🔬 materials science

A variational critical-state theory of friction

This paper presents a variational, finite-kinematics framework for modeling fault gouge as a rigid-viscoplastic, pressure-sensitive material, deriving explicit rate-and-state dependent solutions that align with experimental data and connect to empirical friction laws.

Original authors: Mary Agajanian, Nadia Lapusta, Anna Pandolfi, Michael Ortiz

Published 2026-02-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Mary Agajanian, Nadia Lapusta, Anna Pandolfi, Michael Ortiz

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Earth's crust as a giant, jigsaw puzzle. Sometimes, the pieces don't fit perfectly, and there are cracks between them called faults. When these cracks rub against each other, they don't just slide smoothly like ice on ice. Instead, the grinding creates a fine, powdery dust known as fault gouge.

Think of this gouge like a layer of sand between two heavy books. If you try to slide the books past each other, the sand doesn't just sit there. It squishes, it spreads out, it heats up, and it changes how hard it is to move the books.

This paper is a new, sophisticated "instruction manual" for understanding exactly how that layer of sand (fault gouge) behaves when you push it. Here is the breakdown in simple terms:

1. The Problem: Why is this hard to predict?

Scientists have known for a long time that earthquakes happen when stress builds up and the fault suddenly slips. But the "sand" in between is tricky.

It changes shape: When you push it, it might get thinner (squish) or thicker (dilate/spread out).
It remembers: How hard you pushed it yesterday changes how it reacts today.
It's fast: During an earthquake, things happen so fast that the sand behaves differently than when you push it slowly in a lab.

Old models were like using a crystal ball: they guessed the behavior based on patterns they saw in the past, but they didn't really explain why the sand was acting that way. They were "empirical" (based on observation) but lacked a deep physical foundation.

2. The Solution: A New "Physics Engine"

The authors built a new mathematical framework. Instead of guessing, they built the model from the ground up using variational principles.

The Analogy: Imagine you are rolling a ball down a hill.

Old way: You watch the ball roll and write down, "It goes faster when the hill is steep."
This paper's way: They start with the fundamental law that "nature always chooses the path of least resistance" (or maximum energy loss, in this case). They let the math figure out exactly how the ball moves based on the shape of the hill and the texture of the ground.

They treat the fault gouge as a rigid-viscoplastic material.

Rigid: It doesn't stretch like a rubber band; it's hard rock.
Viscoplastic: It flows like honey when pushed hard enough, but acts like a solid when pushed gently.

3. The "Cam-Clay" Connection

The authors borrowed a famous model used for clay (called Cam-Clay) and adapted it for earthquake faults.

The Clay Analogy: Think of a wet sponge. If you squeeze it hard, water comes out, and it gets smaller (compacts). If you squeeze it and then pull it apart, it might spring back a bit.
The Fault Analogy: The fault gouge acts like that sponge.
- If the rock above is very heavy (high pressure), the sand gets compacted and hardens.
- If the rock is lighter, the sand dilates (spreads out) and softens.

The paper creates a mathematical "map" (called a yield surface) that shows exactly when the sand will start to flow and how much it will spread out.

4. The "Speed" Factor (Rate-and-State)

One of the most famous ideas in earthquake science is "Rate-and-State Friction." It basically says:

Rate: How fast you slide matters.
State: The "history" of the slide matters (how long it's been sitting there, how rough the surface is).

This paper takes that idea and gives it a physical body. Instead of just saying "friction depends on speed," they show how the speed changes the internal structure of the sand grains.

Fast sliding: The sand grains don't have time to settle, so they might jam or spread out, changing the friction.
Slow sliding: The grains have time to rearrange and settle into a stable pile.

5. The Big Discovery: Stability vs. Earthquakes

The most exciting part of the paper is what they found about stability.

Rate Strengthening: If you slide faster, the friction gets higher. This is like a car with good brakes; if you speed up, the brakes grab harder, and you stay stable. This leads to slow, safe "creep" of the fault.
Rate Weakening: If you slide faster, the friction gets lower. This is like a car with bad brakes; the faster you go, the less grip you have, leading to a skid. This is how earthquakes start.

The authors found that whether a fault is stable or prone to earthquakes depends on a specific ratio: how much the rock above is pressing down vs. how much the sand has been squished in the past.

If the sand is very "pre-squished" (consolidated) and the pressure is low, it tends to weaken and cause earthquakes.
If the conditions are different, it stays stable.

6. Why Does This Matter?

This isn't just math for math's sake.

Better Predictions: By understanding the physics of the sand, we can build better models to predict when a fault might slip.
Connecting the Dots: It bridges the gap between simple lab experiments (pushing sand in a box) and the complex reality of the Earth's crust.
Future Tools: The authors suggest this framework could eventually include things like water pressure in the rocks (pore fluids), which is a huge factor in triggering real earthquakes.

Summary

Think of this paper as upgrading the software for a video game.

Old Game: The sand was a static texture. You pushed it, and it slid based on a simple rule.
New Game (This Paper): The sand is a living, breathing physics object. It squishes, it expands, it remembers its history, and it reacts differently depending on how fast you move.

By understanding these deep mechanics, scientists hope to better understand the "ticking time bomb" of earthquakes and perhaps one day predict them more accurately.

1. Problem Statement

Earthquakes, landslides, and volcanic eruptions are driven by the slip of fault surfaces, which is accommodated within a thin layer of comminuted granular material known as fault gouge. Characterizing the mechanical behavior of this gouge is a major challenge in earthquake source physics due to:

Large Deformations: Strains can be extremely large (meters of slip in millimeter-thick layers).
Complex Physics: The material exhibits inelasticity, rate dependence, and significant volumetric changes (dilation/compaction).
Limitations of Existing Models: Current approaches often rely on empirical rate-and-state friction laws or ad-hoc continuum models (e.g., Drucker-Prager) that prescribe yield surfaces a priori. These models often lack a rigorous thermodynamic foundation, struggle with finite kinematics, and fail to naturally link micro-mechanical processes (like grain rearrangement) to macroscopic friction laws.

The authors aim to develop a systematic, variational, finite-kinematics framework for fault gouge that derives the effective stress-strain response from first principles (maximum dissipation) rather than empirical fitting.

2. Methodology

The authors construct a theoretical framework based on critical-state plasticity adapted for finite deformations, specifically utilizing a rigid-viscoplastic assumption.

Variational Framework: The theory is built upon the principle of maximum dissipation. Instead of prescribing a yield surface, the authors define a dissipation potential $\Psi$ dependent on the rate of deformation $D$ . The stress-strain relations are derived as the gradient of this potential ( $\sigma = \partial \Psi / \partial D$ ).
Finite Kinematics: They employ a multiplicative decomposition of the deformation gradient ( $F = F_e F_p$ ). Crucially, they assume rigid elasticity ( $F_e$ is a pure rotation), which is justified because elastic strains in rock are negligible compared to the finite plastic strains in fault gouge during earthquakes.
Cam-Clay Specialization: The general framework is specialized to the Cam-Clay model, a critical-state model originally developed for soils. This involves:
- Defining an elliptical yield surface in the mean effective stress ( $p$ ) vs. deviatoric stress ( $q$ ) plane.
- Introducing a state variable representing the consolidation pressure ( $p_0$ ), which evolves with plastic volumetric strain.
- Formulating rate potentials ( $\psi$ ) to capture rate-dependent viscous effects.
Analytical Derivation: For a shearing and dilating granular layer under constant normal stress, the authors derive closed-form analytical solutions for the evolution of shear stress and layer thickness. This is achieved by assuming a separation of scales where shear rates dominate dilation rates.
Validation: The model is validated through:
1. Numerical Simulations: Solving the governing equations at a material point using an implicit backward-Euler scheme.
2. Comparison with Experiments: Matching results against standard laboratory tests (oedometer consolidation and velocity-step shear tests) and comparing them to existing theoretical models.

3. Key Contributions

Variational Derivation of Friction Laws: The paper derives the effective stress-strain law for a shearing granular layer directly from a dissipation potential, rather than postulating it. This provides a rigorous thermodynamic basis for rate-and-state friction laws.
Unified Treatment of Consolidation and Shear: The model naturally couples shear resistance with volumetric changes (dilation/compaction) via the critical-state concept. It explains how the pre-consolidation pressure dictates whether a material hardens/compacts or softens/dilates.
Closed-Form Solutions: The authors provide explicit analytical expressions for the evolution of shear stress and layer height during simple shear, including transient responses to velocity steps and steady-state behaviors.
Parameter Identification: The framework demonstrates that the model can be calibrated using only standard consolidation experiments and convexity considerations, avoiding the need for complex inverse fitting of shear data.

4. Key Results

Oedometer (Consolidation) Tests: The model successfully reproduces the stiffening of granular materials with increased pre-consolidation pressure. It captures the linear relationship between vertical strain and the logarithm of normal stress, as well as the rate-dependence of the consolidation curve.
Shear Tests (Constant Rate):
- The model predicts a transition in behavior based on the ratio of normal stress to pre-consolidation pressure ( $\hat{\sigma}_n$ ).
- Loosely consolidated materials ( $\hat{\sigma}_n < 2$ ) exhibit hardening and compaction.
- Highly consolidated materials ( $\hat{\sigma}_n > 2$ ) exhibit softening and dilation.
- The steady-state friction coefficient converges to a value consistent with Byerlee's law (~0.6–0.85).
Velocity-Step Tests:
- The model reproduces the characteristic non-monotonic response to step changes in slip rate: an instantaneous rate-dependent jump in shear stress followed by a gradual evolution to a new steady state.
- It captures the transition from steady-state rate strengthening (stable creep) to steady-state rate weakening (potential for dynamic rupture/earthquakes) as the confinement ratio changes.
Rate Sensitivity: Two different rate-potential forms (linear and activation-theory motivated) were tested. Both captured the qualitative experimental trends, though the linear case showed stronger rate dependence.

5. Significance

Bridging Scales: The work bridges the gap between micro-mechanical granular physics (critical state, grain rearrangement) and macroscopic fault friction laws.
Rationalizing Rate-and-State Friction: It offers a physical interpretation for the empirical "state variable" in rate-and-state friction laws, identifying it as the effective consolidation state (or pre-consolidation pressure) of the fault gouge.
Earthquake Mechanics: By linking the transition from rate strengthening to rate weakening to the pre-consolidation pressure, the model provides a mechanistic explanation for why some fault segments slip stably while others generate earthquakes.
Future Applications: The variational framework is extensible. The authors note it can be easily adapted to include pore fluid effects (Terzaghi effective stress) and temperature-dependent shear heating, offering a path toward more comprehensive earthquake source models.

In summary, this paper presents a rigorous, thermodynamically consistent framework that replaces empirical friction laws with a derived constitutive model capable of capturing the complex, coupled shear-dilation behavior of fault gouge under finite deformations.