Divergence detected in AMG solver: temperature
 

Hi,

Answers on this forum have helped me a lot to get started with Fluent, and thank you for that! I hope someone can help me with my simulations, as I keep getting this temperature divergence instability.

To summarize my simulation:
I want to simulate 2D axisymmetric transient conduction, with constant conductivity and heat capacity, and a heat generation that is a function of temperature.

Boundary conditions:
Bottom edge is adiabatic, sides have free convection, and top edge has constant heat flux of about 10^7 W/m2. Radiation is also included, with external emissivity of 1 and temperature of 300 K.

For the heat generation, I interpret the following UDF:
#include "udf.h"

DEFINE_SOURCE(ecomb,c,t,dS,eqn)
{
real source;
real temp = C_T(c,t);

int A = 29.8e18;
int B = -12779.78;

source = A*exp(B/temp);
dS[eqn] = -A*B*exp(B/temp)/(temp*temp);

return source;
}

For the initial conditions, I just assume 300 K temperature. For RMS I have 10^-6, and timestep is 10 us.

I would like to add an emissivity of 0.9 for the material, but I can't find where to enter this property.

I have tried different numerical solvers and initialization, but soon after starting the simulations I get a temperature divergence.

Any ideas what could cause this? Thank you in advance.
Also, can I solve this problem with an UDS equation? I'm just interested in the temperature increase with time.

any advice on what I should do, or what i should change in my model?
i have tried lower heat flux on the boundary, even refined the mesh but no luck

got it to work... seems that fluent wants the x-axis to be the axis of symmetry

hello Florin Saceleanu
how you solve your problem with Divergence detected in AMG solver: temperature error?