I am currently trying to model a surface to surface radiation problem to see the change in view factor between two geometries; however, I first attempted a very simple control case. The case was just radiation between two parallel, finite rectangles. I just set two temperature boundary conditions.
In this case, there exists an analytical expression for the view factor. I modeled the expression in MatLab and it showed me that the change in view factor when I move from 10 um in gap size to 100 nm would be 15% of the value at 10 um.
Since I defined the temperature of the two planes in comsol, a simple thermal resistance network analysis can reveal that for two similar materials (the case here), that the change in radiative heat flow (W) would be directly proportional to the change in view factor. However, when I do the simulations in comsol for different gap distances and integrate the heat flux (W/m2) over the surface for total heat flow, I get a change of 21%, which is an error of 40% compared to the analytical solution.
Has anyone run into the same problem before? This makes me doubt any results I calculated in my previous models that involved radiation.
In this case, there exists an analytical expression for the view factor. I modeled the expression in MatLab and it showed me that the change in view factor when I move from 10 um in gap size to 100 nm would be 15% of the value at 10 um.
Since I defined the temperature of the two planes in comsol, a simple thermal resistance network analysis can reveal that for two similar materials (the case here), that the change in radiative heat flow (W) would be directly proportional to the change in view factor. However, when I do the simulations in comsol for different gap distances and integrate the heat flux (W/m2) over the surface for total heat flow, I get a change of 21%, which is an error of 40% compared to the analytical solution.
Has anyone run into the same problem before? This makes me doubt any results I calculated in my previous models that involved radiation.