9
NATURAL CUBIC INTERPOLATING SPLINE FOR THE HEAT
CAPACITY OF GADOLINIUM
B. Siddikov
Ferris State University
Abstract: A time-dependent one-dimensional model of the Active Magnetic
Regenerator (AMR) that takes into account most of the physical and practical design
problems for the AMR is given as a highly nonlinear system of partial differential
equations. The accurate approximation function for the heat capacity of the
magnetic material (gadolinium) is obtained by using the natural cubic spline and the
least squares curve fitting techniques.
Key-Words: magnetic
refrigeration, heat capacity of gadolinium, numerical
simulation, natural cubic spline
1
Introduction
Magnetic Refrigeration (MR) is
rapidly developing and becoming
competitive with conventional gas
compression
technology, primarily
because
the
most
inefficient
component of the refrigerator – the
compressor – is eliminated. In
addition, MR operating near room
temperature
provides
important
environmental benefits. MR uses a
solid magnetic material as the cooling
source and water (perhaps with
antifreeze additives) as the heat
transfer medium. There is no need to
use volatile
chemicals with potential
environmental problems.
MR
is
based
on
the
magnetocaloric effect, where a
magnetic
material
changes
its
temperature
with
variations
of
magnetic field. One of the key
components of MR is the Active
Magnetic Regenerator (AMR), which
produces
refrigeration without gas
expansion by using the magnetocaloric
effect. An AMR cycle consists of four
operations:
bed
magnetization,
warming of
the magnetic material;
fluid flow from cold to hot reservoirs
through the bed, transferring heat to
the Hot Heat Exchanger, HHEX (this
semi-cycle is called the Hot Blow
Period); bed demagnetization, cooling
of the magnetic material; fluid flow
from hot to cold reservoirs through the
bed, and absorption of heat at the Cold
Heat Exchanger, CHEX (called the
Cold
Blow
Period)
[1,
2].
Regeneration occurs during fluid
flows. A schematic of the AMR is
illustrated in Figure 1.
Figure 1: Schematic of
the AMR
To predict operating characteristics
and improve the design of the Active
Magnetic Regenerative Refrigerators
(AMRR) it is desirable to develop
mathematical models for AMR and
accurate,
stable numerical solvers of
the model.
10
In [3, 4], we developed
numerical scheme for the model [5] to
obtain a computer simulator of AMR.
One of the difficulties in this work is
obtaining a dependable approximation
function for the heat capacity of the
magnetic material (gadolinium). In [3,
4], we used the least squares curve
fitting
technique to obtain the
approximation function for the heat
capacity of gadolinium. Further
research in this field indicated that we
need more accurate approximation
function for the heat capacity of
gadolinium to improve the simulation
results.
This paper reports on
obtaining of such a highly accurate
approximation function.
Do'stlaringiz bilan baham: