Evaluation of liquefaction potential of soil using genetic programming

Abstract

Out of the various seismic hazards, soil liquefaction is a major cause of both loss of life and damage to infrastructures and lifeline systems. Soil liquefaction phenomena have been noticed in many historical earthquakes after first large scale observations of damage caused by liquefaction in the 1964 Niigata, Japan and 1964 Alaska, USA, earthquakes. Due to difficulty in obtaining high quality undisturbed samples and cost involved therein, in-situ tests, standard penetration test (SPT) and cone penetration test (CPT), are being preferred by geotechnical engineers for liquefaction potential evaluation with limited use of other in-situ tests like shear wave velocity tests and Baker penetration tests. The liquefaction evaluation in the deterministic framework is preferred by the geotechnical engineering professionals because of its simple mathematical approach with minimum requirement of data, time and effort. However, for important life line structures, there is a need of probabilistic and reliability methods for taking risk based design decisions. In recent years, soft computing techniques such as artificial neural network (ANN), support vector machine (SVM) and relevance vector machine (RVM) have been successfully implemented for evaluation liquefaction potential with better accuracy compared to available statistical methods. In the recent past, evolutionary soft computing technique genetic programming (GP) based on Darwinian theory of natural selection is being used as an alternate soft computing technique.The objective of the present research is to develop deterministic, probabilistic and reliability-based models to evaluate the liquefaction potential of soil using multi-gene genetic programming (MGGP) based on post liquefaction SPT and CPT database. Here, the liquefaction potential is evaluated and expressed in terms of liquefaction field performance indicator, referred as a liquefaction index (LI) and factor of safety against the occurrence of liquefaction (Fs). Further, the developed LIp models have been used to develop both SPT and CPT-based CRR models. These developed CRR models in conjunction with the widely used CSR7.5 model, form the proposed MGGP-based deterministic methods. The efficiency of both the developed SPT and CPT-based iv deterministic models has been compared with that of available statistical and ANN-based models on the basis of independent database