Geotechnical design often involves high, hard-to-control parameter uncertainties, which result in high variability in the system response. The variability in system response, which is typically addressed by satisfying a minimum safety measure in the form of a factor of safety or reliability index, forces the geotechnical designer to compromise between safety and efficiency (i.e., cost). When robustness of the geotechnical design against such system response variability is not evaluated during the design process, the tradeoff between over-design for safety and under-design for cost-savings is exacerbated. This paper introduces a novel design approach, Reliability-based Robust Design Optimization that considers explicitly the reliability, robustness, and cost. This design methodology is demonstrated with the design of a cantilever retaining wall. System reliability index is used as the performance measure and the tradeoff among the computed reliability index, the variance of the reliability index (as a measure of the robustness), and the cost are investigated. The results show that for some designs (with reliability index between 3 and 3.65), no tradeoff exists between the reliability index and its variance; hence, the design with the greatest reliability index also has the highest robustness (smallest variance) for a given cost. For other designs, a tradeoff relationship exists between the reliability index and its variance for a given cost.
|Number of pages||11|
|Journal||Geotechnical Special Publication|
|State||Published - 2013|