Help! Total heat transfer rate doesn´t sum to zero
 

Hello,

i´m modelling a natural convection and radiation (do) problem. After 6000 Iterations (steady case) the residuals have flatten down (continuity =1e-2, the velocity, k, epsilon are at about 1e-4, energy and discrete ordinate 1e-6). But the sum of the "total heat transfer rate" for all boundaries is far away from zero. It is at about 5000W. My inlets and outlets have a total heat transfer rate between 4000 and 6000W. Is this a indicator that my solution isn´t total converged? After 17000 iterations it´s still the same problem. After that I built up a small problem to investigate this behavior. The result: The total heatflux sum to zero.

I would be pleased for any kind of help! Thank you!