www.elsevier.com/locate/cryogenics

Cryogenics 46 (2006) 111–117

Aerogel insulation systems for space launch applications

J.E. Fesmire *

NASA Kennedy Space Center, Cryogenics Test Laboratory, Kennedy Space Center, Mail Code: YA-C2-T, KSC, FL 32899, USA

Received 7 November 2005; accepted 8 November 2005

Abstract

New developments in materials science in the areas of solution gelation processes and nanotechnology have led to the recent commer-cial production of aerogels. Concurrent with these advancements has been the development of new approaches to cryogenic thermal insu-lation systems. For example, thermal and physical characterizations of aerogel beads under cryogenic-vacuum conditions have beenperformed at the Cryogenics Test Laboratory of the NASA Kennedy Space Center. Aerogel-based insulation system demonstrationshave also been conducted to improve performance for space launch applications. Subscale cryopumping experiments show the thermalinsulating ability of these fully breathable nanoporous materials. For a properly executed thermal insulation system, these breathableaerogel systems are shown to not cryopump beyond the initial cooldown and thermal stabilization phase. New applications are beingdeveloped to augment the thermal protection systems of space launch vehicles, including the Space Shuttle External Tank. These appli-cations include a cold-boundary temperature of 90 K with an ambient air environment in which both weather and flight aerodynamicsare important considerations. Another application is a nitrogen-purged environment with a cold-boundary temperature of 20 K whereboth initial cooldown and launch ascent profiles must be considered. Experimental results and considerations for these flight systemapplications are discussed.Published by Elsevier Ltd.

Keywords: Polymers; Thermal conductivity; Transport properties; Cryostats; Space cryogenics

1. Introduction

Much has been made in recent years of the applicationof aerogels for thermal insulation. The roots of aerogelmaterials extend back to the 1930s when Kistler inventeda process for making a porous inorganic silica product[1]. About 50 years later modern aerogels were developed,coinciding with the advent of solution–gelation (sol–gel)chemical processing methods. These new aerogels are dis-tinguished by their very fine pore sizes (from about 1 to20 nm). The very small pores are synonymous with veryhigh-surface areas (such as from 800 to 1100 m2/g) of theinternal structures. These new aerogels can be made tohave ultra-low densities. The modern sol–gel processing

0011-2275/$ - see front matter Published by Elsevier Ltd.

doi:10.1016/j.cryogenics.2005.11.007

* Tel.: +1 321 867 7557; fax: +1 321 867 4446.E-mail address: James.E.Fesmire@nasa.gov

techniques behind these materials allows for tailoring themolecules for a given application. The modern aerogelsare often characterized as having an ultra-low density inthe monolithic (solid) form, but this is not an essential dis-tinguishing feature. Application of aerogels for thermalinsulation were studied by Hrubesh [2], Smith et al. [3],and others.

In the last five years large-scale commercial productionof the modern, nanoporous aerogel materials has beenachieved. The financial motivation for this business devel-opment was due, in large part, to the potential applicationsin thermal insulation. For example, products now availableinclude aerogel beads manufactured by Cabot Corporation[4] and aerogel composite blankets manufactured by AspenAerogels, Inc. [5]. Because they are fully breathableand hydrophobic, these materials are ideal candidates forthermal insulators in a number of space launchapplications.

0.1

1

10

100

0.01 0.1 1 10 100 1000 10000 100000 1000000

Cold Vacuum Pressure (millitorr)

App

aren

t T

herm

al C

ondu

ctiv

ity

(mW

/m-K

)

Aerogel Blanket, 16 mm, 51 kg/m3

Spray Foam, 51mm, 38 kg/m3

Aerogel Beads, 25 mm, 81 kg/m3

Perlite Powder, 25 mm, 115 kg/m3

MLI, 60 layers, foil & paper, 25 mm

Fig. 1. Apparent thermal conductivity as a function of cold vacuumpressure for aerogels compared to different insulation materials. 78 K and293 K boundary temperatures; residual gas nitrogen.

112 J.E. Fesmire / Cryogenics 46 (2006) 111–117

2. The insulation problem on space launch vehicles

As a general rule, cryogenic thermal insulation systemsshould be either completely sealed or fully breathable withthe surrounding environment. Systems that are partially orimperfectly sealed can lead to reliability or safety problemsbecause the residual matter within is not fully known. Onemajor concern with space launch vehicles is the migrationof air or water molecules from the warm side to the coldside during the transient cooldown processes of the cryo-genic propellant loading operations. Dense air, ice, water,and (in the case of liquid-hydrogen tanks or pipes), lique-fied air, can accumulate within the insulation materialsbecause of this mass-transport effect (commonly referredto as cryopumping). These accumulations can degradevehicle performance in several ways, including increasedheat transfer through the thermal insulation, increasedlift-off weight, and the potential for damaging ice debris.

The problem of accumulated matter within the insula-tion is further compounded when the launch vehicle mustascend from the ambient pressure environment of thelaunch site to the high-vacuum environment of space.The molecules trapped inside the insulation materials mustthen escape to the lower pressure side without damagingthe thermal protection system, adjacent hardware, or flightsurfaces. The rapid heating effects due to the aerodynamicsat high mach numbers only serve to further worsen theproblem.

3. Cryopumping effects in aerogels

Cryopumping is strictly and simply defined as the pro-duction of a vacuum using low temperatures. The fourbasic mechanisms of cryopumping are cryocondensation,cryosorption, cryotrapping, and cryogettering. The mecha-nisms cryocondensation and cryosorption, both broughtabout by van der Waals’ forces, are most relevant to theaerogels used on liquid hydrogen or liquid oxygen systems[6]. When gas molecules impinge a low-temperature surfacethey lose much kinetic energy and can become attached tothat surface. Eventually an equilibrium point will bereached where the rates of particles impinging and depart-ing the surface are equal. Applied to a thickness of porousinsulation material, this point would be one of an infinitenumber of isotherms. Cryopumping effectively ceasesfor a well-insulated system after initial cooldown isachieved.

Aerogels are high-performance thermal insulators inambient or partial vacuum environments because of theirextremely small pore sizes which greatly impede the com-munication of gas molecules by thermal conduction. Aproperly designed thermal insulation system based on aero-gel material will also minimize all interstitial spaces withinthe system, thereby lowering the heat transfer caused bygas convection. During the transient process of cooldownfrom ambient to cryogenic temperature, the gases withinthe system will, to some degree, tend to migrate to the cold

side and condense or freeze. But these cryopumping effectsare lessened because of the very fine pore size and extraor-dinarily low thermal conductivity of the aerogel. Further-more, because the aerogels are such good thermalinsulators, the isotherms between the warm side and thecold side are more readily established. The result is thatcryopumping essentially stops after the transient cooldownperiod. The near steady state condition typical of ambientpressure systems must be maintained for approximately 8 hto be sufficient for the launch vehicle propellant condition-ing and hold period.

The main work of the cryogenic insulation system isdone at lift-off. The transient warm-up phase, starting atlaunch and continuing during the several minutes of ascentinto the vacuum of space, allows the additional trappedmolecules to readily escape through the porous aerogel intothe lower pressure environment with no harm to the vehiclethermal protection systems.

4. Aerogel thermal performance characterization

Looking in closer detail at the physical characteristics ofaerogels shows the possibilities for new thermal insulationsystems for both low-temperature and high-temperatureuses. Fig. 1 presents the variation of the apparent thermalconductivity with the cold vacuum pressure for an aerogelblanket (two layers, 16 mm total thickness) and an aerogelbead (1 mm nominal diameter, 25 mm total thickness)

50

75

100

125

150

175

200

225

250

275

300

0 1 2 3 4 5 6 7 8

Time (hours)

Tem

pera

tute

(K

)

500

550

600

650

700

750

800

Vacuum

Cham

ber Pressure (torr)

Cold BoundaryTemperatureLN2 Feedthrough TemperatureUpper Chamber TemperatureWarm Boundary TemperatureVacuum Chamber Pressure

Fig. 3. Cryopumping experiment using aerogel beads. Residual gasnitrogen. Closed mass system shows stabilized pressure approximately2 h after liquid-nitrogen filling is complete.

J.E. Fesmire / Cryogenics 46 (2006) 111–117 113

material. Comparisons with three commonly used cryo-genic insulation materials, polyurethane spray foam, perlitepowder, and multilayer insulation (MLI), are also given.These experimental data were obtained using a liquid nitro-gen boil-off cryostat in a cylindrical configuration. Detailson this new standardized equipment and method are givenin the literature [7,8].

The foam material, type BX-250, was tested at a thick-ness of 51 mm and with a measured density of 38 kg/m3.The thermal performance data for the aerogel beads andperlite powder has been previously reported [9]. Detailson the 60-layer MLI system shown here have also beenpreviously reported [10].

Both aerogel materials are silica-based, fully breathable,and treated to be hydrophobic. The aerogel blanket wascomposed of two layers, 8 mm thickness each, with anoverall bulk density of 51 kg/m3. Details on aerogel com-posite blanket materials are given in the papers by Fesmireand Ryu [11,12]. The aerogels beads are 90% porous with amean pore diameter of 20 nm and an internal surface areaaround 800 m2/g. With a bead density of 140 kg/m3 and atypical bulk density of 80 kg/m3, the interstitial space isapproximately 60%. The beads have a fully elastic com-pression capability of well over 50% with no significantdamage. The yield strength of an individual aerogel particleallows for volumetric compression of up to 30% with fullreversible behavior [13]. This packing feature supportsthe application of this material as a thermal insulator forambient pressure environments. Further information onthe aerogel bead materials can be found in the paper byAckerman et al. [14].

To investigate the different mass-transport effectsthrough the nanoporous aerogel materials and show theirrelevance to low-temperature thermal insulation perfor-mance, a number of cryopumping experiments weredevised. Two of these experiments using aerogel beadsare described in the following sections.

Fig. 2. Cryopumping experimental test apparatus us

5. Cryopumping experiment using bulk-fill cryostat

A research test cryostat was filled to a level of approxi-mately 200 mm above the top of the liquid nitrogen coldmass tank as depicted in Fig. 2. The vacuum chamberwas stabilized with room air at ambient conditions to sim-ulate the launch vehicle environment. The closed mass sys-tem was then cooled from approximately 293–77 K whilethe drop in pressure from 101 kPa (760 Torr) was moni-tored. The system pressure was lowest at approximately80 kPa (600 Torr) in less than 1 h from the start of cool-down. Pressure stabilized about 2 h after liquid-nitrogenfilling was completed (Fig. 3). The feedthrough tempera-ture trace during cool-down and warm-up explains thetemporary increase in pressure between the 2-h and 3-hmarks. It is interesting to note that this same base pressurewas reached upon filling, indicating that the cryopumpingoccurs only in the close proximity to the cold mass surfaceand only at the beginning of the process.

ing cryostat loaded with bulk-fill aerogel beads.

50

100

150

200

250

300

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Time (hours)

Tem

pera

ture

(K

)

TC6 (125 mm)

TC5 (100 mm)

TC4 (75 mm)

TC3 (50 mm)

TC2 (25 mm)

TC1 (0mm)

Fig. 6. Cryopumping with liquid-nitrogen reservoir. Isotherms in aerogelbeads column form after approximately 3 h. Room air on warm side.

114 J.E. Fesmire / Cryogenics 46 (2006) 111–117

6. Cryopumping experiment using cold-column apparatus

A cold-column apparatus was devised to simulate theinternal launch vehicle environment of a large purged cav-ity of gaseous nitrogen at 293 K in contact with a cold sur-face at 20 K. The thermal similarity between the flightvehicle case and the experimental case is depicted inFig. 4 showing the liquid-hydrogen tank surface, the thinlayer of solid nitrogen, and the thin layer of liquid nitrogen.A 150-mm-tall column of aerogel beads was then inter-posed with temperature sensors at 25-mm intervals (seeFig. 4). The aerogel was held within the G-10 tube columnusing a fiberglass cloth at the bottom. The top surface wasopen to room air. The column was set into a pair of dewarflasks and temperatures were monitored at the differentheights within the column. The liquid-nitrogen level wasmaintained manually by transfer from a third dewar flask.The bottom of the column was inserted 12.5 mm below thesurface of the liquid nitrogen (Fig. 5).

Several long-duration runs of the experiment were madewith similar results. The aerogel-bead-filled column tem-peratures as a function of time are presented in Fig. 6. Iso-

Fig. 4. Thermal similarity between the flight vehicle and the experimental case otemperature of 20 K.

Fig. 5. Cryopumping experimental apparatus with liquid-nitrogen reservoir(R) overall apparatus of aerogel-filled column set into dewar flask.

therms formed in approximately 3 h from the start ofcooldown (initial immersion of the lower end of the col-umn). The results show that liquid quickly filled to somelevel between 0 mm and 25 mm, and that there was noappreciable change even after the 12-h runtime. The

f pooled liquid nitrogen within a large purged cavity with a cold-boundary

and cold column: (L) portion of thermocouple sensors inside column,

J.E. Fesmire / Cryogenics 46 (2006) 111–117 115

isotherms at 50 mm, 75 mm, and so forth were as expectedaccording to previous thermal conductivity measurementsof the aerogel beads [9]. In this case, cryopumping withinthe entire column ceased after about 3 h, at which timestate–state conditions were observed and kept for another9 h. Note that the periodic variations in temperature areattributed to the manual method of liquid-nitrogen levelcontrol. The slow warm-up process is seen from hours 12to 18 as the sequestered nitrogen molecules are liberatedfrom the fine pores of the aerogel.

7. Flight vehicle application: liquid-oxygen feedline

bellows

A liquid-oxygen feedline with exposed metal bellows is agood candidate for an aerogel-based insulation systembecause the bellows must be free to move without collect-ing any ice. Ice is a primary hazard for the vehicle surfacesand could also impede the mechanical articulation of the

Fig. 7. Views of the Space Shuttle External Tank. The right photo shows a portank.

Fig. 8. Liquid-oxygen feedline bellows shown schematically on the left. The raerogel-based thermal insulation system.

bellows assembly. A photo of the feedline area on theExternal Tank of the Space Shuttle is shown in Fig. 7.The aerogel insulation system has undergone developmenttesting as seen in Fig. 8.

This testing demonstrated the aerogel-based system tobe fully effective in eliminating frost or ice. Another keyaspect demonstrated was the hydrophobic nature of theaerogel such that no ice formations occurred within theinternal cold cavities during the 8–10 h of simulatedground-hold conditions. Cryopumping effects were stoppedupon reaching thermal stabilization through the thicknessof the insulation materials. The aerogel beads, being nano-porous, tightly packed, and hydrophobic, further inhibitedthe movement of moisture or air into the cavity.

8. Flight vehicle application: liquid-hydrogen tank dome

A liquid-hydrogen tank dome with an exposed peripheryof cold metal is a good candidate for an aerogel-based

tion of the liquid-oxygen feedline and upper portion of the liquid-hydrogen

ight photo shows full-scale thermal–mechanical testing in progress of the

116 J.E. Fesmire / Cryogenics 46 (2006) 111–117

insulation system because the steady accumulation of liq-uefied nitrogen from the large purged cavity above thedome will lead to additional flight mass and could causecollateral damaging effects to the flight thermal protectionsystem on the vehicle’s outside surface. Fig. 7 shows theupper portion of the liquid-hydrogen tank (the dome toflange area) on the Space Shuttle. A photo of the one-tenth-scale test article for demonstrating another aerogel-based insulation system is given in Fig. 9.

The unit is sprayed with foam insulation except for onenarrow area about the circumference of the upper dome ofthe tank. Liquid helium is used to obtain the necessarycold-boundary temperature of approximately 20 K whilethe pressure inside the nitrogen-purged cavity above thetank dome is monitored. The mass liquefied is calculatedfrom knowing the pressure drop inside the fixed volumeof the cavity. The experiment may also be conducted bysupplying a nitrogen purge gas flow into the cavity to main-tain a constant pressure. Again, the totalized mass flow is

Fig. 9. A demonstration unit for a liquid-hydrogen tank application isshown being prepared for evaluating another aerogel-based insulationsystem. The photo on the top shows the overall test apparatus includingthe data acquisition system and the liquid helium supply dewar. Thecutaway view on the bottom shows the main structural features of thesimulated liquid-hydrogen flight tank.

directly related to the amount of nitrogen liquefaction onthe exposed cold metal surface of the tank dome. Witheither method, constant mass or constant pressure, thecryopumping effects as a function of propellant loadingtime can be determined. The objective of this demonstra-tion testing is to show that the aerogel-based insulation sys-tem stops liquefaction of the nitrogen purge gas aftercooldown is complete, as shown in the research experimentwith the cold-column apparatus.

9. Conclusions

Aerogels provide new possibilities for thermal insula-tion systems for both low-temperature and high-tempera-ture applications on space launch vehicles. Experimentalresearch testing shows that the cryopumping effectsthrough the aerogel-based material systems, even thoughthey are open-cell porous molecular networks, effectivelycease upon completion of cooldown. Further testing hasshown these materials to be fully hydrophobic with notendency to produce ice-laden debris that could be haz-ardous to a vehicle during flight. Current system demon-stration tests at the Cryogenics Test Laboratory confirmproblem-solving applications for existing space launchvehicles including the Space Shuttle. All launch vehiclesmust have thermal protection systems that function reli-ably in both ambient pressure and high-vacuum environ-ments. Future vehicles that must be robust or reusablewill likely be enabled by the aerogel technologies nowbeing developed.

Acknowledgments

The author thanks Stan Augustynowicz for performingthe thermal characterization and analysis of the materials.The author thanks Charles Stevenson and Trent Smith fortheir valuable insight and support throughout this researchwork.

References

[1] Kistler SS. The relation between heat conductivity and structure insilica aerogel. J Phys Chem 1935;39:79.

[2] Hrubesh LW. Aerogel applications. J Non-Cryst Solids 1998;225:335.

[3] Smith DM, Douglas M, et al. Aerogel-based thermal insulation.J Non-Cryst Solids 1998;225:254.

[4] Cabot Corp. Available from: http://www.cabot-corp.com/nanogel.[5] Aspen Aerogels, Inc. Available from: http://www.aerogel.com/

products.htm.[6] Haefer RA. Cryopumping, theory and practice. Oxford: Clarendon

Press; 1989. p. 54–5.[7] Fesmire JE, Augustynowicz SD, Heckle KW, Scholtens BN. Equip-

ment and methods for cryogenic thermal insulation testing. AdvCryog Eng 2004;49:579–86.

[8] Fesmire JE, Augustynowicz. Methods of testing thermal insulationand associated test apparatus. US Patent 6,742,926 June 1, 2004.

[9] Fesmire JE, Augustynowicz SD, Rouanet S. Aerogel beads ascryogenic thermal insulation system. Adv Cryog Eng 2004;47:1541–8.

J.E. Fesmire / Cryogenics 46 (2006) 111–117 117

[10] Fesmire JE, Augustynowicz SD, Demko JA. Thermal insulationperformance of flexible piping for use in HTS power cables. AdvCryog Eng 2004;47:1525–32.

[11] Fesmire JE, Rouanet S, Ryu J. Aerogel-based cryogenic superin-sulation. Adv Cryog Eng 2004;44:219–26.

[12] Fesmire JE, Ryu J. Improved aerogel-based thermal insulationsystems, NASA Tech Briefs, June 1999. p. 56.

[13] Woignier T et al. Plastic densification in silica aerogels. Eur Phys JAppl Phys 1999;6:267–71.

[14] Ackerman WC, Vlachos M, Rouanet S, Fruendt J. Use ofsurface treated aerogels derived from various silica precursors intranslucent insulation panels. J Non-Cryst Solids 2001;285(1–3):264–71.