Reliability-Based Design Optimization Software Package for Broader Simulation-Based Design Applications

The Reliability Analysis and Multidisciplinary Design Optimization (RAMDO) software developed during the current SBIR Phase II (A14-075) provides very accurate results given: (1) accurate input distribution models obtained from large numbers of input test data; and (2) accurate simulation models. However, in industrial and defense applications, often only limited numbers of input test data are available. The challenge is that insufficient input test data induce uncertainties in input distribution models and consequently reliability assessment. In addition, the simulation model needs to be validated by comparing the simulation model output distribution against the target output distribution obtained using a large number of output test data, which is prohibitively expensive. Thus, practically only a limited number of output test data are used, which induces uncertainty in the target output distribution. In the proposed SBIR Subsequent Phase II, computational methods and processes will be developed to obtain a validated simulation model for the confidence-based uncertainty quantification (UQ) and reliability assessment using limited number of input and output test data; and will be integrated with the existing RAMDO. As in the SBIR Phase II, for broader multidisciplinary applications, RAMDO will be integrated with selected commercially available Process Integration (PI) softwares and CAE solvers.