(535g) PSRI’s Web-Based Benchmark Data Repository for Validating Computational Fluid Dynamics

Modelling fluid-solid-particle multiphase flow, such as fluidized beds, with computational fluid dynamics (CFD), is the focus of significant research as the availability of reliable model predictions would impact increase revenue with better designed, optimized designs, and better troubleshooting tools. Currently, the lack of consensus on the validity of physical models is one of the biggest hurdles with the application of modeling fluid-solid-particle multiphase flow unit operations.

This study reviews the current state of CFD as applied to multiphase flow and the experimental data available that has been available for validation. Previous studies have demonstrated reasonably accurate validation of CFD models; but, under a specific set of conditions (material properties, boundary conditions, flow regime, etc.). These same models have provided less accurate results for a different set of material properties or operating conditions. Specifically, changes to various physical models, such as the drag model formulation, were necessary to achieve an accurate comparison against measurements reported under varied flow regimes or using different material properties. A possible reason for the inconsistency is many commercial CFD tools lack of available comprehensive validation data sets need to move such modeling efforts from interpolation to extrapolation. Nor should it, as little data exist that include a wide range of system sizes, and fluidization regimes for a gamut of materials. What is missing is a validation study that compares the accuracy of CFD simulations with a specific set of physical models to predict the flow of the same materials in different experimental systems and/or different materials in the same experimental system.

To fill this data gap, PSRI is launching a new, publicly accessible, web-based repository for measurements from PSRI fluidized bed experiments. The main goal of this work is to provide experimental data such that CFD models can be validated against the measurements for the same material in various experimental units and various materials in the same experimental unit. Users will be able to access experimental data for more than 30 different materials with at least five flow regimes (operating conditions) per material. Instructions to access the experimental data will be provided. Moreover, details of the experimental units, materials, and operating conditions used to generate the benchmark data will be presented.