Restarting from existing result

[michiw]

Hi,

I have added the following line to my SIF file:

Code: Select all

Simulation
  Restart File = "prerun.ep"
  Restart Position = 0
End

I'm performing a coupled simulation of Navier-Stokes and HeatEquation. After simulation was run with a higher viscosity, I copy the result file <Post File> to the <Restart FIle> to get it converged.

My Question:
I would expect that it only need a few steps for a converged result starting from a converged solution to reach convergence? But if I do so without any changed parameter, the calculation starts with a high relative change (error) of around one and need approximatly the same number of steps.

Is it the right way of doing the restart?

By the way I'm using the current version of ElmerGui under Windows.

Thanks,
Michael

[raback]

From the FAQ:

Q: I want to restart using the .ep file, how do I do that?
A: Unfortunately Elmer uses the "Post File", with the usual suffix .ep, only for visualization. If you want to restart you need to save the data in a "Result File" which has some additional information required.

-Peter

[michiw]

Ahh I see,

sorry for my question! I have searched the web and looked in the forum but not the FAQ

The first sentence in the ElmerSolverManual.pdf (section2), I was not able to understand. Now I know what was meant. Mayby a comment about not using the Post Format should be added

Now I have added

Code: Select all

Output File = "prerun.result"
Restart File = "prerun.result"
Restart Position = 0 

to my simulation section and everything is as expected!!

Thank you!!
Michael

P.S: why does the solver do not complain about the wrong file? It only says ASCII 0 instead of ASCII 1?