Search

Abstract This second part of the review on defects as root cause of fatigue failure comprises the origin, the nature and the effects of non-metallic inclusions. Topics addressed are the different kinds of inclusions formed during the manufacturing process, various types of mis-match causing local stresses and, as a consequence, fatigue crack initiation, and effects of characteristics such as size, morphology, localization, spatial distribution and orientation of the defects on the fatigue behavior. Methods for inclusion counting and sizing are discussed along with statistical aspects necessary to be considered when evaluating structural components. Highlights • Part of a comprehensive review on material and geometrical defects as root cause in failure analysis. • Discussion of the background as well as of practical effects of non-metallic inclusions in fatigue. • Discussion of the various parameters which affect the influence of inclusions on short and long crack propagation. [ABSTRACT FROM AUTHOR]

Abstract This third part of the review on defects as root cause of fatigue failure addresses cavities (pores, micro-shrinkages, unmelted regions), defective microstructures and microcracks as material defects and defects due to local damage during manufacturing, service and maintenance such as dents, scratches and localized corrosion. In addition, damage due to contact fatigue and the effect of surface roughness are discussed in the context of fatigue failure. Also addressed is the competition between different kinds of defects in controlling the initiation and early growth of fatigue cracks. Highlights • Part of a comprehensive review on material and geometrical defects as root cause in failure analysis • Discussion of defects such as porosity, dents, mechanical impacts, scratches corrosion pits, etc. • Discussion of the competitive situation of different kinds of defects with respect to the effect on fatigue [ABSTRACT FROM AUTHOR]

Abstract According to the definition of the ASM handbook [ 1,3 ], a defect is "an imperfection ... that can be shown to cause failure by a quantitative analysis and that would not have occurred in the absence of the imperfection". The topic of the present three-part review is a discussion of defects which can cause failure in cyclically loaded structures. The features discussed comprise material defects such as non-metallic inclusions, pores or micro-shrinkages, etc. and geometric defects such as surface roughness and secondary notches which have their origin in manufacturing, and defects such as surface damage due to scratches, impact events or contact fatigue as well as corrosion pits which arise in service. In this first part, the discussion is prefaced by an introduction to basic aspects which are essential for a deeper understanding of the characteristics and mechanisms how the defects influence fatigue crack initiation and propagation. These include the life cycle of a fatigue crack from initiation up to fracture, crack arrest, multiple crack initiation and coalescence, and the material and geometrical properties affecting these. Highlights • Part of a comprehensive review on material and geometrical defects as root cause in failure analysis • Discussion of basic aspects such as crack propagation, crack arrest, multiple cracks, cracks and notches, etc. • Fracture mechanics background to the discussion of the influence of defects [ABSTRACT FROM AUTHOR]

The work presents the procedure developed within the German research project IBESS , which allows for the fracture mechanics-based prediction of the fatigue strength of welded joints under constant amplitude loading. Based on the experimental observations of the crucial failure mechanisms, the approach focuses on the short crack propagation, where elastic-plastic fracture mechanics and the build-up of closure effects must be considered, as well as the variability of the local geometry at the weld toe and the modelling of multiple crack interaction. Analytical solutions are provided for the approximation of the through-thickness stress profiles at the weld toe and for the determination of the crack driving force in the form of a plasticity-corrected stress intensity factor range Δ K p . Proposals for the determination of the initial crack size and the crack closure factor are also included. The approach is validated against a large number of experimental data, which comprises fatigue tests on individual cracks monitored by heat-tinting and beach-marking techniques, as well as stress life curves. Three kinds of welded joints, two steels of significant different strength, two welding techniques and three stress ratios are considered. The results show that the procedure provides good estimations of the statistical distribution of the fatigue strength of welded joints both for the finite and infinite life regime. Furthermore, the predictions are compared with available benchmark data for structural steels. [ABSTRACT FROM AUTHOR]

Any fracture mechanics based determination of the fatigue strength of weldments requires different input information such as the local weld geometry and material data of the areas the crack is passing through during its propagation. The latter is so far not a trivial task as the fatigue crack is usually initiated at the weld toe at the transition from the weld metal to the heat affected zone. Furthermore, the crack propagates through the different microstructures of the weldment even into the base metal and causes final fracture. This paper describes how the material input information has been gained particularly for heat affected zone material by thermo-mechanically simulated material specimens for two steels of quite different static strength. The data comprise the cyclic stress-strain curve, the crack closure effect-corrected crack growth characteristics, fatigue threshold values for long cracks, the dependency of the parameter on the crack length and monotonic fracture resistance. The substantial experimental effort was necessary for the validation exercises of the IBESS approach, however, within the scope of practical application more easily applicable estimating methods are required. For that purpose, the paper provides a number of appropriate proposals in line with its check against the reference data from the elaborate analyses. [ABSTRACT FROM AUTHOR]

The paper provides a review on analytical flaw assessment methods with the focus on fracture under monotonic loading and fatigue crack propagation. The first topic comprises linear elastic as well as elastic-plastic fracture mechanics approaches. It essentially follows their historical development. Topics which are separately discussed are reference/limit loads, the treatment of secondary stresses, strength mismatch, constraint, unstable crack propagation (monotonic R-curve analyses) and statistical aspects. With respect to fatigue crack propagation the analytical treatment of crack closure and constraint and the determination of the cyclic elastic-plastic crack driving force is discussed. Finally, cyclic R-curve analyses are briefly addressed. [ABSTRACT FROM AUTHOR]

• Fail-safe design of a steering knuckle based on analytical flaw assessment. • The effect of stiffness and constraint is investigated. • Overconservative results are obtained when using substitute geometries. • A component-based numerical determination of the model parameters is proposed. • The procedure finds application in parameter variation analyses. Using the case study of a fail-safe design criterion for a steering knuckle of a commercial vehicle, the effect of stiffness/constraint on the crack driving force in the component is discussed. The problem arises when assessment procedures such as R6, BS 7910 or SINTAP/FITNET are applied in conjunction with substitute geometries for determining the model parameters K -factor and limit load (or a substitute for the latter), as it is common practice. In the example, the conservatism was so pronounced that this procedure was in fact unusable. A way out could be the finite element-based determination of the model parameters and their use in the analytical framework. This procedure is useful and effective if the finite element-based calculations are used in parameter sensitivity analyses. [ABSTRACT FROM AUTHOR]

Recently two of the authors of the present paper proposed a model for a fracture mechanics based prediction of the S – N characteristics of metallic components with large microstructural defects and supported this by a validation exercise on tensile plates made of an aluminium alloy AL5380 H321. Here the authors extend the study using a number of further data sets from the literature for which data were available at different R ratios. These data include two aluminium alloys, Al 2024-T3 and Al 7075-T6, and a ductile cast iron, EN-GJS-400-18-LT. Despite of necessary assumptions for the compensation of partially missing input information the results were fairly reasonable with the exception of one data set. The authors identify high applied stress levels in combination with potential multiple crack initiation as the probable root of the problem and propose a scheme how the model can be extended for taking into account crack initiation. [ABSTRACT FROM AUTHOR]

In microelectronic packaging, wire bonding is the predominant method for making electrical connections. Copper is increasingly substituting gold as interconnection material since it is a much cheaper alternative and it also offers several physical advantages. Adequate and reliable mechanical integrity of the connection is usually checked by process controls based onto “wire pull” and “ball bond shear” tests. In this paper the two methods are compared in terms of sensitiveness in detecting a latent weakness of the bond-pad structure, either induced by inappropriate wire bonding process or cumulated during reliability ageing. The failure modes (in terms of frequency and maximum test load) observed at the ball bond interface have been investigated on two different batches of a same chip, obtained from different wire-bonding recipes and including both unstressed and aged units. Cross-sections of the samples, submitted to pull and shear both in destructive and non-destructive tests, have allowed us to investigate the relationship between the bond morphological characteristics (metal deformation and potential micro-damages induced by copper bonding) and the weak points for fracture propagation inside the bond-pad inner layers and the silicon substrate. Besides the experimental activities, fracture mechanics and the finite element method have been employed to model the pull and shear tests. The aims of the finite element modelling have been to predict the reduction of test maximum load in defective ball bonds and the crack growth angle adopting a mixed-mode criterion. Good results have been obtained by the numerical fracture analysis, which can then support the reliability characterization and mechanical improvement of the bond. [ABSTRACT FROM AUTHOR]

The determination of the fatigue life in technical alloys containing large and small defects must rely on a propagation model which accounts for short and long crack growth. Recently an analytical model which incorporates propagation in the short crack regime and plastic correction for the crack driving force has been presented by two of the authors. This work is intended to show further validation of the model, taking into account data sets for different materials with different testing conditions. Despite the assumptions about missing parameters, the value of which had to be taken from the literature, the predictions showed a fairly good approximation of the fatigue lives. A possible interpretation of the results in terms of multiple crack initiation and propagation at higher loads is proposed. [ABSTRACT FROM AUTHOR]

Highlights: [•] We carried out tensile and small punch tests on two different lots of 1CrMoV steel. [•] One lot consisted of new material, the second one underwent 195,700 operating hours. [•] Differences in the small punch tests were not confirmed by tensile tests. [•] We carried out inverse analysis on small punch test to derive the mechanical properties. [•] We proposed a new method for the identification of the material parameters. [Copyright &y& Elsevier]

Abstract: An analytical fracture mechanics model for predicting the finite life fatigue strength of components is presented which combines a number of well established and newly developed approaches such as Murakami’s and McEvily’s approach for describing the transient behaviour of crack closure of short cracks, the analytical (long) crack closure function of Newman which became part of the widely used NASGRO approach, the R6 procedure, a method for improving the ligament yielding correction f(Lr ) of R6 proposed by the authors of the present paper and other elements. Basic assumption is the pre-existence of initial flaws such that the crack initiation or nucleation stage is small and can be neglected. The application of the model is demonstrated for small tension plates of aluminium Al 5380 H321 with artificial initial defects generated by FIB technology, the size of which was fixed on the basis of fractographic investigations on broken, smooth specimens. [Copyright &y& Elsevier]

Abstract: The present work focuses on the application of the inverse methods on the small punch test (SPT) in order to predict the behavior of turbine rotor steel upon in-service loading. A numerical framework has been implemented in which the small punch test has been simulated by means of finite element analyses and compared with the experimental results in order to assess the material parameters. The comparison has been carried out relying on the load-displacement curve of the SPT. The material behavior has been represented through an elastic-plastic constitutive law and a micro-mechanical damage model to account for softening and material failure. The assessed material parameters have been employed in the simulation of the tensile test, showing a good approximation of the basic mechanical properties of the material. [Copyright &y& Elsevier]

Abstract: The aim of this paper is a collection of stress intensity factor solutions for cracks in railway axle geometries which the authors of the present special issue developed and/or used for damage tolerance analyses. These solutions comprise closed form analytical as well as tabled geometry functions and they refer to solid as well as hollow axles and various crack sites such as the T- and V-notch and the axle body. [Copyright &y& Elsevier]

Abstract: Numerical methods are mostly used in the field of fatigue to derive the stress intensity factor (SIF) or J-integral solutions to be employed in damage tolerance analysis of cracked components. In this frame, simple assumptions about material properties are taken into account. More refined approaches try to describe the plasticity-induced crack closure in order to account for retardation effects under variable amplitude loading. In these approaches, the cyclic plasticity is used and cyclic finite element analyses are carried out. In the present work, a novel strategy is presented for the calculation of the relevant parameters to the fatigue crack growth, based on the evaluation of local field parameters (J-integral, T-stress) and cyclic material properties. It is demonstrated that, in case of mild steels and under the assumption of a stress ratio R =−1, the global constraint factor αg widely employed in fatigue crack growth algorithms such as the strip-yield model, can be calculated in a closed-form on the basis of the expression of the crack-tip fields. Moreover, αg provides a reasonable explanation of the fatigue crack growth behaviour of the A1N steel for different geometrical and loading configurations. Further investigations carried out on different medium and high strength steel grades show that the plastic radius ahead of small and long cracks at their fatigue limits can be considered as a constant for the material. [Copyright &y& Elsevier]

Abstract: The present paper summarizes the results of fatigue crack growth investigations on a hollow railway axle which were undertaken as a joint research project between the Politecnico di Milano, Italy, and the GKSS Research Centre Geesthacht, Germany, within an activity of ESIS-TC24 (Technical Committee on the Integrity of Railway Structures). Since no analytical stress intensity factor (SIF) solutions for semi-elliptical surface cracks in the S-transitions of railway axles were available, the first step in the research was to generate them for both reverse and rotary bending. The results were then used to investigate both the effects of rotary bending, and, of the press fits at the wheel and gear, on the fatigue crack growth and residual lifetime so as to provide essential input information for setting up inspection intervals. It was found that the effect of rotary bending, although existent, was rather moderate whereas the press fits had a large and detrimental effect. [Copyright &y& Elsevier]

[Display omitted] • Oxide-induced crack closure influences near-threshold crack propagation greatly. • Lower stress ratios and higher absolute air humidity enhance oxide formation. • Oxide formation and accumulation is larger in martensitic and bainitic steels. • Oxide-products formation does not depend on the testing procedure. • Optimal closure-free data are determined for stress ratio R ≥ 0.8. The experimental determination of the resistance to fatigue crack propagation in steel, particularly in the so-called near-threshold regime, is a crucial issue for safety-relevant components which are designed to operate for a large number of loading cycles before undergoing periodic inspections. This work summarizes the most relevant results of extensive experimental campaigns conducted at BAM and IPM over the last years, which have been devoted to the determination of fatigue crack propagation characteristics on medium and high strength steels. The influence of environment, oxide-induced crack closure and test methodologies in the near-threshold regime are discussed. [ABSTRACT FROM AUTHOR]

The paper provides an application of the IBESS approach to the investigation of the influence of various parameters of the global and local weld geometry as well as material defects on the fatigue strength of weldments. For this purpose, the global weld parameters, such as the weld toe radius, the flank angle, the excess weld metal, local secondary notches (in the present study as a measure of surface imperfections) and inclusions sizes have been determined as statistical distributions for different joint types and geometries and two steels of different strengths. The results are in line with literature data and reveal the potential of the theoretical approach to predict the correct trends. The combination with an advanced weld quality system has been demonstrated to be possible. [ABSTRACT FROM AUTHOR]

Abstract: Two of the authors of the present paper proposed a definition of a reference load F 0 which can be used as an alternative option to the common limit load for cases where the definition of the latter might be problematic in the frame of flaw assessment procedures such as R6, BS 7910, SINTAP or FITNET. The reference load is defined as that load at which the ligament yielding parameter L r approaches one. F 0, in general, varies along the crack front. It has to be obtained by finite element analyses but the results can be approximated by analytical expressions. In a previous study of the authors such a solution was presented for tension loaded plates with semi-elliptical surface cracks. In the present paper it is extended to pure bending and to combined bending and tension. [Copyright &y& Elsevier]

Abstract: The fatigue of mechanical components can be explained by the growth of very small fatigue cracks, which lead to the final fracture of the component. It is, therefore, essential to try to describe the growth and the thresholds of small cracks. In this paper, the Strip-Yield model is used to analyse the fatigue thresholds of a mild-steel widely used for manufacturing railway axles. Furthermore, the need for a precise definition of the constraint factor, under cyclic, non-linear response of the material, is addressed, together with the choice of an appropriate constraint formulation. This results in good fatigue threshold predictions at R =−1 for long cracks (obtained from SE(B) specimens), whereas the description of the Kitagawa diagram (obtained by fatigue tests on specimens with artificial defects) is relatively poor. The analysis of the cyclic stress ahead of small cracks enables us to compare the results with the threshold models for small cracks by McEvily, Usami and Shida. [Copyright &y& Elsevier]

The cyclic R-curve describes the resistance of a material to fatigue crack propagation from the short to the long crack regimes and it is therefore an essential ingredient in any fracture mechanics-based fatigue assessment procedure. This work presents different testing procedures employed in the experimental determination of the cyclic R-curve, especially focusing on the comparison with long fatigue crack propagation thresholds obtained by means of the compression precracking load reduction (CPLR) procedure. The tests were performed on the EA4T steel considering different stress ratios. The results show a good reproducibility of the cyclic R-curves at every stress ratio and for any testing procedure. In addition, the cyclic R-curves were used in a fracture mechanics-based assessment to predict the fatigue limits of specimens containing micro-notches. • The crack propagation threshold depends on the crack length. • A new experimental methodology to determine the R-curve is presented. • The R-curve is used to predict the endurance limits of micro-notched specimens. [ABSTRACT FROM AUTHOR]