Experimental Relationship Between Compressional Wave Attenuation and Surface Strains in Brittle Rock

Abstract

Linear ultrasonic testing (LUT) has been extensively used as a tool for the evaluation of damage processes in various materials ranging from synthetic metals to natural geomaterials, such as rocks. A key limitation of LUT‐based damage studies to date is the lack of explicit evidence used in associating material damage with the changes in measured LUT attributes (e.g., ultrasonic wave amplitude and velocity). In this study, the evolution of the full‐field strains in brittle rock specimens (Lyons sandstone) subjected to failure are analyzed in real time and linked with the changes in the ultrasonic wave amplitude in localized areas illuminated by ultrasonic beams, termed as the ultrasonic image areas. The noncontact optical full‐field displacement measurement method of 2‐D digital image correlation is implemented in combination with the LUT procedure to continuously track changes in the ultrasonic wave amplitude with the evolution of strains across the surface of the uniaxially loaded intact rock specimens. The ultrasonic amplitude showed near‐linear correlation with the intensity of inelastic tensile strain recorded in the rock specimens. The results from the study corroborate that the ultrasonic changes are in fact influenced by the regions of tensile cracking, which is the primary inelastic deformation mechanism in brittle rocks.

Illumination of Damage in Intact Rocks by Ultrasonic Transmission‐Reflection and Digital Image Correlation

“…2-D DIC is a widely employed non-destructive non-contact optical method for the computation of full-field deformations on a planar surface (Hedayat et al, 2014a;Sutton et al, 2009). 2-D DIC is fundamentally based on the comparison of the digital image of the material surface in its undeformed (or reference) state with the digital images of the same material surface, acquired during different stages of deformation (deformed states) (Shirole et al, 2019a(Shirole et al, , 2019b. For computation of the full deformation field, the reference image is subdivided into square-shaped arrays of pixels defined as subsets (the number of pixels defines the "subset size"), and the distance between each subset (specifically, space between the DIC tracking points) is also specified in terms of pixels ("step size").…”

“…Accordingly, for accurate tracking of the subsets and subsequent deformation measurements, it is crucial that each subset is uniquely identifiable. For that reason, a stochastic pattern of grayscale values (speckle pattern) is typically applied on the specimen surface (Hedayat et al, 2014a;Shirole et al, 2019aShirole et al, , 2019bSutton et al, 2009). To achieve this, a multi-color paint from Rust-Oleum was sprayed on the specimen surface (for details, read Hedayat, 2013, and Shirole et al, 2019a, 2019b.…”

“…For that reason, a stochastic pattern of grayscale values (speckle pattern) is typically applied on the specimen surface (Hedayat et al, 2014a;Shirole et al, 2019aShirole et al, , 2019bSutton et al, 2009). To achieve this, a multi-color paint from Rust-Oleum was sprayed on the specimen surface (for details, read Hedayat, 2013, and Shirole et al, 2019a, 2019b. Following this, images of the speckled surface of the rock specimens were captured in real time during the entire period of the uniaxial compressive loading experiments ( Figure 1a).…”

“…Step size is usually selected such that it is smaller than half the subset size (Hedayat et al, 2014a;Shirole et al, 2019a). Considering the required computation time for the 2-D DIC analysis, a step size of 5 pixels was selected (Shirole et al, 2019a).…”

“…In this study, the damage progression in uniaxially loaded prismatic intact rock specimens was monitored by concurrently applying the T-mode and R-mode LUT techniques. In-sync with the LUT procedure, the stress-induced changes in the strain-field of the rock specimens were also computed using the 2-D DIC technique so as to explicitly correlate the changes in T-mode and R-mode waveforms with the damage in the specimens (Harnett et al, 2018;Shirole et al, 2019aShirole et al, , 2019b).…”