[VLima]

Hi there.

I am new to OpenFOAM and started testing some examples from the tutorial.
In the dam break problem my results are different from what is documented (see plots attached).
I was wondering if someone has any idea why these differences are observed and if all is ok with my installation or if something is missing.

Many thanks in advance for your time and help.

Vânia

[chrisvandijk]

Hi,


I have the same issue as you. My results are different for one and four cores.


Note that your figures look a lot like the ones in the tutorial (https://cfd.direct/openfoam/user-gui...k/#x7-520002.3), so I assume your results are right. However I don't know why the results are different, I'd also like to know the answer.


Greetings,
Chris

[dlahaye]

Do you obtain the same residual values for 1 and 4 processors?

[chrisvandijk]

Originally Posted by dlahaye

Do you obtain the same residual values for 1 and 4 processors?

The residual values are a bit different. The average of the final residuals for the last five timesteps for 1 core are 8.264e-8, for 4 cores it's 9.267e-8. This is probably correlated with the amount of iterations needed for convergence (again as an average of the last five timesteps: 52 for 1 core, 114 for 4 cores).


However, I'd say an error of less than 1e-7 shouldn't cause such a big discrepancy between the results for 1 and 4 cores.

[dlahaye]

I agree. No idea what else might cause the discrepancy.

[chrisvandijk]

Hi, I managed to find the answer (together with help from my thesis supervisors) in the meantime. The problem is the tolerances. If I edit my fvSolutions file and change all tolerances to 1e-18 or so, my results for one and four cores are much more similar.


Greetings,
Chris