Enhanced Monte Carlo for Reliability-Based Design and Calibration

Abstract eng:
This paper extends the recently developed enhanced Monte Carlo approach to the problem of reliability-based design. The objective is to optimize a design parameter(s) so that the system, represented by a set of failure modes or limit states, achieves a target reliability. In a large majority of design and/or calibration contexts, the design parameter α itself can be used to parameterize the system safety margin M (α). The lower tail of this random variable behaves in a regular way and is therefore amenable to straightforward parametric analysis. In contrast to the original Naess et al. method [1], the intention is to estimate the value αT that corresponds to a (very) small target system failure probability pf T . Monte Carlo sampling occurs at a range of values for α that result in larger failure probabilities, and so the design problem essentially amounts to a statistical estimation of a high quantile. Bounds for αT can easily be constructed. Several examples of the approach are given in the paper.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket