An automatic approach for the multi-objective shape optimization of micro gas turbine heat exchangers is presented. According to the concept of multidisciplinary optimization, the methodology integrates a CAD parametric model of the heat transfer surfaces, a three-dimensional meshing tool, and a CFD solver, all managed by a design optimization platform. The repetitive pattern of the surface geometry has been exploited to reduce the computational domain size, and the constant flux boundary conditions have been imposed to better suit the real operative conditions. A new approach that couples cold and warm fluids in a periodic unitary cell is introduced. The effectiveness of the numerical procedure was verified comparing the numerical results with available literature data. The optimization objectives are maximizing the heat transfer rate and minimizing both friction factor and heat transfer surface. The paper presents the results of the optimization of a 50 kW MGT recuperator. A final optimal primary surface shape has been obtained, which differs remarkably from the usual geometries found in literature. It improves all the objectives, in particular, pressure drop and, more slightly, cost, without loss in wall heat flux. It must be noted that such results have been obtained with constant hydraulic diameter and, consequently, constant mean flow velocity, and Reynolds number. The values of these parameters were chosen according to best practice design rules. The design procedure can be effectively extended and applied to any industrial heat exchanger application.
