CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Post-Processing (https://www.cfd-online.com/Forums/openfoam-post-processing/)
-   -   paraFoam crash(es) (https://www.cfd-online.com/Forums/openfoam-post-processing/232426-parafoam-crash-es.html)

JayDeeUU December 13, 2020 21:23

paraFoam crash(es)
 
I have run several of the standard OpenFoam tutorials, and in most cases (e.g. pitzDaily, buoyantCavity and hotRoomBoussinesq) I've been able to display the solution fields in paraFoam, as expected.

However an exception is the case "iglooWithFridges". Without altering any of the system files, I

1) copy the case folder from tutorials
2) cp -r 0 0.org
3) run blockMesh
4) run snappyHexMesh
5) run buoyantSimpleFoam (to time 4000)
6) paraFoam. Select time 3800, check "T" and Apply:
7) crash:
--> FOAM FATAL IO ERROR:
size 4000 is not equal to the given value of 11274

I have had this outcome repeatedly, without fail, over a couple of weeks. A crash with similar messaging occurs when I run my own case, also using buoyantSimpleFoam. In my case I start off with a blockMesh that is 50 x 50 x 50 (=125000) but snappyHexMesh builds a refined mesh (lying entirely within the original blockMesh; see my post at https://www.cfd-online.com/Forums/op...b-volumes.html). The solver proceeds to a normal stop. I can take a look at the ascii file (say "T" ,at the end of the run)... it looks like (showing a few key lines, with the line numbers shown at left):

Code:

20 internalField  nonuniform List<scalar>
21 125000
22 (
23 290.2886
24 290.29108
.
.

124021 289.94858

When I attempt to look at the T data in paraFoam, so soon as I load T and "Apply" I get a crash:

--> FOAM FATAL IO ERROR:
size 125000 is not equal to the given value of 369502

I don't know where the "given" size comes from... the size 125000 concurs with the number of temperature lines in the data file. I wonder whether perhaps openFoam has taken the flow domain to be that defined by the original blockMesh instead of the inner, refined grid that snappyHexMesh set up.

I would anticipate these crashes only reflect my own inexperience, however, the fact that it happens for a tutorial case causes me to wonder. Any suggestions appreciated! Thanks. John.


Resolution of the problem: It turns out that the problem was to have (re-) built the mesh without using the overwrite flag, i.e. I ought to have been issuing "snappyHexMesh -overwrite"

See related forum discussions:

https://www.cfd-online.com/Forums/paraview/152151-size-not-equal-given-value.html
https://www.cfd-online.com/Forums/op...ven-value.html


All times are GMT -4. The time now is 07:39.