[Nek5000-users] turbJet problem
nek5000-users at lists.mcs.anl.gov
nek5000-users at lists.mcs.anl.gov
Mon Dec 9 17:23:45 CST 2013
Hi Barak,
Your err-run.log clearly shows CFL number increases considerably. So, it is
indeed a problem of time-stepping stability. So, as you suggest a variable
time-stepping might be necessary. you could even write a small routine in
.usr file to control your timestep DT. If you decrease your error of
divergence, it has nothing to do with simulation time. You can control
simulation time with NSTEPS of .rea file. However for explicit method the
region of stability of higher accuracy methods are smaller. So, 3rd order
accurate method is a little less stable. You could also try changing TORDER
= 2 , when your jet develops considerably.
I heard lagrangian formuation could be used in nek5000, but unfortunately I
do not know that. It will be nice if you post this question of lagrangian
part in an open format, so that the best person in the knowledge of this
may answer you.
Thanks,
Tanmoy
On Mon, Dec 9, 2013 at 3:59 PM, <nek5000-users at lists.mcs.anl.gov> wrote:
> Hi Tanmoy,
> It is the 4th time that the program stops (probably because of two big
> dt). See the end of the log of the run. The time step is dt=1e-3. The first
> test it was 0.02. It seems that the time step has to be reduced when the
> jet develops considerably.
> Perhaps, the best thing to do is to simulate till the time limit when the
> time step becomes not enough small, to stop, and then restart with the next
> time step (decreased) till the next stage and so on.
> Now I have further questions:
> 1. I noticed that there is a parameter to control the error of divergence.
> If I choose smaller value, how it ncrease the simulation time?
> 2. I am interested in the Lagrangian properties of the flow, i.e.
> calculating velocity, vorticity or other velocity derivatives at the
> Lagrangian orbit. Is it possible and which are the available tools?
>
> Barak
>
>
>
> On 12/09/2013 02:56 AM, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Hi Barak,
>>
>> I am not aware of variable time stepping in nek5000. The time stepping
>> used is BDFk/EXTk with CFL constraint for stability. You can change
>> TORDER in .rea file to 2 or 3 to make the time stepping 2nd or third
>> order accurate. There is a characteristic based time-stepping though...
>> IFCHAR in .rea file.
>>
>> However, if for a value ot DT you are getting convergence, essentially
>> you should get the same for smaller value of DT. What is the error you
>> are getting after the program aborts after 1200 iterations? It is still
>> stored in your log file I suppose.
>>
>> Thanks,
>> Tanmoy
>>
>>
>> On Sun, Dec 8, 2013 at 12:26 PM, <nek5000-users at lists.mcs.anl.gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
>>
>> Hi Tanmoy,
>> I have divided by 4 and again my run stopped at iteration 1200. My
>> question is: does nek5000 have variable time-stepping?
>> Thanks,
>> Barak
>>
>>
>> On 12/07/2013 08:37 PM, nek5000-users at lists.mcs.anl.__gov
>>
>> <mailto:nek5000-users at lists.mcs.anl.gov> wrote:
>>
>> Hi Barak,
>>
>> it is because of the fact that the generated file format is
>> probably not
>> readable in visit. In the *usrchk* subroutine in your .usr file
>> write
>>
>> these 2 lines
>> param(66) = 4
>> param(67) = 4
>>
>> before the end of the routine
>>
>> Then compile and execute your program. Now you will get output
>> files in
>> the form of turbJet.fld001, turbJet.fld002, ...and so on.
>> type in visnek turbJet in your terminal, and generate
>> *turbJet.nek5000*.
>>
>> This .nek5000 file can be read by visit.
>>
>> Thanks,
>> Tanmoy
>>
>>
>> On Sat, Dec 7, 2013 at 10:33 AM,
>> <nek5000-users at lists.mcs.anl.__gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>
>> <mailto:nek5000-users at lists.__mcs.anl.gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>>> wrote:
>>
>> Hi Tanmoy,
>> Many thanks. It seems that after dividing by 4 (dt=0.005)
>> it runs
>> and I get output files in the form jet0.fxxxxx .
>> Now I am trying to visualize them using the precompiled
>> ubuntu11
>> version of visit. Unfortunately, I get the following error:
>>
>> VisIt could not read from the file
>>
>> "/home/galanti/COMPUTATION/____NEK5000/nek5_svn/TEST-____
>> examples/turbJet/jet0.f*
>>
>>
>> database".
>>
>> The generated error message was:
>>
>> There was an error opening
>>
>> /home/galanti/COMPUTATION/____NEK5000/nek5_svn/TEST-____
>> examples/turbJet/jet0.f*
>>
>>
>> database. It may be an invalid file. VisIt tried using the
>> following file format readers to open the file: Nek5000
>>
>> The following error(s) may be helpful in identifying the
>> problem:
>> When attempting to use "Nek", the file matched the general
>> format
>> expected by the reader but eventually encountered this
>> error: "A tag
>> called filetemplate: must be specified".
>>
>> What's wrong?
>> Many thanks again,
>> Barak
>>
>>
>>
>> On 12/07/2013 12:42 AM, nek5000-users at lists.mcs.anl.____gov
>>
>>
>> <mailto:nek5000-users at lists.__mcs.anl.gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
>>
>> Hi Barak,
>>
>> It is a problem with numerical Stability. Decrease DT
>> in your
>> .rea file
>> until the simulations become stable. You could start
>> with 1/2 DT to
>> check if simulation is stable, other wise go for 1/4
>> DT, and
>> check with
>> this simple test.
>>
>> Thanks,
>> Tanmoy
>>
>>
>> On Fri, Dec 6, 2013 at 3:31 PM,
>> <nek5000-users at lists.mcs.anl.____gov
>> <mailto:nek5000-users at lists.__mcs.anl.gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>>
>> <mailto:nek5000-users at lists.
>> <mailto:nek5000-users at lists.>____mcs.anl.gov <http://mcs.anl.gov>
>>
>>
>> <mailto:nek5000-users at lists.__mcs.anl.gov
>> <mailto:nek5000-users at lists.mcs.anl.gov>>>> wrote:
>>
>> Hello,
>> I am testing the example turbJet, and it seems
>> that I have some
>> problem with output. This are the output files:
>> jet0.f00001
>>
>> Moreover, the ran stops with emergency exit at
>> time 5.64
>> (CFL too
>> large?):
>>
>>
>> 282 DNORM, DIVEX 6.2789891850172769E-004
>> 8.4967723454134898E-006
>> 282 5.6400E+00 1.1065E+01 Fluid done
>> 282 5.6400000E+00 1.1955953E-03 6.4269380E-04
>> 9.9999991E-01 e2
>> t max/min: 0.0000000000000000
>> 0.0000000000000000
>> CFL, Ctarg! 22.504301079496056
>> 2.0000000000000000
>> call outfld: ifpsco: T
>>
>> 283 5.6400E+00 Write checkpoint:
>> 0 283 OPEN: jet0.f00001
>>
>> 283 5.6400E+00 done :: Write checkpoint
>> file size =
>> 55. MB
>> avg data-throughput =
>> 572.1MB/s
>> io-nodes = 1
>>
>>
>> 7 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 3 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 5 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 0 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 6 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 4 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 2 Emergency exit: 283 time =
>> 5.6399999999999677
>> Latest solution and data are dumped for
>> post-processing.
>> *** STOP ***
>> 1 Emergency exit: 283 time =
>> 5.6399999999999677
>>
>>
>>
>>
>> Thank you,
>> Barak
>>
>>
>> ______________________________
>> _______________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.______gov
>> <mailto:Nek5000-users at lists.
>> <mailto:Nek5000-users at lists.>____mcs.anl.gov <http://mcs.anl.gov>
>> <mailto:Nek5000-users at lists.__mcs.anl.gov
>> <mailto:Nek5000-users at lists.mcs.anl.gov>>>
>> https://lists.mcs.anl.gov/______mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/____mailman/listinfo/nek5000-users>
>>
>> <https://lists.mcs.anl.gov/____mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users>__>
>>
>>
>> <https://lists.mcs.anl.gov/____mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users>
>>
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>__>__>
>>
>>
>>
>>
>>
>>
>> ___________________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.____gov
>> <mailto:Nek5000-users at lists.__mcs.anl.gov
>> <mailto:Nek5000-users at lists.mcs.anl.gov>>
>> https://lists.mcs.anl.gov/____mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users>
>>
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>__>
>>
>>
>> ___________________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.____gov
>> <mailto:Nek5000-users at lists.__mcs.anl.gov
>> <mailto:Nek5000-users at lists.mcs.anl.gov>>
>> https://lists.mcs.anl.gov/____mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users>
>> <https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>__>
>>
>>
>>
>>
>> _________________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.__gov
>> <mailto:Nek5000-users at lists.mcs.anl.gov>
>> https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>
>>
>> _________________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.__gov
>> <mailto:Nek5000-users at lists.mcs.anl.gov>
>> https://lists.mcs.anl.gov/__mailman/listinfo/nek5000-users
>> <https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>
>>
>>
>>
>> _______________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.gov
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>>
>>
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20131209/779d0564/attachment-0001.html>
More information about the Nek5000-users
mailing list