[Users] Synchronisation and boundary prolongation in NSNS gallery example with slypp EOS

Tue Mar 12 12:55:55 CDT 2019

Wilke

If you use symmetry_rotating_180, then there cannot be any refinement
boundaries that are too close to the symmetry boundary, hence there will be
slightly larger fine grids.

-erik

On Tue, Mar 12, 2019 at 1:46 PM Wilke van der Schee <
wilke.van.der.schee at cern.ch> wrote:

> Dear Erik,
>
> Thanks for the reply, such might indeed be the case. We use the same
> settings as in the gallery, i.e. a gridsize of 396 with dx=18 and then 7
> levels of refinement. Could the symmetry also be of influence here?
> (both reflection_z and symmetry_rotating180 are used). Looking at the grid
> (it was attached as visit0000.png) it looks quite fine to me, but I don't
> have that much experience.
>
> Thanks again, best,
>
> Wilke
>
> On Tue, 12 Mar 2019 at 18:38, Erik Schnetter <schnetter at cct.lsu.edu>
> wrote:
>
>> On Tue, Mar 12, 2019 at 11:26 AM Wilke van der Schee <
>> wilke.van.der.schee at cern.ch> wrote:
>>
>>> Dear users,
>>>
>>> We have been trying for a while to get neutron star mergers for
>>> different equations of state, with only partial success. While the gallery
>>> example with the polytrope runs fine, we already encounter problems with
>>> the SLy_tabulated EOS. (we're using the current release, ET_2018_09)
>>>
>>> One particularly puzzling error we don't understand is the following
>>> (full log+par file attached)
>>>
>>> "INFO (NSTracker): Found star at (-10.125,-5.90625,0)
>>>     21256   373.641 |   52.3035497 |    0.0001505    0.0053630 |
>>> 0.0014340 |    1.0717264 |    0.0019186 |      3591
>>> INFO (CarpetRegrid2): Enforcing grid structure properties, iteration 0
>>> INFO (CarpetRegrid2): Enforcing grid structure properties, iteration 1
>>> WARNING level 1 from host n0086.compute.hpc process 0
>>>   while executing schedule bin (none), routine (no thorn)::(no routine)
>>>   in thorn CarpetLib, file
>>> /hpc/local/CentOS7/uu_itp_strings/Cactus/configs/sim/build/CarpetLib/dh.cc:161:
>>>   ->
>>>
>>> /hpc/local/CentOS7/uu_itp_strings/Cactus/configs/sim/build/CarpetLib/dh.cc:980:
>>>    [ml=0 rl=6 c=0] The following grid structure consistency check failed:
>>>    Synchronisation and boundary prolongation: All points must have been
>>> received
>>>    needrecv.empty()
>>>
>>
>> This is an internal error in Carpet. Most of the time, this indicates
>> another error that wasn't caught properly before the details of the grid
>> structure are generated. What resolution are you using? If you are using a
>> resolution that is too coarse, then the refined regions don't fit into the
>> simulation domain, and this could lead to such an error.
>>
>> -erik
>>
>>
>>
>> ml=0 rl=6
>>> baseextent=([756,504,756]:[6408,11784,6408]:[4,4,4]/[189,126,189]:[1602,2946,1602]/[1414,2821,1414]/5640296116)"
>>> (shortened log attached, full log here:
>>> https://www.dropbox.com/s/avgrshl8chhzp6i/atmostr025full.out?dl=0)
>>>
>>> and then much more information about the grid structure (see full log).
>>> Indeed when looking into visit at this time the grid structure is a bit
>>> strange (also attached, at t=373.5, just before the crash). It containts
>>> the 3rd central mesh already, which is not really supposed to be triggered,
>>> but on the other hand that by itself should not lead to this crash (and I'm
>>> not quite sure why the trigger was activated at t=180).
>>>
>>> The main difference of the parameter file wrt the gallery is of course
>>> the equation of state and hence the initial Lorene file. We also needed to
>>> go to dtfac=0.25 (otherwise even earlier con2prim errors appear), and for
>>> testing and stability purposes we now use an atmosphere density of 10^{-9}.
>>>
>>> Any ideas how to look for more information or on similar problems would
>>> be very welcome, we do have more output files of course. Looking at the
>>> source code didn't help me much for now.
>>>
>>> Many thanks in advance for suggestions,
>>>
>>> Wilke van der Schee (post-doc Utrecht University)
>>> _______________________________________________
>>> Users mailing list
>>> Users at einsteintoolkit.org
>>> http://lists.einsteintoolkit.org/mailman/listinfo/users
>>>
>>
>>
>> --
>> Erik Schnetter <schnetter at cct.lsu.edu>
>> http://www.perimeterinstitute.ca/personal/eschnetter/
>>
>>

-- 
Erik Schnetter <schnetter at cct.lsu.edu>
http://www.perimeterinstitute.ca/personal/eschnetter/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.einsteintoolkit.org/pipermail/users/attachments/20190312/6253d647/attachment.html 