Mechanical, micro-structure and contour mapping analyses of highly-porous intermediate-texture analog rock from instrumented micro-indentation in conjunction with statistical/machine learning tools

Exploring the spatial variability of the mechanical properties of rocks and cementitious/porous materials, provides an understanding on the distribution of the mechanical phases and the surface characteristics of the material. This type of analysis often necessitates instrumented indentation experiments to be performed, which comprises a highly promising technique in the characterization of rocks in the fields of unconventional reservoir and energy geomechanics, as the indentation test results can offer an alternative and cost-effective assessment of the bulk behavior of rocks and help to interpret geophysical and stratigraphic analyses. In the present study, we investigated through micro-indentation experiments the elastic properties of a crystalline-to-vesicular intermediate-texture analog rock and the relationship of modulus and hardness with the different phases of the material. For this purpose, alumina-based ceramic was used as a paradigm analog rock, which conforms with the columnar class displaying a structure similar to that of some types of extrusive igneous rocks, resembling a porous structure which can be of interest in many areas of geo-energy and resources engineering. An instrumented experimental program adopting the grid method was implemented and the samples were tested in their original (non-abraded) and in abraded states. The results revealed high spatial variability of hardness and modulus and significant influence of the position the indenter penetrates the sample. This provides a qualitative inference that a single value of hardness (or modulus) may not adequately represent the inherent properties of rocks. Clustering analysis, Weibull statistics and contour mapping were some of the major tools used to characterize the analog rock. The Weibull analysis could provide better qualitative inferences on the influence of abrasion on the elastic properties of the material compared with the direct use of mean values, and the mapping showed, on a qualitative standpoint, a good correlation between elastic properties, chemical composition and textural characteristics of the samples. These results provide some new directions in the application of instrumented indentation, in conjunction with Weibull and machine learning analyses, in the characterization of rocks and solids of complex structures, dictating that interpretation of mean values may not be the most effective way to analyze indentation test data, but an approach such as Weibull statistics can provide a more effective means to deal with the large discrepancy of the test results, which is expected to be the case for many rocks. As porous rocks are often encountered in sedimentary basins and have been identified as a potential path for expanding geo-energy production, the present study provides promising directions in the application of instrumented indentation for the characterization of such materials with direct interest in oil/gas exploitation and design of the production process.