COMPARING FISH-INSPIRED RAM FILTERS FOR COLLECTION OF HARMFUL ALGAE

In this work, several different bioinspired filter geometries are proposed, fabricated, and tested in a flow tank. A novel approach is explored that mimics how filter-feeding fish efficiently remove small food particles from water. These filters generally take the form of a cone with water entering the large end of the cone and exiting through mesh-covered slots in the side of the cone, which emulates the rib structure of these filter feeding fish. The flow in and around the filters is characterized and their ability to collect algae-scale, neutrally-buoyant particles is evaluated. Filter performance is evaluated by using image processing to count the number of particles collected and studying how the particles are deposited on the filter. Results are presented in the form of particle collection efficiencies, which is a ratio of particles collected to the particles that would nominally enter the filter inlet, and images of the fluorescent particles deposited on the filter at different time intervals. The results show little sensitivity to the filters' inlet geometries, which was the major dfference between filters tested. Comparative results are also presented from a 2D CFD model of the filters generated in COMSOL. The dfferent geometries may differentiate themselves more at larger Reynolds numbers, and it is believed that a fluid exit ratio, or ratio of inlet area to exit area, is the most critical filter parameter. Field testing has demonstrated collection of real algae (i) with this bioinspired filter, and (ii) from a robot platform, but using a more conventional plankton net. The larger vision is to develop these filters and mount them on a swarm of autonomous surface vehicles, i.e. a robot boat swarm, which is being developed in parallel.