PSI - Issue 57

2 2

Laurent Dastugue et al. / Procedia Structural Integrity 57 (2024) 355–364 Michael Klein et. al./ Structural Integrity Procedia 00 (2019) 000 – 000 Michael Klein et. al./ Structural Integrity Procedia 00 (2019) 000 – 000

356

The full integration of the fatigue analysis into a general FEM solver solves the process issue (Fig. 1). Both parts of the analysis, the classic stress analysis, and the fatigue analysis, are based on a common data model and use the same resources. This ensures data efficiency (large data stays internal, double data handling is avoided) and the import/export of intermediate results is no longer necessary. Stress gradients, which were previously additionally calculated by the fatigue software based on its own model data, are now available in better quality directly within the FEM software based on the original FEM model. This can be used to increase the quality of the results. Due to the elimination of data exchange and the HPC orientation of the FEM software, the integrated fatigue analysis benefits significantly in terms of performance. New analysis classes are made possible by reduced amounts of data. Thanks to the integration, the complete stress results for all calculation steps no longer must to be saved, but are used directly ("on-the-fly") by the fatigue analysis in each calculation step and then immediately deleted again. This significant process improvement means that new classes of model sizes and significantly more result steps can be considered. The results are then much more accurate. The focus is on process simplification and reducing computing time for industrial applications. © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Fatigue Design 2023 organizers 1. Motivation The existing fatigue analysis process is typically done as two or more step simulation with several software. This process is time consuming, interface dependent, limited by disk space, includes manual steps, and is not maintained. A solution is the complete integration of the fatigue analysis into a general FEM solver. Then one common data model and resources are used, the process benefits from the high-performance infrastructure of the solver, the import/export or archiving of intermediate results is no longer necessary, and more accurate stress gradients can be calculated. 2. Typical classic process An industrial process, see reference Bernd et. al. (2022), that previously consisted of three steps (Fig. 2) serves as an example. The aim is to investigate the fatigue life of a structure which has initial assembly loads and many repeated load cases. This class of fatigue analysis is called load collective. The full integration of the fatigue analysis into a general FEM solver solves the process issue (Fig. 1). Both parts of the analysis, the classic stress analysis, and the fatigue analysis, are based on a common data model and use the same resources. This ensures data efficiency (large data stays internal, double data handling is avoided) and the import/export of intermediate results is no longer necessary. Stress gradients, which were previously additionally calculated by the fatigue software based on its own model data, are now available in better quality directly within the FEM software based on the original FEM model. This can be used to increase the quality of the results. Due to the elimination of data exchange and the HPC orientation of the FEM software, the integrated fatigue analysis benefits significantly in terms of performance. New analysis classes are made possible by reduced amounts of data. Thanks to the integration, the complete stress results for all calculation steps no longer must to be saved, but are used directly ("on-the-fly") by the fatigue analysis in each calculation step and then immediately deleted again. This significant process improvement means that new classes of model sizes and significantly more result steps can be considered. The results are then much more accurate. The focus is on process simplification and reducing computing time for industrial applications. © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Fatigue Design 2023 organizers Keywords: FEA; Fatigue Analysis; Process simplification; Process acceleration; Integration; Accurate Gradients; Data efficiency 1. Motivation The existing fatigue analysis process is typically done as two or more step simulation with several software. This process is time consuming, interface dependent, limited by disk space, includes manual steps, and is not maintained. A solution is the complete integration of the fatigue analysis into a general FEM solver. Then one common data model and resources are used, the process benefits from the high-performance infrastructure of the solver, the import/export or archiving of intermediate results is no longer necessary, and more accurate stress gradients can be calculated. 2. Typical classic process An industrial process, see reference Bernd et. al. (2022), that previously consisted of three steps (Fig. 2) serves as an example. The aim is to investigate the fatigue life of a structure which has initial assembly loads and many repeated load cases. This class of fatigue analysis is called load collective. © 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0 ) Peer-review under responsibility of the scientific committee of the Fatigue Design 2023 organizers Keywords: FEA; Fatigue Analysis; Process simplification; Process acceleration; Integration; Accurate Gradients; Data efficiency