Fault friction

Thus, the behaviour of massive earthquakes is dependent on the properties of single molecular irregularities or asperities.

If this water is removed, by extreme drying, the rock minerals do not behave at all as expected:[2] they exhibit no fault healing or dynamic friction.

For a fault being stressed to the point of an earthquake, these bonds begin to stretch and break.

Once the critical distance has been achieved, there is a significant strength loss, and the fault begins to slide.

It could be that the earthquake "skids" are greased by silica gel,[4] the water acts as a standard bearing lubricant, or that there is a "lift and separate" mechanism at work.

This means that between the individual grains of the rock, there are small pores which can be filled with a gas (usually air) or a fluid.

The rock type along a fault can have a large effect on the amount of frictional resistance present.

Laboratory experiments have proved that the presence of water will promote the rupture of a fault in carbonate rocks (marble).

[5] However, these experiments also showed that in silica-bearing rock types (microgabbro), the presence of water may delay or even inhibit the rupture of a fault.

[5] In other words, the microscopic grain contacts which hold the fault in place instantly melt due to high stresses.

The presence of water delays this "flash melting" basically by cooling the contacts, and keeping them in solid form.

This molten rock (frictional melt), can then expand and work its way into the pores and imperfections on the fault surface.

[8] Fault ruptures generate massive amounts of heat, which usually result in frictional melting.

These pseudotachylites can form at pressures at or above roughly 0.7 GPa, which equates to deep crustal faulting.