Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Magnetic heating and refrigeration:A new technology of heat and cold production
University of Applied Sciences of Western SwitzerlandYverdon-les-Bains, Switzerland
Peter W. Egolf
Journée du CUEPE 2006Geneva, 6th of April 2006
Thermique Industriel et Systèmes:O. Sari, N. Erbeau, J. Brulhart,M. Zaghoudi, G. Wyss, G. Chollet,K. Martin-Drezet
Simulation des Systèmes Thermique: P.W. Egolf, F. Gendre, A. Kitanovski, D. Ata-Caesar, H. Bajpai, K. Winkelmann
The Divisons TIS and SIT
ExperimentsTheory
Collaboration and Help
• Prof. Dr. O. Sari• F. Gendre• N. Erbeau• N. Alber• F. Brun• J. Hu
University of Ljubljana
Prof. Dr. Alojz PoredosDr. Andrej KitanovskiDr. Alen Sarlah
New name:
Table of contents
• Brief history of magnetocaloric materials and systems• The magnetocaloric effect• Magnetic heating and refrigeration at room temperature• Existing prototypes• Magnetic forces and thermodynamic potentials• Possible systems• Porous rotary magnetic refrigerators• Thermodynamic cycles, cascades, regeneration• Numerical method and simulations• Alternative machines• Magnetic heat pumps for Switzerland?• International activities• Conclusions and outlook
„Giant“ magnetocaloric effect
New materials ... may bring practical magnetocaloric cooling a step closer. A large magnetic entropy change has been found to occur in MnFeP0.45 As0.55 at room temperature, making it an attractive candidate for commercial applications in magneticrefrigeration.
Nature Highlights10 th of January 2002
„Giant“ magnetocaloric effect
In an article published in the journal Nature last month, scientistsat the University of Amsterdam reported that they had created an iron-based compound that also exhibits a large warming effect in a magnetic field. Iron and other ingredients in the compound areconsiderable less expensive than gadolinium.
The New York Times9 th of February 2002
„Giant“ magnetocaloric effect
• 1881 : Warburg discovers the magnetocaloric effect in iron
• 1926 : Debye proposes the first practical application at low tempera-tures. The process is named adiabatic magnetization/demagnetization
• 1927 : Giauque proposes the same process independently of Debye
• Since 1930 : Application of magnetic cooling becomes standard in low temperature physics (T<<1 K)
• At present : Discovery of magnetocaloric materials with Currie tem-peratures TC around room temperature
• Development of new magnetic cooling systems operating at roomtemperature: Promising new technology !
Brief history of magnetocaloricmaterials and systems
The magnetocaloric effect
For the entropy of rare earth alloys (no electron phonon interactions)a superposition of the partial entropies is possible (Tishin, 2003):
The entropy of the electronic (e) and the lattice (l) subsystemsshow no dependence on the magnetic field! Then:
Some basic thermodynamic relations and definitions:
),()()(),( THSTSTSTHS mle ++=
dTTHTsdHTHH
sdTTdTdsdTTdT
dsTHdsH
m
T
mle ),(),()()(),( ⎟⎠⎞
⎜⎝⎛∂∂
+⎟⎠⎞
⎜⎝⎛∂∂
++=
χχ
δ⎟⎠⎞⎜
⎝⎛∂= T
qcTdsq=δ ( )χχ
δTsTc ∂=
The magnetocaloric effect
Combining these equations:
With the Mayxwell relation (e.g. Kitanovski & Egolf, Int. J. Refr. 29, 3-21, 2006:
because:( ) ( )HT T
MHs
∂−=∂δμδ
0TT
mHs
Hs
⎟⎟⎠
⎞⎜⎜⎝
⎛∂
=⎟⎠⎞
⎜⎝⎛∂
δδ
dHTHHsdTT
THcdTTTcdTT
TcTHdsT
m
H
mle ),(),()()(),( ⎟⎠⎞
⎜⎝⎛∂∂
+++=
The magnetocaloric effect
and:
mleH cccc ++=
In an adiabatic magnetization or demagnetization process:
0=ds
Therefore, it follows that:
( ) ( )HHH
HTM
cTdT
TMdTT
c∂=⇒=∂− δμδμ 0
0 0
Moving magnetocaloric material into (out of) a magnetic fieldincreases (decreases) the magnetisation and the temperature.
dH
The magnetocaloric effect occurs at a phase transition betweendifferent magnetic states.
The temperature of phase change is the Curie temperature.
The magnetocaloric effect
Data from:
Pecharsky VK, GschneidnerKA Jr. Tunable magnetic regenerator alloys with a giant magnetocaloric effect for magnetic refrigeration from ~ 20 to ~ 290 K. Applied Physics Letters 1997; 70 (24): 3299-3301.
Still limited temperaturedifferences!
The magnetocaloric effect
Tegus, PhD thesis
Van der Waals-ZeemannInstitute
University of Amsterdam
2003
Magnetic heating and refrigerationat room temperature
0
50
100
150
200
250
260 280 300 320 340
T (K)
q (J
/kg)
Gd
Gd5Ge2Si2
MnAs1-xSbx
x=0
x=0.05
x=0.1
x=0.5
x=0.55
x=0.65
x=0.9
x=1
x=1.1
MnFeP1-xAsx
Mn2-xFexP0.5As0.5
Mn1.1Fe0.9P0.47As0.53
Field change: 0 - 2 TTc-12<T<Tc+12
200
300
400
500
600
700
260 280 300 320 340
T (K)
q (J
/kg)
Gd5Ge2Si2
Gd
x=0.1
x=0.05
x=0
MnAs1-xSbx
MnFeP1-xAsx
x=0.65x=0.55
x=0.5
Gd5(Si1.985Ge1.985Ga0.03)
Field change: 0 - 5 TTc-25<T<Tc+25
Heating and refrigeration „capacities“:
Zimm et al., 2001Tegus et al., 2002
Magnetic heating and refrigerationat room temperature
Magnetocaloric materials should (partly by Yu et al., 2003):
• Modest Debye temperature (a high temperature makes the fraction of latticeentropy small correspondingly at a high temperature)
• High temperature difference at phase transition• Large width of transition• No magnetic hysteresis (irreversibility of cycle)• Small specific heat and large thermal conductivity to guarantee fast heat exchange• Large electric resistance to avoid the eddy current loss• Fine molding and processing must be possible• No poison substances (arsenic, etc.) for direct contact applications
Magnetic heating and refrigerationat room temperature
Hallbach Magnet > 4 Tesla
Development of magnets
High flux line density
Magnetic heating and refrigerationat room temperature
Magnetic heating and refrigerationat room temperature
Four stage process each:
A) Magnetic heat pump:1) Increase of magnetic flux density2) Rejection of heat for heating purpose
at high temperature level3) Decrease of magnetic flux density4) Injection of heat at low level to the
heat pump
B) Conventional heat pump:1) Compression (Increase of pressure)2) Rejection of heat for heating purpose
at high temperature level3) Expansion (decrease of pressure)4) Injection of heat at low level to the
heat pump
Advantages:• Green technology (no use of conventional refrigerants)
• Noise-less technology (no compressor)
• Very high thermodynamic efficiency (20-30% higher than in conven-tional cycles due to the reversibility of the magnetocaloric effect)
• Lower energy consumption
• Simple construction
• Low maintenance costs
• Low pressures (operation at atmospheric pressure, advantage e.g. for air-conditioning applications and refrigeration systems in automobiles)
Magnetic heating and refrigerationat room temperature
• Strong magnetic fields with not completely known influences on living creatures
• Protections to avoid disturbances of electronic components
• The field strengths of permanent magnets are still limited, superconducting magnets are too expensive
• The temperature differences are still not so high
• In moving material applications the precision of the machinemust be high (small gap between magnet and magnetocaloricmaterial)
Disadvantages:
Magnetic heating and refrigerationat room temperature
Existing prototypes
A development in Japan
Chubu – Toshiba (2000)
This machine consists of a cascade with three pistons of differentmagnetocaloric materials. A low magnetic field of B=0.7 T is in-duced by permanent magnets.
Existing prototypes
AMES Laboratory - 1997
The kernel are two pistons of packed beds of gadolinium spheres. A superconducting magnet with a high magnetic field is applied.
Development in the USA
Existing prototypes
AMES Laboratory - 2001
Rotary magnetic refrigerator with permanent magnets(with complicated fluid switches)
Development in the USA
Existing prototypes
University of Victoria
Developments in the USA
dVHMF iiK ∇∫=Γ
0μr
dVBMF iiV ∇∫=Γ
0μr
dVMHF iiV ∇∫=Γ
00μr
extensive vector analyticaltreatment
Kelvin (K):
Liu & Odenbach (V):
FVgrad H
N
S
Kitanovski & Egolf, 2005 (in print)
A. Lange, Kelvin force in a layer of magnetic fluid.J. Magn. Mat. 2001; 241: 327-329:
Magnetic forces and thermodynamic potentials
Dilute limite only (ferrofluids)!
N
S
What is the difference?We still may have the external work:
But now additionally a flux of magnetisationoccurs. Its infinitesimal change is:
Then the technical work is:
In turbo machinery this work is also namedshaft work
dMHdw ext00
)(1 μ−=
( )MHddw abs00
)(1 μ−=
( ) 00)(
1)(
1)(
1 MdHdwdwdw extabstech μ=−−=
Magnetic forces and thermodynamic potentials
pdvdw ext −=)(1 dMHdw ext
00)(
1 μ−=
Quantity Gas thermodynamics
Magneto-thermodynamics
Driving “force”(stress parameter)
Pressure p Field H0
Order parameter(reaction of system)
Specific volume v Magnetization M(orientation of spins)
External work(closed system)
(related to force of Liu)
Technical work(open system)
(related to Kelvin force)
vdpdw tech =)(1 00
)(1 MdHdw tech μ=
H0p
v M
Magnetic forces and thermodynamic potentials
p, V
dW1
S
Fluid
p.
N
S
dW1
H0
H0
M, H
Magneto-caloric
material
.
.
Possible Systems
p(1), v(1) p(2), v(2)
W1abs
W1tech
Fluid
p H0
H(1), M(1)H(2), M(2)
W1tech
W1abs
Magnetocaloricmaterial
Possible systems:Open systems
1) Static system with two storage blocks and valves2) Linear movement3) Rotational movement
Solution A: Rotary wheel, magnet fixed, axial current.
Solution B:Wheel fixed, rotating magnet, axial current.
Solution C:Rotary wheel, fixed magnet, radial current.
Solution D:Fixed wheel, rotary magnet, radial current.
Possible systems
Porous rotary magnetic refrigerator
A. Kitanovski, P. W. Egolf,O. Sari, 2003.Procédé et dispositif pourgenerer en continue du froidet de la chaleur par effetmagnetique.
PCT BR – 10’463 - IN
Four functions arerealized in a verysimple manner!
Porous rotary magnetic refrigerator
[ ][ ] [ ])3()4()1()2(
)4()1(
hhhh
hhCOPBrayton
−+−
−=
A. Kitanovski, P.W. EgolfInternational Journal of RefrigerationThermodynamics of Magnetic Refrigeration (in print):
Internal energy:
Legendre Transformation:
„Coefficient of performance“:
MdHdsTdurr
001 μ−=
MHuhrr
0011 μ+=Work
Fast process(adiabatic)
0001 ),( MdHdsTHsdh μ+=
Heating energy
Thermodynamic cycles, cascades, regeneration
To improve the available temperaturedifferences cascade systems..
Thermodynamic cycles, cascades, regeneration
...or regeneration systems are required!
Thermodynamic cycles, cascades, regeneration
Numerical method and simulations
z
rzvt
n
i
n
ii
zi
∂+∂+⎟⎠⎞
⎜⎝⎛
∂Λ∂+∂
Λ∂+∂Λ∂
∑∑== 11
)()(
1χχ
φφωρφ
&&
πφπφφ 2,),(,),( )()()( ≤<Δ+== hothotinRRhot
inFF TTLTTLT
πφφφ ≤<Δ+== 0,)0,(,)0,( )()()( coldcoldinRRcold
inFF TTTTT
LzzTzT RR ≤≤= 0),,0(),2( π
Energy conservation:
( ) 0=−+∂∂
RFF
pF TTLzTcv
ξαρ
( ) 0=−+∂∂
FRR
HR TTTc δαψφωρ
Mapping of a cylinder to a square:
(Periodicity condition)
Thermal inertia and conductivities negligible!
Numerical method and simulations
20
25
30
35
40
0 30 60 90 120 150 180
Azimuth pos ition (de gre e s )
Tco
olin
g flu
id (°
C)
00.250.50.751
Downstream position
(x/L)
20
25
30
35
40
0 30 60 90 120 150 180
Azimuth pos ition (de gre e s )
Tw
arm
whe
el (°C
)
00.250.50.751
Downstream position
(x/L)
10
15
20
25
30
180 210 240 270 300 330 360
Azimuth pos ition (de gre e s )
Tch
illed
flui
d (°C
)
00.250.50.751
Downstream position
(x/L)
10
15
20
25
30
180 210 240 270 300 330 360
Azimuth pos ition (de gre e s )
Tco
ld w
heel (
°C)
00.250.50.751
Downstream position
(x/L)
Temperature distributions in a steady state condition: L=0.2 m, D=0.198 m., vcold=vwarm=6 m/s, ω=0.6 s-1,=704 W, mwheel=8.09 kg, ΔB= 5 Tesla. Material: Gd5(Si1.985Ge1.985Ga0.03) Q&
Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Q&Withoutoptimi-sation:
COP=5.8
Azimutal position (°) Azimutal position (°)
Azimutal position (°) Azimutal position (°)
Newest work:Analytical optimization
( ) )()(
)(
1
)(
)(,
1
)(
)(
,1121 )(
coldR
hotR
hotR
CarnotCarnothotR
hotinF
hotF
hotR
TTTCOPCOP
TT
eCOPhot
F
−=⎟
⎟⎠
⎞⎜⎜⎝
⎛−−=
≤≤
−
4342144 344 21
χ
χχ
More simple (also analytical) models:
1,
,1
<<==
==
vLf
Svc
hSt
SStL
cfh
rHR
R
r
R
HRR
ρ
δψρδψχ
vchSt
NTUStcv
h
pFFF
FpFF
F
ρ
ξρ
ξχ
=
=== ,
Numerical method and simulations
Verticalslabs arevery bad!
Ferromagnetic kernel
Numerical method and simulations
Alternative machines
Second generationof a rotative proto-type with regeneration.
Alternative machines
A system withradial flow isbeing built andwill be presen-ted at the Hann-over industrialexhibition 2006.
Heat pumps for new buildings with floor heatingFluids: Brine/water (brine: 25 % ethylen-glycol)Temperature primary side: 0 °C (e.g. water of a river)Temperature secondary side: 35 °CPower: 8 kW
Renovation building with radiator heatingFluids: Air/waterTemperature primary side : -7 °CTemperature secondary side : 50 °CPower: 15 kW.
Magnetic heat pumps for Switzerland?
Conventional heat pump application Analogous design of a magnetic heat pump
Simple magnetic heat pump system with only two pumps. A requirement is identicalfluids in the two circuits, e.g. water/glycol.
Magnetic heat pumps for Switzerland:New building with floor heating
The system shows an air/waterheat exchanger, a pump in thesmall heating cicuit and two airventilators.
Schematic drawing of a complete heating system with a magnetic heat pump. By the heat recovery system therequired temperature difference becomes smaller and a magnetic heat pump appplication becomes realistic.
Magnetic heat pumps for Switzerland:Renovation building with radiator heating
New IIR Working Party on Magnetic Refrigeration
See on the web:www.iifiir.iifiir.org
First Int. Conf. on MagneticRefrigeration at Room Temperature
27-30. September 2005in Montreux Switzerland
140 participants
Interested:Coca Cola, Néstle,Daewoo, Danfoss, Peugeot, Citroen, Axima, Arcelik,…
Next conference:Protorosz, Slovenia11-13 April 2007
Swiss Technology Award 2006First prize
Conclusions and Outlook
• Magnetic refrigerators at room temperature have been designed(and several patents for different systems submitted)
• The basic theory has been established, but will be further prepared
• First numerical simulations results have been made available
• Experimental validation is planned to be performed
• Optimisation calculations are very important to reach high COP values and are still missing
• More powerful systems as cascades and systems with regene-ration must be designed, their behaviour measured and modeled.
Thank you for your attention!
Magnetic coolingat room temperature:
A future major evolution in refrigeration?