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

Something wrong in UEqns.H within twoPhaseEulerFoam

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

Reply
 
LinkBack Thread Tools Display Modes
Old   March 15, 2011, 13:35
Exclamation Something wrong in UEqns.H within twoPhaseEulerFoam
  #1
Member
 
Charlie
Join Date: Dec 2010
Location: 415 Kinross Dr. Newark, DE 19711
Posts: 80
Rep Power: 7
cheng1988sjtu is on a distinguished road
I'm right now trying to understand how twoPhaseEulerFoam works, and get what they are erally solving. However, There seems something odd in the code.

Following is the source code for Ua and Ub in twoPhaseEulerFoam:

Quote:
fvVectorMatrix UaEqn(Ua, Ua.dimensions()*dimVol/dimTime);
00002 fvVectorMatrix UbEqn(Ub, Ub.dimensions()*dimVol/dimTime);
00003
00004 {
00005 {
00006 volTensorField gradUaT = fvc::grad(Ua)().T();
00007 volTensorField Rca
00008 (
00009 "Rca",
00010 ((2.0/3.0)*I)*(sqr(Ct)*k + nuEffa*tr(gradUaT)) - nuEffa*gradUaT
00011 );
00012
00013 if (kineticTheory.on())
00014 {
00015 Rca -= ((kineticTheory.lambda()/rhoa)*tr(gradUaT))*tensor(I);
00016 }
00017
00018 surfaceScalarField phiRa =
00019 -fvc::interpolate(nuEffa)*mesh.magSf()*fvc::snGrad( alpha)
00020 /fvc::interpolate(alpha + scalar(0.001));
00021
00022 UaEqn =
00023 (
00024 (scalar(1) + Cvm*rhob*beta/rhoa)*
00025 (
00026 fvm::ddt(Ua)
00027 + fvm::div(phia, Ua, "div(phia,Ua)")
00028 - fvm::Sp(fvc::div(phia), Ua)
00029 )
00030
00031 - fvm::laplacian(nuEffa, Ua)
00032 + fvc::div(Rca)
00033
00034 + fvm::div(phiRa, Ua, "div(phia,Ua)")
00035 - fvm::Sp(fvc::div(phiRa), Ua)
00036 + (fvc::grad(alpha)/(fvc::average(alpha) + scalar(0.001)) & Rca)
00037 ==
00038 // g // Buoyancy term transfered to p-equation
00039 - fvm::Sp(beta/rhoa*K, Ua)
00040 //+ beta/rhoa*K*Ub // Explicit drag transfered to p-equation
00041 - beta/rhoa*(liftCoeff - Cvm*rhob*DDtUb)
00042 );
00043
00044 UaEqn.relax();
00045 }
00046
00047 {
00048 volTensorField gradUbT = fvc::grad(Ub)().T();
00049 volTensorField Rcb
00050 (
00051 "Rcb",
00052 ((2.0/3.0)*I)*(k + nuEffb*tr(gradUbT)) - nuEffb*gradUbT
00053 );
00054
00055 surfaceScalarField phiRb =
00056 -fvc::interpolate(nuEffb)*mesh.magSf()*fvc::snGrad( beta)
00057 /fvc::interpolate(beta + scalar(0.001));
00058
00059 UbEqn =
00060 (
00061 (scalar(1) + Cvm*rhob*alpha/rhob)*
00062 (
00063 fvm::ddt(Ub)
00064 + fvm::div(phib, Ub, "div(phib,Ub)")
00065 - fvm::Sp(fvc::div(phib), Ub)
00066 )
00067
00068 - fvm::laplacian(nuEffb, Ub)
00069 + fvc::div(Rcb)
00070
00071 + fvm::div(phiRb, Ub, "div(phib,Ub)")
00072 - fvm::Sp(fvc::div(phiRb), Ub)
00073
00074 + (fvc::grad(beta)/(fvc::average(beta) + scalar(0.001)) & Rcb)
00075 ==
00076 // g // Buoyancy term transfered to p-equation
00077 - fvm::Sp(alpha/rhob*K, Ub)
00078 //+ alpha/rhob*K*Ua // Explicit drag transfered to p-equation
00079 + alpha/rhob*(liftCoeff + Cvm*rhob*DDtUa)
00080 );
00081
00082 UbEqn.relax();
00083 }
00084 }
in the code it seems very different for solving Ua and Ub (Noted as red in the quoted codes, compare line 10 and 52, line 24 and 61, line 41 and 79), However, We can define which phase is the dispersed phase in the "constant" files. Which means that we Ua and Ub should be solved in the similar way.

Or does openfoam has assigned use Ua to calculate dispersed phase and Ub for continuous phase? for example I define b as dispersed phase, then openfoam will solve Ua for b?

Also, note that only in Ueqn for a we can find the kinetic theory option. Is that true?

What's the momentum equation to guide this code? I'm relatively new to twoPhaseEulerFoam. Any help will be highly appreciated!
cheng1988sjtu is offline   Reply With Quote

Old   June 24, 2011, 09:30
Default
  #2
New Member
 
Ulisses
Join Date: Mar 2009
Location: Brazil
Posts: 24
Rep Power: 9
ulisses is on a distinguished road
Hi Charlie,

I am using kinetic theory in twoPhaseEulerFoam but my solution diverges. Maybe it could be because of that you showed up. Have you achieved a conclusion about that?

Thanks
ulisses is offline   Reply With Quote

Old   June 24, 2011, 10:48
Default Try to reduce the timestep
  #3
Member
 
Charlie
Join Date: Dec 2010
Location: 415 Kinross Dr. Newark, DE 19711
Posts: 80
Rep Power: 7
cheng1988sjtu is on a distinguished road
Hi,

There are two options to simulate the particle phase turbulence (In the code).

If we turn on the kinetic theory, then use kinetic theory to calculate the effective viscosity of particle phase.

If we turn the kinetic theory off, then we use turbulent response coefficient Ct to model the turbulent viscosity of particle phase.

As far as I'm concerned, the alphaEqn is more of significance leading to numerical divergence, since the alphaEqn in twoPhaseEulerFoam is nonlinear and also need to be bounded on both ends. try to refer to PhD thesis of Henrik Rusche, 2002.

About the phase selection, I just confirm to the convention, that is, use default setting, a as disperse phase, b as continuous phase.

Any question are welcomed.

Charlie

Quote:
Originally Posted by ulisses View Post
Hi Charlie,

I am using kinetic theory in twoPhaseEulerFoam but my solution diverges. Maybe it could be because of that you showed up. Have you achieved a conclusion about that?

Thanks
cheng1988sjtu is offline   Reply With Quote

Reply

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
Warning: Dynamic zone with wrong CG using 6DOF Manoj Kumar FLUENT 1 August 11, 2012 04:03
A huge mistake that could ruin my project: wrong turbulence specification Cheeko FLUENT 1 June 12, 2010 07:02
BuoyantBoussinesqSimpleFoam and axial-symmetric results wrong mass flow Thomas Baumann OpenFOAM 6 December 21, 2009 11:31
udf error srihari FLUENT 0 February 9, 2009 10:00
Pressure contour seems wrong??? Harry Qiu FLUENT 1 June 29, 2001 05:53


All times are GMT -4. The time now is 09:58.