Recovering waste heat through thermoelectric (TE) technology is critical for enhancing energy efficiency. However, commercial devices often require n- and p-type thermoelectric units to have similar cross-sectional areas and conductivities, complicating optimization of their figure of merit (ZT). Moreover, real-world heat sources are not constant temperature sources, and only heat flux can be measured. Given these constraints, selecting appropriate parameters for TE devices to maximize their output power for specific heat sources is of significant scientific and practical importance. In this work, we studied a series of n- and p-type Bi2Te3-based TE materials, prepared using commercial hot extrusion techniques, with average conductivities of approximately 650 S cm-1, 820 S cm-1, 1050 S cm-1, and 1300 S cm-1 over the temperature range of 50-250 °C. These materials were assembled into four different thermoelectric devices (HE1, HE2, HE3, and HE4). Finite element simulations for devices HE1, HE2, HE3, and HE4 (TE leg: 1.4 mm × 1.4 mm × 1.6 mm, device area: 40 mm × 40 mm) showed maximum power densities of 3953.5 W m-2 and 1300.8 W m-2 at 67,000 W m-2 and 33,500 W m-2 heat fluxes, respectively. The fabricated devices achieved maximum power densities of 2907.1, 3014.0, 2910.9, and 2525.4 W m-2 at 67,000 W m-2 heat flux, with discrepancies attributed to interface thermal resistance. Based on these findings, we have plotted output performance roadmaps for TE devices with different average electrical conductivities under various heat flux densities (67,000 W m-2 and 33,500 W m-2) as functions of external load and TE leg height. These roadmaps enable rapid design of TE devices tailored to specific external heat source conditions and cost requirements, thereby maximizing waste heat recovery. This study's use of electrical conductivity as a performance indicator for final device output power density offers direct guidance for practical waste heat recovery applications.
Read full abstract