M. W. Brown

Shear mode crack growth and rolling contact fatigue

Abstract Mixed mode crack growth direction criteria have been applied to calculations of the loading experienced by surface-initiated rolling contact fatigue cracks growing at a shallow angle to the surface. It was demonstrated that the cracks grow on the plane of the maximum shear stress rather than perpendicular to the maximum tangential stress as occurs during conventional fatigue testing. Mixed mode fatigue crack growth tests were then performed on rail steel specimens in an investigation of the effect of loading cracks in shear. It was demonstrated that while conventional mixed mode testing always produces cracks that either arrest after at most a millimetre of coplanar growth or branch to grow perpendicular to the tensile stress, under certain conditions similar to those experienced by rolling contact fatigue cracks, growth on the plane of the maximum shear stress range does occur. Additionally, under these conditions the cracks can grow up to 10 times faster than would be predicted by conventional fatigue testing.

Modelling the three-dimensional behaviour of shallow rolling contact fatigue cracks in rails

Abstract A model of the fatigue behaviour of three-dimensional (3D) semi-elliptical shallow-angle rolling contact fatigue (RCF) cracks was developed by combining numerically obtained (3D FEM) linear elastic fracture mechanics (LEFM) crack front loading histories with mixed-mode fatigue crack growth rate data. The model was applied for the “squat” type of crack—typical of rail RCF failure, to predict (i) shape development, (ii) co-planar extension, (iii) spalling, and (iv) branching to transverse defects. The model demonstrated that propagation of “squats” under the rail surface is feasible when a fluid entrapment mechanism is introduced, encouraging a mixed-mode shear dominated growth. However, cracks may branch to a Mode I direction when the residual stress, crack inclination and braking force create favourable conditions.

A branch criterion for shallow angled rolling contact fatigue cracks in rails

Abstract A well-developed ‘squat’ grows at about 10° to the upper surface of the rail. It is observed that this shallow angled rolling contact fatigue crack may change its growth direction; either to branch downwards and grow at about 55° to the rail surface or to branch upwards. Various research workers have shown that the crack is subjected to a characteristic sequence of tensile and shear stresses. The downward branches are more detrimental to the integrity of rails because they can lead to fracture. Therefore it is necessary to identify the condition that transforms the stable shallow-angled crack to a branch. Experimental results on a cruciform specimen subject to the characteristic load sequence show that the crack growth direction is determined by the effective mode I stress intensity factor range and degree of overlap in mixed mode loading.

A model for sliding mode crack closure part II: mixed mode I and II loading and application

Abstract The physical model proposed in Part I has been extended to quantify sliding mode crack closure (SMCC) under mixed mode I and lI loading conditions, with predominantly mode lI loading. The results reveal that a small nominal mode I component tends to reduce the local wedging mode I stress intensity and consequently reduces friction attenuation to increase the effective mode II stress intensity. Local stress fields are significantly disturbed due to crack morphology with a clear increase of mixed mode ratio, leading to a predominantly mode I loading condition at the crack tip. The extended model has been used to rationalize the mixed mode fatigue threshold results under predominantly mode II loading conditions for a structural steel BS4360 50D. The effect of load ratio observed experimentally is discussed in the light of fretting-which is believed to be the main mechanism contributing to sliding mode crack closure.

The formation and propagation of mode I branch cracks in mixed mode fatigue failure

Abstract Fatigue crack propagation behaviour under mode II and mixed mode I and II loadings has been investigated in a weldable structural steel BS4360 50D, using an asymmetrical four point bend arrangement. The results show that the conditions for the formation and propagation of mode I branch cracks are of decisive importance for fatigue failures under combined loading conditions. Various aspects of the subject, including fatigue thresholds, branch angle, path and branch crack growth rate, have been studied and the results are reported in this paper. While cracks have been observed to propagate in the maximum ΔKI direction as widely reported, complications associated with the conditions to initiate such branch cracks are far less well understood. In addition, simple tools are in demand to measure and predict branch cracking in this well tried test system. Attempts towards these objectives have been made in this work.

Some aspects of fatigue thresholds under mode III and mixed mode III and I loadings

Abstract Fatigue threshold behaviour under mode III and in-phase mixed mode III and I loading conditions has been studied in a 3.5% NiCrMoV steel, using V-notched and precracked cylindrical bar specimens at load ratios R = 0.1 and −1. Comparisons have been made between the results from V-notched, precracked specimens and the results from earlier work using slit specimens. While little influence of R -ratio was found for slit specimens, R -ratio appears to be a decisive component in defining mixed mode fatigue thresholds in precracked specimens. Ambiguities arise as to the best strategy for mixed mode fatigue tests. Crack surface interference appears to be the predominant force to modify fatigue crack growth behaviour under mixed mode loading conditions. In such circumstances the measured mixed mode fatigue thresholds are generally geometry dependent and failure analyses based on LEFM can not be applied directly.

Characterization of 2024-T351 friction stir welding joints

Characterization of macrostructure, microstructure, hardness, precipitate distribution, residual stress, and cyclic deformation behavior of 2024-T351 friction stir welded joints has been conducted. Inhomogeneous microparameters governing the nonuniform residual stresses and cyclic strength are discussed. The cyclic strength of the weld microregimes is controlled by grain size and distribution of precipitates achieved during the weld process. The comprehensive information of micro-and macromechanics is used to assist in understanding the mechanism that governed the fatigue crack initiation, propagation, and life of the welded joints.

A Two Dimensional Analysis of Mixed-Mode Rolling Contact Fatigue Crack Growth in Rails

ABSTRACT In rolling contact fatigue of rails, cracks form at a shallow angle, and grow primarily in the direction of travel underneath the rail surface. Once nucleated, cracks may branch to a Mode I direction, whereas shallow cracks grow when the crack is filled with fluid. This paper attempts to model the response of shallow-angle cracks through fracture mechanics analysis, using finite element stress analysis and multiaxial fatigue tests to simulate the rolling contact history. The evolution of the history of mixed- Mode I and II stress intensity factors is derived from a 2D finite element model. Biaxial fatigue crack propagation experiments have been conducted on BS11 rail steel to investigate the effect of sequential mixed-mode loading on angled cracks. The effects of fluid in the crack, residual stress in the rail head and braking are considered to demonstrate that “squat” rolling contact fatigue cracks can only develop under prescribed loading conditions.

Rail vehicle wheel wear prediction: a comparison between analytical and experimental approaches

There are a number of theoretical and practical techniques to compute rail vehicle wheel wear. For instance, the Archard equation is a well-known tool to determine the worn volume in sliding contact, as a function of normal load, sliding distance and the surface hardness. Of course, the wear coefficient (called K) used in this equation to differentiate the wear models implicitly comprises the conditions that govern the contact surface. Two situations can be taken into account when considering a sliding contact in a rail vehicle wheels, particularly along a curved track: (i) when the radial force prevails the lateral tangential force, which is mainly the frictional force but before flanging and (ii) during flange contact. Also, the Archard equation is employed within the tread and flange regions separately, both the regions being of interest in this paper. A number of approaches are then used to find the distance slid. The authors compare the field test results and the outcome of the analytical approaches. When the wheel wear results acquired from the two test bogies on Iranian Railways, all technical (rigid frame bogies with new assemblies and components) and operational items were identical, except for changing the bogie orientation in the second test trial for a short period. Good agreement was found between the analytical and practical investigations.

Application of Biaxial Plasticity and Damage Modelling to the Life Prediction and Testing of Automotive Components

ABSTRACT A method has been developed for making life predictions for engineering components subject to multiaxial loadings using the local strain approach. The method has been incorporated into a computer program, which uses a Mroz-Garud (1,2) cyclic plasticity model and the Wang-Brown cycle counting and damage models (3,4) . The plasticity, rainflow cycle counting and damage models have been generalised to deal with any free-surface loading conditions. The program makes calculations based on strain gauge rosette measurements and its application is illustrated by calculations from three typical automotive components. Some interesting methods for visualising the analysis results are explored. In addition to life prediction, the approach also has applications in accelerated durability testing.