Optimizing the laser diode ray tracing model for LERP system simulation based on likelihood image sampling

Ray tracing software is a powerful tool for optical system analysis that provides built-in laser source models. However, these models prove to be inadequate for laser beams with a factor of M 2 far from one, as is the case for high-power multi-mode laser diodes. In certain applications, accurate prediction of the beam shape can be critical to simulation model performance; in opto-thermal simulations of laser-excited remote phosphor (LERP) systems [1] , [2] , overestimation of the high intensity region of the beam can lead to overestimation of the phosphor temperatures calculated. For this reason, a ray tracing laser diode source model was developed based on random sampling from the weighted (likelihood) image of the experimentally measured beam profile.