One finite volume kills all.

Propanotriol · September 26, 2020, 11:19

Hi there,

I am simulating an injection nozzle. Everything is set correctly and the simulation do work properly (a priori).

The problem is that as the velocity in the injector increases, I have ONE finite volume close to a wall that its CFL number goes towards unity (while for the rest of the mesh it is very low and OK). This unique element is making my time step to decrease and decrease until it is just impossible to continue with the simulation.

I tried to change the mesh, remesh, etc, add BL elements, but nothing, I have always, one or two elements that make the simulation impossible.

The simulation do not crash, it is just that the time step becomes extremely small (1e-15) so I am obliged to stop the simulation.

I do use a corrected scheme in my fvScheme. Is there any way to make the simulation more robust?

Thanks.

simrego · September 28, 2020, 04:38

Hi!

You can use limited instead of corrected.
"limited 0.33" is really robust.
And you can use pure 1st order schemes (upwind) for a trial run and your schemes will be robust like hell. (but for the final run change back to second order schemes!! If it'll crash with full upwind you should check your mesh and the boundary conditions.
And what is the solver? If you use the PIMPLE algorithm you can go beyond CFL 1.

September 26, 2020, 11:19	One finite volume kills all.	#1
Propanotriol New Member S A Join Date: Oct 2019 Posts: 8 Rep Power: 7	Hi there, I am simulating an injection nozzle. Everything is set correctly and the simulation do work properly (a priori). The problem is that as the velocity in the injector increases, I have ONE finite volume close to a wall that its CFL number goes towards unity (while for the rest of the mesh it is very low and OK). This unique element is making my time step to decrease and decrease until it is just impossible to continue with the simulation. I tried to change the mesh, remesh, etc, add BL elements, but nothing, I have always, one or two elements that make the simulation impossible. The simulation do not crash, it is just that the time step becomes extremely small (1e-15) so I am obliged to stop the simulation. I do use a corrected scheme in my fvScheme. Is there any way to make the simulation more robust? Thanks.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Matlab code for Finite Volume Method in 2D	coagmento	Main CFD Forum	5	October 7, 2022 14:47
multiphaseEulerFoam (OF2.3.0) : Courant number explodes when running in parallel	Mehrez	OpenFOAM Running, Solving & CFD	10	May 18, 2016 12:44
multiphase turbulance case floating error	harsha_kulkarni	OpenFOAM Running, Solving & CFD	3	February 18, 2016 06:06
Problem with SIMPLEC-like finite volume channel flow boundary conditions	ghobold	Main CFD Forum	3	June 15, 2015 12:14
influence of convex or concave control volume in finite volume discretization	openfoammaofnepo	Main CFD Forum	1	March 24, 2014 03:13

September 28, 2020, 04:38		#2
simrego Senior Member anonymous Join Date: Jan 2016 Posts: 416 Rep Power: 14	Hi! You can use limited instead of corrected. "limited 0.33" is really robust. And you can use pure 1st order schemes (upwind) for a trial run and your schemes will be robust like hell. (but for the final run change back to second order schemes!! If it'll crash with full upwind you should check your mesh and the boundary conditions. And what is the solver? If you use the PIMPLE algorithm you can go beyond CFL 1.