N A S A TECHNICAL NASA TM X-52404

MEMORANDUM

<
z Hard COPY (HC)

Microfiche (MF)

ff 653 J d V 65

EXPLORING I N AEROSPACE ROCKETRY

17. ROCKET TESTING AND EVALUATION IN GROUND FACILITIES - -.-

by John H. Povolny
Lewis Research Center
Cleveland, Ohio

Presented to Lewis Aerospace Explorers

Cleveland, Ohio
1966-67

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION * WASHINGTON, D.C. * 1968

 EXPLORING I N AEROSPACE ROCKETRY

17. ROCKET TESTING AND EVALUATION IN GROUND FACILITIES

John H. Povolny

Presented to Lewis Aerospace Explorers

Cleveland, Ohio
1966- 67

Advisor, J a m e s F. Connors

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

 D.
NASA
Technical
Chapter Memorandum
1 AEROSPACE ENVIRONMENT
John C. Evvard . . . . . . ...................... X-52388
2 PROPULSION FUNDAMENTALS
J a m e s F. Connors . . . . . ..................... X-52389
4 3 CALCULATION OF ROCKET VERTICAL-FLIGHT PERFORMANCE
John C. Evvard . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-52390
4 THERMODYNAMICS
Marshall C. Burrows ........................ X-52391
5 MATERIALS
William D. Klopp ........................... X-52392
6 SOLID-PROPELLANT ROCKET SYSTEMS
Joseph F. McBride . . . . . . . . . . . ............... X-52393
7 LIQUID-PROPELLANT ROCKET SYSTEMS
E. William Conrad . . . . . . . . . . . . .............. X-52394
8 ZERO-GRAVITY EFFECTS
William J. Masica . . . ....................... X-52395
9 ROCKET TRAJECTORIES, DRAG, AND STABILITY
Roger W. Luidens . . . . . . . . . . . . . . . . . ......... X-52396
10 SPACE MISSIONS
Richard J. W e b e r . .......................... X-52397
11 LAUNCH VEHICLES
Arthur V. Zimmerman ........................ X-52398
12 INERTIAL GUIDANCE SYSTEMS
Daniel J. Shramo . . . . . . . .................... X-52399
13 TRACKING
John L. Pollack. ........................... X - 52 400
14 ROCKETLAUNCHPHOTOGRAPHY
William A. Bowles . . . . . . . ................... X - 5240 1
15 ROCKET MEASUREMENTS AND INSTRUMENTATION
Clarence C. Gettelman . . . . . . . . . . . . . . . . ........ X - 52 402
16 ELEMENTS O F COMPUTERS
Robert L. Miller . . . . . ...................... X-52403
17 ROCKET TESTING AND EVALUATION IN GROUND FACILITIES
John H. Povolny. . . . . . . . . . . . . . . . . . . . . . . . . . .. X - 52 404
18 LAUNCH OPERATIONS
Maynard I. Weston .......................... X-52405
19 NUCLEAR ROCKETS
A. F. Lietzke. . .. ........................ X-52406
20 ELECTRIC PROPULSION
Harold Kaufman. . . ......................... X-52407
21 BIOMEDICAL ENGINEERING
KirbyW. Hiller. . . . . . ...................... X-52408

iii
 17. ROCKET TESTING AND EVALUATION IN GROUND FACILITIES
by John H. Povolny*

Rocket engines and vehicle stages must operate in a variety of environments. Some
components need to perform well in space, others must be effective on the launch pad,
still others must respond during atmospheric flight, but many need to function satisfac-
torily under all conditions from launch throngh orbit. Of these conditions, vibration,
pressure, vacuum, temperature, humidity, mechanical stresses, and gravity forces are
the most important ones affecting performance. Before NASA will commit any engine or
other component to flight, they must be s u r e that it will perform perfectly. To achieve
this, extensive testing is necessary. Ideally, test facilities for this purpose should be
able to reproduce many of these environmental factors at the same time, but, practically,
this is seldom possible, so the effects of environment a r e usually examined one or two at
a time, and testing is often limited to those considered most significant.
Although the investigations usually range from tests of the smallest component to
tests of the complete system in a simulated environment, this discussion ignores the
smaller research setups and concentrates on the larger test facilities used by NASA at
the Lewis Research Center.

AMBIENT FACILITIES

Back in the early 1940's, when rocketry became a serious study, engine research
and development facilities consisted primarily of small (several hundred pounds thrust
capacity), horizontal o r vertical, sea-level test stands such as the.one illustrated in fig-
u r e s 17-1 and 17-2. Then there w a s so much to learn about the fundamentals of rocket
propulsion that these small-scale rigs were satisfactory. In fact, small test stands are
still useful for basic research purposes. As the size of the engines increased, larger,
vertical, sea-level test stands were built, such as the one illustrated in figure 17-3. This
facility, located at the Lewis Research Center, will support experimental rockets having
thrusts up to 50 000 pounds and using exotic propellants such as liquid hydrogen and liquid
fluorine. The largest test stand built to date f o r liquid-propellant systems is for the M-1
engine and is located in Sacramento, California; the largest for solid-propulsion systems
is for the 260-inch-diameter engine and is located near Homestead, Florida. The stand
*Chief, Engine Research Branch.
 1
Figure 17-1. - Simple rocket thrust stand.
' CS-33832

Figure 17-2. - Small sea-level thrust stand.

2
a,
.-m
S

S
a,
L
0
n
2
a,
0
vt
-
u
._
5
3
V
m
S
c
VI

c
.VI
,-
a,
I
a,
u
%
.

3
 Figure 17-4. - M-1 rocket test complex.

Figure 17-5. - M-1 rocket test stand.

4
 Figure 17-6. - 260-Inch-solid-rocket test stand.

and complex for the M- 1 engine, which develops 1.5 million pounds of thrust, is illustra-
ted in figures 17-4 and 17-5, and the stand for the 260-inch engine, which will develop
5.0 million pounds of thrust, in figure 17-6. The two stands are basically different in
that the liquid-rocket stand consists of a tower from which the engines are fired down-
ward, while the solid-rocket stand is a hole in the ground from which the engines are fired
upward. The reason for this is that the solid engine performance is not influenced by
gravity, and thus it can be fired in any attitude; furthermore, it is cheaper to dig a hole
in the ground than to build a tower.

ALTITUDE FACILITIES

The facilities discussed so far are only useful for first-stage engines or engines
which operate where altitude o r space effects a r e not significant. Where this is not true,
as in the case of upper-stage engines o r engines with large-expansion-ratio exhaust
nozzles, then high-altitude facilities are required. There are various ways of simulating
5
 (a) Without flapper valve.

FLAPPER
I I VALVE 7

. ~ CD-9430 -

(b) With flapper valve.

Figure 17-7. - Rocket-exhaust ejector.

the desired altitudes; one of the simplest and least expensive is illustrated in figure 17-7.
In this case, the entire test stand is enclosed in a tank which has one end left open so that
the rocket exhaust can escape. The opening is fitted with a cylindrical tube called an
ejector, which utilizes the energy of the exhausting gases to reduce the pressure in the
tank. Pressures approaching 1 pound per square inch absolute, corresponding to an
altitude slightly over 70 000 feet, have been obtained by this method. Although this
technique provides altitude simulation once the engine is operating, it cannot simulate a
high altitude for testing engine starting characteristics. This can easily be remedied,
however, by adding a flapper valve to the exit end of the ejector tube and evacuating the
system. When a high-altitude start is to be made, the vacuum pump is turned on and the
pressure in the tank and ejector tube is thereby reduced, while the higher atmospheric
pressure pushes on the outside of the flapper valve and gives a tight seal. When the
desired pressure condition is achieved, the engine is ignited; exhaust from the engine

6
forces the flapper valve open and the operation is the same as before. If higher altitudes
are required during engine operation, they are made possible by the addition of a steam
ejector pump o r by the installation of the entire engine and rocket exhaust ejector assem-
bly inside a vacuum chamber.
The steam ejector pump is the method used at the B-1 facility located at the NASA
Plum Brook Station (fig. 17-8). This installation has a vertical test stand, 135 feet high,
currently capable of testing hydrogen-fluorine rockets with thrusts up to about 6000 pounds;

(a) Overall view of test facility. (b) Test enclosure. (c) Engine cross section.
Figure 17-8. - B-1 test facility.

with some modification, it can accommodate engines with thrusts up to 75 000 pounds.
The test engine is installed with the exhaust discharging down at about the 68-foot level,
leaving a space above the engine for a 20 000-gallon propellant tank. This arrangement
allows testing the propulsion system of a complete stage. Run time is limited to several
minutes by the capacity of the propellant tanks o r by the capacity of the storage system
that supplies steam to the ejectors. The B-1 facility has no vacuum chamber for com-
pletely enclosing the rocket engine.
The vacuum chamber is used to simulate altitude at the Propulsion Systems Labora-
tory (PSL) at Lewis. Rocket engines installed in the PSL are illustrated in figures 17-9
to 17-11. The Centaur engine shown in figure 17-9 is using the PSL tank itself as the
vacuum chamber and the flame tube as the exhaust ejector. The hot gases leaving the

7
 FUEL TANK OXIDIZER TANK

AFERBURNING
02 TANK FARM
VENTS

CD-9432

Figure 17-9. - Sketch of Centaur engine installed in Propulsion Systems Laboratory.

TEST CHAMBER
LALTITUDE

ROTATING
EXHAUST
EQUIPMENT-

EXHAUST EJECTOR

PROPELLANT SERVO-
CONTROL VALVE

TEMPERATURE CONTROLLED CD-9433

PROPELLANT ENCLOSURES
Figure 17-10. - Sketch of engine with exhaust ejector.

8
 Figure 17-11. - Engine being installed in exhaust ejector in Propulsion Systems Laboratory.

flame tube are discharged into an evacuated system where they are first cooled and then
removed by several banks of high-capacity pumps. Although satisfactory for many inves-
tigations, the vacuum obtainable by this method is limited by leakage through the PSL tank
hatch. When the ultimate in vacuum is desired, as f o r a large-expansion-ratio rocket
nozzle program, the engine is completely enclosed within an exhaust ejector as well
(figs. 17-10 and 17-11). Engines having up to about 40 000 pounds thrust can be investi-
gated in this facility.

COMBINED ENVlRONMENTS

Engine Testing

Testing rocket engines under a vacuum is significant because the thrust and effi-
ciency of the rocket is determined as much by the pressure acting outside the engine as
by what is going on inside. The latter, of course, is determined by how well the com-
plete propulsion system (consisting of valves, meters, pumps, controls, tanks, etc. )
functions, and this, in turn, is affected by other factors such as the thermal balance
(and ultimately the temperature) of the various components and how long they have been
in space. Obviously, this is of much greater concera for an upper stage that has to

9
function after being in space for some time than it is for the lower stages of a booster.
With this in mind a new facility was designed with the capability of investigating the
effects of thermal factors as well. This facility, which is approaching completion, is
designated as the B-2 Spacecraft Propulsion Research Facility and is located at the NASA
Plum Brook Station. Cutaway illustrations of this facility a r e presented in figures 17-12
and 17-13. Resembling the B-1 facility in that it is downward firing with the engine gases
being pumped by both exhaust and steam ejector, the B-2 differs in having the exhaust
ejector and cooling systems below ground; however the principal difference between the
two facilities is that in the B-2, the complete stage, including the engines, can be ex-
posed to a space environment for as long as desired before firing, whereas the B-1 in-
stallation can only produce a vacuum while the engine is running.
The space environment in the B-2 is simulated in a 38-foot-diameter chamber that
surrounds the test vehicle. The inner wall of this chamber is lined with liquid-nitrogen
panels (-320' F) that simulate the cold of space. Mounted near the inside wall is a n
a r r a y of quartz, infrared lamps that can be used to simulate solar heating. Proper
coordination of these heaters with the liquid-nitrogen system will provide a satisfactory
model of the space thermal environment. The space-vacuum environment that is re-
quired during testing is provided by a four-stage vacuum system that is connected to the
chamber. This system will reduce the chamber pressure t o 5X10m8millimeter of mer-
cury (equivalent to a n altitude of about 200 miles) as long as the engines a r e not opera-
ting. Starting the engines destroys the vacuum and increases the pressure t o an equiva-

10
 Figure 17-13. - Cross section of 8-2 test chamber.

lent altitude of slightly less than 100 000 feet (about 20 miles); this is sufficient, however,
for engine performance evaluation. The actual value of the equivalent altitude obtained
during this phase is a function of engine size and becomes lower as the engines become
bigger. The exhaust system is capable of handling total engine thrusts up to about
100 000 pounds for periods as long as 6 minutes.

Component Testing

In addition to rocket engine testing, space facilities in which the engines are not
fired can be useful in many ways, such as determining the operating temperatures of

11
 Figure 17-14. - Cross section of Space Power Chamber.

various components after a period of exposure o r checking the function of electrical or

mechanical components such as a power generating system, a guidance system, or the
separation of a nose cone or insulation panels. One such facility that has been useful to
the Centaur Project is known as the Space Power Chamber (SPC). This facility
(fig. 17-14) was created by partitioning off and modifying a section of an old altitude wind
tunnel and by installing liquid-nitrogen panels, solar heat simulators, and high-vacuum
pumping equipment.
During space environment tests conducted on a complete Centaur stage in one end of
this chamber, all the systems were actuated except for firing of the engines. Even the
telemetry system was exercised, with data being transmitted to the Lewis telemetry
station located in another building. A subsequent comparison of flight thermal data with
that obtained in the test chamber showed excellent correspondence.
This chamber was also used f o r jettison tests of the Centaur nose fairing. In this
case a real Centaur nose fairing with all its flight systems was installed in the opposite
end of the chamber. During these tests an altitude of 100 miles was simulated, and
although the nose cone had been successfully tested a number of times at sea-level pres-
sure, it was not able to take the higher forces that were generated when the separation
occurred in a vacuum. Needless to say, a redesign was required. When the redesigned
nose cone was finally flown, a comparison of the flight data with that obtained in the vac-
uum chamber again showed good correspondence.

12
 Figure 17-15. - Cross section of Space Propulsion Facility.

CONCRETE ENCLOSURE

DOME 1-1/4*' 5083 ALUMWUM PLATE

TON TROLLEY CRANE WITH l/8" 3003 ALUMINUM CLADDING INSIDE

1/4" ALUMINUM LINER PLAT

38-09, CONCRET

WALLS & FLCQR 7/8" 5083 ALUMINUM PLATE

WITH 1/8" 3003 ALUMINUM C L A D m G INSIDE

TEST CHAMBER ASSEMBLY AREA

DISASSEMBLY AREA
(SPACE SIMULATION)

MBER DOOR CIXISED CHAMBER W O R OPEN

VACUUM PUMPS

CD-8391

Figure 17-16. - Cross section of Space Propulsion Facility test chamber.

13
 Another Lewis space environmental facility of note is known as the Space Propul-
sion Facility (SPF), which is under construction at the Plum Brook Station and will be put
into operation in early 1968. This facility, which is illustrated in figures 17-15 and
17-16, differs from the preceding one in that it will be used to test nuclear power genera-
tion and propulsion systems as well as larger, chemically propelled vehicles and space-
craft. The SPF will have an aluminum test chamber (fig. 17-16), 100 feet in diameter
and 1212 1feet high, surrounded by a heavy concrete enclosure for nuclear shielding and
containment. It will have facilities for assembly and disassembly of experiments and
will be able to vibrate the system within a vacuum environment (ultimate capacity
6X10-8 mm Hg). It will also have experiment-control and data-acquisition systems.
Rdther than building in a thermal simulation system of heaters and cryogenic panels,
these systems will be built for the particular experiment being conducted. The facility,
of course, is designed to comply with all the AEC safety regulations applicable to reac-
t o r s as large as 15 megawatts. The concrete shielding walls are approximately 6 feet
thick so that the radiation levels experienced by people working nearby will b e less than
the levels specified by AEC. This is one of the most advanced space environmental cham -
b e r s under construction and should be useful in future investigations.

STRUCTURAL DYNAMICS

In addition to the large facilities for evaluating the effects of the space environment
on upper stage and propulsion system performance, large facilities a r e also necessary
for determining the structural characteristics and capabilities of complete boosters. The
reason for this is that, although it is generally possible to calculate the natural frequen-
cies of the first and second bending modes of a complete launch vehicle as well as the
dynamic loads that would be encountered at these frequencies, it is impossible to cal-
culate these for the higher modes. Calculating the damping of the vehicle is also im-
possible. Further, there are additional factors such as the interplay between the pro-
pellant system and the structure which cannot be computed and which have a significant
effect. Thus, the surest way to assess the structural capabilities of a vehicle is to test it
on a dynamic test stand like that which has been successfully used for the Atlas-Centaur-
Surveyor vehicle. This stand (fig. 17-17) is known as the E-stand and is located at the
Plum Brook Station. A s illustrated in figure 17-18, the method of installation is to sus-
pend and position the complete vehicle by means of springs with natural frequencies (in
combination with the masses involved) lower than those of the vehicle so that it can re-
spond to the electrodynamic shaker without being influenced by the suspension and posi-
tioning systems. No environmental factor is simulated in these tests other than the dy-
namic force inputs.

14
 Figure 17-17. - E-stand.
HORIZONTAL STABILIZATION

LY STY RENE BALLS

CENTAUR LH2 TANK
WATER IN CENTAUR
LOX TANK

i
LOAD CELL 2
HYDRAULIC CYLINDER^
WATER IN ATLAS

SUSPENSION SYSTEM;
1
28- INCH- DIAMETER
STEEL CABLE
(AT FOUR PLACES) I
i'll

Y
WATER IN ATLAS

,-I-BEAM SUPPORT FRAME

(ALONG X- AND Y-AXES)

15 000-POUND-FORCE SETDOWN STANCHIONS

ELECTRODYNAMIC SHAKER
SEISMIC MASS CD-8359

Figure 17-18. - Atlas-Centaur installed in E-stand.

15
 Perhaps the following brief discussion will explain the nature of the problem better.
Chapter 11 mentioned that the performance of a booster system is highly dependent on its
weight , with the lighter systems having superior performance. Accordingly, structural
weight is kept to a minimum and usually ranges between 6 and 10 percent of the total
launch weight f o r the better systems. In addition, minimum drag requirements for the
flight path through the atmosphere dictate a long slender vehicle. The result is a
highly elastic vehicle with a continually changing natural frequency that is caused by the
m a s s change due t o propellant consumption and varying G-forces during flight. The prob--
lem is Purther complicated by the marly different disturbances and forces that can be en-
countered:
During engine ignition
By the sudden launcher release at lift-off
By the ground winds
By the high altitude gusts (jetstream)
By vectoring the engines
A s a result of coupling between the engine, propellant system, and structure
By sloshing of the propellants
During engine shutdown
During separation of the stages
During insulation-panel o r nose -cone separation
A s a result of the aerodynamic and shock wave pressures generated during flight
through the atmosphere
By the firing of attitude control engines
These forces, acting singly or in combination, can produce one o r more of the fol-
lowing types of deflection of the vehicle:
(a) Lateral - where the vehicle is deflecting normal to its centerline axis (bending)
(b) Longitudinal - where the vehicle is deflecting parallel to its centerline axis
(becoming alternately shorter and longer); this can be either a nonreinforced
or a reinforced oscillation which is augmented by the engine and propellant
systems (called pogo) which results in much greater deflections
(e) Torsional - where the vehicle is rotating about its centerline axis in alternately
opposite directions
Generally one o r more modes of each type of deflection may develop during a flight,
so the vehicle should be tested through at least the third mode, if possible. Inasmuch

-
as a vehicle in-flight is in free-free condition (no restraint at any point), the character-
istic free-free deflection curves are used to define the modes of oscillation. Thus, a
vehicle deflection curve that looks like defines the first mode, one that looks
like \ the second mode, and one like the third.

16
(a) Overall view following test. (b) Closeup of wrinkle patterns.

Figure 17-19. - Atlas tested to ultimate load capacity.

17
 In addition to the mode and type of deflection, another factor that must be considered
is the amount of damping inherent in the vehicle. If it is equal to or greater than the
critical damping, then a single, suddenly applied load will not make the vehicle oscillate
at any of its natural frequencies. If it is less than critical, then the vehicle will oscillate
at one of its natural frequencies but with a decreasing amplitude as follows: rv\rvv--
Vehicle damping is a difficult thing to predict and is best revealed by a full-scale experi- .
ment o r by comparison with similar vehicles for which it is known.
A complete determination of the structural characteristics of a rocket booster in
flight is a complex affair. The engineering approach that is generally employed is as
follows: First, the structural equations defining the vehicle deflection modes at any
point in time a r e derived (with the use of the spring-mass method), then the damping is
estimated, and finally the effects of all the various disturbances are calculated. The
vehicle is then tested in a stand similar to the E-Stand, and the experimental results are
compared with those predicted. If they are the same, that is fine, but if not, then the
equations must be modFfied until they represent the actual event. Once agreement is
obtained, then flight performance can be reliably predicted.
In addition to dynamic response, the E-Stand is also valuable for determining the
ultimate load capability of a launch vehicle. An experiment of this type was conducted on
an Atlas booster (fig. 17-19) which revealed that the ultimate load capability of the Atlas
w a s about 50 percent greater than had been previously assumed. This is a significant
result because it means that the Atlas still has a substantial growth potential for future
space missions.

RELIABILITY AND QUALITY ASSURANCE

All the foregoing discussion of facilities and environmental testing is concerned pri-
marily with the performance evaluation of complete propulsion systems and stages.
Every stage, of course, is made up of thousands of parts (over 300 000 in the Atlas), and
it is difficult to ascertain that all these parts will satisfactorily function at one time, so
that the intended mission will be successful. This was recognized as a problem a r e a in
the aerospace industry in the early 19507s, and it was then that reliability and quality
assurance engineering, as known today, began. It combines the elements of engineering,
statistics, and good sense for evaluating the probability that a given system, subsystem,
o r part will perform its intended function for a specified time under specified conditions.
The reliability field can b e broken down into two basic areas: (1) design goal reli-
ability and (2) use o r operational reliability. During the design of a component an esti-
mate can be made of its reliability if the reliabilities of the individual p a r t s are known.
This can be calculated from the mathematical expression for individual reliability,

18
 R = e-t /MTB F

where R is the reliability (or probability of success), e is a constant, t is the mission

time, and MTBF is the mean time between failures or operating hours divided by num-
b e r s of failures. For the more complex case where the failure of any one part will cause
II failure of the entire component, the total reliability equation is

Rsystem = RIXR2XR 3X...%

It is thus evident that for high system reliability it is necessary to have extremely high
part reliability.
Once the component has been built, it is still necessary to evaluate its reliability
experimentally because manufacturing and assembly processes vary and also because the
environment that the parts experience in this component may be somewhat different than
that for which they were designed. This is usually done in a series of design evaluation
and proof tests. If the failure rates from these tests are plotted against total operating
time (for all components) a curve similar to the one in figure 17-20 is usually obtained.

23 t A R L V.+U
FAILURE
PERIOD
"3 LW
+ E A R 0 UT -.+
PERIOD

O P E R A T I N G TIME-

Figure 17-20. - Idealized failure rate curve.

The high failure rate that usually occurs in the early period is generally a result of
initially poor parts, marginal design, or both. The high rate that is obtained in the later
period is usually a result of wearing out. The fairly low, constant rate that falls
between the two high rates is defined as the useful life. For a high reliability, the useful
life should be long compared with the time a part has to operate, and the failure rate
during the useful life should be as low as possible. Inasmuch as testing is the primary
indicator of reliability, the more tests that are run and the more data that are obtained,
the more confidence there will be in the results. Confidence can be reduced to a statis-
tical value which reflects the degree of probability that a given statement of reliability is

\
19
correct. Of course, in order to achieve and maintain a given reliability, it is necessary
to originate designs with sufficient operating margins, provide specifications for the pro-
cesses as well as the finished parts, and enforce a comprehensive system of quality con-
t r o l o r assurance. Constant vigilance and attention to detail is the price of high reliabil-
ity.

20 NASA-CLEVELAND, OHIO E-3364- 17