Statistical Validation of Engineering and Scientific Models: Application to the Abnormal Environment; SAND2004-1029

x, 101, [1] pages. Tables, Figures. References. Appendix A: Model Parameters. Appendix B: Measurements, Predictions, and Sensitivities. CPUF/COYOTE model predictions were compared to experimental observations for the progression of a foam decomposition front due to heating from one side at 600 C, 750 C, and 900 C. We began with the characterization of the uncertainty for the model predictions due to uncertainty in model parameters. A first order sensitivity analysis and Monte Carlo simulation were used to estimate the structure of the uncertainty in the model predictions for the front motion due to parameter uncertainty. The resulting distributions were found to be normally distributed, correlated, with non-uniform variances. The metrics of Hills and Trucano (1999, 2001) were used to evaluate the consistency of the model predictions with the experimental observations for eight experiments. Mixed results were found depending on which experiment was considered. The model predictions for the 900 C experiment were found to be inconsistent with the experimental observations for the uncertainty models developed here. A locally linear analysis for the 600 C experiments indicated that the corresponding model predictions had very large uncertainty, and predictions and experimental observations were found to be consistent within this large uncertainty. Experimental results were analyzed for multiple orientations of the 750 C test cases. The results were mixed with the model tending to perform better for vertical orientations (heated from below or above) of the foam-in-can experiment.…