Using steady state results to initialize flow in unsteady simulation
 

Hello everyone. I have to simulate an unsteady simulation for a short pipe(because of some phenomena that I need to observe which is unsteady in
nature). But I want to initialize the flow under steady-state conditions. So I already ran the simulation in a steady state and needed to use the resulting flow(At the outlet) to initialize the unsteady simulation.

I understand I need to use user-defined field functions, but I do not know how to write the syntax for the same. Is there any resource that I can use to learn the syntax? Any example which is similar in nature will also be a big help.

Don't you want to use the steady solution everywhere in the domain? Why only at the outlet?


Unless you clear the solution, starccm+ remembers it, so if you solve the steady problem and then switch to unsteady, you continue from the steady solution. Alternatively you can save the steady solution using a mapper, load it into the unsteady simulation as a file table and use interpolation to provide field functions that can be selected in the initial conditions. But this becomes impractical for large problems.

That is correct, I want to use the steady state solution for the entire domain to initialize the unsteady solution. I decoupled both legs of the simulation into two different simulation files. That way I don't have to run the steady state case every time I make some changes to the unsteady problem. Moreover I will be doing some mesh refinements, introducing injectors in the unsteady problem which is not necessary for the steady case. So if I have to run it multiple times, it might get cumbersome.

Using the mapper or the user-defined field function, which one would be more suitable? I am new to starccm so I am unable to make a judgement. The problem with udff is I don't know the syntax. I will explore the mapper and see if that works.

Thank you for your reply.

So you have run the steady state part of the simulation and it has converged. And I'm assuming you have saved the simulation file with the solution.

Is there anything stopping you from copying the simulation file with the converged solution, make any changes you want and run from that point as unsteady?
(I'm assuming that the solution interpolation should also be enabled in the physics models)