Lighting simulations using smoothed LED profiles compared with measured profiles

Ray tracing simulations of LED lighting systems typically use the smooth angular intensity profiles supplied by the LED manufacturer. However, measurements of a range of 5 mm LEDs presented in this paper show bright regions in the LEDs' angular distributions. The intensity patterns and bright regions vary between different LEDs, even when the measured angle for 50% integrated light output (as measured using an integrating sphere) is similar. When non-diffuse or partially diffuse optical elements such as clear light guides are part of a lighting system design, this source profile unevenness is intensified, so that bright caustic rings are formed. We have performed lighting simulations using coloured LEDs coupled into a clear light guide, and compared the light output using smoothed LED profiles with that using actual measured profiles. The simulated light patterns projected from the end of a light guide onto a screen are compared with that obtained by experiment. It is shown that the uniqueness of individual LED beam patterns needs to be taken into account for simulation accuracy. This is particularly important when the lighting system combines the output from several LEDs. It is also shown to be crucial in optimising the amount and type of diffusion required for homogenising the light distribution.