Quantification of pre-peak brittle damage: Correlation between acoustic emission and observed micro-fracturing

https://doi.org/10.1016/j.ijrmms.2006.11.003Get rights and content

Abstract

Acoustic emission (AE) is a useful non-destructive technique to determine whether damage in rock material has occurred, even though the interpretation of AE can be complex. In this paper, two laboratory situations are studied, to investigate the correlation between micro-fractures and AE. In the first case, damage is induced by pure macro-compressive stresses and in the second case by macro-tensile stress (in one direction). The specific evolution of the recorded cumulative AE energy as a function of the applied load leads, in both cases, to a subdivision of the loading in different phases. Thresholds for the transition of these phases are defined. A thorough petrographical analysis of thin slices of samples, damaged to these different thresholds, allows a physical interpretation of damage phases. This methodology leads to a damage evolution model both for macro-compressive and macro-tensile stresses. This paper demonstrates the advantages of the combined use of AE and detailed petrographic study of thin rock slices. Furthermore, insight is provided on the pre-peak damage evolution caused by macro-compressive and macro-tensile stresses in rock material.

Introduction

A change in the stress-state of rock influences its micro-scale state through the development of cracks and fissures. A change in the global stress state is also manifested as a variation in local stresses within the sample. If the local stresses exceed the local strength (compressive or tensile), cracks occur, grow and interact [1]. This progressive fracturing of rock results in the formation and growth of the macro-fracture(s). The pre-peak damage process remains the subject of extensive research in order to fully describe and better understand the process [2], [3]. A wide variation of tools such as scanning electron microscopy [4], resistivity measurements [5], acoustic emission [6] have been used.

Some things are known about the pre-peak damage process caused by macro-compressive stresses. Different stress thresholds, based on stress–strain curves and AE are defined [7], [8], [9]. Discussion of the physical mechanisms such as the systematic initiation of cracks and the crack damage (i.e., the occurrence of unstable crack growth) are given for these stress thresholds. A corresponding mathematical formulation of the systematic crack initiation and crack damage thresholds as a function of the uniaxial compressive strength during uniaxial and triaxial tests is described by [10]σ1-σ3=Aσucswhere σ1, σ3, σucs are, respectively, the major and minor principal stress at the threshold, and the uniaxial compressive strength. The material parameter A varies between 0.4 and 0.6 for crack initiation, and between 0.7 and 0.9 for crack damage.

The damage process caused by macro-tensile stresses is generally considered using the mathematical description of the growth of a single crack. This crack is present in the material in the form of micro-cracks, which propagate towards a single macro-crack, finally leading to failure. In this approach, the local stress redistribution around the tip of the crack determines the growth process [11]. Some authors also discuss empirical observations of fracture patterns caused by macro-tensile stresses [4].

This paper describes several damage phases defined by laboratory recorded cumulative AE energy. Furthermore, the micro-damage, corresponding to the damage phases, is described using a systematic observation of thin slices. As some insight of AE and failure process due to macro-compressive stresses is given in literature [7], [8], [9], this methodology is firstly used to describe and verify the failure due to pure macro-compressive stresses. Furthermore, this methodology is extended to describe the failure process due to macro-tensile stresses (in one direction). In order to create macro-compressive and macro-tensile stresses in a localised zone, a specific sample geometry is used, as described in the following paragraph.

Section snippets

Methodology

The rock used in this laboratory study is Belgian crinoidal limestone of Carboniferious age (Tournesian period: 368–359 million years ago, [12]). This material is predominantly composed of calcite, in various forms. Apart from crinoids, the limestone also contains additional bioclasts. Another major constituent of the crinoidal limestone is micrite. This very fine-grained microcrystalline calcite makes up the matrix of the limestone [4], [13]. Stylolites are frequently encountered. Test

Results and interpretation

Five and seven samples are loaded in the compression and tension configurations, respectively. In the evolution of the recorded cumulative AE energy, four (compression configuration) and three (tension configuration) different phases can be distinguished during the loading. Three additional samples, which are not loaded, are used as reference samples. In total, seven thin slices are prepared of which five are digitised and further analysed. Since the maximum applied force differs in individual

Conclusions

Using the evolution of the recorded cumulative energy of AE and a thorough micro-petrographic study of thin slices, two damage models (in the case of macro-compressive and -tensile stresses) are proposed.

In the case of macro-compressive stresses, four phases during loading can be distinguished based on the recorded cumulative AE energy evolution: (1) no AE are recorded, (2) the phase of the fae, (3) the phase of the systematic crack formation and growth, (4) the phase of the crack interaction

Acknowledgement

The financial support of the Research Council of the Katholieke Universiteit Leuven (OT-project OT/03/35) is gratefully appreciated.

References (26)

Cited by (0)

View full text