Search

This work presents the design and performance evaluation of a novel magnetic refrigeration unit capable of controlling the temperature of a 31-bottle wine cooler cabinet. The main components of the system are: an active magnetic regenerator composed of eight porous beds of Gd/Gd-Y spheres, a two-pole magnetic circuit responsible for the magnetic field changes, a set of solenoid valves for flow control, and fan-supplied finned tube heat exchangers on the cold and hot sides of the wine cooler apparatus. Synchronization of the magnetic field variation and fluid flow processes was performed using real-time magnetic field measurements. The system was capable of reaching a mean temperature of 10.8 o C inside the retrofitted wine cooler cabinet by generating 27.9 W of cooling capacity for an ambient temperature of 25 o C. The coefficient of performance and second-law efficiency for this operating point were estimated as 0.21 and 1.0%, respectively. When operating at a cabinet temperature of 12.5 o C, the coefficient of performance and second-law efficiency achieved by the magnetic system were 0.38 and 1.6%.

Void (dead) volumes represent a penalty to the thermal performance of Active Magnetic Regenerators (AMR). While much effort has been put in designing AMRs with small void volumes, in most designs, penalties associated with inlet flow maldistribution and pressure loss may increase with the reduction of the void volumes. In this work, an experimental evaluation of the combined effect of void volume and inlet flow maldistribution is carried out. The analysis is performed based on the regenerator thermal effectiveness determined experimentally in a passive regenerator test apparatus. The results show that the negative impact of inlet flow maldistribution on the regenerator effectiveness is more significant than that of the void volume.

The development of efficient active magnetic regenerators (AMR) is highly dependent on the regenerative matrix thermal performance. Matrix geometries should have a high thermal effectiveness and small thermal and viscous losses. In this study, we present a systematic experimental evaluation of three different regenerator geometries: parallel-plate, pin array and packed bed of spheres. All matrices were fabricated with approximately the same porosity (between 0.36 and 0.37). The cross sectional area and length of the regenerator beds are identical, resulting in the same interstitial area. Hence, any difference in performance between the matrices is due to interstitial heat transfer between the solid and the fluid and losses related to thermal, viscous and magnetic effects. As a means to quantify these losses individually, experiments were first conducted using stainless steel matrices without the application of a magnetic field (passive regenerator mode). Later, gadolinium matrices made with the same characteristics as the stainless steel ones were evaluated in an AMR test apparatus for which experimental results of cooling capacity, temperature span between the thermal reservoirs, coefficient of performance and second-law efficiency were generated as a function of utilization for different operating frequencies. Parallel plates had the poorest performance, while the packed bed of spheres presented the highest cooling capacity. On the other hand, the packed bed also had the highest viscous losses. For this reason, the pin array exhibited the highest COP and second-law efficiency.

In this first part of a two-part paper, a new active magnetic regenerator (AMR) laboratory apparatus is presented and evaluated. The setup is composed of a nested Halbach cylinder magnetic circuit (maximum magnetic flux density of 1.69 T) assembled in phase with a double effect displacer that provides the cold and hot blows to the regenerator. A single packed-bed regenerator with 195.5 g of gadolinium spheres is used in a discontinuous (i.e., reciprocating) cycle. The system performance is evaluated in terms of characteristic curves (i.e., cooling capacity as a function of temperature span), coefficient of performance and second-law efficiency as a function of the utilization factor and operating frequency. Maximum values of COP have been identified for a given temperature span. The maximum values of second-law efficiency were obtained at system temperature spans between 15 and 20 K.

In this second part of a two-part paper, a mathematical model of active magnetic regenerators is applied to identify and quantify the main losses taking place in the AMR evaluated experimentally in Part I. Among those losses, the heat interaction with the external environment and the presence of dead (void) volumes between each end of the regenerator and the hot and cold heat exchangers were found to be the main factors that affect the AMR performance. Demagnetizing losses were considered as a function of the matrix geometry, temperature and applied magnetic field. In addition to predicting the time-dependent behavior of the fluid temperature exiting the regenerator during each blow and the cycle average cooling capacity, the model was able to quantify the impact of each loss mechanism on the thermal performance of the AMR.

An experimental and numerical assessment of multilayer active magnetic regenerators (AMR) composed of gadolinium (Gd) and gadolinium–yttrium (Gd–Y) alloys (Gd[Formula: see text]Y[Formula: see text], Gd[Formula: see text]Y[Formula: see text] and Gd[Formula: see text]Y[Formula: see text]) is presented. First, by calculating the adiabatic temperature change and the isothermal entropy change from the experimental data for the above materials, we show that, with reasonable accuracy for engineering design purposes, these properties can be determined by shifting the properties of pure Gd to the Curie temperature of the Gd–Y alloy — a common but not yet validated assumption in the design of Gd–Y AMRs with a low Y content. Next, we show that the optimal Gd–Y layer fraction in multilayer AMRs can be determined using the figure of merit known as the material refrigerant capacity (RC), which agrees well with the results from a more complex one-dimensional thermal non-equilibrium porous medium AMR model. Finally, the performance of the latter model is verified against the experimental cooling power data for two- and three-layer Gd–Y regenerators at temperature spans of 25, 30 and 35[Formula: see text]K.