A methodology to consider local material properties in structural optimization

摘要

We propose a finite element methodology to consider local material properties for large cast iron components in shape optimization. We found that considering local strength instead of uniform strength within shape optimization brings to different results in terms of safety-cost balance for the same component. It is well known that local mechanical properties of large cast iron components are defined by their microstructure and defects, which locally affect the strength of the components. Considering or not local mechanical properties can dramatically change a component reliability evaluation during its design. Since a typical industrial aim for shape optimization is trying to get the optimal solution in terms of component quality and cost, considering local material properties is even more important than in traditional design process where no optimization techniques are used. We compute solidification process parameters via finite element solidification analysis, and then we exploit experimental correlation between these parameters and ultimate tensile strength to evaluate the local reliability of the finished component under its static loading conditions. We believe that this methodology represents an opportunity to better design casting components when mechanical properties are deeply affected by their production process as described in the provided examples. In these examples, we wanted to minimize casting cost constrained by a target reliability and we get component cost reduction by considering local material properties. Future research will address the problem of using dedicated casting simulation software instead of general purpose finite element analysis software to compute solidification analysis and then introducing fatigue analysis and correlation between fatigue material properties and casting process output variables.