Design of segmented off-diagonal thermoelectric generators using topology optimization

A density-based topology optimization methodology is used to optimize the off-diagonal figure-of-merit and off-diagonal electrical power output of thermoelectric generators by distributing two different thermoelectric active materials in a two dimensional design space. Off-diagonal thermoelectric generators are characterized by converting a vertical thermal heat flux into a horizontal electric current, and may be useful in applications where the electrodes connected to the generator are prone to thermo-mechanical stress and wear. With basis in the topology optimization framework and a sequence of numerical examples, we discuss critical and important model parameters such as objective functions, heat transfer rates and device lengths. All results are supported by exhaustive crosschecks and validations, and it is shown that the off-diagonal figure-of-merit and the off-diagonal electrical power output may be improved by 233% and 229%, respectively, compared to other optimization approaches available in the literature.