Numerical and experimental study of the flow distribution inside slender packed beds of spherocylindrical particles

Abstract

In this paper, Magnetic Resonance Imaging (MRI) and a Pore Network Model (PNM) are used to characterize the flow in packed beds of spherocylindrical particles. PNM is chosen as it is a relatively fast numerical approach, which provides local information on the bed flow pattern. MRI scans of packed bed reactors provide detailed information on the bed structure and the flow. In this study, the packed beds are reconstructed from the MRI images. The impact of the image quality on the PNM’s flow field prediction is assessed. It is shown that improved image quality significantly enhances prediction accuracy. With a sufficient image quality, PNM is able to closely match MRI in predicting the flow fields and capture important characteristics such as wall channeling.