ABSTRACT
We consider variants of Successive Quadratic Programming (SQP) applied to Simultaneous Analysis and Design (SAND) implementations for PDE-based engineering modeling systems. In this study we consider in particular large scale P D E solvers for fluid flow, reactions and interactions with solid surfaces. We sketch the SQP algorithm and options that need to be compatible with the information obtained from the modeling system. Here we consider four levels of implementation and highlight each level with challenging PDE-based optimization problems. These implementation levels also need to be supported with object oriented software platforms that take advantage of advanced computing environments and lead to flexible modifications in solution and optimization algorithms. The benefits of this approach wi l l be demonstrated through numerical results on challenging optimization problems with P D E models.
