Abstract
Stick-slip along precut surfaces in stressed foam rubber is similar to earthquake faulting, stick-slip in rock specimens, and theoretical predictions. An additional feature is the common occurrence of multiple events. A significant amount of slip occurs as fault creep.
For simple one-slip events on a long fault, the peak particle velocities near the center of the fault average about half the value Δσβ/μ, with Δσβ/μ apparently being a good upper bound. The variation is probably due to focusing by rupture propagation. On a circular fault, the peak values average 3 to 4 times less, partly a result of a greater amount of fault creep, and probably, partly a result of focusing by rupture propagation. Total stress drops are typically about 10 to 20 per cent of the absolute stress, and during individual events about 80 to 90 per cent of the released strain energy is dissipated as friction.
For multiple events the cumulative source time function is usually much longer than the source dimension divided by the shear-wave velocity. Thus, the far-field spectrum would have the shape for fractional stress drop, but the low-frequency spectral corner would not correspond to the fault dimension; thus, the inferred stress drops would be too low. Multiple events may explain some anomalously low inferred stress drops for small earthquakes and may partly explain the success of surface-wave excitation as a method of distinguishing underground nuclear explosions from earthquakes. Foam rubber models may be used to study strong motion around various types of faults and, thus, aid in the problem of microzonation.
Transition process from nucleation to high-speed rupture propagation: scaling from stick-slip experiments tonatural earthquakes