CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   scalarTransportFoam erased my original data and has very slowly transport results (http://www.cfd-online.com/Forums/openfoam-solving/104816-scalartransportfoam-erased-my-original-data-has-very-slowly-transport-results.html)

foamWang July 16, 2012 17:10

scalarTransportFoam erased my original data and has very slowly transport results
 
Hi foamers,

I used other solver (e.g. pisoFoam) to generate a series of velocity fields.
Then, I wanted to use scalarTransportFoam to calculate the scalar transport by the given velocity fields.
So I just simply type scalarTransportFoam after setting up BC for T in 0 directory and adding DT in transportProperties, as well as some more corresponding modification in the fvSolution and fvSchemes to make it run.

It runs without error, but my original U data has been replaced by the U file in 0 directory and the transport is actually slow, which looks like the U is not read in at all.

Could anybody help me?

Thanks.

foamWang

akidess July 17, 2012 03:23

Did you start the simulation from latestTime?

vishalsacharya October 30, 2013 18:20

Hi,

I think starting the simulation from latestTime means that only the U from latestTime is used. But what if for every time-step, I want scalarTransportFoam to use the U from that time-directory that was obtained from a different simulation?
Say I have time-directories 0, 0.1, 0.2 etc all with a U file in them for that time, then can i copy these to scalarTransportFoam and get it to use the U from that time if I set my time-step to 0.1 and startTime to 0? I would have to technically move the code block that reads U, into the time-loop in the solver code. But this should work right?

Vishal

Lieven October 31, 2013 03:51

Hi Vishal

Quote:

I would have to technically move the code block that reads U, into the time-loop in the solver code
That is indeed one option that will work. Note though, that you limited the choice of time step strongly in this case and... it will be rather slow since IO-operations are quite time-demanding.

I would opt to include the scalar transport equation directly into U-field solver and solve the scalar transport on the fly. For 3 reasons:
1. It gives you more flexibility towards choice of time step in the scalar transport.
2. It will go faster than storing all U-field and reading them afterwards
3. You need far less storage space

Cheers,

Lieven


All times are GMT -4. The time now is 16:01.