TY - CONF A1 - Blasón, Sergio A1 - Madia, Mauro T1 - Determination of the fatigue limit and initial crack length by means of fracture mechanics N2 - The fatigue design of metallic components is usually based on two different approaches, namely total life and damage tolerance. The former takes into account both the initiation and propagation stages. The latter is based on fracture mechanics and needs an initial defect, which propagates until the critical size under cyclic loading, provided it is large enough. In this case, a residual life rather than a total life is obtained. The overall lifetime of a cyclically loaded structure involves four consecutive stages: a) crack initiation; b) propagation of microstructurally short cracks; c) propagation of physically/mechanically short cracks; d) propagation of long cracks. Considering the propagation from the short crack regime, the damage tolerance approach can be extended to adequately calculate the total life of a component. The fatigue strength of metallic materials containing defects depends on the non-propagation condition of small cracks emanating from these defects. In this work, the presence of mechanically short cracks from the beginning of the component's life is considered. Consequently, the methodology of analysis must be able to treat adequately local ligament yielding effects typical for short cracks and must include the thorough description of the crack closure effect up to the long-crack regime. In the analysis based on long cracks, the linear-elastic condition is mostly satisfied which allows the use of the linear elastic parameter ΔK for describing the crack driving force. On the contrary, this assumption is not adequate for mechanically short cracks because the crack depth is in the order of the plastic zone. Instead, an elastic-plastic driving force should be considered. Furthermore, the gradual build-up of the plasticity-induced crack closure effect must be considered, which implies a transition from the intrinsic (effective) threshold value, ΔKth,eff, to the long-crack threshold, ΔKth,LC. The effective component is a material parameter which is dependent on the elastic properties and crystal lattice. The gradual build-up of crack-closure is described by ΔKth,op which is a function of the plastic properties, grain-size, environment conditions, load ratio and crack-depth. The crack closure can be characterized experimentally by the so-called cyclic R-curve. Some other crack closure effects, such as roughness or oxide-debris induced, might be incorporated as well. The knowledge of the cyclic R-curve can be useful to determine the largest non-propagating crack size at the material fatigue limit. This is realised by means of the so-called cyclic R-curve analysis: The tangent criterion between the driving force and the cyclic R-curve define the transition between crack arrest and propagation. In this regard, a0 is defined as that crack depth which will only grow into a non-propagating crack due to the development of crack closure. The procedure outlined briefly here has been successfully applied to the determination of the fatigue limits of steel weldments. Considering the stochastic distribution of the weld geometric parameters (namely, weld toe radius, the flank angle, and the excess weld metal) it is possible to perform a full probabilistic cyclic R-curve analysis and determine this way the statistical distribution of the initial crack size at the weld toe at the fatigue limit. T2 - Virtual Conference on Mechanical Fatigue CY - Online meeting DA - 09.09.2020 KW - Cyclic R-curve KW - Fracture Mechanics KW - Fatigue Strength KW - Short Crack Propagation PY - 2020 AN - OPUS4-51238 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - JOUR A1 - Schubnell, J. A1 - Konidena, S. K. A1 - Jung, M. A1 - Braun, M. A1 - Ehlers, S. A1 - Madia, Mauro A1 - Kannengießer, Thomas A1 - Löschner, D. T1 - Approach for the probabilistic fatigue assessment of welded joints based on the local geometry of the weld seam N2 - Welded joints show large variation of the weld toe geometry along the weld seam, which is one important reason for the comparably large scatter in fatigue life. Therefore, it is crucial to take the local geometry at the weld toe into account, to reduce the conservatism in fatigue assessment of welded joints. This study is based on the IBESS procedure for the calculation of the fatigue strength, whereby the evaluation of local geometrical parameters is carried out by means of 3D surface scans. The approach is validated against 26 fatigue test series. The fatigue life is in general overpredicted, whereas good agreement is achieved for high stress ratio (R = 0.5). A sensitivity analysis conducted with IBESS shows that weld toe radii ρ < 2 mm and flank angle α < 30° have a significant influence on the calculated fatigue strength. In contrast to this, no strong correlation between ρ and the fatigue strength was determined experimentally in this study. KW - 3D Scanning KW - Fatigue Strength KW - Fracture Mechanics KW - IBESS Approach KW - Local Weld Geometry KW - Welded Joints PY - 2023 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:b43-585700 SN - 8756-758X SP - 1 EP - 20 PB - Wiley AN - OPUS4-58570 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Weichert, D. A1 - Kister, Alexander A1 - Houben, S. A1 - Ernis, G. A1 - Wrobel, S. T1 - Robustness in Fatigue Strength Estimation N2 - Fatigue strength estimation is a costly manual material characterization process in which state-of-the-art approaches follow a standardized experiment and analysis procedure. In this paper, we examine a modular, Machine Learning-based approach for fatigue strength estimation that is likely to reduce the number of experiments and, thus, the overall experimental costs. Despite its high potential, deployment of a new approach in a real-life lab requires more than the theoretical definition and simulation. Therefore, we study the robustness of the approach against misspecification of the prior and discretization of the specified loads. We identify its applicability and its advantageous behavior over the state-of-the-art methods, potentially reducing the number of costly experiment. T2 - 2nd Annual AAAI Workshop on AI to Accelerate Science and Engineering CY - Washington DC, USA DA - 07.02.2023 KW - Active Learning KW - Gaussian Process KW - Fatigue Strength PY - 2022 AN - OPUS4-56684 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -