Computational design and performance assessment of microvascular composite panels for battery cooling

Batteries for electric vehicles require active cooling to extend their lifespans and protect them from structural damage. To satisfy these requirements, conventional battery packaging consists of layers of aluminium to provided cooling and fiberglass to provide protection. With the goal of reducing weight while providing both cooling and superior damage protection, a new battery packaging scheme utilizing microvascular composite panels has recently been proposed. In these panels, complex 2D or 3D micro-channel networks can be embedded using a novel manufacturing technique based on a sacrificial fiber or template approach. Because most heat is dissipated via the coolant, the micro-channel design in the composite is crucial to the cooling performance of the panel. In this work, we develop a non-uniform rational B-spline (NURBS)-based interface-enriched generalized finite element method (IGFEM) for the thermal analysis of the composite. The thermal impact of the micro-channels is captured using a simplified model, which is validated with experiments and verified with FLUENT. By combining our IGFEM solver with a gradient-based optimization algorithm, we optimize the parallel channel design of various network topologies defined by the number of branches. A recently developed adjoint method and sensitivity analysis is applied to speed up the shape optimization. The optimal designs obtained using these techniques are shown to have superior performance to the reference network configurations.