Active Magnetic Regeneration - Magnetocaloric Energy Conversion: From Theory to Applications - page 145

AMR cycle. Furthermore, in the Hybrid AMR cycle, the regeneration factor
remains almost constant for a wider range of operating frequencies, whereas in a
Brayton-like, and especially an Ericsson-like AMR cycle, the regeneration factor
drastically decreases with the increased frequency. However, at low operating
frequencies (i.e. a longer duration of the

ow period), the Hybrid AMR cycle
approaches the regeneration factor of the Brayton-like and Ericsson-like AMR
cycles.
By changing the working regime of the device and performing the Hybrid AMR
cycle, the performance of the test device was improved (e.g. for the maximum no-
load temperature span and consequently the regeneration factor) compared to the
Brayton-like AMR cycle. A similar conclusion with respect to combining the
process of (de)magnetization and

fl uid fl

fl uid fl ow was also obtained by Bj ø rk and

Engelbrecht [ 18 ].
Next, the experiments were performed at a xed frequency of the operation

(f = 0.37 Hz) and with a constant utilization factor (U = 0.3). The cooling load was
varied in order to obtain different temperature spans, related to a particular AMR
thermodynamic cycle (Fig. 4.31 ).
The experimentally predicted COP of the AMR cycle was calculated using the
numerical programme [ 16 , 17 ] as the ratio between the experimentally obtained
cooling power and the sum of the numerically calculated input work, which can be
further divided into the work required to pump the fluid and the magnetic work (in
this particular case, the related mechanical losses were neglected). The magnetic
work input was calculated as the integral of the performed thermodynamic cycles in
the T - s diagram. A detailed description of the experimentally predicted COP can be
found in Plaznik et al. [ 9 ]. For the particular case of Fig. 4.31 , the ratios between the
magnetization period and the

fl uid fl ow period were the following: Brayton-like

AMR ( ˄ mag : ˄ f = 0.75:0.5); Ericsson-like AMR ( ˄ mag : ˄ f = 0.75:1.25); Hybrid

( ˄ f = 0.75:0.87).

Based on the results of Fig. 4.31 , we can estimate that the cooling performances
and the maximum temperature span (under no-load conditions) of the Brayton-like
and Hybrid Brayton

˄ mag :

Ericsson-like AMR cycles are similar. On the other hand, the

-

Fig. 4.31 The experimentally
obtained maximum
temperature span and COP as
a function of the cooling load
for a parallel-plate gadolinium
AMR ( 0 H = 1.15 T,

U = 0.3, f = 0.37 Hz)

ʔʼ