Configuration design of thermoelectric modules via topology optimization

Abstract

The structural configuration of a thermoelectric module crucially dictates the overall performance of a thermoelectric generator (TEG), which can be significantly enhanced through geometrical optimization. However, conventional approaches predominantly rely on the dimensional scaling of predefined geometries, such as standard rectangular or cylindrical thermoelectric legs. Aiming to maximize the output power of the module, this study employs a density-based topology optimization method to autonomously evolve the geometric configuration of the thermoelectric legs. Furthermore, the performance of the optimized design is subsequently benchmarked against modules with common conventional configurations. The results demonstrate that, under an identical volume constraint, the topology-optimized structure exhibits a wavy profile characterized by a concave midsection and widened ends. This non-uniform cross-sectional geometry achieves a global optimum by optimally balancing the trade-off between local thermal and electrical resistances. Consequently, the topology-optimized design outperforms conventional thermoelectric leg configurations in terms of output power, reaching a maximum of 0.488 W—a 2.3% enhancement compared to the baseline rectangular counterpart. These findings substantiate the effectiveness of the topology optimization approach in the structural design of thermoelectric generators.