This award supports Professor Billie Spencer of Notre Dame University and two faculty members and a graduate student from the University of Illinois at Urbana-Champaign to collaborate in computational engineering with Professors G.I. Schueller and H.J. Pradlwarter of the Institute for Mechanics of the University of Innsbruck, Austria. Their goal is to improve the calculation of failure probabilities of certain categories of dynamical systems. They propose to approach this by developing the concept of continuously evolving sample distributions based upon satisfaction of certain `fitness` criteria. The key feature of their approach is investigating the use of Monte Carlo simulation methods, and enhancing them through `genetic algorithms.` This approach shows great promise for enhanced computational efficiency and applications to larger and more complex systems. The research effort will benefit from the complementary expertise of the two groups. The US group will focus on development of the genetic algorithm and parallelization of the Monte Carlo procedures, areas in which the US investigators are well qualified. The Austrian group will concentrate their efforts on the development of advanced Monte Carlo procedures, including the determination of proper fitness functions.