Using Statistical Calibration for Model Verification and Validation, Diagnosis of Model Inadequacy, and Improving Simulation Accuracy
June 11 - 8:40 AM to 9:20 AM GMT+2 - Room: E-8
Presented by Dr. Mark Andrew, SmartUQ Technology Steward
The growing use of simulations in the engineering design process promises to reduce the need for extensive physical testing, decreasing both development time and cost. Statistical calibration optimizes tuning of model parameters to improve simulation accuracy, and estimates any remaining discrepancy which is useful for model diagnosis and validation. Also, because model discrepancy is assumed to exist in this framework, it enables robust calibration even for inaccurate models. The presentation will show an application of statistical calibration to a bracket design fatigue model simulated in COMSOL. The calibration will tune the bracket’s material properties and fatigue characteristics. For illustrative purposes, the physical test data will be generated using the same simulation routine, but with an intentionally applied bias and random noise to replicate physical testing errors. The accuracy and conclusions from the statistically calibrated model will be compared with the uncalibrated model as well as a model calibrated with conventional error minimization methods. The metrics used in the comparison will include optimization, sensitivity analysis, and propagation of uncertainties motivated by sources like manufacturing variations during bracket fabrication. The results will illustrate the importance of calibrating a model before drawing design conclusions. Multiple metrics will be shown which can be used for model validation, including a discrepancy map which characterizes inadequacies in the simulation.
