Hi Taiko,


I will (partly) skip the first part of your message and let others reply on that, as I don't have experience with NS simulations.


Regarding the gravitational wave extraction, I think you are accessing the data correctly (your example gives the time series for one mode at one radius), and there should be more than one element in general. From your parameter file, it seems the output should go to an ascii file (it's the default), so you can double check in that file too.

However, since you have `Multipole::out_every = 128`, does your simulation reach iteration 128 and later? In particular, because you're getting nans at iteration 0, does the simulation actually go past t=0? Given what you're saying, I guess your single element of `SimDir.gws[100][(2,2)].x` is 0 (i.e t=0).


Best,


Jordan



From: Users <users-bounces@einsteintoolkit.org> on behalf of MAKINO Taiki via Users <users@einsteintoolkit.org>
Sent: Thursday, July 2, 2026 12:39:00 AM
To: users@einsteintoolkit.org
Subject: [Users] Question about IllinoisGRMHD C2P failure and kuibit Psi4 extraction
 

CUIDADO: Email de um sistema externo. Cuidado com links, anexos e pedidos de dados/senhas.
CAUTION: Email from an external system. Be careful with links, attachments, and requests for data/passwords.
Dear Einstein Toolkit Users,

I am a new Einstein Toolkit user, and I have some questions about a BNS simulation.

I am trying to run a BNS simulation using bns.par, which I downloaded from Einstein Toolkit Gallery (https://einsteintoolkit.org/gallery/bns/index.html) and slightly modified.

I can successfully create and submit the simulation but the following message appears during the simulation:

INFO (IllinoisGRMHD): C2P: Iter. # 0, Lev: 5 NumPts= 132300 | Backups: 0 0 0 | Fixes: VL= 0 rho*= 0 Averaged pts = 38250 Font1D 38250 | Failures: 2250 InHoriz= 0 / 0 | 1.00 iters/gridpt Error, Sum: rho 6.378e-08, 2.433e-05 | tau nan, -nan | Sx nan, nan | Sy nan, nan | Sz nan, nan

To investigate this problem, I added the following parameters:
Multipole::interpolator_name = "Lagrange polynomial interpolation"
Multipole::interpolator_pars = "order=4"

I also changed the parameter Driver::ghost_size  from 3 to 4, but the same error occurred.

I would like to ask:
  • What could cause tau, Sx, Sy, and Sz to become NaN in this situation?
  • Does this indicate a failure of  C2P conversion?
  • Are there any parameter settings or initial data that I should check first?

Additionally, I have another question. When I use kuibit to extract Psi4, I expected to obtain the full time series (or all available modes) but the returned object contains only one element, for example, 
      SimDir.gws[100][(2,2)].y=[-9.79586237e-15-5.27476073e-14j].

Other attributes such as
       SimDir.gws[100][(2,2)].x 

also contain only one element. Is this a correct behavior or could it indicate a problem with my .par file? 

For reference, I have attached my parameter file.

Thank you very much for your time and any suggestions.

Best regards,

Taiki Makino

Graduate School of Science, Nagoya University

makino.taiki.g6@s.mail.nagoya-u.ac.jp


取得 Outlook for Mac