Discussion Forum

Hi everybody, I want to calculate the temperature variation in a building with partially transparent walls caused by irradiation by the Sun. For this I used the Heat transfer with surface-to-surface radiation module with “surface-to-surface radiation” and “radiation in participating media” activated. As a simple example, I constructed a cylindrical glass plate underlain by an opaque cylindrical plate, with a gap in between. This ensemble is illuminated from above by the Sun (using the “External Irradiation” feature). Top and bottom surface of the transparent plate were declared as “Incident intensity” to avoid that they are opaque. The example (see attachment) produces solutions, but they appear to be unphysical: 1) First, I would expect that with a very small absorption coefficient, most of the (visible) radiation from the Sun would not be attenuated by the glass plate and thus the second plate below should be illuminated with almost the same intensity as the top glass plate. This seems not to be the case in the calculation. The result indicates that even with a very small absorption coefficient of the upper plate (with this the transmissivity of the glass plate should be practically 100%?) and a scattering coefficient of zero nothing of the solar irradiation hits the lower (opaque) plate. It seems that the lower plate is shadowed by the upper plate “as if” the upper plate would be also opaque, at least on the lower side. 2) Another strange result is that the “incident intensity in W/m^2” along the central axis of the upper plate is much higher than the surface irradiation in W/m^2 on the top surface of the glass plate. This also seems unphysical. This incident irradiation intensity inside the glass plate should always be smaller than the maximum irradiation intensity at the top surface, I would expect.

**Can anyone tell me what is wrong with my model? **

I also tried to use the Heat transfer with radiation in participating media module with “surface-to-surface radiation” and “radiation in participating media” activated, but it makes no difference. Moreover I have tried all offered radiation transfer methods (DOM, P1, Rosseland), but this also does not change the strange results. For the spectral band in the Heat transfer with surface-to-surface radiation module, “solar and ambient” was used. Thus, the light source radiates most of its power in the visible spectral range. In the application library I could not find an appropriate example, I started with the “glass plate” model and extended it accordingly.