[ET Trac] #2621: keep running into nan ,need help

Roland Haas trac-noreply at einsteintoolkit.org
Fri Jul 22 07:23:18 CDT 2022


#2621: keep running into nan ,need help

 Reporter: Artectek
   Status: new
Milestone: ET_2022_05
  Version: ET_2022_05
     Type: bug
 Priority: minor
Component: EinsteinToolkit thorn

Comment (by Roland Haas):

Two items were brought up:

1. the resolution may well be too low. The gallery example \(which your parfile dns.par seems to be base on\) sets the resolution to a very coarse value, namely 18Msun on the coarsest level and 18Msun/2^6 = 0.28 for the level that covers the star\(s\) initially. This corresponds \(using Msun ~~ 1.48km\) to a resolution of 416m which is too coarse for any serious simulation. It should be no more than about 280m so all resolutions should be scaled to dx=dy=dz=12
2. Looking at your log output then at t=0 there is: “INFO \(CarpetTracker\): Setting position of refined region #1 from surface #0 to \(13.5,0,0\)” indicating that the star was found with center at x=13.5 however you initial grid structure for the refined regions \(the CarpetRegrid2::position\_x parameters\) centers the boxes around x=22.5 which is not correct \(probably not the cause for the failure at the end but incorrect and will lead to an initially incorrect grid structure\). You can open eg 2D output files \(hydrobase-rho.xy.h5\) in VisIt and add a “Mesh” plot to check what things look like at iteration=0.

‌

--
Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2621/keep-running-into-nan-need-help
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.einsteintoolkit.org/pipermail/trac/attachments/20220722/be9be8e7/attachment.html 


More information about the Trac mailing list