Mid Sweden University

Wireless sensor network (WSN) nodes cannot directly used by the traditional power grid, a most common way to supply power to WSN nodes is to use batteries. However, the battery has a disadvantage that is the limited lifetime, adding an energy harvester to extend the lifetime is a good way. The thermoelectric generator is a good choice used for indoor condition since the temperature difference is easy to get in indoor condition, and a higher power density can be obtained, which can reach to 60 μW/cm2 at 5 C gradient. The thermal energy generator (TEG) consists of the peltier element and the heat sink. The aim of this thesis is to model a TEG in COMSOL Multiphysics and study the characteristic of thermal energy harvesters and the influence of environment variables on thermal energy harvesters in indoor condition. The results of the thesis show that higher air flow speed at the same temperature gradient or bigger temperature difference at the same air flow speed both can lead to obtain better output performance of the TEG with 1.6 ohm purely resistive load. Using heat sinks also can lead to obtain better output performance of TEG. Different heat sink geometries have been investigated and it is shown that the square angled fin has better performance for cooling down than squared fin and pin fin when comparing the influence of different heat sink structures . Compering the influence of different air flow directions, the direction of air flow has negligible influence on the pin fin heat sink, but has great influence on the square angled fin heat sink and the squared fin heat sink. Compering the influence of different heat sink materials, the copper heat sink has an almost negligible influence on performance for cooling down than the aluminum heat sink, it is obvious that it has been improved by 4%. For the pin fin heat sinks investigation, with the same size of pin in the same area, the air gap between two pins properly increases, the cooling down performance of pin fin heat sink can be improved by 30%.