Abstract
In this paper, several new models are developed based on a soft computing technique for predicting undrained shear strength (su) of marine clays at South Pars field, Persian Gulf, Iran. For this purpose, numerous piezocone penetration tests (PCPTs) data are used along with laboratory and in-situ measurements of su. In this regard, evolutionary polynomial regression (EPR) is used instead of conventional simple linear regression analysis in order to correlate su and cone tip resistance more precisely. According to five developed models (models I to V), it is found that the tip resistance corrected for pore water pressure effect (qt) yields slightly better estimation of su, compared to the raw cone tip resistance (qc) obtained directly from PCPT. Additionally, the cone tip resistance is correlated much more precisely with su data obtained from torvane test than that of unconsolidated undrained (UU) triaxial test. The EPR-based models offer a considerable improvement in estimating su based on cone penetration test data and can predict su with minimum error in comparison with other theoretical and empirical models. The models are also verified successfully by the data obtained from another survey area, outside the main survey area which has been used to develop the models.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call