reactingtwophaseeulerfoam?

[kimy]

@东岳 Thanks, if the sum of that two term is zero, so there cannot exist any problem during the simulation. why I use the sign "plus", it crashed while 'minus' run well? And if I changed the phase2.turbulence.nut() to be nu() a constant, "plus" has no problem?

[李东岳]

why I use the sign "plus", it crashed while 'minus' run well? And if I changed the phase2.turbulence.nut() to be nu() a constant, "plus" has no problem?

Do you obtain physically reasonable results? or just not blowing up.

[kimy]

@东岳 Frankly, I use the minus sign and I got the results however they doesn't agree well with the results from PHOENICS. The velocity profile seems shift in some degree. So I would like to try "plus" this term. I am not sure the place I added is right or not. When I change one to be plus and the other to be minus, it runs well and the results seems better.

[李东岳]

You cannot implement equations as you want. What you implement should obey physics. Otherwise even you get "reasonable" results, its just coincidence.

[kimy]

@东岳 I see. That is the reason why I post here. I wanna be sure that term should be implemented in which place. This is what I expect. Do you have any idea?

[kimy]

@东岳 I am using IPSA model (lauder and spalding). In all of equation used, the alpha term has been introduced inside.
http://www.cham.co.uk/phoenics/d_polis/d_lecs/ipsa/ipsa.htm#11

[kimy]

@东岳 Sorry. I am still confused that how the equation(24) mentioned in your page could be solved in OpenFoam. In the source code:

        solve
        (
            pEqnComp1() + pEqnComp2() + pEqnIncomp,
            mesh.solver(p_rgh.select(pimple.finalInnerIter()))
        );

why pEqnComp1() + pEqnComp2() + pEqnIncomp ranther than - pEqnComp1() - pEqnComp2() + pEqnIncomp. Because in your equation(24), pEqnComp1() + pEqnComp2() in the right side while pEqnIncomp in the left side.

[李东岳]

Thanks for your feedback. The R.H.S. of the following equation should be a - sign. It was corrected.
\begin{equation}\label{comp_nablaU}
\underset{\mathrm{pEqnIncomp}}{\underbrace{\nabla\cdot\left(\alpha_\rd\bfU_\rd+\alpha_\rc\bfU_\rc\right)}}=\underset{\mathrm{\color{red}{-}pEqnComp1\color{red}{-}pEqnComp2}}{\underbrace{-\frac{\alpha_\rd}{\rho_\rd}\frac{\rD\rho_\rd}{\rD t}-\frac{\alpha_\rc}{\rho_\rc}\frac{\rD\rho_\rc}{\rD t} }}.
\end{equation}

What is the first term? Its a second order tensor which is not consistent with the second term (scalar).

[kimy]

@东岳 So if I wanna add the term related to the gradient alpha in the continuity equation, in which place should I put the term?

[kimy]

@东岳 Thanks, Dongyue. I am still confused that why I add the same term in the pEqnComp1 and pEqnComp2, or pEqnIncomp leads to the different results?

[李东岳]

So if I wanna add the term related to the gradient alpha in the continuity equation, in which place should I put the term?

You should implement it in the alpha equation.

[kimy]

@东岳 Sorry, to my knowledge alpha equation dosn't exsist in the twophaseeulerfoam? How can I realize it?

[李东岳]

alpha field was solved by MULES algorithm. You can add something after MULES procedure as follows:

solve(fvm::ddt(alpha) - fvm::laplacian(D, alpha) = fvc::ddt(alpha);

[kimy]

@东岳Thanks. however MULES is not implemented in twophaseeulerfoam. As mentioned before, the term must be added into the continuity equation rather than alpha equa, which is the Spalding's IPSA solver in PHOENICS.

[kimy]

@kimy Sorry, I mean where is the MULES in the twophaseeulerfoam, which I did not find but I found MULES in the terminal during calculation.

[李东岳]

alpha was solved in twoPhaseSystem.C