Risk and Uncertainty Management for Multidisciplinary System Design and Optimization

Computational methods and simulation based methodologies can be combined for optimizing the design of undersea weapons with respect to cost and performance. Multiple engineering disciplines such as: warhead performance, structural-acoustics, sonar performance, guidance and control, electromagnetic and thermal analyses of the electric motor, computational fluid dynamics, and structural dynamics must be considered during undersea weapon design. Therefore an organized multi-disciplinary optimization (MDO) strategy must be employed for the design. Due to the large number of iterations required during a MDO process and in an effort to include high fidelity simulations in a MDO process, theories for developing metamodels have been established. In any optimal design the final system operates close to several of the constraints since the optimal point is driven against the active constraints by the optimization strategy. Deterministic optimization methods do not account for variability introduced due to manufacturing, material properties, and equipment performance, nor for uncertainties in the mission profile, human operators, the environment, and the simulation models themselves. The proposing firm (MES) will develop a network of probabilistic metamodels and incorporate them in a multi-disciplinary design optimization environment. The new developments will be demonstrated through a case study relevant to undersea weapon design. Existing MES products in MDO under uncertainty, and in time dependent metamodeling, will comprise the foundation of the proposed research, thus maximizing the impact of the sponsor’s funding.