ERROR AND UNCERTAINTY QUANTIFICATION AND SENSITIVITY ANALYSIS IN MECHANICS COMPUTATIONAL MODELS

摘要: Multiple sources of errors and uncertainty arise in mechanics computational models contribute to the final model prediction. This paper develops a systematic error quantification methodology for models. Some types are deterministic, some stochastic. Appropriate procedures developed either correct prediction deterministic or account stochastic through sampling. First, input error, discretization finite element analysis (FEA), surrogate output measurement considered. Next, which arises due use sampling-based methods, is also investigated. Model form estimated based on comparison corrected against physical observations after accounting solution approximation errors, experimental (input output). Both local global sensitivity measures investigated estimate rank contribution each source result. Two numerical examples used demonstrate proposed by considering mechanical stress fatigue crack growth analysis.