Pattern Recognition Applicability of Artificial Neural Networks in Rock Abrasiveness Determination Using Rock Strength and Brittleness Data

Asadi, A. (Islamic Azad University)

OnePetro 

ABSTRACT: Cerchar abrasion index (CAI) is commonly used to represent rock abrasion for estimation of bit life and wear in various mining and tunneling applications. The test is simple and fast, but there have been some discrepancies in the test results which are related to the type of equipment, condition of the rock surface, operator skills, testing procedures, and measuring the wear flat. This paper focuses on the estimation of CAI and investigates the impact of various parameters on that. Results of a limited Cerchar tests on a set of rock samples from different laboratories are analyzed to correlate rock properties data to CIA value, which every value indicate an abrasiveness classification. As a result of a literature review, it is concluded that the abrasiveness of a rock sample based on the CAI value is strongly correlated with uniaxial compressive strength (UCS) and brittleness of rock samples. Rock brittleness is a function of UCS and Brazilian tensile strength (BTS). Thus, collected data of these parameters were hired to develop and train artificial neural networks (ANN) as an artificial intelligence (AI) method for estimation of drilling tool wear using data of rock strength and brittleness as inputs. It is pursued by the application of pattern recognition which is achieved by ANNs.

1. INTRODUCTION

In this research, artificial neural networks (ANN) are applied to predict drilling bit wear using collected and calculated data of rock abrasiveness, including uniaxial compressive strength, Brazilian tensile strength and rock brittleness.

This purpose is pursued by performance of pattern recognition as it is one of the major applications of ANNs. The final proposed neural networks produces the abrasiveness classification by receiving data of UCS, BTS and brittleness.

1.1. Wear and Rock Abrasiveness The wear is in many ways similar to the effect of harder minerals on softer ones and is easily represented by the scratch that the hard objects engrave in soft minerals. Plinninger et al., 2002, depicted that “Abrasive wear” is the predominant wear process in most rock types.