Multidisciplinary Design Optimization for High Reliability and Robustness

摘要

Optimization and uncertainty analysis considering multiple design criteria involves seamless integration of often conflicting disciplines. Over the last 3 years Wright State University has been applying analysis tools to predict the behavior of critical disciplines to produce highly robust torpedo designs using robust multi-disciplinary design optimization. Multidisciplinary optimization and system reliability analysis were emphasized during the third year of the grant period and more reliable torpedo designs were obtained for all the different disciplines. An optimal configuration of a supercavitating torpedo model that fits in a cavity generated by the cavitator was obtained. An evidence theory based method to determine the reliability of the cavitator is presented. Structural optimization of the lightweight torpedo model was done for it to be safe from underwater explosions and to reduce the acoustic signature. Also, reliability based optimization was performed for the lightweight torpedo model using system reliability constraints.