Experimental study on strain localization and slow deformation evolution in small-scale specimens

Fayuan Yan, Enzhi Wang, Xiaoli Liu, Chengzhi Qi, Weijia JiaInternational Journal of Coal Science & Technology 2025 · DOI

This academic paper, published in the International Journal of Coal Science & Technology, explores the micro-mechanics of how rocks deform and fail. By using high-resolution imaging on small sandstone samples, the researchers bridge the gap between laboratory observations and the massive "slow deformation waves" that precede and follow earthquakes in the Earth's crust.

TL;DR

Researchers used Digital Image Correlation (DIC) to visualize how "slow deformation waves" move through red sandstone during compression. They discovered that strain localizes into specific "flow channels" and "Lüders bands" long before the rock actually breaks, providing a laboratory-scale model for understanding seismic migration and earthquake precursors.

What Problem Does This Paper Address?

Earthquakes are often thought of as sudden, catastrophic snaps. However, geophysicists have long known that the Earth’s crust also experiences "slow deformation waves"—movements that propagate at speeds far lower than standard seismic waves (tens of kilometers per year rather than kilometers per second). These slow waves play a critical role in how stress is transferred across tectonic plates and how earthquakes are "bred."

While these waves have been observed in large-scale field data and in ductile materials like metals, studying them in brittle rock at a laboratory scale is difficult. There is a significant knowledge gap regarding how these waves initiate, how they relate to the internal microstructure of the rock, and how they eventually lead to "strain localization"—the point where deformation concentrates into a single fracture.

Methods

The authors performed uniaxial compression tests on uniform red sandstone specimens (50 × 50 × 100 mm). To capture the minute movements of the rock surface, they employed two key techniques:

  1. Speckle Photography: Samples were spray-painted with a random pattern of black and white dots.
  2. Digital Image Correlation (DIC): A high-definition industrial camera recorded the sample at 14.2 frames per second. DIC software then tracked the movement of the speckles to calculate the full-field strain (longitudinal, transverse, and shear) with high precision.

The tests were conducted at three different displacement-controlled loading rates (0.1, 0.5, and 1.0 mm/min) to see how speed influenced the propagation of the deformation waves.

Datasets

The study does not use a public machine learning dataset. Instead, it generates primary experimental data from:

  • Red Sandstone Samples: Chosen for their relatively low stiffness, which makes slow deformation easier to observe compared to harder rocks like granite.
  • Stress-Strain Time Series: High-resolution data capturing the transition from elastic loading to the stress-holding stage.
  • Strain Cloud Maps: Spatiotemporal maps of strain components ($\epsilon_{xx}$, $\epsilon_{yy}$, $\epsilon_{xy}$) across the specimen surface.

Key Insights and Ideas

  • The "Particle Motion" Link: The researchers found that the velocity at which the deformation wave moves through the rock is proportional to the loading rate. They conclude that "slow deformation transfer velocity" is essentially the transfer of the particle's motion state, driven by local stress gradients.
  • Flow Channels: Deformation doesn't happen uniformly. It first organizes into "flow channels"—a mesh-like distribution of strain that crosses the specimen symmetrically.
  • Nucleation Pairs: Strain "nuclei" (the starting points of failure) appear in pairs—one in a compressive region and one in a tensile region. This represents the microscopic origin of dislocations within the rock structure.
  • Lüders Bands in Rock: Usually discussed in metallurgy, the authors identified Lüders bands (concentrated deformation bands) in the sandstone. These bands form before the rock enters its macroscopic plastic deformation stage and are directly influenced by the maximum values within the flow channels.

Key Results

  • Wave Propagation: The study confirmed that deformation propagates as a slow wave from the active loading end toward the fixed end of the specimen.
  • Velocity Calculations: At a loading rate of 0.5 mm/min, the deformation speed at specific coordinates was measured between $4.68 \times 10^{-5}$ m/s and $7.25 \times 10^{-5}$ m/s.
  • Localization Mechanism: The researchers mapped the evolution of failure: it begins with random particle movement, organizes into flow channels, develops into Lüders bands, and finally concentrates into a fracture surface.
  • Seismic Analogy: The laboratory-scale deformation transfer showed characteristics similar to seismic migration waves, suggesting that small-scale rock tests can validly simulate large-scale tectonic processes.

Limitations and Future Work

The authors acknowledge that while these small-scale tests provide a "reference significance," the Earth's crust is far more complex than a uniform sandstone block.

  • Scale Effects: The transition from a 10cm lab specimen to a 10km crustal segment involves massive changes in pressure, temperature, and material heterogeneity that weren't modeled here.
  • Texture Uniformity: The study used "uniform texture" sandstone to isolate the wave mechanics. Future work could introduce pre-existing faults or joints to see how they disrupt or accelerate slow deformation waves.
  • Application: The next step is using these DIC-derived patterns to create more accurate predictive models for seismic migration in specific geographic regions.

Summary updated

Abstract

Abstract A kind of slow deformation wave is produced in the crust under the action of internal and external factors, which plays an important role in the formation and occurrence of earthquakes. In this paper, uniaxial compression tests are carried out on red sandstone samples with uniform texture. Displacement controlled loading methods are adopted, and the loading rates are 0.1 mm/min, 0.5 mm/min and 1.0 mm/min, respectively. The micro-characterization method of speckle photography and DIC processing technology are adopted. The transfer characteristics of slow deformation and strain localization of red sandstone specimens during loading and deformation are discussed. The results show that the boundary advance velocity is proportional to the slow deformation transfer velocity with the change of position, so it can be considered that the slow deformation transfer velocity is equal to the particle motion transfer velocity. The formation and development of sample strain localization may be determined by the flow channel, nucleation and Luders zone evolution. The formation of the Luders band is related to the maximum value of the flow channel, and as deformation increases, the Luders band merges and develops with the maximum value of the nearby flow channel. By applying different loading rates, the influence of loading rate on the average transfer velocity of slow deformation was obtained; the slow deformation wave during seismic migration has similar characteristics to the deformation transfer in the experiment, therefore the research results have reference significance for further studies on the evolution characteristics of slow deformation waves and seismic migration.
BibTeX
@article{yan2025experimental,
  author = {Fayuan Yan and Enzhi Wang and Xiaoli Liu and Chengzhi Qi and Weijia Jia},
  title = {{Experimental study on strain localization and slow deformation evolution in small-scale specimens}},
  journal = {International Journal of Coal Science \& Technology},
  year = {2025},
  doi = {10.1007/s40789-025-00771-8},
}
Figure