AIAA96-2749

Copyright ©1996 by Pressure Systems, Inc. Published by American Institute of Aeronatutics & Astronautics with permission.

Page 1

AIAA96-2749

DESIGN AND DEVELOPMENT OF A COMMUNICATIONS SATELLITE PROPELLANT TANK

Michael J. Debreceni, William D. Lay, & T. K. Kuo Pressure Systems, Inc., Commerce, CA

and Donald E. Jaekle, Jr.,

PMD Technology, Andover, MA

ABSTRACT

In order to provide a lighter weight, less complicated, more easily manufactured, and more capable propellant tank for a high capacity communications satellite, a new design was conceived from two previously successful communications satellite programs. A light weight tank shell coupled with re-oriented pressurant and propellant ports and a redesigned propellant management device (PMD) fulfilled this requirement. The resultant tank is fabricated in an all-welded titanium configuration with a sponge, vane and trap PMD to store and to provide pressurized gas-free propellants to the main satellite thrusters and reaction control system thrusters in a zero g environment. Each spacecraft contains two tanks - one for nitrogen tetroxide (N2O4) oxidizer and the other for monomethylhydrazine (MMH) fuel. The tanks are mounted in tandem in the spacecraft at 32 tabs with nut plates located near the tank equator. The newly analyzed and designed PMD, with no moving parts, a minimum of surface tension elements, and titanium alloy components, minimizes weight and complexity. A stress and fracture mechanics analysis on the tank assembly, using imposed environmental conditions, operational requirements, and spacecraft interface loads, verified the structural adequacy of the PMD and tank assembly. Successful acceptance and delta qualification tests on one flight-type unit empirically validated the analysis and design of the tank assembly. A

number of flight units have been delivered on time. The first launch of this propellant tank design is tentatively scheduled for October 1996.

INTRODUCTION Pressure Systems, Inc. (PSI) was contracted to analyze, design, fabricate, assemble, test, qualify and deliver propellant tanks to support an initiative to build a cost competitive, high capacity commercial communications satellite. These satellites feature C-Band, KA-Band, and KU-Band transponders. The program initially provided for one delta qualification unit and eight flight units. Eight more tanks were delivered on a follow-on contract. Presently, additional tanks are in production at PSI. Each of the satellites includes two propellant tanks -- one for the nitrogen tetroxide (N2O4) oxidizer and one for the monomethylhydrazine (MMH) fuel. They are mechanically mounted in tandem inside the satellite’s central cylinder with 32 equally spaced tabs and welded into the propulsion fluid system. The tanks are of an all-welded titanium construction. They incorporate passive propellant management devices for propellant acquisition and to supply gas-free propellants to the system thrusters throughout the required mission life. This tank assembly reflects the latest in design innovation and state-of-the art technology for improved manufacturabilty, performance, reliability and testing. The PMD was specifically analyzed and designed to meet the defined mission profile and requirements. This tank assembly is compatible

Copyright ©1996 by Pressure Systems, Inc. Published by American Institute of Aeronatutics & Astronautics with permission.

Page 2

with launches on Atlas-Centaur, Ariane, Long March, and Proton launch vehicles. DESIGN, DEVELOPMENT, AND ASSEMBLY The tank and PMD assemblies were also designed for producibility since additional tanks are required to support the initial programs with options for follow-on units.

The propellant tank assembly summary of capabilities is shown in Figure 1.

FIGURE 1: SPACECRAFT PROPELLANT TANK SUMMARY OF CAPABILITIES

Tank Outside Diameter, Pressurized (in.) 49.30 Tank Length, Pressurized (in.) 49.23 Tank Fluid Capacity, Minimum Liters 1000 in3 61,025 Maximum Initial Propellant Load 96% of tank volume 3027 lbs N2O4 1858 lbs MMH Fluids MMH, N2O4, IPA Distilled Water, Ar, He, GN2, LN2 Temperature Range 19-122oF Pressures (Psia at 122°F) MEOP * 250 Collapse 3.0 Proof 312.5 Burst 375 Minimum Expulsion Efficiency, % 99.5 Leakage Zero liquid leakage ≤ 1 x 10-6 scc/sec He at MEOP * Pressurant Leakage Maximum Weight , lb. 70.0 Minimum Design Life, Years 16.4

Stiffness Tank (96%) Propellant Fill Fraction > 50 Hz Natural Frequency PMD Lateral (Immersed in NTO or MMH) > 20 Hz Natural Frequency PMD Axial (Immersed in NTO or MMH) > 100 Hz Natural Frequency

Launch Vehicle Compatibility Atlas, Ariane, Long March, & Proton * Maximum expected operating pressure.

Page 3

Figure 2 outlines the vibration requirements for the propellant tank. FIGURE 2: VIBRATION REQUIREMENTS

Test (Sine)

Axis

Frequency (Hz)

Acceleration (g, 0-Peak)

Limit Acceleration (g)

5 - 13 0.19 in. S.A. * 3.3 Acceptance 13 - 17 3.3 3.3 (2094 lbs. of Lateral 17 - 45 2.31 2.31

water at 250 psig) (X or Y) 45 - 100 1.54 2.31 4 Octaves/Min 5.0 - 22.3 0.15 in. S.A. 7.6

22.3 - 24 7.6 7.6 24 - 26 Ramp Down 7.6 Axial 26 - 50 2.0 3.46 (Z) 50 - 52 Ramp Down 3.46 52 - 100 1.0 2.0

Test

(Sine)

Axis Frequency

(Hz) Acceleration (g, 0-Peak)

Limit Acceleration (g)

5 - 13 0.25 in. S.A. 4.3 Delta 13 - 17 4.3 4.3

Qualification Lateral 17 - 45 3.0 3.0 (2094 lbs. of (X or Y) 45 - 100 2.0 3.0

water at 250 psig) 5 - 22.3 0.20 in. S.A. 10.2 2 Octaves/Min 22.3 - 24 10.2 10.2

24 - 26 Ramp Down 10.2 Axial 26 - 50 2.6 4.5 (Z) 50 - 52 Ramp Down 4.5 52 - 100 1.3 2.6

Test

(Sine)

Axis Frequency

(Hz) Acceleration (g, 0-Peak)

Delta 5.0 - 10.8 0.25 in. S.A. Qualification 10.8 - 20 3.0 (Dry & Un- All 20.0 - 100 1.5 pressurized) (X, Y, Z) l00.0 - 400 1.0

2 Octaves/Min

Frequency

(Hz)

Delta

Qualification

Acceptance (Reference only -- test not required)

- Random (2 minutes/axis) - 2094 lbs of water and pressurized to 250 psig.

90 to 100 +30 dB/OCT +30 dB/OCT 100 to 800 0.054 g2/Hz 0.027 g2/Hz

800 to 2000 -3 dB/OCT -3 dB/OCT

OVERALL grms 8.7 6.2 * Single amplitude.

Page 4

DESIGN, DEVELOPMENT & ASSEMBLY (cont’d) The tank shells are machined from annealed

titanium alloy Ti-6AL-4V hemispherical forgings. During processing the hemispheres are rough machined, solution heat treated and partial aged, skim machined, final aged and then final machined.

The tank is mounted into the spacecraft structure

by circumferential tabs with nut plates located on one of the hemispheres near the girth weld. The mounting tabs are machined from a ring forging that is Electron Beam (EB) welded to one of the hemispheres in the partial machined state. Individual tabs are finish machined from the ring when the hemisphere is completed. The propellant and pressurant ports are 1/2 inch diameter 6AL-4V titanium-to-304L CRES transition tubes with a flow control venturi in the propellant line.

A shell structural analysis and fracture mechanics analysis, utilizing the NASA FLAGRO program, were used to design the tank shell thicknesses and reinforcements while a stress analysis was used to analyze the PMD details, subassemblies, and assemblies including its installation into the tank. An initial fracture mechanics plan with the tank's histogram governed the former analysis while the imposed design and environmental test conditions, as listed in Figures 1, 2, and 4 dictated the structural and stress analysis parameters and boundary conditions.

Figure 3 depicts an outline of the propellant tank assembly. FIGURE 3: OUTLINE OF PROPELLANT TANK ASSEMBLY

The all-titanium propellant management device is a combination radial sponge, lightweight trap, and vane design. This surface tension PMD has been designed to provide gas-free propellant delivery throughout mission including, but not exclusively, apogee and on-orbit maneuvers. The design utilizes a safety factor of approximately two (2) on all required volumes and a safety factor of three (3) on all required porous material bubble points. These design margins coupled with conservative analyses have yielded a PMD design which easily meets the mission requirements and, in addition, provides for some off-design capability. The PMD is for use in a 49.3 inch diameter spherical tank. The same design can be used in the NTO and the MMH propellant tanks. Additional features were incorporated into the

Page 5

design to provide optimal service. First, because the PMD is a passive device with no moving parts, the design is inherently reliable. Second, the design is constructed entirely of titanium. The PMD is lightweight and offers exceptional compatibility, long life, and reliability. Finally, the PMD is designed not only to provide propellant during steady flow conditions but also to ensure gas-free delivery throughout the mission. The PMD is designed to suppress vortexing, to suppress surface dip, and to accommodate fluid transient motion (which could cause premature ingestion of gas). These additional design considerations have led to a reliable and efficient PMD design. The trap is cylindrically shaped with internal fins and a closed, domed top. The trap allows gas-free propellant during off design operation, provides a flow path to the bulk space propellant during all phases of demand, and minimizes propellant residuals. The trap’s bottom platform is machined from a ring forging and incorporates four screened window assemblies. These window assemblies, consisting of 30 x 160 mesh CP titanium screen and a support plate, are EB welded to the bottom platform and are located at 90o intervals around the platform. This platform also supports the sponge panels at the bottom. The sponge panel support plate is machined from a ring forging and provides the supports for the sponge panels and vane panels at the top of the trap. Finally, the trap is closed at the bottom by a trap base plate which is also machined from a ring forging. Sandwiched between this base plate and the bottom platform is a pick-up assembly consisting of a thin titanium sheet with three welded-in perforated sheet disks at 90o intervals. The -X/-Y quadrant is lacking a perforated sheet window to allow for horizontal handling of the tank assembly. The vertical direction must be maintained at 45o to the -X and -Y axes. The perforated sheet windows are positioned so that propellant access during lateral and setting maneuvers is assured throughout mission. The trap has been sized to contain propellant for contingency maneuvers during which the trap inlet windows will not have access to propellant. No such maneuvers are currently planned. In

addition to the trap, a sponge comprised of titanium sheet separated by a tapered gap, has been incorporated into the design. The sponge has been designed to provide liquid to the trap during the repetitive on-orbit maneuvers. These include attitude control, East-West station keeping, North-South stationkeeping, and deorbit. The sponge also will provide propellant during transient fluid reorientation at apogee ignition. The sponge and trap are connected by four titanium screened windows located on the four thrust axes. These windows allow propellant flow into the trap while preventing gas penetration. Four large major vanes, which are laser cut from thin titanium sheet, are spaced 90o apart and follow the inside contour of the tank. They interface into the trap at the bottom of the tank. They re-supply the sponge during the coast between burns. The major vanes and window assemblies are oriented on the major spacecraft axes. Figures 5A and 5B present schematic views of the PMD. The PMD is fabricated from 6AL-4V, 3AL-2.5V, and commercially pure (CP) titanium sheet, tube, and bar stock. All details are resistance welded, tungsten inert gas (TIG), or EB welded into subsequent assemblies. The perforated sheet disks are EB drilled while the sponge panels are chemically etched to provide the hole pattern. After assembly, the PMD is mounted and welded into the outlet hemisphere while the vane-to-tank wall gaps are maintained. The expulsion assembly is bubble point tested and accepted before the tank assembly is closed. For this configuration there is one automatic TIG girth weld to assemble the final tank assembly. The girth weld is radiographic and dye penetrant inspected and the final assembly is stress relieved in a vacuum heat treat furnace.. The mounting tabs are then final machined and the nut plates are installed. The tank is then ready for final inspection, testing, cleaning, and delivery.

PMD ANALYSIS AND DESIGN I. PMD Introduction & Requirements

Page 6

The PMD is a passive, all titanium, surface tension device designed to provide gas-free nitrogen textroxide (NTO) and monomethylhydrazine (MMH) during all mission accelerations with a minimum expulsion efficiency of 99.5% and a safety factor of two.

The PMD design incorporates a small trap, a large radial sponge, and four vanes as illustrated in Figures 5A and 5B. The device consists of a cylindrical trap housing attached to the tank with bolts. The PMD attachment was designed to allow the PMD to be installed in an existing tank shell. One of the existing bolt holes is used as the outlet. The trap housing has a four screen covered entrance windows for propellant acquisition. Above these

As with most PMDs, this PMD is designed specifically for the defined mission. The mission requirements include upright ground operations and launch, followed by three axis stabilized perigee and apogee burns to achieve orbit, and lateral thruster firings of varying duration to maintain orbit. In addition, horizontal handling will be required for Proton launch processing. The mission requirements are summarized in Table 1.

windows is the radial sponge. Also attached to the trap housing are four vanes which extend from the trap to the tank girth. Within the trap is an annular pick up assembly utilizing electron beam drilled perforated sheet windows to allow gas-free propellant acquisition within the trap. This trap, sponge and vane PMD concept easily meets the mission requirements.

II. PMD General Design Description

FIGURE 4: FACTORS OF SAFETY FOR STRUCTURAL ANALYSES

Component Load Condition Factor of Safety

Pressure Vessel MEOP 1) 1.25 on Yield, Peak Stress 2) 1.5 on Ultimate, membrane Stress

Protoflight Test 1.4 on Ultimate, Membrane Stress 1.25 on Yield, Peak Stress

Qualification Test 1.0 on Yield, Peak Stress

Proof Test 1.0 on Yield, Peak Stress

Burst Test 1.0 on Ultimate, Membrane Stress

External Pressure 1.10 on Buckling

PMD, Mounting Tabs, & Other Components

All 1.25 on Ultimate 1.00 on Yield 1.10 on Buckling``

NOTES: 1) Peak Stress is defined as primary (membrane) plus secondary (bending and stress concentrations) stress at any location.

2) Membrane Stress is defined as the Average Stress across the thickness.

Page 7

Page 8

Page 9

TABLE 1

PROPELLANT MANAGEMENT DEVICE PERFORMANCE REQUIREMENTS SUMMARY

Ground Operations and Launch - Unlimited burn duration.

- Ending fill fractions of 15% to 96%. • Tanks are filled, drained, and launched outlet

down . • Apogee failure and orbit augmentation redundant modes are identical to N/S or E/W Stationkeeping. • Tank fill fraction is between 70% and 96%.

• Tanks are upright or horizontal during handling. During horizontal handling, the tank fill fraction will be maintained sufficiently high to cover the PMD and the pressure will be 215 psia minimum.

• Safe mode requires spin about a horizontal axis (between the two PMDs) at up to 0.3 rpm. Propellant required to enter and leave safe mode is 5 in3 maximum.

Orbital Operations • During launch, booster attitude control can

produce up to 0.01g in any direction. No propellant is required.

• During orbital operations, all stationkeeping

maneuvers are in one of four lateral directions - east, west, north, and south. Stationkeeping maneuvers will occur repeatedly and at most once every hour.

Ascent Operations

• Separation from the booster produces a 0.02g

lateral acceleration for a maximum of 10 seconds.

• The tanks are in blow down mode with a minimum pressure of 170 psia.

• North-South (N/S) stationkeeping. • System line priming occurs 100 sec after separation. - requires 131 in3 of each propellant at a

maximum rate of 0.61 in 3/s. -The priming flow rate is 20 in 3/s maximum.

- produces 0.004g maximum laterally. - Priming occurs during coast and while spinning about a horizontal axis through

• East-West (E/W) stationkeeping and Reposition maneuvers. the PMD at up to 0.35 rpm.

• Pulsed thruster operation occurs prior to apogee and throughout mission. - requires 52 in

3 of each propellant at a maximum rate of 1.22 in 3/s.

- 32 ms duration at 1000/hr maximum. - produces 0.008g laterally. - Maximum time averaged flow rate of 0.015

• Pulsed thruster operation occurs throughout mission (See Ascent Operations for description).

in 3/s.

- Maximum time averaged acceleration of 1 x 10-4 g

• Loss of Lock requires operation during a spin about a horizontal axis at up to 1.0 rpm. Recovery requires 10.5 in3 per tank maximum.

• Multiple three axis stabilized apogee engine firings consume the majority of propellant in a typical mission. Firing specifics include:

• The trap shall provide a minimum of 85 in3 for contingency. - Maximum flow rate of 5.2 in 3/s per tank.

- Axial accelerations between 0.01 and • Lastly, an End-of-Life (EOL) De-Orbit is required which 0.05g.

-produces 0.004g laterally,

Page 10

- uses 157 in3 of propellant at a rate of 0.98 in 3/s.

- alternatively uses the apogee engine.

Miscellaneous

• MMH and NTO are the propellants.

• Gas-free propellant delivery is required throughout mission.

• Residuals shall be limited to 0.5% of the tank volume maximum.

• Propellant temperature is between 40 and 120oF.

The trap housing consists of a centrally located cylindrical region and an annular region below the sponge. The trap housing encloses a volume of propellant which is available for contingency maneuvers. The screened trap inlet windows are located along the thrust axes below the sponge and near the cylindrical region of the trap. All of the screen in the PMD lies in a single plane simplifying bubble point testing. The screen is constructed from commercially pure titanium and is supported by a window pane grid. The sponge consists of thin sheet metal panels positioned in proximity to one another and forming a tapered gap between each panel pair. The taper ensures that the sponge is full in zero g and that sponge draining is efficient. The sponge is located over the trap inlet windows and provides a refillable reservoir of propellant available for stationkeeping maneuvers. The four sheet metal vanes are positioned on the east, west, north, and south axes and follow the tank contour from the sponge to the tank girth. The vanes provide a flow path to the sponge from the propellant pool settled by the lateral operational accelerations. The vanes are notched at the sponge to minimize sponge leakage and are designed to refill the sponge during periods of zero g coast. The trap pick up assembly consists of an annular channel positioned outboard within the trap. Three electron beam drilled perforated sheet

windows are positioned in the upper portion of the pick up channel. The windows are 90o apart and are rotated 45o off the thrust axes. This position reduces residuals while allowing horizontal handling with no risk of gas ingestion. All of the porous elements in the design are titanium and prevent gas from penetrating into the trap and into the outlet lines prior to depletion. The minimal area of porous element greatly increases reliability. The entire design uses less than 3.5 square inches of screen and less than 4.0 square inches of perforated sheet. Several key characteristics make the PMD robust, reliable and able to provide optimal service. First, because the PMD is a passive device with no moving parts, the design is inherently reliable. Second, the design is constructed entirely of titanium. Thus, the PMD is lightweight and offers exceptional compatibility, long life, and reliability. Third, the design contains a minimal quantity of screen and perforated sheet; providing increased strength and reliability. Reducing screen area dramatically increases reliability. Finally, the design is implemented to rely minimally on the porous elements within it. During many nominal operations all porous elements are submerged. This detail to design robustness is a key feature of this PMD. The extremely simple and robust PMD provides low cost, low mass, and high reliability. PMD performance will exceed all requirements. Typical PMD safety factors are listed in Table 2. III. PMD Operational Description This section describes the PMD function during each phase of the vehicle’s life. The operation is separated into its three logical phases: Ground Operations, Ascent Operations, and Orbital Operations.

Page 11

The various phases of the mission that the PMD will encounter and how the PMD will affect the propellants are illustrated in Figure 6, The Operational Sequence.

The PMD is designed to be launched in the outlet down position. Similar to upright ground handling, launch is illustrated in the operational sequence with the propellant position identical to ground handling.

Ground Operations TABLE 2

The ground operations can be divided into three parts; filling, handling, and draining. These are important not only from a flight standpoint but also from a testing standpoint. One must be able to fill the tank in a reasonable time when following a standard procedure. Similarly, handling and ground draining must be accomplished without excessive effort. The operational sequence depicts these ground operations.

TYPICAL CALCULATED PMD SAFETY FACTORS

• Trap Usage: Allocation of Volume Mission Stage Volume at EOL (in3) Contingency 85 Gas Trapped During Fill 31

Total Trap Use 116 Filling occurs with the tank upright in the outlet down position. The tank pressure is initially at atmospheric pressure to reduce the amount of gas retained in the trap above the trap inlet windows. The trapped gas will be compressed upon pressurization. The filling process is straightforward and should introduce no difficulties either to the technician or to the PMD.

Total Trap Volume 270 Trap Residual Vol. 35 Available Volume 235 Safety Factor (SF) 2.0 • Perforated Sheet Loads Event SF* (NTO) SF (MMH) Ascent (Worst) 7.3 16

Apogee (Worst) 4.4 4.7 The PMD was designed for handling with the tank in the outlet down or horizontal orientation. Gas will not enter the trap nor the pick up assembly during handling. Both upright and horizontal handling are illustrated in the operational sequence.

Orbital 8.6 9.6 • Loads on Porous Elements During Boost Element SF (NTO) SF (MMH) Trap Inlet Screen 19 41

Perforated Sheet 15 32 Ground draining may have to be accomplished with propellants and certainly will occur with test fluids. The liquid remaining in the tank at the end of ground draining will have to be evaporated from the tank. The tank will be drained in the outlet down position as shown in the operational sequence.

• Loads on Porous Elements During Separation Element SF (NTO) SF (MMH) Trap Inlet Screen 10 21 Perforated Sheet 7 16

• Sponge Volume (Worst Case) - 120oF NTO Ascent Operations Acceleration (g) SF Ascent operations can be divided into five stages: ascent, separation, system priming, coast, and apogee engine firing. In addition, the spacecraft may enter and leave safe mode; a mode where the vehicle slowly spins about a horizontal axis.

.004 2.1 .008 2.3

*Safety Factor

Page 12

Page 13

After the initial ascent, the spacecraft is separated from its booster. Separation produces a lateral acceleration of up to 0.02g for up to 10 seconds. The propellant will begin to settle during the 10 second acceleration but will regain its zero g configuration soon afterward. As early as 100 seconds after separation, the system is primed. High flow rate propellant demand occurs during system priming. System priming may occur while spinning about a horizontal axis at up to 0.35 rpm. In the NTO tank the propellant will be settled by the centripetal acceleration. In the MMH tank, the propellant will move to the top of the tank. The surface tension forces will maintain propellant over the porous elements. With all of the porous elements submerged in propellant, gas ingestion is not possible. System priming will dynamically load the trap housing. The trap housing has been designed to accommodate the system priming transient structural loads. During zero g coast propellant occupies its minimal energy state. At high fill fractions, the gas in the bulk space will be spherical. The gas bubble in the trap will also be spherical. The fins in the trap keep the gas far from the porous elements. Zero g coast is illustrated in the operational sequence. During apogee engine firing, significant quantities of propellant are consumed but the trap inlet windows will remain submerged during the settling maneuver allowing gas-free acquisition. The PMD’s only function during apogee is to provide propellant during the ignition transient (and subsequent slosh) and prevent vortexing which could lead to premature gas ingestion. Steady state apogee is illustrated in the operational sequence. Oribital Operations Once on-orbit, the PMD has been designed to provide gas-free propellant delivery during four types of maneuvers: 1) short duration, low propellant consumption, repetitive maneuvers such as attitude control, 2) intermediate duration, intermediate consumption, repetitive burns such as stationkeeping and reposition maneuvers, 3) low horizontal axis spin rate maneuvers such as

loss of lock and safe mode, and 4) the depletion burn in a lateral or settling direction. Between these maneuvers a zero g coast is encountered and the propellant will occupy a position minimizing its surface area (and thus surface energy). The sponge is filled with propellant and the propellant resides primarily in the outlet hemisphere. The low consumption maneuvers are of sufficiently short duration to ensure that the propellant will not displace far from its equilibrium condition. Sponge propellant is easily consumed without any risk of gas ingestion into the trap. Likewise the trap pick-up assembly will always be in contact with propellant making gas ingestion impossible. The operational sequence shows the zero g propellant position with all porous elements submerged. During a stationkeeping or reposition maneuver, the propellant in the sponge is consumed. The burn is long enough to reorient most of the propellant into a pool on the tank wall where it is inaccessible. The sponge will retain propellant between sponge panels during the lateral acceleration. The sponge has been sized to hold enough propellant to meet each maneuver’s propellant demand. Subsequent to a lateral maneuver, the propellant in the bulk space will reorient into its coast configuration refilling the sponge thus readying it for the next maneuver. At any time, operation during a low rate spin about a horizontal axis is required. The spin produces a centripetal acceleration which settles propellant in the NTO tank and unsettles propellant in the MMH tank. However, the acceleration is very small and will not significantly deform the gas bubble in the trap and will not empty the sponge. The sponge will easily provide the propellant required to enter and leave this mode. The last maneuver is de-orbit where depletion will occur. As the remaining propellant in the sponge is consumed, the area of screen covered with propellant decreases. This decrease in area causes an increase in the flow losses. Once the head differential, caused by the flow losses and the hydrostatic head, increases to the bubble point of the screen, gas will be drawn into the

Page 14

trap. The consumption of the trap propellant will proceed similarly until gas penetrates the perforated sheet. Because there are no porous elements downstream of the perforated sheet, the gas will enter the outlet and the tank will be depleted. The PMD has been designed to provide gas-free propellant to the tank outlet during the required conditions until the fill fraction of the tank falls below 0.5%. IV. PMD Design and Analysis Design The mission requires that the PMD provide gas-free propellant during settling burns and during subsequent non-settling burns. The maneuver duration is extremely long during settling operations and of limited duration during nonsettling operations. PMDs can be classified into two basic categories: control devices and communication devices.1 Control devices are able to deliver a fixed quantity of propellant while communication devices offer unlimited duration operation. Because this mission requires fixed quantity propellant delivery for most maneuvers, a control PMD is feasible. A communication PMD could meet the mission requirements but the PMD chosen for this mission is the most robust, reliable and lightweight design available. The apogee phase is most easily accommodated with a pickup assembly positioned in the propellant pool. This is accomplished by using the trap inlet windows as the pickup assembly. The sponge acts as a reservoir to accommodate transients and as a vortex suppressor. The limited duration of the on-orbit maneuvers, allows the use of control devices which are more reliable, smaller, and simpler than communication devices. Two control devices were incorporated into this PMD: the sponge and the trap. The sponge can provide the propellant required for each of repetitive maneuvers such as stationkeeping. The size of the sponge and the number of sponge panels were determined based upon the stationkeeping volume and acceleration

requirements. The sponge is a refillable control device and the vanes were incorporated to refill the sponge between maneuvers. The trap is not refillable and is designed to accommodate once-in-a-lifetime contingency maneuvers which the sponge cannot accommodate. Examples include a longer than anticipated station change maneuver and/or off-design recovery operations. The trap is sized to provide 85 in3 of propellant for these types of maneuvers as well as to allow the gas ingested into the trap to expand upon blowdown. Inside the trap is a communication device to allow unlimited burn duration for the propellant in the trap. Analysis The principle method of PMD performance verification is analysis coupled with component and tank assembly bubble point and flow loss testing. The analyses examine, in detail, the fluids’ reaction to all phases of the mission. Propellant location, reorientation, and flow characteristics are determined and evaluated. The PMD is analyzed to ensure adequate control and delivery of propellant. The porous elements are shown to demonstrate the required margins. In addition, flow losses and flight depletion residual propellant quantities are analytically determined. Because PMDs have been extensively proven in flight, and drop tower tests have verified the analytical techniques used to design them, no test verification program is required as such testing would yield no new information. Because each maneuver in the mission can directly affect the PMD, each performance analysis addresses a phase of mission. First, the impact of Ground Operations on the PMD is examined. Second, the impact of and operation during Ascent Operations is examined. And finally, the operation of the PMD during all Orbital Operations is analyzed. The specific analyses are listed in Table 3. Due to the summary nature of this paper, no results are presented. The detailed process of vane,

Page 15

sponge and trap design and analysis can be found in the series of papers titled Propellant Management Device Conceptual Design and Analysis: Vanes, Sponges, or Traps and Troughs by D.E, Jaekle, Jr. 2,3,4 The analyses conducted verify that the PMD will meet all requirements of the specification by providing gas-free propellant upon demand.

TABLE 3 PMD PERFORMANCE ANALYSES

I. General Design Analyses A. Trap Sizing B. Pickup Assy Sizing C. Trap Inlet Window Sizing D. Sponge Sizing E. Flow ∆P F. Thermal Effects II. Ground Operations A. Filling B. Draining C. Handling III. Ascent Operations A. Boost B. Separation C. System Priming D. Apogee Engine Firing E. Safe Mode F. 3 Axis Attitude Control IV. Orbital Operations A. Sponge Use B. Sponge Refill C. Trap Use D. Loss of Lock E. De-orbit & Depletion

QUALIFICATION TESTS

One test specimen was subjected to a series of “delta” qualification tests, so named because the tests were specifically devised to qualify the redesigned PMD only. The tank shell, which is identical to the one used for a previous propellant tank except for port locations, had previously been qualified. The delta qualification unit included a flight-type PMD mounted into reworked tank shells. These shells

were from a previously tested qualification tank which had been burst tested but not taken to destruct. The shell had the center section removed, had the internal test system and original PMD removed, and had the outlet (propellant) port modified to accept the new PMD. This test tank was subjected to the tests listed in Figure 7.

FIGURE 7: DELTA QUALIFICATION TESTS FOR THE

PROPELLANT TANK Preliminary Inspection of Product Mass Measurement Pre-Proof Volumetric Capacity Ambient Proof Pressure Test Visual Inspection Post-Proof Volumetric Capacity Bubble Point Test Sine & Random Vibration Tests Visual Inspection Bubble Point Test Visual Inspection Expulsion & Pressure Drop Test Bubble Point Test Visual Inspection Data Review The qualification test starts with an inspection for dimensions, verification of previous inspections and tests, identification, successful completion of all manufacturing planning, damage, material certifications, and completed rework instructions.

The dry tank assembly is weighed on a precision scale. The actual weight was 71.5 lbs. versus a maximum allowable of 70.0 lbs. The actual weight was measured for reference only since the test tank shell is a modified shell, which is thicker, and it contains additional closed off test and outlet ports. The shell was not optimized for this test. The tank is subjected to an ambient hydrostatic proof pressure of 200 psig for 60 seconds and the mass of water required to fill the tank is measured to determine the actual tank volume before and after the application of the proof pressure. Again, because this was a previously used shell the actual flight unit proof pressure of 322.5 psig was not used.

Page 16

The above test series merely proved that the reworked shell was structurally adequate for the ensuing vibration tests. Earlier the reworked shells had been dimensionally and dye penetrant inspected to “re-zero” them after the burst test. No discrepancies were noted. The pre-proof volumetric capacity was 60,837.2 in3 while the post-proof volumetric capacity was 60,843.3 in3. This results in a volume increase of 11.1 in3 or a difference of 0.02% which is typical. It is well within the maximum allowable of 0.2%. A visual inspection revealed no damage or deformation. The tank was drained and dried. A bubble point test with Isopropyl Alcohol (IPA) as the test medium was accomplished on the PMD trap screens. Bubble points are verified for 60 seconds minimum to verify PMD screen integrity then the capillary breakdown value is measured. The tank bubble point was successfully tested to a level of 2.0 in. of head and the breakdown head was measured to be 2.5 in. Vibration tests are performed on the tank in two orthogonal axes (Y and Z). The tank and PMD are symmetrical in the lateral axes (X and Y); therefore, only one lateral axis has to be tested. The test fixture simulates the tank-to-spacecraft installation interface but is also sufficiently stiff to be considered a rigid mass and not introduce spurious inputs to or responses from the tank assembly. Figure 2 presents the vibration test parameters. The tank was tested first in the axial (Z) axis and than in the lateral (Y) axis. The water load simulates the worst case oxidizer load condition. The tank was pressurized to 100 psig for this series of tests since it is a reworked shell. The dry sine test was performed first followed by the wet sine and wet random tests. Up to 13 accelerometers are installed on the tank to measure the response levels and to characterize the response spectrum. Up to four control accelerometers are mounted on the test fixture to provide a closed loop control system with the vibration driver equipment to insure that the specification requirements are met. The wet sine vibration spectrum input is notched to limit the acceleration loads on the tank. Natural

frequency results are presented in Figure 8. They are consistent and meet the specification limit of >50 Hz for all axes.

FIGURE 8: TANK NATURAL FREQUENCIES, 96% FILL

FRACTION (HZ) Axis Calculated Measured

Z 80 80 - 90 Y 70 70 - 80 The tank was visually inspected during and after the vibration tests and no damage or deformation was noted. Bubble point tests were performed after the Z axis tests and at the completion of the vibration tests to verify that the screens had not been damaged by the vibration tests. A bubble point of 2.0 in. of IPA was held for 60 seconds in each instance, and the IPA test heads at breakdown were 2.4 in. and 2.5 in., respectively - all consistent values. An expulsion and pressure drop test was performed. The tank was initially filled with 2197.4 lbs. of water and pressurized to 90 psig. The water was initially discharged into an evacuated surge tank. This simulates the surge start operation and flow out of the propellant tank during system priming. The tank is drained in the vertical attitude using a 90 psig ullage pressure and a flow rate of 1.35 gpm. Pressure drop and residual water volume are measured. The tank demonstrated residuals well within the mission requirement of 305 in3. The resultant expulsion efficiency exceeded the minimum required of 99.5% as far as ground residuals are concerned. During the final bubble point test using IPA, the PMD trap screens held a test head of 2.0 in. for 60 seconds. The test head at capillary breakdown was measured at 2.5 in. of IPA. A final visual inspection revealed no damage, deformation, or leakage. Based upon the above tests and inspections, the propellant tank assembly was qualified for flight use.

Page 17

ACCEPTANCE TESTS The acceptance tests for the propellant tanks are listed in Figure 9.

FIGURE 9: ACCEPTANCE TESTS FOR THE PROPELLANT TANK

Preliminary Inspection of Product Mass Measurement Pre-Proof Volumetric Capacity Ambient Proof Pressure Test Visual Inspection Post-Proof Volumetric Capacity Cryogenic Proof Pressure Test Visual Inspection Post-Cryo Proof Volumetric Capacity External Leakage Test Bubble Point Test Sine Vibration Test Visual Inspection Pressure Drop Test Bubble Point Test External Leakage Test Radiographic Inspection Dye Penetrant Inspection Visual Inspection Cleanliness Check Data Review The acceptance test starts with an inspection for dimensions, verification of previous inspections and tests, identification, successful completion of all manufacturing planning, damage, and material certifications. The dry tank assembly is weighed on a precision scale. The tank is subjected to an ambient hydrostatic proof pressure of 322.5 psig for a maximum of 66 seconds. The mass of water required to fill the tank is measured to determine the actual tank volume before and after the application of the proof pressure. Permanent set must not exceed 0.2%. The drained and dried tank is then exposed to a cryogenic proof pressure test at 440 psig for five seconds maximum. The tank is filled with liquid nitrogen and totally submerged in a stainless steel lined pit filled with liquid nitrogen. The tank is vented, drained, and allowed to reach ambient conditions without heating. Another volumetric capacity test is performed with water to insure that the test specimen has not experienced a permanent set

greater than 0.2%. An external leakage test is run at 240 psig with helium gas for 15 minutes minimum in a vacuum chamber evacuated to 0.2 microns of mercury or less. Leakage must not exceed 1 x 10-6 scc/sec of helium gas. To verify the integrity of the PMD screened elements after the proof pressure tests, a bubble point test with IPA as the test fluid is performed. The bubble point is held for a minimum of 60 seconds and then the capillary breakdown is measured. Vibration tests, configured as described in Figure 2, are performed on the tanks in two orthogonal axes (Y and Z). Sufficient response accelerometers are installed on the test specimen to characterize the tank responses and control accelerometers are placed on the test fixture to insure compliance with the specification requirements. The sine vibration spectrum input is notched to limit the acceleration loads on the tank. Next, a pressure drop test, using distilled, deionized water, is completed. Water is discharged at 1.35 gpm (worst case mission flow rate) with a 220-240 psig ullage pressure while the tank pressure drop is measured and compared to the specification requirements. The bubble point test, external leakage test, fracture critical radiographic inspection of the girth weld and internal PMD views, fracture critical dye penetrant inspection of the external surface, and visual inspection are performed to verify the tank has no damage as a result of the environmental tests. Finally, the units are cleaned to Figure 10 levels, purged to a -65oF dew point, pressurized to 15 psig with dry GN2, the ports capped, and packaged in two sealed plastic bags for shipment. Figure 11 depicts an outline of the propellant tank showing the port orientations. The top view illustrates the propellant port while the bottom view shows the pressurant and propellant outlet ports.

Page 18

Page 19

Page 20

Page 21

Figures 12 through 15 show photographs of the PMD, the propellant tank, a unit in the lateral vibration fixture, and a tank installed in the proof pressure fixture, respectively.

FIGURE 10: TANK CLEANLINESS

LEVEL

Particle Size (Microns)

Maximum Amount of Particles

> 100 Microns None 76-100 Microns 8 51-75 Microns 20 10-50 Microns 500 < 10 Microns No Silting

NOTES: 1. No metallics >50 microns 2. Non-volative residue (NVR) less than 1mg/100 ml.

CONCLUSIONS

This propellant tank has met all design objectives that can be verified by analysis and the delta qualification test specimen has successfully passed all tests. Sixteen propellant tanks have been delivered on time. Additional tanks are in work at Pressure Systems, Inc. (PSI). The delta qualification unit is in storage at PSI. The first satellite with this propellant tank is scheduled for launch in October 1996 on a Proton booster from the Baikonur Cosmodrome, Kazakhstan. The second is scheduled for an early 1997 launch on an Atlas-Centaur booster from CX36, Cape Canaveral, Florida.

ACKNOWLEDGMENTS The authors wish to thank the following individuals for their assistance on the successful completion of the development and qualification

of the propellant tank assembly and delivery of the first production units: Don Bond, Ron McClellan, Wayne Rogers, Carlos Solis, Tso-Ping Yeh of Loral Engineering and John Radkowski and Tom Thrailkill of Loral Subcontract Administration, who provided much support and assistance. In addition, thanks are expressed to Harold Braund, Ken Foster, and Larry Hicks of Foster Engineering, Agoura, CA who performed the Stress & Fracture Mechanics Analysis and supported the customer design review. Finally, thanks are expressed to all of PSI’s subcontractors who provided materials, detail parts, manhours, and services to make the program a success.

REFERENCES

1 Rollins, J. R., Grove, R. K., and Jaekle, D. E., Jr., “Twenty-Three Years of Surface Tension Propellant Management System Design, Development, Manufacture, Test, and Operation”, AIAA-85-1199, 1985.

2 Jaekle, D. E., Jr., “Propellant Management Device Conceptual Design and Analysis: Vanes”, AIAA-91-2172, 1991.

3 Jaekle, D. E., Jr., “Propellant Management Device Conceptual Design and Analysis: Sponges”, AIAA-93-1970, 1993.

4 Jaekle, D. E., Jr., “Propellant Management Device Conceptual Design and Analysis: Traps and Troughs”, AIAA-95-2531, 1995.

ABOUT THE AUTHORS

Michael J. Debreceni is a Senior Program Manager at Pressure Systems, Inc. (PSI), Jerry Kuo is a Senior Engineer at PSI, and Bill Lay is the Engineering Manager at PSI. Donald E. Jaekle, Jr. is the owner of PMD Technology and PSI’s PMD Analyst & Consultant.

Page 22

NOTES