Aircraft-Fuel-Tank Design for Liquid Hydrogen

RESEARCH MEMORANDUIVr

FOR LIQUID HYDROGEN AIRC WFT -FUEL -TANK DESIGN By T. W. Reynolds Lewis Flight Propulsion Laboratory C Ieveland, Ohio Declassified October 2.5,1* e#m.

FOR AERONAUTICS

WAS HI NGTOFi August 9, 1955 Reclassified May 29, 1959 TECH LIBRARYKAFB,NM NACA RM E55F22 Iflll[llillllllllln[illlllll[ilill ‘--- cllJ44n4i “ NATIONAL AIJVIBORY CmMITTm FOR AERONAUTICS RESEARCH MEMmWmM AIRC’ME’T-FUEL-TANK DESIGN FOR LIQUID HYIROGEN * By T. W. Reynolds SUMMARY Some of the considerations involved in the design of sircrst% fuel tanks for liquid hydrogen are discussed herein. Several of the physicsl properties of metals and thermal insulators in the temperature range from ii ambient to liquid-hydrogen temperatures are assembled. Calculations %ased on these properties indicate that it is possible to build a large-size liquid-hydrogen fuel tank wblch (1) will weigh less then 15 percent of the fuel weight, (2) will have a hydrogen vaporization rate less than 30 percent of the cruise fuel-flow rate, and (3) can be held in a staud-by condition and readied for flAght in a short time.

rmRoIxIcTIoN Recent performance analyses have indicated that liquid hydrogen offers promising advantages as a fuel for both long- and short-range, high-altitude aircraft operation (ref. 1) . b order to obtain these ad- vsmtages, however, the fuel-tank weight must be kept to a small fraction of the fuel weight.

To keep the tank weight light with a fuel of the very low density (4. 42 lb/cu ft) and low boiling-point temperature (37° R) of liquid hydro- gen causes problems not encountered with ordinary liquid fuels. Recent experience in the liquefaction of hydrogen and in the storsge and transfer of liquid hydrogen has indicated that the problems involved in its use as an aircraft fuel are not insurlmuntable. Concepts of fuel-tank design different from those used for ground storage sre required for aircrstt fuel tanks. Ground storage DeWar vessels, designed for low heat-leak rates, are multiple-shell units and have much too high a ratio of tank weight to fuel weight to be considered for flight use.

In this report are assembled some of the physical properties of materials and other considerations involved in the design of a liquid- In addition, some details of the design and hydrogen aircraft fuel tank.

.

anticipated performance of a particular fuel tank for a long-range air- craft are reported.

.

liACARM E55F22 “ GENERAL DESIGN CONS=~ONS Values of some of the physical properties of hydrogen which are “- 1, presented in references 2 ead 3 have been reprodu+d in table I and fig- -< _ uresland2.

The major considerations involved in the desi~ of a fuel tank for an airplane using hydrogen as fuel are (1) it must be light weight, (2) it should not have a heat-leak rate such that fuel will vaporize faster then the engines will burn it at cruise conditions, snd (3) it should le capable of storing fuel on the ground fol”a reasonable length of time without loss of fuel. .,— Present designs of ground storage Dewsr vessels f~r liquid hydrogen employ a triple-jacketed construction. The inner shell, which contains the liquid hydrogen, is surrounded by a vacuum space, a liquid-nitrogen- This cooled radiation shield, another vacuum space, andan Guter shell.

construction, besides having multiple shells, requires_the outer wall, at Such construction,.whilenecesssry to least, to withstand a full vacuum.

maintain the low losses required for long-time storage of liquid hydrogen, is obviously too heavy for a flight fuel tank.

l In order to maintain the tank weight ta~ small f=action of the fuel weight, it will be impossible to use more thsn a single-shell tank with an l extremely light-weight insulation around it. Such a tank cannot be built to have a very low heat-leak rate or fuel-loss rate such as the storage ‘ - = Dewar vessels have.

!i?he,following principal specifications were assumed for the tank -.

design reported herein: (1)”The fuel-tank weight shall be no more than about 3.5percent of_ - the fuel weight.

.

(2) The heat-leak rate willbe such that fuel tiifivaporize no faster than about one-third of the fuel-flow rate for normiiLz?atedengine speed.

Rather thsn specify a no-loss the on the ground after filling a tank, the time was determined after meeting the preceding two requirements.

MATERIALS OF COlWURUX17~N Metals Only limited data are available on the ktiengtfi of metals and alloys l at the temperature of liquid hydrogen (37° R). The tefiile and yield However, strengths of nmst metals increase as the temperature is lowered.

NACARM E55F22 some materials, for exemple the ferritic steels, become very brittle at subzero temperatures and sre unsatisfactory as structural materials.

Recent work (refs. 4 to 6) has indicated that the austenitic stainless steels, monels, and aluminum alloys have satisfactory properties for use at liquid-hydrogen temperatures. Yield strength of some of the metals increases fran50 to 100 percent of their room-temperature value h de- creasing to these low temperatures. At the same time the ductility re- mains at usable values and the impact stren@h does not change much from room temperature values.

Recent data on the yield strength end ductility of stainless steels 303, 310, and 316, 24ST aluminum, smd monel (ref. 5) are shown in fig- ures 3 and 4. Data for stainless steel type 301 (ref. 6) are also shown on figure 3. These extra~lated data indicate that yield-strength values in excess of 150,000 pounds per square inch may be attained by certain hardened stainless steel alloys at 37° R.

At a yield strength of 150,000 pounds per squsre inch for steel, the strength-to-weightratio for the staimless steel is still only about three-fourths of that for the 248T aluminum.

A lighter weight tank could be made from the aluminum alloy. However, stainless steel may still be the desirable material for construction because of its superior impact and welding properties.

When a fuel tank using liquid hydrogen is cooled from ambient tem- peratures to 37° R, the heat capacity of the metals will be of interest in calculating the cooling load. Figure 5 shows the heat capacity of several metals in this temp~ature rsmge (ref. 7) . These data have been integrated to give enthal.pyvalues above 0° R (fig. 6) for more conven- ient use in heat-load calculations.

The large temperature range through which the tank materials operate causes thermal expansion and contraction problems. Linear coefficients of thermal expansion of several metals (refs. 8 to 10) in the r-e from 0° to 300° K (0° to 540° R) sre shown in figure 7. The integrated values of expansion coefficients from 300° to 0° K are presented in figure 8.

Data on thermal conductivities =d emissivities of various metals in this temperature rsnge maybe found in references 11 to 14.

Insulating Materials A discussion of the problems of low-temp~ature insulation is pre- sented in references M to 20. Ih general, the characteristics of an insulator which are desirable for this particular application are l (1) Low thermal conductivity .

.

NACA RM E55F22 (2) Low vapor penetration --* - .

(3) Good compressive strength .

(4) Low density The thermal conductivity of a nwdber of insulators at liquid- hydrogen terr.rperature and at liquid-nitrogen temperature both in vacuum . . _ — and in the presence of hydrogen and nitrogen gas j.sgiven in reference “= 21.

Santocel has one of the lowest thermal conductiv~ties of Jmown ti- ..- .— sulators. It is used extensively in evacuated sh~lls=for insulation o“f liquified-gas storage containers.

Santocel is a &ranular insulator, how- — ever, requiring a supporting container.

It is sulject to settling and has a higher density-conductivityproduct than several other insulators of interest. Kapok and several fibrous glass insulations have been pr6- duced that have both low thermal conductivity and’low-density. However, the fibrous nature of these materials, which gives them poor vapor barrier qualities and lack of compressive strength, makes them poor choices for this application.

.— The @e of material that seems to offer the be$J insulating possi- bilities is a fosmed plastic. Typical of this t e i.sStyrofoam (apoly-” e styrene fosm). This plastic is cellulsr in strut ure with the cells T discontinuous. St~ofoam has negligible valor pehetr-tition end good corn-““ pressive strength. It csm be made in vsrious shapes %d densities, Its physical properties are dependent on the density of the material. Scnie ‘“-- of the physical properties of Styrofosm at 77° F (ref. 22) sre given in -. — the fOliOwi~ table: Foam density, lb/cu ft 1.6 2.0 Compressive yield strength, psi “1O-2O .15-25 25-35 Tensile strength, psi 30-45 50-’70 80-100 Shear strength, psi 15-25 25-35 35-45 Impact strength, 3/8” by 1/8” section, 0.5-1.2 1.1-1.8 2.1-2.7

T

in l -lb —.

0.23-O.ti Btu/(sqft)(hr) —.

-. _ ‘~~fin:;~uctimty’ ., 2-5X1.o -5 Coefficient of linesr &qxansion, (in.) /(in.)(°F) - - 0.27 Specific heat, Btu/(lb)(°F) at 40° F 1.5-3.0 Water vapor transmission, grains/(sq ft) .— (hr)(in. of thickness)(b. Hg vapor —.

pressure difference) The compressive strength is about 50 percent ~eater at liquld- The vsria- nitrogen temperature (78° R) thm it is at room t’emp=ature.

tion of the thermal conductivity with temp=aturd foF Styrofoam (ref. 23) , is shown in figure 9.

— .- . .

.

NACA RM E55F22 Other types of expanded plastic team might also prove of int~est.

Isocyanate foams, for exsmple, can be fosmed in place ‘(ref.24), and may have the advantage of ease of installation.

DESIGN CALCULATIONS The characteristics and performamce of a tank to hold nominally 25,000 pounds of liquid hydrogen are presented in the followhg paragraphs.

This tsak is the approximate size selected for the 5500-nautical-mile- ramge subsonic bomber discussed in reference 1. The cruise fuel-flcrw rate Is assumed to be 1350 pounds per hour.

Tank Shell The tank will be pressurized, the pressure aiding in maintaining the fuel-tank shape. As noted in figure 2, the density of the saturated liquid varies considerably with the equilibrium pressure. Liquid expsa- sion may be considerable as the liquid heats until the vapor pressure reaches the temk design pressure. For ex~le, at 30 pounds per square inch absolute> expansion is about 5 percent, at 60 pounds per square inch it is 12 percent, and at 150 pounds per squsre inch it is ne=ly 40 per- cent. Allowance must be made for a expansion volume greater than the anticipated liquid expansion. Since the fuel-tamk weight till also in- crease as the design pressure is increased, it is desirable to keep the tank pressure as low as possible.

It was pointed out in reference 1 that a tsnk pressure of 2 atmos- pheres would be adequate for pumping the fuel at the high-sltitude cruise condition. A cylindrical tank 10 feet in diameter and 81.6 feet long, including hemispherical ends, has a volume of 6151 cubic feet. This is nearly 9 percent OV= the nominal volume of 25,000 pounds of liquid at l-atmosphere pressure. This expansion volume is adequate for a working pressure of 2 atmospheres.

If a @eld strength of 150,000 pounds per square inch for a hsrdened 300-series stainless steel at 3P R and a safety factor of 2 are assumed, the shell thickness will be 0.024 inch. The surface srea of this shell is 2564 squsre feet snd the weight 2564 pounds.

Heat-Leak and Surface Temperatures The tsmk is insulated with a layer of Styrofoam, which is covered with laminated Mylar-aluminum foil. A schematic cross section of the tank is shown in figure 10. An estimate of the heat leak and surface temperatures was made in the fol.loting manner: The heat transferred through the Styrofoam insulation was set.equal to that transferred from the surroundings to the surface by radiation and natural convection: =+ ?WICARllE55F22 ; = hcA&ta - t~) + hreA~(ta - t~) = ~ ~[ts - (-423)] (1) .

-— where .

logarithm of mean srea of inner and outer insulation surfaces, sq % ft (~ differs from As by only shout 2 percent &d is therefore - “- “~ assumed equal to As in the calculations.)

As surface area of tank, sq ft .- — diameter, in.

‘o — ta - t~ At e emissivity of surface, assumed equal to 0.06 h= free convection coefficient, Btu/(hr)(sq ft)(°F) (Estimated from” .- -.

0.25 hc=0.27~ (ref. 25)) u .

— radiation coefficient for emissivity of 1.0 -.

% k thermal conductivity of insulation, Btu/(hr)(sq ft)(°F/ft) (A mean value of 0.015 was assumed for the calculations; see fig. 9) .— L thickness of insulation, f% heat leak, Btu/hr q -.

ta ambient-air temperature, ‘F — - e time, hr t~ equilibrium swface temperature, OF .- Surface temperatures were determined for insulation .thicknessesof 3/4, ~, 2*, ~d 3 inches, end for ambient temperatures ox 80° ~d -67° F.

.— At ther”-67b F anibienttemperature, a pressure of 0.1 atmc-ephere was assumed. The natural convection coefficient waE assumed “tovary with the .

.— — At 0.25 ~o.5 Wher: ~ .J5pre5~we ““-:- square root of the pressure, hc = 0.27 — Do () .

in atmospheres.

l .

., _= -.=... ,: ~ NACARM E55F22 The following table sumnsrizes the results of these surface.

temperature calculations: Insulation thickness, Total insulation Surface temperature, ‘F, b in. weight, lb at ambient temperature of - 80° F -67° F 3/4 209 -70 -271 417 -15 -209 G 625 10 -182 z 3 835 23 -162 These data are plotted in figure 1.1.

Once the surface temperatures have been determined, the heat-leak rate is easily calculated from the equation ~ (2) -=:~[t~ - (-4=)] e Heat-lesk rates expressed in Btu per hour and also as pounds of hydrogen evaporated per hour as a function of insulation thickness are shown in figure 12. The vaporization rate at the altitude condition with ~ inches of insulation is about 255 pounds per hour. The smount of hytlrogengas necesssry to maintain the tsnk pressure while removing liquid at the rate of 1350 pounds per hour is obtained by multiplying the liquid-flow rate by the ratio of vapor b liquid densities at the tsnk pressure. At 30 pounds per square inch, the ratio of vapor to liquid densities (fig. 1) Therefore, about 49 pounds per hour of hydrogen gas will is about 0.036.

maintain the tsmk pressure. This smount is only about 20 percent of the vaporization rate due to heat leak for the 2~-inch insulation thickness.

The heat leak at higher anibient temperatures canbe approximatedby multi- plying the heat leak cskulated for the 80° F saibient by the ratio of the over-all temperature differences involved.

These heat-leak calculations have not assumed any additional insulat- ing value that might be obtained through installation of this tank in an The heat-leak calculations should therefore be con- aircraft fuselage.

servative for this type of installation.

Externally munted tanks would have higher convective heat-transfer coefficients than the free-convection coefficients assumed in evaluating equation (l). Consequently, with sn external installation, these heat- leak calculations will be somewhat low for conditions dur~ flight.

Filling Tank with liquid hydrogen, when it is at The problem of filling the tank initially, is essentially one of both room temperature and filled with air .

8 NACA RM E55F22 .

COOl@ and purgi~ . The amount of heat that must be removed in $ooling the tank and tisulatim from en smbient temperature of 80° F to llquid- hydrogen temperatures canbe estimated by using the integrated heat- capacity curve for iron (fig. 6) and the heat-capacity value for Styro- For an insulation thickness of 2* inches, this heat smounts to foem.

less than 1.50,000 Btu.

.

In order to prevent the formation of any solid materials, which might plug fuel lines or orifices, it willbe neceqs~to purge the t&lc and lines with either helium or hydrogen gas, since sny other material — will be solid at these temperatures.

— .

,- There are several alternative schemes for filling the tank: (1) Flushing with gaseous helium, then filling with liquid hydrogen directly (2) Cooling with liquid ~trogen, flushlng with helium, then filling — with liquid hydrogen —.

(3) Cooling and flushing with helium that has been_cooled by liquid # nitrogen — (4) Flushing with helium cooled by a mechanical refrigeration cycle Since the heat capacity of liquid hydrogen is rather high (fig. I-3 snd ref. 26), the cooling load could be a%sorbed as!sen%ible heat in the .

The tank is designed for a working pressure of 2 atmospheres.

liquid.

The temperature of the saturated liquid at this pressu??-e is 41° R (fig.

2) or about 4° R above the normal boiling point. we average heat capaZ- .- .-!

ity of the liquid over this temperature interval Is about 2.4 Btu per The heat sink available in the liqtid.i_& then, pound per %.

wc#t = (25,000)(2.4)(4) = 240,000 Btu where w is mass in pounds and ~ is heat capacity at constsmt pressme in.Btu p&r p~d per %. T’hi~_ —___ is more than enough to absorb the full cooling load. = — It will be noted from figure 6 that about 95 p;erc”Ej3t of the cooling .- load couldbe extractedby using liquid nitrogen to:CW1 the tank. When the tszikis cooled by some other means thsn using the h~at capacity of .—- the liquid hydrogen to absorb the heat load, the fille~ tank csn stand for a longer time without losing any hydrogen thro~h evaporation. The- heat sink available in the liquid is used ta absorb, the heat leak into the tank until the liquid temperature rises to a point where the vapor pressure is equal to the tank design Pre8s~e. ~i~,t~e has been The variation of no-loss time ““- called the no-loss time in this report.

for the 30-pound-per-squsre-inchworking-pressure t;snls reported herein is plotted against insulation thickness in figure 14. ~or the ~ inch ,<..

insulation, the no-loss time is about 165 minutes. ~ .

,- , NACARM E55F22 .

It may be desirable to maintain the tsnks in a refrigerated, stand- by condition, ready to be filled with liquid hydrogen. Therefore, it is suggested that a cycling system using helium gas, cooled by liquid nitro- W gen, maybe cycled through the tmk until such time as it is desired to The introduction of liquid hydrogen, then, fill with liquid hydrogen.

does not impose a great thermal shock on the tsnk, the loss of hydrogen through vaporization in cooling from liquid-nitro&n to liquid-l&dro~en temperature is relatively small, and the system is full of helium gas prior to the introduction of the liquid hyirogen.

a SUMMARY OF CALCUIA3!IONS The results of the foregoing calculations are summarized in the following table: N & Size: . . . . . . . . . . . . . . . . . . . . . . . . . .

Diameter, ft . 10 . . . . . . ...* . . . . . . . . . . . . . . .

Length, ft . . 81.6 Volumej cu ft . . . . . . . . . . . . . . . . . . . . . . . . . 6X51a gal . . . . . . . . . . . . . . . . . . . . . . . . . 45,800— h Surfacearea, sqft . . . . . . . . . . . . . . . . . . . . . . “2564 . . . 2 Working pressure, atm . . . . . . . . . . . . . . . . . . . . .

. .

lbsqin: . . . . . . . . . . . . . . . . . .

( Styrofoam insulation : . . 2$ Thickness, in. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

Density,lb/tuft . . . . . . . . . . . . . . . . . . . 1.3 Weight of tank: Shell)lb . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564 . . . . . . . 625 Insulation, lb. . . . . . . . . . . . . . . . . . .

. . . . . . . 64 Covering, lb. . . . . . . . . . . . . . . . . . . .

. . . . . . . 247 Allowsmce for baffles and stiffeners, lb . . . . . .

. ..*. . 3500~f- Approximate total weight, lb . . . . . . . . . . . . .

Estimated performance at amibienttemperature, F: . . . . 80 -67 ,. 10 -182 Outer surface temperature, OF . . . . . . . . . . .

88,000 49,500 Heat-leakratejBtu/br.. . . . . . . . . . . . . .

. . 454 255 .Hydrogen-vaporization rate, lb/hr . . . . . . . . .

. .

No-loss time onground, min . . . . . . . . . . . .

%olds 25,000 lb liquid hydrogen with 9 percent exp~ion volume.

bCovered with layer of Mylar-aluminum foil.

weight. ~ cAbout 14 percent of fuel CONCLUDING REMARKS of the physical properties of matecials and from a consideration thermal insulators now available, it appears feasible to design a .- NACA RM E55F22 liquid-hydrogen fuel tanl.for aircraft use which (1) will weigh less .

— than 15 percent of the fuel weight, (2) will have a h@&ogen va@tizatio~ rate less than 30 percent of the cruise fuel-flow rate, and (3) can be held in a stand-by condition and readied for flight in a short time.

Lewis Flight Propulsion Laborat&y -— National.Advisory Committee for Aeronautics — Cleveland, Ohio, June 22, 1955 REFERENCES .-.

.

1. Silverstein, Abe, and Hall, Eldon W.: Liquid Hydxogen as a Jet Fuel for High-Altitude Aircraft. NACARM E55C28a, 1955.

— -.

2. Woolley, Harold W., Scott, Russell B., = Brickwedde, F. G.: Com- pilatk o~ The&l pro~erties of H@Qgen in Its V&rious Isoto2ic .,.” and Ortho-Para Modifications. Jour. Res. Nat. Bur. Standard~, vol..

.

41, no. 5, Nov. 1948, pp. 379-475.

3. White, David, and Johnston, Herrick L.: The Thermodynamic Properties of Liquid Normal Hydrogen Between the Boiling Point and the Critical Temperature snd up to 150 Atmospheres Pressue. TR 264-23, Cryogenic Lab., Dept. Chem., The Ohio State Univ., Feb. 1,-1953. (AF Contract W33-038 ac 14794 (16243), Proj. RF-264.)

4. Krivobok, V. N.: Properties of Austenitic Stainless Steels’at Low Temperatures. Mechanical Properties at Low Temperatures, Circular 520, U.S. Dept. Comm~ce, Nat. Bur. Standards, May 7, 1952, pp.

13.2-134; discussion, pp. 134-136. ‘“’ Low 5. ICropschot, R. H., Psrkerson, C. R., O’Donel, J,, and Crum, M. G.: Temperature Tensile Testing Equipment and ResultA”(300°-200 K). Rep.

Bur. Standards, JUIY 1, 1953. (~ 2708, U.S. Dept. Commerce,-Nat.

Proj. 0306-20-2667.)

—.— __ 6. Anon.: Stainless Steel Handbook. Allegheny Ludlum Steel Corp., .— Pittsburgh (Pa.), 1951. — — 70 Anon.: International Critical Tables. ‘Vol. V. McGraw-Hill Book Co., Inc., 1929, p. 83.

—.

8. Altman, Howard W., Rubin, Thor, and Johnston, Herrick L.: Coeffi- cients of Th=mal Ex@nsion of S.olids.at Lowjllsm~eratures. II - The Thermal Expansion of Monel, Inconel, Contracid, SAE 1020 Low ; Carbon Steel and Type 410 Stainless Steel. ‘T!R 264-19, Cryogenic ‘- “--A - Lab=, Dept. Chem., The Ohio State Univ., May 1, 1952. (AP Contract W33-038 ac 14794 (16243), pro~. RF-264.) — w’ NACA RM E55F22 9. Altman, H. W., Rubin, T., and Johnston, H. L.: Coefficient of Thermal Expsnsion of Solids at Low Temperature. III - The Thermal Expansion of Pure Metals, with the Data for Aluminum, Nickel, Titanium, and Ztionium. TR 264-27, Cryogenic Lab., Dept. Chem., The Ohio State Univ l , Feb. 10, 1954. (AP Contract W33-038 ac 14794 (16243), Proj.

RF-264.)

10. A1.tmm, Howard W., Rtd)injThor, and Johnston, Herrick L.: Coeffi- cients of Thermal Expansion of Solids at Low Temperatures. IV - The Thermal Expansion of Type 304 Stainless Steel and Yellow Brass.

TR 264-28, Cryogenic Lab., D=pt. Chem., The Ohio State Univ., Mar.

8, 1954. (AF Contract W33-038 ac 14794 (16243), Proj. RF-264.)

11. Powers, Ro%ert W., Ziegler, John B., and Johnston, Herrick L.: The Thermal Conductivity of Metals and Alloys at Low Temperatures.

11 - Data on Iron and Several Steels Between 25° and 300° K. In- TR 264-6, Cryogenic Lab., Dept.

fluence of Alloying Constituents.

Chem., The Ohio State Univ., Apr. 25, 1951. (AF Contract W33-038 ac 14794 (16243), Proj. RF 264.)

Powers, Robert W., Ziegler, John B., W Johnston, Herrick L.: The 1.2.

Thermal Conductivity of Metals and Alloys at Low Temperatures.

III - Data for Aluminum Alloys Between 25° and 300 K. TR 264-7, (Al Contract Cryogenic Lab., Dept. Chem., The Ohio State Univ.

W33-038 ac 14794 (16243), Proj. RF-264.)

1.3. Powell, R. L.: Thermal Conductivities of Solids at Low Temperatures.

TM No. 17, Cryogenic Eng. Lab., Nat. Bur. Standards, Boulder (Colorado), May 26, 1953.

14. Reynolds, M. M., Fulk, M. M., Weitzel, D. H., and Park, O. E.: A Preltd.nary Report on the Infrared Absorption of Metals at Low Tem- perature. -TM ~o. 16, Cryogenic ~. Lab., Nat. Bur. Standards, Boulder (Colorado), May 26, 1953.

and Develop- White, John F.: Low-Temperature Insulation - Problems 15.

pp. 647-652.

ments. Chem. Eng. Prog., vol. 44, no. 8, Aug. 1948, Insulation at 16. Palmer, Burton M., and Taylor, Robert B:: Glass -Fiber no. 8, Aug.

Liquid-Air Temperatures. Chem. hg. Prog., vol. 44, 1948, pp. 652-654.

Low-Temperature Insulation - 17. Bradley, C. B., and Sime, J. F.: Thermal Insulation for Low-Temperatur~ Applicatio~. Chem. Eng.

Prog., vol. 44, no. 9, Sept. 1948, pp. 723-726.

Styrofoszuf~. - Low-’Temperature 18. McIntire, O.,R., and Kennedy, R. N.: Insulation. Chem. Eng. Prog., vol. 44, no. 9, Sept. 19482 PP.

.

727-730.

.- -. — — Ii?

NACA RM E55F22 ‘ — 19. Van Buskirk, E. C., snd Surland, C. C.: Low-Temperature Insulation - .

A New Plastic Low-Temperature Insulation. Chem. Eng. Prog.,“vol.

44, ?10.10, Oct. 1948, ~. 803-804.

20. Senders, Victor: Cellular-Glass Insulations for Low-Temperature —w Equipment. Chem. E%. Prog., vol. 44, no. 10, Oct. 1948, pp.

804-806.

21. Powers, R. W., Johnston, H. L., Hansen, R. H., and Ziegler, J. B.: Thermal Conductivity of Heat hwlating Materials. TR 264-16, Cryogenic Lab., The Ohio State Univ., May 14,1953. (AF Contract W33-038 ac 14794 (16243), Proj. RF264.)

22. Anon.: plastics Dept., me DO-WChem.

st~ofosml - Technical Data.

Corp., Midland (Mlch.), I-851. .

— . . .

23. Waite, H. J.: Styrofoam (Expanded Polystyrene) Msulation at Low Eng. Conf., U.S.

Temperature. Rep. 3517, Proc. 1.954Cryoge@c Dept. Commerce, I?at.Bti. Standards, Feb. 1955,–pp. 158-163.

24. Witte, P. J.: Foaming Plastics - Cellular Cellulose Acetate. Modern _ Plastics, Encyclopedia Issue, Sept. 1954, pp. 168-169.

.- 25. McAdsms, William H.: Heat Transmission. Second cd., McGraw-Hill Book CO., ~C., 1942, p. 241.

l 26. Smith, A. Lee, Hallett, Nathan C., and Johnsttin, Herrick L.: Con- - The Heat Capacity of Liquid Para- densed Gas Calorimetry. VI hydrogen from the Boiling Point to the Critical Point. Jour. Am.

Chelll. Sot., VOL. 76, no. 6, Mar. 20, 1954, pp. 1486-1488.

..

.

— NACA RM E55F22 TABLE I. - PHYSICAL PROPERTIES OF HYDROGEN Molecular weight . . . . . . . . . . . . . . . . . . . . . . . . 2.016 Heating value, Btu/lb . . . . . . . . . . . . . . . . . . . . .51,571 Boiling point atlatm, OR. . . . . . . . . . . . . . . . . . . 37 25.2 Melting point, OR..... . . . . . . . . . . . . . . . . . .

59.6 Critical temperature, OR.. . . . . . . . . . . . . . . . . . .

Critical pressure, atm . . . . . . . . . . . . . . . . . . . . . 12.8 1.95 Critical density, lb/tuft.. . . . . . . . . . . . . . . . . .

25.2 Latent heat ofnelting, Btu/lb . . . . . . . . . . . . . . . . .

Iatent heat of vaporization, Btu/lb . . . . . . . . . . . . . .

4.42 Density, liquid at 1 atm and 37° R, lb~cu ft . . . . . . . . . .

Density, vapor at 1 atm and 492° R, lb/cu ft . . . . . . . . . . 0.0056 0.014 Viscosity, liquid, centipoises . . . . . . . . . . . . . . . . .

0.695 . . . . . v = 0.0084 & Viscosity, vapor, centipoises at TO K () Btu/(lb)(OR) . . .“ 2.25 Heat capacity, liquid parahydrogen at 370 R Heat capacity, vapor at 519° R, Btu/(lb)(OR\ . . . . . . . . . . 3.4 ~ , I , , 1 1 I II I I 1 I I % (’3?)

lo~ . . #- ,yt’”””’% 42 (75.

4?

XL I 1 I lllM’1L7/1 /vxlmt’11# I I I 111111 11111 I . .

.(x .1 1 10 Km Km Fn3ssure, ah !2’ - FTamUre-temperdmre-rknsity relntim for Qdrqg3n.

m 1. Data obtained frm references 2 and3.

,i ,, ..

< . d .

d I 1 ,,, 1 ’11 NACA RME55F22 4.G \ 4.4 - 60 / \ 4.2 0 4.0 c 1, g .

> ~ 3. e / If -50 F /7 \ .

z!

!!7 3. 6 i!

B 5. 4 -40 3. 2 -35 3. 0 4 6 8 10 12 0 2 R-sssure, atm Fressure-tsmperatuI= -density relation for saturated liquid hydrogen.

Fi@re 2. - IELta obtainecl fromreference 3.

.

.

... 1 NACA RM E55F22 .

, 1 r 24 ST Ahu@ll : o q Mo.rwl “= -.

180x103 v 303” Steln146s- steel A 310 stdnlqsc s_@el O 316 Stainl&sssteel IS 301 Sttin~ss steel annealed) ,A 301Stainless steel 1/2hsrd) [ v 501stafdss s%eel fullhard) -.

-- .

.

) ‘- _ 0 30L Stainlessskeel 3/4 hqrd) - ‘- -- -- - -- - ,- _ - - - .

-.

“ -o- \ .

\ 0- - \ n.

.

l .- m A Liquid lxquid hydrogen -nitrogen =2 Sal 600 o 100 200 300 400 Temperature,% Figure 3. - Yield strengthof several alloys at low temperature. Data obtained — from references5 and 6.

NACA RM E55!?22 0 24 ST Alumlnum t3 Monel v 303 Stainless steel A 310 Stainlesssteel ~ 316 Staiulesssteel / \ .

d y R — — / . - / — — -_ Q d “ o 400 100 200 300 500 600 Temperature,OR Fiwx! 4. - Elongationof severalalloys at lowtemperatures. ~ta obtaine~ ~ reference5.

.

.

.

.— NACA RM E55F22 .- -.

.— .26.

M3gmesium --- -- Aluminum —-— Iron —--— Nickel ue&leBium Zinc / - .24 - -–——— copper ~ - ——— Tin ———— Lad / ‘ Al uminum.

< / / / / .20 / / / / / .

~ / / ~, / v .16 / -.

/ / / i / / / .I.2 i / $ Iron / { J!

/ 4 ) .08 .

/ / / Tin -, / .04 f Lee6.

Liquid .

nitrogen .- o 40 30 120 160 m ’246 2.93 32a Temperature, OK I I I I I . . .. .

o 1(N 200 300 400 503 ““ 603’ Tempraturej % -.-— mgul’e 5. - Eeat capacity of several mtaleat low timpetitures. Data obtained frcm reference 7.

“ J NACA RM E55F22 1 I Magnesium / .- / Magnes Iutn I I l— / 80 — — — — Aluminum 44~ ‘-1 I I ‘-– %% r 75 — Aluminum / I I i —-— .~~n 14 70 — 65 — 60 — 55 — 50 — .

q 45 — ~ la * J 40 — ~ ~ 35 — & 30 — / / / / 25 — / Tin / , , , I / 1/ / / ‘ . I 20 — 4 / Lead— 15 — / / / ‘ / _ .- / 10 ‘— I Llq;ld h~oge~ t--l 5 “ I 1/ r--l .

o— 120 160 200 240 280 320 0 40 80 Temperature, oK I I I o 100 200 300 400 500 Temperature, ‘R Figure 6. - Integrated heat-capacity curves for several metals above 0° K.

,-.

NACA RME55F22 .

$.

2 .&lo-s I I Aiminum 2.0 — 304 8tainless Ste elm — 1.6 / 1.2 410 Stainlese } .8- { l / I .

A ~ .4 d - ~ -. 4 0 50 100 1.50- ‘2cx — 250 306— Tauprature, ‘% :, — 1 I o 103 2CX) 303 ‘ %3 500 Temperature , % : z F— ”:,,... ._” - Lineaxcoefficients of eqmmslonj of several metalsat low R@.n’a 7.

_- temperatures. Data obtained frcn-keferenc@ 8 .~o10.

— .-.

NACA RM E55F22 21 I G 1 A steel NI II M I I l\ll N .

H%#xNS

\ I I I I I \l l\ I .

I I , . t I I - t 11 I 1 .

I I 1 - I E i I \ I \ 0 50 250 303 KC) 150 200 Temperature, % I I I .— I I o 100 200 300 403 500 .

Temperature, % Figure 8. - Integrated values of linearcoefficients of e~sion for severalmetala. lkta obtelnedfrom references8 to 10.

.

—.

.

NACA RM E55F22 b .026 .022 .018 I .014 .

l 010 ..

-400 -300 -200 -1oo 0 100 200 Temperature, OF Figure 9. - Variation of thermal conductivity of styrofoam with temperature. Data obtained from reference 23.

.

d T.

.- .= NACA RM E55F22 .

fl~nsion space Mylar-aluminum foil covering Foamed plastic insulation _ Liquid hydrogen .

.

Figure 10. - Liquid-hydrogen fuel tank.

— NACA RM E55F22 .

— .

50 % Ambient t&uperature, w A o - .

/ -50 - — -.

_.. .

-1oo

I

.

!$

$ .

-200 -250 -300 .

1 2 3, - — Insulation thickness, L, {n. = Figure 11. - Variation of surface temperature with in- “. , sulation thickness for ambient temperatures of 80° — ;- and -67° F.

.

!J “.

— -.

300,000 250,000

I

!

Ambient 2c0,000 ~ -temperature ,

-a

d OF t I \ \ 80 \ ~ ~ ~ o 1 2 Insulation thickness, L, in.

- Variation of heat-leak and hydrogen- Flglrre 12.

evaporation rates with insulation thickness.

-NACA RM E55F22 .

— .— .- -.

I I I I 1 I 26 30 34 38 42 46 30 54 58 62 — !leqperature, %?

.

Figure 13. - Eeat cawity of liquld parahydrogen. Jxbta - .- .- 4.

obtained from reference 26.

J NACA RM E55F22 20C / ~ +’ 150- m m ; K #?

.

/ 0 1, 3 2 4 Insulation thickness, L, in.

Figure 14. - Variation of no-loss time with in- sulation thickness. Ambient temperature, 800 F.

.

~~cA - LamleyFlel& va.