Many structural engineering problems involve incremental/iterative solution procedures and require significant amount of computations. The proposed investigation concerns with developing efficient reanalysis strategies in advanced multiprocessing environments. This study involves application of advanced mathematical concepts including graph- theoretic techniques and multilevel algorithms for computational intensive engineering problems. The two application areas identified for this research are adaptive finite element analysis and optimization using substructures. The objective is to develop efficient parallel solution procedures for these two computational intensive problems. The algorithms developed will be important to the next generation engineering software, and the work will enrich the knowledge in structural engineering in advanced computing environments.
