[g_beka]

Hello all,

We are a team that is struggling to deal with a fire simulation in a room for a course in our Master's program.

After the 600th iteration some cells tend to hit 5000K temperature making the simulation and the temperature plots unrealistic.

I attached the Udf file to this thread. Seems that heat flux is the main reason.

any ideas?

Of course our boundary conditions are set to heat flux while the outer walls are convection. Solution is transient. And the professor of the course has already checked the model + materials + all the settings and it runs smoothly for the first seconds. Then some cells ascend to 5000 K for no reason. The heat is coming from from a surface.

Thank you for your time.

[EmmTheof84]

I had a similar problem with a heat transfer simulation in the past.
The issue was the time step. As your temperature rises, I assume buoyancy forces increase and you have stronger convection currents.
If your time step is not small enough to capture those variations then you will see increased temperature in some cells. You should try reducing your timestep after the 600th iteration and see if this fixes things
I haven't done fire simulations though so I can't give you more guidance.

[AlexanderZ]

your heat flux is time dependent

heatflux = 250000*(time/150)*(time/150)/0.25;

do you think, is it possible, that heatflux has become huge enough to raise temperature to 5000K?

what is actual time after 600 iterations?

also i recommend you to put line

time = RP_Get_Real("flow-time");

out of face loop