Filling Process

The graphical user interface of Elmer
Post Reply
hoppiw07
Posts: 3
Joined: 24 Aug 2021, 12:03
Antispam: Yes

Filling Process

Post by hoppiw07 »

Hi everyone,

i have a problem with elmer gui and paraview. First to my project, I want so simulate a filling process of a liquid in a glas. The first eight steps looks not too bad, but the drop goes not to the bottom of the glas. Also an error code is always written:
I change the BDF from 1 to 2 to 3, integrate an other solver, change the timesteps and intervall, change constants but nothing gets me ahead and so the error code was the same. Did someone have an idea whats the problem in my sif or mesh is?

Thanks in advance.
Attachments
case.sif
(4.81 KiB) Downloaded 193 times
2D.msh
(682.25 KiB) Downloaded 188 times
2D.geo
(809 Bytes) Downloaded 188 times
hoppiw07
Posts: 3
Joined: 24 Aug 2021, 12:03
Antispam: Yes

Re: Filling Process

Post by hoppiw07 »

as an complement to my first message
Attachments
mesh.elements
(367.76 KiB) Downloaded 187 times
mesh.boundary
(12.68 KiB) Downloaded 187 times
egproject.xml
(152.28 KiB) Downloaded 185 times
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: Filling Process

Post by Rich_B »

Hi,

I ran your case, here is the last section of the solver output:
FlowSolve: -------------------------------------
FlowSolve: NAVIER-STOKES ITERATION 20
FlowSolve: -------------------------------------
FlowSolve:
FlowSolve: Starting Assembly...
FlowSolve: Assembly:
: ..Assembly done
FlowSolve: Dirichlet conditions done
ComputeChange: NS (ITER=20) (NRM,RELC): ( 0.87747935E+40 1.5579146 ) :: navier-stokes
FlowSolve: iter: 20 Assembly: (s) 2.59 49.28
FlowSolve: iter: 20 Solve: (s) 0.26 6.35
FlowSolve: Result Norm : 8.7747935060334869E+039
FlowSolve: Relative Change : 1.5579145737967666
SaveResult: -----------------------------------------
SaveResult: Saving results to file: ./2D/previousrun.result
SaveResult: Done writing results file
SaveResult: -----------------------------------------
ResultOutputSolver: -------------------------------------
ResultOutputSolver: Saving in unstructured VTK XML (.vtu) format
ResultOutputSolver: -------------------------------------
LevelSetTimestep: Levelset timestep 1.235E-42
ERROR:: ExecSimulation: Timestep smaller than epsilon: 1.235E-42
If you search the forum for
levelset peter
look for items such as this one:

viewtopic.php?f=7&t=4&p=21833&hilit=lev ... ter#p21833
Unread post by raback » 03 Dec 2019 16:42
Hi Brando,

This might be outside the capabilities of Elmer. Sloshing has been done only for fluids that do it in a civilized manner (using moving mesh & ALE formulation). Severe sloshing would be better of with VOF or SMH methods. We have levelset but it is pretty simple one and does not conserve mass. Elmer does have particles but there can be no shear between them currently. I would probably look at codes that were built on particles in mind. FEM excels on PDEs that have some minimization principle associated and you've none of that here.

-Peter
Rich.
hoppiw07
Posts: 3
Joined: 24 Aug 2021, 12:03
Antispam: Yes

Re: Filling Process

Post by hoppiw07 »

Thanks Rich_B for your help.

In the time, i optimize the sif and it looks a little bit better. But the problem is also that the gas go not realistic out and the drop reach not correct the buttom

I also had the idea to solve the problem with freesurface but i think, for a liquid is levelset the correct part.

Does someone see the mistake in the sif and can help?

Big thanks in advance.

hoppiw07
Attachments
case.sif
here my sif.
(4.92 KiB) Downloaded 179 times
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Filling Process

Post by raback »

Hi

This is pretty old but I haven't looked at before. I just studied the geometry (a long tube leading to a rectangular container in 2D about to filled).

This cannot be solved with Lagrangian methods since obviously mesh deformation will not suffice. So for proper fluid solvers we are left with the level-set strategy. Levelset methods would be much more ideal if the two fluids would be pretty similar. Liquid replacing gas is not easily presented by the smeared down material parameters. Also, you probably have turbulence in the system and there is really nothing in Elmer on turbulence & two phase flow.

Now, the width of your mesh was 7.3 metres. It could be that those really should be in mm and this is a microfluidic thing. And maybe turbulence is not important. Even in that case there is a problem with the boundary conditions & levelset. The profile on the pipe would be then paraboli vanishing to the edges. This means that there is nothing transporting the levelset function close to the edges. This is a problem. Maybe there should be slip-conditions for the fluid.

I would suggest looking for some other software for this problem.

-Peter
Post Reply