Emulating Filter Feeding Structures in CFD to Efficiently Separate Oceanic Particles

Traditional filtration methods used to separate solid particles from liquids typically have high energy consumption, or are plagued by high pressure drop or fouling that requires regular intervention or filter replacement. In contrast, many freshwater and marine organisms, e.g. paddle fish and manta rays, have evolved to effectively remove suspended particles from water in energy-efficient ways. These ram-feeding organisms rely on intricate feeding structures that induce flow separation and vortex formation at small scales to facilitate particle retention. Termed cross-step filtration or ricochet filtration in the literature, these highly-structured gill rakers have proven quite effective at filtering particle-laden flows with low pressure drop while also avoiding clogging. Here, we investigate the effect of the shape of filter feeding structures through computational fluid dynamics (CFD) in the pursuit of the design of an efficient oceanic particle collection device. Filter feeding structures inspired by cross-step and ricochet filtration geometries are simulated using turbulent multiphase flow models. The shape and orientation of the structures, as well as their scale relative to the particles, are investigated across a range of Reynolds numbers and their ability to concentrate suspended particles are reported. The accuracy of the simulation results are compared to preliminary particle tracking experiments.