A novel comprehensible non-intrusive sensitivity-driven additive Aerodynamic Shape Optimization (AASO) and its implementation using the Lattice Boltzmann Method (AASO-LBM)

Shape optimization is a relevant topic in many fields such as e.g. fluid energy harvesting, passive mixer design or pressure loss reduction in channels. Although the literature is rich in applications of surrogate-based, adjoint-based or topology-based optimization methods, there are no methods for easy non-parametric non-intrusive optimization. Also, some methods such as the adjoint shape optimization allow only minor shape modifications and are rare in unsteady flows. In this paper, a comprehensible sensitivity-driven Additive Aerodynamic Shape Optimization technique (AASO) is proposed, which aims to optimize iteratively the shape of an object by aggregation/removal of small to large-sized pieces of material to areas where they impact the most. This method can be applied to steady and unsteady flows without any code modifications. In this work the Lattice Boltzmann Method (LBM) is combined with the (AASO) technique to deal with irregular geometries. This leads to the here named Additive Aerodynamic Shape Lattice Boltzmann Method for optimization (AASO-LBM). The method has been successfully tested for the optimization of an object in a laminar channel (in both steady and unsteady regimes), which is representative of many applications such as micromixing or bladeless microturbine design.