Limiting possible internal fluid pressure in Navier-Stokes simulation

Numerical methods and mathematical models of Elmer
Post Reply
thomasatelmer
Posts: 47
Joined: 18 Jan 2019, 18:18
Antispam: Yes

Limiting possible internal fluid pressure in Navier-Stokes simulation

Post by thomasatelmer »

Hello, Elmer community,

I am sending in my question after undergoing some Research in this Forum here, and I did not find possible Solutions unfortunately. For modelling a 3D case of shearing a thin layer of low-viscosity fluid, I am undertaking Simulation using the N-S solver with incompressible material model. Due to the nature of the Problem, symmetry and using the Reynolds solver are no Options. However, as the fluid is passing over local steps in the geometry, I get quite some pressure drop in the fluid at These Locations. The pressures obtained are much lower than any Kind of vapor pressure that could ever be assumed in reality. Now I am wondering how I could These low pressures to be "forbidden" in the Solution - I think for the start, just limiting the Minimum pressure might be helpful. I know that the Reynolds solver has such Feature, however for the Navier-Stokes, what are my Options?

Looking Forward to your hints,
Thomas.
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Limiting possible internal fluid pressure in Navier-Stokes simulation

Post by raback »

Hi Thomas

I don't think the general limiters will work in Navier-Stokes. It might be possible to have them but I would guess some coding would be needed to limit the soluton. You could of course try. Take a case of elasticity where a component is limited and mimic it for Pressure in Navier-Stokes.

As finite elements is not conservative by construction you may get these artefacts. It could be also be affected by mesh density or the compressibility model. It is quite hard to suggest a remedy without actually working with the case.

-Peter
thomasatelmer
Posts: 47
Joined: 18 Jan 2019, 18:18
Antispam: Yes

Re: Limiting possible internal fluid pressure in Navier-Stokes simulation

Post by thomasatelmer »

Hello Peter,

thank you for your assessment, I seem to be coming to the more difficult Areas now... :-). I wonder if using a user-defined compressibility model would do the trick, much like in viewtopic.php?t=748#p2528

Example from this post:

Code: Select all

Material 1
  Name = "Lubricant oil (coarse values)"
  Heat Conductivity = 0.128
  Reference Temperature = 293
  Heat Capacity = 2093
  Density = Variable Pressure
  ...
    Real MATC "max(0.0001 866/(1+0.09*(1/tx)))"
  Compressibility Model = User Defined
  ...
  Viscosity = 20.0e-3
End
Thank you for your continued Support!

Best Regards,
Thomas
thomasatelmer
Posts: 47
Joined: 18 Jan 2019, 18:18
Antispam: Yes

Re: Limiting possible internal fluid pressure in Navier-Stokes simulation

Post by thomasatelmer »

Hello everybody,

I managed to use soft limiters for the pressure in the Navier-Stokes solver. Per section 8.2 of the solver Manual:

In the solver Control section:

Code: Select all

Apply Limiter = Logical True
Limiter Value Tolerance = Real 1.0e-9
Limiter Load Tolerance = Real 1.0e-9
This generally enables the limiter functionality.
Now the fluid in which the pressure should be limited, receives a Body force:

Code: Select all

Body Force 1
  Name = "Pressure Limit"
Pressure Lower Limit = Real -6.13e4
End
... end of Story :-)

the solution is a Little rough at times and takes a Little longer to converge, but convergence still can be achieved.

Thanks for the Support,
Thomas.
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Limiting possible internal fluid pressure in Navier-Stokes simulation

Post by raback »

Hi Thomas

Great! Well done. We certainly didn't plan this but great if it works.

So what is being done here is that a lower limit is added to pressure. This means that the continuity equation is no longer honored but pressure is replaced by the limit value when we go below the limit. In practice it is a contact set problem and at contact Dirichlet conditions are applied.

If we think about physics this could some kind of cavitation model i.e. pressure can never go below the some limit where the fluid would start to evaporate. One might use it also as a numerical trick but it may be a remedy for other problems and might not be physically what one is expecting.

-Peter
Post Reply