeSi) plates; in contrast, only β-phase plates are observed in the 0.4 pct Mg alloys. The yield stress increases with the Mg content, although, at 0.7 pct Mg, it is less than expected, possibly because some of the Mg is lost to π-phase intermetallics. The tensile ductility is less in the higher Mg alloys, especially in the Sr-modified alloys, compared with the lower Mg alloys. The loss of ductility of the unmodified alloy seems to be caused by the larger Si particles, while the presence of large π-phase intermetallic particles accounts for the loss in ductility of the Sr-modified alloy.     

Fatigue crack growth mechanisms in Ti6242 lamellar microstructures: Influence of loading frequency and temperature

Fatigue crack growth experiments were carried out on Ti6242 alloy with large colony size. The alloy was heat treated to provide three different lamella size; fine, coarse, and extra coarse. Tests were conducted at two temperatures, 520 °C and 595 °C, using two loading frequencies, 10 and 0.05 Hz. The latter frequency was examined with and without a 300-second hold time. All tests were performed in air environment and at a stress ratio of 0.1. This study shows that at 520 °C, the Fatigue crack growth rate (FCGR) is not significantly influenced by changes in the microstructure. For 0.05 Hz/low ΔK, however, the FCGR is higher in the fine lamellar microstructure and is accompanied by- the appearance of a plateau, which disappears in the extra large lamella microstructure. Furthermore, the addition of a 300-second hold time does not alter the crack growth rate. At 595 °C, while the general level of the FCGR is higher than that at 520 °C, the effects of loading frequency and hold time remain similar to those reported at the lower temperature. Unlike the results at 520 °C, however, the FCGR at low δK is not influenced by variations in lamellar microstructure. Under all test conditions, the fatigue process is predominantly controlled by one single mechanism associated with transcolony fracture and formation of quasi-cleavage facets. The fatigue crack growth results and the associated fracture behavior as obtained in this study are correlated to the crack-tip shear activity and transmission at the α/β interfaces. A general hypothesis accounting for the role of loading frequency, temperature, and microstructure on the observed cracking mechanisms is presented.     

Subcritical crack growth at bimaterial interfaces: Part III. shear-enhanced fatigue crack growth resistance at polymer/metal interface

Fatigue crack growth along an Al/epoxy interface was examined under different combinations of mode-I and mode-II loadings using the flexural peel technique. Fatigue crack growth rates were obtained as a function of the total strain energy rate for G_II/G_I ratios of 0.3 to 1.4, achieved by varying the relative thickness of the outerlayers for the flexural peel specimen. Fatigue crack growth resistance of the interface was found to increase with increasing G_II/G_I ratio. Such a shear-enhanced crack growth resistance of the interface resulted in a gradual transition of crack growth mechanism from interfacial at the low G_II/G_I ratio to cohesive at the high G_II/G_I ratio. Under predominantly mode-I loading, the damage in the polymer took the form of crazing and cavitation. In contrast, laminar shear occurred under highly shear loading, resulting in a larger amount of plastic dissipation at the crack tip and improved fatigue crack growth resistance.     

Fatigue crack growth behavior in inconel 706 at 297 K and 4.2 K

The near-threshold fatigue crack growth rate (FCGR) behavior of Inconel 706 was investigated at ambient (297K) and liquid helium (4.2 K) temperatures, respectively. Specimen orientation did not affect the FCGR properties of Inconel 706. At 297 K, a significant influence of R-ratio on the rates of crack propagation was observed while at 4.2 K, the R -ratio effect was minimal. The extent of oxide-induced crack closure was shown to be insignificant in influencing near-threshold crack growth kinetics of Inconel 706 at both temperatures of 297 and 4.2 K. Roughness-induced crack closure was believed to be the dominant mechanism responsible for the influence of R -ratio on the FCGR properties of Inconel 706; this was proved quantitatively by direct crack closure measurements conducted at 297 and 4.2 K. A greater degree of roughness-induced crack closure was observed at 297 K than at 4.2 K; this correlated with the more pronounced R-ratio effect at 297 K. Decreasing the temperature from 297 to 4.2 K decreased the growth rates of fatigue cracks in Inconel 706. The effect of temperature on crack propagation behavior increased with increasing R-ratio. Crack closure could not rationalize this temperature effect. Moreover, the increase in material strength or Young's modulus on cooling from 297 to 4.2 K could not totally account for the influence of temperature on the near-threshold FCGR properties. Dislocation dynamics appears to offer a qualitative explanation for this temperature effect.     

Fatigue crack growth through partially stabilized zirconia at ambient and elevated temperatures

Fatigue crack growth through magnesia stabilized zirconia at 20, 450 and 650°C has been observed dynamically in a high temperature loading stage for the scanning electron microscope. Crack tip micromechanics parameters were measured using the stereoimaging technique. Fatigue crack growth at ambient temperature was found to be very similar to crack growth through metallic alloys. With increasing temperature, the stress intensity levels in which stable fatigue crack growth could be sustained were found to narrow significantly, until fatigue is expected to not be a valid mechanism of crack growth above about 750°C. Measured crack tip parameters were used to derive the low-cycle fatigue and the stress-cycles to failure characteristics. The latter agreed with measured SN curves. Deformation within the plastic zone was shown to account for the measured value of fracture toughness. The mechanisms of crack growth are discussed.     

Fatigue crack propagation in rapidly solidified aluminum alloys at 25 °C and 300 °C

The fatigue crack propagation performance of two rapidly solidified aluminum alloys was investigated in air at 25°C and 300°C. The results show that the crack propagation rates for continuous cycling tests of Al-8Fe-4Ce and Al-4.7-Fe-4.7Ni-0.2Cr alloys were similar at 25°C. Although the crack propagation rates of both alloys were increased at 300°C, the Al-Fe-Ce alloy exhibited the greater resistance to crack propagation. The inclusion of a tensile hold time in the fatigue loading cycle at 300°C produced an increase in the crack propagation rates for both alloys over the rates for continuous cycling. The fatigue crack propagation performance of the rapidly solidified alloys was not found to be superior when compared with the fatigue crack propagation performance of a wrought aluminum alloy tested under the same conditions. Transmission and scanning electron microscopy study of the tested specimens revealed that the crack propagation mode was primarily transgranular, with the metastable dispersoid particles providing impenetrable barriers to dislocation motion.     

Fatigue crack propagation behavior of titanium alloys 6242S and 5621 S at elevated temperature

The influence of test environment on the fatigue crack propagation performance of titanium alloys Ti-6A1-2Sn-4Zr-2Mo-O. 1Si (6242S) and Ti-5A1-6Sn-2Zr-0.8Mo-0.25Si (5621 S) was investigated at 450 °C. The results show that the crack propagation resistance of 6242S and 5621S was reduced in air compared to that in vacuum at a loading frequency of 0.17 Hz. A similar environmental behavior was observed when a one-minute hold period was included in the fatigue cycle. Both alloys exhibited excellent creep-fatigue resistance at 450 °C as little effect of hold time was observed. Overall, the fatigue crack propagation resistance of 6242S was found to be slightly superior to that of 5621S. These results are discussed in terms of the effects of thermal-mechanical history and microstructure on the fatigue crack propagation characteristics of titanium alloys. This paper is based on a presentation made in the symposium “Crack Propagation under Creep and Creep-Fatigue" presented at the TMS/AIME fall meeting in Orlando, FL, in October 1986, under the auspices of the ASM Flow and Fracture Committee.     

Fatigue crack propagation in aluminum-lithium alloy 2090: Part II. small crack behavior

A study has been made of the mechanics and mechanisms of fatigue crack propagation in a commercial plate of aluminum-lithium alloy 2090-T8E41. In Part II, the crack growth behavior of naturallyoccurring, microstructurally-small (2 to 1000μm) surface cracks is examined as a function of plate orientation, and results compared with those determined in Part I on conventional long (≳5 mm) crack samples. It is found that the near-threshold growth rates of small cracks are between 1 to 3 orders of magnitude faster than those for long cracks, subjected to the same nominal stress intensity ranges (at a load ratio of 0.1). Moreover, the small cracks show no evidence of an intrinsic threshold and propagate at ΔK levels as low as 0.7 MPa{ie563-01}, far below the long crack threshold ΔK_TH. Their behavior is also relatively independent of orientation. Such accelerated small crack behavior is attributed primarily to restrictions in the development of crack tip shielding (principally from roughness-induced crack closure) with cracks of limited wake. This notion is supported by the close correspondence of small crack results with long crack growth rates plotted in terms of ΔK_eff (i.e., after allowing for closure above the effective long crack threshold). Additional factors, including the different statistical sampling effect of large and small cracks with microstructural features, are briefly discussed.     

Fatigue crack initiation and microcrack growth in 4140 steel

Initiation and growth of fatigue microcracks were studied in vacuum degassed 4140 steel in three conditions: as-quenched, tempered at 400°C, and tempered at 650°C. Micro-scopic examinations were made of specimens with metallographically polished notches using a 400 times long working distance microscope with an x-y micrometer base mounted directly on an MTS machine. Following Barsom and McNicol, the cycles to fatigue crack initiationN _i were plottedvs ΔK/√p and threshhold values of gDK√p were determined. The data of logN _i vs log [ΔK/√p-ΔK/√p|_th] fit on a straight line. Microcracks grew most rapidly in as-quenched specimens and least rapidly in 650°C tempered specimens at the same ΔK/√p. In as-quenched specimens, fatigue cracks initiated at grain boundaries but in the 400 and 650°C tempered specimens they initiated at intrusions-extrusions. They are also associated with Northwestern’s Materials Research Center.     

Solid solution decomposition mechanisms in cast and microcrystalline Al-Sc alloys: I. Experimental studies

The structure and the electrical and mechanical properties of Al-Mg-0.22 wt % Sc-0.15 wt % Zr alloys with various magnesium contents (0, 1.5, 4.5 wt %) are experimentally studied during the decomposition of the solid solution of scandium and zirconium in the states after solidification from a melt (cast ingots) and after subsequent multicycle equal-channel angular pressing (microcrystalline structure). The dependences of electrical resistivity ??, microhardness HV, macroelasticity limit ??₀, and yield strength ??_y on the annealing temperature and time are analyzed.     

Fracture and fatigue behavior of sintered steel at elevated temperatures: Part II. Fatigue crack propagation

Fatigue cracking resistance of sintered steel as a function of temperature is characterized by crack growth rate vs the stress intensity range, ΔK. The stress ratio effects on fatigue crack propagation (FCP) are investigated from room temperature to 300 °C. The crack closure effects on FCP are evaluated by both theoretical and experimental approaches. We found that the crack closure cannot be fully responsible for the observed increase of fatigue resistance with low stress ratio. Experimental results support that both K _max and ΔK control near-threshold crack growth. Fatigue crack resistance at high ΔK regime decreases with temperature. The apparent increase of fatigue resistance at the near-threshold regime at elevated temperatures might be attributed to microcrack toughening.