Efficient Heat Exchange and Radiation Simulations using Coarse-Grained CFD-DEM
Please login to view abstract download link
Coupled computational fluid dynamics and discrete element method (CFD-DEM) [1] simulations are essential to predict complex interactions in particulate systems, including fluid flow, heat transfer, and chemical reactions. They play a critical role in industrial process optimization, although challenges remain related to (i) computational efficiency, (ii) model accuracy, or (iii) coarse graining to allow large-scale systems to be simulated. Our present study introduces three key improvements to CFD-DEM simulations: (i) a heat exchange limiter to stabilize heat transfer predictions, (ii) a view factor correction for radiative heat transfer in coarse grained simulations, and (iii) mesh refinement guidelines for the well-known P1 radiation model. Our enhancements allow for more accurate and efficient simulations of particle-fluid systems. Specifically, the introduction of our heat exchange limiter allows the CFD time step to be chosen O(104) higher at a minimal increase of the computational cost per step. This limiter does not affect the prediction quality in our use cases, and is easy to implement in existing CFD-DEM tools. Also, we demonstrate that coarse graining introduces significant errors in traditional view factor calculations. The correction developed by us for view factors ensures that the radiative heat exchange rate calculated between parcels (and between parcels and walls) is consistent with classical (i.e., non-coarsened) results. Finally, we demonstrate that the mean free radiation path sets the Eulerian grid resolution in coarse grained CFD-DEM simulations that use the P1 radiation model [2]. To allow such simulations in an industrial context, adequate smoothing algorithms must be used to decouple the parcel size from the Eulerian grid size. We highlight that combining smoothing and adequate grid size, excellent predictions of the P1 model in dense fluid-particle systems can be achieved.