# Dynamic rupture

An important domain which is currently taking off is the problem of simulating 3D dynamic rupture and the associated wave propagation and their interaction. In principle, faults can be defined inside a 3D grid (e.g. using finite differences) and appropriate boundary conditions and frictional behaviour are implemented at the fault. In the absence of possibilities to carry out laboratory rupture experiments under appropriate crustal conditions the phenomenological studies using numerical modelling are essential for progress. As is the case for the other domains there are various approaches to the problem of dynamic rupture propagation, notably methods which discretise only the fault itself (e.g. boundary integral elements) or methods based on 3D grids (e.g. finite differences) and fault planes defined within them. As complete 3D calculations are only beginning to be possible in interesting frequency domains there is an urgent need to define benchmarks and define domains of applications of the various approaches.

There are many open questions on the more technical side, e.g. (1) how can we incorporate **geometrical complexity** of the fault plane; (2) how can we accurately model rupture including the **free surface**; (3) what is the optimal implementation of **fault boundary conditions**? Recent results have shown that important insight can be gained with these techniques. An example is the discovery of the importance of rupture at material discontinuities which are likely to exist in regions with large deformation rates (e.g. San Andreas Fault, North Anatolian Fault, etc.). Breakthroughs in the understanding of earthquake processes may be possible in the near future with further extensions and comparative studies proposed in this project.