Speeding up of Gas-Particle Flow Simulations Using Non-Iterative Time Advancement (NITA) Solver

Kinetic theory based Two-Fluid model (TFM) for gas-particle flow simulations is routinely used from initial conceptual designs to improving the efficiency of the particle handling equipment. There has always been a demand for simulations to be faster and scalable, especially with recent advances in high performance and cloud computing. In this study we demonstrate the considerable speed-up of gas-particle flow simulations using the NITA solver and scalability of TFM over a large number of compute cores. The National Energy Technology Laboratory (NETL) small-scale fluidized bed, bubbling fluidized bed and circulating fluidized bed challenge problems were used to test the speed-up of the simulations using NITA solver. It was observed that the NITA solver both qualitatively and quantitatively captured distinct gas-particle flow patterns as predicted by the classical iterative time advancement solver while speeding up the simulations by as much as a factor of six. The scalability of kinetic theory based TFM was tested using NETL circulating fluidized bed problem and the model achieved excellent scalability over a large number of compute cores.