I do understand the situation where the outer wall temperature is fixed. The color gradient difference in the wood is not very large, but when inspected with a probe, there is a temperature difference between the inner and outer surfaces of the wood. This means that heat transfer occurs between the fluid and solid domains in that area, even though at a quick glance it may seem otherwise. The effect of natural convection initially appears not to work, but on closer inspection, the temperature above the aluminum domain is higher than below it, so natural convection works at least somehow.
What puzzles me is why the K-residual doesn't converge. Can the result still be correct even if not all residuals converge? I refined the mesh for greater accuracy, but I didn’t notice any significant difference. Could it be that the calculation model is not suitable for analyzing such a small fluid domain, or is it really the case that natural convection behaves like this in reality?
I also created a model with the same parameters but with a larger fluid domain, where natural convection behaves at least in my opinion “correctly”. In this case, the denser and lighter air masses clearly redistribute under the influence of gravity, and the K-residual also converges now.
You are navigating away from the AIS Discovery experience