Calculation of absorption efficiency

Hi all,

I confront a problem about calculation of absorption efficiency. Just one film of gold is used to test.

I put one surface above the film, then set the surface "power outflow, time average (emw.nPoav)" to calculate the absorption of gold. Next, the gold is abandoned to set the surface "power outflow, time average (emw2.nPoav)" to calculate the incident power. This equals to a reference. Then I can get the absorption efficiency easily when "emw.nPoav" divided by "emw2.nPoav". This method works well when the incident angle is normal.

However, when I turn to get absorption efficiency as a function of wavelength and the angle of incidence, the absorption efficiency reaches around 50% (refer to the attached picture). As we know, there should be almost no absorption for ideal gold. So I am afraid if this method works well for different angle. Is there other effective method to calculate the absorption efficiency?

PS: I find that the absorption efficiency calculated through this method is above far away from 100% when I calculate other complex structure. That's why I use just gold to test.

Thank you.