insufficient precision on multigird level
 

Hello, I have a warning about the insufficient precision on multigird level.

WARNING: insufficient precision on multigrid level 1, nRows = 11374
AMG coarsening halted.
This may indicate double precision version is needed.
Temperature limited to minimum value 931 times on model-part "Body 1"


After several times iterations then I have got an error and the Simulation stops.

Error: AMG solver diverged. A floating point exception has occurred: floating point exception [Invalid operation].


Can anyone tell me how to inprove the precision of multigrid Level?
I am using multiphase segregated flow to simulate an Interphase reaction.
Thanks!!

Is your mesh sufficiently fine and how's the quality. I had faced such error previously. Try to split region by non-contagious. If it doesn't split then it is OK.Check other mesh quality parameters. Such error may cause when solver is not robust (new physics). Need details of problem for further comments.

:):cool:

Ignore the AMG solver warnings, this is your problem. Your case is blowing up. Check the mesh and boundary conditions. Where is the temperature limited as the output says above?

Hello Ashok, thanks for your reply.
I checked my mesh, it contains 79920 cells, I am not sure if it fine enough for a 0.0015m3 region?
The lowest cell quality is 0.4, and other mesh quality parameters don't indicate there are bad cells in my mesh.
The simulation is about a porous region with a heterogeneous reaction in laminar flow. Following models are selected:
• Cell quality remediation
• Eulerian multiphase
• Gradients
• Laminar
• Multiphase equation of state
• Multiphase interaction
• Multiphase segregated flow
• Phase coupled fluid energy
• Steady
• Three dimensional

Thank you for reply. How can I confirm where is the temperature limit?
In my simulation there is a reactor heated up by an inner tube, and I set the outer wall of the heating tube constantly 320 degrees. So I assumed it leads to the temperature limit?