Learning Forum

[mert.berkman]

I am working on a transient CFX model which gives Newton method failed to converge warning at the end of each time step with status:STALLED. After some time steps the run suddenly stops (midway iterations) without much more information. Any suggestions to address the issue is welcome.

[NickFL]

We will have to know more about your problem before we can help. Are you solving a fluid flow with a mixture? Newton's Method is used when using the lookup tables. It could be that, or it could be something else. If you are simulating a mixture, does the error go away when you switch to a single material?

[C N]

Hello Mert,

I will need more information on the physics you are dealing with for the model to assist you.

Thanks,

[C N]

Hello Mert,

I need more information about this error but typically in cfx there can be convergence issues 

Occasionally you may find that jobs submitted in serial will converge while those in parallel fail. This can be due to the different internal structure of the multigrid solver. The partitioned mesh leads to different coarse mesh blocking than the serial mesh, and if you have selected a timestep size that is close to the critical convergence limit, this can cause convergence problems.

Usually a reduction in the timestep size alleviates this problem.

I hope this helps you .

Thanks,

[mert.berkman]

Yes, the fluid is a mixture of CO2 liquid and gas with a RGP table for saturation. I'll try going to a single material.

[NickFL]

Some things to think about:

1. Make sure the gas tables are computed over the potential region of interest. There could be areas high/low pressure areas where the component properties are outside where the tables have been computed. Usually it would give you a different error in addition to above, but I have seen stranger things.
2. Another thing would be if you have a recirculation area with low pressure, you could, counterintutively, coarsen the mesh in this region. Now do this coarsening only if it isn't in an area where the solution critical. The reason the coarsing would work here is that it would artificially diffuse the low pressure region over a larger number of cells increasing the minimum value.

[mert.berkman]

Thanks all for the suggestions. I'll try reducing the time step size, changing the material and may be remeshing and will provide updates. Please keep suggestions coming!