Earth Science · Seismology
June 2026 · 13 min read · Faults · Stick-Slip · Magnitude · Last updated: 22 June 2026

Fault Mechanics: From Stress to Earthquake

Written by MySimulator Team · Reviewed by MySimulator Editorial Review

An earthquake is the sudden release of energy that has been quietly accumulating in the Earth's crust for years, decades, or centuries. Tectonic plates grind past one another, but faults do not slide smoothly — friction locks them in place while elastic strain builds in the surrounding rock, until in seconds the rock snaps back and shakes the ground. This article develops the mechanics of that cycle: Reid's theory of elastic rebound, the stick-slip friction captured by the Burridge-Knopoff model, the Mohr-Coulomb criterion for fault failure, the seismic moment that defines modern magnitude, the remarkable Gutenberg-Richter scaling law, and the physics behind earthquake early-warning systems.

1. Elastic Rebound: The Earthquake Cycle

After studying ground deformation across the 1906 San Francisco earthquake, Harry Fielding Reid proposed the elastic rebound theory, still the foundation of earthquake physics. The crust on either side of a locked fault deforms elastically — like a bent spring — as the plates move, storing strain energy. When the accumulated stress exceeds the fault's frictional strength, the fault slips, the rock springs back to its relaxed shape, and the stored elastic energy radiates away as seismic waves.

This gives the earthquake cycle: a long interseismic period of slow strain accumulation, a brief coseismic rupture lasting seconds to minutes, and a postseismic relaxation as the crust adjusts. The strain that drives it follows from elasticity:

Elastic strain energy stored per unit volume: U = τ² / (2μ) where τ = shear stress on the fault μ = shear modulus of the rock (~30 GPa for crust) Slip deficit accumulates at the plate rate v over recurrence time T: D ≈ v · T (the slip released in the next event)

The crucial insight is that the energy released in an earthquake was stored gradually over the entire interseismic period. A fault accumulating strain for 150 years can release it all in under a minute — a power amplification of roughly eight orders of magnitude.

2. Mohr-Coulomb Failure

When does a fault actually slip? The Mohr-Coulomb failure criterion states that a fault slides when the shear stress acting along it exceeds the frictional resistance, which depends on the normal stress clamping the fault closed.

Failure occurs when: τ ≥ C + μ_f · (σ_n − P) where τ = shear stress on the fault plane C = cohesion (often ~0 for a pre-existing fault) μ_f = coefficient of friction (~0.6–0.85, "Byerlee's law") σ_n = normal stress clamping the fault P = pore fluid pressure (σ_n − P) = effective normal stress

The role of pore fluid pressure P is profound. Fluid in the rock pushes the fault walls apart, reducing the effective normal stress and therefore the frictional resistance. This is why injecting fluid underground — wastewater disposal, geothermal stimulation, or dam-reservoir filling — can trigger induced seismicity: it does not add stress so much as unclamp faults already close to failure. The same mechanism, natural pressurisation of trapped fluids, contributes to many tectonic earthquakes.

3. Burridge-Knopoff Stick-Slip

Faults exhibit stick-slip behaviour: they stick under friction, load up stress, then slip suddenly. In 1967 Burridge and Knopoff built a simple mechanical model that captures the essential dynamics and remarkably reproduces real earthquake statistics.

Picture a chain of blocks resting on a rough surface, each connected to its neighbours by springs and to a slowly moving driver plate by a loader spring. The blocks stick until the accumulated spring force overcomes static friction; then they slip, and the slip of one block can load its neighbours past their own thresholds, propagating a rupture along the chain.

Equation of motion for block i (1D Burridge-Knopoff): m·ẍ_i = k_c·(x_{i+1} − 2x_i + x_{i−1}) + k_p·(v·t − x_i) − F_friction(ẋ_i) where k_c = coupling-spring stiffness between blocks k_p = loader-spring stiffness to the driving plate v = plate loading velocity F_friction = velocity-weakening friction (key to instability)

The decisive ingredient is velocity-weakening friction: friction that drops as sliding speeds up. This is what makes the slip unstable and explosive rather than a smooth creep, and it is formalised in modern rate-and-state friction laws. The Burridge-Knopoff model is a celebrated example of self-organised criticality: without any tuning, it spontaneously produces a spectrum of slip events from tiny to huge, following a power-law size distribution — the same Gutenberg-Richter scaling seen in real seismicity.

4. Seismic Moment and Magnitude

How big is an earthquake? The most physically meaningful measure is the seismic moment M₀, which captures the actual work done by the rupture.

Seismic moment: M₀ = μ · A · D where μ = shear modulus of the rock (~30 GPa) A = area of the ruptured fault surface (length × width) D = average slip across that surface Units: newton-metres (N·m)

Older scales like the Richter (local) magnitude saturate for great earthquakes — they simply stop increasing once the rupture grows beyond the wavelengths the instrument measures. The moment magnitude scale M_w, defined directly from M₀, does not saturate and is used for all significant events today:

Moment magnitude: M_w = (2/3)·log₁₀(M₀) − 6.07 (M₀ in N·m) Logarithmic consequences: +1 magnitude unit → ~31.6× more energy (10^1.5) +2 magnitude units → ~1000× more energy So an M8 releases roughly 1000× the energy of an M6.

Because M₀ scales with rupture area times slip, the largest earthquakes require enormous faults: the 2011 Tohoku (M9.0) and 2004 Sumatra-Andaman (M9.1) events ruptured subduction megathrusts hundreds to over a thousand kilometres long. There is an upper limit set simply by the longest continuous fault the planet can offer.

5. The Gutenberg-Richter Law

One of the most robust empirical laws in all of geophysics, the Gutenberg-Richter law describes how earthquake frequency depends on magnitude. Plot the logarithm of the number of events against magnitude and you get a straight line over many orders of magnitude.

Gutenberg-Richter relation: log₁₀ N = a − b·M where N = number of earthquakes with magnitude ≥ M a = productivity (total seismicity of the region) b = slope, typically ≈ 1.0 for tectonic regions With b ≈ 1: each unit drop in magnitude → ~10× more earthquakes. (Roughly: one M7 for every ten M6 for every hundred M5...)

This power-law distribution is the statistical signature of a system in self-organised criticality — there is no characteristic earthquake size, just a smooth scaling from the smallest microquakes to the largest megathrusts. The b-value itself is informative: it tends to drop in highly stressed regions and may change subtly before large events, making it a subject of active research. Together with the Omori law (which describes how aftershock rates decay as ~1/time after a mainshock), Gutenberg-Richter underpins modern probabilistic seismic hazard assessment.

6. Earthquake Early Warning

We cannot yet predict the day an earthquake will strike, but once a rupture begins we can warn people seconds to tens of seconds before the strong shaking arrives. The physics is simple: earthquakes radiate two main body waves at different speeds.

P-waves (primary, compressional): v_P ≈ 6 km/s — fast, weak shaking, arrive FIRST S-waves (secondary, shear): v_S ≈ 3.5 km/s — slower, strong, DAMAGING shaking Warning lead time at distance d: Δt ≈ d · (1/v_S − 1/v_P) Plus: electromagnetic alerts travel at light speed — far faster than any seismic wave.

An earthquake early warning system — such as Japan's nationwide network or the ShakeAlert system on the US West Coast — detects the fast, harmless P-wave at stations near the epicentre, rapidly estimates the location and magnitude, and broadcasts an alert that races ahead of the slower, destructive S-wave. Even ten or twenty seconds of warning is enough to stop high-speed trains, halt surgeries, open elevator doors, shut gas valves, and let people take cover. The warning works precisely because data travels at the speed of light while the damaging shaking crawls along at a few kilometres per second — a rare case where the laws of physics give us a head start over a natural disaster.

🌐
Earthquake Fault Simulator
Load a fault with tectonic strain and trigger stick-slip rupture and elastic rebound
📈
Seismic Wave Simulator
Watch P-waves and S-waves propagate and see how early warning buys time
🗺️
Plate Tectonics Simulator
Explore the plate boundaries that load the faults where earthquakes are born