Robustness evaluation for forming simulations
Forming processes are naturally affected by scatter which influences significantly the forming results. Keeping scatter within tolerable limits is a decisive quality criterion for a robust forming process. At BMW AG, optiSLang and SoS is used to conduct stochastic analysis of simulation results in order to achieve a sufficient prognosis. For this purpose, robustness evaluations and sensitivity analyses are applied in virtual product design as well as during the production process.
Especially by the projection of statistical measures onto FE-structures, engineers can evaluate the "hot spots" of scattering result variables in their routine environment. Using a linear analysis, correlations between input and resulting scatter can be determined. In addition, further correlations on the local level (element level) from quadratic correlation hypothesis as well as of result variable scatter can be obtained from histograms.
Will, J.; Bucher, C.; Ganser, M.; Grossenbacher, K.:
Computation and visualization of statistical measures on FE structures for forming simulations
Proceedings Optimization and Stochastic Days 2.0, Weimar, Germany, 2005