دانلود رایگان مقاله کاربرد دیود حرارتی پلتیر در پمپ حرارتی مگنتوکالری

Abstract

The major hurdle for commercialization of room temperature magnetic refrigeration is the inadequate power density of a device due to the low operating frequency. Limitation in the heat transfer rate imposed by solid-fluid convection is the primary cause. Applying Peltier thermal diodes in combination with microchannel heat exchangers has been investigated as a possible solution. This study improves in realism upon earlier work, mainly by extending the calculations with a modeled thermoelectric effect. After reverse engineering the Micropelt MPC-D701 Peltier module, behavior of a 2D single-stage device is examined, consisting of two Peltier modules with heat exchangers enveloping a thin layer of magnetocaloric material. The near optimal switching frequency is determined and performance characteristics are calculated for several configurations of varying load, field strength and fluid velocity. A better performance is observed without the magnetocaloric effect due to reduced heat leakage through the passive thermal diode. This behavior extends to multiple devices in series where active magnetic regeneration is induced. Two possible solutions were explored, but these did not show significant improvement in device performance when applying the magnetocaloric effect.

5. Conclusion and recommendations

Contrary to the work published by Tomc and others, this article shows a less optimistic outlook on application of thermoelectric thermal diodes in a magnetocaloric heat pump. The expected improvement in operation frequency and therefore power density is observed, but this happens in conjunction with an unwanted side effect. Heat leaking through the passive thermal diode causes a reduction in the temperature span when comparing configurations with and without the MCE enabled. In other words, a heat pump solely based on the Peltier effect shows better performance than one employing both the MCE and the Peltier effect. Next to a slightly larger temperature span, the COP differs by an order of magnitude caused by the absence of magnetic work input. The above statement is based on purely numerical work with numerous assumptions, so the first and foremost recommendation is to validate if this conclusion also holds experimentally. The numerical model itself can be improved by incorporating a thermodynamically sound implementation of the magnetocaloric effect based on the mean field theory. The power density at a realistic temperature span (e.g. a household fridge) can be determined by increasing the number of devices in series and investing more computational time. Thermal diodes of some form are essential for successful commercialization of magnetocaloric refrigeration. Based on the results of this article, Peltier thermal diodes (i.e. Micropelt MPC-D701) have lost some of their potential. Thermal switches employing geometric isolation, by for example electrowetting, might be a promising alternative (Cha [23]).