[American Institute of Aeronautics and Astronautics 37th Joint Propulsion Conference and Exhibit - Salt Lake City,UT,U.S.A. (08 July 2001 - 11 July 2001)] 37th Joint Propulsion Conference and Exhibit - Historical perspective of combustion instability in motors - Case studies

  • Published on

  • View

  • Download


(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.A01-34508AIAA 2001 -3875Historical Perspective of CombustionInstability in Motors: Case StudiesDr. Fred S. BlomshieldNaval Air Warfare CenterChina Lake, CATIME,th37m AIAA/ASME/SAE/ASEE Joint PropulsionConference and ExhibitJuly 8-11, 2001Salt Lake City, UtahFor permission to copy or republish, contact the American Institute of Aeronautics and Astronautics,1801 Alexander Bell Drive, Suite 500, reston, VA 22091-4344.(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.Historical Perspective of Combustion Instability in Motors: Case Studies*Dr. Fred S. BlomshieldNaval Air Warfare CenterCode 4T4320D, Propulsion Research Branch, China Lake, CA 93555blomshieldf@navair.navy.milAbstractOver the last 40 years, many solid rocket motors have experienced combustion instability. Combustion instability is theinteraction with a motors inherent oscillatory modes with the combustion and/or fluid dynamic processes inside the motorcombustion chamber. If only a small amount of available energy, less than one percent, is diverted to an acoustic mode,combustion instability can be generated. Standard Missile, Sidewinder, Harm, Trident, Hellfire and Minuteman just to name afew have experienced pressure oscillations sometime during their development and life. Combustion instability can lead toballistic pressure changes, can couple with other motor components such as guidance or thrust vector control, and in the worstcase, cause motor structural failure. Combustion instability in a motor can cost motor developers millions of dollars to eliminate.Sometimes, combustion instability, although present, does not cause system problems and can be tolerated in a motor. In thesecases, it still must be characterized and understood. The purpose of this paper will be to review 28 of these motors and whatsolutions were used. It is hoped that some of the lessons learned in the past can be applied to future motors.Introduction And BackgroundSolid rocket instability is the interaction among theinternal combustion and flow processes with a motor'snatural acoustic modes. The most common acoustic modesexcited are longitudinal and tangential modes. See Fig. 1.Combustion instabilities can be classified into two groups,linear and non-linear. Linear instabilities are those that areusually characterized by low amplitude sinusoidaloscillations that have very simple frequency content. Mostmotors have low-level linear instabilities that often cause noproblems. When linear instabilities become stronger innature and cause DC pressure shifts (changes in the meanballistic pressure trace), have relatively high limitingamplitudes, and more complicated frequency content, thenthey become non-linear in nature. Other forms of non-linearinstability behavior include pulsed instability. This is when amotor is pulsed by something passing through the nozzle orsudden increases in burning surface area due to propellantvoids. These oscillations are characterized by steep frontedhigh amplitude pressure waves. Fig. 2 shows both types ofinstabilities.1 In Fig. 2, linear tangential oscillations occur ataround one second and later, at two seconds, the motor waspulsed into high-level non-linear longitudinal oscillations.No oscillations resulted from Pulse 1. Details of theseoscillations are shown in Figs. 3 and 4, respectively. Motorproblems from combustion instability can be manifested in aFig. 1. Example of Tangential Mode (Left) and LongitudinalMode (Right)variety of ways including thrust oscillations, ballistic pressurechanges, adverse coupling with guidance and thrust vectorcontrol systems, and in some cases catastrophic motor failure.1.0 1.5 2.0Time - secondsFig. 2. Example of Motor Experience Combustion InstabilityAcoustic stability is determined by the variousacoustic gains and losses in the system. Acoustic gainsinclude pressure coupling (coupling between acousticpressure oscillations and combustion processes, verycommon), velocity coupling (coupling between acousticvelocity and combustion), distributed combustion (couplingwith burning metal particles away from the propellantsurface), and flow field interactions (phenomenon like vortexcoupling with the motor acoustics). Acoustic lossmechanisms include nozzle damping (acoustic energy lost outthrough the nozzle), particle damping (for systems withparticulates the particles remove acoustic energy), acousticflow losses (sometimes called flow turning or boundary layer* This effort was sponsored by Independent Research Funding of the Naval Air Warfare Center.# Research Scientist, Research and Technology Division, Senior Member AIAAApproved for Public Release; Distribution is Unlimited.(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.losses), structural damping, and wall losses. Nearly allresearch in the area of combustion instability is involved inunderstanding these gains and losses.1.0 1.2 1.4 1.6Time - secondsFig. 3. Close-up of Linear Tangential Oscillations800600^ 4001 200& 0-200-4001.: In 13 Cycles, the'- Oscillations Grew tolOOOpsiPeak-to-Peak-Initial Pulse j; AbTuaUrt*'VH^vi/y188 1.90r r rf \1 f I1.92 1.94Time - secondsFig. 4. Close-up of Pulsed Non-Linear OscillationsIn order to solve combustion instability problems,either the gains must be reduced or the losses increased.Ways to accomplished this include 1) changing the oxidizer,metal and additive particle sizes which can effect both thepropellant response and the particle damping; 2) modifyingthe internal motor geometry which can effect nozzledamping, acoustic flow field and velocity coupling; 3) theaddition of a stability additive to modify the particulatedamping; and 4) the addition of a burning rate catalyst whichcan lower the response.The purpose of this paper is to concisely documentsome motors that have experienced combustion instabilityand what was done to solve the phenomenon. The amount ofdata on these past motors varies substantially. As such, somemotors only have a brief description with limited discussion.Others have considerably more documentation and moredetail is presented. Some general observations will bediscussed at the end of the paper. Unfortunately, this paper isby no means a complete list. Readers are encouraged toshare with the author any other experiences for futurepublication or any inaccuracies in the following. Some of themotor technical details came from the CPIA Rocket MotorManual.2 As a minimum, the following details on each motorare presented:1) Name of system.2) How instability manifested, i.e. DC-shifts, coupling withother motor components, structural failure, etc.3) Where in the program did it happen, i.e. development,deployed, aged, etc.4) Approximately when did it happen.5) What type of instability, i.e. tangential, longitudinal,linear, non-linear, etc.6) Propellant family, i.e. AP/HTPB, aluminized, doublebase, CMDB, etc.7) Fix, i.e. added stability additive, changed geometry,particle size changes, program cancelled, etc.Combustion Instability Event Descriptions1. SERGEANT Sounding Rocket.3 This polysulfide-AP(ammonium perchlorate) based rocket developed in the early50s was cancelled at the Jet Propulsion Laboratory due tocost overruns because of tangential mode instability. It had acast AP propellant with a star perforated charge. Theoscillation frequencies were probably around 1000-2500 Hz.It was at this time that the Jet Propulsion Laboratory changedfrom KP (potassium perchlorate) to AP and was before theuse of aluminum. During this same period, Aerojet washaving tangential combustion instability trouble with its"Aeroplex" propellant, AP-acrylate, with oscillatoryfrequencies between 2500 and 6000 Hz.2. RVA-10 Prototype.3 This star perforated motor wasbuilt to assist in the scale-up of cast propellant motors at theJet Propulsion Laboratory. It also used polysulfide-AP basedpropellant and experienced tangential mode instability. It hadother problem besides combustion instability. It is unknownif the instability was solved. The technology underdevelopment was for the Sergeant and Hermes motors in theearly 50s.3. SERGEANT Theater Ballistic Missile.3"6 Developedby the Jet Propulsion Laboratory for the Army in the 1953-62time period. Application of RVA-10 technology using a castStar perforated grain with polysulfide-AP (no aluminum)propellant. Tangential combustion instability on somesubscale preliminary motors produced roll torque so strongthat motor cases were scored on the outside due to rotationunder the chain restraints of the static test stand. Flight testsalso exhibited excessive roll torques due to combustioninstability. This and other problems delayed flightcertification so long it was obsolete by the time it went intoservice. The instability was never completely fixed. Thisprogram gave birth to the whole Brownlee era of combustioninstability pulsed motor research (Brownlee was a Canadianmilitary graduate student at the California Institute ofTechnology).7(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.4. SUBROC.3'8 During development of the Subroc motorin the late 1950s, preliminary tests were run on quarter andhalf-length motors to keep down costs on hardware testing.No combustion instability was encountered in these tests orthe first few tests on full-length motors. Then, very severeaxial mode instability occurred when the aluminum particlesize was increased to achieve a more desirable burning rate.This happened at a time when studies on aluminumcombustion's effect on combustion instability were beinginvestigated by Rohm and Haas.9 Also early studies onpulsed nonlinear instabilities were first being conducted,particularly by Brownlee and Price et al at NAWC, ChinaLake.7'10 The motor had a 120 inch long, six pointed stargrain 20.5 inches in diameter made with an aluminizedpolyurethane propellant. These multi-harmonic longitudinalinstabilities had significant 500+ psi DC pressure shifts andhad peak-to-peak oscillatory pressure amplitudes of over 400psi.Several small mechanical changes were made to themotor, but none proved to be successful in eliminating theinstability. A long list of possible fixes for Subroc wasproposed, but it was urged that a) the particle size of thealuminum be decreased, and b) that all sources of debris inthe motor be eliminated (to avoid pulsing by ejection ofdebris through the nozzle). The first response to therecommendation was to replace 50% of the aluminum withfiner aluminum. This seemed to fix the problem, until the35th static firing exhibited severe instability. After somediscussion, it was decided to go to 100% fine aluminum. Thefinal change increased the aluminum content from 13 to14.5%, 10% 10|im and 4.5% 35|im fine aluminum particles.After that change, no further encounters with instabilityoccurred. No mechanistic explanation was offered for theimproved behavior, but the increased particle damping waslikely the major factor in the improved behavior.This is a case where on going research provided thefix. Concurrently, Price and Waesche had been studying theeffect of aluminum particle size on pulsed instability.9'10 Thisresearch was sponsored by ARPA and the Navy StrategicPrograms Office as part of an effort to forestall unforeseenproblems in the new era long range rockets. The Subrocteam was under great pressure to have submarine basedservice-qualified weapons ready. The program delays ofthree or four months due to Cl problems increased costs, butthe missiles were ready on time.5. IROQUOIS.8 The Iroquois motor was developedaround 1960 with an aluminized propellant. It was about 87inches long and had a 7.3-inch diameter. The motor wassubjected to pulsing during development and sometimeswould go unstable with DC ballistic pressure shifts.Although that exact fix for the stability is somewhat unsure,the formulation was changed by increasing the aluminumcontent from 16 to 18% and the aluminum particle size wasreduced from 30 to 5 Jim.6. MK-27 TARTAR.8 The early 1960s Aerojetdeveloped ship launched anti-aircraft Tartar missile, Fig. 5,was a dual boost-sustain motor with a polyurethanealuminized AP boost propellant and a polyurethanenitroguanidine AP sustain propellant. The 12-inch diameter80 inch long motor was connected to a 20-inch long blasttube. Low level 270 Hz first mode longitudinal oscillationswere observed early in burn during the boost phase andoccurred much stronger near the end of the sustain phase,Fig. 6. These later oscillations were associated withsignificant DC pressure shifts. To reduce the amplitudes, theexhaust nozzle was moved from the upstream end to thedownstream end of the blast tube. Later an additionalimprovement program was conducted to enhance the long-term storage characteristics of this motor. The propellantremained unchanged, but the grain was altered to reduce themechanical stresses. The first three tests of the new geometryoscillated during the sustain phase and concurrently the meanpressure increased significantly. It was hypothesized that thismotor was being pulsed into instability by ejecta. Therefore,several changes were made to reduce the possibility ofejecting small pieces from inert motor components and toincrease the acoustic damping. Specifically, the propellantweb was increased in the head-end, the volume of the aftplenum was increased, the material in the aft collar waschanged and the exterior surface of the igniter was insulated.These changes eliminated the problem.Fig. 5. MK-27 Tarter Improved Tartar MotorBoost Phase Very Slight Oscillationsat 270 cpsDate 17 June 1960Temperature 85FSustainer Phase Free of Oscillations.Oscillations at ZJQ cpsFig. 6. MK-27 Tarter Pressure Trace7. TOW.8 This tactical design was done in 1964. Themotor was about 6 inches long and 5 inches in diameter. Thegrain was bonded to the head end and burning occurred onthe inside diameter, outside diameter and end face of thegrain. Although the pressure instrumentation available hadlimited frequency response, this motor experienced strong 1stand 2nd tangential mode instability with associated DCpressure shifts. Adding metal additives to suppress theinstability was not an option since the motor requirementswere to remain smokeless and flashless. Using baffles tobreak up the tangential oscillations eventually controlled theinstability. These baffles were placed around the outside ofthe grain between the grain and the case. Numerous designswere tried until a solution was obtained.(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.8. Improved GENIE.8 The mid 1960s the Aerojet Geniemotor was a 17-inch diameter motor containing apolyurethane AP aluminized propellant in an aft-end finocylgrain geometry for air-to-air applications. Pressureoscillations at around 450 Hz were encountered in theprogram as shown in Fig. 7. Reducing the aluminum particlesize in the booster propellant from 27 to 8 um on the theorythat particle damping would be increased eliminated theinstability. Although little evidence supported thismodification, the solution eliminated the oscillations.2000 -1000 -114psi Oscillationsat 450 cyclesTIMEFig. 7. Improved Genie Pressure Trace9. MINUTEMAN Wing I Stage III Missile.3 Inapproximately 1968, the US Air Force experienced five flightfailures in the Minute Man Wing I missiles, Figs. 8 and 9.The tests were on missiles returned from service for routineevaluation. This rate of failure raised concerns about thecondition of all the missiles in place in silos, and aninvestigation was started. The first objectives were todetermine what proportion of the missiles in the ICBM fleetwere defective and identify which ones. This led to amassive review of production documentation in search ofsingular data that correlated with the five third stage motorsthat had flown out of control. It was found that all fivemotors were from the last half of the production run. It wasalso noted that the powdered aluminum in the propellant wassupplied from a different plant (same company, samedesignation) during the second half of the production run.This led to suspicion that the "new" aluminum might not beequivalent to the original aluminum. The propellant had adouble base binder and contained both HMX and AP oxidizerwith around 17% aluminum.Fig 9. MinutemanMissileBy this point therewas also suspicion in theminds of some members ofthe investigating team thatthe failures might have beenlinked to combustioninstability, which canproduce vibration in themissile structure that impairguidance and controlsystems. This motor hadalways exhibited instabilityand appreciable pressureoscillations, but the staticfiring test data did not showmuch difference betweenmotors from the first andsecond half of the productionrun. Some question wasraised about the frequencyresponse of the data system,and the point was raised thatthe frequency response forthe telemetry data in flighttests was better than thatfrom static firings. Exami-nation of the flight datashowed that there was achange in the spectrum of pressure oscillations about halfwaythrough the production run. The most severe oscillationsnow occurred in a different acoustic mode of the motorcavity. This meant a different coupling between the pressurewaves and the missile structure and flight control system.Since the missile failures were loss of flight control, thechange in vibration frequency appeared to be a plausiblecause of the failures. Further, the members of the team thatwere specialists on combustion instability believed that achange in the source might be the cause of the change incombustion instability. Based on this, attention was focusedon those motors that were loaded with propellant containingthe "new" aluminum.At this point two efforts were mounted, one to find outhow the vibration environment caused malfunction of thecontrol system, and the other to find out why the "new"aluminum changed the combustion instability. The first issuewas particularly important because it offered a hope for aquick fix by modification of the control system ("quick", butvery likely requiring recall of half of the ICBMs formodification). Otherwise the motors would have to bereplaced or "regrained." This would require a combustioninstability study in order to assure that reworked or newmotors would be acceptable and establishment of aproduction line for the motors. The decision was to go forthe "quick fix" and also to initiate studies of the aluminumeffect on the combustion instability.It was quickly found that the "new" acousticmode frequency was resonant with a hydraulic linefrequency in the control system. This line contained adamper consisting of a spring-loaded piston. It wasfound that when oscillation in the line became tooFig. 8. Stage III Motor(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.large, the piston became unseated in its guide and stuckin that position, and the control system could not returnto neutral position. The problem was fixed byreplacing the piston by a longer one that would not beunseated. Although this was a quick fix solution to theproblem, it still required that half of the Minutemanforce be removed from their silos, shipped to themaintenance base, modified and returned to silos.By the time the control system problem had beenpinpointed, combustion photography had showed markeddifferences in the aluminum combustion of the original and"new" propellants. Previous studies at China Lake hadindicated that such differences were probably due todifferences in the amount or condition of the oxide coating onthe aluminum particles. Both the Air Force and Navycontinued to study the effect of oxide coatings on steady andunstable propellant combustion. The Air Force establishedan in-house program on combustion instability of solidrockets at it Rocket Propulsion Laboratory, and fundeddevelopment toward a computer program for predictingcombustion stability of proposed motor designs.1 U210. MANPADS Sustainer Motor.8 This motor wasdeveloped by Rohm and Haas in the late 1960s. Combustioninstability was encountered in a number of tests of theoriginal MANPADS motor during transition from high to lowpressure operation. This original motor contained a lowburning rate (0.24 in/sec at 1000 psi) 16% aluminized CTPBformulation and used the aft-end finocyl geometry shown.These oscillations reached 1800 psi peak-to-peak and had acharacteristic frequency of 700 Hz. A program to reduce themagnitude of these oscillations was then initiated using analternative propellant from another program. This newHTPB formulation contained 18% aluminum, a AP bimodaldistribution and increased solids content (88% vs. 86%), allof which were thought should reduce the oscillations. Severalmotor tests were conducted using the heavyweight geometry.This propellant exhibited even higher amplitude oscillationand produced a 100% over pressure during the boost phase.The 700 Hz oscillations also continued during the sustainphase. Thus, changing to this new propellant was a stepbackwards. To correct this problem, the propellant wasmodified again by reducing the aluminum content to 16%,converting to a trimodal AP distribution and returning to the86% solids loading. A total of five motors with these latestmodifications were tested at temperatures from -40F to +140F and no instabilities were observed during the boosterphase of the motor's operation. The exact mechanism for theimproved stability is still uncertain. Figs. 10 and 11 showsome details of the observed oscillations and a stable motorfiring.11. Advance Tactical Rocket (ATR).13 This LockheedPropulsion research and development motor experience DC-shifts in 1975 due to tangential mode combustion instability.It had a reduced smoke AP/HTPB propellant. Modifying thepropellant by increasing the AP particle size and adding acatalyst solved the problem. This kept the burning rateconstant. These changes lowered the pressure-coupledresponse function. This was confirmed by T-Burner responsetesting by Jim Crump at China Lake. The T-burner is adevice for measuring the pressure-coupled response of a solidpropellant using laboratory scale testing.Note: Peak-to-Peak Pressure is 1800 psi.Oscillations are shown for illustration andare not representative of the Frequency.Figure 10. Pressure History of Unstable MANPADS MotorFigure 11. Pressure History of Stable MANPAFS Motor12. Advanced Air Launched Motor (AALM).8 This 53inch long research and development motor with a 4 inchdiameter was developed by Thiokol in the mid 1970s toevaluate and demonstrate concepts intended to reducesignificantly the cost of solid propellant motors for air-to-ground applications. Fourteen motors were manufacturedwith the identical five slot forward cross section with an aftinternal cylinder all containing a reduced smoke AP/HTPBpropellant. The differences among these motors were in thecomposition of the propellants. The variations in propellantcomposition were concerned with oxidizer distribution andcontent, plasticizer, bond agent, combustion stabilizer contentand iron oxide content. Motors were also fired both hot andcold. Many of the motors were pulsed and often the secondpulse led to non-linear combustion instability. During thestudy, it was observed that higher concentrations of ZrCwould reduce the motors sensitivity to pulsing. These motorswere linearly stable.13. MK 12 NAVY STANDARD Missile Booster.8 Thismotor which was developed in the mid 1970s experiencedflight failures that were traced to guidance system relayssubjected to excessive vibration that were eventually tracedto pressure oscillations in the solid rocket. This system useda metallized AP containing propellant. The motor wasexperiencing tangential mode instability and also had thirdmode longitudinal oscillations. The fix was to merely replacethe relays with more robust units. No detailed stabilityanalyses were conducted nor were alternative propellant and(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.grain options considered. This program serves as anexcellent example of solving the problem at the system levelwhile living with the motor instability behavior.14. SLUFAE.14 This 5-inch diameter AP/HTPBdevelopment motor had a first mode longitudinal modeoscillation. To eliminate the oscillations, one percent 12 jimA12O3 was added to the formulation. This fix worked and thefirst mode longitudinal oscillations disappeared.Unfortunately, a 1500 Hz longitudinal oscillation thenappeared corresponding to a higher longitudinal mode. Arather unique fix was employed. A softball sized helmholtzresonator was added to the forward end of the motor that hadthe same frequency. The oscillations disappeared. The fixwas unique, but could only be accomplished since the motorhad available space for the resonator.15. MK-36 Reduced Smoke SIDEWINDER.8'13 The air-to-air Sidewinder missile is a 5 inch diameter motorcontaining a AP/HTPB propellant that is approximately 60inches long developed concurrently by Thiokol and Herculesbetween 1976 and 1977, Figs. 12 and 13. Surprisingdifficulties from nonlinear axial mode unstable combustionwere encountered during the conversion of the MK-36sidewinder motor from an aluminized to a reduced smokepropellant. These concerns were addressed duringdevelopment when motors experienced DC-shifts due tolongitudinal instabilities. Fig. 2 above looks very similar tothe kind of unstable pressure traces encounter in the MK-36development. The motor was required to be stable to pulsing.A number of alternative grain designs were considered.Initially, the keyslot geometry was the primary candidatedesign. However, when minor propellant changes were madeto increase the burning rate, this design was also found to beunstable when pulsed. The star grain was found to beunstable when pulsed and therefore dropped fromconsideration. The aft-fmocyl was considered in the analysisphase, but dropped because the stability of the forwardfinocyl design was predicted to be superior. The principalproblem with this aft-finocyl grain was the large plenum inthe aft-end and its effect on stability. These geometricstability trends were confirmed in more recent studies.15 Thefinal solution was to design a unique forward finocyl graingeometry that balanced fin and cylinder burning areas inaccordance with theory of linear velocity coupling as appliedin the SSP code, combined with ballistic analysis.11'12 Thistheory requires that the ratio of the burning area upstream ofthe oscillatory pressure node (point of zero pressureoscillation) to that towards the aft-end remains relativelyconstant and near unity throughout the motor operation. Ageometry with three unique cross sections; cylinder, 10 slotsand oval; was designed and tested. Here is an example wherestability analyses correctly assisted in designing a stablemotor. An alternative fix, and one that was eventuallyadopted, was to replace some of the AP with RDX to reducethe low frequency pressure-coupled and velocity-coupledresponse functions at constant burning rate. T-Burnerresponse tests confirmed the reduction in pressure-coupledresponse. It should also be noted that this propellant alsocontained one percent ZrC as a stability additive. Thesechanges led to stable combustion in 70 tests conducted duringdevelopment, preflight readiness tests and flight.Fig. 12. Sidewinder Fired from a F-14 TomcatFig. 13. Modern Sidewinder Missile16. Reduced Smoke AGM-65 MAVERICK.8'13 Thisdevelopment motor, Figs. 14 and 15, experienced DC-shiftsin 1977 due to tangential mode instabilities. Programs atboth Thiokol and Aerojet were undertaken to reduce andeliminate the problem. It had a reduced smoke AP/HTPBmFig. 14. Maverick MissileFig 15. Maverick Firing from a F-16(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.propellant. Lowering the pressure-coupled response functionby increasing the AP particle size, keeping the burning rateconstant by adding a catalyst and adding particulate additivessuch as ZrC and aluminum oxide solved this motorinstability. The lower pressure-coupled response wasconfirmed by T-Burner tests by Wil Andrepont at AFRPL.17. Low Cost Missile Motor (LCMM).13 This Thiokolresearch and development motor experienced DC-shifts in1978 due to longitudinal mode acoustic instability. It alsohad an AP/HTPB reduced smoke propellant. The solution tothe DC-shifts was to replace the nozzle throat with a low costeroding throat whose increase in throat area as the motorburned increased the nozzle damping as the web burned.This was combined with a grain design modification tomaintain the thrust-time trace. The solution was guided byuse of the Standard Stability Code (SSP) combined withballistic analysis.11'1218. Minimum Smoke Motor (MSM).13 This Thiokolresearch and development motor had a rough and rovingpressure trace in 1978 due to a tangential mode instability. Ithad a Nitramine CMDB (composite modified double base)propellant. This propellant had a lead salt catalyst thatdeposited on the surface of the propellant as the propellantregressed. A very high port flow speed disrupted thiscatalytic layer at the propellant surface causing the catalyzingeffect to vary unpredictably. The fix was to increase the portto throat area ratio at a slight penalty to total impulse.19. AGM-88 HARM.8 This Thiokol air-to-surface anti-radar missile motor was designed in 1976 and went intoproduction in 1980. The 10-inch diameter 80-inch longAP/HTPB grain geometry consists of forward cylindricalsection with two radial stress relief slots which transitions toa five-point star in the aft end. See Fig. 16. This high L/Dmotor uses a single low burning rate non-aluminizedpropellant in conjunction with the aft-star grain to produce aboost/sustain mode of operation. This motor had lowamplitude pressure oscillations during static test firings.Static test motors were instrumented with strain gages toassess the combustion stability behavior. The data indicatedfirst mode longitudinal instability that occurred three timesduring the transition from boost to sustain. Since little highermode frequencies were found the oscillations were assumedto be linear in nature. This is another case where theinstability did not affect the missile system and, therefore, nochanges were necessary. It was necessary to quantify theinstability.20. EX-70.13 This research and development motorexperienced excessive amplitude pressure oscillations in1979 due to longitudinal mode oscillations. Theseoscillations resulted in unacceptable thrust oscillations. Thissystem used a reduced smoke AP/HTPB propellant.Thiokol's fix was to use finer AP to increase nominal burningrate combined with increased nozzle throat size to maintainthe thrust time trace. Finer AP reduced the pressure-coupledand velocity-coupled response functions at the lowlongitudinal mode frequency. The increased nozzle throatsize had two stability purposes. First, it increased the portflow speed reducing the velocity-coupled response functionand, second, increased the throat size increasing nozzledamping. The solution was assisted by the use of the SSPcode combined with ballistic analysis.21. EX 104 NAVY STANDARD Missile.8 The EX 104Thiokol Rocket motor, Fig 17, was developed to replace theMK-56 Standard Missile motor in the mid 1980s. This motorused different AP/HTPB aluminized propellants for boostand sustain phases in a multiple star configuration. The twopropellants had widely different burning rates at 1000 psi, 1.7and 0.54 in/sec, for the boost and sustain propellants,respectively. This motor exhibited acoustic oscillations in thefirst, second and third axial modes throughout the motorsoperation. The largest oscillation amplitudes were on orderof 8 psi peak-to-peak. Response data was measured in theChina Lake T-Burner and stability analyses were performedon the motor. Since no components in the missile systemFig. 17. EX 104 Standard MissileFig. 16. HARM Air to Surface Missile(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.were affected and the motor consistently performed asdesigned, no changes were made to the motor and theinstability was left unchanged.22. Vertical Launch ASROC Booster Motor.8 ThisNavy/Thiokol deve-loped vertically launched anti-submarinemissile was developed for NAVSEA in 1985. The 14-inchdiameter 69 inch long motor contained a non-aluminizedAP/HTPB propellant. The motor exhibited oscillatorypressures during the development. Because combustionstability issues could have been a significant problem, mosttest motors were instrumented to detect oscillatory pressures.Specifically, signals from the chamber pressure transducerswere high pass filtered at 80 Hz to remove the mean pressurecomponent and then amplified to increase their sensitivity.Accelerometers were also used to provide additionalcombustion stability information. All the test motorsexhibited oscillatory pressures between 1.5 and 3.0 seconds.These oscillations had a characteristic frequency ofapproximately 270 to 275 Hz, which corresponds to the firstlongitudinal acoustic mode frequency. Fig. 18 shows thefrequency spectrum. However, all 15 test motors behaved aspredicted and designed. No deviations in chamber pressureresulted from these oscillations. Furthermore, the associatedvibration and thrust oscillations did not cause any failures insystem components. Thus, the program decided there was noreason to eliminate the oscillations. Fig. 19 shows ashipboard launch.0 100 200 300 400 500 600 700FREQUENCY, HzFig. VLA Frequency Spectrum23. SENTRY.16 This research and development ballisticmissile defense motor developed by Thiokol in the Mid1980s used a very high burning rate propellant. ThisAP/HTPB/AL propellant was catalyzed by Catacene toachieve rates of over 5 in/sec at 1000 psi. During testing,violent longitudinal combustion instability was encountered.Initial stability predictions indicated the motor would bestable due to the high aluminum content in the formulation.It was determined, although the conclusion was notunanimous, that distributed combustion of the burning metalparticles coupling with the motor's acoustics was the cause ofthe problem. The reasoning being that the high ratepropellant injected the aluminum particles into the motor atsuch a high rate of speed that the majority of the metalcombustion took place away from the propellant surface. Thisled to an extensive research program looking at high ratealuminized propellants. Response function testing wasperformed by China Lake at pressure with numerousFig. 19. Vertically Launched ASROCpropellant formulations including ones that systematicallyvaried the aluminum particle size. The conclusion reachedwas that smaller particles drove distribution more than largerparticles.17 In conjunction with this, theoretical modelingverified this conclusion and showed that the increase surfacearea of greater numbers smaller particles was moresignificant in acoustic driving than fewer large particles thatmay burn a longer time.17 The Sentry program wasterminated, in part due to combustion instability; however,much was learned about high burning rate aluminizedpropellant combustion stability.24. Low Contamination Propellant Motor (LCPM).13 Thisresearch and development motor experienced DC-shifts in1988 due to tangential mode instability. The purpose of thismotor was to minimize exhaust particulates. This motor alsoused an AP/HTPB propellant. The fix was to use coarser APwith a higher AP solids loading to reduce the high-frequencypressure-coupled response function at constant burning rate.This change achieved stability with only 0.5% ZrC, acommon stability additive. The motor achieved a factor ofthree reduction in exhaust particulates from prior productionpropellant. The solution was guided by SSP as applied tomotor instability data with no additive.25. SPACE SHUTTLE BOOSTER Motor.16'18'19 TheThiokol produced Shuttle main booster motor has a 12-feetdiameter and is 126-feet long containing over 1.1 millionpounds of AP/A1/PBAN propellant producing 2.6 millionspounds of thrust, Fig. 20. Combustion instability wasconsidered in the system from the very early days ofdevelopment. Measurements on the booster detected lowlevel first and second mode longitudinal instability on theorder of one to three psi peak-to-peak. One psi of pressureoscillations equals 33,000 Ibs of thrust oscillations. No DCpressure shifts were encountered and system impacts wereminimal. These oscillations are flow driven by vortexesgenerated due the segmented nature of the booster design.After the Challenger disaster in which an O-ring failed in the(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.Fig. 20. Space Shuttle Launchjoint between segments the joint design was modified. Theredesigned segments were much stiffer. The effect oninstability was to slightly increase the oscillatory levels. Twocomments about the shuttle instability, first, the first modefrequency is only 14 Hz. It turns out that this is a frequencythat is very close to the human brains reaction time toexternal stimuli. There was some worry by NASA that thebooster oscillations might interfere with the astronaut'sabilities. Fortunately, this was not the case, althoughastronautics are warned and feel the vibrations during boosteroperation. Second, the oscillations are carefully monitoredand using statistical techniques, the worst-case scenario isdetermined. This, in turn, is used to predict the vibrationalloads on the shuttle and it's payloads.26. Development Booster Motor, DBM.8'16 Both linearand nonlinear combustion instabilities were observed duringthe initial development tests of a development booster motordeveloped by United Technologies Chemical Systems(UTCS) in 1994. These high L/D development motorscontained a low burning rate aluminized HTPB compositepropellant and an aft-end fmocyl grain geometry to minimizeerosive burning effects. Head-end Kistler pressuremeasurements revealed low amplitude oscillations started atabout 10% of Web Action Time (WAT) and lasted to about50% of WAT. Waterfall frequency spectra from this testshowed the oscillatory energy was concentrated in the firstlongitudinal mode. There was very little energy in the higherharmonic frequencies. This linear instability occurredconsistently in each of the development tests, but was judgedto have no negative impact on the vehicle performance. Thus,the missile system could "live" with the vibration and thrustloads resulting from this instability.During three motor tests, however, step increases inthe mean chamber pressure of the type shown in Fig. 21 wereobserved. Fig. 22 shows a waterfall plot from one of thesetests. Note the high harmonic content begins simultaneouslywith the pressure step. The frequencies of these higherharmonic components are integer multiples of thefundamental frequency and do not match the higher acousticmode frequencies of the motor chamber. It should be notedthat high-speed movies from these tests were all examined ingreat detail. Pulsing produced by inert components beingejected through the nozzle was observed in one test, but couldnot be identified in the other two tests. It should also be notedthat propellant formulation changes to enhance themechanical properties were being considered at this sametime.(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)1 Sponsoring Organization.Fig. 21. DBM Unstable TraceFig. 22. Waterfall Plot from Unstable TestThis behavior was judged to have a significantnegative impact on the vehicle performance and studies wereinitiated to eliminate this behavior. Experts were called inboth within the UTCS, government agencies, academia,private consultants and retired engineers. It should be notedthat this approach was extremely effective and UTCS shouldbe applauded for this. Multiple meetings were held and theentire team did considerable work between meetings. Thisincluded numerical analysis of the flow field, stabilityanalysis, propellant response function measurements onnumerous propellants by both China Lake and UTCS andconsiderable grain design analyses. Altering the graingeometry was the primary approach. In particular, linearstability predictions suggested that changing the grain fromthe aft-finocyl to a radial slotted configuration could increasethe nozzle damping significantly. The basic hypothesis wasthat increasing the linear damping would also enhance thenonlinear stability and therefore make the motor lesssusceptible to step pressure increases. Additionally, thischange would decrease the limiting amplitude of the linearoscillations and the resulting vibration and thrust oscillationloads. Subsequent motor tests validated these changes.Pulsing the radial slot motor at the time at which themaximum oscillation occurred and near burnout resulted inexponentially decaying oscillations. Both test motorsperformed as designed. Based on these results, andsubsequent motor tests results, this problem was fixed anddevelopment proceeded.27. MARS PATHFINDER DECENT Motors.16 The MarsPathfinder program used three rocket motors manufacturedby Thiokol to stop the descent vehicle just before the air bagencased lander was dropped on the surface of the planet.These motors were originally AP/HTPB motors with twopercent aluminum and were about 30 inches long, Fig. 23.Three motors are fired all at one time as shown in Fig. 24during a suspended test at China Lake. During an early droptest of the vehicle assembly from a helicopter, longitudinalinstability was detected in all three motors. However, no DCshifts were observed and the motors operated properly. Inthis program no chances were to be taken and the aluminumcontent was increased from two to 16 percent. Subsequenttesting of the vehicle at China Lake with highly instrumentedFig. 23. Mars Pathfinder Descent MotorFig. 24. Mars Pathfinder Descent Vehicle Test Firing10(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.motors showed no instabilities, Fig. 24. Because of the tightschedule and budget of the program, detailed investigationswere not performed. The cause of the instabilities was neverfully understood, but the added aluminum solved theproblem.28. NWR Research Motors.1'15'16 In this study, 33 5-inchdiameter 60-inch long motors containing reduced smokeAP/HTPB propellants were cast and fired at China Lakebetween 1985 and 1998. The majority of these motors werepulsed several times during burn. These were instrumentedwith redundant high frequency gages. Various stabilityadditives were also used. These motors were easily pulsedinto violent non-linear longitudinal instability. See Figs. 2, 3and 4 at the beginning of this paper. Certain geometries, fullcylinder and full star, were more easily pulsed. Motors withstability additives were much more stable to pulsing. Motorswere more likely to go unstable later in burn. As pressurewas increased, tangential mode instability and associated DCpressure shifts were observed. It was also easier to pulse amotor into non-linear instability as pressure was increased.References 1 and 10 have detailed accounts of the study.TrendsTable 1 shows a summary of the motors discussedabove. This table includes a few details of each of the motorsabove including name, approximate date, type of motor andpropellant. The type of instability is also presented and it isimportant to note than often more than one type of instabilitywas observed. Those types are longitudinal, tangential, wereDC pressure shift observed and/or was the instability due tosome sort of motor pulsing. Finally, what was done to fix themotor? Was the propellant modified, metal content or sizeadjusted and/or was the geometry adjusted. As with the typeof instability, often multiple fixes were used. Under details isa brief description of what was done. For a completedescription, the text in the body of this report must beexamined for a particular motor.From the table it is difficult to see absolute trends inthe data. Both tangential and longitudinal instabilitiesoccurred with all propellant families, although tangentialinstabilities were rare with aluminized propellants. Solutionsto fix the problems varied greatly and were no doubtdependent upon the constraints imposed in the engineers incharge of fixing the problems. The application of the motorappears to have no bearing on the type and solution of theinstability. There were about equal numbers of fixes basedon propellant changes, metal modifications and geometry.Often a combination was used. Two solutions were verycommon to mitigate the combustion instability. The first wasto increase in the fine fraction particle size of AP in thepropellant formulation. This was often complimented byusing a burning rate catalyst to maintain the propellantburning rate. The second was to reduce the size of thealuminum particles.ConclusionThe above 28 motors is by no means a complete list ofmotors that have experienced combustion instabilityproblems. Other motors that were known to have combustioninstability technical issues during development deploymentinclude and:Polaris (A-3)Poseidon (C-3)Trident I (C-4)Trident II (D-5)Shuttle Igniter MotorShuttle Separation MotorStandard Missile (Mk-56)Aegis ER BoosterTomahawk BoosterHarpoon Booster (MK-22)AMRAAM BoosterAAAMBomarcSealanceAgileSparrow (AIM-7F)ShrikePenguinChaparral (MK-50)2.75 Smokeless (MK-66)ZuniCondorASW/SOWMEW Ejection SeatArrowSea SparrowThis paper is an ongoing exercise in documentingmotors that experienced combustion instability issues andwhat was done to solve the problems. Engineers andscientists reading this papers are urged to inform the authorof other "motor stories" and also point out any inaccuraciesin the description above. It is hoped that documenting thesemotors will enable future motor designer to, not only solveproblems as they do arise with greater confidence, but also toavoid combustion instability in the first place.AcknowledgementsThe author wishes to thank Norm Cohen of CohenProfessional Associates, Redlands, CA; Professor EdwardPrice of the Georgia Institute of Technology; James Crump ofthe Naval Air Warfare Center, China Lake (retired); Dr.Robert Brown of Brown Associates, retired from UnitedTechnologies Chemical Systems for their many inputs to thislist. Particular thanks to Dr. Robert Brown and his Reference8. Inputs for about half the motors discussed came from thisreference. The document is well documented with manyreferences to the source material.References1. F.S. Blomshield, "Pulsed Motor Firing," NAWCWDTP 8444, Naval Air Warfare Center Weapons Division,China Lake, California, August 2000.2. "CPIA/M1 Rocket Motor Manual," Volumes I and II,Chemical Propulsion Information Agency, July 1994.3. Personal Communications with Professor EdwardPrice, Georgia Institute of Technology, Atlanta, Georgia.4. Personal Communications with Professor GaryFlandro, University of Tennessee Space Institute, Tullahoma,Tennessee.5. G.A. Flandro, "Roll Torque and Normal ForceGeneration in Acoustically Unstable Rocket Motors," AIAAJournal, Vol. 2, No. 7, pp. 1303-1306, July 1964.6. E.M. Landsbaum and F.W. Spaid, "ExperimentalStudies of Unstable Combustion in Solid Propellant RocketMotors," TR 32-146, Jet Propulsion Laboratory, pp 13-14,August 1961.11Table 1. Summary of MotorsMotor DetailsNo.12345678910111213141516171819202122232425262728NameSergeantRVA-10SergeantSubrocIroquoisTartarTowGenieMinutemanManpadsATRAALMMK-12SlufaeSidewinderMaverickLCMMMSMHarmEX-70EX- 104ASROCSentryLCPMSHUTTLEDBMPathfinderNWRDate1951195119571961196019611964196519681969197519751975197519771977197819781978197919851985198519881990199419961997App.SoundingTBMTBMSAS7SAASTSATABALSLATARESSAASTSATAASRESRESASSAASAASASTBM???BOSSLSPARESPropel lant TypePolysulfide/APPolysulfide/APPolysulfide/APAl/PolyurethaneAluminizedDuel GrainDouble BaseAP/Al/PolyurethaneDouble Base/AP/AlAP/A1/HTPBAP/HTPBAP/HTPBAP/A1/HTPBAP/HTPBAP/HTPBAP/HTPBAP/HTPBDouble Base/CMDBAP/HTPBAP/HTPBDuel AP/A1/HTPBAP/HTPBAP/A1/HTPBAP/HTPBAP/A1/PBANAP/A1/HTPBAP/A1/HTPBAP/HTPBTypeLongXXXXXXXXXXXXXXXXXXXTangXXXXXXXXXXXDC shiftsXXXXXXXXXXXXXPulsedXXXXFixPropXXXXXXXXXMetalXXXXXXGeo.XXXXXXXXXXDetailsNot fixedUnknownNot FixedReduced Al sizeIncreased Al%, reduced SizeNozzle moved downstreamAdded BafflesReduced Al sizeChanged system, not motorLowered Solid LoadingIncreased AP size, added catalystZrC containing motors were betterChanged system, not motorAdded helmholtz resonatorGrain design change, some RDX for APIncreased AP size, added catalyst, ZrCAdd eroding Nozzle, changed geometryIncreased port areaNo changes requiredSmaller AP, increase nozzle sizeNo changes requiredNo changes requiredProgram ended, smaller AlSmaller AP, higher loadingNo changed requiredGrain design changeIncreased Al from 2% to 16%Stability additives, geometry, pressure8CD8c?5-88OTlTBM - Theater BallisticATA Air-to-AirMissile SAS - Surface Anti-Submarine SAA - Ship Launched Anti-AircraftBAL - Ballistic Missile SL - Surface Launched AS - Air to Surface RESSTS - Surface to Surface SPA - Space Motor- Research Motor BOS - Space BoosterQ.O">5O^^COoo(QO(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.7. W. G. Brownlee. "Nonlinear Axial CombustionInstability in Solid Propellant Motors," AIAA Journal, Vol.2, pp. 275-284, February 1964.8. R.S. Brown, "Combustion Stability Of InterceptorRocket Motors: A Practical Approach To ManagingInstability Problems," CPTR 95-57, Chemical PropulsionInformation Agency, June 1995.9. R. H. Waesche, "Mechanisms And Methods OfSuppression Of Oscillatory Burning By Metallic Additives,"35th JANNAF Combustion Meeting, CPIA Pub. 680, Vol. I,pp. 605-615, December 1998.10. E.W. Price, "Status of Solid Rocket CombustionInstability Research," NOTS TP 4275, U.S. Naval OrdnanceTest Station, China Lake, CA, February 1967.11. "The Solid Propellant Rocket Motor PerformanceComputer Program (SPP), Version 6.0, Volume V: User'sManual for the SPP Grain Design and Ballistic Module,"AFAL-TR-87-078, Air Force Astronautics Laboratory,Edwards Air Force Base, California, December 1987.12. "The Solid Propellant Rocket Motor PerformanceComputer Program (SPP), Version 6.0, Volume VI:Standard Stability Prediction (SSP) Method for Solid RocketMotors," AFAL-TR-87-078, Air Force AstronauticsLaboratory, Edwards Air Force Base, California, December1987.13. Personal Communications with Norman Cohen, CohenProfessional Services, Redlands, California.14. Personal Communications with James E. Crump,Retired, Naval Air Warfare Center Weapons Division, ChinaLake, California.15. F.S. Blomshield, J.E. Crump, H.B. Mathes and M.W.Beckstead, "Stability Testing and Pulsing of Full ScaleTactical Motors," NAWCWPNS TP 8060, Naval Air WarfareCenter Weapons Division, China Lake, California, February1996.16. Personal Recollections of the author, Fred S.Blomshield, Naval Air Warfare Center Weapons Division,China Lake, California.17. F.S. Blomshield, K.J. Kraeutle, R.A. Stalnaker, M.W.Beckstead and B. Stokes, "Aluminum Combustion Effects onCombustion Instability of High Burn Rate Propellants," 28thJANNAF Combustion Meeting, CPIA Publication No. 573,Vol. Ill, pp. 419-438, Brooks Air Force Base, San Antonio,Texas, November 1991.18. F.S. Blomshield and H.B. Mathes, "PressureOscillations in Post-Challenger Space Shuttle RedesignedSolid Rocket Motors," AIAA Journal of Propulsion andPower, Vol. 9, No. 2. pp. 217-221, March-April 1993.19. F.S. Blomshield and C.J. Bicker, "PressureOscillations in Shuttle Solid Rocket Motors," 1997 AmericanInstitute of Aeronautics and Astronautics Joint PropulsionMeeting, Paper No. AIAA-97-3252, Seattle, Washington,July 1997.13cover: cover_a: 1: 4: 6a: 6b: 6c: 6d: 7a: 7b: 8: 9: 10a: 10b:


View more >