LED dedoming has been an important advancement for flashlight enthusiasts because it significantly increases the luminance of the LED which increases the throw. See here for an explanation. However, the mechanism for the luminance increase is not completely understood. For example some LEDs (219C, XP-G3) do not exhibit as high a luminance increase as others (XP-G2, XM-L2, XP-L) upon dedoming.
These measurements do not necessarily answer the above question, but I think these measurements are interesting and could be useful to explain or predict results when slicing (incomplete dedome).
For these measurements I used a XPL V6 1A at 4.0A and measured the lux 38cm directly above the LED. I then sliced small thicknesses of the dome with a razor blade and measured the removed silicone with calipers. After each slice I took a picture from above with the LED slightly lit and used this to estimate the active lit area of the LED. To estimate the area I selected the lit area (using photo editing software) and used the 3.5mm x 3.5mm base as a reference. There was significant uncertainty in this method, which is why there are error bars on the luminance graph below.
This graph shows the luminous intensity of the LED as a function of the cumulative silicone thickness removed. The lux goes way down upon the first slice because the apparent area decreases. Then as the silicone is made thinner, the lux increases even though the area is roughly constant, which means the luminance is increasing.
The last point is with the LED dedomed. I used the warm gas method, but I’ve found a lot of the XPL V6 1A emitters I’ve had recently do not dedome well and leave a layer of silicone that is well-attached to the phosphor. I ended up carefully scraping the silicone away.
This graph shows the LED luminance, which is the luminous intensity divided by the LED area.
Here is the data.
After the first slice, the LED looked like this:
Just the top of the dome was sliced off. The LED is unmagnified since the top of the silicone is flat, but some light from the edges of the dome is still refracted towards the detector. These lit areas were included in the area calculation.
