Reliability-Based Robust Design of Raft Foundation and Effect of Spatial Variability

Abstract

Geotechnical design considers different limit states of design. Conventional deterministic design methodology uses the factor of safety approach, while reliability-based design and robust design methodology use probability distribution and statistics of properties. In this study, different design methodologies were compared. Efficient probabilistic methods and optimization techniques such as the first-order reliability method and genetic algorithm were used to arrive at a robust and optimized design. Estimation of system reliability helps in attributing importance to both the limit states involved, simultaneously. In addition, a combination of random field theory with numerical modeling in the framework of the Monte Carlo simulation method and response surface method was used in considering spatial variability. In this way, the risk involved while using conventional deterministic design methodology is brought out; it is apparent from the larger foundation sizes obtained using reliability-based and robust design methodologies. This emphasizes the importance of considering the variability of soil and the robustness of design for critical structures using probabilistic methods.