Hi there
I’m studying the subroutines ‘NavierStokeBoundary’ and ‘NavierStokesWallLaw’ , but there are many statements in the source-code that I can’t understand (form the subroutine itself as I’m a good programmer in Fortran ). Is there any document that can help me in this subject?
Thanks,
Diana Carlson, NY
‘NavierStokeBoundary’ and ‘NavierStokesWallLaw’
-
diana.carlson
- Posts: 24
- Joined: 21 Oct 2009, 09:49
-
raback
- Site Admin
- Posts: 4838
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: ‘NavierStokeBoundary’ and ‘NavierStokesWallLaw’
Unfortunately the documentation here is rather sparse. The Boundary subroutine assembles the Neumann and Robin type of BCs i.e. forces, slip coefficient etc. Their equations may be at least for most parts found in the ModelsManual under N-S. The WallLaw is related to turbulence modeling and is even more poorly documented. Generally Elmer has quite little documentation on the implementation. The convention to favour descriptive variablenames saves a lot, but I can understand the agony in having to decipher the numerical formulation from the code.
-
diana.carlson
- Posts: 24
- Joined: 21 Oct 2009, 09:49
NavierStokeBoundary
Hi
Is there any book, paper, thesis or any other text document that may help me understand the subroutine ‘NavierStokeBoundary’? I’m really confused what variables NodalAlpha, NodalBeta and NodalSlipCoeff are.
Regards,
Diana
Is there any book, paper, thesis or any other text document that may help me understand the subroutine ‘NavierStokeBoundary’? I’m really confused what variables NodalAlpha, NodalBeta and NodalSlipCoeff are.
Regards,
Diana
-
raback
- Site Admin
- Posts: 4838
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: ‘NavierStokeBoundary’ and ‘NavierStokesWallLaw’
Hi Diana,
NodalAlpha and NodalBeta are related to free surface flows that depends on surface curvature. I guess NodalAlpha refers to the normal part of the stress tensor as in Eq. 2.16 of Models Manual and NodalBeta to the tangential dependent part as in Eqs. 2.17-2.19. This has been used, for example in the melt flows when modeling the growth of silicon. These coefficients are relevant only at free surfaces.
The NodalSlipCoeff refers to slip condition as opposed to no-slip condition. The tangential part of the stress tensor is assumed to be proportional to the velocity difference between wall and fluid with this particular coefficient. The slip flows are relevant at least when the local Knudsen number is in the range of ~0.01 or higher. This number gives the ratio of mean free path between collosions and the characteristic dimension of the flow. For ideal gases the m.f.p. is around ~72 nm and thus this physical feature is typically considered in micrometerscales only (MEMS). The slip coefficient could perhaps be used also for other types of problems where a thin boundary layer could be expressed using a similar formulation.
-Peter
NodalAlpha and NodalBeta are related to free surface flows that depends on surface curvature. I guess NodalAlpha refers to the normal part of the stress tensor as in Eq. 2.16 of Models Manual and NodalBeta to the tangential dependent part as in Eqs. 2.17-2.19. This has been used, for example in the melt flows when modeling the growth of silicon. These coefficients are relevant only at free surfaces.
The NodalSlipCoeff refers to slip condition as opposed to no-slip condition. The tangential part of the stress tensor is assumed to be proportional to the velocity difference between wall and fluid with this particular coefficient. The slip flows are relevant at least when the local Knudsen number is in the range of ~0.01 or higher. This number gives the ratio of mean free path between collosions and the characteristic dimension of the flow. For ideal gases the m.f.p. is around ~72 nm and thus this physical feature is typically considered in micrometerscales only (MEMS). The slip coefficient could perhaps be used also for other types of problems where a thin boundary layer could be expressed using a similar formulation.
-Peter