Binary Black Hole Evolutions with Mesh Refinement

The problem of modelling black hole mergers is one of many length scales, from the tiny distances within the horizons and around the singularities, to the far away regions where the spacetime is essentially flat with small perturbations. Modelling such systems accurately requires both high resolution and very distant boundaries, resulting in very large numerical grids and computationally expensive simulations.

The last year, the numerical relativity group has begun making effective use of mesh refinement methods in their simulations. These methods place high resolution only in the places where they are needed, resulting in great savings in computational cost. A demonstration that the technical challenge of implementing such methods is reliably possible was presented in a recent paper by Schnetter, Hawke, and Hawley [1]. Since then, the methods have been carried over to the binary black hole and neutron star case. Simulations which used to run on an entire supercomputer can now be run on only a few nodes with similar accuracy.

As a result we are able to perform more simulations out faster, making experiments with initial data sets and parameter studies possible. That removes an important barrier to the advancement of black hole simulations and opens up many new avenues for future work.

Bibliography

[1] E. Schnetter, S. H. Hawley, I. Hawke, Evolutions in 3D numerical relativity using fixed mesh refinement, Class. Quantum Grav. 21, 1465-1488, 2004
Eprint: gr-qc/0310042