Deep Analytics | Case Study
Improved Recovery of Spilled Oil
Deep Analytics was hired to develop novel geometries for drum skimmer surfaces that are more effective for recovering spilled oil from water, specifically for thin slicks of oil. This is an ongoing contract R&D effort that has relied heavily on mechanical design.
DA developed a multiphase technical approach in which scale drums with innovative surfaces were evaluated in tanks with controllable levels of oil and sea water. Phase I required rapid prototyping of many small-scale drums, featuring a wide array of potential surfaces and materials inspired from available literature. These drum designs, as well as the apparatus, were drafted and optimized in 3D-CAD software. The drums and multiple components of the experimental apparatus were 3D-printed, via SLS and FDM processes, both in-house and out.
In phase II, we integrated features of successful drums into a streamlined selection of larger test drums, then tested in yet another, larger custom apparatus. Development led to surface designs complex enough to tap out the resources common 3D-CAD software. To solve this issue, DA used Python to model the critical surface mathematically and create a 3D-printable version. Using mesh-editing software, the surface was combined with connecting features designed using 3D-CAD.
The phase-II drums and test tank were each large enough to require external manufacturing. The drum sections were 3D-printed in industrial SLS machines, and the tank was fabricated by a local shop using DA’s fabrication drawings. In addition, DA designed and sized the necessary pumps, piping, and instrumentation to supply and control oil replenishment.