Multiphysics Model And Performance Optimization For Miniature Thermoelectric Module

Researches on different types of renewable energy and new energy have become a hot spot in recent years because of environmental protection and efficient use of energy. The thermoelectric conversion is one of the researches, and it is the technique to convert thermal energy into electrical energy, or convert electricity into heat by means of the thermoelectric effect of semiconductor materials and has been gained widely attention. Thermoelectric devicesdo not include fluid substances and moving parts driven by compression and expansion of the fluid substances, and therefore they do not create pollution problems and have no damages to the environment. The application of the thermoelectric conversion will be very wide, and it is necessary for the industrial application of this technology to carry out the basic theory research on thermoelectric conversion so as to optimize structures of thermoelectric devices for better performance.Thermal resistance model and one-dimensional heat conduction model have been developed for the thermoelectric performance investigation, however, there are some key issues needed to be solved due to limitations on the use of the models. For example, constant or temperature-averaged materials properties can only be used and the heat loss cannot be involved, and, more importantly, the current density is not worked out and the coupling effect of the temperature field and electric potential field is not taken into account. Therefore, it is necessary to develop a general, three-dimensional, steady-state and transient models for a thermoelectric unit/device with coupling of the temperature field and the electric potential field, by means of theory for electron and hole transport in semiconductor thermoelectric materials and deep analyze on thermoelectric effects involved in the semiconductor thermoelectric technology. All the effects related to the thermoelectric technology, such as Fourier heat conduction, the convection and radiation heat, Seebeck effect, Peltier effect, Thomson effect, Joule heating effect, and effect from variable properties of semiconductor materials.The detailed analysis was made for the impacts of variable properties of materials and heat loss on the thermoelectric unit/device performance for dynamics response characteristics of thermoelectric unit/device in variable operation conditions and the variable external environment. The quantitative comparison was done between the three-dimensional multiphysics model and the thermal resistance model for the accuracy of predicting the thermoelectric unit/device performance in this paper. The comparison showed that the conversion efficiency of the thermoelectric generator from the prediction of the thermal resistance model was overestimated by up to39%. The reason for the overestimation of the thermoelectric performance was given based on the internal mechanism of the heat conduction. The preference criteria of thermal protection design were presented from the ultimate performance of the thermoelectric unit/device. The mechanism behind undershoot phenomenon and phenomenon overshoot in cooling capacity or in output power were revealed from the fundamental of thermoelectric effects.The geometrical structure of thermoelectric device has a strong impact on its thermoelectric performance. For the previous researches on a single parameter, the optimization of a parameter was fulfilled in the conditions of seeking the optimal value for the parameter and keeping values of other parameters unchangeable. However, all of the parameters have their impacts on the thermoelectric device performance. The optimal performance of the thermoelectric device cannot be obtained only by a single optimal parameter. Therefore, a multi-objective and multi-parameter optimization should be developed so that the real optimization of the operating conditions and structure can be accomplished for thermoelectric devices. For this purpose, the three-dimensional multiphysics model is treated as a direct problem solver and the simplified conjugated-gradient algorithm is treated as an inverse problem, both of them are conjugated through improving interface technology. The output power of a thermoelectric generator can increase by2.69to8.93times when the multi-objective and multi-parameter optimization is performed.There are distinct differences in the performance among different thermoelectric units when there is non-uniform temperature in heat source forms of a thermoelectric device/system. Results from research on a single thermoelectric unit are insufficient to indicate the overall performance of a thermoelectric device/system. Therefore, the research on modeling the thermoelectric device/system must be carried out. By putting the fundamental point of departure for improvement of overall performance of a vehicle thermoelectric generation system, a practical thermoelectric generation system for the vehicle exhaust waste heat recovery was introduced. The vehicle exhaust was considered as a real form of the heat source and a water-cooling heat sink was used as the cold source. In this way, the number of thermoelectric unit pairs and optimization of thermoelectric unit arrangement under the conditions of two external constraints were mainly discussed when there was non-uniform temperature in the heat source.