Hızlandırılmış Monte Carlo ışını izleme tabanlı radyasyon ısı transfer çözücüsünün geliştirilmesi

The Monte Carlo ray tracing (MCRT) is a semi-analytical stochastic method for the solution of radiative transfer equation (RTE). However, its high computation cost prevents its extensive use. In this study, an accelerated MCRT based radiation heat transfer solver has been developed that is capable of handling complex geometries in participating and nonparticipating medium with the capability of simultaneous 3d visualization of the results. Use of the graphic processors resulted in a 40x speed-up as compared to the modern high-end CPU's and an additional 10x speed-up is also achieved with the use of efficient data structures. The RTE is coupled with the energy equation in a way that the solver treats predefined critical regions very precisely while the remaining regions are solved with mediocre accuracy. Using this adaptive technique in the coupling of the equations, the error in the calculations has dropped by 5x compared to the standard MCRT method. The proposed method requires 1.5x more memory compared to the standard MCRT and does not have any penalty in terms of computational time.