[FLASH-USERS] 2D Spherical and Cartesian Geometry
Michiel Bustraan
michiel.bustraan at astro.su.se
Tue Sep 27 04:04:54 EDT 2016
Klaus,
I'm sorry for being unclear earlier.
When I made the changes to the unsplit hydro you suggested, it did not seem to make a difference.
Changing from 0-180 to 0-90 just caused those issues to happen around 90 degrees. I think the issue lay more in the structure of the grid or the use of hydro solver than the actual boundaries.
I ran a simulation overnight that seemed to be working when I left, with split hydro. It produced a few plotfiles that show reasonable results.
However, when coming in about an hour ago, I saw that it crashed overnight with the same error message about missing neighbors.
I'm attaching a logfile from an identical simulation I just ran, with no changes apart from increasing the number of AMR levels and changing the y boundaries from 0-90 to 0-180.
The AMR levels I changed to make it crash sooner so I could provide a clean logfile as soon as possible, and the boundaries because I first thought that might help.
The original simulation, which ran at least a couple of hours before crashing, had only 3 refine levels.
In the terminal, the error message is as follows:
refined: total blocks = 24596
Missing neighbors on PE 3, lb= 859,surrblks=( 843, 3, 1)( 853, 3, 2)( 858, 3, 1)( 845, 3, 1)( 859, 3, 2)( 864, 3, 1)( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)
Missing neighbors on PE 7, lb= 2989,surrblks=( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)(2907, 7, 1)(2989, 7, 2)(2994, 7, 1)(2909, 7, 1)(2995, 7, 2)(3000, 7, 1)
Missing neighbors on PE 0, lb= 1424,surrblks=(1409, 0, 2)(1414, 0, 2)(1430, 0, 2)(1419, 0, 2)(1424, 0, 2)(1440, 0, 2)(1451, 0, 1)(1452, 0, 1)( -1, -1,-1)
Missing neighbors on PE 0, lb= 1440,surrblks=(1414, 0, 2)(1430, 0, 2)(1435, 0, 2)(1424, 0, 2)(1440, 0, 2)(1445, 0, 2)(1452, 0, 1)( -1, -1,-1)( -1, -1,-1)
Missing neighbors on PE 0, lb= 1445,surrblks=(1430, 0, 2)(1435, 0, 2)(1462, 0, 1)(1440, 0, 2)(1445, 0, 2)(1464, 0, 1)( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)
Missing neighbors on PE 6, lb= 2936,surrblks=(2927, 6, 1)(2928, 6, 1)( -1, -1,-1)(2931, 6, 2)(2936, 6, 2)(2952, 6, 2)(2941, 6, 2)(2946, 6, 2)(2962, 6, 2)
Missing neighbors on PE 6, lb= 2952,surrblks=(2928, 6, 1)( -1, -1,-1)( -1, -1,-1)(2936, 6, 2)(2952, 6, 2)(2957, 6, 2)(2946, 6, 2)(2962, 6, 2)(2967, 6, 2)
Missing neighbors on PE 6, lb= 2957,surrblks=( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)(2952, 6, 2)(2957, 6, 2)(2976, 6, 1)(2962, 6, 2)(2967, 6, 2)(2978, 6, 1)
Missing neighbors on PE 3, lb= 859,surrblks=( 843, 3, 1)( 853, 3, 2)( 858, 3, 1)( 845, 3, 1)( 859, 3, 2)( 864, 3, 1)( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)
Missing neighbors on PE 6, lb= 2936,surrblks=(2927, 6, 1)(2928, 6, 1)( -1, -1,-1)(2931, 6, 2)(2936, 6, 2)(2952, 6, 2)(2941, 6, 2)(2946, 6, 2)(2962, 6, 2)
Missing neighbors on PE 6, lb= 2952,surrblks=(2928, 6, 1)( -1, -1,-1)( -1, -1,-1)(2936, 6, 2)(2952, 6, 2)(2957, 6, 2)(2946, 6, 2)(2962, 6, 2)(2967, 6, 2)
Missing neighbors on PE 6, lb= 2957,surrblks=( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)(2952, 6, 2)(2957, 6, 2)(2976, 6, 1)(2962, 6, 2)(2967, 6, 2)(2978, 6, 1)
Missing neighbors on PE 7, lb= 2989,surrblks=( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)(2907, 7, 1)(2989, 7, 2)(2994, 7, 1)(2909, 7, 1)(2995, 7, 2)(3000, 7, 1)
Missing neighbors on PE 0, lb= 1424,surrblks=(1409, 0, 2)(1414, 0, 2)(1430, 0, 2)(1419, 0, 2)(1424, 0, 2)(1440, 0, 2)(1451, 0, 1)(1452, 0, 1)( -1, -1,-1)
Missing neighbors on PE 0, lb= 1440,surrblks=(1414, 0, 2)(1430, 0, 2)(1435, 0, 2)(1424, 0, 2)(1440, 0, 2)(1445, 0, 2)(1452, 0, 1)( -1, -1,-1)( -1, -1,-1)
Missing neighbors on PE 0, lb= 1445,surrblks=(1430, 0, 2)(1435, 0, 2)(1462, 0, 1)(1440, 0, 2)(1445, 0, 2)(1464, 0, 1)( -1, -1,-1)( -1, -1,-1)( -1, -1,-1)
*** Wrote plotfile to rad_blast_wave_castro_hires2_forced_hdf5_plt_cnt_0000 ****
DRIVER_ABORT: missing neighbor!
DRIVER_ABORT: missing neighbor!
I've also attached the flash.par file and the Config file used during setup. I based my simulation on a different existing test problem, so a lot of it has been commented out.
Regarding the EOS, what I meant was that I was running previous simulations (in 1D spherical and then in 2D cartesian) by defining eosMode, eosModeInit, and hy_eosmodeafter with some variation of dens_temp (some of which is still there in the comments of flash.par). When I ran simulation with the split hydro in 2D spherical, the code remarked that that wasn't an appropriate choice of EOS, and would only run if I selected dens_ei. I was curious what might have caused this.
I hope this information is more helpful, and I'm sorry for the lack of information in my last message.
- Michiel
________________________________________
From: Klaus Weide <klaus at flash.uchicago.edu>
Sent: Monday, September 26, 2016 11:54 PM
To: Michiel Bustraan
Cc: flash-users at flash.uchicago.edu
Subject: Re: [FLASH-USERS] 2D Spherical and Cartesian Geometry
On Mon, 26 Sep 2016, Michiel Bustraan wrote:
> Klaus,
> Those changes didn't seem to help,
Michiel,
Surely something must have changed at least upon changing the angle
coordinate range from 0-180 to 0-90 ? At least, a plot corresponding to
the one you sent earlier should look different, since you would be
eliminating the part of the domain that appears to have the largest
problem!
> but switching to split hydro did seem to make the simulation work.
Thanks for checking...
> However, when I tried to increase the resolution, either through more AMR levels or a higher base resolution, I get the following error message.
> Missing neighbors on PE 1, lb= 243,surrblks=( 232, 1, 2)( 237, 1, 1)( 249, 1, 1)( 238, 1, 2)( 243, 1, 2)( 251, 1, 1)( 254, 1, 1)( 255, 1, 1)( -1, -1,-1)
> Do you know what might cause this?
When does this happen? Did you just get this message, or were there
additional messages either on standard output or in the log file? Did
FLASH abort immediately after this or not?
The underlying PARAMESH Grid implementation seems to be confused, having
inconsistent information about neighboring blocks. It is hard to say more
without more info from you.
For example, which physics units other than Hydro have you compiled in?
(E.g., Gravity?) What does the Grid look like when this happens?
> Also, I was unable to run the simulation with the 'dens_temp' EOS. Is
> split hydro only able to run with 'dens_ie', or is the temperature EOS
> supposed to work?
FLASH has several Eos implementations (corresponding to, roughly,
different source directories under physics/Eos/EosMain). You haven't said
which one you are using. The setup command line is not enough to tell us,
since the set of included units and implementations also depends on the
REQUESTS and REQUIRES directives in your simulation's Config file and
other Config files.
There are many calls of Eos() and Eos_wrapped() (and Eos_guardCells)
throughout the code, some with a hardwired mode like MODE_DENS_EI, others
depending on some runtime parameter. There are several runtime parameters
that specify an Eos mode (like eosMode, eosModeInit, hy_eosModeGc,
hy_eosModeAfter). So it is not clear what it means to "run the simulation
with the 'dens_temp' EOS".
Sending a full log file would make it much clearer what you are doing.
Also standard output with warnings and messages, parfile, Config file
and any modified source files form your Simulation directory.
Klaus
-------------- next part --------------
A non-text attachment was scrubbed...
Name: logfile.log
Type: text/x-log
Size: 39253 bytes
Desc: logfile.log
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160927/89028487/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: flash.par
Type: application/octet-stream
Size: 11794 bytes
Desc: flash.par
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160927/89028487/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Config
Type: application/octet-stream
Size: 4236 bytes
Desc: Config
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160927/89028487/attachment-0001.obj>
More information about the flash-users
mailing list