In createFields.H "g*h" is calculated at the cell and face centres so
that relative heights are automatically taken into consideration, assuming
that is you specify g correctly in the "constant" directory of your case folder.

I wonder if this is what you mean?

Regards,

Richard K.

(with reference to Andrea's runtime error above)

I had a similar error message a while back (running OF1.6.x using openmpi ver. 1.2.8) but updating openmpi to version 1.4.1 seems to resolve the issue.
Perhaps it might work for you too,

Regards,

Richard K.

I don't know about OF1.7.1 but the git version 1.7.x works fine for parallel

Duong

Hi Foamers,

I have the same problem as Andrea had.

Running my case in parallel with version 1.6. works fine. But with version 1.7.1 it doesn`t work in parallel. It works only on one processor.

Is there a fix for that? Or is it a problem with openmpi as richpaj said?

Found it and it works.

https://github.com/OpenCFD/OpenFOAM-...7bc8646e01526d

Thanks
Stefan

Hi everyone;

I have been testing the "capillaryRise" test case that is available.
Using OF2.3.x and bc's similar to test case.
I would like to measure the flow length of the meniscus turning off the influence of 'g'.

I know one analytical solution which is the "Washburn equation (Wiki)" that gives the capillary rise over time but i observed quite a big difference between the computed flow length and analytical equation.
NOTE: Wiki discusses relation for a cylindrical channel but I changed the flow resistance to a rectangular channel.

https://en.wikipedia.org/wiki/Washbu...ion#Derivation

Is there an other analytical solution for this sort of case? If yes, could you please let me know regarding it.

Thanks;
Saideep