Automatic Generation of Robustness Knowledge for Selected Crash Structures
In the design process of crash-loaded structures, one has to know the structural behavior under typical uncertainties, like the scatter of material data or changes in the load direction. In conclusion it is necessary to find crash structures which are robust against these uncertainties. The first questions is, how to find the major robustness relevant information? The second question is, what are the best design canges, in order to make the crash structure more robust? Usually, crash engineers need a lot of time to answer these questions. In this contribution we present a method and a corresponding software implementation, which extracts robustness knowledge mostly automatically from a design of experiments study. The software interprets this large amount of data by projection of the results onto a simplified geometry. These simplified results can be saved and compared very efficiently. As a result we not only get information about unintentional buckling behaviour, but also a feeling for the rates of occurrence and possible causes. Besides a generic rail, different additional structures are examined. For all these examples the automatic interpretation of the robustness behavior will be shown.
https://www.dynamore.de/de/download/papers/2016-ls-dyna-forum/Papers%202016/dienstag-11.10.16/optimazation-and-robustness/automatic-generation-of-robustness-knowledge-for-selected-crash-structures/view
https://www.dynamore.de/@@site-logo/DYNAmore_Logo_Ansys.svg
Automatic Generation of Robustness Knowledge for Selected Crash Structures
In the design process of crash-loaded structures, one has to know the structural behavior under typical uncertainties, like the scatter of material data or changes in the load direction. In conclusion it is necessary to find crash structures which are robust against these uncertainties. The first questions is, how to find the major robustness relevant information? The second question is, what are the best design canges, in order to make the crash structure more robust? Usually, crash engineers need a lot of time to answer these questions. In this contribution we present a method and a corresponding software implementation, which extracts robustness knowledge mostly automatically from a design of experiments study. The software interprets this large amount of data by projection of the results onto a simplified geometry. These simplified results can be saved and compared very efficiently. As a result we not only get information about unintentional buckling behaviour, but also a feeling for the rates of occurrence and possible causes. Besides a generic rail, different additional structures are examined. For all these examples the automatic interpretation of the robustness behavior will be shown.