Advanced Materials Research Vols. 891-892

Abstract: This paper describes the results of a research study conducted to improve the understanding of fatigue crack growth rate behavior in the threshold growth rate region and to answer a question of the validity of threshold region test data. The validity question relates to the position held by some experimentalists that using the ASTM load shedding test method does not produce valid threshold test results and material properties. The question involves the fanning behavior observed in threshold region of da/dN plots for some materials in which the low R data fans out from the high R data. This fanning behavior or elevation of threshold values in the low R tests is generally assumed to be caused by an increase in crack closure in the low R tests. Also, the increase in crack closure is assumed by some experimentalists to result from using the ASTM load shedding test procedure [1-3]. The belief is that this procedure induces load history effects which cause remote closure from plasticity and/or roughness changes in the surface morphology. However, experimental studies performed by the authors have shown that the increase in crack closure is more a result of extensive crack tip bifurcations that can occur in some materials, particularly in aluminum alloys, when the crack tip cyclic yield zone size becomes less than the grain size of the alloy. This behavior is related to the high stacking fault energy (SFE) property of aluminum alloys which results in easier slip characteristics. Therefore, the fanning behavior which occurs in aluminum alloys is a function of intrinsic dislocation property of the alloy, and therefore, the fanned data does represent the true threshold properties of the material. However, for the corrosion sensitive steel alloys tested in laboratory air, the occurrence of fanning is caused by fretting corrosion at the crack tips, and these results should not be considered to be representative of valid threshold properties because the fanning is eliminated when testing is performed in dry air.

Abstract: In this paper, an analytical method is presented for crack growth threshold criterion based on calculation of dynamic stress intensity factors (DSIF) in the mixed mode. The basis of the method is comparison of DSIF and fracture toughness. The analytical method is grounded on solving two-dimensional wave equations in the frequency domain, using complex functions and mapping technique. Using this method, the J-integrals are calculated and based on relation between the J-integrals and DSIF, these factors are derived. The proposed criterion for crack growth threshold is defined as a comparing calculated DSIF (K_I, K_II) with fracture toughness (K_c), using complex functions and mapping technique. The mapping technique helps to convert the complicated boundaries geometry into a simple one (unit circle). Finally, since using complex functions theory is an analytical method, the results have a high precision compared with numerical methods.

Abstract: The very high cycle fatigue behaviour of high-strength steels is mostly affected by cracks which were initiated at non-metallic inclusions inside the fatigue specimens and hence under the surface. By separating the data points in the S-N-diagram by the chemical composition of the inclusion at which a crack is initiated the authors recognized that the lifetime depends strongly on the inclusion type. The authors further deduce that the chemical composition of an inclusion as well as its geometry influences the mechanism by which such an inclusion triggers crack initiation. Thus, titanium nitride and homogeneous calcium oxide inclusions have been observed to show fractures caused by the stress concentration in the inclusion. The broken inclusions then present sharp notches in the steel matrix and cause further damage from there. In contrast, aluminium calcium oxide inclusions decay or detach from the steel matrix during loading on account of a rather low interface stress. This detachment process results in holes in the steel matrix, which act as logical starting points of fatigue cracks. Both processes described above occur at different stress intensity factors and lead to failure before an ultimate number of cycles of 10^9. Furthermore, it was possible to determine threshold values of failure for each inclusion type by stressing run out specimens on a higher stress level. These threshold values are in accordance with those of specimens which failed during one-level stressing. The fatigue tests were performed with 100Cr6 in martensitic and bainitic condition. Tension/compression tests at a load ratio of R = -1 were conducted on an ultrasonic fatigue testing facility. Failures were only initiated at non-metallic inclusions. The fracture surfaces were analyzed by scanning electron microscopy and by energy dispersive X-ray spectroscopy.

Abstract: The two most promising approaches to determine Stress Intensity Factor (SIF) developedover the past decade are the Symmetric Galerkin Boundary Element Method - Finite Element Method(SGBEM-FEM) based alternating method and the Extended Finite Element (XFEM) method. Thepurpose of this paper is to determine the SIFs for a number of 2-D crack problems by the two ap-proaches and measure their relative effectiveness in terms of accuracy, speed and computational re-sources.In the SGBEM-FEM alternating method, a finite element analysis is carried out on the un-crackedbody using the externally applied loading and next a boundary element analysis is performed byreversing the stresses found on the crack location from the finite element analysis, and the residualstresses on the boundary of the finite body are determined. The steps are repeated until convergenceis achieved where the residual stresses on the boundaries and traction on the crack surfaces are closeto zero.In the XFEM method, the mesh is created without considering the topology of the crack configura-tion and the discontinuities are handled by special discontinuity enrichment functions. The enrichmentfunctions increase the degrees of freedom and the regular stiffness matrix is augmented by additionalterms corresponding to the extra degrees of freedom but the increase in computational requirement isoffset by not having the burden of remeshing the finite elements.Both SGBEM-FEM alternating method and XFEM method are used to solve a number of crackproblems and the example cases clearly show the computational efficiency of the SGBEM-FEM al-ternating method over the XFEM method.

Abstract: The railway axles are subjected to cyclic loading, therefore there is a risk of fatigue failure. For reason that possible crack could not be detected by non-destructive testing method an existing crack in the railway axle must be considered. This is conservative approach commonly used in applications where potential fatigue failure has unaccepted consequences. This paper deals with retardation effect caused by overload cycles and compares results obtained by no retardation approach and results obtained by generalized Willenborg model, which takes into account the retardation effects due to plastic zone around the crack tip. Results obtained can contribute to the better understanding of fatigue crack behavior in railway axles.

Abstract: Pariss law of fatigue crack propagation rate is well applied in the defect-tolerance fatigue approach. When carry out same approach in the very high cycle fatigue domain, the understanding of mechanism about fatigue crack propagation threshold which is obviously important, is helped. In the present work here, the fatigue crack propagation threshold of a surface crack for an Armco iron loaded in the VHCF regime was investigated by a new approach which combines the fracture surface analysis and the temperature recording on the surface during the test by an infra-red camera. The experiments were carried out on a sheet specimen under a 20 kHz ultrasonic frequency loading with IR images registration. Three stages of fatigue crack were identified with different mechanisms. It is found that the transition between initiation and crack propagation corresponds to the intrinsic fatigue threshold. It takes more than 99% of the gigacycle fatigue life to achieve this transition size.

Abstract: Crack propagation tests of lead-free solder were conducted at room temperature in air using center-notched plates under load-controlled conditions with three waveforms: triangular pp waveform having fast loading and unloading rates, cp-h waveform having a hold time under tension, and cc-h waveform having a hold time under tension and compression. The J integral was evaluated from load-displacement curves. For fatigue loading of pp waveform, the crack propagation rate was expressed as a power function of the fatigue J-integral range. The creep component due to the hold time greatly accelerated the crack propagation rate. The creep crack propagation rate was found to be a power function of the creep J integral range for each case of cp-h and cc-h waveforms. The creep crack propagation rate for cp-h waveform was higher than that for cc-h waveform. Displacement-controlled tests were also performed under four triangular strain waveforms: pp, cp, cc and pc. For the case of pp waveform, the crack propagation rate was also expressed as the same power function of the fatigue J integral range as in the case of load-controlled tests. The creep crack propagation rate was expressed as a power function of the creep J integral range for each case of cp, pc and cc waveforms. Microscopic observations were conducted to clarify micromechanisms of creep-fatigue crack propagation.

Abstract: A novel test method is presented in this paper for determination of the creep crack growth threshold. In this method, a wedge is placed in the notch of a standard compact tension (CT) specimen with a pre-crack, which provides an evaluable stress fields characterized with stress intensity factor (K) at the crack tip. Then the specimen is kept in the test temperature for a certain time. The creep crack growth threshold of the material at a given temperature can be obtained by extrapolating the curves of the crack growth rate vs. K in the double logarithmic coordinates. Since it does not need mechanical testing machine and several specimens can be tested simultaneously, the proposed method is economical in both costs and time. Tests on a nickel-based powder metallurgy superalloy FGH97 are conducted using this method at 600°C, 650°C and 750°C, which indicate its feasibility.

Abstract: Austenitic stainless steel grade UNS S31035 (Sandvik Sanicro® 25) has been developed for the next generation of 700°C A-USC power plant. This paper will mainly focus on the study of low cycle fatigue behavior and damage mechanisms of the material at room temperature, 600C to 700C by using electron back scatter diffraction and electron channeling contrast image techniques. At room temperature, the material shows a hardening and softening behavior as usual. At high temperature, however, it shows only a cyclic hardening behavior. Dynamic strain ageing can be one of the mechanisms. The damage and fatigue crack initiation mechanisms due to cyclic loading at different temperatures and loading conditions have been identified. The interactions between dislocations or slip bands with grain boundary or twin boundary are the main damage mechanism at low temperature or at high temperature with large strain amplitudes. Strain localization due to dislocation slipping is the main mechanism for the damage in grain.

Abstract: Reduced activation ferritic/martensitic (RAFM) steels are candidate materials for the test blanket modules of ITER. Several degradation mechanisms such as thermal fatigue, low cycle fatigue, creep fatigue interaction, creep, irradiation hardening, swelling and phase instability associated irradiation embrittlement must be understood to estimate the component lifetime. The current work focuses on the effect of tungsten and tantalum on low cycle fatigue (LCF) and creep-fatigue interaction (CFI) behavior of four RAFM steels with varying W and Ta contents. Total strain controlled LCF experiments were performed under various strain amplitudes in the range +0.25% to +1% and temperatures (300 K to 873 K) in air at a constant strain rate of 3×10^-3s^-1 using a servo hydraulic fatigue testing system. CFI experiments were carried out at total strain amplitude of +0.6% and by applying strain hold of different durations (10 min and 30 min) in peak tension and peak compression. Both LCF and CFI life of the RAFM steels improved with the increase in tungsten and tantalum contents. Based on the amount of softening during continuous cycling, tungsten content was optimized at 1.4 wt. % and the tantalum content at 0.06 wt%. Stress relaxation obtained during creep-fatigue interaction studies showed close relation with the chemical composition of the RAFM steels. Other damaging parameters influencing fatigue life were dynamic strain ageing (DSA) occurring in the intermediate temperature regime and oxidation at elevated temperatures. Keywords: RAFM steel, low cycle fatigue, dynamic strain ageing, creep-fatigue interaction, oxidation