A Free Simulator Software for Vacuum Field BVPs | Tom

[TGlad]

Hi, I am interested in simulating the vacuum field equations, but solving a full boundary value problem rather than the initial value problem. i.e. I might have boundary conditions in all spatial and temporal extents/extremes, rather than just an initial 3D surface.

Does anyone know any free software that could perform such a simulation? I have started looking at the Einstein Toolkit docs and it doesn't really look like it does this, and I don't have much info on GRCombo.

many thanks,

Tom.

 

[Paul Colby]

I might have boundary conditions in all spatial and temporal extents/extremes, rather than just an initial 3D surface.

Well, I have no experience with numerical relativity, however, specifying all boundary conditions on all extremes need not be consistent, right? Doesn't sound like a well posed problem numerical issues aside.

 

[TGlad]

It seems quite likely I would be over-constraining the vacuum equations, in which case I would need to provide fewer constraints, or allow for the solution closest to a vacuum rather than pure vacuum. But basically I'm after a solver that doesn't solve the usual initial value problem forward in time, but instead solves a boundary value problem for the whole 3+1 block.
It is totally OK if no-one has any recommendations, that is in itself helpful information.

 

[PeterDonis]

It seems quite likely I would be over-constraining the vacuum equations, in which case I would need to provide fewer constraints

I'm after a solver that doesn't solve the usual initial value problem forward in time, but instead solves a boundary value problem for the whole 3+1 block.

The point that @Paul Colby was suggesting (which I think is valid) is that any attempt to go beyond providing initial value conditions will be over-constraining the equations. Knowing the initial values is already sufficient to constrain the boundary value problem for the whole 3+1 block.