Mechanism of Compaction Grouting in Sand and Efficiency Prediction Equation Based on a DL-Extended Database

Abstract

Previous studies on compaction grouting focused more on clay but paid less attention to sand. In this paper, using the Nor-sand model, an finite difference method (FDM) simulation was carried out to study the grout efficiency and grout pressure of compaction grouting in sand. It is inferred that the deformation of sand during grouting consists of dilation, contraction, and compression. The grout efficiency depends mainly on the state parameter rather than the stress state of the sand, and the peak grout pressure is negatively related to the state parameter and positively related to the stress state. Furthermore, the effect of soil parameters of different sand types and grout volume on grout efficiency is discussed. It is found that the grout efficiency is obviously influenced by the critical state, yield, hardening, and dilatancy parameter of the Nor-sand model and grout volume. Finally, a simple dimensionless equation is proposed based on 36,750 cases obtained from simulations combined with a deep-learning (DL) extended database. The large amounts of cases enables the equation to cover the common values of the main influencing factors of compaction grouting in sand. It is proven that the equation can predict the grout efficiency of compaction grouting in sand well under different conditions.