Hi CFD community

I have been working of a project to model the flow in a equipment wit a domain similar to a pipe, but with obstacles inside of the pipe. This equipment is unique because the diameter is 1.2m and a lenght 4.5m.

As far as I know, I have not seen any work that try to perform a CFD analysis for such a large domain.

This domain offers to problems:

First the demand of cells in order to obtain a iterative convergence to a residuals less than 1e-03, and at the same time represent the flow field accurate. Second difficulty to obtain grid independence.

If someone believe that can help keep reading the description of the simulations.

Two mesh have been designed to attack the problem the first one very coarse, the second one fine in the areas of expected high velocity differences.

the first mesh provided a iterative continuity residual of 1e-04, where as the second mesh provided just a residual of 1e-02.

I have try improve the mesh till the point to mantain progresive distribution of the cells. However, I still having the same problem.

It seems that the PDE's ignore the obstacle an can predicted the flow field fro the coarse mesh. However, when the mesh is refiend the PDE's do not ignore the obstacle and the model can not handle the velocity diferences between adjanced cells.

If someone can help me

Thank you in advance

Regards

Alex Munoz

Grid independence depend on your mesh system,you can mesh your model in servel size,maybe the first one is very coarse,and the size of it is 15K,the next will be 30K,and the third is 60, ... ,the last one maybe is 200k.where the difference between the two continuouse object,you get the Grid independence.