Divergence in AMG-solver (enthalpy and p1)
 

Hi everyone!
I simulate a steady turbulence flow combustion flow with the non-premixed combustion.
state relatation is equilibrium and its non-adiabatic
energy-model on
turbulence model ist k-omega (SST).

the simulation works fine and had a good convergence.
the i activated the P1-Radiation model.

for the mixture i set absorptions coefficient = wsggm-domain based.
scattering coefficient = 0
refractive = 1

the thermal wall boundary conditions i set to:
mixed: with free stream temp = 100°C
external and internal emmisivity = 1
and transfer coefficient = 5 w/m2 K

the feed inlet has a temp of ca 1000°C and the oxidator of 800°C

as i said the simulation converged good but with activated the p1-radiation
the comes the message after aproxx 25 iteration steps
divergence detected in amg-solver: p1
divergence detected in amg-solver: enthalpy

thank you for help!

Hi

Can you solve your problem ?
I got the same problem like you.
Please share me if you solved

Thanks

hi
yes got the problem solved.
set the absorptions coefficient to wsggm-cell based, but i think this was only minor change.

But i decreased the under-relaxation factors. especially for p1.

now i calculate with these factors:

pressure: 0,3
density: 0,8
momentum: 0,4
turbulent kinetic energy: 0,7
turbulent dissipiation rate: 0,7
energy: 1,0
P1: 0,5

the first factors are so low, because i have i high swirled turbulence flow with a mixture.
i think it is important that you set the factor for p1 at beginning so low. So the enthalpy produced from p1 can build up slowly.

when the residuals got stabil, you could enhance the factors.

Hope i could help you! Good Luck