Support vector machines in mechanical properties prediction of jet grouting columns

Abstract

Strength and stiffness are the mechanical properties currently used in geotechnical works design, namely in jet grouting (JG) treatments. However, when working with this soil improvement technology, due to its inherent geological complexity and high number of variables involved, such design is a hard, perhaps very hard task. To help in such task, support vector machine (SVM), which is a data mining algorithm especially adequate to explore high number of complex data, can be used to learn the complex relationship between mechanical properties of JG samples extracted from real JG columns (JGS) and its contributing factors. In the present paper, the high capabilities of SVM in Uniaxial Compressive Strength (UCS) and Elastic Young Modulus estimation of JG laboratory formulations are summarized. After that, the performance reached by the same algorithm in the study of JGS are presented and discussed. It is shown, by performing a detailed sensitivity analysis, that the relation between mixture porosity and the volumetric content of cement, as well as the JG system are the key variables in UCS prediction of JGS. Furthermore, it is underlined the exponential effect of the age of the mixture in UCS estimation as well as the high iteration between these two key variables.