CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > OpenFOAM Running, Solving & CFD

interFoam fvSchemes for Convection Equation

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   February 7, 2012, 23:29
Question interFoam fvSchemes for Convection Equation
  #1
Member
 
carowjp's Avatar
 
Jim Carow
Join Date: Apr 2010
Location: Michigan, USA
Posts: 40
Rep Power: 7
carowjp is on a distinguished road
I have been implementing a solver for one phase of fluid and a second phase of viscoelastic solid based on interFoam. The constitutive equation for the solid depends on the left Cauchy-Green deformation tensor (B) which is calculated by an equation (similar to an upper convective derivative):

Code:
   gradU = fvc::grad(U);
   L = gradU.T();

   fvSymmTensorMatrix BEqn
   (
       fvm::ddt(B)
     + fvm::div(phi, B)
    ==
     twoSymm(B & L)
   );
As a test case I am solving a lid drive cavity case with a viscoelastic solid disk as the second phase. The fluid velocities and behavior seems correct however I am observing 'ripples' on the surface of the disk which seem artificial (please see the attached). The solid also orbits around the cavity faster than the benchmark.

I suspect numerical diffusion from the upwind differencing scheme used in the solution of BEqn.

Sorry for the discourse but can anyone suggest alternative schemes or PISO settings which might improve this?

I am using the following:

Code:
divSchemes
{
    default         none;
    div(rho*phi,U)  Gauss limitedLinearV 1;
    div(phi,alpha)  Gauss vanLeer;
    div(phirb,alpha) Gauss interfaceCompression;
    div(phi,B) Gauss upwind;
}
and

Code:
PISO
{
    momentumPredictor yes;
    nCorrectors     3;
    nNonOrthogonalCorrectors 0;
    nAlphaCorr      1;
    nAlphaSubCycles 2;
    cAlpha          2;
}
thank you,

James
Attached Images
File Type: png wiggles.png (13.2 KB, 61 views)
File Type: png wiggles2.png (17.4 KB, 59 views)
carowjp is offline   Reply With Quote

Old   February 8, 2012, 04:06
Default
  #2
Senior Member
 
akidess's Avatar
 
Anton Kidess
Join Date: May 2009
Location: Delft, Netherlands
Posts: 919
Rep Power: 17
akidess will become famous soon enough
I don't know about your specific case, but if you want to limit numerical diffusion MUSCL and SuperBee schemes work quite well.
__________________
*On twitter @akidTwit
*Spend as much time formulating your questions as you expect people to spend on their answer.
*Help define the OpenFOAM stackexchange Q&A site: http://area51.stackexchange.com/prop...oam-technology
akidess is offline   Reply With Quote

Old   April 19, 2012, 23:17
Question 'Wiggles' and 'Wrinkles'
  #3
Member
 
carowjp's Avatar
 
Jim Carow
Join Date: Apr 2010
Location: Michigan, USA
Posts: 40
Rep Power: 7
carowjp is on a distinguished road
Thanks for the input on the convection schemes. I can run with UMIST and SFCD applied to the convection term in the equation above but QUICK, MUSCL, and SuperBee lead to segmentation faults.

Since then I've learned the issue shown in the attached images wasn't numerical diffusion, although maintaining a sharp interface is always a good thing.

I am still looking for an explanation for the presence of the artificial 'wiggles' or 'wrinkles' which appear on the interface using interFoam's interfaceCompression scheme and a non-zero value for cAlpha.

If you have a look at a nice comparison of VOF methods in:

Gopala, V. R., & van Wachem, B. G. M. (2008). Volume of fluid methods for immiscible-fluid and free-surface flows. Chemical Engineering Journal, 141(1-3), 204-221. doi:10.1016/j.cej.2007.12.035

You will see the same artificial wiggles present in the results of their simulations performed with interFoam, though there is no specific mention of this issue.

In my cases there were no forces due to surface tension and I believe that such forces stabilize the interface an counteract the artificial distortion. Does Weller's scheme used in interFoam depend on the presence of surface tension?

thanks and regards,

Jim
carowjp is offline   Reply With Quote

Old   April 23, 2012, 08:42
Default
  #4
Senior Member
 
akidess's Avatar
 
Anton Kidess
Join Date: May 2009
Location: Delft, Netherlands
Posts: 919
Rep Power: 17
akidess will become famous soon enough
Section 4.2 of this paper mentioned grid-alignment of the interface: http://powerlab.fsb.hr/ped/kturbo/op...InterTrack.pdf

It looks like the only thing you can do is reduce the time step.
__________________
*On twitter @akidTwit
*Spend as much time formulating your questions as you expect people to spend on their answer.
*Help define the OpenFOAM stackexchange Q&A site: http://area51.stackexchange.com/prop...oam-technology
akidess is offline   Reply With Quote

Old   July 15, 2013, 15:51
Default Left-
  #5
New Member
 
bryan
Join Date: Nov 2012
Posts: 5
Rep Power: 4
bcgooder is on a distinguished road
I've also been trying to use the left-cauchy green tensor in a new solver with this code.

I was just wondering what you used for your boundary and initial conditions for B??
bcgooder is offline   Reply With Quote

Old   July 15, 2013, 16:08
Default BC and IC's for B
  #6
Member
 
carowjp's Avatar
 
Jim Carow
Join Date: Apr 2010
Location: Michigan, USA
Posts: 40
Rep Power: 7
carowjp is on a distinguished road
Hello,

I used the initial conditions B=I, identity since the body was initially undeformed, and I used zeroGradient conditions for the BC's.

If you are advecting the tensor B you will likely face the challenge of diffusion based on the convection schemes I mentioned in this post.

I ended up having best success with the using the CICSAM scheme (available in OF extend version ShipHydroSIG) for advecting the scalar field alpha.

Curious to know what you are working on.

regards,

Jim
carowjp is offline   Reply With Quote

Old   July 15, 2013, 16:36
Default
  #7
New Member
 
bryan
Join Date: Nov 2012
Posts: 5
Rep Power: 4
bcgooder is on a distinguished road
I'm actually working on a viscoelastic fluid model for blood where the left-cauchy tensor is needed for the stress constitutive equation.

I'm still not entirely sure if this method for finding B will work in my case but I was planning to test it out as a first try.
bcgooder is offline   Reply With Quote

Reply

Tags
diffusion, fvschemes, interfoam, viscoelasticfluid

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Calculation of the Governing Equations Mihail CFX 7 September 7, 2014 06:27
Momentum equation of interFoam? cfdfans OpenFOAM 9 January 5, 2013 04:09
Constant velocity of the material Sas CFX 15 July 13, 2010 08:56
interFoam + a scalar transport equation CYMa OpenFOAM 1 May 11, 2010 03:50
mass flow in is not equal to mass flow out saii CFX 2 September 18, 2009 08:07


All times are GMT -4. The time now is 02:24.