[Abubakar786]

Hi all,

I am trying to run a von-Karman vortex FSI simulation using CONVERGE CFD but when I am running it gives me an error as shown below. I tried all the possible ways but was not able to resolve the request you to help me with this.

***********Making an embedding change********************



MAX CFL= 9.9687e-01, MAX VISCOSITY CFL= 4.5718e-01, MAX CONDUCTION CFL= 3.3836e-01, MAX MASS DIFFUSION CFL= 5.8613e-01, MAX MACH CFL = 4.3183e+01
[ 2.80%] ncyc 267; Time 1.327753412e-01 (sec); dt= 5.151927594e-04 (dt_cfl)

ncyc= 268, time= 1.332905339e-01, dt= 5.151927594e-04, time-step limit = dt_cfl(Rank 1) WARNING: ERROR IN CELL PAIRING. CELL WITH NO ALIGNED NEIGHBOR
(Rank 1) WARNING: ERROR IN CELL PAIRING. CELL WITH NO ALIGNED NEIGHBOR
(Rank 1) WARNING: [3321] RANK1 During cell pairing, no viable neighbor found. 682 775 765 0. The volume specified by min_vol_ratio may not be met. This may adversely affect time-step control and simulation stability.
(Rank 1) WARNING: [3321] RANK1 During cell pairing, no viable neighbor found. 682 775 770 1. The volume specified by min_vol_ratio may not be met. This may adversely affect time-step control and simulation stability.
(Rank 1) FATAL_ERROR: [ 366] Something is wrong in function get_temp_from_table_massfrac: upper_energy_value must be greater than lower_energy_value!
upper_energy_value is -nan(ind), and lower_energy_value is -nan(ind).


Thanks in advance.

[jetcheve]

Good afternoon! There are quite a few things that could drive this error. That being said, I do see your mesh is changing and that cell pairing is struggling. I would try to get a post.h5 file (if you are using 3.0 or later) at the time the solver crashes and take a look at the quality of the mesh. In particular look for small gaps where large cells are trying to be fit into and correct this. Check the embedding change - is it coarsening something that may be problematic? Once you've done that try again.

Good luck and let us know if we can help!

- John

Originally Posted by Abubakar786

Hi all,

I am trying to run a von-Karman vortex FSI simulation using CONVERGE CFD but when I am running it gives me an error as shown below. I tried all the possible ways but was not able to resolve the request you to help me with this.

***********Making an embedding change********************



MAX CFL= 9.9687e-01, MAX VISCOSITY CFL= 4.5718e-01, MAX CONDUCTION CFL= 3.3836e-01, MAX MASS DIFFUSION CFL= 5.8613e-01, MAX MACH CFL = 4.3183e+01
[ 2.80%] ncyc 267; Time 1.327753412e-01 (sec); dt= 5.151927594e-04 (dt_cfl)

ncyc= 268, time= 1.332905339e-01, dt= 5.151927594e-04, time-step limit = dt_cfl(Rank 1) WARNING: ERROR IN CELL PAIRING. CELL WITH NO ALIGNED NEIGHBOR
(Rank 1) WARNING: ERROR IN CELL PAIRING. CELL WITH NO ALIGNED NEIGHBOR
(Rank 1) WARNING: [3321] RANK1 During cell pairing, no viable neighbor found. 682 775 765 0. The volume specified by min_vol_ratio may not be met. This may adversely affect time-step control and simulation stability.
(Rank 1) WARNING: [3321] RANK1 During cell pairing, no viable neighbor found. 682 775 770 1. The volume specified by min_vol_ratio may not be met. This may adversely affect time-step control and simulation stability.
(Rank 1) FATAL_ERROR: [ 366] Something is wrong in function get_temp_from_table_massfrac: upper_energy_value must be greater than lower_energy_value!
upper_energy_value is -nan(ind), and lower_energy_value is -nan(ind).


Thanks in advance.

[Abubakar786]

Originally Posted by jetcheve

Good afternoon! There are quite a few things that could drive this error. That being said, I do see your mesh is changing and that cell pairing is struggling. I would try to get a post.h5 file (if you are using 3.0 or later) at the time the solver crashes and take a look at the quality of the mesh. In particular look for small gaps where large cells are trying to be fit into and correct this. Check the embedding change - is it coarsening something that may be problematic? Once you've done that try again.

Good luck and let us know if we can help!

- John

Hi John,

Firstly, thanks for replying to my question.

Yes, you were correct mesh quality was the problem the large cells were trying to be fit into the small area to get rid of this I refined the mesh from 0.2 m to 0.05 m, and for simplicity, I removed the fixed embedding but I am facing the same error the simulation is crashing at the time like the previous case.

For your reference, I pasted the mesh image where I can see the mesh quality is poor,


Thanks in advance.

[jetcheve]

Hello, again!

First off - can you provide more details about your simulation? Geometry, cell count, boundary conditions, when this error occurs and if something is happening at that time, etc.?

Next, if you don't see any further mesh issues I would look at your time-step and CFL numbers. Make sure these aren't set too high. Without more to go on, it's hard to troubleshoot the error you are seeing, unfortunately.

Thanks and let me know your thoughts.

- John

Originally Posted by Abubakar786

Hi John,

Firstly, thanks for replying to my question.

Yes, you were correct mesh quality was the problem the large cells were trying to be fit into the small area to get rid of this I refined the mesh from 0.2 m to 0.05 m, and for simplicity, I removed the fixed embedding but I am facing the same error the simulation is crashing at the time like the previous case.

For your reference, I pasted the mesh image where I can see the mesh quality is poor,


Thanks in advance.

[jetcheve]

Originally Posted by Abubakar786

Hi John,

Firstly, thanks for replying to my question.

Yes, you were correct mesh quality was the problem the large cells were trying to be fit into the small area to get rid of this I refined the mesh from 0.2 m to 0.05 m, and for simplicity, I removed the fixed embedding but I am facing the same error the simulation is crashing at the time like the previous case.

For your reference, I pasted the mesh image where I can see the mesh quality is poor,


Thanks in advance.

Also, just so you know, your image didn't come across properly so I couldn't get a look at your mesh.

- John