Modeling Thermoelectric Device Enhancement of Heat Sink Performance

Increased power density is straining the ability of air-cooled heat sink technologies to provide adequate cooling for heat-generating components. Several technologies are under investigation as replacements for air-cooling. Under specific conditions, a well-selected thermoelectric device [TED] can act as an enhancement to a heat sink’s heat removal capacity or allow it to achieve lower temperatures. Such improvements to heat sink performance using a thermoelectric device are possible without increasing airflow or heat sink dimensions. Proper sizing of this kind of optimized thermoelectric system involves consideration of multiple conditions, including the amount of heat being generated, the temperatures involved (typically, target case temperature and expected ambient temperature), and available voltage and current. Although thermoelectric devices are often thought of as inefficient, with Coefficients of Performance [COP] of less than 1, a well-selected TED can have a COP of much greater than 10. Existing methods for thermoelectric optimization, for the sake of simplicity, often ignore the thermal resistance of the heat sink or ignore the effect of temperature dependence of the thermoelectric material parameters of resistivity, thermal conductivity, and thermopower. To correctly include these factors in the design of the TED, a methodology has been developed to determine an optimum device while simultaneously considering the input parameters of θ_CA (case to ambient thermal resistance), heat load, target cooling temperatures, and available DC power. The method is iterative, involving the use of given input conditions to yield an estimate for expected final temperature conditions, which are used to produce an initial estimate of the thermoelectric material parameters, which in turn are used to calculate the optimized device. The performance of this device is calculated to determine a new estimate for temperatures and material parameters. The process is repeated until convergence occurs for the device design. The methodology can also demonstrate the performance benefits of integrating a TED into an existing conventional fan/sink system, and also describes conditions that are unsuitable for the use of TED’s. Graphical representation of the information can be readily generated as an aid to design.