A fast model for the optimal composition sensitivity based on regression analysis of superalloy properties calculated with CALPHAD approach.
Alexander Müller, Friedrich-Alexander Universität Erlangen-Nürnberg, Erlangen, Germany
Alloy-by-design is usually dealt with MOOP (Multi Objective Optimization Problem) resulting in a Pareto-front with a set of optimal compositions. But, scattering of element concentrations during the alloy production process causes property deviations to an optimal alloy composition. The sensitivity of such compositions should be taken into account during the optimization process, to be able to determine robust and optimal alloys. Established models for sensitivity require a high computational effort, due to a high number of necessary calculations. So we derive a fast model for such sensitivity, based on a regression analysis of mean variations of given properties. This approach allows the prediction of sensitivity values and the violation fraction for given constrains with one calculation only.
Session T1: Tuesday, 26 June 2018
End: 10:45 a.m.