Development of a Hierarchical Data Format for Modeling, Simulation and Postprocessing in Structural Mechanics and its Ecosystem

In order to develop ecologically and economically sustainable products, especially in the aerospace industry, it is necessary to reduce conservatisms in order to be able to exploit further lightweight potentials. One way to achieve this is to use more realistic representations of components in virtual testing and challenge knock-down factors. For this purpose, all significant influences on the component behaviour must be determined simulatively before it is tested virtually. Additionally, modern product development requires shorter development times and thus more agile approaches. At the German Aerospace Center (DLR e.V.) Virtual Product House (VPH) a virtual end-to-end-process is developed that couples design elements, sizing, virtual manufacturing and simulation-based testing of components. This leads to a significant risk mitigation of virtual product development processes by providing more realistic, simulative test results earlier in the product development process. For the design of complex structures, their manufacturing processes and testing procedures it is neither feasible nor possible to use a single solver for this approach. Therefore, the development process poses great challenges to the consistency of numerical models and simulation results between assessment steps. In the worst case, each individual simulation step takes place in a different solver with the respective proprietary format and may use different model idealizations. Depending on the number of tool interfaces it is beneficial to store the necessary information in a central solver-agnostic format and provide interfaces to the required solver formats. The presentation gives an overview about the motivation, implementation and eco-system around a novel approach for the complete model and simulation result description of computational structural mechanics based information, e.g. for the finite element method. A programming-language independent representation is developed using machine-readable formats for the model description as well as the VMAP standard for the numerically heavy data storage [7, 8]. In the presentation the automated hierarchy development and programming-language independent representation are described and shown on examples. It is shown how model generators use these language elements to provide models as well as post-processing capabilities that leverage the approach and reference implementation.