CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   simpleFoam time step continuity errors (https://www.cfd-online.com/Forums/openfoam-solving/124922-simplefoam-time-step-continuity-errors.html)

CUBoulderPhDStudent October 15, 2013 16:39

simpleFoam time step continuity errors
 
I'm running a simple laminar case using simpleFoam and low Reynold's number (around .01) with a very large domain. My system is 140000nm by 8000nm with mesh points around every 5nm. I've got a rectangular grid imprint on the bottom of the membrane and the domain is decomposed using scotch method into 48 domains.

I keep on running into errors where the "time step continuity error" goes to infinity. I've run "checkMesh -allGeometry" and the mesh comes back fine.

I have no slip on the imprinted membrane.
Pressure uniform 0, and zeroGradient velocity on the outlet.
The top is open to the rest of the fluid but I know the velocity profile along the surface and set freestream velocity and pressure on the surface.
Inlet I only had 6 data points but it was pretty uniform so I have uniform fixed value velocity there and zeroGradient on pressure.

I've tried a bunch of other boundary conditions, but no matter what the "time step continuity error" always explodes. Does anyone have any ideas or suggestions to try?

Thank you

Daniel_Khazaei October 15, 2013 17:06

As far as I know simpleFoam is a steady-state solver! (1.6-ext)

How could you have time step continuity error?


regards

CUBoulderPhDStudent October 15, 2013 17:14

It is a steady state solver. That's particularly what's disturbing. From the other threads I looked at, when this error occurs with simplefoam its either a mesh problem (which I checked) or its a turbulence model problem (and I'm laminar) and there was one boundary problem that involved periodic boundaries.

Daniel_Khazaei October 15, 2013 17:19

Quote:

Originally Posted by CUBoulderPhDStudent (Post 457157)
It is a steady state solver. That's particularly what's disturbing. From the other threads I looked at, when this error occurs with simplefoam its either a mesh problem (which I checked) or its a turbulence model problem (and I'm laminar) and there was one boundary problem that involved periodic boundaries.

I think that should mean the continuity error or someone correct me?:rolleyes:

Have you checked the boundary conditions?

it would be better if you could share a log file of your simulation.

cdunn6754 October 20, 2016 15:15

I am in a similar situation now, did you ever make any progress?

shrirang November 7, 2016 23:09

Time step continuity error
 
Time step continuity errors are due to improper mesh or boundary conditions.

After modifying the simpleFoam solver for T eqn, I was getting time step continuity errors in the range of 1+e11 and the solver crashed every time.
Changing the model to laminar did not solve the problem.
The problem was solved by changing the boundary conditions for Temperature and taking relaxation factors 0.2 for all variables.

Though the solution is not converging yet, but the time step continuity problem is not an issue now.

Hope that will help someone.

cdunn6754 November 8, 2016 10:51

Yep you are right, I think in my case I fixed it by decreasing the relaxation factors. Even with appropriate BC I was still having trouble but the under-relaxation fixed it.


All times are GMT -4. The time now is 11:49.