IDENTIFICATION OF THE GOVERNING PARAMETERS RELATED TO ROCK INDENTATION DEPTH BY USING SIMILARITY ANALYSIS

This paper focuses on the indentation depth in rocks caused by a hemis pherical indenter. The problem is approached by a combination of simil arity methods with an artificial neural network. The similarity method s offer a profound understanding of the physical problem and help to i dentify the most important governing parameters or factors that reflec t the essence of the rock indentation events, thus simplifying the tar get problem. The artificial neural network provides an advanced comput ing model, which allows more factors to be involved. The predictions o btained using this combined approach are in better agreement with the experimental results than predictions using other methods. (C) 1998 El sevier Science B.V. All rights reserved.