Bayesian Approach to Prototype Testing

by P. Hauge Madsen, Research Assoc.; Risoe National Lab., 4000 Roskilde, Denmark,
Niels C. Lind, Prof. of Civ. Engrg.; Univ. of Waterloo, Ontario, Canada, N2L 3G1,

Serial Information: Journal of the Structural Division, 1982, Vol. 108, Issue 4, Pg. 753-770

Document Type: Journal Paper


The strength of series-produced structures is uncertain because of modelling error and material parameter uncertainty. By testing of prototypes, information on both uncertainties is obtained. Because of strength correlation in a series, sampling is considered non-random. The Bayesian Method is used to consider the problem of relating prior knowledge and test data to a prescribed characteristic value of the load carrying capacity, i.e., a prescribed fractile of the strength distribution, or a target value of a reliability index. Examples show application to prototype testing structural assemblies or parts. It is concluded that even a slight correlation may have a significant influence on the estimated strength.

Subject Headings: Bayesian analysis | Structural strength | Strength of materials | Errors (statistics) | Uncertainty principles | Correlation | Load tests | Load bearing capacity

Services: Buy this book/Buy this article


Return to search