Dynamic instabilities of frictional sliding at a bimaterial interface

Efim A. Brener, Marc Weikamp, Robert Spatschek, Yohai Bar‐Sinai, Eran BouchbinderJournal of the Mechanics and Physics of Solids 2016 · DOI · arXiv

TL;DR

This paper provides a rigorous 2D elastodynamic stability analysis of frictional sliding between dissimilar materials, accounting for finite-size effects and complex friction laws. The authors discover a "universal instability" at long wavelengths and identify specific high-speed wave modes that drive instability even when the friction interface is inherently velocity-strengthening.

What Problem Does This Paper Address?

Frictional sliding is a fundamental process in both geophysics (earthquakes) and engineering (brake squeal, machining). A major challenge in modeling these systems is the "bi-material effect": when two different materials slide against each other, changes in slip velocity are coupled to changes in normal stress, which can trigger instabilities.

Previous research often relied on two simplifications:

  1. Quasi-static approximations: Ignoring inertia and wave propagation, which is only valid at very low speeds.
  2. Infinite geometries: Assuming the sliding bodies are infinitely large, ignoring how the finite height of a block affects its stability.

The authors tackle these gaps by performing a full linear stability analysis that includes both elastodynamics (wave effects) and a finite system height ($H$), using the realistic "rate-and-state" friction framework.

Methods

The researchers modeled a 2D system consisting of a deformable elastic solid of height $H$ sliding on a perfectly rigid substrate.

  • Linear Stability Analysis: They applied small spatiotemporal perturbations to a steady sliding state and derived a complex-variable "spectrum equation." This equation relates the wave-number ($k$) of a perturbation to its growth rate ($\Lambda$).
  • Constitutive Framework: They primarily used the standard rate-and-state friction law, which accounts for the "age" of contact points and instantaneous velocity changes.
  • Generalized Models: They also tested the Prakash-Clifton and Linker-Dieterich models, which introduce a finite-time (non-instantaneous) response of friction to changes in normal stress.
  • Asymptotic Analysis: The authors analyzed two limits: the small $kH$ limit (long wavelengths relative to block height) and the large $kH$ limit (short wavelengths or infinite system size).

Datasets

This is a theoretical and analytical paper; no new experimental datasets were generated. However, the authors cite a large compilation of experimental data from Bar-Sinai et al. (2014) to justify their focus on "velocity-strengthening" friction, a regime previously thought to be stable but shown here to be susceptible to dynamic instabilities.

Key Insights and Ideas

  • Universal Instability: Perhaps the most striking finding is that for any finite height $H$, homogeneous sliding is universally unstable at long wavelengths. This instability is mediated by "waveguide-like" modes that are strongly coupled to the system's boundaries.
  • Symmetry Breaking: Friction breaks the directional symmetry of wave propagation. The analysis shows that instabilities traveling in the direction opposite to sliding (at nearly the dilatational wave speed) are far more dominant than those traveling in the sliding direction.
  • The Role of Velocity-Strengthening: Conventionally, friction that gets "stronger" as you slide faster (velocity-strengthening) is seen as stabilizing. This paper proves that elastodynamic coupling at a bimaterial interface can overcome this stabilization, leading to "super-shear" instabilities.
  • Mathematical Well-Posedness: While simple Coulomb friction models often lead to "ill-posed" math (where perturbations grow infinitely fast at short wavelengths), the authors show that rate-and-state friction keeps the problem mathematically sound even when the system is physically unstable.

Key Results

  • Growth Rate Scaling: In the small $kH$ limit, the maximum growth rate of the universal instability scales as $1/H$. This means thinner sliding bodies are more unstable.
  • Dominant Modes: In the large $kH$ limit, the dominant instability travels at nearly the dilatational wave speed ($c_d$) in the "un-preferred" direction.
  • Stability Phase Diagrams: The authors produced comprehensive phase diagrams using dimensionless parameters ($\Delta, \beta, \gamma, f$). These diagrams clearly delineate the boundaries where quasi-static, shear-wave, and dilatational-wave instabilities emerge.
  • Regularization Effects: Incorporating a finite-time response to normal stress (as in the Prakash-Clifton model) was shown to significantly promote stability, delaying the onset of instabilities and reducing the range of unstable wave-numbers.

Limitations and Future Work

  • Strong Material Contrast: The study assumes one of the two solids is infinitely rigid. While this simplifies the math and provides an upper bound for instability, real-world cases involve two deformable bodies with finite material contrast.
  • Linear vs. Non-linear: As a linear stability analysis, the paper describes the onset of instability but cannot predict the final form of the resulting rupture (e.g., whether it becomes a crack-like front or a pulse-like slip).
  • Lateral Boundaries: The model assumes the system is infinite in the sliding direction ($x$). Future work could incorporate a finite length $L$, which would introduce additional "global" instability modes.

Summary updated

BibTeX
@article{brener2016dynamic,
  author = {Efim A. Brener and Marc Weikamp and Robert Spatschek and Yohai Bar‐Sinai and Eran Bouchbinder},
  title = {{Dynamic instabilities of frictional sliding at a bimaterial interface}},
  journal = {Journal of the Mechanics and Physics of Solids},
  year = {2016},
  doi = {10.1016/j.jmps.2016.01.009},
}
Figure