The Interplay Between Science and Statistics in Modeling Stress–Strain Curve Data

ABSTRACT

Data obtained from an archived 9-year aging study on S5370 foam were used to develop a stress–strain aging model using a strategy that first fits separate stress–strain curve data based on a form from polymer science. Then, exploratory data analysis of the model coefficients from these separate fits was used to develop a full model. This article illustrates this strategy and demonstrates how the full model can be fit and subsequent predictions can be made.

KEYWORDS:

• Bayesian analysis
• exploratory data analysis
• prediction

ACKNOWLEDGMENTS

We thank C. C. Essix for her support and encouragement of this work. This work was funded by the National Nuclear Security Administration through the Enhanced Surveillance Campaign at Los Alamos National Laboratory. We thank an anonymous referee for helpful comments that improved the exposition of this article.

Additional information

Notes on contributors

K. Sham Bhat

K. Sham Bhat is a Scientist and received his Ph.D. in Statistics from Penn State University in 2010. His research interests include analysis of computer models, Bayesian statistics, space-time data, and applications to energy, climate, and defense.

Stephen A. Birdsell

Stephen A. Birdsell is an R&D Engineer and holds an M.S. In Chemical Engineering from the University of Colorado, Boulder. His research interests are in understanding the effects of aging on the performance of polymer materials.

Michael S. Hamada

Michael S. Hamada is a Scientist and holds a Ph.D. in Statistics from the University of Wisconsin-Madison. He is a Fellow of the American Statistical Association. His research interests include design and analysis of experiments, measurement system assessment, quality control, and reliability.