Economic Aspects of the Magnetocaloric Energy Conversion - Magnetocaloric Energy Conversion: From Theory to Applications - page 367

Fig. 9.3 Comparison of the
speci c costs for a 50-W c
compressor and the equivalent
ideal magnetic refrigerator

Table 9.4 The mass-production cost estimation of the magnet assembly in Fig. 9.3
Material

Eurokg − 1

Mass
(kg)

COST
(Euro)

Manufacturing cost of
assembly

Total cost
(Euro)

NdFeB 40 6 240 +25 % 300

Soft iron 5 22.2 111 +25 % 139

AMR
LaFeCoSi

60 0.60 36 / 36

475

NdFeB 40 6 240 +25 % 300

Soft iron 5 22.2 111 +25 % 139

AMR Gd 150 0.66 99 / 99

538

comparison with the whole magnetocaloric device, a frequency of operation of
above 3 Hz will actually be required. At this frequency, the magnetic refrigerator
will have to operate with an ef

ciency at least as high as a compressor refrigerator,
as well as with an equivalent temperature span between the heat source and heat
sink (i.e. 40 K or more). Note again that the 40 K temperature span with the
predicted same or higher ef ciency of the magnetic refrigerator represents an ideal

case. This also holds true for the frequency of operation. For such an ideal case the
AMR should actually consist of a layered magnetocaloric material and produced as
a structure with very tiny plates (a honeycomb or similar ordered structure) with
thickness of, e.g. 30

100 ʼ , porosity of about 30 45 %, and with a similar thickness

- -

of equally sized fl uid voids. For frequencies above 5 Hz, using water as the heat

transfer uid in such an AMR will also not be a good option.

Note also that any complication related to the magnet structure will increase the
costs. In Fig. 9.2 , the magnet assembly applies simple rectangular prisms of per-
manent magnets. If we wish to make such magnets more complex, the

fl

rst question
is of course why is this required? For instance, we have already shown in the
chapter on AMRs (Chap. 4 ) that the homogeneity of the magnetic

eld is perhaps