Reliable Thermoelectric System Design under Optimal Requirements from the Structural and Performance Perspective

Thermoelectric generation (TEG) is an eco-friendly, solid-state energy conversion technology that can directly convert heat to electricity via Seebeck effect. The structural reliability of thermoelectric systems is one of the major issues hindering the widespread adoption of this technology in terrestrial applications despite its success in space applications. Traditional design and optimization techniques were focused on performance without much consideration for reliability. The requirements for performance often contradict the requirements for reliability which complicates the design of a TEG for both performance and reliability. Currently there are no established standards for evaluating the thermomechanical reliability of TEGs in the industry. The objective of this dissertation is to investigate the factors that affect the structural integrity of TEGs and develop a metric that quantifies their reliability. For this purpose, the parameters that contribute to stresses in a TEG were identified and studied using numerical simulations. We implemented a procedure for evaluating the probabilistic thermomechanical reliability of TE materials based on the common practice in the ceramics industry which considers the statistical nature of strength of the brittle materials using Weibull analysis. Finite element simulations were carried out to study the influence parameters such as leg geometry, dimensions, spacing, metallization thickness, and processing conditions etc. on the stresses and associated reliability. The effect of design modifications such as module configuration change was also investigated.;Results showed that the boundary conditions and processing temperatures significantly influence the TE material and metallization stresses and the device reliability. The external compressive load enhanced structural reliability. Among the geometrical parameters, the leg length has a dominating influence on the stresses compared to the cross-sectional shapes. The cross-section geometry of legs showed an effect that depends on fixed/free end conditions.;This research also focused on the geometry optimization of a TEG for performance and reliability. For this purpose we incorporated design of experiments (DOE) based response surface methodology (RSM) in conjunction with FEA and optimized the geometry of a TEG for performance and reliability using desirability functions. The results from optimization clearly demonstrated the effectiveness on this methodology in determining the tradeoffs between performance and reliability.