Selection of magnetic materials for an active magnetic regenerative refrigerator

Abstract

High efficiencies of active magnetic regenerative refrigerators (AMRR) are strongly dependent on the correct selection of the magnetic regenerator working material. The selection process involves the thermodynamic analysis of the AMRR cycle to determine the adiabatic temperature change profile (ΔT versus T) over the temperature span of interest for an ideal material, and then the matching of real materials whose magnetocaloric effect (MCE) as a function of absolute temperature best fits this profile. This paper develops the calculation of the ideal magnetic material ΔT versus T profile for a real AMRR operating between 110 and 300 K and with 1 kW of cooling power. The ideal profile was a function of the constant entropy flux from the cold end heat load, the irreversible regenerator entropy production, and other real effects. To accommodate the large temperature span, several magnetic materials were chosen and layered in the regenerator from the cold to the hot end by increasing the Curie temperature. The resultant ΔT versus T curve of the combined material provided only a rough approximation of the calculated ideal material curve. To improve this approximation, physical mixing of magnetic refrigerants was investigated. This procedure diluted the magnetic moment, thereby reducing magnetic entropy available for the cycle. Further, under adiabatic conditions, mixing produced an intolerable amount of entropy during cycle execution. Simple segmentation of the regenerator with more magnetic materials that better match the ideal profile is an easier way to approximate the ideal ΔT versus T curve with real materials. Optimum segmentation will be determined by regenerator complexity.