Catastrophic Failure: what controls precursory damage localisation in rocks?

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Geosciences


Catastrophic failure is a critically-important phenomenon in the brittle Earth on a variety of scales, from human-induced seismicity to natural landslides, volcanic eruptions and earthquakes. It is invariably associated with the structural concentration of damage in the form of smaller faults and fractures on localised zones of deformation, eventually resulting in system-sized brittle failure along a distinct and emergent fault plane. However, the process of localisation is not well understood - smaller cracks spontaneously self-organise along the incipient fault plane, often immediately before failure, but the precise mechanisms involved have yet to be determined. Many questions remain, including : Q1 - How do cracks, pores and grain boundaries interact locally with the applied stress field to cause catastrophic failure to occur at a specific place, orientation and time?; Q2 what dictates the relative importance of quasi-static and dynamic processes?; and Q3 - why can we detect precursors to catastrophic failure only in some cases?

Here we will address these questions directly by imaging the whole localisation process, using a newly-developed x-ray transparent deformation cell and fast synchrotron x-ray micro-tomography. We will visualise the nature and evolution of the localisation process structurally and seismically together for the first time at high resolution in a synchrotron. We will deliberately slow the process to image its evolution, and to investigate the strain-rate dependence of the underlying mechanisms, using rapid electronic monitoring and feedback control. This will provide unprecedented direct observation of the relevant mechanisms, including the contribution of seismic (local cracking producing acoustic emissions) and aseismic (elastic loading and silent irreversible damage) processes to the outcome. This innovative combination of techniques is timely, feasible, and is likely to transform our understanding of the role of microscopic processes in controlling system-size failure. The results will provide interpretive models for similar processes in natural and human-induced seismicity, including scale-model tests of strategies for managing the risk of large induced events.

Planned Impact

The potential for impact is significant. In the main part of the proposal we will combine acoustic data with time-lapse micro-CT images to develop a process-based understanding and new theories for the micro-mechanics of strain localisation. This will increase our understanding of localisation processes and our ability to forecast and possibly to control material failure in Earth materials. This is hugely important for facilitating NERC's mission to "Manage our environment responsibly", from natural resource extraction to building resilience to geological hazards. Our results will feed directly into the study of induced seismicity, where the partition of strain between seismic and total strain, and its implications for supporting operational forecasting of seismic risk for engineering decisions, is a pressing and outstanding research question with direct and immediate practical application (Bourne et al., 2014).

Seismic waves and velocity measurements are central in many applications of geophysics. It is now common in the oil and gas industry to use a combination of field scale seismic information and laboratory-derived physical parameters of different rocks to develop and test models for the subsurface structure, and any changes in the reservoir due to production and injection of fluids. These applications require the same 'time-lapse' approach we will use in the laboratory tests, with the advantage that we will know the deformation field independently. In a recent review of trends and challenges in in the industry the authors conclude that "acceptance and application of 4D seismic techniques in both exploration and production indicates that time-lapse 3D exploration and reservoir monitoring are coming of age as a tool to minimize drilling risk and to maximize the return on investment". However, there is often no way to 'ground-truth' such models, as direct imaging of Earth's subsurface is impossible. The improved understanding on the controls on measured velocity and its changes (including subtle changes detected by coda-wave interferometry) is likely to significantly improve our models of subsurface structure and its evolution, and hence reduce the significant costs associated with drilling risk. For example the current cost of a drilling a well in the North Sea is currently just under $10M.

A better understanding of the velocity structure will also allow us to improve the accuracy of earthquake location algorithms for the quantification of seismic and volcanic hazard and risk. The ground-truth data obtained from our proposed blend of technologies in a high-pressure rock deformation apparatus will provide the first direct means to test location algorithms, in particular their ability to detect more localised correlated structures that may pose a greater risk of generating large events.

More generally, the localisation of deformation is a fundamental physical process with a variety of applications in a diverse range of applications, from developing robust materials, including those used in bio-engineering, to non-destructive testing and monitoring of buildings and infrastructure.


10 25 50
Description Data obtained from deforming rock samples in the Diamond Synchrotron has been used to infer the location and size of micro-cracks that develop before catastrophic failure and lead to precursory signals that might be used for forecasting the failure time. The results confirm the gradual localisation of deformation onto a shear band prior to the formation of a thoroughgoing fault, and demonstrate the frequency-size distribution of cracks is a power-law whose exponent decreases in time, similar to those of the radiated acoustic emissions. This is a breakthrough because the power law previously was only inferred - this is the first time it has been tracked live in a test. These results, and a paper describing the design of the new deformation cell, are now published.

In another first we observed simultaneously the development of damage and local strain while also recording acoustic emissions, revealing new details on the process of shear band formation in brittle porous rocks by combining acoustic emission monitoring ('sound') with x-ray 'vision'. This is the first time this combination has been used, providing proof of concept of a breakthrough in experimental capacity, and revealing new insight into the relationship between local deformation processes, the resulting seismic radiation, and bulk velocity changes. The experiments also showed that controlling deformation rate by feedback from acoustic emission rate provided an effective means of controlling the maximum magnitude, with significant potential implications for regulating induced seismicity. These results are now published in a general interest journal (Nature Communications) and one more is due to be submitted in March 2023 concentrating on the acoustic results.

This body of work has generated several invited lectures to international audiences: the 2001 'Fracmeet conference, the SEG passive seismics group (2002), the Geoscience and Geo-Energy Group (2002), and the preliminary results were presented as the AGU 2019 Lorenz lecture in non-linear geophysics.
Exploitation Route Could be used in other applications of non-destructive testing in engineering, or in imaging the progression of damage in bone material for medical applications. The output data has been used as a platform to win a new 'Pushing the Frontiers' grant from NERC.
Sectors Environment

Description They have been discussed forllowing a presentation to the SEG (society of Exploration Geophysicists), including industry practitioners, who contributed to the Q&A session afterwards. The key finding that feedback on seismic event rate may be more effective in suppressing large magnitude earthquakes that the current 'traffic light system' based on maximum magnitude alone may have significant impact on regulation of induced seismicity in the future.
First Year Of Impact 2022
Sector Energy
Title Micromechanics of shear failure in a porous rock: a combined dataset of high-resolution time-resolved 3D x-ray micro-tomography volumes and local 3D strain fields with contemporaneous acoustic emissions and ultrasonic velocity survey waveforms 
Description This dataset shows both the micro-scale mechanisms and acoustic response involved in shear failure of a deforming porous rock. To our knowledge, this is the first such dataset to combine simultaneous acoustic measurements and x-ray tomography imaging. It comprises a time-series of 3D in-situ synchrotron x-ray microtomography (µCT) volumes showing a Clashach sandstone sample (CL10) undergoing triaxial deformation to failure under a constant acoustic emissions (AE) event rate. Use of a constant AE event rate slowed down the failure process after peak stress, enabling shear failure to be captured in unprecedented spatio-temporal detail by the µCT volumes. These volumes are accompanied by the local incremental 3D strain fields and simultaneously acquired waveforms from acoustic emissions and ultrasonic velocity surveys, as well as mechanical bulk stress and strain. These data are fully explained in Cartwright-Taylor et al. Seismic events miss important grain-scale mechanisms governed by kinematics during shear failure of porous rock, in review at Nature Communications. We also include an equivalent time-series of the same data types showing a second Clashach sandstone sample (CL04) undergoing triaxial deformation to failure, this time under a constant strain rate where failure happened abruptly, shortly after peak stress. Both collections were acquired in-situ on the beamline I12-JEEP at the Diamond Light Source, Didcot, UK, in September 2019. Each 3D µCT volume of the sample is contained in a .zip file labelled with the sequential scan number. Each volume comprises reconstructed 16-bit grey-scale data in a sequence of 2D image files (.tif), each numbered according to the depth at which it lies within the sample volume. The file dimensions are pixels, with an edge length of 7.91 µm. Two further .zip files contain the incremental 3D volumetric and deviatoric strain fields, obtained from digital volume correlation between neighbouring µCT volumes. Each strain field consists of a 32-bit 3D image file (.tif) in pixels with an edge length of 316.4 µm, labelled with its scan increment. Also included are (i) .csv files, containing the mechanical stress and strain time-series, the time and mechanical data at which each µCT volume was scanned, and the acoustic emissions event rate data, and (ii) .zip files containing times and waveforms for the acoustic emissions and ultrasonic velocity surveys as .ascii files. The .zip and .xlsx files are labelled with the sample name, the data type (grey-scale, strain-volumetric, strain-deviatoric, seismic, mechanical, mechCT, eventrateAE) and the sequential scan number (grey-scale only) according to the following convention: sample_datatype_scan#. We acknowledge Diamond Light Source for time on beamline I12-JEEP under proposal MG22517. This work is supported by the UK's Natural Environment Research Council (NERC) through the CATFAIL project NE/R001693/1 Catastrophic failure: what controls precursory localisation in rocks? 
Type Of Material Database/Collection of data 
Year Produced 2022 
Provided To Others? Yes  
Description Invited presentation on Earthquake forecasting to the Chinese Earthquake Authority in Beijing. This is the main organisation charged with assessing seismic hazard in China. 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Invited presentation on Earthquake forecasting to the Chinese Earthquake Authority in Beijing to an audience of around 30. This is the main organisation charged with assessing seismic hazard in China. This generated a debate on the uncertainties in time-independent and time-dependent seismic hazard maps.
Year(s) Of Engagement Activity 2019
Description SEG Webinar on implications of project results for interpretation of induced microseismicity for practitioners 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Several hundred attended the webinar, with a following Q&A. HOsted by the SEG passive seismics group.
Year(s) Of Engagement Activity 2022