[ar.miraliakbar]

Hello everyone,

I would like to ask for help for my thesis validation simulation where I am reproducing the results of Tsai and Wu. The reference is given below in the link:

An efficient micromixer based on multidirectional vortices due to baffles and channel curvature

I have added the passive scalar as a representation of molar fraction to the simpleFoam solver and ran 2 simulations for Re = 1 and Re = 81 for comparing the results with the experimental ones. Every thing is Ok for the Re=1 but for the Re = 81 the passive scalar is diverging despite the convergence of U and p. Based on my calculation and the data given in the article the Re = 81 is equal to Normal Velocity Inflow velocity of 0.606 m/s.
I did the meshing with Ansys Mesh module with hexahedral elements of 4 million cells.

I would like to know what would be the main cause of the divergence?
how can I diagnose my simulation?

I have attached the related information:
Monitoring Passive Scalar with probe.JPG

Mesh.JPG

Residuals.JPG

Velocity Monitoring with probe.JPG

[numb]

I also calculated this case.
I think your case is diverged, maybe is fvScheme is wrong. I try some divSchemes for div(phi,C), some Schemes is convergent, but some others is diverging.
And I also have several other problems.
1. I'm confused with the inlet velocity. Only the Re number at the inlet is proposed in this paper. The velocity is missing. I also calculated the inlet velocity is 0.606 with the method in this paper. But the max velocity and is the degree of mixing totally different with this paper if u=0.606. And I change some other velocity and try to find the correct inlet velocity. And I find if the velocity at the inlet is 0.81, then the max velocity and the degree of mixing is same with the result in the paper. The result is same as the paper when Re = 9. But when I change the Re to 1, the degree of mixing in my simulation is only about 0.6 rather 0.2 in this paper.
2. And I tried some other divSchemes, but the scalar field can't be limited between 0 to 1 although I have added "bounded" before the Scheme. The scalar field in my simulation is about -0.3 ~ 1.3.
Have others simulated this case? Or someone know how to limited the scalar field?
The first figure is from the paper, and the second one is my simulation result when I set the inlet velocity as 0.81m/s