[vishal_s]

Hi...

I am trying to simulate jet in cross-flow using sprayFoam (water jet coming into air). I have used patchInjection and LISA atomization model. The phase change model, turbulence, and the heat transfer model are turned off. However, I am getting this error:

#0 Foam::error:rintStack(Foam::Ostream&) at ??:?
#1 Foam::sigFpe::sigHandler(int) at ??:?
#2 ? in "/lib64/libc.so.6"
#3 ? at makeBasicSprayParcelSubmodels.C:?
#4 ? at ??:?
#5 ? at ??:?
#6 ? at ??:?
#7 ? at ??:?
#8 ? at ??:?
#9 ? at ??:?
#10 ? at ??:?
#11 ? at ??:?
#12 __libc_start_main in "/lib64/libc.so.6"
#13 ? at ??:?
/var/spool/slurmd/job01562/slurm_script: line 22: 120606 Floating point exception(core dumped) sprayFoam > log

I understand it is something to do with division by zero . I have tried but unable to resolve the issue. Can anybody please tell me how to make this case work. I am attaching my case file here. I am using OF-2.4.0.

[piu58]

I just had a look at your case without running it. You u file has no real outlet. You should have at least one boundary wit zeroGradient. Else you put material in your volume baut there is no chance for it to get out.

[vishal_s]

Hi Uwe, thanks for the suggestion. I tried it out but still not working. I also tried by increasing/reducing the number of parcels per second but of no help. Can you suggest me something, I am really struck here.

I have one more doubt, which injector in OF-2.2.2 represents a blob type injector?

Thanks.

[piu58]

Your case is a rather complicated one. If anybody should help you it would be a fine idea to explain something more:
- a sketch of the setup and what to be calculated
- how are the properties of your case set? Where can one find which physical/chemical properties?
- What is the base of your case? Is there a fundamental one which you modified?
- How did you make safe, that the physical and chemical properties are correct and complete? What may be taken for safe because it is based on another well checked case?
- Which modifications did you try? Which results did they have?
- Do you write a protocol of your runs? I recommend this. Protocol a run, even if it crashes. Make sketches form the result or print them out.

[vishal_s]

Hello Uwe,
1) a sketch of the setup and what to be calculated
click here link for the diagram of my case set-up. It's a jet in cross-flow simulation, where the cross-flow fluid is air and the jet fluid is water. I am trying to calculate the velocity and the droplet diameters (Sauter diameter and mean diameter) due to primary (LISA) and secondary atomization (ReitzDiwakar) using patchinjection model. My jet volume flow rate is 10e-6m3/s.
2) how are the properties of your case set? Where can one find which physical/chemical properties?
As a first step, I am trying to run the simulation without combustion, turbulence or phase change. For the rest, U: both the inlets have fixed value, outlet zeroGradient (Changed after you suggested).
p: zeroGradient at the inlet and 1e5 at the outlet
T: both inlet have temperature of 300 k , oultet ZG
N2 and O2: 0.766 and 0.234 at the cross-flow inlet and internal field. For jet inlet: zero Gradient was used.
Ydefault: zeroGradient for everything.

The rest variables. I have given some flag value as they are not being calculated.
The physical and chemical properties are mainly copied from lagrangian/sprayFoam/aachenbomb.

The sprayCloudProperties are copied from lagrangian/sprayFoam/aachenbomb tutorial and lagrangian/reactingParcelFoam/verticalchannel tutorial.

3)What is the base of your case? Is there a fundamental one which you modified?
The base of my case is lagrangian/sprayFoam/aachenbomb tutorial. Here I have modified the injector type, primary atomization and geometry (blockMesh). Turned off combustion, heat transfer and turbulence. Also, I have changed the injection fluid from C7H16 to water.

4) How did you make safe, that the physical and chemical properties are correct and complete? What may be taken for safe because it is based on another well checked case?

I consider the physical and chemical properties are correct and complete as they work fine for the aachenbomb tutorial. Moreover, I never found any one reporting any bug in this forum regarding the physical or chemical properties.


5) Which modifications did you try? Which results did they have?
As I have mentioned in (3), I have changed the injection type, fluid, primary atomization and the numerical domain. The results obtained from the base case by running aachenbomb looks fine and correct. However, after making my changes, the simulation is blowing up.

6) Do you write a protocol of your runs? I recommend this. Protocol a run, even if it crashes. Make sketches form the result or print them out.
Yes, I did write the protocol. They are attached in the .tar.gz file attached along with my case in the first post.

If you have any more query, please let me know.

Thanks.

[piu58]

I'll have a closer look at it. Please have some patience.

[piu58]

I used your model as you delivered it, only blockMesh before.

At least one scheme is missing:

keyword div(((rho*nuEff)*dev2(T(grad(U))))) is undefined in dictionary
...at2_1/system/fvSchemes.divSchemes

[vishal_s]

Thanks Uwe. I will incorporate this scheme and run it. Will get back to you with the result.

[vishal_s]

Hi Uwe, I incorporated the div(((rho*nuEff)*dev2(T(grad(U))))) in fvScheme, the result is same. The simulation is blowing off after few time-steps.

Please click here for the log file of my simulation, along with the new case set-up.

Thanks.

[piu58]

I tried it again. Again, the case is not complete. I get in the first round:

--> FOAM FATAL ERROR:
previous iteration field
IOobject: volScalarField rho "C:/OpenFOAM/17.02/Z0Pilz-3.0.x/run/T2/0"

  not stored.  Use field.storePrevIter() at start of iteration.

    From function GeometricField<Type, PatchField, GeoMesh>::prevIter() const
    in file /opt/OpenFOAM/OpenFOAM-3.0.x/src/OpenFOAM/lnInclude/GeometricField.C at line 839.

FOAM aborting

I cannot help you if I don't get a valid case form you. So I give up.

One last hint: Run the simulation without redirecting the log. Perhaps you don't have the last lines of the log stored, because the solver crashes irregularly. On the screen the log is complete.

All in all, it seems to be a problem with the solver or one of it's libraries.