Predicting fracture network development in crystalline rocks

Jessica Mcbeck,Yehuda Ben-Zion,Francois Renard,John Mark Aiken,Ben Cordonnier

doi:10.5194/egusphere-egu21-2965

Abstract

&lt;p&gt;The geometric properties of fractures influence whether they propagate, arrest and coalesce with other fractures. Thus, quantifying the relationship between fracture network characteristics may help predict fracture network development, and hence precursors to catastrophic failure. To constrain the relationship and predictability of fracture characteristics, we deform eight rock cores under triaxial compression while acquiring in situ X-ray tomograms. The tomograms reveal precise measurements of the fracture network characteristics above 6.5 microns. We develop machine learning models to predict the value of each characteristic using the other characteristics, and excluding the macroscopic stress or strain imposed on the rock. The models predict fracture development more accurately in the experiments performed on granite and monzonite, than the experiments on marble. Fracture network development may be more predictable in these igneous rocks because their microstructure is more mechanically homogeneous than the marble, producing more systematic fracture development that is not strongly impeded by grain contacts and cleavage planes. The varying performance of the models suggest that fracture volume, length, and aperture are the most predictable of the characteristics, while fracture orientation is the least predictable. Orientation does not correlate with length, as suggested by the idea that the orientation evolves with increasing differential stress and thus fracture length. This difference between the observed and expected predictability of orientation highlights the significant influence of local stress perturbations on fracture growth within brittle material in laboratory-scale systems with many propagating and interacting fractures.&lt;/p&gt;

Highlights

Mechanical weaknesses, such as fractures, control the macroscopic strength of brittle solids (Griffith, 1924)
Consistent with the idea that increasing differential stress should increase the length of fractures and change the fracture orientation (Fig. 2c), the models that predict the fracture orientation depend on the length (Fig. 6)
We develop the models to predict fracture network characteristics using other characteristics, without knowledge of the macroscopic stress or strain imposed on the rock

Summary

Introduction

Mechanical weaknesses, such as fractures, control the macroscopic strength of brittle solids (Griffith, 1924). Machine learning analyses of data from laboratory experiments on crystalline rocks under triaxial compression indicate that whether a fracture propagates or closes depends on the factors that control the stress intensity factor (fracture volume, length, aperture, orientation), as well as the distance to its nearest neighbor (McBeck et al, 2019). These factors, along with the shape anisotropy of the fracture (one minus aperture/length), control the timing of catastrophic failure (McBeck et al, 2020). This approach thereby allows assessing the applicability of existing conceptualizations of fracture network development (Fig. 2)

Objectives

Methods

Results

Discussion

Conclusion