N A T I O N A L AERONAUTICS A N D SPACE ADMINISTRATION

Technical Report 32-7227

The Status of Monopropellant Hydrazine Technology

T. W. Price
D. D. Evans

JET PROPULSION LABORATORY

CALIFORNIA INSTITUTE O F TECHNOLOOY

PASADENA, CALIFORNIA

February 15, 1968

 N A T I O N A L AERONAUTICS A N D SPACE ADMINISTRATION

Technical Report 32-7227

The Status of Monopropellant Hydrazine Technology

T. W. Price
D. D. Evans

Approved by:

J E T PROPULSION LABORATORY
C A L I F O R N I A INSTITUTE OF T E C H N O L O G Y
PASADENA, CALIF O R NIA

February 15, 1968

 TECHNICAL REPORT 32-7227

Copyright @ 1968
Jet Propulsion Laboratory
California Institute of Technology

Prepared Under Contract No. NAS 7-100

National Aeronautics & Space Administration
 Acknowledgment
The authors would like to acknowledge the cooperation of the Hamilton-
Standard Division of United Aircraft Corp., the Marquardt Corp., the Rocket
Research Corp., the Shell Development Co., and TRW Systems for permission
to reproduce data and for discussing various technical areas with us.

JPL TECHNKAL REPORT 32-1227 iii

 Contents (contd)

Figures

1. Micropore surface area vs run time for Shell 405 aatalyst (20-30 mesh) . . . . 4
2. Theoretical characteristic velocity and adiabatic gas temperature of
monopropellant hydrazine . . . . . . . . . . . . . . . . . . . 6
3. Theoretical vacuum specific impulse of monopropellant hydrazine . . . . . 6
4 . Steady-state axial profiles in a catalyst bed . . . . . . . . . . . . . 7
5. Steady-state axial profiles of reactant concentrations in a catalyst bed . . . . 7
6. Ignition delay vs the reciprocal of bed temperature . . . . . . . . . . 11
7. Characteristic reactor response vs chamber pressure . . . . . . . . . . 12
8 . Injector head temperature as a function of duty cycle . . . . . . . . . . 13

9. Effect of pulse width on specific impulse . . . . . . . . . . . . . . 14

10. Effect of pulse number and pulse width on specific impulse . . . . . . . . 14

11. Mariner C post-injection propulsion system schematic . . . . . . . . . . 15

12. Ranger-Mariner monopropellant hydrazine 50-lbf thrust rocket engine
(nonspontaneous catalyst) . . . . . . . . . . . . . . . . . . . 16

13. Mariner IV fuel tank pressure increase as a function of flight time . . . . . 17

14. Mariner 1969 spontaneous catalyst 50-lbf thrust rocket engine . . . . . . . 18

vi JPL TECHNICAL REPORT 32-7227

 Contents

.
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 1

II. Hydrazine Properties . . . . . . . . . . . . . . . . . . . . . 2

.
111 Catalysts for Hydrazine Decomposition . . . . . . . . . . . . . . 3
A . Nonspontaneous Types . . . . . . . . . . . . . . . . . . . . 3
B. Spontaneous Types . . . . . . . . . . . . . . . . . . . . . . 3

.
IV Catalytic Reaction Chambers . . . . . . . . . . . . . . . . . . 5
. . . . . . . . . . .
A . Continuously Operated Thrust Producing Devices 7
1. Throttling . . . . . . . . . . . . . . . . . . . . . . . . 8
2. Problem areas . . . . . . . . . . . . . . . . . . . . . . 8
B. Performance of Continuously Operated Devices . . . . . . . . . . . 9
C. Intermittently Operated Thrust Producing Devices . . . . . . . . . . . 10
1 . Start transient . . . . . . . . . . . . . . . . . . . . . . 10
2. Catalyst bed lifetime . . . . . . . . . . . . . . . . . . . . 12

3. Heat transfer to the propellant feed system

. . . . . . . . . . . . 12

D. Performance of Intermittently Operated Thrust Producing Devices . . . . . 13

E. Gas Generators . . . . . . . . . . . . . . . . . . . . . . 14

.
V Monopropellant Hydrazine Thruster Use on Flight Systems . . . . . . . 15

.
A Ranger-Mariner . . . . . . . . . . . . . . . . . . . . . . . 15
B. Mariner 1969 . . . . . . . . . . . . . . . . . . . . . . . 17
.
C lntelsat . . . . . . . . . . . . . . . . . . . . . . . . . 17

D. Transtage . . . . . . . . . . . . . . . . . . . . . . . . . 17
E. ATS-C . . . . . . . . . . . . . . . . . . . . . . . . . . 17
F. NRL Satellite . . . . . . . . . . . . . . . . . . . . . . . . 18

VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 18

References . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Table
1. Some physical properties of hydrazine . . . . . . . . . . . . . . . 2

J P l TECHNICAL REPORT 32- I227 V

 Abstract
The use of hydrazine as a monopropellant for thrusters and gas generators
has several outstanding advantages. The associated systems are simpler, as
compared to bipropellant systems, and the exhaust products are clean and
relatively cool, allowing utilization of simple, radiation-cooled decomposition
chambers. The development of the Shell 405 catalyst, which spontaneously
initiates hydrazine decomposition at room temperature, has greatly increased
the interest in monopropellant hydrazine systems.

This report gives a brief account of the history of hydrazine as a monopro-

pellant, as well as a few comments on the handling and physical properties of
the propellant. The major emphasis, however, is given to a comprehensive
survey of the current status of this technology, including catalyst availability
and handling, reactor design and operation, system considerations, and the
limitations and problems of monopropellant hydrazine. The current and antici-
pated applications are described, and a brief account of non-propulsion applica-
tions is included. Finally, current work relative to new catalysts and propellant
additives is summarized.

JPL TECHNICAL REPORT 32-1 227 vi i

 The Status of Monopropellant Hydrazine Technology

1. Introduction propellant (Refs. 1 and 2). It was found that homoge-

neous, thermal decomposition of hydrazine proceeded
Rocket system designers have often longed for a magic very slowly, requiring a residence time of about 35 m~
liquid which, upon opening a single valve to a simple that corresponded to a characteristic length L” of about
“decomposer,” would instantly change state into a clean, 800 in., far too large to be practical in most applications
energetic gas that could then be used to do work- (Ref. 3 ) .
propel or stabilize a vehicle, pump liquids, drive turbo-
machinery, etc. Monopropellant hydrazine systems, par- Subsequent work evolved an efficient catalyst that
ticularly those using Shell 405 catalyst, are currently would sustain hydrazine decomposition if the catalyst
the closest thing to such a dream. The purpose of this temperature could be brought to about 600-800” F ini-
paper is to acquaint a non-specialist with the current tially (Refs. 4 and 5 ) . It was necessary, therefore, to
status of this technology. Little new data will be pre- employ a technique that preheated the bed to initiate and
sented; rather, the intent is to discuss what is known, as sustain decomposition. Preheating methods used include:
well as what is not known, the existing applications, and
the potential of this technology. (1) Using electrical or radioisotope techniques.
(2) Operating in the liquid bipropellant mode for a
The Germans used a fuel mixture of hydrazine-hydrate short period of time.
in methyl alcohol ( B-Stoff ) with hydrogen peroxide as
the oxidizer for the power plant of their manned inter- (3) Utilizing a liquid-solid hybrid reaction (Ref. 6).
ceptor ME-163. As anhydrous (i.e., less than 5% water) The complications of such techniques, together with early
hydrazine became available, it was studied extensively doubts about the safe handling of the propellant, se-
in this country, initially by JPL in the late 1940s, for riously impaired the acceptance of monopropellant hy-
potential usage as either a bipropellant fuel or as a mono- drazine systems.

JPL TECHNICAL REPORT 32-1227 1

 Meanwhile, a broad base was being laid at JPL and The first operational usage of monopropellant hydra-
elsewhere as the feasibility of monopropellant hydrazine zine was in the Able-4 lunar spacecraft, developed dur-
devices was demonstrated for a variety of applications, ing 1959 by Space Technology Laboratories. Although
including turbo-alternator power supplies (Ref. 7), liquid launch vehicle difficulties prevented the attainment of a
guns (Ref. 8), and generated gas pressurization sys- successful mission, the Able4 20-lbf thrust hydrazine en-
tems (Ref. 9 ) . Simultaneously, industry acceptance of gine was fired in space as an experiment. Next came the
hydrazine-based fuels evolved with the earth-storable JPL Ranger and Mariner series spacecraft. All these
ballistic missile development. spacecraft employed a 50-lbf thrust hydrazine engine for
postinjection trajectory correction maneuvers. These en-
Monopropellant hydrazine systems offer several advan- gines employed the JPL nonspontaneous catalyst and
tages over the bipropellant systems or cold-gas systems. used a bipropellant ignition scheme. Design details of the
Monopropellant systems are much simpler (which usually Mariner system will be discussed later.
implies greater reliability ) than a bipropellant system be-
cause, obviously, only one fluid system is required and,
moreover, the necessity to maintain mixture ratio is not
required. But, at the same time, monopropellants offer II. Hydrazine Properties
many of the advantages of bipropellants, such as carrying
At ordinary temperatures and pressures, hydrazine is a
the desired energy in a compact liquid form, easy on-off
clear, colorless liquid with an odor similar to that of
control, throttlability, etc. Nonreactive, cold-gas systems
ammonia. Some of its physical and thermodynamic prop-
also offer many of these advantages, but hydrazine offers erties, pertinent to this discussion, are presented in Ta-
significantly greater quantities of energy in much smaller ble l. An extensive discussion of hydrazine may be found
and lighter packages. A further advantage of monopro- in Ref. 10.
pellant hydrazine is its low operating temperature allow-
ing chambers to be built of low cost materials and to be
Hydrazine is miscible only with polar solvents such as
operated with no cooling. So, basically, the advantage of
water, alcohols, ammonia, and amines. References 11-15
monopropellant hydrazine is energetic simplicity. The
discuss some of these mixtures and their properties.
added reliability (i.e., simplicity) cannot be overempha-
sized.
Hydrazine is thermodynamically unstable with respect
to decomposition into the elements and is exothermic
The engineering problems associated with hydrazine
when decomposed. It readily decomposes catalytically
systems are few. In the course of this paper, those de-
and also decomposes thermally in the vapor phase if
ficiencies that are unique to monopropellant hydrazine
heated to a high enough temperature. In either case, de-
systems will be pointed out, but no attempt will be made
composition is accompanied by release of considerable
to discuss those problems which are common to all liquid
amounts of energy.
propellant systems such as valves, tanks, regulators, etc.
Although hydrazine is easy to decompose, the initiation
Table 1 . Some physical properties of hydrazine energy required is high enough so that it can be handled
quite safely. It is not shock sensitive and can be heated to
Property Quantity
approximately 500 F before thermal decomposition be-
Molecular formula NZHI
comes a problem. It should be noted, however, that many
materials begin to act as catalysts well below 500°F;
Molecular weight 32.04 awu
therefore, a more practical limit is about 350"F. Mixtures
Density 1.0045 g/cm' at 25°C
of air and hydrazine vapor are extremely flammable and
Melting point +34.7'F at 1 atm care must be taken to exclude air from any hydrazine sys-
Boiling point +236OF at 1 atm tem.
Surface tension 0.004568 Ibf/ft at 77OF

Viscosity 0.000605 Ibm/ft-s at 77°F

Many common materials of construction are readily
compatible with hydrazine. Several of the 18-8 series
Heat of formation 12.05 kcal/mole
stainless steels, most aluminums and titaniums, Teflon,
Heat of vaporization 54.0 Btu/lbm at +236"F
and butyl rubber are materials that have been used exten-
Heat capacity 0.737 Etu/lbm-'F at 77OF sively. Considerable compatibility data are given in Refs.

2 JPL TECHNICAL REPORT 32-1227

16 and 17. Hydrazine systems must be free of rust and A. Nonspontaneous Types
scale or other contaminants which may act as a catalyst.
As mentioned previously, hydrazine decomposition
However, even this is not very stringent compared to the
must be promoted by catalysis for reasonable sized re-
cleanliness requirements for oxygen and fluorine.
actors to be attained. Extensive early work by A. F. Grant
(Refs. 3 and 5 ) evolved a series of catalysts using me-
One of the principal disadvantages of hydrazine is its tallic iron, nickel, and cobalt deposited on an alumina
relatively high freezing point ( 2 3 5 ° F ) . Most military carrier. When 1/8 to 3/16-in. cylindrical pellets of com-
specifications require a freezing temperature of -65"F. pressed, activated aluminum oxide are used as the carrier,
For many applications, the hydrazine freezing point is the catalyst is known as JPL Type H-AS. If higher-
acceptable, but a lower value would provide an extra strength, but lower surface area, fused aluminum oxide
margin of safety; this is still an area for much experimen- spheres are used, it is designated JPL Type H-7. Both
tation. The problem is to find an additive that does not types of alumina carriers are used in the chemical process
adversely affect the desirable properties. For monopro- industry and are readily available. These catalysts are
pellant applications, mixtures of water, ammonia, and resistant to chemical poisoning (although free carbon or
hydrazinium nitrate appear to be most promising. sulfur will act as poisons), have a long shelf life
( > 5 yr), and are available commercially." The use of
The long term storage stability of hydrazine has always non-critical metals and a relatively simple manufacturing
been in doubt, because of its tendency to undergo hetero- procedure results in an almost unlimited supply at a
geneous catalytic decomposition. At near-ambient tem- modest cost ($2 to $6/lbm). Their primary disadvantage,
peratures, long duration ( 2 4 yr) storage of hydrazine in as stated before, is that the catalysts will not sponta-
aluminum tanks was demonstrated at JPL (Ref. 18). neously decompose hydrazine at room temperature.
Long term storage in a bladdered titanium tank is dis-
cussed in Section V of this report. Relatively short dura- B. Spontaneous Types
tion ( ~ 1 0 h0) storage of hydrazine in titanium vessels at
Over the years, efforts were expended at JPL (Ref. 3)
293°F was also demonstrated at JPL (Ref. 19).
in searching for a spontaneous catalyst, but with no real
success. In January 1962, a search began in earnest when
So, while hydrazine is not something to be handled
the Shell Development Co. won a NASA competition to
casually, for the rocket industry, which routinely handles
attempt to develop such a hydrazine catalyst. The results,
chemicals such as liquid oxygen, liquid hydrogen, hydro-
reported'in Ref. 20, must be considered a true techno-
gen peroxide, nitric acid, fluorines, boranes, etc., hydra- logical breakthrough. This developed catalyst, known as
zine is relatively straightforward.
the Shell 405, is not only spontaneous, but is capable of
many spontaneous starts. This capability is one of the
primary reasons for the interest it has aroused.
111. Catalysts for Hydrazine Decomposition
The availability of this -catalyst has greatly expanded
The differentiation between nonspontaneous and spon- the potential application of hydrazine as a monopropel-
taneous catalysts is somewhat arbitrary and may well vary lant. Before the advent of Shell 405, interest in monopro-
depending on the interpreter. When hydrazine catalysis pellant hydrazine was centered principally at JPL, but
is the topic, spontaneous implies three things : now at least half a dozen industrial concerns are pursuing
The decomposition can be initiated while the cata- in-house or contracted programs based on the Shell 405.
lyst and hydrazine temperatures are no higher Both NASA and DOD are now heavily committed to
than 70 F. hydrazine systems.

No other energy source is required. The Shell 405 catalyst is composed of only two compo-
nents. These are the high surface area (160 m2/gm), alu-
The time between contact of hydrazine with the minum oxide carrier and the active metal, iridium. The
catalyst and the initiation of the decomposition iridium constitutes slightly over 30% of the total catalyst
process is very short ( 10-100 ms ) . mass. The actual manufacturing process is classified. The
If any of these restrictions are relaxed, then catalysts "Type Ni 1600 S (Type H-7) and Ni 1601 T (Type H-A-3) are
that ordinarily would be considered nonspontaneous, supplied by the Harshaw Chemical Div. of Kiewanee Oil Co.,
could be considered spontaneous. Cleveland, Ohio.

JP L TECHNlCAL REPORT 32-1 227 3

catalyst is currently available in several particle sizes and
shapes. The current cost of Shell 405 is about $2000/lbm,
and this cost is one of its major disadvantages. This factor,
along with the present fairly limited supply of iridium,
restricts its use somewhat. Shell estimates that there is
enough iridium currently available to make 1000 lbm of
catalyst a year. The current rate of catalyst production is
approximately 100 lbm per year. Iridium is presently a
by-product of platinum mining, but there are substantial,
unexploited iridium deposits in Africa, so that its avail-
ability should not necessarily be the limiting factor in
preventing major usage of the Shell 405.

As with any rapidly developing field of technology,

many potential problem areas are bypassed or explored in
only enough detail to solve the current pressing problems.
Handling of the Shell 405 catalyst is one of these poten-
tial problem areas. For instance, it has been known, since
the first introduction of the Shell 405, that it is hygro-
scopic. Also, upon exposure to air, the catalyst is oxidized
as well. What harmful effects these problem areas
might create, if any, has been mainly a matter for specu-
lation. It has been evident that, at ordinary ambient tem-
RUN T I M E , min
peratures and pressures, the effects were slight.
Fig. 1. Micropore surface area vs run time for Shell 405
Small scale reactor work at Aerospace Corporation led catalyst (20-30 mesh) (reprinted from Ref. 21)
to the first realization that there could be a potential
vacuum ignition problem. A broader study examining the catalyst. Exposure to alcohol or acetone may lead
area of catalyst handling confirmed the Aerospace obser- to a long ignition delay on the first firing following
vations relative to vacuum ignition. These results are re- exposure. However, even these two solvents had
ported in Ref. 21 and are based on the five phenomena no permanent effect.
investigated: exposure of catalyst to vacuum, surface area
degradation during reactor operation, exposure to com- (4) Cyclic operation of a reactor in a vacuum, followed
mon solvents, oxidation-reduction cycling, low tempera- by exposure to air, revealed no discernible catalyst
ture catalyst bed ignition. Summarized briefly, the degradation attributable to the air exposure.
results and conclusions of Ref. 21 are: ( 5 ) Catalyst bed temperature has a marked effect on
(1) Appreciable ignition delays may be encountered ignition delay, particularly below the freezing
in a vacuum environment. However, this effect is point of hydrazine.
strongly dependent on reactor design, temperature,
and the prior history of the catalyst; thus, the Some recent work by the Hamilton-Standard Div. of
effect cannot be adequately predicted at this time. the United Aircraft Corp. indicates that the ambient
pressure effect on ignition delay is associated with freez-
(2) The catalyst surface is reduced very rapidly dur- ing of the hydrazine, and not catalyst activity. Their
ing the first few minutes of reactor operation, and experiment consisted of observing the start transient for
then changes only slightly thereafter. Figure 1 different ambient pressures down to 8pm Hg (approxi-
illustrates this point. Although a properly designed
mately 0.0002 psia). At 17pm Hg, a noticeable and repro-
reactor will operate satisfactorily on degraded
ducible change occurred in the ignition delay. They
catalyst, this point is one of the factors that affect
attributed this change to a freezing of a portion of the
ignition delay.
hydrazine initially introduced into the reactor. The delay
(3) Exposure to methylene chloride, freon, trichloro- then occurs because it takes longer to decompose enough
ethylene, or distilled water has no effect on the hydrazine to raise the bed temperature.

4 JPL TECHNICAL REPORT 32-1227

 Although the Shell 405 catalyst has excited the most leave the catalyst bed. More commonly, one minus this
interest and seems to possess a very good combination quantity, or the amount of ammonia that has been disso-
of activity and durability, other spontaneous catalysts are ciated, is used.
available. These catalysts are generally less active, par-
ticularly at low temperatures, than the Shell 405 and, To present the significance of this percent-ammonia
in addition, may be less durable. Most of these are de- dissociated parameter, the chemical kinetics of the hydra-
scribed in Ref. 20. However, none have received a great zine decomposition process will be discussed briefly. The
deal of attention and their usefulness is still in question. overall decomposition mechanism was represented in
The military, in particular, has a requirement for lower Refs. 2 and 5 as a two step process, The first step is the
cost catalysts, even at the expense of longer cold bed exothermic decomposition of hydrazine into NH, and N,.
response times, less efficiency, etc.

IV. Catalytic Reaction Chambers

This step is followed by the much slower endothermic
A hydrazine, catalytic decomposition chamber consists decomposition of the ammonia:
basically of an injector for distributing the hydrazine and
a catalyst bed enclosed in a suitable container. (See Fig.
12, a typical reactor.) The injector and the geometry of
the catalyst bed may vary widely depending on the pur-
pose of the reactor and the imagination of the builder. When initiated, the first reaction generally goes to com-
Regardless of the design and use, all reactors operate in pletion, while the second reaction is much slower. Thus,
basically the same way. the amount of ammonia dissociated can be controlled by
varying the length of the catalyst bed.
The purpose of the injector is to atomize the liquid
hydrazine and to distribute it to the catalyst bed as uni- Equations ( 1) and ( 2 ) were never intended to repre-
formly as possible. A second objective is to place the sent the details of the decomposition mechanism, but
maximum amount of catalyst in contact with the hydra- were intended to relate the initial and final chemical
zine in as short a time as possible to minimize the start species. n u s , the equations could be viewed as a pseudo-
transient time. stoichiometric relationship. However, Eqs. (1) and (2)
have been the source of some confusion, since they im-
Within the catalyst bed, there appear to be four more ply that a suitably short catalyst bed would result in no
or less distinct regions. In the bed region nearest the in- molecular hydrogen (i.e., no ammonia dissociated) in the
jector, the liquid is vaporized. The energy for this phase reaction products.
change comes from the catalytic decomposition of a small
amount of hydrazine. In the next region, moving from The heterogeneous, hydrazine decomposition process
the inlet of the bed to the outlet, the heterogeneous cata- has not been extensively investigated, but the homo-
lytic decomposition of the vaporized hydrazine assumes geneous process has been studied. References 22 and 23
a dominant role. In the third region, hydrazine continues show that the homogeneous mechanism is much more
to decompose, but principally in a homogeneous reaction. complicated, as would be expected, than that repre-
In the last region, a portion of the initially-formed am- sented by Eq. (1). Furthermore, hydrogen is produced
monia is dissociated endothermically to nitrogen and in a step that does not involve decomposition of ammo-
hydrogen. The first three regions occupy only about 10% nia. An analogous situation probably exists for hetero-
of the total bed length under steady-state conditions. geneous reactions.

The performance of monopropellant hydrazine reactors An expression that more closely represents the overall
can be measured with the usual rocket parameters of hydrazine decomposition process would be
characteristic exhaust velocity c* and specific impulse
Isp. However, there exists a single parameter, unique to 2N,H, ---+2NH, + N, + H, (3)
monopropellant hydrazine, to completely specify the
performance. This parameter is the amount of ammonia This process would be followed by further catalytic de-
remaining in the decomposition products as these gases composition of the ammonia.

JPL TECHNICAL REPORT 32-7227 5

 If Eqs. ( 1) and ( 2 ) are used to represent the decom- since the energies released in Eq. (1) and absorbed in
position of hydrazine, then there will appear to be at Eq. (2) are specified in advance. Since the temperature
least 25% of the ammonia dissociated regardless of the of the decomposition products is rather low (+1800°F)
catalyst bed length. Operating reactors rarely achieve an and since the only species present are N,, H,, and NHB,
ammonia dissociation of less than 40% because of other all chemical reactions essentially cease at the exit of the
limitations. Thus, the 2!3% minimum has little practical catalyst bed, and subsequent gas flow can be considered
significance. truly frozen. Thus, specifying the amount of ammonia
dissociated also specifies the theoretical reactor perfor-
For the performance of reactors, Eqs. ( 1) and (2) are mance. Figures 2 and 3 show some examples of theoreti-
employed (since these equations are widely used by cal performance as a function of X . A more extensive
those working in the field and the limitation is generally collection of theoretical performance calculations can
understood) to obtain be found in Ref. 24.

4
N2H4 4 - ( 1 - X ) NH,
3
+ 31 ( 1 + 2 X ) N2 + 2 X H, It is convenient to subdivide monopropellant hydrazine
catalytic decomposition chambers into three categories :
(4)
(1) Continuously operated thrust producing devices.
In this case, X represents the fraction of the originally
formed ammonia that is subsequently dissociated. When (2) Intermittently operated thrust producing devices.
Eqs. (1) and ( 2 ) are combined this way, it becomes
obvious why the amount of ammonia dissociated com- (3) Gas generators.
pletely specifies the composition of the decomposed hy-
drazine. Further, it fixes the temperature of the gases,

I CHAMBER PRESSURE - 150 psia I\ 1

3g
FROZEN FLOW

3800 0 20 40 60 80 100 200 I I I I I

0 20 40 60 00 IO
AMMONIA DISSOCIATION,'70 AMMONIA DISSOCIATED, Oo
/

Fig. 2. Theoretical characteristic velocity and adiabatic Fig. 3. Theoretical vacuum specific impulse of
gas temperature of monopropellant hydrazine monopropellant hydrazine

6 JPL TECHNICAL REPORT 32-1227

. A. Continuously Operated Thrust Producing Devices the hardware used with nonspontaneous catalysts were
not directly applicable to the more active spontaneous
In 1954, A. F. Grant (Ref. 5 ) published the results of catalyst. Clearly, a new set of design criteria was re-
the extensive experimental investigation at JPL on mono- quired. As a result of a competitive bidding, the
propellant hydrazine decomposition chambers. Included Rocket Research Corp. was awarded a NASA contract,
in this report are correlations for the prediction of the NAS 7-372, to do an experimentaI characterization of the
amount of ammonia that will be dissociated, the pressure Shell 405 catalyst. Volume I1 of the final report (Ref. 25)
drop across the bed, etc. for several nonspontaneous is similar to Ref. 5 in that it presents in one place the
catalysts. This report still serves admirably as the start- information needed to design a hydrazine catalytic reac-
ing point for reactors that use nonspontaneous catalysts, tor employing the Shell 405 catalyst.
and appreciably reduces the amount of development
needed to obtain an optimized reactor.
As a broader approach to characterizing any hydrazine
catalyst, the problem is being attacked from the analyti-
The first usage of the Shell 405 catalyst was the substi- cal point of view. This work, by the United Aircraft

-
tution of it for the nonspontaneous variety in reactors Research Laboratories, is being conducted under NASA
designed according to Ref. 5. The results, while far from contract NAS 7-458. The first year of work is described in
discouraging, indicated that the design procedures and Ref. 26. Briefly, the approach and results are as follows.
Two computer programs that are based on steady-state
and transient models of a reactor system have been de-
2 5 0 0 7 veloped. These programs have been used to calculate
0 EXPERIMENTAL DATA temperature and reactant concentration distributions for
various initial bed temperatures, feed temperatures,
chamber pressures, mass flow rates, injector locations,
and catalyst size distributions. The results have been
compared, wherever possible, to available experimental
measurements. Generally good agreement between theo-
retical and experimental results is found. Typical results
are shown in Figs. 4 and 5.

!F
W
U
0.61 I I I I I
3
tU
W
a
z
rn
a
O EXPERIMENTAL DATA

0
0 0.05 0.10 0.15 0.20 0.25
A X I A L DISTANCE, ft AXIAL DISTANCE, f t

Fig. 4. Steady-state axial profiles in a catalyst bed Fig. 5. Steady-state axial profiles of reactant concentra-
(reprintedfrom Ref. 26) tions in a catalyst bed (reprinted from Ref. 26)

JPl TECHNICAL REPORT 32-1227 7

 Thus, both nonspontaneous and spontaneous catalysts Under full-thrust conditions, the bed loading must be .
have been well characterized. To date, the largest reactor limited to avoid an excessive pressure drop across the
built and operated as a thruster has produced 300-lbf catalyst bed; this pressure drop could result in crushing
rust. However, there is a recent, serious interest in the catalyst particles. Bed loading must also be limited
higher-thrust levels, and several detailed designs are com- so that channeling, the flow of raw hydrazine between
pleted for reactors that produce 1500 lbf thrust, but hard- catalyst particles or down the walls, does not occur. Chan-
ware is yet to be built. This interest arose principally neling results in poor performance and, with nonsponta-
because of the inherent reliability (i.e., simplicity) of neous catalysts, may even cool the bed so much that the
monopropellants. Based solely on the specific-impulse decomposition reaction cannot be maintained. Channel-
system-weight type of analysis, bipropellants are almost ing does not appear to be a major problem with
always clearly superior. But when reliability becomes a Shell 405 because of its high reactivity. Channeling and
major consideration and when the total impulse require- excessive pressure drop must be considered for all reac-
ments are modest, then monopropellant hydrazine sys- tors, not just those being throttled; however, the design
tems are very competitive. criteria presented in Refs. 5 and 25 take these factors
into account.
At the other end of the thrust scale, 0.5-lbf thrust seems
to be the lower limit for continuously operated devices. 2. Problem areas. As might be expected, because of
There is no inherent reason why such devices must be the rather extensive amount of development work already
pulsed, but most applications seem to require it. In a sim- accomplished, the problems encountered are generally
ilar vein, most thrusters are operated at chamber pres- under control. Some of the more important ones are
sures less than 300 psia. Again, there is no inherent reason chamber pressure overshoot during the start transient,
why the pressure cannot be higher; but this low pressure pressure excursions (spiking) during steady-state opera-
is a natural consequence of the fact that most monopro- tion, and loss of catalyst activity.
pellant hydrazine reactors are a part of pressure-fed
systems. As will be described later, there are other appli-
a. Pressure overshoot. The pressure overshoot during
cations for which higher chamber pressures are required.
startup is a result of the finite time required to initiate
the decomposition reaction and, as previously noted, is a
1. Throttling. A requirement for many thruster applica- strong function of the initial temperature and prior his-
tions is that of throttling. This is a technology area where tory of the catalyst. During this time period, which can
monopropellants in general and hydrazine in particular vary from several to over 200 milliseconds, raw hydrazine
offer several significant advantages over bipropellant is accumulated. When the reaction is finally initiated, the
rockets. With only one propellant, no mixture ratio prob- released energy is rapidly transferred to the unreacted
lem exists. Although reasonably uniform distribution of hydrazine, which then decomposes thermally. This de-
the hydrazine over the catalyst bed must be achieved, composition can be controlled, however.
injection requirements are much less stringent than those
for bipropellant engines. Obviously, mixing considerations Probably the best control is to put a thin layer of gran-
are nonexistent. Consequently, merely providing a vari- ular catalyst at the top of the bed. This layer provides a
able pressure drop in the feed system accomplishes the region of high catalyst surface area per unit volume and
throttling effectively. Several reactors have been throt- significantly reduces the reaction initiation time. Reduc-
tled in this way over a 1O:l range (50- to 5-lbf thrust). ing the bed loading accomplishes the same objective.

Several factors tend to limit the throttling range. At b. Spiking. The spiking problem is characterized by a
the high-thrust end, the bed must be long enough to sudden change from what appeared to be a state of
insure smooth operation; however, as the propellant flow- equilibrium to another steady-state condition. This change
rate is reduced, the absolute level of performance will can occur anytime from a few milliseconds to a minute
drop noticeably even though a large percentage of the after start of the reactor and usually is not a violent
theoretical limit is still achieved. This performance drop change, but it is undesirable, particularly when thrust-
occurs because a bed sized for high chamber pressure time predictability is important. This change is believed
will yield a higher fraction of ammonia dissociation at to be the result of liquid N,H, initially penetrating deeply
the low chamber pressure conditions. into the bed; then, as the entire reactor warms, a part of

8 JPL TECHNlCAL REPORT 32-1227

- this liquid region reaches a temperature such that the The problem of fines is far from being completely
liquid vaporizes, decomposes, and changes the tempera- solved. The origin can be from at least two sources:
ture distribution, establishing a new equilibrium condi- mechanically induced vibration, such as from a launch
tion. The entire process requires only a few milliseconds. vehicle environment, and abrasion during operation.
Unequal thermal expansion of the catalyst and the reac-
Once the second steady state is established, it will tor results in an apparent decrease in bed volume during
persist as long as the reactor is in operation. This spiking reactor operation, and further loosens the catalyst, ac-
is one of the difficulties discovered when Shell 405 was centuating the abrading process. The top of the bed is
used in a reactor that operated very satisfactorily with the particularly susceptible to the unequal thermal expansion.
nonspontaneous catalyst. An obvious approach is to hold the catalyst securely in
place. Spring loaded devices have been tried with some
success. A NASA-funded investigation is just beginning
Again, this problem can be alleviated by a layer of to explore the possibility of deliberately fusing the indi-
granular catalyst at the top of the bed. This layer appar- vidual particles together (Contract NAS 7-583, Rocket
ently provides enough surface area so that the incoming Research Corp.). If this fusing can be accomplished
hydrazine is completely vaporized and the decomposition without adversely affecting other bed properties, it could
process firmly established in the first 0.1-0.2 in. of the be very successful. The Shell Development Co., under an
bed during the initial start transient. Air Force contract, is attempting to develop carriers with
improved high temperature properties.
c. Loss of catalytic activity. Loss of catalytic activity
could conceivably result from any of several causes. Some
of these causes are: decrease of the total catalyst surface The rates at which catalyst is lost have been measured
area, crystalline growth of the active metal, poisoning for many reactors. However, it is meaningless to list these
of the catalyst, and physical loss of catalyst. Poisoning is rates without also giving details of the reactor, the op-
not a problem with neat hydrazine, but can occur from erating conditions, prior history of the catalyst, etc. Two
free carbon or sulfur. Attempts at promoting the cataIytic examples will be cited, however, to illustrate the magni-
decomposition of mixed hydrazine fuels (added MMH or tude of the problem. During the course of the work
UDMH ) have been almost totally unsuccessful because described in Ref. 25, widely varying loss rates were noted.
of the presence of carbon. Amounts as high as 10% of the original amount of cata-
lyst loaded were lost per minute. Shortly after this, an
improved carrier was found for the granular catalyst,
The surface area loss and active metal recrystallization reducing the loss rate by nearly an order of magnitude.
are known to occur, but, as was noted in Ref. 21, the only More recently, one concern operated a 0.1-lbf thrust re-
observed effect has been a longer ignition delay for sub- actor continuously for 3.5 h. During that period, 10%
sequent cold bed starts, particularly when the bed is near of the original catalyst was lost, representing a rate of
the freezing temperature of hydrazine. One of the diffi- just slightly under O.O5%/min.
culties is the lack of a sensitive, quantitative, and defini-
tive measurement of catalyst activity. Ignition delay can
be used, but it is too heavily dependent on the reactor
design to be very useful. B. Performanceof Continuously Operated Devices
Theoretical performance of these devices has already
d. Physical loss of catalyst. By far the most serious been presented in Figs. 2 and 3. Always a crucial criterion
problem is the physical loss of catalyst that can result in in the rocket engine industry is the fraction of the theo-
very rough operation. This loss is the result of relative retical performance that can be attained. Since mixing
motion of adjacent catalyst particles with the subsequent and atomization are not key requirements, 95-98% of
abrasion of the particles and the crushing of individual the theoretical vacuum specific impulse lspvat, shown in
pellets from pressure drop loads. The net result is the Fig. 3, is commonly achieved over a wide range of op-
generation of fines, which are loosely defined as particles erating conditions. Small size reactors may not do as well
smaller than 1/3 the size of the smallest particles orig- because of higher heat losses through the reactor walls.
inally loaded into the reactor. These fines can easily pass Reactors that operate at a high bed loading may likewise
between the larger particles; many pass completely out of suffer because of channeling. Nevertheless, it is not un-
the reaction chamber. reasonable to expect to achieve 95% of theoretical I,, vat.

JPL TECHNICAL REPORT 32-1227 9

 A separate criterion is the absolute level of perfor- for an expansion area ratio of 50:1, are realized easily .
mance that can be expected or the fraction of dis- within the present state of the art.
sociated ammonia that can be achieved. Reference 25
reported a minimum value of about 55% of the ammonia C. Intermittently Operated Thrust Producing Devices
was dissociated for the Shell 405 catalyst, the limitation
being a bed length long enough to achieve an acceptable Development of the Shell 405 catalyst has perhaps its
level of roughness. A random variation in chamber pres- greatest impact in the area of intermittently operated
sure of *4% was used as the criterion. A reactor (see (i.e., pulsed) devices, because of the capability of mak-
Fig. 12) using a nonspontaneous catalyst operates at ing repeated cold bed starts. Since repeated starts with
an ammonia dissociation level between 50 and 55%. nonspontaneous catalysts is possible only if the bed is
The measured Isp at E = 44:l for this reactor, is hot, hydrazine did not receive serious consideration for
235 lbf-s/lbm and represents approximately 97% of pulse use until the advent of the Shell 405.Pulse opera-
theoretical performance. tion is characterized by very short periods of propellant
flow to the reaction chamber, as well as by many starts
In general, reactors that use the Shell 405 will produce and stops of the reactor operation.
more dissociated ammonia than those packed with a
nonspontaneous catalyst, because of the greater activity Potential applications are for attitude control devices,
of the former type catalyst. Thus, reactors using Shell maintaining a satellite in a prescribed orbit, etc. These
405 and designed according to the criteria of Ref. 25 can kinds of applications can require widely varying on-times
be expected to yield slightly less performance (about (i.e., the time during which the reactor is operating), as
5 lbf-s/lbm in specific impulse). well as various duty cycles (i.e., the on-time fraction of a
complete on-off cycle).
However, some recent developmental work shows
promise in reducing this performance differential. Sev- Because of the extensive backlog of experience with
eral reactors have been tested with beds shorter than nonspontaneous catalysts, converting from nonsponta-
those recommended by Ref. 25. A second advance is the neous catalysts to the Shell 405 for continuously operated
successful use of higher bed loadings than those covered reactors was relatively straightforward,Such was not the
by Ref. 25. Values of G (propellant flow rate/cross- case for pulsed reactors. Thus, this entire field is less
sectional area of catalyst bed) as high as 0.08 lbm/in.% extensively explored than the continuously operated re-
have been reported. Both of these advances are possible actors, but it is being actively pursued by industry. Parts
primarily because of improved catalyst retention tech- of the information needed to characterize pulse-mode
niques. reactors have been gathered by individual companies, but
as yet there has not been a report in the open literature
A third improvement being developed is the use of of a systematic program directed at this technology area.
beds composed of a mixture of spontaneous and non-
spontaneous catalyst. The reactor (see Fig. 14) as de- Pulse operation of a hydrazine monopropellant reactor
scribed in a later section uses this type of bed. Not only creates a set of problems not encountered during steady-
is the ammonia dissociation reduced, but this reactor state operation. Very short start and shutdown transients
runs smoother and the catalyst costs are reduced. Some are generally required over a varying range of environ-
investigators have reported success in using layered beds mental conditions. Reproducibility is often more im-
composed largely of nonspontaneous catalyst below a portant than a short pulse requirement, especially
thin layer of Shell 405. reproducibility of the impulse centroid. One common
approach to shortening the start transient is to increase
The performance of continuously operated reactors can the bed loading. But this approach can drastically affect
be summarized as follows. A reactor can be built for the catalyst loss rate. Repeated pulses also appear to be
most applications to deliver a minimum of 95% of theo- more detrimental to the catalyst than continuous op-
retical vacuum specific impulse. The use of a nonsponta- eration.
neous catalyst will result in less ammonia dissociated
than the use of the Shell 405. However, this difference 1. Start transient. The start transient is usually de-
will most likely be reduced in the near future. In any scribed in terms of the time from the propellant-valve
case, delivered specific impulses of 230-235 lbf-s/lbm, electrical-on signal until 90% of the final steady-state, or

10 JPL TECHNICAL REPORT 32-1227

rated, thrust is reached. Included in this overall re- principal factor seems to be high temperature, since the
sponse time are: same result can be obtained in an oven and the principal
result is a decrease in surface area (see Fig. 1).For a
(1) The valve opening time. duty cycle where the bed temperature remains above
(2) The time to fill the propellant feed system. approximately 300”F, this surface area change presents
no difficulty. However, there are applications where the
(3) The time between introduction of hydrazine to
bed will cool well below 300°F between periods of op-
the catalyst bed and the beginning of the decompo-
eration. As noted in Ref. 21, this low temperature is
sition reaction (i.e., the ignition delay).
precisely the condition for which surface area change is
(4)The time required to generate enough gas to raise important and will significantly affect the start transient.
the chamber pressure to near its final value. It would be very desirable to control the effect of surface
The valve actuation and feed system fill times can, of area on cold bed starts, but, most likely, only how to
course, be reduced by proper design. The ignition delay predict the effect will be known.
and the rise time are strongly affected by the injection
scheme, the catalyst bed geometry, the history of the The initial temperatures of the catalyst bed and the
catalyst, and the initial catalyst and propellant tempera- propellant have a very strong influence on the start tran-
tures. sient and the shape of the entire pulse. A preliminary in-
vestigation of this effect is included in Ref. 21. Figure 6
shows the effect of bed temperature on ignition delay.
Whatever the application, injectors generally receive a
Figure 7 illustrates the effect of bed temperature on the
great amount of attention. This is primarily the result of
pulse shape. The experiment that Fig. 7 illustrates con-
injector design being largely a “black art,” even for a
sisted of firing a reactor for 10 pulses (60 ms on and 340
system as simple as monopropellant hydrazine. It would
appear that every group that works with hydrazine de-
velops a favorite injection scheme which each feels has
1000
advantages over any other. Usually these techniques are
considered proprietary and will not be described in any
detail here. In general, the objective of these different
injectors for pulsed applications is to contact as much
catalyst in as short a time as possible. This goal can lead
to an injector that does not inject all or even most of the
20 0
hydrazine in an axial direction. It can also lead to high ~~~

G-loading approaches. The majority of these different

schemes achieve a total response time on the order of n 100
10-20 ms for a warm catalyst bed, and several can ap- E
proach this time for an initial bed temperature of less i
a
-I 60
than 100°F. W
n
40
z
The catalyst bed geometry for pulsed reactors is 0
h
dominated by the need to place a maximum catalyst z
(3 20
surface area adjacent to the injector. This requirement
presents no problem for low-thrust devices since a small-
sized granular catalyst is required anyway. For some IO
larger reactors, the bed is composed of two sizes of
catalyst particles to reduce the pressure drop. The small 6 W
catalyst is placed nearest the injector. Many different I
4
geometric approaches to this problem have been tried; 0 2.10 2.20 2.30 2.40
most reasonable ones appear to work satisfactorily. F) (17’F) (-5OF) (-25*F)(-43*F)

RECIPROCAL OF BED TEMPERATURE x 103

As was noted in the description of steady-state thrust-
ers, the internal surface area of the catalyst can change Fig. 6. Ignition delay vs the reciprocal of bed temperature
significantly depending on the history of the catalyst. The (reprinted from Ref. 21)

JPl TECHNICAL REPORT 32-1227 11

 Ir
VALVE CUI7
2. Catalyst bed lifetime. As was the case for con- .
tinuously operated reactors, the physical loss of catalyst
can be a problem for pulsed reactors. However, the prob-
lem does appear more severe for pulsing. Apparently,
200
many periods of short operations impose a more severe
environment than an equivalent amount of continuous
~ VALVE OPEN operation.

.-0 Several companies have built reactors capable of sev-

v)
n ~
eral hundred thousand pulses with a total on-time on the
CHAME R
w' order of 50,000 s (14 h). These reactors generally operate
E
at a low bed loading.
3
v)
w 100
E
n High bed loading is certainly a more stringent condi-
tion. A long train of pulses is a much less difficult re-
quirement than many starts with a cold bed, a condition
most likely to be affected by catalyst degradation.

To summarize, it is possible to achieve long lifetimes

for pulsed reactors. Long-life requirements and severe
operating conditions require careful design and the use
C of more sophisticated catalyst retention techniques.
f- TIME

3. Heat transfer to the propellant feed system. Any

Fig. 7. characteristic reactor response vs chamber
system that uses hydrazine must be designed so as to
pressure (reprinted from Ref. 21)
severely limit the amount of heat conducted to any re-
gion where the fluid is stagnated. The consequences of
not observing this rule are generally disastrous, since the
hydrazine will surely decompose with the release of
ms off), followed by 10 s of continuous operation, and
large quantities of gas and energy. This fact has been
then another 10 pulses. In contrast to the results shown
demonstrated many times in both monopropellant and
in Fig. 7, the 10 pulses after the continuous operation are
bipropellant rocket engines. Balancing this tendency of
nearly identical in both the start transient and the pulse
hydrazine to decompose is the fact that it is a relatively
shape.
good coolant and can absorb fairly large quantities of
energy before vaporizing and decomposing.
It would appear, from the evidence to date, that build-
ing a reactor producing identical pulse shapes for dif- Since monopropellant hydrazine reactors normally op-
ferent combinations of initial bed and propellant erate without any cooling, the metallic chamber can easily
temperatures and catalyst histories is not likely in the reach a temperature of 1700-2000°F. Heat from the
near future. It is much more likely that the ability to chamber can be conducted into the propellant feed sys-
predict the effect of these variables accurately will be tem. This conduction presents no great problem when the
achieved. reactor is operated continuously, since, unless there has
been a gross design error, the flowing hydrazine can easily
For applications where a long series of pulses are re- absorb this energy. With pulsed reactors, this heat transfer
quired, the error from the nonreproducibility introduced problem is not easily avoided. During periods of no
by the first several pulses becomes negligible. In recog- propellant flow it is quite possible to transfer enough
nition of the difficulties involved in making all pulses in heat to the upstream side of the propellant valve to lo-
a series match later pulses, the reproducibility require- cally decompose the hydrazine. Even the injector may
ment is often written in terms of a later pulse. Often the become hot enough to decompose the hydrazine there.
sixth pulse is used. This requirement is similar to that Figure 8 shows typical injector temperatures that a
encountered with hydrogen peroxide reactors. properly designed reactor may reach.

12 JPl TECHNICAL REPORT 32-1227

 0 2 4 6 8 IO 14 16 18

DUTY CYCLE, %

Fig. 8. Injector head temperature a s a #unction of duty cycle (reprinted by permission of Rocket Research Corp.)

The solution is to thermally isolate the critical parts of D. Performance of Intermittently Operated Thrust
the liquid system from the chamber and catalyst bed, and Producing Devices
to make sure that there are no parts of the feed system
Pulsed operation of a reactor may result in less of the
where hydrazine is heated while stagnant. The most
potentially available chemical energy being converted to
common technique used is to make the conduction path
sensible heat during a single pulse than for an equivalent
as Iong and thin as possible to introduce a large thermal
period of continuous operation. This energy Ioss is the
resistance. Often a long capillary tube (or tubes) con-
result of a significant part of the released chemical en-
nects the propellant valve to the injector, and the valve
ergy being lost first to heat the catalyst particles and
is mounted on a thin-walled structural member. This
chamber walls, and then to the external environment by
technique has the disadvantage of increasing the system
convection and radiation. This effect is illustrated in Fig.
liquid volume and, hence, the transient times. Seal ma-
9 which shows vacuum specific impulse vs the on-time
terials having very low thermal conductivities are often
for a specified duty cycle. The shorter the pulse and the
used, thus doing double duty as an insulator as well.
longer the time between pulses, the lower the impulse.
The problem of heat conduction to the feed system is Also, Fig. 9 shows the impulse after the reactor has
one that can be handled by proper design. Each reactor reached steady-state following a series of pulses that
must be treated according to its intended use. The duty bring the reactor to some equilibrium temperature with
cycle and on-time must be considered, as well as the its surroundings. Another illustration of this effect is
external environment. The use of computer programs for shown in Fig. 10, where vacuum specific impulse is
solutions to transient heat transfer problems has proved plotted vs the pulse number (in a series) for a given duty
to be an important tool to this end. cycle and various values of on time.

JPL TECHNICAL REPORT 32-1227 13

 250 hydrazine will produce about 30 ft3 of gas at standard ,
E
2 temperature and pressure.
3I 225
'c
0 The gas generator problems are the same as those al-
g.
-I
200 ready noted for thrusters. In fact, almost any thruster
3 could also serve as a gas generator. The gas generator
a
175 would differ only by the absence of an expansion nozzle
2 5-1 bf FLIGHTWEIGHT and by the catalyst bed being slightly longer to promote
LL
0 ammonia dissociation, thus lowering the molecular weight
E 150
and gas temperature and increasing the volume of gas
(I)

produced.
I3 125
0
5 An application of a monopropellant hydrazine gas gen-
100 erator is described in Ref. 9. This gas generator employed
0.01 0.10 1.0 IO
PULSE WIDTH, s a nonspontaneous catalyst and was equivalent to a
thruster producing about 20 Ibf. This generator fulfilled
Fig. 9. Effect of pulse width on specific impulse (reprinted all the design requirements and was capable of pumping
by permission of Hamilton Standard Div. of 11 lbm/s of liquid propellants. Another gas generator
United Aircraft Corp.) built and tested at JPL to drive the Jupiter turbopumps
was equivalent to a 1000-lbf thrust rocket. Current work
250 at Rocket Research Corp. under Hercules subcontract
E involves a large-throttled gas generator.
0
> 225
I
'c
0 Some of the more interesting applications of hydra-
k i 200
zine and hydrazine mixtures are being made with gas
cn
1
3
generators. The Naval Ordnance Test Station at China
a Lake, Calif., has been especially active in this field. The
2 175
0 Navy is interested in the deep-ocean recovery of objects,
LL
- and has successfully applied monopropellant hydrazine
150
to this problem. By expelling the gases from a reactor
(I)

I into a rubber buoyancy bag, the Navy has made free-

2
0
125 ascent-lifts of 4000 Ibm from a depth of 2800 ft. Another
B successful test has been a lift of 53,000 lbm from a depth
100 of 85 ft (Ref. 27). As a prelude to work at even greater
I 2 4 6 IO 20 40 60 100
PULSE NUMBER depths, the Navy has operated reactors at chamber pres-
sures as high as 11,500 psi. These latter tests were con-
Fig. 10. Effect of pulse number and pulse width on ducted at sea level.
specific impulse (reprinted by permission of
Hamilton Standard Div. of Other applications currently under development will
United Aircraft Corp.) only be mentioned to show the variety of uses possible.
A large-throttled hydrazine gas generator is being used to
E. Gas Generators supply the gas for a mass-excited solid-propellant rocket
(Rocket Research Corp. under Hercules subcontract).
Monopropellant hydrazine is ideally suited for gas Turbo-alternator systems for the generation of electrical
generation. A hydrazine gas generator differs from a re- power are being developed using hydrazine as the energy
actor, which produces thrust, only in the use made of the source. Attitude control systems that combine the ad-
decomposition products. The major goal for any gas vantages of cold gas thrusters with the compactness of
generator is to produce a large volume of cool, clean gas monopropellant hydrazine have been tested (Ref. 28).
that will not condense. Hydrazine and catalytic decompo- The gas generator products are stored in a plenum which
sition chambers meet these requirements extremely well. then supplies cold gas on demand. These are only a few
If 75% of ammonia is decomposed, each Ibm of liquid of the many applications of hydrazine gas generators.

14 JPL TECHNKAL REPORT 32-1227

V. Monopropellant Hydrazine Thruster Use on ridges provided the capability of two separate course
Flight Systems
corrections.

A. Ranger-Mariner A cross-section view of the Haynes Alloy No. 25 reactor

assembly is shown in Fig. 12. The injector consists of five
In the past, 50-lbf thrust, nonspontaneous catalyst
reactors were employed on the JPL Ranger and Mar- swirl-type atomizers, four for fuel and one for oxidizer, the
latter one being employed only during the first second
iner spacecraft for trajectory correction maneuvers
of operation. The catalyst bed consists of JPL Type H-7
(Refs. 29-31). These rocket engines used a 15-cm3 start
spherical particles (3/16-in. diameter).
slug of nitrogen tetroxide (N,O,) for ignition and ini-
tial catalyst bed heating. A schematic diagram of the
During qualification, this engine was subjected to:
Mariner-C propulsion system is shown in Fig. 11. This
system was used for both Mariner IV, now in solar orbit, (1) Humidity testing consisting of a soak at 21°C and
and Mariner V, the successful Venus probe. The use of 95% humidity, followed by exposure to 38°C and
dual explosively-actuated valves and oxidizer start cart- 95% humidity for 30 min.

NITROGEN TANK
TEMPERATURE TRANSDUCER

NITROGEN

PROPELLANT
TEMPERATURE
TRANSDUCER

NITROGEN PRESSURE PROPELLANT TANK

REGULATOR, PRESET
NITROGEN VALVE, SHUTOFF No. I THRUST CHAMBER
PRESSURE PROPELLANT VALVE PRWELLANT VALVE,
SHUTOFF No. I

PROPELLANT VALVE, PROPELLANT VALVE,

START No. 2

ROCKET ENGINE

OXIDIZER VALVE, START No. 2

CHECK VALVE
OXIDIZER VALVE, START NO.&-I OXIDIZER FILL VALVE No. 2
IGNITION CARTRIDGE
2-WAY VALVE, EXPLOSIVELY
0 OPERATED, NORMALLY
W OPEN
IZER
-VALVE No. I -I
I
II
@ 2-WAY VALVE, EXPLOSIVELY
OPERATED, NORMALLY
W CLOSED
OXIDIZER RESERVOIR

0
W 2-WAY VALVE, MANUALLY
OPERATED IGNITION PARTRlnGF
I "n.1
I GNlT ION CARTRl DGE
I 0.. 1"-

N2 RESE:RVOIR PRESSURETRANSDUCER>
N2 RESERVOIR, IGNITION CARTRIDGE

N2 FILL VANE, IGNITION CARTRIDGE

Fig. 1 1. Mariner C post-injection propulsion system schematic

JPl TECHNICAL REPORT 32-1227 15

 , - _ .

The most significant problem encountered was the

generation of catalyst fines during vibration. These fines
consisted of a relatively small quantity ( a gram or less)
of metal impregnated aluminum oxide with individual
particle sizes ranging from those of dust to those of coarse
sand (1-500 pm). Although the loss of these particles has
no measurable effect on the reactor performance, their
presence is potentially detrimental to sensitive optical
instruments. During the Ranger and Mariner programs,
an analysis was performed which indicated that the solar
pressure would effectively blow away these particles
within a matter of hours. No problems have ever been
identified that could have been caused by these catalyst
fines, nevertheless, their presence cannot be ignored. An
obvious solution to contain them within the reactor until
it is fired would be to use a frangible diaphram to close
the nozzle. Similar devices are used on solid propellant
rockets.

Perhaps our best space storage data on hydrazine sys-

tems is being obtained from Mariner ZV, launched during
Nov. 1964. Although its primary mission of obtaining
Mars scientific data has been completed, Mariner ZV is
still a fully operational spacecraft and data is obtained
about once weekly at this time.
Fig. 12. Ranger-Mariner monopropellant hydrazine 50-lbf
thrust rocket engine (nonspontaneous catalyst) The Mariner system uses a butyl bladder within a
titanium propellant tank. This propellant containment
A shock test of 100 g, 0.5- to 1.5-ms duration, five scheme .has been quite effective. Since the propellant
times in each of three orthogonal directions. tank is isolated from the high pressure nitrogen source
by an explosive valve assembly, any gas formation
Static acceleration of 14 g in three orthogonal di- resulting from the propellant either reacting with the
rections for 5 min each. bladder, or undergoing homogeneous or heterogeneous
decomposition, would result in a propellant tank pressure
Low frequency vibration of 3-g peak from 4.4 to increase.
15 Hz followed by complex wave vibration consist-
ing of 14-g-RMS noise for 36 s, and a total of 600 s
A plot of propellant tank pressure increase due to de-
of 5-g-RMS noise, 2-g-RMS sinusoid, and 9-g-RMS
composition is shown in Fig. 13. The data have been cor-
sinusoid.
rected for pressure changes resulting from temperature
For margin limit purposes, the engine was then test fired effects. A moderate ( 4 7 0 psi) pressure increase has
for 200 s during two separate tests, both tests at a occurred over a 2yZ-yr period. Also, the propellant tem-
chamber pressure that was 1.25 times the normal operat- perature has been quite moderate ( 4 6 4 7 ° F ) .
ing pressure (normal run duration and pressure are 100 s
and 190 psia, respectively). Two tests, each of 100-s Extrapolation of the Eberstein and G'lassman rate con-
duration, were conducted with the engine installed in an stants (Ref. 23) to these temperatures reveals that notice-
inverted position. The purpose of these tests was to dem- able gas phase homogeneous decomposition should not
onstrate that, after shutdown, the propellant contained be expected over a 3-yr period. Thus, the observed pres-
within the injector will vaporize without undergoing sure rise is concluded to be only due to reaction with the
explosive decomposition as a result of heat soak back bladder material or to heterogeneous decomposition, cat-
into the injector from the hot catalyst bed. All of these alyzed either by the bladder or, more unlikely, by the
tests were successful. titanium tank walls.

16 JPL TECHNICAL REPORT 32-1227

 -
BUTYL BLADDER
W
(I) I
I / 6Al-4V TITANIUM TANK
2 30- I / -
a: / TEMPERATURE RANGE :
0
l 0
45OF AT APHELION TO
I. -
67OF AT PERIHELION
I
I A A A &,
f PHELION PERIHELION APHELION PERIHELION
11 I I I I I I I I I
-1- _I_

Fig. 13. Mariner IV fuel tank pressure increase as a function of flight time

Higher temperatures, especially those in excess of equivalent reactor but for two exceptions. First, the quan-
100"F, are more detrimental because of marginal com- tity of catalyst fines generated during vibration appears
patibility with the butyl bladder (Ref. 31). A second to be somewhat greater; and second, this reactor appears
burn of the propulsion subsystem was accomplished to degrade faster as run time is accumulated. This degra-
in October of 1967. This successful operation conclu- dation takes the form of rougher operation. For example,
sively established the space storability of the nonspon- the Mariner 1969 prequalification engine roughness, after
taneous catalyst. all environmental tests, was +4% at the end of 330-s
accumulated time (individual tests of 30, 100, and 200 s ) .
B. Mariner 1969
For the Mariner 1969 mission to .Mars, JPL has elected c. /ntelsat
to fly a spontaneous catalyst engine in lieu of the pre-
vious nonspontaneous types. Although multiple opera- TRW Systems is currently fabricating 3.5-lbf thrusters
tion is not a requirement (only two starts are required), and the associated feed system for Intelsat-3. These re-
the inherently simpler system and increased reliability actors are capable of either Pulsed 0' continuous Opera-
were overpowering reasons to upgrade the system. tion. A complete description of this system is contained
in Ref. 32.
A cutaway of this 50-lbf thrust reactor is shown in
Fig. 14. A showerhead type injector is used and 20-mesh
D, Transtag=
Shell 405 catalyst is packed directly against a 60-mesh
screen that is adjacent to the injector face. The remainder Rocket Research Corp. is developing 25-lbf thrust en-
of the catalyst bed consists of a mixture of 75% Shell 405 gines under Martin-Denver subcontract for the Transtage
1/8-in. cylindrical particles and 2S% JPL Type H-A-3 vehicle. The significant point with regard to this work
1/8-in. cylindrical particles. The Type H-A-3 nonsponta- is that this monopropellant system will replace an exist-
neous catalyst was added after tests showed that it pro- ing bipropellant system to gain higher reliability.
moted smoother reactor operation. Roughness values,
based on chamber pressure excursions, now run about
E. ATS-C
&2%, equivalent to those values obtained on the pre-
vious Mariner nonspontaneous catalyst engines. Hamilton-Standard is delivering 5-lbf thrust reactors
to Hughes Aircraft for use on the ATS-C satellite. Pre-
This engine has undergone a prequalification test series viously, two redundant peroxide systems have been used;
similar to that described previously. It appears to be an now, one of these systems will use hydrazine.

JPL TECHNICAL REPORT 32- 1227 17

F. NRL Satellite ORIFICE PLATE 'YPE MOL SI
\ .STEEL
Rocket Research Corp. has delivered a 0.5-lbf system FINE CATALYST -INLET
to the Naval Research Laboratory for use in one of their -
small research satellites.

These programs summarize most of the known unclas-

sified flight project usage of hydrazine systems. There
are many other applications now being studied. One of
the more interesting of these is for the Voyager cap-
sule descent propulsion system. Although a relatively
high thrust is required (up to 2000 Ibf total), the mod-
erate total impulse requirement, together with the pos-
sible use of multiple engine configurations, make the use
of monopropellant hydrazine feasible for this purpose. 0,062-in
The recent success at JPL in heat sterilizing a simple HAYNES ALLOY
No. 25
Shell 405 system (275"F, 6 cycles), using a non-
bladdered titanium Ranger tank (Ref. 33), has demon-
strated the feasibility of heat sterilizing these systems.

VI. Conclusion
Monopropellant hydrazine is a rapidly expanding tech-
nology area. Only now does it appear that the full
potential is about to be realized. The development of the
spontaneous Shell 405 catalyst has greatly expanded the
interest in the very simple monopropellant hydrazine sys-
tems. The capability of many spontaneous ignitions and Fig. 14. Mariner 1969 spontaneous catalyst 50-lbf
the system simplicity possible with the Shell 405 are the thrust rocket engine
primary reasons for this growing interest.
very near future, reactors using the Shell 405 catalyst will
A monopropellant hydrazine reactor has flown on each be an important part of several vehicles. The next few
of the JPL Ranger and Mariner spacecraft and, in the years should see many more applications.

18 JPL TECHNICAL REPORT 32-1227

 References
1. Altman, D., and Thomas, D. D., Evaluation of Hydrazine as a Monopropel-
lant and Gas Generant, Progress Report 9-36. Jet Propulsion Laboratory,
Pasadena, Calif., Apr. 1949.
2. Thomas, D. D., The Thermal Decomposition of Hydrazine, Progress Report
9-14. Jet Propulsion Laboratory, Pasadena, Calif., Aug. 1947.
3. Grant, A. F., Jr., Catalysts for the Thermal Decomposition of Hydrazine
W h e n Used as a Monopropellant or as a Gas Generant, Publication 15. Jet
Propulsion Laboratory, Pasadena, Calif., Feb. 1953.
4. Grant, A. F., Jr., Development of Hydrazine as a Monopropellant and Gas
Generant, Report 9-1. Jet Propulsion Laboratory, Pasadena, Calif. Mar. 1950.
5. Grant, A. F., Jr., Basic Factors Involved in the Design and Operation of
Catalytic Monopropellant-Hydrazine Reaction Chambers, Report 20-77. Jet
Propulsion Laboratory, Pasadena, Calif., Dec. 1954.
6. Evans, D. D., and Lee, D. H., Zgnition Means for Monopropellant, US.
Patent 3,298,182 (issued to J. E. Webb, NASA). U.S. Department of Com-
merce, Washington, June 1964.
7. Lee, D. H., and Foster, C. R., A Monopropellant-Hydrazine Turboalternator
Auxiliary Power Unit, Publication 81, Jet Propulsion Laboratory, Pasadena,
Calif., Jan. 1957 (Confidential).
8. Foster, C. R., Gorden, E., and Van de Verg, N., Experimental Liquid-
Propellant Gun Firings at the Jet Propulsion .Laboratory, Publication 29.
Jet Propulsion Laboratory, Pasadena, Calif., Mar. 1954.
9. Lee, D. H., and Evans, D. D., The Development of a Heated-Hybrid Gen-
erated Gas Pressurization System for Propellant Tanks, Technical Report
32-375. Jet Propulsion Laboratory, Pasadena, Calif., Feb. 1963.
10. Audrith, L. F., and Ogg, B. A., The Chemistry of Hydrazine. John Wiley
and Sons, Inc., New York, 1951.
11. Thomas, D. D., Liquid-Vapor-Solid Equilibria for Ammonia-Hydrazine SYS-
tem, Memorandum 9-12. Jet Propulsion Laboratory, Pasadena, Calif., Apr.
1948.
12. Elverum, G. W., Jr., and Cole, L. G., Some Physical-Chemical Studies of
Hydrazine-Hydrazine Nitrate-Water, Memorandum 20-79. Jet Propulsion
Laboratory, Pasadena, Calif., Dec. 1952.
13. Petker, I., Wilford, D. B., and Mason, D. M., A Study of Some Freezing
Point Depressants for Hydrazine and Their Properties, Progress Report
20-238. Jet Propulsion Laboratory, Pasadena, Calif., Nov. 1954.
14. Vango, S. P., and Krasinsky, J. B., Density, Vapor Pressure, and Viscosity of
Solutions of H ydrazine Mononitrate in Hydrazine, Technical Memorandum
33-103. Jet Propulsion Laboratory, Pasadena, Calif., Oct. 1962.

JPL TECHNlCAL REPORT 32-1227 19

 References (contd)

15. Vango, S. P., and Krasinsky, J. B., Hydrazine Mononitrate in Hydrazine,

Supplemental Physical Data, Technical Memorandum 33-122. Jet Propulsion
Laboratory, Pasadena, Calif., Mar. 1963.
16. Lee, D. H., A Survey of the Compatibility of Various Materials with Hydra-
zine and Mixtures of Hydrazine, Hydrazine Nitrate, and Water, Memoran-
dum 20-152. Jet Propulsion Laboratory, Pasadena, Calif., 1957.
17. Toth, L. R., “Material Compatibility,” in Supporting Research and Advanced
Development, Space Programs Summary 37-44, Vol. IV, p. 175. Jet Propulsion
Laboratory, Pasadena, Calif., Apr. 1967.
18. Hollywood, L. P., Metz, T. R., and Porter, R. N., Storage Tests of Nitrogen
Tetroxide and Hydrazine in Aluminum Containers, Technical Report 32-1039.
Jet Propulsion Laboratory, Pasadena, Calif., Jan. 1967.
19. Stanford, H. B., “Hydrazine Sterilization Tests,” in Supporting Research and
Advanced Development, Space Programs Summary 37-35, Vol. IV, p. 155.
Jet Propulsion Laboratory, Pasadena, Calif., Oct. 1965.
20. Voge, H. H., et al., Development of Catalysts for Monopropellant Decompo-
sition of Hydrazine, Final Report 3-13947, Contract NAS 7-97. Shell
Development Company, Emeryville, Calif., Apr.-Dec. 1964 (Confidential).
21. Carlson, R. A,, and Baker, W., Space Enuironmental Operation of Experi-
mental Hydrazine Reactors, Final Report 4712, Contract NAS 7-520. TRW
Systems Group, Redondo Beach, Calif., Apr. 1967.

22. McHale, E. T., Knox, B. E., and Palmer, H. B., “Determination of the De-
composition Kinetics of Hydrazine Using a Single-Pulse Shock Tube,” in
Tenth Symposium (International) on Combustion, University of Cambridge,
Cambridge, England, Aug. 17-21, 1964. Combustion Institute, Pittsburg, Pa.,
1965.

23. Eberstein, I. J., and Glassman, I., “The Gas-Phase Decomposition of Hydra-
zine and its Methyl Derivatives,” in Tenth Symposium (International) o n
Combustion, University of Cambridge, Cambridge, England, Aug. 17-21,1964.
Combustion Institute, Pittsburgh, Pa., 1965.

24. Lee, D. H., Performance Calculations for Monopropellant Hydrazine and

Monopropellant Hydrazine-Hydrazine Nitrate Mixtures, Technical Report
32-348. Jet Propulsion Laboratory, Pasadena, Calif ., Dec. 1962.

25. Development of Design and Scaling Criteria for Monopropellant Hydrazine

Reactors Employing Shell 405 Spontaneous Catalyst, Final Report RRC-66-
R-76 Volumes I and 11, Contract NAS 7-372, Rocket Research Corp.,
Seattle, Wash., Jan. 1967.

26. Kesten, A. S., Analytical Study of Catalytic Reactors for Hydrazine Decom-
position-First Annual Report, UACRL F910461-12, Contract NAS 7-458.
Research Laboratories, United Aircraft Corp., East Hartford, Conn., May
1967.

20 JPL TECHNICAL REPORT 32-7227

 References (contd)

27. Witcher, J., and MiLIer, D., “Recovery of Deep-Ocean Objects Using Hydra-
zine-Powered Buoyancy Devices,” paper to be presented at the Monopropel-
lant Technology Symposium, Johns Hopkins University, Silver Spring, Md.,
Nov. 28-30,1967.
28. Groudle, T. A., “Reaction Control Gas Supply System,” in Supporting Re-
search and Advanced Development, Space Programs Summary 37-44, Vol.
IV, p. 171. Jet Propulsion Laboratory, Pasadena, Calif., Apr. 1967.
29. Lee, D. H., Development of the Midcourse Trajectory-Correction Propulsion
System for the Ranger Spacecraft, Technical Report 32-335. Jet Propulsion
Laboratory, Pasadena, Calif., Mar. 1963.
30. Evans, D. D., Groudle, T. A,, and Mattson, R. F., Development of the
Ranger Block III Spacecraft Propulsion System, Technical Report 32-829.
Jet Propulsion Laboratory, Pasadena, Calif., Mar. 1966.
31. Schmitz, B. W., Groudle, T. A., and Kelley, J. H., Development of the Post-
injection Propubion System for the Mariner-C Spacecraft, Technical Report
32-830. Jet Propulsion Laboratory, Pasadena, Calif., Apr. 1966.
32. Mosley, V., et al., “Development of the Monopropellant-Hydrazine Propul-
I
sion System for Intelsat-3,” Paper L67-60 in Proceedings of the Ninth Liquid
1 Propulsion Symposium, St. Louis, Mo., Oct. 25-27, 1967, sponsored by the
Interagency Chemical Rocket Propulsion Group.
33. Groudle, T. A., “Sterilizable Monopropellant System,” in The Deep Space
Network, Space Programs Summary 37-47, Vol. 111, Jet Propulsion Labora-
tory, Pasadena, Calif., Oct. 1967.

JPL TECHNICAL REPORT 32- 7 227 21