Dear all,
I have recently changed my hydro simulations to the Refluxing strategy implemented by Erik Schnetter. For this I have changed the standard configuration to a *cell centred refinement, staggered boundaries and low order spatial prolongation for the hydro variables*. Since then I experience problems with BSSN or, more specifically, the QuasilocalMeasures thorn (a=0.9 black hole). After an initially fine behaviour, *no Eigenvector with value one is found and the spacetime gradually crashes*.
I am still puzzled where this is coming from. Does anybody have some experience with the combination spacetime evolution + hydro refluxing, i.e. the somewhat different grid setup? I would be glad for any input to see where I start to fix my work.
Thank you very much and my best from Valencia
Jens
Jens
If a horizon is not found, then I suspect insufficient resolution. You should be able to choose different prolongation operators for spacetime and hydrodynamics. Since the geometry is not refluxed, you could use a higher order operation there, while staying with linear interpolation for hydro.
-erik
On Wed, Jan 9, 2019 at 3:05 PM Jens Mahlmann jensmahl@alumni.uv.es wrote:
Dear all,
I have recently changed my hydro simulations to the Refluxing strategy implemented by Erik Schnetter. For this I have changed the standard configuration to a *cell centred refinement, staggered boundaries and low order spatial prolongation for the hydro variables*. Since then I experience problems with BSSN or, more specifically, the QuasilocalMeasures thorn (a=0.9 black hole). After an initially fine behaviour, *no Eigenvector with value one is found and the spacetime gradually crashes*.
I am still puzzled where this is coming from. Does anybody have some experience with the combination spacetime evolution + hydro refluxing, i.e. the somewhat different grid setup? I would be glad for any input to see where I start to fix my work.
Thank you very much and my best from Valencia
Jens _______________________________________________ Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
Dear Erik, thank you very much for your mail. I hope I may bother you again with an Amateur question - I am surely misunderstanding something. The parameter Carpet::prolongation_order_space gives the interpolation order for the reconstruction of coarse grid points from the finely refined grids. In the par-examples of Refluxing I find Carpet::prolongation_order_space set to 1 and 3 (i.e. different values). If for the refluxing the parameter Carpet::refinement_centering is set to "cell", then gridpoints should be aligned from fine to coarse grids. My question is if the prolongation_order_space makes a difference for "cell" centered refinement? This is the parameter which costs me a lot of time when changing to higher order, so I want to understand this.
Thank you very much and my best!
Jens -------- Original message --------From: Erik Schnetter schnetter@cct.lsu.edu Date: 1/10/19 01:59 (GMT+01:00) To: Jens Mahlmann jensmahl@alumni.uv.es Cc: Einstein Toolkit Users users@einsteintoolkit.org Subject: Re: [Users] Spacetime evolution with hydro refluxing Jens If a horizon is not found, then I suspect insufficient resolution. You should be able to choose different prolongation operators for spacetime and hydrodynamics. Since the geometry is not refluxed, you could use a higher order operation there, while staying with linear interpolation for hydro. -erik
On Wed, Jan 9, 2019 at 3:05 PM Jens Mahlmann jensmahl@alumni.uv.es wrote:
Dear all,
I have recently changed my hydro simulations to the Refluxing strategy implemented by Erik Schnetter. For this I have changed the standard configuration to a cell centred refinement, staggered boundaries and low order spatial prolongation for the hydro variables. Since then I experience problems with BSSN or, more specifically, the QuasilocalMeasures thorn (a=0.9 black hole). After an initially fine behaviour, no Eigenvector with value one is found and the spacetime gradually crashes.
I am still puzzled where this is coming from. Does anybody have some experience with the combination spacetime evolution + hydro refluxing, i.e. the somewhat different grid setup? I would be glad for any input to see where I start to fix my work.
Thank you very much and my best from Valencia
Jens
_______________________________________________
Users mailing list
Users@einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Jens Mahlmann -
jensflo