Apollo Experience Report Ascent Propulsion System

8/8/2019 Apollo Experience Report Ascent Propulsion System

1/29

N A S A T ECHN I CAL NO T E N A S A TN D-7082Y

IN000h

APOLLO EXPERIENCE REPORT -ASCENT PROPULSION SYSTEM

Muntzed Spacecrdft CenterHoz4st012, Texus 77058

by Clurence E . Humphries und Reuben E . Taylor

N A T I O N A L A E R O N A U T I C S A N D S PA CE A D M I N I S T R A T I O N W A S H I N G T O N , D. C. M A R C H 1973


2/29

3. Recipient's Catalog No.I. Report No . 2. Government Accession No.1 NASA TN D-70827. Author(s)

Cla ren ce E. Humphries and Reuben E. Taylor, MSC

4. Title and SubtitleAPOLLO EXPERIENCE REPORTASCENT PROPULSION SYSTEM

8. Performing Organization R eport No.MSC S-341

10. Work Unit No.

5. Report DateMarch 1973

17. Key Words (Suggested by Aut ho r(s))'Apollo* Lunar Module' Lunar Module System Tests* PropulsionPropul sion Development

18. Distribution Statement

9. Performing Organization Name and AddressManned Spacecraft CenterHouston, Texas 77058

19. Security Classif. (o f this report) 20 . Security Classif. (o f this page) 21 . NO . of PagesNone None 29

11. Contract or Grant No.

22. Price'$3.00

1 13. Type of Report and Period Covered12. Sponsoring Agency Name and AddressNational Aeronautics and Space AdministrationWashington, D. C. 20546

I Technical NoteI 14. Sponsoring Agency Code15. Supplementary Notes

The MSC Di rector waived the use of the International System of Uni ts (SI) or this ApolloExperi ence Report, because, in his judgment, the use of SI Units would impair theusefulness of the report o r res ult in excessive cost.The development of the Apollo lu nar module a sce nt propulsion subsys tem is documented.ascent propulsion subsystem was designed, built, and tes ted to provide propulsive power fo rlaunching the ascen t stag e of the lu nar module fro m the surfa ce of the moon into lunar orbit forrendezvous with the orbiting command and ser vic e module. Because total redundancy was pro-hibitive in this sub system, specia l emphasis had to be placed on reliabili ty in design with ade-quate demonstra tion short of an extens ive flight-demonstration progr am.technical problems encountered during all phases of the program ar e identified and the correc-tive actions a re discussed. Based on this experience, sev era l recommendations are made forany future program with si mi la r subsystem requirements.

16. AbstractThe

The significant

* Fo r s a l e by t he Na t iona l Tec hn ic a l I n fo rmat ion Serv ice , Sp ring f ie ld , V i r g in ia 22151


3/29

CONTENTS

Section PageSUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2SYSTE,M DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3DEVELOPMENT AND QUALIFICATION . . . . . . . . . . . . . . . . . . . . . . 5

Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Ascent Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

1 6SYSTEM CHECKOUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20FLIGHT PERFORMANCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Apollo 5 Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Apollo 9 Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Test A rticles and T est Rigs . . . . . . . . . . . . . . . . . . . . . . . . . . .

Apollo 1 0 Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Apollo 11 Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Definition of Realistic Requirements . . . . . . . . . . . . . . . . . . . . . . 24Flexible Program Flow Plan . . . . . . . . . . . . . . . . . . . . . . . . . . 24Leakage and Contamination Control . . . . . . . . . . . . . . . . . . . . . . 24Inspection and Manufacturing Repeatability . . . . . . . . . . . . . . . . . . 25

iii


4/29

TABLES

PageableI

I1MAJOR TEST ARTICLES AND RIGS . . . . . . . . . . . . . . . . . . . 16THE PA - 1 WHITE SANDS TEST FACILITY TEST

SERIES SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . 13FIGURES

Figure Page1 Ascent propulsion sys tem with an ambient- helium-pres surizat ionsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Ascent-engine assembly . . . . . . . . . . . . . . . . . . . . . . . . . 43 Ascent-engine injector and valves . . . . . . . . . . . . . . . . . . . . 44 Major phases of the APS development prog ra m . . . . . . . . . . . . . 55 Squibvalve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Ascent propulsion system propellant-tank to ru s ring . . . . . . . . . . 1 07 Types of injector grid patte rns . . . . . . . . . . . . . . . . . . . . . . 118 Stability devices on the production inject or . . . . . . . . . . . . . . . 1 29 Final-design injector . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1 0 Propellant-inlet section of the injector manifold . . . . . . . . . . . . 1 511 Injector filte r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 512 Modified engine fuel duct . . . . . . . . . . . . . . . . . . . . . . . . . 1513 Tes t r ig PA-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

iv


5/29

APOLLO EXPE Rl ENCE REPORTASCENT P ROPU LS ION SYSTEM

By C l a r e n c e E. H u m p h r i e s a n d R e u be n E. T a y l o rM a n n e d S p a c e c ra f t C e n t e r

S U M M A R YIn June 1962, the Manned Spaceflight Management Council selected the lunar-orbit- rendezvous mode for the Apollo Program. From thi s concept the requirementevolved for a separate lunar module ascent stage propulsion subsystem to return thetwo crewmembers to a lunar-orbit rendezvous with the command and service modules.A propulsion system in which hypergolic bipropellants and a pressure-fed ablativelycooled engine wer e to be used was developed. The system was to provide 3500 poundsof thrust.During the development program, emphasis w a s placed on the use of provenmanufacturing techniques , design integrity, and ground-based testing. This approachresulted in the discovery and correcti on of all significant problems in the engine and inthe pressurization and feed system before using the equipment on manned flights.The following principal problems were encountered and resolved. Engine com-bustion instability was resolved ultimately by developing a new injector under contractto a second manufac turer. Repeated failures and discrepanc ies on the originalpressurization- system regulator were resolved by modifying and adapting the lunarmodule reaction control system regulator to the ascent propulsion subsystem. Br az eand weld- joint design prob lems on severa l pressurization and feedline componentswere resolved. Vehicle syst em leakage lead to redesign of feed system mechanicaljoints to provide redundant sealing capability.Major vehicle-integrated propulsion system testing and certification were accom-plished on the prototype 'PA-1 test rig at the NASA White Sands Test Facility usingflight-type engine and component hardware. Early flight-vehicle checkouts, accom-

plished at Bethpage, N. Y., and at the NASA John F. Kennedy Space Center, involvedprob lems tha t necessit ated frequent component replacements because of failures on thespacecraft caused by component failures during qualification. The lunar module ascentpropulsion subsystem flight sy st ems performed within nominal predicted ranges andonly on one series, lu na r module-3, was a system problem indicated. A fa il ur e of theasce nt propulsion subsystem helium pri ma ry regulator was indicated by a shift i n oper-ating press ures to a secondary regulator and was attributed to contamination duringreplacement of a component in checkout.


6/29

A review of the Apollo asce nt propulsion subsystem experience is indicative thatfut ure system development can be augmented by earl y definition of reali stic require-ments (including leakage and contamination control), a more flexible program-flow planand improvement in design and certification techniques to en sur e adequate braze- andweld- joint inspection and pa ram et ric control,c

I N T R O D U C T I O NThe ascent propulsion subsys tem (APS) had to opera te satisfactorily to ret ur n thecrewmen safely fro m the lunar surfac e. It wa s imperat ive that the system be as simplas possible to in crea se its reliability. Therefore, the prim ary considerations in thedesign and development of the APS were simplicity and reliability, whereas the sec-ondary considerations were performance and weight.The propellants wer e se lected on the bas is of experience with other progra ms,storage requirements, hypergolicity, performance, and density requirement s for pack-

aging. Pre ssure-f ed engines were selected to decr ease complexity by eliminating theneed for pumps, associat ed moving pa rts , and control s. High-pressure gaseous helium(stored at ambient temperature ) w a s selected for t h e pres suri zing fluid. Ablativelycooled thrust chamber s (as compared with regeneratively cooled chamber s) were usedto decrease t h e possibi lity of the propellant freezin g. Wider operational lim its of mix-tu re ratio, propellant tempera ture, and chamber pre ss ur e als o were possible withablatively cooled chambe rs than with regeneratively cooled chambers. A single-engineconcept was chosen to simplify the vehicle control system requirements and therebyincrease total syst em reliability. Subsystem reliabi lity had to be achieved with asingle system. Earl y studies showed that the majori ty of propulsion-system fa ilu reswer e caused by fail ures of contro ls, valves, and solenoids, and were not the re sul t ofinjector o r thrust -chamber failures. Therefore, redundancy of these componentsw a s used, where practical, to incr ease reliability in these failure -sensitive categories

REQU I REMENTSTo meet guidance requirements, the APS was req uired to produce 3500 pounds ofthrust, to fire for a total duration of 550 seconds, to develop 90 percent of the rat edthr ust within 0.450 second after the st a rt signal, and to decay to 1 0 percent of the ratedthr ust within 0. 500 second after the cutoff signal. The maximum allowable combustionchamber pressure during start transients was 177 percent of the nominal combustion-chamber pressure. This w a s a vehicle stru ctur al limitation. To minimize the heattran sfer to the chamber, the magnitude of any period ic o r uniformly cyclic chamber-pres sure fluctuation o r oscillation that oc curr ed at a frequency of 400 hertz or lesscould not exceed k3 psi, and those variations that occurred at a frequency gre ate r than400 hertz could not exceed 26 psi. In addition, the engine wa s required to be stabledynamically in the presence of all self-induced o r artifica lly induced disturbances,thereby causing fluctuations of 175 perc ent of nominal chamber pr es su re in the combustion process. Recovery time to a stable steady-state operation could not exceed0.020 second.

2


7/29

The required oxidizer-to-fuel mixture ratio w a s 1.6 to 1, with a 170 psia requiredpressure at the in terface between t h e propellant-feed section and the engine. Therequired propellant bulk temperature was 70" k 20" F, and the fuel and oxidizer tem-peratures were to be within 10" F of each other.

To R C SA

The liquid propellants fo r the APS were nitrogen tetroxide (oxidizer) and an equalmixture of hydrazine and unsymmetrical dimethylhydrazine (fuel). A total of5213 pounds of propellant w a s required, of which 196 pounds w a s residual. -

T o R C S

SYSTEM D E S C R I P T I O NA schematic of the APS is shown in figure 1. The propellants a re pre ssurizedby gaseous helium at ambient temperature, supplied fr om two identical tanks and routedthrough redundant-flow li nes into the pr es su re regula tors. The helium is stored at anominal pr es su re of 3150 psia and a nominal temperature of 70" F.

I ,Bvoass

O x i d i z e ru w-n g i n e

@ T e m p e r a t u r e t r a n s d u c e r I] xp los i ve va l vea o l e n o i d v a l v e Qxl P r e s s u r e - r e l i e f v a l v eB O u a d h e c k va lve @ B u r s t d i s k C o u p l in g d i s c o nn e c t

@ F i l t e r 6 es t p o i n t 0 r i f i c eP r e s s u r e r e g u l a t o r 0 B r a z e d c a p @ P r e s s u r e t r a n s d u c e r

Figure 1.- Ascent propulsion systemwith an ambient- helium-pressurization system.

Before initial ascent-engine opera-tion, explosive valves are used to isolatethe helium-storage tanks. A filter in eachhelium-flow path tr ap s debri s from theexplosive valves. Downstream from thefilter, each helium-flow path has a normal-ly open latching solenoid valve and twoseries-connected pressure regulators.The primary and secondary helium-flow paths merge downstream from theregulators to for m a common helium mani-fold. The manifold rou tes the helium intotwo flow paths: one path leads to the oxi-dizer tanks and the other path leads to thefuel tank. A quadruple check-valve assem-bly isol ates the upstream components fro mthe corrosive propellant vapors and pre-vents possible hypergolic action in thecommon manifold that may result from mix-ing of propellants o r propellant vapors as aresu lt of backflow from the propellant tanks.Two paral lel explosive valves, down-str eam from each quadruple check-valveassembly, provide a positive seal upstreamfrom the propellant tanks to isolat e the fueland oxidizer (liquid and vapor) before theinitial ascent-engine start. This configura-tion reduces compatibility problems involv-ing helium components and prolongs the lifeof the pres suri zat ion- sys tem components.

3


8/29

Immediately ups tre am f rom the fuel and oxidiz er tanks, each helium-flow pathcontains a burst -disk relief-valve assembly. The relief valve on each helium-flowpath can pas s the e nti re helium flow fro m a failed-open pa ir of r egulators withoutdamaging the propellant tanks. A low-level sensor in each propellant tank causes acabin caution light to turn on when the remaining propellant in ei the r tank is limited toapproximately 10 seconds of engine operation... Each propellant flows through a tr im orifice to the propellant filter in th e engineassembly, then to the isolation and bipropellant valve a ssembl ies (propellant- shutoffvalves). The t ri m orifice provides an engine-interface p re ss ure of 170 psia for properpropellant utilization by the engine. A secondary distribution line is interconnected tothe reaction control system (RCS). A serie s-par alle l arran gemen t of RCS/APS in ter-connect solenoid valves (part of the RCS) permit s the RCS to burn APS propellants(providing the APS is pre ssu riz ed and the propellant is settled during t h e t ime theinterconnect valves are opened).

.

The asce nt engine, shown in figure 2, is a fixed-injector, res tar tab le, bipropellantrocke t engine tha t has an ablatively cooled combustion chamber, throa t, and nozzleextension. Propel lant flow to the ascent-engine combustion chamber is controlled bya valve-package assembly, trim orifices, and an injector assembly (fig. 3). The valve-package assembly is equipped with dual passages f o r both the fuel and the oxidizer andhas two seri es-connected ball valves in each flow path.

P r e s s u r e

O x i d i z e r - s h a f t -

F u e l - s h a f t - s e a l v e n t

F u e l - a c t u a t o r v e n t

Figure 2. - Ascent-engine assembly.

A c t u a t o r -p r e s s u r eline1uel

s h u t o f fvalve-P r o p e 1 a n t -v a l v ea c t u a t o r-

F r o m a c t u a t o r -i s o l a t i o n v a l v esFuel i n O x i d i z e r in

' F i l t e r

- O x i d i z e r -s h u t o f fva Ive

-T r im o r i f i c e

C h a m b e ra b l a t i v em a t e r i a l

N o z z l eab1 at vemater ia l -

I n s u l a t o r

- i b e r g l a s sf i l a m e n tw i n d l n g

Figure 3. - Ascent-engine injector andvalves.4


9/29

DEVELOPMENT AND QUAL1 F I CAT1ONDevelopment of the A P S through an extensive light-test pr og ram was not feasibleeconomically. Therefore, emphasis w a s placed on using proven manufacturing tech-niques, design integrity, and ground-based testing. The pr ogra m was designed to

achieve a fully qualified system in time to support the first manned lunar module (LM)flight. The plan w a s to te st and evaluate materials, components, and assemblies inprogress ively integrated configurations, using various te st r ig s and prototype str uc -tu ra l simulators. Major phases of the development program to accomplish this planare shown in figure 4.component level, progressed through the modular level, and, finally, moved intofull-scale system tests.The development and qualification of the A P S star ted at the

PS breadboard-

De l iver ies

Figure 4. - Major phases of the APS development program.

5


10/29

In addition to the mainst ream hardware-test programs , other tests were designedto identify and reso lve potential problems and to acqui re test data to support the flights.Te st s to determine the effects of a vacuum on fuel venting f ro m the actua tor duringengine-valve shutdown, the ef fects of a vacuum on the engine-valve fuel actuator whena la rg e leakage occurred, and the restart limitations on the ascent engine in space wereamong the other tests.

ComponentsHelium tanks. - The contract that was awarded for the helium-tank pr es su re ves-s e l s w a s delayed because of a delay in the final propulsion- system-concept definition.Spacecraft-weight problems necessitated detailed trade-off studies on the possible us eof a supercritical-helium-storage system.supercritical-helium-storage system, the ambient- helium-storage syst em was retainedeven though there w a s a potential weight penalty.

To avoid the complexity of the

Ten ascent helium tanks that were approximately 0.1 inch undersize ac ro ss themounting base we re fabricated. This fabrication er r o r was caused by the failure toconsider weld shrinkage when the assembly drawings we re made. The affected tankswe re used in the design-feasibility te st program, the design-verification test (DVT)program, and some of the ground te st s. Qualification of the ascent helium tank wascompleted successfully; no significant problem was noted.

Helium solenoid latching valve. - The original subcontractor had difficulty meet-ing the specified differential- p ressu re and internal- leakage requirements. The sub-contractor requested and was granted a release fr om the contract; therefore, a secondso ur ce was selected fo r the development and qualification of the flight solenoid valve.Originally, the solenoid valves we re designed without a requirement fo r fuel andoxidizer compatibility; and nylon sea ts , in combination with Butyl O-rings, were used

to seal high-pressu re cavities within the valve. When it was determined that propellantvapor would migrat e because of helium pressuri zation, the requirement fo r compati-bility was imposed. The subcontractor was directed to ca rr y on parallel effort's thatconsisted of test ing the existing valve design with propellant vap ors fo r 3. 5 days todetermine whether the solenoid valves (with compatibility squib valves in the systemdownstream f rom the check valves) were acceptable and developing a valve to meetrequirements for a 44-day period of exposure to propellant vapors. A t that time, useof nylon and Butyl O-rings was considered adequate fo r a 3.5-day exposure to propel-lan t vapors. However, a design-f easibilit y investigation (DFI) test program wasinitiated to determine nylon and Butyl compatibility with propellant vapors for 44 days.Before completion of the compatibility testing, the subcontractor was directed to stopall work related to the 44-day propellant-compatibility redesign f o r the solenoid valvesbecause the decision had been made to r et ai n the compatibility squib valves locateddownstream of the check valves.The DFI testing indicated that backup rings and lubricants w ere nec ess ary to

. prevent severe chafing of the Butyl O-rings during cycling of the solenoid valves. Thecompatibility tests showed that nylon disintegra tes when it co me s in contact with oxi-dize r vapor, proving the nylon seat was inadequate; therefore, the seat material was

6


11/29

changed to the present mate rial (Teflon). Other design changes brought about duringtesting were the replacement of Teflon backup rings with Teflon "cap seals " and theincorporation of electromagnetic interference (EMI) suppressors .Cold-temperature testing during the DVT program showed the need to reduceO-ring friction to permit reasonable closing times. The use of a Teflon cap seal and acompatible lubricant (PR 240AC) solved the problem.When the solenoid valves were deenergized, back electromotive force in the sole-Zener diodesoid co ils caused voltage spik es that damaged other electrical equipment.wer e added between coil leads to redir ec t this energy until it dissipated. The qualifi-cation program was completed successfully; however, the change to the dual-diodeEMI- suppression package required a supplemental qualification program.The solenoid latch valve installed in te st rig PD-2 (a descent propulsion test rigat the NASA White Sands Test Facility (WSTF)) demonstrated a problem in its ability torema in in the open-latch position after open-close commands in rapid sequence. Theproblem w a s one of residual magnetism in t h e latch plunger, causing it to exceed the

sprin g-ret urn for ce when the valve was cycled rapidly. The design changes to elim-inate this problem were a higher force latch spring, a modified lat ch plunger (to housethe new spring), and shims to maintain the required latch-spring for ce on assembly ofeach valve. Two valves wer e successfully subjected to a requalification test series.The solenoid valve on LM-3 leaked externally through the brazed joint at thevalve body and an inlet tube. Upon review of the brazed-joint design, it w a s found t h a tthe difference in the thermal expansion of the valve body and the tube eliminated theclea rance required fo r the nickel-brazed material to flow properly. The decision w a smade to redesign the solenoid valve for LM-5 and subsequent vehicles by using a brazedjoint with the proper clearances and gold-nickel alloy as the braze material. A t es tpro gram was initiated, with the original nickel-brazed configuration, to cert ify thesolenoid valves fo r LM-3 and LM-4. These tests were completed successfully.The ext ernal leakage reported at the prime-contractor plant for the LM-3 sole-noid valve la te r increased when the valve was checked at the NASA John F. KennedySpace Center (KSC). The inc rea se was partly attributed to the difference in measuringtechniques; however, the decision at that time was to replace the leaking solenoidvalve with the new-configuration solenoid valve at the KSC. Subsequently, the LM-4valves were changed to the gold-nickel-brazed configuration. A delta-qualificationte st pro gram wa s successfully conducted to certify the gold-nickel-brazed configurationfo r flight.Helium pressure regulator. - The original subcontractor had difficulty meetingthe speci fications for external-leakage control and cont rol bandwidth. The require-ments we re determined to be unnecessarily stringent and were modified.Continued fa ilures of the r egu lators to meet the lockup and cold-helium require-ment led to the sear ch for a suitable backup. The helium regulato r used for the LMand the ser vic e module reaction control sys tems w a s chosen and w a s modified to meetthe flows that we re consistent with the APS requirements. A backup subcontrac torwas engaged to provide the helium regulator. Testing of this regulator configurationshowed low-frequency oscillations at nominal APS flow rates. Evaluation of the test

7


12/29

re su lt s indicated that the regula tor was ve ry sensitive to the APS modular configurationand, possibly, the oscilla tions were caused by coupling between the check valve and theregulator. Te sts were run at the WSTF on test r ig PA-1 to determine the effects ofthe oscillation on the sys tem. It w a s determined that the oscillations did not propagateto the engine interface; therefore, the oscillations did not affect engine operation. A sa resul t of these te st s, the specification was changed to allow oscillations less than1 5 ps i peak to peak, which we re r epre sen tive of those obse rved during PA - 1 testing.Both subcontractors submitted regu lato rs to the qualification program. Theoriginal subcontractor submittal had many discrepancie s and underwent many fai lu re sthat would re qu ir e fu rthe r developmental work. The backup subcontrac tor submittalunderwent one failure; the regulator failed to lock up after a long blowdown. This fail-ure w a s attributed to freezing of excessive moisture in the helium used during testing.The moisture was verified as the failure sou rce during subsequent tests; therefore,the backup unit was selected.The pr im e contractor in- house investigation of the cause and effect of oscilla tionsin regulated pr essur e was continued, and a muffler to damp the oscilla tions was devel-oped and added to the pres suri zat ion syst em fo r LM-3 and subsequent vehicles in the

spring of 1968.Squib valve. - During the DVT pro-gra m for the squib valve, it was noted thatthe piston of the valve did not trav el to thefully open position. The cause of the fail-ur e was determined to be a design defi-ciency i n the piston seal (fig. 5). Thisdesign deficiency allowed pyrotechnicblowby. The valve was redesigned and, bymeans of subsequent testing, it was veri-fied that the system was acceptable.In late 1966, a squib valve that hadpreviously met all acceptance-test leakagerequirements w as found to be leaking at abr az e joint while it was instal led on one ofthe WSTF est r igs (PA-1) . An investiga-tion of the failu re revealed an inadequate

-Figure 5. - Squib valve.

bra ze joint between the inlet tube and the valve body. The faulty bra ze joint probablyresulted fro m excess ive porosity o r poor brazing flow. The design of the bra ze jointdid not lend the joint to adequate inspection procedures; ther efore, the pri me cont rac-to r initiated an inline braze change fro m an internal br aze joint to an external br az ejoint that could be inspected. The modified design was not incorporated at this pointin the program because the failure was believed to be an isolated case.

During the Apollo 9 (LM-3) mission in March 1969, the descent-engine pressur-ization system had a leak that was sim ila r to that observed at the WSTF on test r igPA-1 . Later, af te r the LM-2 drop test in May 1969, another squib valve was found tobe leaking af te r operation. As a resul t of the se additional failures , a decision wasmade to replace all valves of the ear ly br az ed configuration on the lunar modules..

8


13/29

Quadruple check valve. - Only minor problems wi th the quadruple check valvewere experienced in the development program. The qualification-tes t program wa ssuccessful, even though some of the valves did not meet the leakage specification of1 0 scc/hr. The leakage rat es experienced were determined to be acceptable for theLM missions because squib valves were retained for helium-system isolation purposesand operated just before ascent-engine operation.Propellan t tank. - During development testing, a production LM ascent- stagepropellant tank w a s mounted in a fixture, placed in a temperature-controlled environ-mental chamber, filled with oxidizer (nitrogen tetroxide) to 9 8 percent of the total tankvolume, tem perat ure conditioned at 1 0 5 " F, pressurized to 245 psi with 1 0 0 percentgaseous helium, and maintained at these test conditions for 47 hours. A t the end oft h i s period, the test vessel started losing pressure at the r at e of 2. 5 psi/min, indi-cating a tank fai lure. Visual inspect ion of the tank revealed a 0.5-inch crack in themembrane area of the lower hemisphere. A meta llurg ical evaluation of the tank mate-rial disclosed the presence of stress corrosion similar to that experienced at about thesame time period with other titanium oxidizer tanks. The amount of nit ric oxide in theoxidizer had been reduced recently in the proc ess specification to procure pu rer nitro-

gen tetroxide. The resolution (addition of n itr ic oxide to the nitrogen tetroxide ) w a scommon to all oxidizer tanks in the program. The amount of ni tri c oxide requ ired innitrogen tetroxide to eliminate t h e stress corrosion w a s 0 . 4 5 to 0 . 9 5 percent.In an effort to save weight, all-welded tank configurations were developed inwhich the manhole cl osu res at the bottom of the tank and on the diffuser at the top ofthe tank wer e modified. Th is alte ration saved approximately 8 pounds per tank byeliminating the heavy bolted flanges. Titanium pipe elbows wer e welded to the closureand the diffuser, and bimetall ic joints were used to make the transition fro m the tita-nium elbows to the type 304L sta inless-s teel lines. The zero-gravity can and antivortexbaffle were redesigned using titanium rather than aluminum. The modified tanks suc-

cessfully completed qualification in May 1969 .Propellan t feedlines. - Originally, the propellant feedlines were designed to incor-pora te flex hoses to abso rb the movement of the lines. During design-verificationtesting, a fa ilu re of the flex hose was experienced. In June 1966 , the flex hose wa sreplaced wi t h a bellows assembly. The bellows assembly was modified fur the r toincorporate rest raining ba rs a fter misalinement of the line s w a s noted on LM-1. ForLM- 3 and subsequent vehicles , ball joints were installed instead of the bellows.Because of the weight problem of the LM , a weight-saving program w a s initiatedthat resulted in reducing the propellant-feedline wall thickness. The new feed lineswere installed on L M-6 and subsequent vehicles, and the subcontractor initiated a

qualification pro gra m to veri fy the acceptability of the thin-wall line and ball- jointassembly. A leak w a s detected during vibration qualification testing on the lightweightlines. The leak w a s located at the bracket t h a t holds the li ne to a torus ring at thebottom of the tank (fig. 6) . The propellant-feedline design required welding of thebracket to the line; during subsequent vibration testing, cr ac ks wer e noted in thewelded area. Examination of the fa ilu re revealed that the problem resu lted fr om aconcentration of stresses in th e welded area during vibration. The solution w a s todistribute the stresses by incorporating another supporting member in the design.

9


14/29

iI

Figure 6. - Ascent propulsion systempropellant-tank torus ring.

During the setup for the tank-qualificationtes t, which incorporated the section of pro-pellant line where the leak had been experi-enced, crac ks wer e noted in the weldedareas. A thorough investigation revealedthat the cr ack s we re induced during thewelding of the bracket to the line. The fol-lowing decis ions were made as a res ult ofthe investigation.

1. The LM-5 (heavyweight config-uration) w a s to be flown with no modifica-tions because the 0.049-inch wall thicknesswas provided al ready .2. The bracket f or all lines less thanthe 0.049-inch wall thickness was to beground off and, i f the re were no cracks, aclamshell bracket and saddle were to beinstalled with epoxy for str uct ura l support.This modification was made on the LM-6oxidizer l ine and on both oxidizer and fuellines fo r LM-7 to LM-9.3. For LM-10 and subsequent vehi-cles, a new propellant-feedline sectionwithout the weld w a s to be used.-

Ascent EngineThe contr act to develop the asce nt engine was awarded in July 1963. By ear ly1967, it was apparent that the injector development probably would cons train the Apollomissions as a res ult of failu res that had occurr ed during stabi lity testing. Consequent-ly, a backup con trac tor wa s selec ted to develop an injector that would satisf y therequi rements of t he Apollo miss ions and that would be compatible with the other enginecomponents developed by the original engine contractor.The original ascent-engine development w a s divided into four categories : theinjec tor, the ablatively cooled thrust chamber, the bipropellant valve, and the engineassembly.Injector. - Thre e different types of injector patterns w ere relea sed fo r fabrica-tion. The types wer e an alternating-triplet gr id pattern, an alternating-triplet gridand radial patterns, and a radial-triplet pattern (fig. 7). Each pattern used a barr ierdoublet (oxidizer and fuel) for chamber cooling.In the s umme r of 1964, an inj ector that had a radial-triplet pattern was selectedf o r the prototype because of its characteristic compatibility with the ablatively cooled. thrust chamber. Performance was still a problem; there fore, modifications to the

10


15/29

injector pattern that increased the film-bar rier propellant-mixture ratio to achievehigher performance and still maintain thrust-chamber compatibility wer e pursued.

Fuel p r i m a r y0.0242 0 . W 5O x i d i z e r p r i m a r y0.0419 0.0005

Fuel b a r r i e r

144 p r i m a r y t r i p l e t s56 b a r r i e r d o u b l e ts

A l t e r n a t i n g t r i p l e t g r i d

Fuel p r i m a r y 0.0250 0.0005O x i d i z e r p r i m a r y 0 .0 4 3 0 + 0.0005;ypo r e r

+ 0.0005O x i d i z e r b a r r i e r0 . 0 3 % 0.00050.0310 ? 0.0005

0.03% 0.0005

O r i f i c e s i z e sB 3 t ype B 3 - L t y p e

P r i m a r y o x i d i z e r 0.0492 0.0571P r i m a r y fuel 0.0292 0.0330B a r r i e r o x i d i z e r 0,0210 0.0236B a r r i e r fuel 0.0260 0.0295

120 p r i m a r y t r i p l e t s ( f u e l o n o x id i z e r14 p r i m a r y t r i p l e t s ( o x i d i z e r o n lue l l

D i m e n s i o n s a re i n i n c h e s32 b a r r i e r d o u b l e ts

A l t e r n a t i n g g r i d a n d r a di al R a d i a l t r i p l e t

Figure '7. - Types of injector g rid patte rns.In the fall of 1964, two changes in requirements were imposed: the NASA MannedSpacecraft Center (MSC) combustion- stability c ri te ri a were clarified and the continuous-burn time w a s increased to 460 seconds because of increased vehicle weights.Combustion- stability t es ts on the unbaffled injector wer e delayed because of the lackof availability of instrumentation. In late January 1965, the unbaffled injecto r wasdetermined to be unstable under the new criteria, and the design of a baffled configura-tion w a s initiated.The original cont rac tor investigated several baffle configurations. Using expe-rience from the descent-engine program, a flow-through baffle design for cooling was

implemented. Th is design was selected, even though sev era l combustion instabilitiesoccur red with metal chambers used in bomb test s during the development phase of theprogram. These fail ures wer e explained as hardware failure before instability, loca-tion of the bomb in the throat of the thr ust chamber that was thought to be an unrealistictest condition, o r the resu lt of a gap between the baffle and the chamber. This deci-sion was explained furth er with the hypothesis that an ablatively cooled chamber wouldhelp damp combustion oscillations. Two versions of the baffled-injector configurationwere designed. Because of fabrication problems, both injector s required 6 to '7 monthsto fabricate. A s a result of the time required to develop the baffled injector and therequirement to meet schedules, the production release w a s required before eithe r

11


16/29

configuration w a s hot fired. A decision was made to select the configuration fo r pro-duction that gave the higher per formance and ablative compatibility, as determinedfrom the test results of various development injectors.The init ial fir ing of the production injector exhibited ero sio n of the ablativechamber mate rial ne ar the injector-chamber interface. To eliminate this problem,modifications we re made to the bar ri er orifices; thes e modifications reduced the

eros ion. During the oxidizer- and fuel-calibra tion runs fo r the engine qualificationwith the baffled injector, the engine underwent combustion instability when bombedwith an arti fici ally induced distu rbance in an ablatively cooled chambe r. The enginew a s inadvertently allowed to continue to fire in thi s unstable mode for a tim e sufficientto caus e extensive damage to the injector. Because of the damage, it could not bedete rmin ed whether the hardw are wa s defective before instability o r whether the in-stability ca used the damage. It was concluded that the instabilities were the result ofhardw are failure. Additional testing proved that these conclusions we re erron eous.A plan was devised to evaluate the ability of the engine to withstand combustiondistu rbanc es without becoming unstable and to ch arac teri ze the instabilit ies i f they did

occ ur. The lack of hard war e cau sed th is plan to proceed slowly. When combustioninstability oc cur red in an ablatively cooled chambe r during bomb testing, it was con-cluded tha t the injector was not satisfactory fo r the Apollo vehicle. Lat er , a sponta-neous instability occurred 290 seconds into an accepta nce firing, fur the r substantiatingthe conclusion that the inject or design was unsatisfactory.Numerous modifications were madeto the injector in an attempt to solve thestabi lity problem. A photograph of thefinal configuration (fig. 8) shows the injec-tor and the devices that were used in anatte mpt to ens ure sta ble operation of the

engines. The final configuration exceededthe performance requirement for theengine.Ablatively cooled thrust chamber. -The original design fo r the as cent enginedefined the thrust chamber as being abla-tive, with a radiation-cooled nozzle. Theradiation- cooled nozzle had been shown tobe technically feasible ; however, thi s con-cept had never been applied to an enginethat wa s buried in the vehicle o r that wasrequired to perform fire-in-the-holestarts fr om another s tage (using the

Figure 8. - Stability devic es on theproduction injector.descent stage as a launch platform). Evaluation of the possible proble ms with theradiation-cooled nozzle led to the selection of an ablatively cooled nozzle fo r the ascentengine, even though the engine weight would be slightly grea te r. Some of the re as on sfo r abandoning the radiation-cooled nozzle were high design ri sk , high cos t, lac k ofdevelopment experi ence with buried installations, and modifications require d in thetest facilities.

1 2


17/29

The tota l weight of the chamber with a n ablatively cooled nbzzle was approxi-mately 161 pounds. A weight-saving program caused a change in the ma ter ial of thethrust - chamber insulation and nozzle ablat or and, thereby, the weight was reduced34 pounds. Te st re su lt s showed a marked improvement in the ablative ch arac ter isti csof this chamber and it wa s certif ied fo r us e on manned flights.Bipropellant valve. - The bipropellant-valve concept w a s a series-parallel ball-valve arrangement that w a s actuated hydraulically by an elec tri cal solenoid pilot valve.The development testing of t h e valve wa s performed at the valve level; however,because of direct ion to qualify the component at the highest possible subassembly level,qualification testing w a s performed at the engine level to cer tify the valve fo r LM-1.Out-of -specification leakage w a s observed during qualification testing of the valveactuator. The leakage wa s acceptable fo r the unmanned flights, but design changeswere incorporated for l ate r vehicles.A s a re su lt of long -term usage of bipropellant valves in oxidizer vapors at theWSTF, it w a s found that the needle bearings were subject to cor ros ion . Thus, inmid-1967 the needle-bearing material was changed to stainless steel. In mid-1968,

the fuel actuato rs experienced leakage caused by O-ring twisting during valve operation.Several valves on production engines at the backup-engine cont rac tor plant were re-jected. Top-visual-quality 0- rings were instal led without su ccess. Subsequently, itwas determined that the or iginal leakage requirements were too string ent and that thevalves were satisfactory under more realist ic requirements.Engine assembly . - Analyses of the heat-shield st ructural marg ins during fire- in-the-ho le (FITH) test ing of the engine indica ted a negative margin because of thecombustion-chamber-pressure peaks during the start transient of the engine. The con-tr ac tor investigated sev era l means of reducing the chamber- pres sure overshoot withoutchanging the valve design, but all solutions tried were unsuccessful. Therefore, theheat-shield str uc tu re on the vehicle w a s modified to accommodate the higher chamber

overshoot during FITH testing. Similar problems were experienced by the backupcontractor.Because of the problems i n combustion stability and ablative compatibility asso-ciated with the first production inject ors, the engine-development program of the orig-inal contractor w a s divided into two phases. The first phase was a test program tocer tify that the ear ly engines, equipped with early production inject ors, wer e accept-able for flight on the LM-1 and LM-2 unmanned vehicles , and the second phase wa s atest program to certify that the engine with a stable combustion injector w a s accept-able for the manned LM-3 and subsequent manned vehicles.The first phase, certifying the engine for LM-1 and LM-2 vehic les, consisted ofone DVT sea-level test engine, five DVT altitude-test engines, and th re e DVT altitude-qualification-test engines. Two additional problems were noted during t h i s test pro-gram. The proble ms were the ablatively cooled thrust-chamber erosio n and leakage inthe valve actu ato r. However, it was concluded fro m these tests that the engine wassatisfac tory f or the unmanned vehicles.The second phase of the program consisted of thr ee DVT alt itude engines, oneDVT sea-level engine, and fo ur DVT altitude-qualification eng ines. Because theinjector-fabrication processes were changed to use electrical-discharge machining

1 3


18/29

of the orifices, an additional qualification engine and one DVT sea- leve l engine wer eadded to the test program. Thrust-chamber erosion w a s observed during these tests,Backup engine contractor. - The backup contract was initiated and managed bythe MSC with close inter face with the vehicle cont ractor. The scope of the contract thatwas awarded encompassed the design, development, and qualification of a flight injecto rfor the LM ascent engine. This effort al so included the assembly, test, and production

of complete rocket engines by integrating this qualified injector with thrust ch am bersand propellant valves furnished by the original engine contractor.Candidate injector s we re selec ted, tested, and evaluated to evolve a stable,high-performance configuration that w a s compatible with the ablatively cooled thrustchamber. The basic injector design selected was a baffled, flat-faced, multiring unitwith integral propellant-distribution manifolding and acoustic cavities . The th re einjector-orifice patterns that were evaluated included a triplet, consisting of two fuels tr eams impinging on each oxidizer stream; an unlike doublet, in which a str ea m ofoxidizer impinged on each fuel stream; and a mixed doublet, in which this mixing ofoxidizer and fuel was a secondary res ul t of fuel-fuel and oxidizer-oxidizer impingement.The unlike doublet was selected as the production injector because of its per-formance, compatibility, and stability margins.minor changes in the fuel-film cooling in the bas ic unlike-doublet injector design tooptimize performance and ablative compatibility while maintaining a wide margin ofcombustion stability .ibility requirements.

Efforts were directed toward making

The final configuration exceeded the performance and compat-

Because of a combustion instability during bomb testing on an injector that hadexperienced se ve re face damage, a "soft center" modification was made to improvestability chara cte ris tic s. The modification cons isted of closing off six of the nineimpingement sets at the center of the injector . The change reduced the injection densityin the center of the injector and reduced the amount of propellant available fo r possibleenhancement of radia l combustion waves.

A complete final-design injector is shown in figure 9, and a section through thepropellant i nlet s is shown in figu re 10 to illust rate the flow passages. An ear ly injec-tor inlet screen (filter) is shown in figure 11 installed in the typical manner. A s canbe seen i n figure 11, the filte r is a woven-wire-cloth truncated cone with plea ts. Thefil te r has a solid plug that effectively se rv es as a dynamic-head- suppression device,and the filter p rotects t h e injector orifice s fr om contamination. The pleated filte r,made of all-welded corrosion- res ist ant steel, with a 120-micron length along twomajor axes, was designed to stop all particles.Although the pleated filter successfully completed all the required qualificationtest s, the contaminant-capacity req uirement s of the test plan were not consistent withlev els of potential contaminant in the vehicle propellant tanks. The potential fo r pr es -su re drop ac ro ss the pleated-injector-inlet filters could have a significant effect on theengine mixture ratio and, consequently, on the ascent- sta ge propellant utilization. Forvehicles not landing on the lunar sur face , the propellan ts usually were off-loaded, pro-pellant utilization was not considered crit ical , and the pleated filters were acceptablefor u se in engines fo r these vehicles. A more efficient inlet sc re en that was to have a

'

14


19/29

Figure 9. - Final-design injector.I

i

Figure 11. - Injector filter.

Figure 10. - Propellant -inlet sec tion ofthe injector manifold.lower pre ss ur e drop when contaminated wasdesigned fo r lunar-landing vehicles . Thenew injector-inlet filter (caged filte r) wasdesigned to meet the revised requireme ntsfor contaminant capacity and to provideadditional s tru ct ura l support fo r preventingscreen deformation.

After qualification tes ts w ere com-plete, a structura l fa ilure of a productionfilter o cc urr ed during an engine-acceptancetest. The fail ure was caused by theresistance-weld procedure that was used forattaching the heavy, oute r mesh-suppor tbasket to the upper collar of the filt er. Thestrength developed by the weld was insuffi-cient to withstand the maximum filter loadduring servic e. A modification that elim-inated the need for resi sta nce welding wasmade to the filt er design.

The occurre nce of preigni tion-press ure spikes during engine startu p required achange in propellant ducting to ensure that the oxidizer would reach the combustioncham ber befo re the fuel. The modified duct is shown in figur e 12.Fo rmal demonstration of the accept-ability of the complete engine for lunarflight w a s accomplished by testing sixrocket-engine assemblies through a qual-ification se ri e s of environmental t es ts andhot-firing demonstrations. The engineswere produced as complete production unitsand were fully representat ive of flightengines. Reuse of bipropellant valves dur-ing this progr am was nece ssitated becauseof hardware shortage . The acceptance .

Figure 12. - Modified engine fuel duct.1 5


20/29

testing revealed a problem in production fabrication that required a change in the weld-ing procedure to prevent fa tigue crack s fr om developing in the weld attaching the bafflebase to the injector face. The change in the welding procedure improved the injectorso that it was declared capable of per forming the r equi rements of a manned lunar-landing mission.

Used at the WSTF wi t h test r ig1 H A - 3

The selection of the backup-contractor engine fo r use on the flight vehicles w a smade from an evaluation of each cont ractor production engine. The backup con tracto ral so had complete qualification-testing faci litie s and had production hardware availablefor installation on the vehicle when required.The responsibility fo r the backup engine w a s assigned to the vehicle contr actorsubsequent to t h e decision to use that engine fo r the LM . Formal demonstration of theacceptability of the complete engine wi th the lighte r weight chamber, updated valve,and improved engine fi lt er s w a s accomplished by testing four additional rocket-engineassemblies under the direction of the vehicle contractor.

T es t A r t i c l e s and Tes t R i g sSeveral test a rti cle s and test ri gs were developed and tested to accomplish theAPS testing program. The test articles and test rigs contributing to the APS develop-ment and qualification a r e listed in table I.

TABLE I. - MAJOR TEST ARTICLES A N D RIGS

Development phase

Breadboard

Test articleo r rig Remarks

PressurizationmodulesBA- 1BA- 2

BA- 3BA- 4

Used at the factory with test rigH A - 1

Used at the or iginal engine-contracto r facility with testr ig HA-2

Cold-flow tests at t h e factory

1 6


21/29


22/29

TABLE I . - MAJOR TEST ARTICLES AND RIGS - Concluded

Development phase

Full-configurationtest article

Test art icleo r r igLTA- 3

LTA-4

LTA- 5LTA-6LTA- 7LTA-8

LM- 2

Re mark s

Used for str uc tur al and vibrationtests to define component vibra-tion levelsEnvironmental te st s in the factoryvacuum chamberDescent propulsion rig at the WSTFCanceledCanceledThermal tests at the MSC to definethermal limitsUsed in drop tests at the MSC.Ascent stage sent t o Osaka,Japan, fo r Expo '70. It wastra nsf err ed to the Smithsonianfollowing Expo '70.

Breadboard testing. - The breadboard testing was initiated to achieve ear ly sys temverification of design, inte raction of components, and technica l feasibility of design.Four heavyweight tes t rig s were built; these ri gs were essentially open stru ct ure supon which the var ious propellant-feed- section components and the p ressurizationmodules could be mounted in the s am e arrangement and rela tive locations as on theLM. The rigs were used fo r cold-flow testing at the prime-contractor facility, f ortesting at the engine-contractor facility in the ea rl y engine-development program, fo rhelium-pressure-blowdown tests in the prime-contracto r cold-flow facility, and fo rsyst em testing with preproduction engines and components at the WSTF.Prototype testing. - Most of the prototype tes ting of the APS wa s conducted on theWSTF test rig PA-1. This test rig was an ascent-stage stru ctur e that incorporatedthe APS and RCS equipment in an approximate flight configuration (fig. 13). Simula-tion was used except where it would affect the APS perf ormance . The test rig accom-modated both t h e pressure-fed, 3500-pound-thrust ascent engine fo r firi ng in adownward position and the 16 RCS engines, using a separ ate propellant-supply systeminterconnected with the APS propellant tanks. The PA - 1 ascent stage was constructedof aluminum alloy and consisted of th re e major sections: the fo rw ar d cabin, the mid-section, and the aft equipment bay. The s tr uc tu re included provisions f o r installing

18


23/29

heat shie lds for the FITH testing. A sum-mary of the te st s conducted on te st r ig PA-1is given in table 11. These tests includedall normal mission requirements and anumber of off-limit tests of possible prob-lem area s such as FITH ests , pressureovershoots, and component-abort tests.Although all of these ho t-firing te st s wer enot conducted with a qualified engine injec-tor, the tests proved successfully the in-tegrity of the propellant and pressurizationsections.

Figure 13. - Test r ig P A - 1 .TABLE II. - THE PA - 1 WHITE SANDS TEST FACILITY T EST SERIES SUMMARY

Testseries

12

4

56

7A7C8 R8

EA

5A97B

11

NumberofrunsSeri es features

Cold-flow (substitute and actual propella nts)FITH starts, abort, and normal mode star ts, prep ress ure bal-ance and unbalance, compatibility of APS with RCSLM-1 mission duty cycle, rest art s, en gn e stability, propellantdepletion, off-nominal performanceCold-flow (substitute and actual propellan ts) LM-1Sea-level firings, LM-1 mission duty cycle ( 2 ) , pressure regu-lator evaluation, abort, and normal and off-limit startsFITH start s, abort, and normal mode st ar tsCold-flow (substit ute and actual propellants ) LM- 3Mission duty cycle to depletion, off-nominal st ar tsMission duty cycle with APS/RCS, heat-soak-back tem pe rat ur es ,helium-saturated-propellant evaluationPropellant depletion, pr ess ure regulator malfunction, abort

starts, propellant utilizationPropella nt utilization with off-nominal conditionsBlowdown mission duty cycle, propell ant utilization, cold sta rtFITH st ar t evaluation, adequacy determinati on of t her ma lprotective blankets and blast deflectorHigh- mixture -ratio mission G duty cycle demonstration

aN o hot firing during cold-flow tests

1 9


24/29

Wi th flight-qualified engines, 57 es ts we re conducted over a total firing timeof 3392 seconds. Five engines and six thrust chambers (three heavyweight and threelightweight) were used. The following a r e the re su lt s of these tests.1 . The mixture ratio of the ascent engine could be predicted within 0.6 percentfo r engine operation at chamber pr es su re s of 1 1 2 to 1 3 0 ps ia and temperatures of40' to 100' F with helium-saturated and unsaturated propellants.2 . Combined with RCS operation, the overall APS and RCS mixture rati o couldbe predicted within 0. 75 percent.3. The validity of the propellant-feed- sys tem cold-flow calibrat ions was veri -fied, and the flight -engine ch ara ct eri sti cs we re confirmed.4. The FITH starts, at simulated lunar-launch conditions, were perf ormedwith no s truc tura l damake o r adverse effect on engine performance.5. Abort s ta rt s were performed safely at ullage pre ss ur es of 62 to 215 psia.6. Engine operation was not affected adversely by operation on redundantregulators.7. The transition to the adjusted s yst em-pr es sur e lev els was smooth andgradual.8. The engine operated safely in the tank-ullage-decay (blowdown) mode fro mnominal chamber pressure to 8 psia, and it could be safely shut down by means ofpropellant depletion.

SYSTEM CHECKO UTDuring the functional checks on LM-1 and LM-3, the relief-valve bur st disksruptured prematurely. Investigation revealed tha t the burs t-disk design would notallow any backpressure on the disks without pre ma tu re ruptu re. A s a re su lt of thesefailures, the checkout procedures were reviewed and cor rec ted in an attempt to pre-clude th is occurrence on other vehicles. In spit e of the precaut ions taken, prem at urerupture of relief-valve bur st disks remained a checkout problem throughout theprogram.Quadruple check-valve leakage was a common checkout problem, and mostvehicles required special work and testing to co rr ec t the leakage. Either purging toco rr ec t the leakage, valve replacement, o r waivers to accept the leakage we re requiredon most vehicles. The many debrazing and brazing opera tions that we re requ ired inor de r to replace the check valves also introduced contamination into the system. A s aresult, complete helium- module replacement was ne ce ssar y (for instance, on LM- 3).Design changes were considered to improve the checkout problem but wer e notimplemented.

20


25/29

Ear ly vehicle checkout w a s plagued by component replacement resulting fromfail ures of t h e components either during checkout o r elsewhere in the pro gra m (suchas in qualification syst em test s). Particularly , failure w a s frequent on the heliumregulators, the helium solenoid valves, the helium relief valves, and the engine.

FLIGHT PERF0 RMANCEThe flight prog ram f or the LM consisted of one unmanned flight (LM-1) . TheL M - 2 vehicle was planned fo r an unmanned flight, but this was canceled and L M - 2 w a sused as a ground-based test article. The first manned vehicle was LM-3, whichdemonstrated successful operation in earth orbit. The LM-4 vehicle was used inlunar-orbit-operation tests, and LM- 5 was the first lunar-landing vehicle.

A p o l lo 5 MissionApollo 5 (LM-1) was the first mission with a flight-configuration lunar module.The primary objectives of the Apollo 5 miss ion included verifying the APS and theabort-staging function fo r manned flight.was installed in the APS. An engine fro m the origina l engine contractorA s a re su lt of the premature shutdown of the descent engine, an al te rnate miss ionplan w a s selected that re sult ed in two ascent-engine firings. The first firing wa.s of60 seconds duration. Approximately 1 - 1 / 2 hours later, the engine w a s again com-manded on and fir ed to propellant depletion. The total ascent-engine f iri ng time forthe mission w a s approximately 40 seconds less than predicted. A t least 2 0 secondsof th is time can be attributed to higher-than-expected propellant usage by the RCS

engines through the propellant interconnect as a resu lt of the control configurationaft er staging. Another 10 seconds of th i s time was caused by propellant slosh. Highvehicle-attitude ra te s caused t h e propellant-tank outlet por ts to be uncovered prema-turely . An oxidizer-depletion shutdown had been expected. However, when the pro-pellant in the feedlines (sufficient fo r approximately 1 second of nominal operation)was depleted, as indicated by t h e engine-interface pr es su re , helium was ingested intothe oxidizer and the fuel lines almost simultaneously, causing thrust decay. A s indi-cated by the abbrev iateddecay, an additional 1 0 seconds of normally usable propellant w a s in the tanks at thrustdecay, but w a s unavailable because of the sloshing and high vehicle-ro ll rates.time between low-level sensor uncovering and thrust

The start tran sien t and the beginning of steady-s tate operation showed high-amplitude chamber-pressure oscillations on both the first and the second engine starts.The 400-hertz oscillations , which occurred immediately after the start-transient over-shoot, were ch arac ter ist ic of the original-contractor engine during ground-basedtesting and wer e expected on th i s flight. The oscillations appeared to be a form of low-frequency instability, caused by t h e coupling of the combustion pro cess with the reso-nant frequency of the engine propellant-feed system. Ground-based tests wereindicative that gaseous helium, dissolved o r entrained in the fuel, tended to inducecoupled instab ility.

21


26/29

The high-magnitude 400-hertz oscillations resumed approximately 4. 5 secondsbef ore thrust decay, an indication of propellant g as entrainment caused by the vortexingo r sloshing resulting f rom the high vehicle-attitude rates that occu rred during this por-tion of the mission. The th ru st decay appeared smooth; no spikes o r other detrimentalef fe ct s were apparent af te r the damping of the 400-hertz oscillations and the initiationof the chamber-pressure decay.The engine valves should close immediately af te r fuel depletion because the valvesa r e fuel-actuated. Following depletion, the engine-valve-position indica tor s showedthat the engine valves did not start to close until 85 seconds after the thrust star ted todecay and we re not closed completely when the telemetry signal was lo st by the groundstation 58 seconds lat er. System pr es su re s and temperatur es indicated that someblockage of fluid flow occurred between the propel lant tanks and the engine after theinitial thrust decay. The blockage was most likely caused by smal l amounts of propel-lant that were frozen by cold helium ingested into the feedlines. Although it was not anormal depletion shutdown, no hazardous o r detrimental effects wer e apparent.The performance of the ascent engine was not veri fied to have been within the

expected accuracy because of the l oss of A P (change in pr es su re f rom tank outlet toengine interface) flow-rate measurements and because of the high roll r a te s of thevehicle during propellant depletion; however, the engine-pressure measurements andthe vehicle velocities that were obtained indicated that the ascent-engine performancewas within the nominal predicted to lerances .

Apollo 9 MissionThe Apollo 9 miss ion (LM-3) included the first manned flight test of the LM. Themission was the second manned flight te st of the Saturn V launch vehicle and the thir dmanned flight of the Block II command and ser vi ce module (CSM).The overall mission objective was to evaluate crew operation of the s eparated LMand to demonstrate docked-vehicle functions in an earth-orbital mission, thereby qual-ifying the combined spacecraft for lunar flight. The LM operations included an ascent-engine firin g to propellant depletion af te r final separat ion fr om the CSM.The APS w a s used for two firings: a 3-second firing while the ascen t s tage wasmanned and an unmanned fi ring to propellant depletion. The LM was out of ground-tracking-station range during the first ascent-engine firing; therefore, no data wereavailable. However, when data were first acquired after the firing, system pre ssu re sand tempera tures were normal. The second ascent-engine fir ing was initiated suc-cessfully and last ed fo r 362. 3 seconds. During the second firing, sy st em pre ss ur eswe re lower than expected for the first 290 seconds, thus indicating a malfunction i n thepri mary regulator. The lower operating pr es su re s produced no undesirable effects inthe system. The second ascent-engine fir ing was terminated by the planned oxidizerdepletion. The engine wa s commanded off approximately 10 seconds lat er. The deple-tion shutdown appeared nominal in all resp ects . The transient char acte rist ics that theengine demonstrated during the oxidizer-depletion shutdown mode wer e investigatedand were found to compare favorably with ground-based test data. All applicabletransient- specification requirements we re met.

22


27/29

A detailed asse ssment was made of the lower-than-expected syst em pre ss ur es todeter min e the cause and magnitude of the malfunction. After extensive testing in thethermochemical test area at the MSC and on test rig PA-1 at-the WSTF, it was con-cluded that the most probable cause of the anomaly wa s contamination of the regulatorduring the solenoid-valve replacement. A review of regula tor backflow and sys tem-rep air pro cedur es during checkout on LM vehicles indicated th e possibility of minorbackflow on vehicles through LM-7. Cautions were added to th e changeout proced ures ,and special instruc tions were given to personnel. Incr eased surveillance of regulatorcheckout data w a s maintained by the subsystem office for all subsequent vehicles.

Apollo 10MissionThe Apollo 10 mission was the first flight test of the LM in a lunar environmentand the second manned LM flight. Because the miss ion objective w a s to show a fire-to-depletion firing and not a AV, the amount of propellant loaded w a s approximately50 percent that of the nominal lunar mission.The overall mission objective w a s to duplicate, as closely as possible, a lunar-landing mission, with the exception of lu nar landing and lift-off. Inherent in thi s objec-tive w a s the performance of th e APS during the lunar-orbit -insertion maneuver. Alsoincluded as mission objectives were verifications of LM operation in a lunar environ-ment and of mission suppor t of all spacecraft at lunar distances.Docking of the CSM with the LM and separation of the docked vehicles f rom theSaturn IVB occu rred 4 hours af te r launch. Undocking of the LM from the CSM in lunarorbit occurred 98.5 hours after launch. Approximately 2 hours aft er completion ofthe descent propulsion syst em firing, the descent stage was separated and the APSengine was f ire d for 15 . 6 seconds. Upon completion of this insertion maneuver, the

ascent stage was docked with the CSM, and the crew and equipment tr an sf er was accom-plished. Approximately 6 hours later , t h e ascent stage w a s separated, and the enginewas ignited fo r the 248.9-second burn to propellant depletion.Shortly af ter the ignition signal fo r the first APS burn, the ascent-engine-quantitycaution light came on, triggering a maste r caution and warning al ar m. The ascent-engine-quantity caution light was controlled by the oxidizer and fuel low-level se nsor sin the propellant tanks. Approximately 1 second after the ascent engine "on" signalfor the manned LM-4 APS burn, the oxidizer low-level sensor was activated fo r approx-imately 1 second while the m as ter al ar m was on for 2 seconds befor e being manuallyres et. Low-level sens ors in both oxidizer and fuel tanks operated as expected duringthe remainder of the first burn and the enti re second burn. Based on th is performance,

it was concluded t ha t the signal received during the first burn was valid.The cause of the anomaly was believed to be insufficient set tling of APS propel-lan ts before the burn. The LM-4 propellant tanks were filled to 50 percent, the lea ston any flown vehicle. Based on information contained in the Spacecraft Operation DataBook, an RCS ullage-propellant-settling maneuver of 3 o 4 seconds in duration wouldbe adequate for a.vehic le with the propellant load of LM-4. The actua l RCS ullage-propellant-settling maneuver before APS firs t-burn ignition las ted for a period of4.1 seconds. The propellant-settling maneuver wa s reexamined, and the settling ma-neuver duration was increased fo r increased ullage (decreased propellant) volume.

23


28/29

Apo l lo 11M i s s i o nThe Apollo 11 mission (LM-6) was the th ird manned flight of the LM, the fifthmanned flight of the Block 11 CSM, and the fourth manned flight of a Saturn V launchvehicle. The pr imar y objective of the mission was to perf orm the first manned lunarlanding and return .Burn time for the APS lunar lift-off maneuver was 434.9 seconds. Upon comple-tion of the insertion maneuver, the ascent stage docked with the CSM, and the crewtran sf er and equipment tr an sf er we re effected. Approximately 3. 5 hours later, theascent stage w a s jettisoned. A l l APS performance p ar am et er s wer e well within thei rrespective limits.

1

CONCLUDING REMARKSIn review of the ascent propulsion sy stem development, sever al areas that could

enhance the development of future subsystems a r e discussed in the followingparagraphs.

D e f i n i t i o n o f R e a l i s ti c R e q u i r e m e n t sRequirements for the system and its components should be thoroughly definedbefore the beginning of the design phase. Typical examples of inadequate definitionsat the beginning of the design phase we re the stability requ irement for the engine, thepropellant-vapor compatibility requirement f or the pressu riza tion components, and theleakage requirements. Late definition of the stability and compatibility requi rementscaused redesign, which w a s reflected by cos t in cr ea se and schedule slippage. Overlystringent leakage requirements caused failures and redesign efforts that were notnecessary .

F lex ib l e P r og r am Flow P l a nEarly in the testing program, the pri me contractor and vendors were working toa cost-incentive flow plan. As a result, vendors directed the ir efforts to delivery ofcomponents ra th er than to problem solving and hardware development. This policyresulted in failure s late in the development phase and in component replacement onthe ascent propulsion system. This problem was most apparent in the engine and regu-

lat or component area. A more f lexible flow plan would have allowed component instal-lation at later opportunities.Leakage and Con t am ina t i on Co n t r o l

Lunar module 1 had many leakage prob lems caused by overly stringent leakagerequirements, the design of the sys tem, and nonstandard measurement techniques.The stringent leakage requirement was initiated by the uncerta intie s and by the

24


29/29

corros ive nature of the oxidizer vapor. When stringent leakage control w a s required,design changes (such as installing redundant seals with the capability to mea sure theintegrity of the pri mary se al) were necessa ry. In addition, leakage-measurement tech-niques were different at t h e vendor facility, at the prime- contra ctor facility, and atthe NASA John F. Kennedy Space Center. Attempts were made to standardize leakage-measurement techniques to ensure the ability to mea sure changes in leakage. Wherestringent leakage control may be required, it is recommended that leakage-measurement techniques be standardized.

Also, contamination w a s a problem (especially with check valves). Th is problemw a s magnified by the many component replacements. Where contamination control iscritical, it is recommended that all functional components (such as regulators andcheck valves) have built-in filters.I ns pec t i on and M a nu f ac t u r i n g Repeatab il it y

The sqi ib valve, the helium solenoid valve, the relief valve, and the engin !underwent prob lem s of manufacturing repeatability. The main cau se of the problem wa sinternal b raz e joints o r welds that could not be inspected. In the engine, the injector,when fabricated, had cra ck s in t h e electron-beam welds. These cr ack s were causedby imp roper control of the welding parameter s. Hardware should be designed to per -mit inspection of weld- and braze-joints. When th is prac tice is not feasib le, the weldpar amete rs should be certified and supported by means of test samples from all quali-fied welders.Manned Spacec raft CenterNational Aeronautics and Space Administration

Houston, Texas, October 30, 1972914- 13-00-00-2

Documents

Apollo Experience Report Ascent Propulsion System