1408 Multidisciplinary Design Optimization under Uncertainty

Tae Hee Lee, Hanyang University
Nozomu Kogiso, Osaka Prefecture University
Po Ting Lin, National Twain University of Science and Technology
In engineering design, the traditional deterministic design optimization which considers the variables as only deterministic values has been successfully applied to reduce the cost while satisfying all of design requirements. As a system becomes complex, multi-disciplines are necessarily analyzed to estimate the responses of the complex system. Multidisciplinary design optimization (MDO) technologies for a complex system have been well developed in many areas of engineering since MDO was issued in 1990s from aerospace engineering. MDO can promise a synthetic optimum solution of the system while satisfying complicated design constraints and considering potential synergistic effects of the each discipline.
However, many design variables and parameters cannot be considered as a deterministic value but probabilistic one under the actual design environment. Therefore, probabilistic behaviors of responses are payed attention to count on the probability that satisfies a requirement of response, the so-called probabilistic constraint of design. To assure the reliability of the probabilistic constraints, reliability based design optimization (RBDO) has been developed for last decades. However, RBDO is applied to limited cases of practical design because of complexity of computation. To emerge the difficulties, recent developments in multidisciplinary design optimization under uncertainty will be highlighted in the World Congress of Computational Mechanics.
This MS covers, but not limited to, the following topics
(1) Uncertainty quantification
(2) Extreme environments
(3) Reliability analysis
(4) Robust design optimization
(5) Reliability based design optimization
(6) Multidisciplinary design optimization