Unsteady simulation with steady solution as initial value
 

Dear all

I try to simulate an unsteady case with converged steady solution as initial value. So What I did is copy the steady case as a new unsteady case, and change corresponding parameters for unsteady calculations.
In the controlDict file, I set it like this:
------------------
startFrom latestTime;

startTime 0;

stopAt endTime;

endTime .5;

deltaT 1e-4;
-----------------------



When I run pimpleFoam, it says:


-----------------------
Create time

Create mesh for time = 10000

Reading field p

Reading field U

Reading/calculating face flux field phi

AMI: Creating addressing and weights between 3724 source faces and 3724 target faces
AMI: Patch source weights min/max/average = 1.00025, 1.00025, 1.00025
AMI: Patch target weights min/max/average = 1.00025, 1.00025, 1.00025
Selecting incompressible transport model Newtonian
Selecting turbulence model type RASModel
Selecting RAS turbulence model kOmegaSST
kOmegaSSTCoeffs
{
alphaK1 0.85034;
alphaK2 1;
alphaOmega1 0.5;
alphaOmega2 0.85616;
gamma1 0.5532;
gamma2 0.4403;
beta1 0.075;
beta2 0.0828;
betaStar 0.09;
a1 0.31;
b1 1;
c1 10;
F3 false;
}

No finite volume options present


PIMPLE: no residual control data found. Calculations will employ 2 corrector loops


Starting time loop

End
-------------

No time loop happens, any guys have an idea???

Best regards,
Ye

Hi,

your endTime is 0.5 s, but

your start time is 10000.

Why don't you use mapFields -sourceTime 10000 <steady-state-case-directory> for mapping steady state solution onto your transient simulation initial time?

I forgot to mention, that mapField does not work for me. It gives a lot warning and get stopped. I also have a thread about that.

Dear Alexey

Thank you for you reply. I got my steady state solution for 10000 iterative steps. And I both tried to set 'startTime 10000 or startTime 0', 'endTime 10000.5 or 0.5', since I want to have 0.5s transient simulation, but they do give the same output as I list.

The time loop just does not start. I do not know why...

Best regards,
Ye

Hi,

by default mapFields uses

maybe you can try other methods: direct or mapNearest and they will work? I.e. add -method option to mapFields invocation.

Also if you just copy results from steady state solution, there should be uniform subfolder (i.e. 10000/uniform) where OF stores information on time and time step. Maybe if you delete this subfolder it will do the loop?

Because this is quite redundant if the mesh and the boundary conditions are the same. Just do something like

$ mv 0 0.org
$ cp -r ../steadyStateCase/0 .
$ rm -r 0/uniform

Dear all,

After deleting the file 10000/unform, it works well.

Thank you !

Ye

Hi Ye and all,
I tried to the method you mentioned, and it works for serial running, but I can't run in parallel, the error:

[dang-SVE14A2SGC:17212] [12] /opt/openfoam231/platforms/linux64GccDPOpt/lib/libfiniteVolume.so(_ZNK4Foam16cyclicAMIFvPatch11ma keWeightsERNS_5FieldIdEE+0x1d5) [0x7ff2692f1085]
[dang-SVE14A2SGC:17212] [13] /opt/openfoam231/platforms/linux64GccDPOpt/lib/libfiniteVolume.so(_ZNK4Foam20surfaceInterpolation 11makeWeightsEv+0x27e) [0x7ff2697889ce]
[dang-SVE14A2SGC:17212] [14] /opt/openfoam231/platforms/linux64GccDPOpt/lib/libfiniteVolume.so(_ZNK4Foam20surfaceInterpolation 7weightsEv+0x19) [0x7ff269788be9]
[dang-SVE14A2SGC:17212] [15] pisoFoam() [0x41a35b]
[dang-SVE14A2SGC:17212] [16] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0) [0x7ff266f28a40]
[dang-SVE14A2SGC:17212] [17] pisoFoam() [0x41c82b]
[dang-SVE14A2SGC:17212] *** End of error message ***
--------------------------------------------------------------------------
mpirun noticed that process rank 1 with PID 17210 on node dang-SVE14A2SGC exited on signal 8 (Floating point exception).
--------------------------------------------------------------------------

How do you solve in parallel?
Thanks a lot!
Sophie