ERI: Simulation and modeling of polydisperse gas-solid flows

Oakland University, Rochester MI

Investigators

Abstract

Multiphase flows, such as the motion of solid particles suspended in a gas, are present in many industrial and natural systems. Two examples are the flow of ash and debris following a volcanic eruption and the flow inside circulating fluidized bed reactors, which can be used to convert agricultural waste into biofuels. In both examples, the solid and gaseous phases interact with each other. This interaction is strong enough to alter large-scale behavior, which, in turn, affects key quantities of interest, such as the speed at which volcanic ash propagates or the chemical conversion efficiency in a reactor. Predicting the dynamics of these flows is challenging, because they contain particles with widely variable properties such as size, composition, and shape. This project will use data-driven modeling methodologies to (1) quantify how large-scale behavior of gas-solid systems are affected by variations in particle size; (2) develop a novel, data-driven approach for formulating predictive models of these types of flows; and (3) develop a K-12 outreach demonstration that illustrates how high performance computing can be leveraged to understand and predict multiphase flows that are important to society. Strongly coupled, gas-solid flows occur in a wide range of industrial and natural systems. In systems with substantial mass loading, momentum coupling between the phases gives rise to large-scale heterogeneity, which directly affects quantities of interest such as chemical conversion of biomass, heat transfer between phases, and settling time. Despite the importance of such flows, accurately predicting their behavior at relevant scales remains challenging. As a result, engineering decisions are frequently made based on over-simplifications such as idealized mixing or uniformity in the particulate phase. The objective of this project is to advance knowledge of polydisperse, gas-solid flows and close the modeling gap that exists in characterizing and predicting their behavior at relevant scales. In particular, this work will (1) leverage high-fidelity Euler-Lagrange simulations to quantify the relevant terms that dominate polydisperse behavior, particularly with application to variable particle size; (2) introduce a novel modeling framework capable of distilling complex data into optimally simple and accurate models; and (3) propose models for polydisperse gas-solid flows that are robust across relevant scales and easily interpretable and integrated into existing solvers. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.