Space Propulsion 2018 BARCELO RENACIMIENTO HOTEL, SEVILLE, SPAIN / MAY 2018

Transcription

1 Space Propulsion 2018 BARCELO RENACIMIENTO HOTEL, SEVILLE, SPAIN / MAY 2018 Design and Qualification of Fuel and Oxidizer Tank Assemblies for the James Webb Space Telescope Walter Tam (1) and Don Jaekle (2) (1) ATK Space Systems, Inc., 6033 E. Bandini Blvd, Commerce, CA U.S.A 90040, walter.tam@orbitalatk.com (2) PMD Technology, 5190 W. Indian View Lane, Wilson, WY 83014, don@pmdtechnology.com Keywords PMD, PMD Propellant Tanks Acronyms CoM = Center of Mass PMD = Propellant Management Device Abstract The James Webb Space Telescope (James Webb) is an infrared space observatory scheduled to launch in The tank development for James Webb took nearly ten years to complete. Early exploratory discussions started in May Final tank delivery was in July The James Webb tank development was not typical. Additional program emphases had included mass minimization and propellant center of mass (CoM) control. Designers and engineers spent six and a half years conducting trade studies in support of requirements and scope definition, and another three and a half years designing, fabricating, and testing hardware. In addition to developing and qualifying new tank shells, designing propellant management devices (PMDs) to meet tight CoM control requirements and building structurally durable PMDs to meet severe launch pad slosh environments were all major challenges. There are four sections in this summary paper. Section 1 is an introduction to the tank development program and background information on tank shell and PMD designs. In Section 2, we describe the tank shell development, the PMD development, and hardware manufacture. Section 3 is a summary of the qualification test program, including a longduration slosh test. In Section 4, we conclude with a review of the program qualification, and compare and contrast James Webb tank qualification with other tank qualification programs. 1. Introduction The James Webb Space Telescope is an infrared space observatory currently nearing the completion of a long development cycle. NASA managers plan to launch the James Webb spacecraft into service on board an Ariane 5 rocket in Amongst the many innovative equipment and instruments aboard the spacecraft are two propellant tanks custom designed to enable the James Webb mission. Designers and engineers developed these propellant tanks using state-of-the-art analysis approaches, heritage design and manufacturing processes, and innovative testing equipment and techniques. Most importantly, the adaptation of a holistic tank development approach ensured high reliability, optimal performance, and affordability. Extraordinary care went into program scope development and definition. Development activities started in May 2005 with exploratory trade studies. Our last proposal to the customer was revision U, indicating 20 revisions to the original proposal before finalizing the program scope. After winning the contract under competitive procurement, we started tank development in January Final flight tank delivery was in July The James Webb tank development program was not typical. Hundreds of prior ATK tank development programs had primary emphasis on achieving performance targets within cost, mass, technical, and schedule constraints. In addition to these key constraints, the James Webb tank development must also include a special emphasis on mass minimization. Within the past twenty-five years, only the MESSENGER tank development had a similar emphasis on mass minimization [1]. The tank development cycle was longer than most other tank development programs. On commercial programs, investors usually demand a rapid time to market in order to generate revenue and justify return on investment. On scientific or civil programs, product developments must support fixed launch windows. On the James Webb tank development program, we needed a long product development window to ensure that we optimized all the program objectives. Throughout the development cycle, engineers conducted multiple trade studies before program start and during program execution to maintain focus on key objectives. Finally, engineers and designers subjected the PMD tanks to significant structural testing. In Section 3, we present the test regimes in detail Orbital ATK. All Rights Reserved. 1

2 Our first contact with the customer dated back to May During initial contact, the study focus was to find a tank solution based on available hardware. Conducting a tank trade study has always been a good way to optimize available resources and options in the marketplace [2]. Finding a derivative tank approach is typically the best way to start any tank development effort. We traded existing shells near 711 mm (28 in) and 635 mm (25 in) in diameter. See Figures 1a and 1b. During initial trade, we found few derivative tank options because of multiple constraints such as size and envelope. See Figure 2. Figure 2: James Webb tank envelope constraints Figure 1a: Preliminary trade study focusing on existing 711 mm diameter shells Figure 1b: Preliminary trade study focusing on existing 635 mm diameter shells As requirements matured, PMD trades became part of the trade space. See Figures 3a and 3b. One important trade was to compare different PMD designs and their associated propellant control capabilities. In addition, we also traded PMD configurations with different mass implications. The focus on mass had led to discussions on using a hybrid shell design to reduce mass [3]. We determined that this approach was not suitable for several reasons. First, a hybrid design might be efficient for a cylindrical tank, but it is not optimal for a spherical tank. Second, an optimized approach might be to qualify a hybrid shell for the fuel tank and an all-metal spherical shell for the oxidizer tank. However, this approach would require two stand-alone qualification programs, thus becoming cost prohibitive. Third, a hybrid tank approach would add another layer of design complexity to an ultraconservative program. Finally, the potential mass savings of the hybrid tank approach was minimal and not financially justifiable. The trade study conclusion was to design all-metal tanks to enable a streamlined qualification test campaign for both the fuel and oxidizer tanks. Several factors led to the final decision to develop new tank shells instead of pursuing a derivative shell approach. First, given the different volume requirements for the fuel and oxidizer tanks, the best mass solution would have been a larger spherical fuel tank and a smaller spherical oxidizer tank. However, this solution was prohibitively expensive because of the need to run two independent qualification programs. The solution was also not practical because of a preexisting envelope constraint. The next best solution was to have a spherical oxidizer tank and a cylindrical fuel tank both using existing domes. However, although ATK has several existing domes that might meet the volume requirements, none could meet the envelop constraints. Finally, engineers and program administrators decided upon developing a new common dome for the two tanks. The larger fuel tank would have two 668 mm (26.3 inch) diameter domes and a 294 mm (11.56 inch) cylinder section. The smaller oxidizer 2

3 tank would be a sphere using the same domes as the fuel tank. The tank qualification approach was to conduct a qualification test program to qualify the fuel tank, and perform a qualification-bysimilarity analysis and protoflight testing to qualify the oxidizer tank. This tank solution met the size, volume, and envelope constraints with the lowest possible mass while incurring only one set of nonrecurring expenses on design, tooling, and qualification testing. For the James Webb tank development program, this approach was the optimal solution. The James Webb tank development was an example of forgoing the conventional derivative tank approach in order to optimize tank design and meet program objectives. This approach required a long list of developmental items from new tools to new tank qualification [4]. Adapting this approach ensured optimized fuel and oxidizer tank designs with minimum mass. From the onset of program development, minimum tank mass was the top priority for the customer that set the tone for the tank development program. Figure 3a: PMD trade study to trade CoM control capabilities Figure 3b: PMD trade study to trade PMD options with different mass implications 3

4 At the conclusion of a series of trade studies, the customer consolidated the study results and merged the knowledge into an equipment specification. The fuel and oxidizer tank specification included the following design parameters: Table 1: P/N Fuel Tank and Oxidizer Tank Design Parameters Parameters Operating Pressure Proof Pressure Burst Pressure Material of Construction Vacuum Rated Tank Mount Inside Diameter Requirements 24 bar ( F 30 bar ( F >36.2 bar (> F. (The Qualification Tank bar or 812 psig). Shell: Solution Treated and Aged (STA) 6AL-4V Titanium Heads and Cylinder Inlet/Outlet Ports: 6AL-4V titanium tubes PMD: 6AL-4V Titanium, annealed, bar stock or sheets Yes (1 atm) Polar bosses Fuel Tank (P/N 80545): 668 mm ( 26.3 in) ID x 962 mm (37.86 in) long Oxidizer Tank (P/N 80546): 668 mm ( 26.3 in) ID sphere Tank Volume Fuel 257 liters (15,685 in 3 ) Oxidizer 131 liters (7,996 in 3 ) Propellant Volume Tank Mass Natural Frequency Shell Leakage Fuel Tank qualified to 178 kg (392 lb m) of N 2H 4, Oxidizer Tank qualified to 132 kg (291 lb m) of NTO. Fuel Tank <16.5 kg (36.4 lb m), (mass of Qual 15.4 kg or lb m, including fittings) Oxidizer Tank <11kg (24.2 lb m) >40 Hz in any direction <1x10-6 std cc/sec He MEOP In Figure 4, we present the conceptual sketches of the James Webb fuel and oxidizer tanks. Some important design features include: The fuel tank has two hemispherical domes and a cylindrical center section The oxidizer tank is a spherical tank with the same hemispherical domes as the fuel tank The fuel tank has a large sponge custom designed to optimize performance in a cylindrical tank The oxidizer tank has a similar sponge PMD, but custom designed to have a profile that optimizes performance in a spherical tank On both tanks, both the pressurant and propellant ports are on the same bottom boss On both tanks, the center posts extend onto the upper domes for structural support Figure 4: The James Webb Fuel and Oxidizer Tank Concepts 2. Tank Development Qualifying a new tank size requires the development of a long list of items such as tools, fixtures, documents, and reports [4]. On the James Webb program, there was an added challenge of developing not one but two shell configurations. The two shells are similar, but not identical. Similarly, the two PMDs are alike, but not identical. When working on the design for one of the tanks, engineers and designers must be mindful of the tooling, cost, mass, performance, and schedule implications to the accompanying tank. On the James Webb program, the new items needing development included the following: Forging dies for domes common to both Mandrels and rings for machining common to both Fixtures for tank shell assembly common to both Fixtures for PMD assemblies for both Fixtures for tank shell and PMD heat treatment for both Fixtures for testing for both fuel and oxidizer tanks 4

5 Fixtures for inspection common to both Fixtures for cleaning for both fuel and oxidizer tanks Fixtures for vibration testing for both fuel and oxidizer tanks Simulators for slosh testing for both fuel and oxidizer tanks Dome forging qualification program common to both Dome forging qualification report Ring forging qualification program Ring forging qualification report Tank heat treatment program common to both Girth weld qualification program common to both PMD quality verification program for both PMD tank qualification program for both PMD tank qualification report for both Design review data packages for both Drawing package for tank shell assembly and PMD assembly for both fuel and oxidizer tanks Manufacturing planning for both fuel and oxidizer tanks Compliance documents for both fuel and oxidizer tanks PMD performance analysis report for both The non-recurring expenses to develop these items were high, but the high price approach was justifiable given the immense emphasis on minimizing overall mass. 2.1 Tank Shell Development Designers and engineers used traditional design approaches to design and analyze the tank shells. Analyses performed including stress analysis, modal analysis, fracture mechanics analysis, strength analysis, buckling analysis, PMD structural load analysis, and PMD slosh fatigue analysis. For all parameters, we analyzed to the worst-case conditions. We used commercial software to conduct the structural analyses, and NASGRO for fracture mechanics analysis. The analyses included published data for both residual stresses and fracture crack growth in hydrazine [5] [6]. We applied conventional safety factors based on requirements and legacy design practices. Additionally, we used test data from past programs when applicable, and continuously balance the needs for conservatism and realism to meet aggressive mass targets. At the conclusion of the analysis effort, we found that the safe-life requirements and the minimum detectable flaw sizes from ATK s nondestructive examination capabilities were the determining factors for minimum membrane thickness. From the analyses results, engineers designed the James Webb shells to minimum wall thicknesses to achieve minimal tank mass. Figures 5 and 6 are some outputs from the structural analyses. Table 2a and 2b are summaries of tank shell margins of safety. Table 3 is a summary of predicted tank growth under pressurization. In Tables 4a and 4b, we summarize the analytical modal frequencies of, respectively. Figure 5: Outputs from tank shell analyses Figure 6: Outputs from PMD Structural analysis 5

6 Table 2a: Summary of Margins of Safety, Strength Tank Part Condition Fuel Cylinder Proof Membrane Burst Fuel & Blind Dome Proof Oxidizer Membrane Burst Ported Dome Proof Membrane Burst Blind Dome Proof Near Boss Taper Burst Ported Dome Proof Near Boss Taper Burst Blind Dome Proof Boss Burst Ported Dome Proof Boss Burst Table 2b: Summary of Margins of Safety, Buckling Buckling Fuel Tank, Combined Ultimate with 320 psig min. Launch Pressure Oxidizer Tank, Combined Ultimate, with 278 psig min. Launch Pressure Table 3: Tank Growth From Internal Pressurization Predicted deflection, in Radial, midcylinder Radial, girth weld Axial, boss-to-boss Fuel Tank M.S Oxidizer Tank MEOP Proof MEOP Proof N/A N/A Table 4a: Summary of Fuel Tank Modal Frequencies from Tank Analysis M.S Mode Wet Dry Min. Requirement 2.2 PMD Development Scientists and engineers have been developing surface tension propellant management devices (PMDs) to enable space flight since the 1960s [7] [8]. PMD designers use an assembly of basic PMD elements to design PMDs that meet both functionality and performance goals [9] [10] [11] [12]. As space programs evolved, new PMD design requirements emerged, including the ability to maintain tight center of mass (CoM) control and the durability to survive slosh environment on the launch pad [7]. On the James Webb program, both tight propellant CoM control and ground slosh durability were design drivers in addition to the basic requirement of providing gas-free propellant upon demand. The James Webb PMDs must meet the following key requirements: Supply gas-free propellant to spacecraft thrusters upon demand Enable tight control over propellant CoM throughout the entire mission Have the structural strength to survive launch pad slosh environment Be extremely lightweight After an extensive trade study in support of multiple proposal options, the customer selected the large sponge PMD concept for PMD development. PMD designers have been designing sponge PMDs since the 1960 s [3]. Most of ATK s Geosynchronous (GEO) communications satellite PMD tanks include sponge as one of the featured PMD elements [4] [13] [14]. We also designed and manufactured low earth orbit (LEO) satellite PMDs using the sponge as a primary element in the PMD [15]. Other notable sponge PMDs include proprietary PMDs [16] [17], the Orbital Express PMD [18], the Fermi Gamma-ray Space Telescope PMD [15], and the Solar Dynamics Observatory PMD [19]. Figures 7 through 12 are some examples of sponge PMDs. Axial 71.2 Hz Hz 40 Hz Lateral Hz Hz 40 Hz Torsion Hz Hz 40 Hz Table 4b: Summary of Oxidizer Tank Modal Frequencies from Tank Analysis Figure 7: A GEO propellant tank sponge PMD [14] Mode Wet Dry Min. Requirement Axial 86.1 Hz Hz 40 Hz Lateral Hz Hz 40 Hz Torsion Hz Hz 40 Hz Figure 8: A LEO propellant tank PMD [15], later used on the Fermi Gamma-ray Space Telescope 6

7 Figure 9: A proprietary tank sponge PMD [16] For example, the detail design of the sponge panels included design considerations for panel shapes, number of panels, perforation, and thickness. Figures 14, 15, and 16 are examples of the detailed trades to customize the PMD configuration for minimum mass. Figure 10: A proprietary tank large sponge PMD [17] Figure 14a, Fuel tank sponge panel trades during PMD design Figure 11: The Orbital Express large sponge PMD [18] Figure 14b, Oxidizer tank sponge panel trades during PMD design Figure 12: The SDO PMD [19] In Figure 13, we show the large sponge PMD concepts for the James Webb fuel and oxidizer tanks. Note each PMD is unique. During the development phase of the program, the primary task for the PMD designer was to optimize each PMD design to meet the program objectives, including low mass. Figure 15: Design phase PMD trade study to trade various sponge options in the fuel tank Figure 13: The N 2 H 4 tank and the NTO tank PMD concepts The PMD design effort included optimization analyses to optimize design and minimize mass. Figure 16: Design phase PMD trade study to trade various options in the oxidizer tank 7

8 The PMD functional validation included examination of all phases of the operational sequence, from ground operations to ascent operations and on orbit operations. See Figure 17. Ground operations consist of filling, draining, and ground handling. Ascent operations consist of launch, separation, priming, and mid-course correction (MCC) burns. The on orbit operations include zero g coast, station keeping burns, momentum dumping, slews, contingency operation, and depletion. Operation Sequence Flow Diagram Filling Handling and launch Draining Separation Early zero g coast MCC Ignition MCC Steady Firing Late Coast Slews (Early) Slews (Late) Contingency Station Keeping Depletion Figure 17: James Webb tanks operational sequence 8

9 Additionally, the PMD analysis included 3D simulations of fluid under dynamic conditions. In Figure 18, we show snapshots of the simulation runs for both the. Fluid simulations are important because they enable the system engineers to visualize fluid motions and understand the impact of fluid dynamics in spacecraft operations. During the PMD design, fluid simulations also enabled validation of CoM control capabilities of the PMD as a function of time. Fuel tank, at 69% fill fraction, lateral acceleration Figure 18: Snapshots of slosh simulation runs, Finally, the PMD design must meet tight CoM control requirements. There were extensive analyses to evaluate CoM control capabilities during dynamic conditions and at various fill fractions. Figures 19, 20, and 21 are some outputs from the CoM control analyses. Figure 19: CoM control vs. propellant load during slew, Fuel Tank Oxidizer tank, at 60% fill fraction, lateral acceleration Figure 21: Sponge CoM control analyses One important non-operational objective of the PMD design was survivability in a ground slosh environment. To prevent overdesign of the PMD structure, tank and PMD designers must ensure that the input loads for the PMD structural analysis are sufficiently high to enable survivability but not overly conservative as to result in an unnecessarily heavy PMD. From the PMD analyses, we derived the correct loads acting on the PMD, and used these loads for the PMD structural analysis. See Figure 22. We validated the adequacy of these loads later during the test phases of the James Webb tank development. Figure 20: Propellant CoM location relative to tank center in zero g, Fuel Tank Figure 22: Propellant CoM location relative to tank center in zero g, Fuel Tank 9

10 Upon conclusion of the PMD design analysis, we configured two custom designed PMDs for the James Webb. Each PMD design optimized tank performance, minimized tank mass, and achieved the design objective of tight propellant CoM control. 2.3 Tank Manufacture The James Webb tank manufacture followed our legacy tank fabrication methodologies. We precision machined the shells from solution treated and aged 6Al-4V ring and dome forgings. Solution heat treatment increased the mechanical properties of the forgings, thus enabling thin wall and low mass pressure vessels. For the fuel tank, we machined two domes and a cylinder section. For the oxidizer tank, we machined two domes only. The time-honored tank shell manufacturing process ensured high quality, high reliability, and low mass. We delivered nearly 6,200 space tanks using the same manufacturing process. We assembled the two PMDs from 6Al-4V components made from sheets or bar stock. The sponge vanes are titanium sheets cut to size and perforated to reduce mass. Both PMDs have sixteen vanes assembled to the PMD center post, as shown in Figure 17. The center post on each PMD attaches to both ends of the propellant tanks. By not having cantilevered attachments in the PMD designs, the designers introduced conservatism and ensured survivability in the slosh environment. Installation of PMDs into the propellant hemispheres formed expulsion assemblies. In Figure 23, we show the expulsion assemblies of both the. Assembling the fuel tank required two hemisphere-to-cylinder girth welds to close the fuel tank. We needed only one hemisphere-to-hemisphere girth weld to close the spherical oxidizer tank. Figure 23: The Fuel Tank & Oxidizer Tank Expulsion Assemblies The tank shell closure welds are tungsten inert gas (TIG) welds performed in an inert environment within a weld chamber to prevent contamination to the thin shell. In Figures 24, we show the completed fuel and oxidizer tank assemblies. Figure 24: Completed Fuel and Oxidizer Tank Assemblies Figure 25 is a view of the tank bottom port. Note both the inlet and outlet ports are on the same boss. The PMD design must accommodate this dual port design. However, we already have several PMD designs based on the same two-port requirement, including another large sponge PMD tank [18]. The design challenge was not significant. Figure 25: The Fuel and Oxidizer Tank Bottom Boss The James Webb flight fuel tank underwent acceptance testing prior to delivery. Table 5 is a list of the fuel tank acceptance test sequence. The test sequence included conventional tests to validate that the tank shell and the PMD have met the structural, functional, and operational requirements. Table 5: Acceptance Test Sequence, Fuel Tank Acceptance Test Title Test Sequence 1 Preliminary Examination of Product 2 Pre-Proof Volumetric Capacity Test 3 Proof Pressure Test 3A Tank Dimensions 4 Post-Proof Volumetric Capacity Test 5 In-Tank PMD Bubble Point 6 External Leak Test 7 Acceptance Vibration Testing 8 In-Tank PMD Bubble Point 9 External Leak Test 10 Visual Inspection 11 Radiograph Inspection of Welds 12 Penetrant Inspection 13 Final Examination of Product and Weight 14 Final Clean 10

11 Unlike the fuel tank, the flight oxidizer tank underwent protoflight testing prior to delivery. In Table 6, we list the oxidizer tank protoflight test sequence. The test sequence is identical to the fuel tank acceptance test sequence, except we conducted the protoflight vibration testing at a higher energy level than the fuel tank acceptance vibration test. Table 6: Protoflight Test Sequence for the Oxidizer Tank Protoflight Test Sequence Test Title 1 Preliminary Examination of Product 2 Pre-Proof Volumetric Capacity Test 3 Proof Pressure Test 3A Tank Dimensions 4 Post-Proof Volumetric Capacity 5 In-Tank PMD Bubble Point 6 External Leak Test 7 Protoflight Vibration Test 8 In-Tank PMD Bubble Point 9 External Leak Test 10 Visual Inspection 11 Radiograph Inspection of Welds 12 Penetrant Inspection 13 Final Examination of Product and Weight 14 Final Clean Delivery of James Webb flight fuel tank and oxidizer tanks was successful. The delivered fuel tank mass was 15.8 kg (34.77 lb m ) against a notto-exceed requirement of 16.5 kg (36.4 lb m ). The delivered oxidizer tank mass was 10.3 kg (22.74 lb m ) against a requirement of 11 kg (24.2 lb m ). Our designers optimized the tank designs, and our manufacturing team fabricated the tanks within specification requirements and met the mass targets. 3. Development and Qualification Testing The James Webb tank qualification approach was qualification by both analysis and test. For the fuel tank design, we manufactured a dedicated Qualification Tank and subjected it to the qualification test sequence summarized in Table 7. For the oxidizer tank, we manufactured only one tank, and subjected it to the protoflight test sequence of Table 6. Table 7: Qualification Test Sequence Conducted on the James Webb Fuel Tank Qualification Test Sequence Test Title 1 Cyclic Pressurization 2 Priming Flow Rate 3 Flow Rate and Pressure Drop 4 PMD Integrity Radiographic Inspection 5 Bubble Point 6 External Leakage 7 Qualification Vibration Testing 8 Launch Accelerations 9 PMD Integrity Radiographic Inspection 10 Bubble Point 11 External Leakage 12 Visual Inspection 13 Penetrant Inspection 14 Radiographic Inspection 15 Final Examination of Product 16 Burst Pressure and Rupture The test engineers followed a pre-determined cyclic pressurization schedule to cycle test the Qualification Tank. See Table 8. Upon conclusion of all other qualification tests, the Qualification Tank had undergone 59 MEOP pressure cycles and 21 proof pressure cycles prior to the final burst pressure test. Table 8: Cyclic Pressurization Schedule Cycle # Pressure Temperature Cycle # Pressure (psig) (psig) Temperature ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) ⁰C (100⁰F) 11

12 Tables 9 through 12 are summaries of the qualification level vibration test environments. Table 9: Qualification Dry Sine Sweep Levels Axis Axial Lateral Frequency Qualification (Hz) Level Units mm g g g g g g mm g g g g Table 10: Qualification Dry Random Vibration Profile Axis Axial Lateral Frequency (Hz) Qualification Level Units g 2 /Hz 20 to db/octave 90 to g 2 /Hz 400 to db/octave g 2 /Hz Overall RMS 10.0 g rms g 2 /Hz 20 to db/octave 90 to g 2 /Hz 400 to db/octave g 2 /Hz Overall RMS 12.7 g rms Table 11: Qualification Wet Random Vibration Profile Axis All Frequency (Hz) Qualification Level Units g 2 /Hz db/octave g 2 /Hz db/octave g 2 /Hz db/octave g 2 /Hz Overall RMS 6.22 g rms Table 12: Qualification Wet Sine Vibration Profile Axis Axial Lateral Frequency (Hz) Qualification Level Units mm g g g g g g mm g g g g For vibration testing, test engineers designed and manufactured a dedicated test fixture. See Figure 26. The high input loads necessitated a dual shaker setup for lateral testing as shown in Figure 26. During wet random and sinusoidal vibration tests, the Qualification Tank held kg (395.9 lb m ) of DI water at test pressure of 24.2 bar (351.4 psig). All vibration test runs were successful. Test results indicate the tank axial frequency at 79 Hz, and lateral frequency at 149 Hz. Axial test setup Lateral test setup Figure 26: The vibration test setups 12

13 To conduct acceleration testing, test engineers installed the Qualification Tank in the same cage fixture as vibration testing, and connected the assembly to a centrifuge as shown in Figure 27. During acceleration test, test engineers filled the Qualification Tank with 178 kg (392.4 lb m ) of DI water and pressurized it to 24.2 bar (351 psig) test pressure. Table 13 is a summary of the qualification acceleration loads. The loads exerted during test were 10.2 g Z-axis, 1.41 g X-axis, and g Y-axis, resulting in a combined load of g. Figure 29 is the test setup for the slosh test. Each PMD must undergo test runs summarized in Tables 14a and 14b. Test engineers designed the linear shaker to oscillate at the required frequencies and amplitudes for the required durations. Consistent with safe-life analysis methodology, the test durations were four times the expected life cycle. We tested each tank to the worst case fill fraction, and periodically inspected the PMDs during testing. Both PMDs survived the endurance test. Figure 30 is a photo of the simulator during slosh testing. Figure 27: The acceleration test setup Table 13: Qualification Acceleration Loads Linear Acceleration Lateral J1 (g) Lateral J2 (g) Axial J3 (g) RJ1 (rad/sec 2 ) Angular Acceleration RJ2 (rad/sec 2 ) Torsion RJ3 (rad/sec 2 ) ± 2.5 ± 2.5 ± 8.0 ± 50.0 ± 50.0 ± 75.0 Increasingly, tank and PMD designers must consider prolonged propellant slosh during ground handling. Primary concerns include scenarios during which the propellant tank is at the launch pad atop the launch vehicle or during launch vehicle transport. While it might be possible to analyze fluid behavior, the best way to qualify structural endurance continues to be slosh testing. During the James Webb tank qualification campaign, engineers conducted slosh test runs to validate the structural integrity of both the fuel and the oxidizer PMDs. Test engineers installed each PMD into a custom designed simulator with a Plexiglas dome. The clear upper dome allowed visual inspection and video recording throughout the test campaign. See Figure 28. Figure 29: The slosh simulation test setup Table 14a: The James Webb Pre-launch Slosh Simulation Testing Runs Run Lateral Displacement (in) ±3.66 (7.32 DA) ±3.23 (6.46 DA) ±3.23 (6.46 DA) Exposure Time (hrs) 4.0 Fuel 3.3 Ox 4.0 Fuel 3.0 Ox 2.0 Fuel 1.5 Ox Fuel Tank Oscillation Frequency (Hz) Ox Tank Oscillation Frequency (Hz) Table 14b: The James Webb Pre-launch Slosh Simulation Testing Fluid Loads Tank Fluid Fluid Mass Fuel Water 141 kg (311 lb m) Ox Oxidizer Simulant 109 kg (240 lb m) Fuel Tank simulator Ox Tank Simulator Figure 28: The slosh simulators Figure 30: The James Webb fuel tank PMD during slosh testing 13

14 After all the environmental tests, we conducted external leak check, radiographic inspection, dye penetrant inspection, and PMD bubble point test to validate the integrity of the tank shell and the PMD. Upon completion of all the validation tests, we conducted a final burst rupture test. The temperature during the burst pressure test was 21.6 C (69 F). During pressure ramp up, there was a 10 second hold at 27.6 bar (400 psig), and a 15 second hold at 37.9 bar (549 psig). The James Webb Qualification Tank ruptured at 56 bar (812 psig). With the temperature corrected burst pressure at 37.5 bar (544 psig). The burst data confirmed the analytical positive margin of safety for burst pressure. Figure 31 is a photo of the ruptured tank after burst test. Figure 31: The ruptured Qualification Tank after burst test 4. Conclusion In this summary paper, the authors describe the primary theme and development activities of the James Webb. Every PMD tank development program has a theme. The theme of the James Webb tank development was a long development cycle to achieve mass minimization with requirements of tight propellant CoM control and prolonged ground slosh. These requirements tend to counteract each other, and designers and engineers must work together to derive an optimal solution that meets stringent and unusual requirements. The thoroughness of the development effort extended to tank shell and PMD components. For example, in addition to tank qualification, we also conducted component-level qualification programs such as dome forging qualification and cylinder forging qualification. The two new shells are excellent additions to our portfolio of qualified tank shells. After the tank qualification, the fuel tank shell became the baseline for another PMD tank development program. Using a derivative tank approach avoided the high cost of a new qualification program. We expect the availability of these shells to benefit many future space programs. References 1. Tam, W., Wiley, S., Dommer, K., Mosher, L., & Persons, D. (2002, July). Design and manufacture of the MESSENGER propellant tank assembly. Paper presented at the 38 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Indianapolis, Indiana. doi: / Tam, W., Ballinger, I., & Jaekle, D. E. Jr. (2008, July). Tank trade studies an overview. Paper presented at the 44 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Hartford, Connecticut. doi: / Tam, W., Hersh, M., & Ballinger, I. (2003, July). Hybrid propellant tanks for Spacecraft and Launch Vehicles. Paper presented at the 39 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Huntsville, Alabama. doi: / Tam, W. & Jaekle, D. Jr. (2016, May). The evolution of a family of propellant tanks containing propellant management devices, SP Paper presented at the Space Propulsion 2016 Conference, Rome, Italy. 5. Haupt, C. W. (1996, July). Residual stress measurement of titanium weld certification rings. Paper presented at the 32 nd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Lake Buena Vista, Florida. doi: / Lewis, J. C., & Kenny, J. T. (1976, July). Sustained load crack growth design data for Ti-6AL-4V titanium alloy tanks containing hydrazine. Paper presented at the 12 th Propulsion Conference, Palo Alto, California. doi: / Tam, W, Behruzi, P., & Jaekle, D. (2016). The Evolutionary Forces and the Design and Development of Propellant Management Devices for Space Flight in Europe and the United States. SP2016_ Paper presented at the Space Propulsion 2016 Conference, Rome, Italy. 8. Rollins, J. R., Grove, R. K., & Jaekle, D. E. Jr. (1985). Twenty-three years of surface tension propellant management system design, development, manufacture, test, and operation. Paper presented at the AIAA/SAE/ASME/ASEE 21 st Joint Propulsion Conference, Sunnyvale, California. doi: /

15 9. Jaekle, D. E. Jr. (1991, July). Propellant management device conceptual design and analysis: Vanes. Paper presented at the AIAA/ASME/SAE/ASEE 27 th Joint propulsion Conference, Sacramento, California. doi: / Jaekle, D. E. Jr., (1993, July). Propellant management device conceptual design and analysis: Sponges. Paper presented at the AIAA/SAE/ASME/ASEE 29 th Joint propulsion Conference & Exhibit, Monterey, California. doi: / Jaekle, D. E. Jr. (1995, July). Propellant management device conceptual design and analysis: Traps and troughs. Paper presented at the 31 st AIAA/SAE/ASME/ASEE Joint Propulsion Conference & Exhibit, San Diego, California. doi: / Jaekle, D. E., Jr. (1997, July). Propellant management device conceptual design and analysis: Galleries. Paper presented at the 33 rd AIAA/SAE/ASME/ASEE Joint propulsion Conference & Exhibit, Seattle, Washington. doi: / Debreceni, M. J., Juo, T. K., & Jaekle, D. E. Jr. (2004, July). Development of a composite wrapped propellant tank. Paper presented at the 40 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Fort Lauderdale, Florida. doi: / Tam, W. H. & Jaekle, D. E. Jr. (2005, July). Design and manufacture of an oxidizer tank with a surface tension PMD. Paper presented at the 41 st AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Tucson, Arizona. doi: / Debreceni, M. J., Kuo, T. K., & Jaekle, D. E. Jr. (2003, July). Development of a titanium propellant tank. Paper presented at the 39 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Huntsville, Alabama. doi: / Benton, J., Ballinger, I., Jaekle, D. E. Jr., and Osborn, M. F. (2007, July). Design and manufacture of a propellant tank assembly. Paper presented at the 43 rd AIAA/ASME /SAE/ASEE Joint Propulsion Conference & Exhibit, Cincinnati, Ohio. doi: / Tam, W. & Taylor, J. R. (1997, July). Design and manufacture of a propellant tank assembly. Paper presented at the 33 st AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Seattle, Washington. doi: / Tam, W., Ballinger, I., & Jaekle, D. E. Jr. (2008, July). Surface tension PMD tank for on orbit fluid transfer. Paper presented at the 44 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Hartford, Connecticut. doi: / Tam, W., Ballinger, I., & Jaekle, D. E. Jr. (2008, July). Propellant tank with surface tension PMD for tight center-of-mass propellant control. Paper presented at the 44 th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Hartford, Connecticut. doi: / Acknowledgments Individuals from multiple functions participated in the tank development program and contributed toward its successful completion. The authors wish to acknowledge their significant contribution and thank them for their continued support. We also wish to acknowledge Mr. Richard Bahng for his outstanding analysis effort throughout the James Webb tank development program. 15