An Enhanced Monte Carlo Method to Calculate View Factors in Packed Beds and its Application Using Distance Approximation

This paper proposes a novel method based on Monte Carlo Ray Tracing for high-precision and low time computation processing of view factors in random assemblies of cylindrical packed beds. The Monte Carlo Ray Tracing method is improved by coupling it with Kowsary’s tangent sphere method under the parallel computation processing through Compute Unified Device Architecture, reducing the computational time in approximately 153 times less than a traditional Monte Carlo Ray Tracing method and obtaining a maximum relative error of 0.39% for configurations evaluated in the literature. Furthermore, a calculation methodology for view factors between particle-particle, particle-wall (considering a discretized wall), and particle-lid is presented and applied on a set of randomly assembled monosized packed beds generated with the LIGGGHTS discrete element method software, covering an average porosity range from 0.38 to 0.52 for the arrays. Finally, detailed analysis and discussion of the obtained results are performed, allowing to find characteristic view factors for the particles according to their positions and correlations for view factors particle-wall and particle-lid as a function of particle-surface distance, considering error intervals for each interaction respectively, defining a view factor calculation methodology based on approximation distance for packed beds within the studied porosity range