CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Fluent UDF and Scheme Programming (http://www.cfd-online.com/Forums/fluent-udf/)
-   -   UDF to transfer results from one case to the next case (http://www.cfd-online.com/Forums/fluent-udf/120719-udf-transfer-results-one-case-next-case.html)

howhs July 12, 2013 21:28

UDF to transfer results from one case to the next case
 
Hi everyone,

I am rather new to UDF. I would like to transfer the results from the current case file (steady state simulation) to the next case file (transient simulation). May I know how can i do it? Is there a need to write the UDF to do so? If there is a need, may i know how can i do it?

Thank you very much in advance.

akarlin July 15, 2013 18:22

I am assuming you are trying to use your steady state results as an initial condition for your transient case. In such a case it is possible to, after running your steady state to convergence, click the general tab and set time to transient. Your steady state results will automatically be carried over to your transient run. Make sure you set an appropriate time step in the run calculation tab. Also, it is a good idea to save the case and data before proceeding.

If for some reason there is a difference in the meshes between your steady state and transient runs you can use Fluent's built in interpolation to bring over the results. In steady state with the results open, go to File>Interpolate>Write data and select all cell zones and fields then click write. Save the .ip file somewhere so that you will be able to find it again. In the transient case go to File>interpolate>read data and select the .ip file. Congrats, you've just brought over the steady state data.

Both of these options assume the geometry/domain between the two are identical. I can't think of a reason why this would not be true. No UDF's are necessary. Of course I'm sure you've probably already figured all this out by now but it cannot hurt to post anyway.

Best regards,
Adam

howhs July 15, 2013 22:15

Thank you so much for your reply!
There will be no difference in the meshes between the steady and transient runs.

I have another question. I understand that we need to use udf to write out a script if we want the flow to be a developed flow. I was thinking if it makes sense to instead of writing a udf, we run a steady state flow to allow the flow to become developed and continue it with the transient flow (as what you mentioned).

If it does not make sense, could you assist me in writing the udf for developed flow?

Thank you so much in advance!

akarlin July 16, 2013 09:25

That seems to be the most logical way forward (starting with steady state and proceeding to transient). Whenever possible that is how I start a transient simulation. This is a good approach to take when you are interested in the long term solution (pseudo-periodic/steady state) of the transient system since it can cut back considerably on the total number of time steps required. During the transient solution monitor the drag or some other relevant parameter to determine if the solution has reached a stable state.

What exactly are you modeling and why is it necessary to model it transient?

howhs July 16, 2013 23:31

Thank you for your reply.
Basically, I am doing a simple simulation on heat transfer due to the subsea sensor (acting as a cold spot) relating to subsea engineering. I will simulating a steady state flow during the oil production phase in a subsea pipeline. Assuming that 10m of the piping will be suffice to achieve adiabatic conditions on both sides of the pipes. A certain inlet velocity is given and a steady state simulation is done.
Now, during a production shutdown (sometime, there is a need to do intervention due to some reasons), the production oil will stop flowing. During this instant when the inlet velocity is zero, I am studying the transient state and temperature drop from the cold spot is noted. That is basically what I have doing.


All times are GMT -4. The time now is 20:07.