Abstract: Full relativistic simulations in three dimensions invariably develop runaway modes that grow exponentially and are accompanied by violations of the Hamiltonian and momentum constraints. Recently, we introduced a numerical method (Hamiltonian relaxation) that greatly reduces the Hamiltonian constraint violation and helps improve the quality of the numerical model. We present here a method that controls the violation of the momentum constraint. The method is based on the addition of a longitudinal component to the traceless extrinsic curvature generated by a vector potential w_i, as outlined by York. The components of w_i are relaxed to solve approximately the momentum constraint equations, pushing slowly the evolution toward the space of solutions of the constraint equations. We test this method with simulations of binary neutron stars in circular orbits and show that effectively controls the growth of the aforementioned violations. We also show that a full numerical enforcement of the constraints, as opposed to the gentle correction of the momentum relaxation scheme, results in the development of instabilities that stop the runs shortly.
Recommendations
- Hamiltonian relaxation
- Why solve the Hamiltonian constraint in numerical relativity?
- The constraints as evolution equations for numerical relativity
- A remedy for constraint growth in numerical relativity: the Maxwell case
- CTTK: a new method to solve the initial data constraints in numerical relativity
Cited in
(3)
This page was built for publication: Momentum constraint relaxation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5290758)
