vega avum gas tank

AIAA 2007-5500

Page 1

OSR 07-S-1347

DESIGN AND MANUFACTURE OF A HIGH PERFORMANCE, HIGHMASS EFFICIENT GAS TANK FOR THE VEGA AVUM

Joe Benton and Ian BallingerATK-Space Systems Inc.

and

Alberto Ferretti and Nicola IerardoAvio S.p.a.

ABSTRACT

A high performance titanium-lined compositeoverwrapped pressure vessel (COPV) with highmass efficiency for either gaseous helium ornitrogen was designed for ESAs Vega LaunchVehicle. The Vega Gas Tank is common forboth the Liquid Propulsion System (LPS) andthe Attitude Control System (ACS) for the fourthstage of the Vega Launch Vehicles Attitudeand Vernier Upper Module (AVUM).

This tank has a nominal volume of 87 liters(5309 cubic inches) and a nominal weight of 23kg (50.7 pounds). The maximum expectedoperating pressure is 310 bar (4496 psi), 50cycles. Proof pressure requirement is 465 bar(6744 psi), 5 cycles, and the minimum burstpressure is 620 bar (8996 psi). The tank isdesigned to hold 25.94 kg (27.24 pounds) ofgaseous nitrogen (GN2).

The Vega Gas Tank design is based on a flight-qualified pressurant tank to take advantage ofits design and flight heritage. To minimize risk,the Vega Gas Tank is designed to use onlyexisting manufacturing technology.Manufacturing cost is minimized by usingexisting tooling.

Nonlinear material and geometric modelingtechniques were used to analyze this tank.Stress analysis showed positive margins ofsafety for pressure cycle fatigue, vibrationfatigue and minimum burst pressure over thedesign requirements. Development andQualification testing verified the design margins

and showed the design analyses to beconservative.

The liner is constructed from commercially puretitanium. This material was chosen due toheritage and for its superb manufacturability,relative high strength, excellent corrosion andoxidation resistance characteristics, good lowand high cycle fatigue characteristics, andcompetitive manufacturing cost.

The overwrap consists of high strength TorayT1000GB carbon fiber and Epon 826 curedresin system. Several composite layers wereapplied, including helical and hoop wraps.

A complete qualification program wasconducted to verify the tank design, including adestructive burst pressure test. The tanksuccessfully completed qualification testing on29 September 2006.

INTRODUCTION

A gas storage pressure vessel with uniquecharacteristics is needed for the Vega LaunchVehicle propulsion system. This tank must behigh performance, light-weight, and designed towithstand severe operational loads.Additionally, this tank must be built with existingtechnology to minimize manufacturing cost andprogram risk. A titanium-lined, carbon fiberoverwrapped tank was designed andmanufactured to meet such a need. A sketchof this tank is shown in Figure 1.

AIAA 2007-5500

Page 2

OSR 07-S-1347

Figure 1: The VEGA Gas Tank

The tank is mounted to the Launch Vehicle bypolar bosses located on the tank centerlineaxis. The ported boss is attached to theLaunch Vehicle by six M6 bolts. The blindstinger boss mounts on a slip joint bearing.The blind end is designed to accommodate the

tanks axial growth during pressurization. TwoGas Tanks are required for the Vega LaunchVehicle.

The VEGA Gas Tank was designed to thefollowing requirements:

Table 1: VEGA Gas Tank Design Requirements

PARAMETERS REQUIREMENTS

Maximum Expected OperatingPressure (MEOP)

310 bar (4496 psi), 10 cycles minimum

Proof Pressure 465 bar (6744 psi), 5 cycles minimum

Burst Pressure 620 bar (8996 psi) minimum

Gas Weight 25.94 kg (57.2 lb) GN2

Size 336.7 mm x 682.8 mm long, (13.26 x 26.9 long), boss to boss

Overall Length 841.8 mm (33.14 inches) nominal

Tank Weight 23 kg (50.7 lb) maximum

Tank Capacity 87 liters (5309 in3) minimum, unpressurized

Natural Frequency >120 Hz

Compatibility Argon, IPA, Helium, Nitrogen, and DI water

Shell Leakage

AIAA 2007-5500

Page 3

OSR 07-S-1347

This tank is also designed to withstand shockand vibration loads and acceleration of 22.5 gsin any direction. All design requirements wereverified by analysis or qualification testing.

DESIGN HERITAGE

ATK-SSI has designed other titanium linedCOPVs that are similar in design andconstruction. These tanks are also titaniumlined and T1000 carbon fiber overwrapped.The VEGA Gas Tank draws its heritage fromthose programs (Reference AIAA 2002-4349).The manufacturing technology established bythose pioneering programs has matured andthe VEGA Gas Tank program did not attempt toestablish any new technology. The focus of the

tank design was mainly to minimize cost andrisk.

To maximize this flight heritage, the design ofthe VEGA tank blind head is nearly identical tothe blind head of the heritage Pressurant Tank,including the mounting features. The Vegatank ported head also exhibits similar featuresas the Pressurant Tank ported head, as shownin Figure 2. Both liner center sections have thesame method of construction.

The liners of both tanks are made of CPtitanium. The selection of CP titanium wasmade to satisfy the design requirements. Thefilament wrap remains the same T1000 carbonfiber.

Figure 2: Design Heritage of the VEGA Gas Tank

DESIGN ANALYSES

The basic approach in designing the Vega GasTank is to maximize heritage by adapting asmany Pressurant Tank design features aspossible while enhancing the manufacturabilityof the liner and overwrap. To minimize riskonly existing manufacturing technology is used.

Several analyses were conducted to designand analyze the VEGA Gas Tank, including:

Finite element analysis to conduct the linermaterial trade study and selection.

Nonlinear axisymmetric analysis to designthe Gas Tank ported and blind heads.Figure 3 shows the ported and blind headsas modeled by the analysis.

Three-dimensional finite element model forthe modal analysis. The analysis isconducted to predict the natural frequenciesof the Gas Tank. The actual frequency ofthe tank is determined at vibration test.Figure 4a shows the first axial mode andFigure 4b shows the first lateral mode.

Random vibration analysis to determinestress and fatigue effects of randomvibration on the vessel. See Figure 5. Forconservatism, only the qualification levelpower spectral density was analyzed.

Shock analysis to determine stressresponses due to shock. The same finiteelement model for the modal analysis isused on the shock analysis.

OSR 07-S-1347

Figure 3: Nonlinear Axisymmetric Finite Element Models

Figure 4a: First Axial Mode

OSR 07-S-1347

Figure 4b: First Lateral Mode

Figure 5: Three-Dimensional Finite Element Model, Axial (Y)Displacement Response

OSR 07-S-1347

Fatigue analysis to determine the cumulativedamage factor due to fatigue. The fatigue liferequirements for the Vega liner consists of 1sizing (autofrettage) cycle and 2 designservice lifetimes, including proof pressurecycles, operating pressure cycles, and axialand random vibration cycles

LINER DESIGN AND FABRICATION

Prior to designing the Vega Gas Tank, a materialtrade study was conducted to compare materialproperties of CP-3, CP-70 and 6AL-4V titanium toaid material selection. It was found that CP-3 isthe only material that can (1) meet the criteria forelastic behavior between zero pressure andMEOP, (2) meet the stated desire for LBBanalysis instead of LBB demonstration test, and(3) meet the overall tank weight requirement.

Typical of most COPVs, the composite overwrapfor the Vega Gas Tank is designed to providemost of the strength for the tank. The liner is alow load-bearing part of the tank shell that servesas a container to carry the gas and provides adefined shape to apply the filament overwrap. Tominimize weight the liner wall is kept as thin aspractical. However, the design of the liner alsotakes into account the mass property of theheavy gas and the high vibration loads duringlaunch, both conditions that result in high bossloads. The high strength, low weight CP-3titanium is ideal for this application.

Other factors that contribute to the selection oftitanium include:

l Good corrosion and oxidation resistance,l Not susceptible to pitting and stress corrosion,l High strength-to-weight ratio,l Good galvanic compatibility with carbon fiber,l Good low cycle fatigue performance,l Good high cycle fatigue performance,l Good manufacturability,l Good weld properties,l Good performance characteristics.

The Vega Gas Tank liner is a four-piececonstruction that consists of two heads, a cylinderand an outlet tube, as shown in Fig 6.

The outlet tube is made from 6.35mm (0.250inch) outside diameter tubing. The ported headand the blind head are machined from raw

forgings. The center section is fabricated from0.8 mm (0.032 inch) thick CP-3 titanium sheet,rolled, formed, and welded into a cylinder withone longitudinal seam weld. This cylinder ismanufactured using the same manufacturingtechnique as the heritage Pressurant Tankcenter section.

The liner is assembled with two girth welds andthe tube assembly is welded using the sameweld technique and weld schedule as thePressurant Gas Tank. Each weld isradiographic and penetrant inspected foracceptance. The completed liner is leak testedprior to the filament wrap operation.

The Vega Gas Tank liner was designed tomirror the construction of the Pressurant Tankliner. The forging is designed such that bothheads can be machined from the same forgingconfiguration. This modification minimizes thenumber of components for processing, andeliminates several operations such as theassembly weld and the post-weld radiographicand penetrant inspection. This makes theVega Gas Tank inherently more reliable byminimizing the number of welds in the liner.

Figure 6: Components of the Gas TankLiner

Outlet tube

Cylinder

Blind head

Ported head

OSR 07-S-1347

COMPOSITE OVERWRAP DESIGN ANDFABRICATION

The Vega Gas Tank composite overwrapcontains several layers of high angle, helical andhoop wraps. The same wet filament windingtechnique used on the Pressurant Gas Tank isapplied to the Vega Gas Tank. This processutilizes dry fiber roving that is in-processimpregnated with a low-viscosity resin. Thematerials used in the composite overwrap includeToray T1000GB high performance carbon fiberand EPON 826 epoxy resin system. The basicresin system has years of commercial heritageand offers excellent characteristics including: lowviscosity; reasonable pot life; high strain-to-failurecapability; good chemical and moistureresistance; and low toxicity. Thousands ofCOPVs have been wrapped using this resinsystem.

The resin system has a 225F cure temperature.The glass transition temperature (Tg) of the curedsystem is 99C (210F), providing a comfortablemargin over the tanks maximum operatingtemperature of 60C (140F).

A computer-controlled filament winding machineis used to perform the composite overwrapoperation. See Figure 7. A computer code wasgenerated to wrap the development tank. Afterdevelopment testing this code was revised andfinalized to wrap the flight tanks. The entire wrapprocess is automated to insure quality andrepeatability.

Figure 7: Automated Filament Winding

The filament wrap is bonded to the liner by athin layer of adhesive. This adhesive is appliedto the liner immediately prior to the filamentwrap operation. After filament wrap, the vesselis placed in an oven and the resin is gelled andcured.

WEIGHT DISTRIBUTION

The Gas Tank weight distribution issummarized in Table 2 below:

Table 2: Gas Tank Weight Distribution

Item NominalWeight (kg)

NominalWeight (lbm)

Liner 4.31 9.50

Adhesive 1.36 3.00

Composite 17.33 40.8

TOTAL 23.0 50.7

The actual weight of the Qualification tank is21.51 kg (47.44 lb).

OSR 07-S-1347

Table 3: Gas Tank Growth

Tank Pressure Linear Growth Radial Growth

Growth at 1000 psig .030 inch .006 inch







TANK GROWTH

The Gas Tank undergoes expansion as it isbeing pressurized. The tank expansion data forthe Qualification tank is summarized in Table 3.The measured tank growth closely matches thepredicted values.

TANK SIZING

The Gas Tank is subjected to a sizingoperation (autofrettage) after the tank iswrapped and the resin system is cured.The autofrettage pressure is selected toachieve the specification requirement ofelastic behavior between zero pressureand MEOP. This pressurization cycle isconsidered part of the manufacturingprocess and is not included in the pressurecycle history. Autofrettage is performedimmediately prior to acceptance proofpressure testing.

QUALIFICATION TEST PROGRAM

A Qualification Tank was fabricated for thequalification test program. The qualificationtesting consists of a series of tests intended toverify the Vega Gas Tank design in thefollowing areas:

l Physical properties such as volume andweight

l Tank shell integrityl Low cycle fatiguel High cycle fatiguel Burst margin

Pass/Fail criteria consist of acceptance typeexternal leak tests conducted at intervalsthroughout the test program. After the tankpasses the final external leak test, it mustundergo a destructive burst pressure test. Asuccessful burst certifies the tank for flight use.

The Qualification Tank is subjected to thefollowing qualification tests:

l Proof pressure testl Volumetric capacityl External leakagel Pressure cyclesl External leakagel Sinusoidal and random vibrationl External leakagel Pressure Surgel Thermal Cyclingl External leakagel NDE Weld Examinationl Final examinationl Destructive burst pressure test

OSR 07-S-1347

Volumetric Capacity Examination: Thevolumetric capacity of the Gas Tank ismeasured using the weight of water method atambient condition. Deionized (DI) water isused to conduct this test. The tank volumesbefore and after the proof pressure test aremeasured to verify that the tank volume meetsthe specification requirement and that the proofpressure test does not significantly change thetank volume. As an example, the internalvolume of the Qualification Tank did notincrease after the proof pressure test, signifyingthat the Gas Tank was manufacturedsuccessfully.

Proof Pressure Test: The hydrostatic proofpressure test is conducted at 465 bar (6744psig) for a pressure hold period of 5 minutes.Successful completion of the proof pressuretest and the subsequent volumetric growth andleakage verification indicate that the tank wasmanufactured successfully.

External Leak Test: The external leak testverifies the integrity of the tank shell and alsoserves to validate the previous series ofpressure testing. The tank is placed in avacuum chamber, which is evacuated to under0.2 microns of mercury, and helium pressurizedto MEOP for 30 minutes. The helium leak ratecannot exceed 1 x 10-6 std cc per second aftera 30-minute stabilization period. For example,the leak rate of the Qualification Tank was 1.9 x10-7 scc/sec.

During pressurization, the compressed gasheats up, thus heating up the tank. To preventoverheating, four thermocouples are attachedto the tank shell to monitor and control thepressurization rate and the tank temperature

during pressurization. The tank temperaturecannot exceed 140F throughout the durationof the test.

Pressure Cycles: The Gas Tank is designedto accommodate a minimum of 20 proofpressure cycles and 50 operating pressurecycles. Additionally, the Qualification Tankexperienced another operating pressure cycleduring volume measurement, 3 operatingpressure cycles for the 3 external leakage tests1 operating pressure cycle in the pressuresurge test, 3 more operating pressure cyclesduring vibration testing and 10 more operatingcycles during thermal cycling. The cumulativetotal of operating pressure cycles is 58, or 8cycles over the minimum requirement.

Pressure Surge: The pressure surge test wasperformed with the tank pressurized with GN2to MEOP 310 bar (4496 psi) the dischargedinto a 120 liter pressure vessel though a 1.9mm orifice. The test noted no anomalies.

Vibration Test, Sinusoidal and Random:Qualification level sinusoidal and randomvibration tests were performed on theQualification Tank in each of the three principalaxes. The vibration test requirements areshown in Tables 4 and 5.

The vibration test fixture is designed to simulatethe tank-to-Launch Vehicle installationinterfaces and orientation. It is also sufficientlystiff to be considered rigid for the testfrequencies. A preliminary test fixtureevaluation was conducted prior to QualificationTank installation to insure the fixture meets thetesting requirements.

OSR 07-S-1347

Table 4: Qualification Level Sinusoidal Vibration Test Environment

Qualification Sine Vibration Environment

Axes Frequency(Hz)

InputLevel (G)

Sweep Rate

5 16 10 mm 1/3 oct/min

X, Y &Z

16 35 10.0 g (11.4 min duration

35 - 70 22.5 g from 5-70 Hz)

70 - 200 22.5 g200 - 2000 10.0 gX, Y &

Z

2 oct/min(2.4 min duration from 70-

2000 Hz )

Table 5: Qualification Level Random Vibration Test Environment

Frequency (Hz) PSD (G2Hz) Input

Level All Axes

Overall Level

(Grms)

20-60 +3 dB/oct

60-1000 0.273

1000-2000 -6 dB/oct20

Total Duration 4 minutes per Axis

Control accelerometers were placed on thevibration test fixture near each end fitting tocontrol the vibration input. Responseaccelerometers were placed on theQualification Tank to measure the tankresponses. The location of the responseaccelerometers were selected to record themaximum stress as predicted in the analyticalmodel. The vibration test included fixturesurvey, resonant frequency search, full level

sinusoidal and full level random runs. Thesame tests were conducted in all three axes.

The test was conducted with the tank loadedwith GN2, pressurized to MEOP, and subjectedto the test environment in Tables 4 and 5. .

A photograph of the vibration test setup isshown in Figure 8.

OSR 07-S-1347

Figure 8: Qualification Tank Vibration Test Setup

Figure 9: Qualification Tank After Burst

OSR 07-S-1347

Destructive Burst: After the completion of thepressure cycles, thermal cycling and vibrationtesting, the Qualification Tank was subjected toa final destructive burst pressure test. TheQualification Tank burst at 904 bar (13117 psi),providing a 45% margin on burst pressure.This data represents a burst factor of 2.92 to 1,and a performance efficiency rating (PV/W) of1.4. Figure 9 shows the Qualification Tank afterburst.

Qualification Tank Pressure Log: Insummary, the Qualification Tank hasundergone the pressure cycles listed in Table6. The tank had been either inadvertently ordeliberately over tested during the rigorous testprogram. The successful completion of thequalification test program is an excellentdemonstration of the tanks robust design.

Table 6: Summary of Qualification Tank Pressure Cycles

Pressure Cycles No. of Cycles

MEOP Cycles: 310 bar (4496 psig) 58

Test Cycle: 414 bar (6000 psig) 1

Proof Cycles: 465 bar (6744 psig) 20

Autofrettage: 466 bar (6760 psig) 1

Burst 620 bar: (8996 psig) 1

OSR 07-S-1347

ACCEPTANCE TESTING

The following acceptance tests areperformed on a flight tank prior to delivery:

l Preliminary examinationl Pre-proof volumetric capacityl Ambient proof pressurel Post-proof volumetric capacityl External leakagel NDE Weld Examinationl Final examinationl Cleanliness measurement

Cleanliness Verification: After the finalexternal leak test, each flight tank is cleanedto the cleanliness level specified in Table 7.

Table 7: Gas Tank Cleanliness Level

Particle SizeRange (Microns)

Maximum Allowedper 100 ml

Less than 5 No Silting

6 to 10 1200

11 to 25 200

26 to 50 50

51 to 100 5

Over 100 0

Photograph of a completed tank is shown inFigure 10.

Figure 10: A completed VEGA GasTank

CONCLUSION

The Vega Gas Tank program hassuccessfully concluded qualification testingwithout failure. The qualification testingshows the Gas Tank having comfortablemargins over all the operationalrequirements.

The Vega Gas Tank is high performance,light weight, and easy to manufacture. Thecomposite overwrap and the linercomponents are made from commerciallyavailable materials. The liner assembly andfilament winding are accomplished usingstandard manufacturing processes andprocedures. Special material and processesare not required.

This tank is also lighter than a typical all-metal tank of the same capacity andcapability. The manufacturing cycle isseveral months shorter than a comparableall-metal tank. Acceptance testing is simpleror equivalent to an all-metal pressurant tank.

The Vega Gas Tank maintains excellentdesign and flight heritage. Its overall designand method of manufacturing are derivedfrom several prior Pressurant Tankprograms. The design of this Gas Tank isextremely conservative and all manufacturingmethods are based upon existing technologywith a resulting high performance and highmass efficiency. .

Most importantly, the successful qualificationof this tank marks the milestone in which aderivative COPV is manufactured efficientlyand inexpensively using existing technology.

OSR 07-S-1347

ACKNOWLEDGMENT

Our sincere thanks to members of the Avioand ATK-SSI team for making this program asuccess.

REFERENCE

1. W. Tam, P. Griffin and Art Jackson,Design and Manufacture of a CompositeOverwrapped Pressurant TankAssembly, AIAA 2002-4349.

ABOUT THE AUTHORS

Mr. Joe Benton a Program Manager at ATK-SSI (Commerce). Mr. Ian Ballinger is ChiefEngineer at ATK-SSI.

Mr. Albert Ferretti is the TechnicalResponsible Engineer at Avio and Mr. NicolaIerardo is a Technical Responsible Engineerat Avio.

.

vega avum gas tank

Documents