[American Institute of Aeronautics and Astronautics 13th Propulsion Conference - Orlando,FL,U.S.A. (11 July 1977 - 13 July 1977)] 13th Propulsion Conference - Theory and intuition in hydrazine reactor technology

W 17-845 Theory and Intuition in Hydrazine Reactor L Technology A.S. Kesten, United Technologies Research Lb) Center, East Hartford, Conn. .r -6- 4 thl PROPULSION Orlando, FloriddJuly 11.13,1977 4 For penlarlon to copy or npubllah, contact the Amwlun tnstltula of Aamnaullcs and A8IIDn1UlICI. 12W Arrnuo of lhr Arnrrlur, Now Yo*, N.Y. 10019. J' ./3.' L ZKECBY AK3 IIVJITION IN €IIDFL&ZII?E REACTOR TECFTOIKGY A. 8. Kcsten* United Technologies Research Center East Hactford, Connecticut Abstract Considerable e f for t has been expended i n recent years t o develop methods for predicting the steady s t a t e and t ransient performance of Qrdrazine catalyt ic reactors. These reactor models have been tested i n a nunber of laborator- i e s i n order t o validate and extend prediction capabili t ies. These studies i n turn have led t o an understandiw of phenomena involving catalyst degradation as w e l l as catalyst breahp. phenomnological insight has pemi t ted refinement of monopropellant t m s t e r s and the growth of This hydrazine. design was taken at UTRC (then UARL) i n the devel- opment of steady s t a t e and transient mdels of the behavior of the hydrazine system. These studies included the development of computer pronams t o calculate the temperature and reactant concentra- t ion distributions as llmctions of time and axlal posit ion i n typical reaction chamber configurations. The computer programs evolved were based upon mdels of the reactor system which consider both thermal a n d catalf i ic decomposition of reactants along with simrltaneats heat and mass transfer between the f ree gas phase and the gas within the A fundamental Wroach t o reactor mnoprogellant intuit ion. the evolution of hvdrazine reactor theom. To This paper w i l l review pores of the catalyst beds. i l l u s t r a t e the present leve l of intuit ion, fa attempt will be rmde t o w e s t how a sinple in- s i t u test t o assess mnapropeUant thnis ter state- of-health might be developed. I. Introduction With the development af Shel l 405 catalyst , Wdrazine monopropellant technology became important for thrusters and ~ S S generators. understand% of how these ca ta ly t ic reactors worked was clearly desirable for (a) effective design, (b) definit ion of reactor l imitat ions and (c) extending the state-of-the-art i n terms of performsnce and life. I n addition it is reasonable t o expect tha t knowledge of how they work d d lead t o t e s t s (hoperully simple) of state-of-health i n order t o elucidate how mch l i f e renains af'ter extended operation. For this kind of understanding t o come about requires extensive i te ra t ion between reactor operation, modeling and labmatory experi- ments. We can t race the h is tor ica l development Of monopropellant "intuition" through the evolution of some of the fundamental s tudies of hydrazine reactors and t h e i r interact ion with thruster and gas generator performsnce r e m l t s . We can a lso examine the current s t a t e of Intui t ion and evaluate the potent ia l for developing a .*le t e s t t o monitor the reactor aging process. &I n. Historical Perspective The first attempts at developing an under- standing of hydrazine reactors came with a Rocket Research study t o establish thruster design criter- ia for Shell 405 catalyst . B e basic design con- cept t h a t emerged featured M upper bed of f ine mesh catalyst t o provide stable decomposition of W r a z i n e and a downstreambed of ccurser catalyst t o promte the decomposition of residual vapor *Mnnager, Combustion Sciences AIAA Associate Fellow 1 ADelysie of Steady-State (beration The W y s i s of a hydrazine engine reaction system pertains t o a reaction chamber of arbi t rary cross s e c t i m packed with catalyst par t ic les into which l iquid hydrazine is injected at a rb i t ra r i ly selected axial locations. hydrazine injection was taken as uniform across t h e cross section of the chamber. Catalyst par- t i c l e s were represented as "equivalent" spheres with a diameter taken as a function of the par t ic le s ize and shape. Both thermal and catalyt ic vapor phase decomposition of hydrazine and m n i a were considered i n developing equations describing the concentration distributions of these reactants. DiI%sion of reactants from the h.ee-gas phase t o the outside rnrrface of the catalyst p e l l e t s was taken in to account. Since the catalyst nvrterikl i s impregnated on the in te r ior and exterior rnrr- faces of porous par t ic les , the diffusion of reac- t a n t s i n t o the porous structure nust a lso be con- sidered. I n addition, the conduction of heat within the porats par t ic les nust be taken in to accwnt since the decomposition reactions are accompanied by the evolution or absorption of heat. A t these locations, A ser ies of calculations was made using the steady-state m d e l * for cases far which experi- mental information was available examine the effectiveness of the mdel and t o establish the r a t e constants associated with the catalyt ic decomposition reactions. these calculations are presented Ref. 3. ture distributions are plotted i n Fig. 1 for one of these cases i n which the catalyst bed packing was taken t o consist of 25-30 mesh catalyst par- t i c l e s for the first 0.2 in. and 1/8 in. x 1/8 in. cylindrical pe l le t s for the rmainder of the bed. Also shown i n Fig. 1 arc temperature measurements obtained by Rocket Research Corporatlon the ccurse of engine firings under the same opera- i n order t o The resu l t s of Tempera- during L tfri conditions. The calculated nole-fraction Fra:iles associated with this case, together with eGeri tental values of Xmle-fraCtions, are shown 1n rig. 2. .- -7. Axial Distance, 2 - Ft F i v e i. . Steady-State Axial Tenperature Profile. P'h77.5 psia. e 3 . E lb/&-sec. X F Y C 2. hact ions of Reactants. P479.5 psia. W3.12 Ib/fi*-sec. Steady-state mal Prot i les of &le- A 8tu* t o assess the effects on the steady- **?e behavior of the system Of honunifonn rad ia l ::;e?tion and of cats lyst bed configurations ex- b.!S!t?w both radial and axial nonuniformities I S E c a c r n d i n Ref. 4. Radial variations i n mass T.m rate or bed packing cauae r a d i a l temperature an: concentration gradients tha t lead t o turbulent --..xion of heat and fads i n the reactor wstem. * CcWJter program was developed t o calculate *..e t a q e r a t u r e and reectant concentration distributions as functions oi SJdnl and radial position i n typi- c a l hydrazine reaction chambers. baaed upon a m o d e l Of t h e reactor system tha t in- cludes treatment of the turbulent diffusion of heat and mss in the iree-gas phase along with heat and mass diffusion within the catalyst par t ic les and between the par t ic les and the free-gas phase. It was shown t h a t a one-dimensional model of the sys- tem based on parameters averaged over the reactor cross section, is not adequate t o describe the behavior of a reactor which h i b i t s significant rad ia l variations i n injection prof i le or bed con- figuration. For these systems, it is necessary t o use t h e two-dimensional model t o effectively pre- dict , tor -le, reactor locations where hydrazine diffusion fromlow- t o high-temperature regions results i n unusually high tenqeratures a t the inter- face between the regions. The program is The effects of the Binarltaneous turbulent diffusion of heat and mss on temperature and reactant concentration prof i les m e indicated i n Fig. 3 for a case i n wbich hydrazine injection is uniform across the i n l e t face of the reactor but additional hydrazine €5 introduced i n t o the reactor through injectors inbedded i n the catalyst bed. I n this case the buried injectors were taken t o dis t r ibute hydrazine uniformly for Ogr/RS0.7 over the flrst 1/2 in. of the reactor. tenperatures are plotted a8 a function of axlal posit ion a t various r a d i a l posit ions i n Fig. 38. Rydrazine d i iKs ion fromlow- t o high-tenperature regions resu l t s i n unusually high temperatures at the interface between the high and low flow r a t e regions. "thermal sheath" which is rare clear ly i l lus t ra ted in Fig. 3b, which is a cross plot of the resu l t s presented i n Fig. 38 . Here temperature i s plotted as a function of radial position at various axlal locations i n the reactor. For camparison purposes, the radial tenperatwe profFle at t h e exi t of a 3-In. bed with a s t e p - a c t i o n (aU i n l e t ) i a e c t i o n profile is a lso plotted i n Fig. 3b. Tha calculated This resu l t s i n the fonmtion of a . e pndysis of Transient Cmeration The anelysis of the steady-state p e r f o r m c e of a catalyt ic reactor based on a model which considers fi lm and pore diffision of heat and mass within porous catalyst par t ic les can be carried aut i n a s t r a i g h t - f o m d manner. However, the analysis of the transiwt behavior of the system using t h i s general mdel is quite complicated since both the reactant concentrations and t h e r a t e constants for decomposition, Which are expo- nentially dependent on temperature, w i l l be func- t ions not only of time, but also of posit ion within the catalyst particles. To circurmrent these complications, the trans- i e n t behavior of the System was analyzed by ~ 1 0 ~ - ing a second model 5 which takes heat and mass aifmsion within the catalyst par t ic lea in to account by considering the i r effects on the r a t e s of the catalyt ic reactions through the "ut i l izat ion factor." r a t i o of actual reaction rate t o t h a t which d d be produced by the Catalyst par t ic le if negligible temperature and concentration gradients exlsted within the par t ic le , can b e calculated as a i\urc- t ion of temperature and reactant concentration a t the outside mrface of a catalyst par t ic le a t steady-state. transient conditions, u t i l i za t ion factors a r e functions only of surface temperatures and concen- trations, the analysis of the t ransient behavior of the reactor Bystem becomes tractable. t h i s method then, the t-erature and keactant concentrations were described as functions of t i m e and position in the reaction chamber. It was assumed tha t during reactor operation l iquid velo- c i t i e s are sufficiently low relative t o other r a t e processes so that, for all pract ical purposes, steady-state in the l iquid and liquid-vapor regions is achieved as soon as the l iquid reaches a given ahlal location in the reactor. No consideration is given t o regression of the liquid-vapor l n t u - face as the chamber pressure builds up since the overall length of the l iquid region in typical reactors is small coqared t o the length of the vapor region. When flow in to the reactor is stoppe4 it is easily shown tha t t h e residual l iquid hydrazine In the reactor vaporizes in just a few lnFlliseconds due t o the very rapid deconposition of hydrazine i n the l iquid region. during reactor shutdown, the l iquid reelons play a slrall role i n determining the transient behavior of the reactor sgstem. This factor, which is defined a8 the If it is assumed that , even under Using Therefore, A ser ies of calculations of the t ransient behavior of a typical continuous n o w reactor for which experimental Wormation was available 6 was s d e in order t o emunine the effectiveness o f t h e transient model. The calfulations pertaln t o a 50 lbf nominal thmst hydrazine reactor 2.4 in. in dismeter into which l iquid hydrazine l a injected a t the upstream a d of the reactor only. catslyst bed packing was taken t o consist of 25-30 The mesh catalyst par t ic lea for the first 0.25 in. and l/8 in. I l/8 in. cyllndrical pellets for the remainder of t h e bed. The steady-state c h d v presme was taken as 200 psia, the steady-state msss flow r a t e as 6.5 lb/f$-sec, and the i n i t i a l chanllrer pressure as 14.7 psfa. these calculations are shown in Figs. 4 through 6. Gas temperatures are plotted as function of time at each of fmr axial posit ions in Fig. 4 for a case in which the initial bed t e q v a t u r e was taken as 530 deg R. prof i les 6 are a l s o s h m i n Fie. 4 for purposes of cowarison. Generally good agreement between theoret ical and experimental r e s u l t s may be noted, par t icular ly duriFg the early stages of the trans- ient. calculated ra tes of response are i n par t due t o the themcouple response time, the use of steady- state u t i l i za t ion factors t o describe heat and s s s d i m s i o n within catalyst par t ic les under transient conditions resu l t s in calculated response r a t e s which m e a l i t t l e tw high. male-fraction prof i les for hydrazine and m n i a , corresponding t o the temperature prof i les shown in Fig. 4, are i l lus t ra ted i n Fig. 5, and the corresponding mole-fraction proflles for nitrogen and hydrogen are i l lus t ra ted in Fig. 6. Here, t h e =le-fractions are plotted as functions of ahlal position at various tlmes. The resu l t s of kasured gas temperature While the differences between meamred and The calculated (AI AXIAL POSITION. 0.08 FT 2100 1700 1300 900 I I ' . 1 2 TIME. t - SEC 3 c c TEMPERATURE IN INTCRSTITtAl PHASE. I, - DE6 R TEMPERATURE IN INTERSTITIAL PHASE. T. - DEG R . I Y 0 a u 0 0 ?2 0 0 0 Q Y P TEMPERATURE IN INTERSTITIAL M I S E , 1, - DEG R Y u 0 0 0 0 0 E 0 0 s 0 9 x r * * a Y? 0 z . P u u, 2 3 . . Figure 6. Fractions of Nitrogen and Hydrogen. Bed Teqxrature=530 Deg R, Steay-State Chamber Pressure-m psia, Steady-State Wss Flow Rate- 6.5 lb/ftz-sec. Transient Axial Profi les Of &le- Initial Additional calculations were made for the reactor configuration noted above i n order t o exandne the effect of pulsed Rw on i n i t i a l t ransient reQonse. The cdculated resu l t s i l lus - t ra ted in Figs. 7 through 13 refer t o a r e a c t o r operated under pulsed now conditions at a eteady- s t a t e chembw pressure of 260 psia, a steady-state mass flow r a t e of 5.8 Ib/ftz-sec, M i n i t i a l c h d e r pressure of 14.7 psia, and an initial bed temperature of 530 deg R. Calculations were made f o r the f i rs t two pulses of a duty cycle consls- t h g of alternate on and off times of 50 maec and 100 m8ec respectively. . The temperature i n the i n t e r s t i t i a l phase l a plotted i n Pig. 7 as a function of time a t varlous a d a l locations In the reaction chambef. m e temperature rises rapidly after reactor start- up, pmticularly i n the upstream portion of the c h d e r . off, the gas temperature in the upstream 6ectFon of the reactor rises extremely rapidly at first because of thermal deconposition of residual hydrazine in t h i s region. reactor downstream of the smell catalyst particles, where the gas temperature I s too low m e r 50 msec t o Pumi t significant t h e n a l decomposition of &&azine, the temperature falls due t o heat trims- fer *om the gas t o t h e m a m catalyst pellets - and t o the chamber w a l l s . eattslyst p e l l e t s i n these regions resu l t s in a very amall temperature change because ol the large =as Of the particles. lhis is i l lust rated i n 8 where catalyst particle tewerature is Plotted as a function of t i m e at the same Ulal locations chosen for Fig. 7. When flow into the reactor i s turned In the regions of the The heat gained by the Figare 7. With Time at Various Axial Positions. State chamber Fressur€=260 psia, Steady-State W s s Flow Rate=T.8 Ib/ft2-sec. Variation of I n t e r s t i t i a l Temperature Steady- .m.. .*Yl_ . / L.I... m r Figure 8. Variation of Catalyst Particle Tempers- h u e v i t h Time at VarioUll kdal Positions. s t a t e chamber Fressurs-260 psia, steady-State bass plow Rate=5.8 Ib/it2-.c~. Steady- 5 m e rapid presmre buildup and decay resul t ing &se operation of the reactor for this case :I i n Fig. 9. The species d e - f r a c t i o n Irc.cue associated with t h i s presrmre variation ad the temperature distributions m u s t r a t e d i n rjGS. 7 and 8 are shown i n Figs. 10 through 13. me variation of nole-f2action of hydrazine with t?m at vmious adal locations i a plat ted in F : ~ . lo. This p lo t i l lus t ra ted the very rapid t h c r r n l deconpositlon Of residual hydrazine i n the hotter upstream regions of the reaction chm- b r r during reactor shutdown as Well as the some- vmt slower catalyt ic decomposition i n the cooler dmstrcam regions of the reactor. The variation of mole-fraction of rtpeaOnia with t i m e is illus- trated ut various axial locations in Fig. L1. Followin& reactor shutdown the residual ammonia near the the upstream end of the reactor decomposes r a t a l j i c a l l y i n the hot catalyst par t ic les , the cooler downstream regions, mmnia is dis- rlnced gradual ly by the nitrogen and hydrogen formed from hydrazine Bnd aamwnia decomposition upstream. These decomposition products flow downstream during s h u t d m as the chamber presrmre decays. These processes lead t o the ammonia mole- fraction prof i les shown l a Fig. 11 and the mle- frnction prof i les of nitrogen and hydrogen shown In Figs. 12 and 13 respectively. I n . I F w. sl; I 10 U 3 ; r E u - a 1 LI Figure 9. Variation of Chdu Pressure with TIM. Steady-State Chmbcr Prcssure=260 psia, Steady- State h s s Flow Rate5.8 lb/f+?-sec. I K I I,.c,”,. I “.ClO.”I 1.1. i 0.2 I I I 0 40 D I,< DMo )m 1 0 I I . . U L l t l l C O " 0 , Figure 13. Variation of Wle-Fraction of Hydrogen v l t h Time at Various Mal Positions. S ta te Chamber Fressure=%O psia, Steady-State Mass Elm Rate=5.8 lb/ft2-sec. steady- The effect of pulse duty cycle on the transient response of t h e erit gas t e q e r a t u r e is i l lus t ra ted i n Fig. 14. 23 lbf nominal thrust engine 1.4 in. i n dlamcter v l t h a packed length of 1.2 In. i n t o which liquid hydrazine l a injected at a temperature of 5 p deg R. A case was chosen i n which hyeaelne in- ject lon was taken at the reactor I n l e t only and i n which t h e steady-state mss flow r a t e was t u e n as 5.76 Ib/fte-sec (0.04 lb/inz-sec), the injector pressure as 150 psia, the i n i t i d chamber pressure a8 14.7 psia, t h e i n i t i a l bed temperature as 530 deg R, and the catalyst bed configuration a8 0.2 in. of 25-30 mesh catalyst par t ic les followed by 1.0 in. of 14-18 mesh catalyst particles. For 8 pulse duty cycle of 60 msec on and 60 mcc off, t rans€ent I n t e r s t i t l a l temperature prof i les are plot ted in Fig. 15 for tn, Mla l positions. one at the end of t h e bed and one at spproldrcatrly t h e mid-point of the bed. Ala0 plotted i n Fig. 1 5 m e the teuperature prof i les coqmted 8t these same points for the reactor operatingunder condl- *io118 of contlmous rather than pulsed Row. t ransient p a r t i c l e temperature proflles associated. With the i n t e r s t i t i a l temperatures S h m i n Fig. 15 are plotted in Fig. 16. The calculated resu l t s re fer t o a '- m e Figure 14. Temperature Prof i les f o r Various Pulse k t y Cycles. Injector Bessure=lW psie. Steady-State Mss Flow Rate=5.76 lb/itz-sec. Comparison of "ransient I n t e r s t i t i a l Figure 15. Transient In t e r s t i t i a l Teqerature Profiles for the Reference Operating Conditions. Injector Pressure=lS psia, Steady-State Mnss Flow R a t ~ 5 . 7 6 lb/tt2-sec. 7 c I , ,> I. *. I. I , 3 , ! I I. 8 . -L,,.L~-....A ....... , . - > ' 1". .-,, . . Similar calculations were performed for l iquid hydrazine temperatures and i n i t i a l catalyst par- t i c l e temperatures between 495 and 580 deg R; all other parameters were fixed a t the values used i n computing the resu l t s shown i n Figs. 17 and 18. The l iquid hydrazine residence time i s plotted i n Fig. l 9 as a function of temperature for a number of pore radi i . hydrazine residence time is shown f o r only a modest change i n the hydrazine temperature. resu l t s p r i m r i l y from the large variation i n hydrazine vapor pressure over the temperature range from 495 t o 580 deg R, which direct ly affects the reactant concentration a n d therefore the r a t e of reaction. b A very marked change i n l iquid This effect L Figure 18. i n Catalyst Pores. Liquid Wdrazine Penetration Depth F i y e 19. Liqu id Wdrazine Residence Time. L. Liquid hydrazine residence times within porous catalyst p8rtlclee, calculated using t h e =themati- cal analysis. were farnd t o exhibit the same depen- dence on terperature as induction eriods neahured in the instrummted reactor t e s t s lo. This s i m i - l a r i t y lends credence t o the l iquid hydrazine pene- t ra t ion mechanism as an explanation for the urmsuaUy long ignition delays and Nbsequent chamber pressure excursions observed during cold s ta r t s . attempt t o provide quantitative measurements of the l iquid penetration phenomenon, an experiment was developed using high speed m t i o n picture techniques t o determine l iquid hydrazine residence times as a function of the adsorbed gas condition of the cata- lyst and the l iquid hydrazine/catalyst temperature U. Nitrogen and carbon d i o a d e were f m d t o have l i t t l e o r no effect on catalyst ac t iv i ty while hydrogen and snrmonia were found t o have effeces which varied aignificantly with the gas conditioning pressure and with temperature. . I n an Poisoning of Shell 405 catalyst par t ic les upon prolonged exposure t o high pressure, low temperature hydrazine decomposition products can occur even i n low pressure reactors when liquid hydrazine blocks t h e pores of catalyst par t ic les until sufficient gas pressure is generated t o expel the liquid. Of couTse, exposure of par t ic les t o high pressure, low temperature hydrogen and snmonia for short times (fractions of a second) should not lead t o exces- sive deterioration. H-wever, repeated exposure fallowing expulsion of l iquid from a pore, caused by R o w of l iquid back t o the pore m t h , can occur i n a catalyt ic reactor. decomposition producta would then occur in an OsciUatOry fashion whenever l iquid wet the muth of a pore. The r a t e of desorption of decomposition products from the pore surface is not l ike ly t o be high as long as the par t ic le is cooled periodically by the l iquid hydrazine. Catalyst poisoning can progress dmet ream Sn a catalyst bed and lead ultimately t o "washout." This mechanism may be prominent under continuma Row conditions as demnstrate6 in Ref. 12. Exposure t o high pressure During the course of the l iquid hydrazine penetration t e s t s fracturing of par t ic les was observed. fundamental studies of catalyst a t t r i t i o n It was t h i s obsemation which prompted and 14. Catalyst Ewe- Wacturing of a catalyst par t ic le can be caused by la rge pressure gradients or thermal s t resses within the par t ic le vhich cannot be m p o r t e d by the porous structure. Breakup caused by large pressure grad- ients is i l lus t ra ted in Fig. x) under conditions i n which l iquid hydrazine wetting the ar ta ide sur- face of a catalyst par t ic le blocks the escape of gaseous decomposition products. Wetting by the very reactive hydrazine resu l t s in gaseous decomp- onition product buildup near the gas-liquid inter- fme. pation of pressure throughout the par t ic le is slow Ressure buildup is very rapid, while dissi- 9 . . b a r u s e of the very #mdl pore sizes and c o r r e p Fccdlngly hlgh pressure drops. Ressu re C(UI be d o v i a t e d I f (a) 68s pushed Uwid Out of the T o r e n , (b) gas escapes from t h e pores which have mt been wet, or (C) the particle i r a c t u e s . Greater the tendency of the p a r t i c l e t o wet and t \ e higher the fraction of the p a r t i c l e which is vet , the w r e l ike ly the particle is t o iracture. I n a study reported in Ref. 15. p a r t i c l e wetting YLB sbom t o be related d i rec t ly t o catalyst re- actfvity in hydrazine. For a given catalyst par- t t c le , react ivi ty increases m k e d l y with tempera- t u r e since the l iquid hydrazine vapor presswe is v e r y sensit ive t o temperature and it is the vapor bwch deco~poses on the catalyst m f a c e s . mr a set of partlcles. wetting is associated with the intr insic ac t iv t ty Of each of the par t ic les -- t\e lower the act ivi ty , the greater the likelihood of wetting. pressure buildup was found t o be an important mechanism which is Innuenced by nmny factors, rnrticutarly l i q u i d temperature. par t fc le ac t iv i ty and par t ic le history. LJ The I(reakup caused by wetting and internal P l w e 20. W r t i r l e Breakup me t o Wetting. In this experimental study 15 ind iv iduo cats- &st par t ic les were muted in a flow reactor i n which the reactor tenperatme, hydrazine f low velo- c i t y and exposure t i m e could be controlled. l y s t par t ic les were tested t o determine the general relationship between wetting and breakup and the effect of par t ic le size md history on these pheno- mena. A logical picture of par t ic le react ivi ty emerged from these observations. Par t ic les which are very low i n react ivi ty although t h e y rimy be extensively wetted do not undergo a high ra te of internal pressure b a d u p because the rate of cacnpe of gas from large pores is adequate t o minimize internal forces. The very reactive particles generate gas ne= the syfacs rapidly enough t o prevent Wetting. In t h i s inatance the internal pressure a lso does not rise. In Lntermediate cases, however, partial wetting can cause the r a t e of evolution of gas t o ex- ceed the ra te of escape a d the internal forces can W d causing breakup. Cata- Pruticles plhich were exposed t o low i n i t i a l t w e r a t u r e s were f m d t o be mre Busceptiblc t o breakup on reactor heating. ?he important effect Of i n i t i a l temperature on breakup v i s e o from the probability of wetting i n discrete weas, rather than uniformly. A t lower i n i t i a l t eqera tures , greater fraction of the -face w i l l be wet #ince fever v e a s of the p a r t i c l e can generate gas rapidly mowh t o prevent capillery intrusion. 1c I If the temperature i s raise& however, the blocked pores tend t o remain blocked u n t i l high internal pressures are attained. n e s e experiments produced several surprising The physical size of par t ic les vas f a d The in t r ins ic ac t iv i ty of larger results. t o be important. par t ic les appear t o be greater.than t h a t of wall part ic les and breakup was significantly lower. This same resu l t had been reported i n a previous investigation a t Bell Aerospace 16. tu re hea t ing and exposure t o mdest hydrazine par- tial pressure produced surprisingly large and favorable effects on breakage for conditions under which neither pore nor structural changes would be expected. cant change i n act ivi ty was evidenced. A hypothe- sis was suggested i n Ref. 17 which involves non- uniform distribution of act ivi ty i n the catalyst par t ic les due, perhaps, t o the techniques eQlOYed for air stabil ization. Ebre specifically, it 18 suggested that the act ivi ty of the external sur- face of catalyst par t ic les nay be reduced during the l a t t e r stages Of marmfacture whve the cata- lyst is exposed t o oxygen under controlled condi- tions. Using the Same methods for air stabil iza- tion, l a rger par t ic les w o u l d be affected less than Wauer ones; 1.e. t h e fraction of the par t ic le which i s reduced i n ac t iv i ty I s larger for a smoll Particle. The larger par t ic les vould then be m r e active, at l e a s t i n the as-received state. For a given Catalyst par t ic le then, it W d be reason- able t o expect tha t external surface react ivi ty could he regenerated i n par t by heating or by ex- P o m e t o a ta'drogen atmsphere. Regeneration of even a very th in region near the surface of a Par t ic le could have a substantial effect on the fraction Of tha t surface which m u l d be wet by l iquid hydrazine. Low tempera- Reduction i n breakage without signifi- Limited tes t ing has been done t o exmine t h i s hypothesis l? and indications are tha t it 1s sound though it nust be considered tentat ive unt i l inde- pendent measurements are made of Intr insic catalyet ac t iv i ty and pore size distribution. If the hypo- thesis is verlfled It vould suggest (a) a change i n the method of air stabil ization ezpleyed i n the manufacturing prFedure and (b) an evaluation of programed l a e c t o r design t o enhance regeneration. This may s i q l y be an examination of M l d e c t o r wlth mular flow under no- circumstance6 and central flow during regeneration. JII. Can the Reactor Agin,q Process be )ibnftaed? Bas monopropellant intui t ion reached the point where a simple test t o m n i t o r the aging process i n a reactor can be developed? mat t e s t is not ret avsileble but there are Indications that one could be developed. I n the normal pulse mode of operation, Catcrly6t a t t r i t ion is the prlmry symp- tom of the aging process and wetting i s a pri- tmd perhap6 the pri- made of breakup. The re- lationship between wetting snd catalyst reac t iv i ty provides t h e clue t o the aging t es t . Catalyst re- ac t iv i ty i n hydrazine has an important influence on the temperature and pressure response of a reactor. Time constants representing temperature and pressure response r a t e s t o a c o l d pulse of hydrazine are l i k e l y t o be vely sensit ive t o catalyst reactivity. As a bed ages, a t t r i t i o n of catalyst which has been exposed t o l iquid hydrazine causes the reaction front t o mve downstream. This w c d d change the rcgponse r a t e for pressure measured at the exhaust and for temperatures measured at various posit ions i n the bed. The quantitative t e s t for wing would take advantage of the pre- dict ive capabili ty of the t ransient model of reactor systems and would require an independent btudy of the change i n catalyst reac t iv i ty with exgosure t o cold &&azine. would require consideration of the fac t t h a t whLle low temperature exposure t o hydrazine may resu l t i n catalyst deterioration, 6ome regeneration of ac t iv i ty resu l t s from decemposition of hydrazine reminine i n the off par t of a duty cycle. This l a t t e r study IV. References 1. Development of Design and Scaling Cri ter ia for Ebnopropellant Wdrazine Reactors Employing Shell 405 Spontaneous Catalyst, Final Report ARC-66-R-76, Vols. I and 11, Contract WLS 7-32, Rocket Research Corp., Seattle, January 1967. 2. Kesten, A. S., "Analfiica Study of Catalytic Reactors for Hydrazine Decomposition - Part I: Steady State Behavior of mdrazine Reactors," Roceedings of the Hydrazine Wnopropellant Technology Synposium, The Johns Hopkins Univer- s i t y Applied Fkysics Laboratory, Silver Spring, b l y l a n d (1967). 3. Kesten, A. B., "Analytical Study of Catalytic Reactors for Hydrazine Decomposition." UACRI. 910461-12, First h a l Progress Report, Contract XM 7-458, my 1967. 4. Kesten, A. 8.. "nirbulent Diffusion of Adat ?ad mas In Catalytic Reactors for Eydrazine Decomposition." Jcyraal of Spacecrafts and Rockets 1, 3 l (1970). 5. Keaten, A. S., "Aaalytical Study of Catalytic Reactors for Hydrazine Decomposition." U A m Wl0461-38, Third Annual Progress Report, Con- tract RA.5 7-49, (1969). 6. Keaten, A. 6. and T. W. Price, "Analytical and Experimental Studies of the Transient Behavior O f Catalytic Reactors for Eydrazine Decomposi- tion." Proceedings of the CPU 10th Liquid Propulsion Synposium, U s Vegas, Nevada (1966). C a r l n w , R. A. and W. Baker, "Space Environment Operation of Experimental Hydrazine Reactors," Report Eo. 4712. 4-67-28, 1?1W, Redondo Beach, 7. a, (1%7). c,. cnm4iOv-i, J. J. and A. S. Kesten, " b 8 b i S cas pressure mildup Within a Porous Catalyst h*icle Which is Wet by a Liquid Reactant," medcal Ewineeriw Science 6, 533 (1571). s a - i o v w i , J. J. and A. S. Kesten, "Study of &3razlne Reactor Vacuum Star t Characteristics," I I ~ L ~ 9 1 ~ 2 1 , Contract NRW-1795 (1969). 13. SL&ovanni, J. J. and A. 8. Kesten. "Analyti- cal and Experimental Studies of the Startup Characteristics of CatalLlytic Reactors for Hydra- r:ne Decomposition," proceedings of the CPIA 12th Liquid Propulsion Symposium, Las Vegas, lievnda (157'0). i c. u. se.ngiovanni, J. J. and A. S. Kesten, "Wtion picture Studies of the Startup Characteristics or Liquid Hydrazine Catalfiic Reactors," Paper Eo. 71-702, AIAA/Shf 7th Propulsion Joint Specialist Conference, Sal t Lake City, Utah (1971) * l?. KrJecer, 0. W. and R. W. Riebling, "Experi- mental Evaluation of Performance Losses in bnopropellant Hydrazine Reactors Using Shell 435 Catalyst," 1572 JAN114F Propulsion Meeting, CPIA hb l i ca t ion No. 228, Vol. IV , (1572). 13. Kesten, A. S. and P. J. lk teney , "Internal Catalyst Breakup Phenomena,'' Contract AFAPL- TH-75-68, Final Report (1576). t u I Breakup Phenomena," Contract AFf(pLTR-76-47, 1L. Taylor, Y. F. e t al., "External Catalyst Final Report (196). I . I-. fhrteney, P. J. and A. S. Kesten, "Catalyst Par t ic le Wetting and Breakup i n Hydrazine System," Journal of Spacecrafts and Rockets 430 (1976). 16. bse ley , V. A. e t el., Zow-Life Wnopropellant irsign cr i ter ia ," contract ~0461~.-73-C-0044, WJI Statu. Reports, BeU. Aerospace Company, (1974-1975). 17. W e n e y . P. J. md A. S. lresten, "Effect of Catalyst Conditioning on Wetting and Breekup," W A Paper No. 76-620, AIAA/W 12th Propul- sion Conference, palo Alto, CA (1976). 12