Abstract

In automotive development the Finite Element Method (FEM) is integrated throughout the virtual product development processes, industrialized, and immanently integrated in the series development. This broad use is supported by simulation process and data management systems (SPDM), which allows extensive automation in modeling, simulation and evaluation of simulation results (post-processing). Regarding post-processing, according to the current state of the art, fixed evaluation scripts are usually executed, and the resulting result values are stored in the databases and reports of the SDMS. Despite the increasing complexity of the simulation models and pure number of daily executed simulations, an appropriate in-depth analysis of individual simulations takes place less and less often and requires the dedicated extraction of the simulations from the SDMS and manual evaluation by experts. Such experts usually require years of training and experience and are therefore increasingly difficult to access. In addition, daily project live seldomly allows time for the necessary depths of the analysis, further resulting in a decrease in expertise and results quality and validity. Hence the development and implementation of suitable expert systems as a special form of Artificial Intelligence promises to be a solution to resolve the discrepancy between analysis experts availability and the immense amount of simulation result being processed in simulation data management. The authors have developed and implemented such expert systems for an AI based, adaptive and “intelligent”, automated evaluation of simulation data results, establishing a new level of simulation data analysis within SPDM. The proposed presentation/paper aims to introduce and illustrate basic concepts, workflows, solution architectures and implementations of automated in-depth, result-dependent, AI-based analyses of Finite Element simulations within SPDM and successful use cases with validated added values, like • automated anomalies detection for result’s plausibilization as simulation data quality management instrument, • automated identification of causal chains in complex structural responses and damage, • decision support system and extended analysis capabilities within in-depth results analysis, • “smart” access to the data in the SPDM databases as fundament sophisticated data mining across different simulations. Certain examples of such implementations will be shown for illustration.