Underground storage tank blowout analysis: Stability prediction using an artificial neural network

Abstract

Most geotechnical stability research is linked to “active” failures, in which soil instability occurs due to soil self-weight and external surcharge applications. In contrast, research on passive failure is not common, as it is predominately caused by external loads that act against the soil self-weight. An earlier active trapdoor stability investigation using the Terzaghi's three stability factor approach was shown to be a feasible method for evaluating cohesive-frictional soil stability. Therefore, this technical note aims to expand “active” trapdoor research to assess drained circular trapdoor passive stability (blowout condition) in cohesive-frictional soil under axisymmetric conditions. Using numerical finite element limit analysis (FELA) simulations, soil cohesion, surcharge, and soil unit weight effects are considered using three stability factors (Fc, Fs, and Fγ), which are all associated with the cover-depth ratio and soil internal friction angle. Both upper-bound (UB) and lower-bound (LB) results are presented in design charts and tables, and the large dataset is further studied using an artificial neural network (ANN) as a predictive model to produce accurate design equations. The proposed passive trapdoor problem under axisymmetric conditions is significant when considering soil blowout stability owing to faulty underground storage tanks or pipelines with high internal pressures.