Fluid flow characterisation in randomly packed microscale porous beds with different sphere sizes using micro-particle image velocimetry

2020 
Abstract An experimental study using pressure-drop and μ -PIV measurements was undertaken to better understand the fluid flow characteristics in different flow regimes within randomly packed microscale porous beds with different sized spheres. Three sintered glass samples were made with glass spheres having a mean diameter of 170 μ m, 430 μ m and 710 μ m by slightly sintering them between two glass slides, forming a sample with a sandwich structure. Four different regimes, pre-Darcy, Darcy, Forchheimer and turbulent were identified in each sintered glass sample using the pressure-drop measurements. The permeability increases with glass sphere size and so does the Reynolds number corresponding to each flow regime boundary. It was found that for a given Re, the pressure drop in the sample with 170 μ m diameter spheres can be ten times higher than the pressure drop in the sample with 710 μ m diameter spheres. Four different pore geometries were identified to be the focus of the measurement zones of the μ -PIV, which were taken across a range of Re spanning all the flow regimes identified in the pressure-drop measurements. The non-dimensional time-averaged velocity distribution was found to be similar in each flow regime for the samples with 170 μ m and 430 μ m diameter spheres, whereas it changed dramatically for the sample with 710 μ m diameter spheres. In general the velocity profiles through the channels within the porous media were found to be near-parabolic, especially in the Darcy and Forchheimer regimes, but in the turbulent regime inertial effects such as localised jets were observed. Detailed observational and statistical analysis of the velocity distributions highlights their very strong dependency on the local geometry with highly localised regions of flow apparently in a different flow regime to that of the bulk flow. However, the global average of the fluctuations throughout the measurement zone does align well with the pressure drop measurements.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    64
    References
    3
    Citations
    NaN
    KQI
    []