CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Divergence detected in AMG solver: temperature -> Increasing relaxation sweeps! (https://www.cfd-online.com/Forums/fluent/224343-divergence-detected-amg-solver-temperature-increasing-relaxation-sweeps.html)

mohammadreza_hj February 14, 2020 11:13

Divergence detected in AMG solver: temperature -> Increasing relaxation sweeps!
 
Hi everybody. i have a question about this error message in the fluent, what does this mean? and How can I eliminate it?

# Divergence detected in AMG solver: :confused Increasing relaxation sweeps!
# You may try the enhanced divergence recovery with (rpsetvar 'amg/protective-enhanced? #t)
Divergence detected in AMG solver: temperature temperature limited to 5.000000e+03 in 75705 cells on zone 7 in domain 1
Error at host: floating point exception
Error at Node 0: floating point exception
Error: floating point exception
Error Object: #f
Registering Udf,

vinerm February 14, 2020 13:07

Divergence
 
What this means is that the case is unable to converge. This could be due to multiple reasons, such as, bad mesh, improper material properties, boundary conditions, or source terms. This could also be due to numerical schemes and time-step (if case is transient).

mohammadreza_hj February 15, 2020 10:37

Quote:

Originally Posted by vinerm (Post 758183)
What this means is that the case is unable to converge. This could be due to multiple reasons, such as bad mesh, improper material properties, boundary conditions, or source terms. This could also be due to numerical schemes and time-step (if the case is transient).

well, thank you. I'm studying material science and I'm simulating the heat transfer and material flow in the friction stir welding.
if you know, it's a steady-state process and I'm working in two cases.
in case 1 I consider the welding tool and the results is this error message.
in case 2 I ignore the welding tool and it is ok.
what do you think about this?

vinerm February 15, 2020 15:20

Source
 
Most likely the reason is source. Since the case without the heat source works alright, rest of the parameters should be alright. Welding tool generates a lot of heat. In mathematical terms, this reduces the diagonal dominance of the matrix and solution diverges. In physical terms, heat is too much to be handled with current schemes and properties. The solution is that you start providing heat slowly. If by the welding tool you mean a heat source, then start by using 10% of this source and then slowly increase it to 100%, may be over 2000 iterations.

Gbeatz February 16, 2020 04:49

under relaxation factor
 
Quote:

Originally Posted by mohammadreza_hj (Post 758170)
Hi everybody. i have a question about this error message in the fluent, what does this mean? and How can I eliminate it?

# Divergence detected in AMG solver: :confused Increasing relaxation sweeps!
# You may try the enhanced divergence recovery with (rpsetvar 'amg/protective-enhanced? #t)
Divergence detected in AMG solver: temperature temperature limited to 5.000000e+03 in 75705 cells on zone 7 in domain 1
Error at host: floating point exception
Error at Node 0: floating point exception
Error: floating point exception
Error Object: #f
Registering Udf,

when i once had almost a similar problem, i played around with the under-relaxation factors in solution controls

vinerm February 16, 2020 05:25

URFs
 
You can play with URFs as well. This will have similar effect as using smaller source value.


All times are GMT -4. The time now is 00:32.