NASA Technical Memorandum 85973
NASA-TM-85973 19840019932
Design Considerations for a 3He Refrigerator for Space Applications Peter Kittel and Andres F. Rodriguez
....... /
"OT 10 !I.E ~A.lEH FROM nms lIOOM
July 1984
NI\SI\ National Aeronautics and Space Administration
- '~':4
..... :c / -~, J \: '\
111111111111111111111111111111111111111111111 NF00829
https://ntrs.nasa.gov/search.jsp?R=19840019932 2018-06-07T15:54:32+00:00Z
NASA Technical Memorandum 85973
Design Considerations for a 3He Refrigerator for Space Applications Peter Kittel Andres F Rodriquez, Ames Research Center, Moffett Field, California
NI\S/\ National Aeronautics and Space Administration
Ames Research Center Moffett Field California 94035
DESIGN CONSIDERATIONS FOR A 3 HE REFRIGERATOR FOR SPACE APPLICATIONS
Peter Kittel and Andres F. Rodr1guez*
Ames Research Center
SUMMARY
The low temperature prov1ded by 3He refrigerators (0.3-3 K) have useful space app11cat10ns. However, the low temperatures and the low surface tens10n of 3 He require special des1gn cons1derations. These considerat1ons include the need for small pores to conta1n the liqu1d in a matr1x, the effects of bubble rucleat10n and growth, and the effects of the thermal conductivity w1thin the matrix. These des1gn considerat1ons are discussed here along with an analysis of a possible conf1nement system.
a
B o
F a
F P
F s
F t
g
h
h
k
K e
accelerat10n
Bond number
B /rh o
eJect10n number
E r/~h J ~
SYMBOLS
p~Va; force due to accelerat10n
force due to a pressure d1fference
force due to surface tens10n
force due to surface tens10n grad1ents
grav1tat1onal accelerat10n
Planck constant
Boltzmann constant
effect1ve conductance
conductance of 11qu1d
*Permanent address: Department of Phys1cs, Un1vers1ty of the Pac1f1c, Stockton, CA 95211.
K m
L
n
N
P
conductance of matrix
latent heat
number of caplllaries
number of liquid molecules
pressure
P s
F /nr2; pressure due to surface tension s
6P
Q
r
R
t
T
T c
T o
T s
T sat
6T
pressure dlfference
heat flux
total heat flux
radlus
gas constant
hold tlme
temperature
condensatlon temperature
operatlng temperature
superheat
saturatl0n temperature
temperature dlfference
volume
x dlstance 1n the dlrectlon of a
a. rh
13 Qh/r
E mass fraction lost durlng pump-down
K effective conductlvlty e
K matrix conduct1v1ty m
n volume vOld fractl0n
2
pv vapor density
o surface tension
INTRODUCTION
The development of a space-compatlb1e 3He refrigerator would provide a slgnlficant lmprovement ln several areas of research such as IR astronomy and the study of crltlca1 phenomena in quantum f1ulds (refs. 1 and 2). 3He can produce refrigeratl0n by evaporatlve coo11ng from 3.2 K (its crltica1 point) to about 0.3 K. The lower 11mlt is set by the vapor pressure and the system pumplng speed.
Whl1e 3He refrigeration has been used for a number of years in laboratories (refs. 3 and 4), on balloons (ref. 5), 1n alrcraft (ref. 6), and in spln-stabi1ized sounding rockets (ref. 7), it has yet to be used in the low-gravity enVlronment of space. Un11ke 4He , WhlCh was used ln IRAS (ref. 8), 3 He lS not a superfluid at these temperatures and cannot be contalned by the fountaln pressure generated by a temperature gradlent.
A proposed 3 He refrlgerator lS shown ln flgure 1. Whl1e thls lS not the only posslb1e conflguratlon, thls des1gn wl1l be used here as a reference 1n discusslng the major features that are also common to other des1gns. ThlS deslgn cons1sts of an adsorptl0n pump, two heat sWltches, and a pot. The pot, the pump, and the lnterconnectlng tublng form a sealed system that need be f1l1ed with 3He only once. The heat sWltches connect the refrlgerator to heat slnks whose temperatures are lndlcated 1n the f1gure. The refrlgeratl0n cycle starts wlth the system below 10 K. The pot sWltch lS closed and the pump sW1tch 1S opened. The adsorptlon pump lS then heated to drlve out the 3 He gas WhlCh condenses ln the pot. When the condensation lS complete, the heater lS turned off and the sWltches are reversed. The pump cools, then readsorbs the gas, causlng evaporat1ve cooling ln the pot. When the pot runs dry, the cycle can be repeated.
Var1atlons of thls type of refr1gerator have been used ln the past whenever gravlty or centr1fuga1 force ensures that the 11qu~d remalns 1n the pot. The purpose of th1S paper 1S to d1SCUSS varlOUS aspects of contaln1ng the 3He In space. If the pot 1S f111ed w1th a porous materlal the llqUld 3 He can be contalned by capll1ary attraction. Although surface tens10n wethods have been used for contro1-11ng other f1ulds ln space, the low temperatures and low surface tenslon of 3 He requlre specla1 conslderat1ons. The use of cap111ary conflnement 1n a 3 He refrigerator was flrst suggested by Ostermeler (ref. 9). Slnce then, Ennls (ref. 10) has demonstrated condensatl0n, conf1nement, and useful refrlgeratl0n of both 4He and 3 He ln an lnverted geometry. In many ways the lnverted operatlon lS a more severe test than a zero-gravlty operat10n.
The maln concerns wlth capl11ary conflnement are whether the vapor can be condensed into the porous matrix wlthout 1eav1ng large vOlds, and whether the liquid can be evaporated from the sponge wlthout forming bubbles that would expel liquid. Expelled 11quid would reduce the quantity of liquid avallable for cooling.
These concerns are dlscussed in the fol10wlng sect10ns, followed by a model of a proposed refrigeratl0n system. The porous matrlx wl1l be modeled as a cluster of smooth cyllndrlca1 caplilarles. These capll1arles wl11 be assumed to be parallel
3
to any acceleration of the refr1gerator. Although real accelerations can occur in any direction, accelerat10ns parallel to the capillary aX1S produce the largest effect. Deficienc1es in this model will be discussed.
We would like to acknowledge the support of the NASA Office of Aeronaut1cs and Space Technology.
CAPILLARY CONFINEMENT
The porous matr1x serves two functions during the refrigeration cycle. It prov1des a place for the vapor to condense during the condensation phase, and 1t reta1ns the llquid dur1ng the evaporat10n phase.
In order for the vapor to condense, several cond1t10ns must eX1st. The part of the system which is 1n contact w1th the vapor must be at a temperature below the critical temperature of the gas, and the pressure in the system must be greater than the correspond1ng vapor pressure. If these condit10ns are met, condensation will occur. If several areas are cold enough for condensation to occur, the condensation can be expected to occur preferent1ally at the coldest point. This effect 1S due to the vapor pressure decreas1ng and the surface tens10n increasing with decreas1ng temperature. Both the vapor pressure grad1ent and the surface tens10n gradlent provide forces that drlve the condensed llquid to the coldest point (cold spot) ln the system. Thus, at the end of the condensation phase, all of the liquld will be condensed at the coldest pOlnt unless there lS some other force, such as gravlty, to drive the fluid elsewhere.
If the cold spot is a porous matrlx, the condensatlon lS enhanced because the surface tensl0n lS lncreased by the small Slze of the pores, and because the lncreased surface area lncreases the condensation rate. However, there may be a dlfflculty when using a porous matrlx - that the condensation mlght take place only on the outside of the matrlx, leavlng vOlds in the lnterl0r. However, experiments by Donnelly (ref. 11) and EnnlS (ref. 10) have not demonstrated th1s effect. They showed that the condensed liqu1d substant1ally fills the matrlX.
In a spacecraft, the porous matrlx must hold the llquid agalnst any lateral acceleratl0ns. ThlS lS requlred durlng both the condensatlon and evaporatlon phases of the refrlgeratlon cycle. These accelerations can, in general, occur in any dlrection and can be of variable magnitude. The effect of the acceleratlon forces lS best descrlbed in terms of the Bond number, B (refs. 12 and 13). The Bond number lS the ratio of the acceleration forces t8 the surface tens10n forces:
B o
F a
F s
(1)
where we have assumed that the llquid wets the caplilary wall w1th a concact angle of 0°; so the COSlne term that normally appears ln the expression for the surface tension force can be ignored. This assumptlon 1S valid for 3He (ref. 9) and w1ll be used throughout this paper. If B > 1, the acceleratl0n forces dominate, and lf B < 1, the surface tension forces odomlnate. Thus, to retain the flu1d, the matrlx mRst be selected such that the stabllity condit10n, B «1 is met for all expected accelerat1ons. Since Pt and cr are propert~es of the refrigerant, and a is determined by the environment, the deslgner lS free to choose only rand h. Th1S restriction can be emphas1zed by separat1ng B lnto two factors:
o
4
B o (2)
where Bl = P£a/2cr gives the system properties and a = rh gives the'deslgn parameters, The stability condition can now be written as a« 1/B1 , Since Bl IS a functlon of temperature, the stabl1ity condltion must be evaluated over the whole operatIng temperature range, WhlCh is the range from the condensation temperature to the evaporat1on temperature. For the case 1n Wh1Ch a = 9.8 m/sec (1 g), the funct10n l/Bl for 3 He is shown 1n f1gure 2. For the curve 1n f1gure 2, the area below the curve 1S the stable reglon. For many applicat1ons, the acceleratlons can occur ln any d1rectlon. Thus h must be the maXlmum d1menslon of the matrlx from an open pore. ThlS would lead one to th1nk that the largest possible volume would be a sphere of rad1us h. However, the matrlx lS not floating free; rather lt lS almost entlrely conta1ned. The only openlng is in the vent tube (flg. 3) and no part of the matrlx can be a dlstance greater than h from the openlng of the tube. Thus the greatest volume lS a sphere of rad1us h. From the stability condltion (a « l/Bl), 1t lS read1ly seen that the smaller r lS, the greater h can be; and thus the more 1Iquld that can be held.
BUBBLE DYNAMICS
The porous matrIX must not only conflne the llqUld durlng the condensatlon phase, but also durlng the evaporatl0n phase. Bubbles that form durlng evaporatl0n tend to expand. At flrst they wl1l expand radlally untll they block the caplilary. If the bubble expands past thls pOlnt, It wlll dlsplace and eventually expel llquld (ref. 14). Slnce the expelled llquld wlll not produce any useful coollng In the system, lt IS deslrable to prevent the expulSIon. This problem can be broken down Into several aspects, starting wlth the Inltlal formation of the bubbles, then thelr growth and movement wlthln the caplllarles. Thls process wll1 be dlscussed In thlS sectlon along w1th effects of real nonldeallzed porous matr1ces.
In a porous matrlx, wlth ItS large lrregular surface, bubbles wlll form by Inhomogeneous nucleatl0n at nucleatl0n sltes on the Interface (refs. 15 and 16). Bubbles wlll form when the llquld temperature, T, IS ralsed sufflclent1y above the saturatlon temperature, Tsat (ref. 17). The quantlty (Ts = T - Tsat ) IS called the superheat. The amount of superheat reqUIred for nuc1eatl0n depends on the surface propertles of the lnterface. Whlle It lS dlfflcult to predlct thlS superheat exactly, an estlmate can be made. Bald (ref. 16) derlved an expreSSlon:
T s
Thls functlon lS shown In flgure 4 as a functlon of temperature for 3 He for 1 mole of 11quld.
(3)
Bubble formation can be prevented If the temperature grad1ents 1n the fluld are kept small and the superheat IS never reached. These lImItations reqUIre the effectIve thermal conductance K across the fluld to be large. In other words, the combIned conductances of theematrlx and the flUId must be cons1dered between the heat source (where nucleatIon IS most llkely to occur) and the fluid's free surface (where evaporative cooling occurs). If the heat flow IS rectIlInear, then
5
(4)
where Krn and K~ are the thermal conductances of the matrix and the 11qu1d, respectively. 6T = 6/Ke gives the temperature drop across the system which must be less than the superheat. Thus for the system to be stable against the formation of bubbles, the following relation must hold:
6/K < T e s
(5)
This formula requ1res that Ke be large. Since K~ is set by the propert1es of 3He , only Krn is a free parameter. A high-conductance matr1x, such as copper, will give the best stability. The area-to-Iength ratios 1n Krn and K£ are also 1mportanto Because the conduct1vity of copper 1S cons1derably h1gher than 3He , thickwalled capillaries will be more stable than th1n-walled ones.
If we cons1der the heat flux into a single cap11lary of radius rand helght h, then equat10n (5) can be expressed in terms of an equivalent conductivity Ke where K = K TIr 2 /h:
e e
h K TITs e
« ---Q
(6)
(The equ1valent conductivlty can also be wr1tten ln terms of the 11quid and matrix conduct1v1t1es, K£ and Km' respectively, and the v01d fraction of the matr1X: Ke K£ + Km(l - n)/n.)
If a bubble does form, it will be free to move within the 11quid. In the absence of grav1ty, wh1ch g1ves rlse to buoyancy, the only force ava1lable to move a bubble comes from surface tens10n grad1ents. Such motion 1S called Marangon1 flow (ref. 18). The surface tension grad1ents are the result of temperature grad1ents and the temperature dependence of the surface tens1on. The surface tensl0n of 3He as a function of temperature 1S shown 1n f1gure 5. Young et al. showed that the buoyancy of a bubble w111 balance the Marangoni effect when
da dT dT dx
(7)
S1nce th1s balance occurs when the surface tens10n gradient force, Ft , 1S equal to Fa for a = g, Ft can be deduced:
da dT --dT dx
(8)
Thus the bubble tends to move toward the h1gher temperature; 1.e., toward the heat source and away from the free surface of the l1qu1d which is being evaporatlvely cooled.
S1nce the bubble most likely forms near the heat source, it will not move far, if at all. Furthermore, at low temperatures da/dT approaches zero as does Ft. Thus at the operating temperature there w111 be no force to move a bubble. At higher temperatures there is another effect that prevents the bubbles from mov1ng. At temperatures at wh1ch there is a substant1al vapor pressure, the bubble acts llke
6
a heat pipe (ref. 19). Vapor is evaporated from the hot side of the bubble and 1S condensed on the cold side. This increases the thermal conductivity of the bubble. The temperature gradient across the bubble is reduced, thus reducing Ft.
We have seen that once formed a bubble is not likely to move; it can only grow. Any growth must be accompanied by the displacement and eventual expulsion of llquid. This effect can be prevented if the growth of the bubbles can be controlled. To illustrate this, we will consider a bubble blocking an otherwise full capillary (f1g. 6). There are two forces act1ng on the llqu1d. One, Fp ' is the force due to the pressure difference, 6P, across the liquid:
F = rrr2 6P p
(9)
The bubble's formation and growth is the result of a heat influx 1nto the bubble. Th1S results 1n a temperature gradient through the liqu1d. If quas1-static process 1S assumed where the vapor in the bubble and the vapor outside the cap1llary are 1n thermal equil1brium w1th the1r respect1ve liquid interfaces, then equation (9) can be wr1tten as
F = nr2 dP I 6T (10) p dt svp
The der1vat1ve, dP/dT, 1S evaluated along the saturated vapor pressure curve. Th1S can be done 1n terms of the Clausius-Clapeyron equat10n (ref. 20):
The 6T in equat10n (10) can be expressed 1n terms of the equ1valent thermal conductance, K
e
6T = Q/K e
where Ke = ~ + K£ and Q 1S the heat flux per cap1l1ary. As before
Ke = Ke Tr2 / h • US1ng equat10ns (11) and (12), equat10n (10) can be wr1tten as
F P
Qhp L v
K T e
The other force act1ng on the 11quld 1S the surface tens10n force:
F s
2nro
(12)
(13)
(14)
To a1d in determ1n1ng Wh1Ch of the two forces dom1ndtes, we will def1ne an e]ect1on number, E , as the rat10 of F to F :
] p s
E J
Qhp L v
2T'rK oT e
(15 )
If E] < 1, the surface tensl0n force w1ll dom1nate and the system w11l be stable aga1nst this form of expuls10n. The expulsion number may be written as E] = SEl where
7
p L v
E 1 = -=--~----2TfOK T
e
contains the system properties and
(16)
S = Qh/r (17)
conta1ns the design parameters. The stability condition now becomes S < l/El. A plot of l/El as a function of temperature is shown 1n figure 7 for 3 He in a 50% copper matr1X. Th1s figure shows the regions of stability and instability.
A real matrix is not likely to be composed of smooth cylindrical capillaries but rather of 1rregular interconnected passages. The effect of the interconnections will be to 1ncrease the system's stab1lity. The pressure in the bubble will be opposed by the surface tens10n of many pores. The pressure difference across the llqu1d is found by combining equat10ns (9) and (13):
llP Qhp L
v
The pressure due to the surface tens10n at n pores is
P s
nF s
Tfr2 2no
r
(18)
(19)
Thus the EJ the system.
llP/Ps 1S decreased by a factor of lin, 1ncreasing the stability of
APPLICATION
In th1S sect10n we w1ll discuss how the above considerations can be applied to the des1gn of a refr1gerator. To illustrate, we w1ll cons1der a hypothetical refr1gerator for coojing an IR detector. The refrigerator requ1rements are shown in table 1. We will assume that the llquid chamber is a sphere (f1g. 3) with a volume V = 4Tfh 3 /3. Since the cap1llary device is used both to contain the liquid and to conduct heat to the llquid-vapor interface, the ent1re volume will be filled with a h1gh-conduct1v1ty matrix such as copper. For our example, we w1ll assume that the v01d fract10n 1S n = 0.5. Since the condensatlon occurs at Tc ' a fract10n, E, of the l1qu1d will be lost in pumping down to To. Furthermore, the volume of llqu1d will change because of. the dens1ty change between Tc and To. After pumpdown, a volume of liquid of Qpt/LP£ must remain to meet the hold time requirement. Thus the volume of the liquld chamber must be
1 V ~ n(l - E:)
or
8
Q t P (20)
h > r 3 - ~'ITn(l
P Q t ] 1/3
(21)
For our example, these are V ~ 7.8 cm 3 and h ~ 1.2 ern.
Bond number considerations lead to another restriction on the chamber specifications. This is the requirement that a« l/Bl, over the entire operating range (Tc to To)' We must use the smallest value of 1/B1 over this range. Since 1/B1 decreases with increasing temperature (fig. 2), the minimum occurs at Tc' Thus we have the restriction that
rh « 1/B1(T ) (22) c
where B1 ~s evaluated at the peak expected acceleration. In our example the requirement hecomes rh« 2x10- 6 m2
•
A th~rd restrlction arises from the desire to avoid bubble nucleation. Equat~on (6) g~ves an expression for h/r ~n t~rrns of the heat flux per capillary. To convert th~s to an expression ln terms of Qp = nQ, we must first estimate the total number of cap~llar~es in the system. One such est~mate ~s the ratio of the void volume to the volume of a s~ngle cap~llary:
Subst~tut~ng th~s into equat~on (6) g~ves
h > 3Q
p 4n'ITK T
e s
(23)
(24)
For our example th~s g~ves h > 0.7 mm. S~nce th~s ~s so much less than the prev~ous constralnt on h (eq. 20), ~t may be ~gnored.
The last restr~ct~on comes from cons~derat~on of the eJect~on number. Agaln we must cons~der the ent~re temperature range. In th~s case 1/E1 decreases with ~ncreas~ng temperature (f~g. 7), so ~ts m~n~mum value occurs at Tc' W~th this consideration, and aga~n taking into account the number of capil1ar~es in the system, the restrict~on that Qh/r < 1/E1 becomes
r/h <
For our example this ~s r/h < 0.02.
4n
3Q E1(T ) P c
(25)
The restrict~ons due to hold t~me, Bond number, and eJect~on number are shown ~n f~gure 8. As can be seen in the f~gure, there is a region of r-h space in whlch a stable refrigerator could be deslgned. The curve for ejection criter~on was calculated assuming non~nterconnectlng capi11ar~es. If lnterconnecting capillaries are used, then thlS curve would move to the rlght and down, increasing the stable reglon. The f~gure shows the case for a slngle capl11ary. In practice, n wl1l be the number of cap~llaries exposed to the pump-out tube.
9
CONCLUSION
It is feasible to design a 3He refrigerator that will operate in the low-gravity environment of space where accelerations can occur in random directions. The key to the design is to fill the liquid chamber with a high-conductivity porous material such as sintered copper. The pores allow surface tension forces to contain the liquid wh~le the conduct~vity ensures that the vaporization occurs at the surface of the matrix rather than internally. The matrix also provides a favorable place for condensation to occur and suppresses bubble nucleation and movement.
10
REFERENCES
1. K~tte1, P.: Refrigeration Below 1 K in Space. Physica, vol. 108B, Aug. 1981, pp. 1115-1118.
2. K~ttel, P.: Sub-Kelvin Temperatures in Space. Adv. Cryo. Eng., vol. 27,1982, pp. 745-749.
3. Chan~n, G.; and Torre, J. P.: A Portable 3He Cryostat for Space Applications. Proc. Sixth Int. Cryo. Eng. Conf., IPC Science and Technology Press, 1977, pp. 96-98.
4. K~tte1, P.; and Brooks, W. F.: Demountable Self-contained 3He Refrigerator. Adv. Cryo. Eng., vol. 27,1982, pp. 727-734.
5. Woody, D. P.; and Richards, P. L.: Spectrum of the Cosmic Background Rad1at10n. Phys. Rev. Lett., vol. 42, Apr. 1979, pp. 925-929.
6. Radost~tz, J. V.; No1t, I. G.; K1ttel, P.; and Donnelly, R. J.: Portable 3He Detector Cryostat for the Far Infrared. Rev. Sci. Instrum., vol. 49, Jan. 1978, pp. 86-88.
7. Gush, H.: Rocket Measurement of the Cosm1C Background Submil11meter Spectrum. Proc. Space Helium Dewar Conference (to be published, U. of Alabama Press, 1984).
8. Urbach, A. R.; and Mason, P. V.: lRAS Cryogenic System Flight Performance Report. Adv. Cryo. Fng., vol. 29,1984, pp. 651-660.
9. Osterme~er, R. M.; Nolt, I. G.; and Radostltz, J. V.: Capillary Confinement of Cryogens for Refrlgeration and LlqUld Control ln Space - I. Theory. Cryogenlcs, vol. 18, Feb. 1978, pp. 83-86.
10. Ennls, D. J.; Kltte1, P.; Brooks, W. A.; Ml1ler, A.; and Spivak, A. L.: A 3He Refrigerator Employing Cap11lary Confinement of Liquid Cryogen. Refrlgeration for Cryogenlc Sensors, NASA CP-2287, 1983, pp. 405-417.
11. Donnelly, R. J.; Klttel, P.; Ostermeler, R. M.; Radostitz, J. V.; Lee, B. R.; and Cooper, J. C.: A Study of Confinement and Heat Transfer Properties of Cryogens, Flnal Report, NASA Grant NSG-2208, 1979.
12. Satterlee, H. M.; and Reynolds, W. C.: The Dynamics of the Free Surface ln Cy11ndrlcal Contalnprs Under Strong Capillary and Weak Gravity Conditlons." Report LG-2, Mechanlca1 Engineerlng Dept., Stanford University, 1964.
13. Alexander, G. E.; Barksdale, T. R.; Rise, R. E.; Lunden, K. C.; and Paynter, H. L.: Experimental Invest1gat10ns of Capillary Propellant Control DeV1ces for Low Gravity Envlronments, vol. II, Flnal Report, NASA Contract NAS8-2l259, Martin Marietta Corp., 1970.
14. Labuntzov, D.; Evdokimov, O. P.; Tishin, I. V., and Ul'ianov, A. F.: Analytical Investigation of the BOlling Process ln Small D1ameter Tubes. Mashlnostroen1e, vol. 7, 1970, pp. 68-73.
11
15. Smith, R. V.: The Influence of Surface Characteristics on the Boiling of Cryogenic Fluids. Trans. ASME. J. of Eng. for Indust., vol. 91, Nov. 1969, pp. 1217-1221.
16. Bald, W. B.: Bubble Nuc1eat1on at Real Surfaces with no Pre-existing Gaseous Phase, Dept. of Engineering Report N-75-29279, University of Oxford, 1975.
17. Kottowski, H. M.: The Mechanism of Nucleation, Superheating and Reducing Effects on the Activat10n Energy of Nucleation. Prog. Heat and Mass Transfer, vol. 7, 1973, pp. 299-324.
18. McGrew, J. L.; Rehm, T. L.; and Griskey, R. G.: The Effect of Temperature Induced Surface Tens10n Grad1ents on Bubble Mechan1cs. App. Sci. Res., vol. 29, June 1974, pp. 195-210.
19. Young, N. 0.; Goldste1n. J. S.; and Block, M. J.: The Motion of Bubbles in Vertical Temperature Gradient. J. Fluid Mech .• vol. 6, Oct. 1959. pp. 350-356.
20. Reif, F.: Fundamentals of Statist1cal and Thermal Phys1cs. McGraw-H1ll, 1965, pp. 304-306.
12
TABLE 1.- REFRIGERATOR REQUIREMENTS
Parameter Symbol Value
Condensation temperature T 2 K c Operating temperature T 0.3 K
0
Hold time t 54 ks
Refrigeratl.on a Qp 40 ~W power
Peak acceleration a 0.1 g
a Includes parasl.tl.c loads.
13
COLD STATION 03K
_HEAT ,SWITCH ,,~~~~
HEAT SINK< 2 K
ADSORPTION PUMP
HEAT SINK<10 K
F1gure 1.- A zero-grav1ty 3 He refrigerator showing the principal components.
4
N 3 E ,....
0 2 ....
.... CO
--- 1
o TEMPERATURE,K
F1gure 2.- Plot of l/Bl as a function of temperature for 3 He for the case 1n Wh1Ch the acceleration 1S 9.8 m/sec. The region below the curve is the region in Wh1Ch surface tension forces dominate.
14
Flgure 3.- The arrangement that allows the greatest volume of enclosed fluid. The eXlt tube penetrates to the center of a spherlcal cavity. The cavity 1S f111ed wlth the porous matrlX.
5
1
o 1 TEMPERATURE,K
Flgure 4.- Plot of the superheat in 3He as a function of temperature.
15
2
z 0 CI) -ZN W E I- ::::; 1 wv U I
« 0 .... u. c: ::J CI)
0 1 3 4
TEMPERATURE, K
Flgure 5.- Plot of the surface tensl0n of 3 He as a functlon of temperature.
16
P + AP T+ AT
p. T
h
F~gure 6.- A vapor bubble block~ng an otherw~se full cap~llary. The cap~llary has a rad~us of r. A column of l~qu~d of he~ght h separates the bubble from free space. A temperature d~fference of 6T and a pressure d~fference of 6P are across the l~quid column. There ~s a heat flux of Q ~nto the bubble.
17
10-2.-------~._------_.--------_.--~
10-7~--------L---------~------__ ~ __ ~ o 1 2 3
T,K
F1gure 7.- Plot of 1/E1 as a funct10n of temperature where 1t has been assumed that half of the thermal conduct1vity path 1S through copper and half 1S through 3He •
100
E E 10
1
1
HOLD TIME
10
O~ t
~" "I V I'
<v~ " ,'1 I
r,l1 m
NUCLEATION
100 1000
F1gure 8.- Plot show1ng the stable and unstable reg10ns of r-h space for var10US types of 1nstab1l1ties. The shading 1nd1cates the reg10n of 1nstab1l1ty.
18
1 Report No 2 Government AccessIon No 3 RecIpIent s Catalog No NASA TM-85973 4 TItle and SubtItle 5 Ropon O.te
DESIGN CONSIDERATIONS FOR A 3HE REFRIGERATOR FOR July 1984
SPACE APPLICATIONS 6 PerformIng Organlz.tlon Code
7 Author(s) 8 PerformIng Orglnlzatlon Report No Peter Kl.ttel and Andres F. Rodn.guez (Unl.versity of
A-9786 the Pacl.fic, Stockton, CA 95211)
10 Work UnIt No 9 Performong OrganIzatIon Name and Address
T-nn?L. Ames Research Center 11 Contract or Grant No
Moffett Fl.eld, CA 94035 13 Type of Repon and PerIod Covered
12 Sponsorong Agency Name and Address Technl.cal Memorandum Natl.onal Aeronautl.CS and Space Adml.nl.stratl.on 14 Sponsorong Agency Code Washlngton, DC 20546
506-54-21 15 Su pplementary Notes POlnt of Contact: Peter Klttel, Ames Research Center, MiS 244-7, Moffett Fleld, CA 94035 (415) 965-6525 or FTS 448-6525
16 Abstract
The low temperature provl.ded by 3He refrlgerators (0.3-3 K) have useful space appllcatl0ns. However, the low temperatures and the low surface ten-Slon of 3He requl.re specl.al deslgn conslderatl0ns. These conslderatl.ons l.nclude the need for small pores to contal.n the l1.qul.d l.n a matrl.x, the effects of bubble nucleatl.on and growth, and the effects of the thermal con-duCtlVl.ty wl.thl.n the matrl.x. These desl.gn conslderatl0ns are dlscussed here along wlth an analysls of a posslble conflnement system.
17 Key Words (Suggested by Author(s)) 18 Distribution Statement
Hellum-3 Unllml.ted Cryogenl.CS Refrlgeratl0n
Subject Category - 34
19 Securoty Oasslf (of thIS report) 20 Securoty Classlf (of thIS page) 21 No of Pages 22 Price-
Unclassl.fl.ed Unclassl.fled 21 A02
·For sale by the NatIonal Technlc.1 InformatIon ServIce Sprongfleld V"glnla 22161
End of Document
