The Problem of Escape from the

E a r t h by Rocket
FRANK J. MALINA* AND MARTIN SUMMERFIELDt
California Institute of Technology

ABSTRACT can achievements in the field of rocketry, as exemplified

The problem of imparting to a given pay load the necessary by the V-2 long-range rocket and the "WAC Corporal"
velocity to escape from the earth is re-examined in the light of high-altitude rocket,10 justify the need for a re-evalua-
recently released information on wartime rockets. T h e funda- tion of the feasibility of escape. I t is on this problem
mental equations of motion of a rocket in outward radial flight that the present paper is concentrated.
are derived and the influence of each of the following design
parameters is examined: c, the effective jet velocity; {', the ratio In general, two methods of attaining the necessary
of propellant mass to initial mass; tVt the time of powered flight; high velocities have been considered: the gun and the
M, the ratio of initial mass t o maximum cross-sectional area; and rocket. Previous investigations have conclusively
C^, the drag coefficient based on the same cross-sectional area.
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

demonstrated, on theoretical grounds, that the maxi-

The exhaust velocities attainable under standard conditions mum velocity that can be imparted to a projectile
with rocket propellants range between 6,200 ft. per sec. for nitric
acid-aniline and 10,200 ft. per sec. for oxygen-hydrogen, the for-
using a gun of unlimited length and propelled by the
mer being representative of dense propellants and the latter of most energetic chemical propellants possible is insuf-
light propellants. On the basis of the design characteristics of ficient to enable it to escape from the earth. The veloc-
the V-2 rocket, it is believed t h a t the maximum practical value of ity attainable by a rocket, on the other hand, ap-
£ t h a t may be realized is about 0.85. Therefore, it is concluded proaches that required for escape provided the mass
t h a t single-step rockets utilizing propellants t h a t develop their
energy b y chemical reaction cannot escape from the earth.
ratio, i.e., the ratio of the contained propellant weight
Escape m a y be accomplished b y single-step rockets, however, if to the gross weight of the rocket, is made sufficiently
nuclear sources of energy can be used to heat gases of low mole- high. Furthermore, by means of multiple-step rockets,
cular weight in order to provide much higher exhaust velocities. the final velocity can be increased even more. On the
The analysis is extended to calculate the velocity achieved basis of data on the exhaust velocity obtainable with
by a multiple-step rocket. The overall mass of the rocket and the
different propellants, as presented below, the conditions
maximum velocity of the last step are related to the number of
steps and to the ratio of the masses of successive steps. required for escape have been calculated. I t is con-
I t is shown t h a t the necessary velocity can be attained with cluded that it is now within engineering feasibility to
any of t h e rocket propellants available b y proper choice of the propel a body with the required velocity.
number of steps. Various examples of multiple-step rockets
capable of projecting instrumental pay loads away from the earth
are presented. NOTATION

a = acceleration of rocket a t any instant

1. INTRODUCTION A — projected area of rocket
c = effective exhaust velocity of rocket jet = F/m

M AN HAS LONG SPECULATED on the possibility of

escaping from the earth, or at least of projecting
bodies from the earth toward other bodies in space.
ca
Ca
D
— average value of c during powered flight
= drag coefficient based on area A
= drag due to air resistance = C&A (1/2pv2)
F = thrust of rocket
Many papers have been written on various aspects of g = acceleration due to gravity a t any point
the subject, including the problems of navigation among go = acceleration due to gravity a t earth's surface
the planets and questions of what may be discovered G = overall mass ratio of step rocket = Mo (1) /Mi
among the heavenly bodies. h = altitude above surface of earth
Of all these questions the problem to be considered K = ratio of Q of rocket to t h a t assumed in Fig. 3
m — mass ejected from rocket per unit time in exhaust jet
first is the means of escaping from the gravitational M = mass of rocket a t any instant
field of the earth. Calculations on various schemes have Me = mass of empty rocket (pay load and propellant ex-
been previously presented by Ziolkowsky,1 Goddard, 2 cluded)
Lorenz,3 Senftleben,4 Oberth,5 Rynin, 6 Campbell,7 Mp = mass of propellant in rocket
Ley,8 Ackeret,9 and others. Recent German and Ameri- Mi = mass of p a y load carried b y rocket
Mo = initial gross mass of rocket = Me + Mv + Mi
Presented a t t h e Sixth International Congress for Applied Me- M</n) = empty mass of wth-step of multiple-step rocket
chanics (Section II) a t Paris, France, September 22-29, 1946. M P (fl) = mass of propellant contained in « t h step
Received February 24, 1947. ilfo(w) = initial mass of n t h step = Me™ + Mp™ + ikf0<n+1>
* On leave of absence with United Nations Educational Scien- n = ratio of acceleration of rocket to go
tific and Cultural Organization (UNESCO), Paris, France. N = number of steps in multiple-step rocket; iVth step is
f Chief, Rockets & Materials Division, J e t Propulsion Labora- last t o fire
tory. r ' = distance from center of earth
471
 472 JOURNAL OF THE A E R O N A U T I C A L S C I E N C E S - A U G U S T , 19 47

— radius of earth celeration due to drag is so small that the velocity loss
S = velocity ratio = (vN + gotv) /ca over the interval of traverse of the atmospheric layer
l = time can be neglected. That this size effect is favorable can
h = duration of powered flight
= velocity of rocket at any instant
be seen from the fact that, for similar bodies of various
v
= escape velocity sizes, the air resistance is proportional to the square of
= velocity of rocket a t end of firing of nth step any dimension whereas the mass varies as the cube of
« structural factor = ' Me™/(M6<» + Mp<») the same dimension. The deceleration, therefore, is in-
A = pay-load ratio = Mi/M^N) versely proportional to the linear size of the rocket.
M = gross mass per unit projected area of rocket = MQ/A
= atmospheric density at sea level
This question is treated quantitatively in Section 3.
Po
= ratio of atmospheric density to standard sea-level The peripheral speed of the earth's surface at the
density equator is 1,530 ft. per sec, less than 5 per cent of the
= propellant loading ratio = Mp/M0 = (1 - X)(l'— e) escape velocity at the surface of the earth. It is con-
sidered permissible to neglect this quantity in the pres-
2. REQUIREMENTS FOR ESCAPE FROM THE EARTH ent analysis, although a more refined trajectory calcula-
tion would perhaps make use of some of this initial veloc-
The familiar problem of the motion of a body in the ity.
earth's gravitational field can be analyzed to derive the The effect of the gravitational attraction of the moon
value of the necessary velocity for escape. The ac- on a rocket leaving the earth in a direction away from
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

celeration of gravity at any distance from the earth's the moon can be neglected. The mass of the moon is
center is given by the following equation : Vsi that of the earth, and its distance is 60 times the
g = go(re*/r*) = go[re2/(re + h)*] (1) radius of the earth; thus the initial potential energy of
the rocket due to the moon's field is only 0.02 per cent
The minimum velocity of escape at any distance is that due to the earth. In the case of a rocket directed
calculable from the following equation, on the basis of toward the moon, the proximity of the moon is sufficient
conservation of energy: to reduce the required initial kinetic energy by 2 per
cent, or the required escape velocity, by 400 ft. per sec.
7 2 v* = fr gdr = fr go(re2/r*)dr = g0(re2/r) (2)
The effects in both cases are considered negligible in the
ve = reV2g0/(re + h) (3) computations in this paper.

The radius of the earth at the equator is 21.0 X 106 3. PERFORMANCE PARAMETERS OF THE ROCKET
ft., and the corresponding acceleration of gravity, cor-
In this section the basic performance parameters of a
rected for the centrifugal effect of the earth's rotation,
rocket launched vertically from the earth's surface will
is 32.2 ft. per sec. per sec. Inserting these values in
be discussed. The parameters have been studied in
Eq. (3), the graph shown in Fig. 1 is obtained. The re-
connection with the sounding rocket. 11,12 However, for
quired velocity for escape at sea level is 36,700 ft. per
purposes of the present investigation, a further con-
sec.
sideration of their significance is desirable. An analysis
The effect of air resistance is neglected in Eq. (3),
of only the powered flight of the rocket is required. I t
The curve of Fig. 1 needs modification because of air
will be assumed that the propulsion system delivers a
resistance only within the height of the atmospheric
constant thrust obtained from the ejection of mass car-
layer, about 50 miles, if at all. However, for rockets of a
ried wholly in the rocket. The energy of the ejected
size sufficiently large to accomplish the flight, the de-
mass is acquired from either a molecular or a nuclear
process.
During powered flight the mass of the rocket at any
instant is given by the equation
d
UJ M = M0[l - f(*/*,)] (4)
^ 40
u. and the fundamental equation of motion has the form J
o 30
a = dv/dt = (F - D - Mg)/M (5)
£ 20
o Substituting F = mc = (Mp/tp)c, D = Cdp0crA(v2/2),
o and Eq. (4) in Eq. (5), the following relation is ob-
tained :
£ 10
0 10 20 30 4 0 (FT) tc n v21
n0 1.9 3.8 5:7 76 (Mi.) dv tp_ 2\x
DISTANCE FROM EARTH'S SURFACE ( I 0 6 FT.)( I 0 3 MILES) = (6)
dt 1 - Of/tp)
F I G . 1. Dependence of escape velocity on distance from the
earth. and on integrating
 ESCAPE FROM THE EARTH BY R O C K E T 473

- — WPOO" "

t(t/tp)
dt
i: gdt + vo (7)

For the case of flight in empty space> the drag term in

o
LU
CO

\
1 3,000

12,000
v°->
V^3^ ,A
>- /
Eq. (7) vanishes, and on integrating one obtains the 1-
11,000
o s
well-known relation for the velocity of the rocket during '
powered flight if the variation of gravitational accelera- a 10,000
y
> CONDITIONS
tion is neglected, 2 9,000
? =0.8
C =7700 FT./SEC.
v = - c l n [1 - f O 0 g ] - got + vo (8) X 8,000
Vo=0
2 Q z
%
At the end of powered flight when / = tv 7,000
500 1,000 1,500 2,000 2,500
V
Vmax. = -Cln (1 - f) - goV+ 0 (9) //7K = MASS PER UNIT CROSS-SECTIONAL AREA DIVIDED BY

It should be noted that Eq. (9) holds even if the thrust DRAG FACTOR (LBS PER SQ FT. )
acting on the rocket is variable. F I G . 2. Variation of maximum velocity with mass per unit cross-
Returning to Eq. (7), it is seen that the performance sectional area of rocket.
of the rocket is determined by six design parameters—
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

c, f, tv, /x, Cd, and v0) which can be varied by the de-
BASED ON
signer, and by two parameters, g and pQ<x, which are inde- DRAG COEFFICIENT
PRC) J E C T E D , FRONTAL A R EA
pendent of the rocket design. OF BODY OF ROCKET
The influence of the design parameters tp, /*, and Cd LLJ

o A N GLE OF ATTACK = 0

on the maximum velocity that a rocket can achieve for NO JET ACTING
UJ
the case f = 0.80, ca = 7,700 ft. per sec, v0 = 0, and o
o
g = g0 i s shown in Fig. 2. The curves have been com- CD

puted using the drag coefficient given in Fig. 3 and at- <
cr
Q
mospheric density and temperature distributions given
for the N.A.C.A. standard atmosphere. 13 The graph
can be used equally well for rockets having a drag char-
acteristic differing from that in Fig. 3 by a multiplica- ) 1 2 3 4
tive constant. The drag factor K therefore has been MACH NUMBER
introduced to permit the use of other drag coefficient F I G . 3. Drag coefficient of typical rocket.
curves that have a shape similar to the one shown in
Fig. 3. In a like manner, by using a factor equal to 1 1
the ratio of the density at sea level to that at the launch- 60,000
CONDITIONS
t = 6u b t U b .
ing altitude, Fig. 2 can be made to apply approximately
to the flight of rockets launched from high elevations 50,000 V°
where the atmosphere density is reduced. 9 = 9o
40,000
In Fig. 4 the effects on the maximum rocket velocity 36,700 fESCAPE
1 SEA
VELOCITY AT
LEVEL ^ ^
^

)00 FT./SEC^
of the propellant loading parameter f and of the effec- 30,000
tive exhaust velocity c are shown for the case tv = 30 ——c^=^l 5 , 0 0 0
sec, ix/K = oo (i.e., drag is neglected), v0 = 0, g = go. 20,000
" c ^ l O ^ O O ^ ^
Several conclusions can be drawn from the curves in
10,000
Figs. 2 and 4. First, the shorter the duration of firing, cn = 6 , 0 0 0

the larger will be the velocity of the rocket at the end of 0.6 0.7 0.8 0.9 1.0
burning. This effect results from the decreased loss
PROPELLANT LOADING RATIO (?)
caused by the constant deceleration due to gravity.
This principle cannot be applied to rocket design with- F I G , 4. Dependence of velocity at end of burning on propellant
loading ratio.
out considering the simultaneous effects of the increased
mass of the propulsion system required for a short firing
duration and the increased structural mass arising from geometrical shape propelled in the same trajectory, the
high acceleration loads. Second, the greater the propel- parameter \x is proportional to the product of the average
lant mass loading ratio f, the higher will be the velocity density and the length of the rocket. Consequently,
at the end of burning, and, furthermore, as f approaches the loss in velocity due to drag is inversely proportional
unity, the terminal velocity becomes infinite. Third, to the length of the rocket. Fig. 2 shows that for tv =
the loss in velocity due to atmospheric drag diminishes 30 S ec, f = 0.8, and c = 7,700 ft. per sec, the loss is 10
as the value of M increases, the loss being inversely pro- per cent at y, = 650 lbs. per sq.ft., and 3 per cent at fx =
portional to ix [cf. Eq. (7)]. For rockets of similar 2,000. An acid-aniline rocket having an average den-
 474 JOURNAL OF THE A E R O N A U T I C A L S C I E N C E S - A U G U S T , 194 7

TABLE 1
Properties of Several Propellant Combinations
Mixture Calculated Exhaust Average
Ratio Combustion Average Velocity at Exhaust Propellant
Propellant Oxidizer/- Temperature Density Sea Level Velocity Loading
Combination Fuel) (°F.) (Lbs./Cu.Ft.) (Ft. per Sec.) (Ft. per Sec.) Ratio
Nitromethane (monopropellant) 3,950 71.0 6,150* 7,240 f 0.995{
Red fuming nitric acid-aniline 5,070 85.9 6,200 7,300 0.995
Hydrogen peroxide (87%)-methyl alcohol 3.6
4.0 4,160 75.5 6,300 7,420 0.995
Oxygen (liquid)-ethyl alcohol 1.5 5,260 60.4 6,870 8,090 0.991
Oxygen (liquid)-hydrogen (liquid) 3.0 4,390 14.8 10,150 12,000 0.960
* The sea-level exhaust velocities are calculated on the basis of a steady combustion pressure of 300 lbs. per sq.in. abs. and an exit
pressure of 14.7 lbs. per sq.in. abs. An optimum nozzle area ratio is assumed (about 4.0). The theoretical values have been reduced
by 12.5 per cent to allow for imperfect combustion and imperfect flow.
f The average exhaust velocities are 18 per cent greater than the sea-level values. This is the approximate average increase due to
reduced atmospheric back pressure during an assumed vertical flight.
J The calculated propellant loading ratios listed in the last column are the minimum values required to enable a rocket to attain the
escape velocity. A constant acceleration of 20g0 is assumed, and drag is neglected. (See Section 4.)

sity of 50 lbs. per c u i t . will suffer a loss of only 3 per neglected during t h e period of powered flight. A treat-
cent in terminal velocity due to drag if it is about 40 ft. m e n t in which the variation of gravitational accelera-
long. This conclusion is significant in the later con- tion is taken into account has been reported. 1 2 I t can
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

siderations of rockets designed to escape from the earth, be shown t h a t the corrections are negligible in t h e cases
since t h e calculations show t h a t such rockets are large examined below. Second, the retardation due to drag
enough so t h a t t h e neglect of t h e drag term is justified. will be considered negligible since the size of t h e con-
In Table 1 t h e practical effective exhaust velocities for templated rocket will be extremely large. Third, t h e
several liquid propellants are listed. I t appears highly initial velocity of the rocket due to t h e peripheral speed
improbable t h a t exhaust velocities higher t h a n those of the earth's surface will be neglected. Finally, the re-
listed in T a b l e 1 can be obtained from molecular reac- quired velocity for escape will be t a k e n as t h a t a t t h e
tions. T h e possibility of obtaining m u c h higher ex- earth's surface, namely, 36,700 ft. per s e c , since it is
h a u s t velocities b y heating gaseous hydrogen with the consistent with t h e first assumption to assume t h a t the
energy released b y a nuclear process has been suggested. altitude a t which t h e propellant is exhausted is small
T h e a d v a n t a g e of using hydrogen as a working fluid is compared with t h e earth's radius.
t h a t , because of its low molecular weight, a large en- T h e equation of motion in this case can be integrated
thalpy drop can be obtained with a temperature n o t ex- as above, regardless of the variation of t h r u s t during the
ceeding t h a t already in use in rocket motors. Calcu- firing period.
lated exhaust velocities for various chamber tempera-
tures are shown in Table 2. = - c m (1 - f) - gotP (10)

F r o m Eq. (10) t h e necessary propellant loading ratio

TABLE 2 to achieve the escape velocity can be determined for
Exhaust Velocities Obtainable with Hydrogen as Working Fluid
each propellant listed in Table 1. First, a consideration
Chamber Heat Exhaust
Temperature Required Velocity of the optimum time of powered flight tp is necessary.
(°F.) (B.t.u. per Lb.) (Ft. per Sec.) I t is evident from E q . (10) t h a t the m a x i m u m velocity
5,000 20,700 23,400 will be obtained when t h e iv = 0; i.e., t h e propellant is
7,000 49,000 29,100
9,000 113,000 36,400 consumed instantaneously. B u t t h e high t h r u s t re-
11,000 152,000 42,300 quired would necessitate a heavy propulsion system,
N O T E : The calculations of the heat required and the exhaust and t h e resulting high acceleration would require a
velocity take into account the dissociation of hydrogen at these heavy structure for t h e rocket, which would seriously
high temperatures. The theoretical values of the exhaust veloc-
ity have been lowered by 5.0 per cent to allow for heat-losses decrease the value of f. A compromise burning time
and imperfect flow. The calculation assumes a chamber pressure m u s t therefore be chosen to obtain the best combina-
of 300 lbs. per sq.in. abs. and an exit pressure of 14.7 lbs. per sq.in.
abs. and flow with constant composition. tion of f and tp.
T h e o p t i m u m combination of f and tv can be deter-
4. S I N G L E - S T E P R O C K E T S D E S I G N E D TO E S C A P E FROM
mined only by considering t h e m a n y factors t h a t enter
THE EARTH
into t h e design of a given t y p e of rocket. T h e choice of
tp determines t h e r a t e of propellant consumption which,
T h e outward radial flight of a rocket starting a t the in turn, determines t h e mass of t h e rocket motor, the
earth's surface with zero initial velocity will now be mass of the hydraulic lines and valves, t h e mass of the
considered with t h e aid of t h e d a t a presented in t h e pre- pumping p l a n t or pressurization system, and t h e mass
ceding section. of t h e structure designed to support these components.
Several approximations will be used. First, t h e varia- Analytically, the relationship m a y be expressed as fol-
tion of t h e gravitational attraction with altitude will be lows :
 ESCAPE FROM THE. EARTH BY R O C K E T 475

f= m (ID LU
C9
where the form of function/ is dependent on the type of <
CC
UJ
rocket being considered. In general, the function is >
such that f increases monotonically with tp and asymp- <
o
totically approaches a limiting value as tv becomes infi- H
"*
O o
nite. By combining Eq. (11) with Eq. (10) for any z V.

specific function f(tP), a unique value of tp can be ob- z: a

cr F
3
tained for which vmax. will be greatest. 0Q >
—-
An alternate approach is usually dictated by the ti.
O > •

maximum acceleration allowable by the particular O

H
O
character of the pay load. In this event, two limiting 2
til
O
_J
UJ
cases are possible: (1) a rocket may be designed to de- h- >
liver a constant thrust, in which case the maximum ac- *<
\-
celeration occurs at the end of burning when the mass >- <0
H
->
of the rocket is least; or (2) the rocket may be designed o
o
<
-L
-J X
to deliver a continually decreasing thrust so as to main- LU UJ
tain a predetermined constant acceleration. In both >
u_
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

cases, the burning time is determined by the accelera- O

1.00
tion selected. O
Y-
For the case of the constant thrust rocket it can be < PROPELLANT LOADING RATIO (?)
shown that Eq. (10) takes the form: FIG. 5. Constant thrust rocket, dependence of performance on
maximum allowable acceleration.

{ h ( 1 -» + (w
i - iy\n + 1/J
(12)
In Fig. 6 the ratio v.maxJc is plotted as a function of f
for various values of n according to Eq. (14). For the
where n = amaxJg^ The minimum value of tv corre- constant acceleration rocket there is no limit to. the
sponding to a given maximum acceleration at the end value of v/c that can be attained regardless of the
of powered flight is: specified magnitude of n. Also, the constant accelera-
tion rocket can attain the same value of v/c, with a
nun. = [f/(l - f ) ] [ * / * o ( n + 1)] (13)
lower value of f than is possible for the constant thrust
In Fig. 5 the ratio vmaxJc is plotted as a function of f rocket. It must be realized, of course, that the con-
for various values of n according to the relationship stant acceleration rocket requires a heavier and more
given in Eq. (12). complicated propulsion system, and that this increase in
It can be seen that the maximum value of the ac- empty mass may actually overcome the advantage in$.
celeration, denoted by n, limits the maximum velocity
that can be obtained. For n = 5, the maximum veloc- UJ
CD
ity of the rocket is approximately equal to the exhaust
velocity of the propellant; for n = 20, the maximum < 4.50
UJ
n = a/g o

i
velocity is slightly more than twice as much. Further- /
4.00
more, for each value of n there is a limiting value of f
above which no improvement in rocket velocity can be 3.50
expected, since in that case the thrust of the rocket does
not exceed its initial weight. The limiting values of f a: 3.00
lie on the dash-dot curve of Fig. 5.
In order to obtain the required high values of vmax./c
it is necessary to tolerate accelerations in excess of 20g
=>
ao
u_
o
Q o
2.50
"\-n=20
^-n= o o //
v\ \

in rockets having propellant loading factors in excess of •z. o 2.00

0.9. It is believed impossible to meet these conditions.
LU
> n=5- \
The second alternative to be considered is the case of 5 1.50
CO
a rocket acted upon by a constant acceleration a = ng0 3
< 1.00
O
during powered flight. For this case Eq. (10) becomes Q X
X
LU
.50
r^
>
/ n
In (1 - f) (14) O
o .65 70 75 .80 .85 90 95 1.00
» The minimum value of tv corresponding to the chosen
< PROPELLANT LOADING RATIO (?)
value of the acceleration is
FIG, 6. Constant acceleration rocket, dependence of perform
U Vmax./ng* (15> ance on allowable*acceleration.
 476 JOURNAL OF T H E A E R O N A U T I C A L SCIENCES-AUGUST, 1947

The optimum thrust-time curve is probably a com- 5. MULTIPLE-STEP ROCKETS DESIGNED TO ESCAPE
promise between the constant thrust case and that de- FROM THE EARTH
signed to produce a constant acceleration. However,
regardless of what curve^is used, it is not possible to The velocity attainable with a single-step rocket is
avoid impractical, high accelerations in the attempt to severely limited by the fact that propulsion energy
obtain high rocket velocities with the propellants listed must be utilized to accelerate the entire empty mass of
in Table 1. the rocket continually even after the bulk of that
empty mass is no longer useful. Earlier investigators
In calculating the performance of a rocket in flight have suggested the use of a rocket made up of a num-
by means of Eq. (10) it is necessary to use the time ber of independent steps, each of which is equipped
average value of the exhaust velocity ca, since the ef- with a propulsion system and a portion of the total
fective exhaust velocity varies with altitude because of propellant load. As each step is exhausted, its empty
the decreasing back pressure at the exit section of the mass is dropped from the rocket, the propulsion of the re-
nozzle. If it is assumed that the chamber pressure of mainder being taken up by the next step. The pay load
the rocket motor is 300 lbs. per sq.in. abs., the effective is carried in the last step of the rocket. In the following
exhaust velocity of the propellants listed in Table 1 can analysis the number of steps and their relative masses
be approximately 50 per cent higher in empty space are considered from the point of view of the overall
than at sea level, provided the* exhaust nozzle is designed mass of the rocket required to project a given pay load
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

with an infinite exit area. Considering the required away from the earth.
weight and size of the nozzle, a fixed and relatively small
A multiple-step rocket is considered in which the
ratio of exit area to throat area is selected. In these cal-
number of steps is N, the pay load of mass Mt being
culations an area ratio of 25:1 is used. The maximum
located in the iVth step. A parameter called the pay-
increase in exhaust velocity over the value at sea level
load ratio can be defined, for each step, as the ratio of
is about 27 per cent. Although a correct evaluation of
the mass of the carried load or pay load to the mass of
the average value ca requires the computation of a
the rocket at the moment that step begins to fire. I t
trajectory for each case, an average increase of 18 per
is assumed in this analysis, and justified below, that
cent is assumed in this paper. Table 1 presents the
the pay-load ratio is the same for each step, as expressed
average values computed in this manner.
by the following equation:
Using these data, the minimum propellant loading
ratio required to attain the escape velocity has been Mi M0(n+1) •Mo ( 2 )
X = 1 < n.< N (16)
calculated for each of the propellants in Table 1. The MJN) M0(n) M0(1>'
maximum allowable acceleration was taken to be 20g,
This assumption simplifies the analysis since all the
and Eq, (14) for flight with constant acceleration was
steps possess equal ballistic properties.
used. The results are presented in the last column of
It is also assumed that all the steps can be designed
Table 1.
with equal structural effectiveness. In accordance with
It is apparent that the propellant loading ratios re- the following equation, the ratio of the empty mass of a
quired for these propellants are all too high to be practi- step to its mass when loaded with propellant (but with
cal. For comparison, the V-2, a typical high-perform- no pay load) is the same for all steps:
ance rocket, has a propellant loading ratio of 0.75 with-
out the pay load. MJn) Me^
,l<n<N (17)
The conclusion can therefore be established that a M0(n) ~ M0(n+l) Me™ + Mv^
single-step rocket utilizing any of the chemical propel-
From Eq. (16) it is readily seen that the mass of the en-
lants heretofore suggested cannot escape from the earth.
tire A'-step rocket is related to the mass of the pay load
A more hopeful performance can be expected of a in the following manner:
rocket propelled by hydrogen heated by a nuclear power
plant. At a chamber temperature of 5,000°F. the ex- Mi = X*JMO(1) (18)
haust velocity at sea level is calculated to be 23,400 ft. Since e and X are the same for each step of the
per sec, and the average value during the entire firing rocket, and if it is assumed that the average value of the
period would be about 27,500 ft. per sec. Under these exhaust velocity applies to each step, then each step
conditions, a propellant loading ratio of 0.73 would per- contributes the same increase in the speed of the
mit escape from the earth, assuming a constant accel- rocket, according to the following equation:
eration of 20g. The corresponding firing time would be
about 60 sec. The required propellant loading ratio 1 (M& + Mo<» + 1}
\ tp
may be reduced to 0.60 if temperatures of the order of Vn - Vn = ln g 0
9,000°F. can be used. A more realistic estimate of the
-^ V w> ) ~ N
performance of such rockets awaits the release of in- (19)
formation regarding the design and operation of nuclear It is assumed in this equation that the deceleration
power plants. due to gravity is constant during powered flight, and
 ESCAPE FROM THE E A R T H BY R O C K E T 477

10 1.0 10 t.o
€ = 0.333
CD
o S=6 o
o
CD C = 0.250
o CD
O
S = 3J
3 —^ .7 ~ Q-R

< £ <
or o cr
co < c
co 5
a: CO
CO
k'H
<
u. 4
S=3
S= 6, y <
o
<
/ .4
S=3
// 1 /

hj
x :>
H
a:
<
CD
1
X
H
cr
<
CD
\A
7=
LL
o o .1
/ _i
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

0 0
6 8 10 12 14 2 4 6 8 10 12 14
NUMBER OF STEPS (N) NUMBER OF STEPS (N)
F I G . 7. Step rocket performance, e = 0.333. F I G . 8. Step rocket performance, e = 0.250.

equal to the sea-level value. The error introduced by 10 1.0

this assumption is small, since the distance at which
CD
burning stops is small compared with the radius of the O
€ = 0.200
earth. This will be illustrated in the numerical ex- CD
O
amples presented below. S !
3^i
By means of Eqs. (16) and (17), and by means of the
S=6
defining equation for M0(n) given in the table of sym-
bols, Eq. (19) can be reduced to the following equation: < 1
tr S =6 / 1
CO
\

y
CO
- ca In [e(l ~ X) + X] go N (20) <
o s« 3 /
It is assumed that the firing of each step commences
1
4
at the instant the preceding step ceases firing and the 2 3
x
2
empty structure is cast away. The velocity of the
rocket at the end of burning is merely N times the in- I
CD
—/
/
crease in velocity of each step. O /
vN = -Nca]n [e(l - X) + X] - g0tp
The effect of atmospheric drag on the first one or two
(21)
0
L _ L /
6 8 10 12 14
0

NUMBER OF STEPS (N)

steps is neglected, since the value of /x for the rocket
F I G . 9. Step rocket performance, e = 0.200.
being considered is at least 1,000 lbs. per sq.ft.
(Fig. 2).
Eqs. (18) and (21) can be put in dimensionless form It is generally considered that the propellant having
by the introduction of two parameters: the highest exhaust velocity is the most desirable.
G = Mo(1) IM% = overall mass ratio This advantage, however, is usually considerably re-
S = (% + gotv)fca = velocity ratio duced by the fact that such propellants have a low den-
sity and therefore necessitate the use of proportionally
G = \~N (22) more tank and structural weight than in the case of the
5 = -Nln [e(l - X) + X] (23) denser, low-performance propellants. The exhaust
velocities and densities of several propellant combina-
In Figs. 7, 8, and 9, G and X are plotted as functions of tions are presented in Table 1. The effect of the in-
N and S for three values of e. creased structural weight has been accounted for,
The parameter 5 is computed from the average ex- in the examples below, by taking e = 0.33 for
haust velocity of the propellant, the burning time of the oxygen-hydrogen rockets and e = 0.25 for acid-aniline
rocket, and the desired final velocity of the rocket. rockets.
 478 JOURNAL OF T H E A E R O N A U T I C A L SCIENCES-AUG'UST, 1947

The burning time of the rocket also enters into the accurate even if practical design considerations demand
value of the parameter 5. For a given type of rocket some degree of inequality in the various values of X
there is an optimum tv that, on the one hand, is small throughout a multiple-step rocket.
enough to avoid an excessive loss in velocity due to Several illustrative cases of multiple-step escape rock- .
retardation by gravity, and, on the other hand, is suf- ets can be calculated by the above method. I t is first
ficiently long not to require an excessive weight of rocket necessary to decide upon the nature and mass of the
motor, propellant feed system, and structure. In the pay load, which, in turn, are dependent on the purpose
cases calculated below, the optimum burning times will of the flight. It is not the intention of this paper to
be estimated on the basis of present experience in the propose a particular mission to be accomplished. How-
design of rocket vehicles. No general analytical ex- ever, the minimum useful pay load may be estimated
pression for the optimum burning time can be derived on the assumption that the purpose of the flight is to
without recourse to specific designs. project away from the earth an instrument consisting
To carry out the performance calculations the proper of a radar beacon transmitter and an electrical pickup
value of € must be estimated. In this case experience for measuring a single quantity, e.g., cosmic ray inten-
with the V-2 rocket and similar devices can be used as sity. Radar signals may be sent to the rocket during
a guide. The gross weight of the V-2, less war head, flight by means of a ground transmitter at least as
is 25,700 lbs. Its empty weight, less war head, is 6,400 powerful as that recently used by the U.S. Army Signal
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

lbs. The value of e for the V-2 is therefore 0.25. In Corps to reflect signals from the moon. The radar
the cases computed below, e is taken to be 0.25 where beacon is intended to "reflect" the signal back to the
the pay-load ratio X is small; and, to account for some receiver on earth, thus providing data on its trajectory.
increase in structure, at larger values of X the value of The values of the measurements made during flight
e is taken to be 0.33. could be incorporated in the return signal from the
One of the assumptions in the foregoing analysis is radar beacon. It has been boldly estimated that, if
that it is best to fire each step immediately after the the utmost care is exerted in the design of the instru-
exhaustion of the preceding step, with no time delay. ment, its weight may be kept to as low as 10 lbs. This-
That this program is the optimum one is evident from figure will be arbitrarily used in the calculations to fol-
the fact that the possibility of escaping from the earth low.
depends on imparting sufficient energy to the last step As a first example, a five-step acid-aniline escape
of the rocket. The energy of the last step is equal to rocket may be considered. As in the case of single-step
the integral, over the duration of firing, of the thrust rockets,"the average value of the exhaust velocity ca can
acting on it multiplied by the velocity. Since the be taken to be 7,300 ft. per sec. Allowing a burning
thrust and the total duration of firing are independent time of 40 sec. for each step, the total burning time tp,
of the magnitude of any intervals between the firing of is 200 sec. The required escape velocity varies with the
successive steps, the maximum energy can be obtained distance from the earth at which the end of burning
only if the firing of each step takes, place at the highest occurs (Fig. 1). In this case, propulsion ceases at about
possible velocity. Therefore, there should not be any 600 miles from the earth's surface where the velocity of
coasting period between steps. escape is 34,300 ft. per sec. The parameter 5 can be.
Another question relates to the assumption in the evaluated in the following manner:
preceding analysis that the pay-load ratio X should be
equal for all steps. A general analytical proof of the ^ve + gotp = 34,300 + 32.2 X 200 =
fact that this is the optimum condition can be derived ca 7,300
by considering any two successive steps of a multiple-
step rocket, and setting up an expression for the dif- T h e use of 32.2 ft. per sec. per sec. instead of t h e t i m e
ference between the mass ratio G when the two values average value of g is conservative, a n d introduces an.
of X are equal and G' for the case of unequal values of X. error of less than 2 per cent in t h e value of S.
To simplify the presentation in this paper, a numerical Assuming a value of 0.25 for t h e structural factor e9t
case can be examined. Let two such steps be designed Eqs. (22) a n d (23) can b e used t o derive t h e pay-load,
to increase the velocity of the rocket by an amount such ratio and the overall mass ratio: X = 0.104, G = 82,300.
that AS = 2.0. Then e = 0.25 and X = 0.157 for each T h e gross mass of t h e rocket is then 823,000 lbs. T h e
step. The overall mass ratio G is calculated to be masses of t h e individual steps, each loaded with propel-
41.0. If two other steps were substituted for these to lant, a r e : 739,000, 75,200, 7,880, 824, a n d 96 lbs., r e -
produce the same value of AS but with unequal values spectively. T h e mass of t h e last step includes t h e 10-
of X, e.g., 0.200 and 0.118, the overall mass ratio would 1b. p a y load. T h e t h r u s t of t h e first step is of t h e o r d e r
be 43.0. Thus, the minimum mass ratio is obtained of 3,000,000 lbs. T h e thrust of each of t h e succeeding;
with equal values of X. It is of interest to note that the steps is in proportion t o t h e mass of each step. Assum-
analytical derivation shows that G increases slowly as ing t h a t t h e average density of acid-aniline rockets is-
the values of X deviate from equality. Therefore, the approximately 50 lbs. per cu.ft., t h e overall dimensions,
mass ratios computed from Eq. (22) are sufficiently of t h e rocket can be estimated: length = 130 ft.;
 ESCAPE FROM' THE EARTH BY R O C K E T

E3
Characteristics of le-Step Escape Rockets
Number Structural Pay-load Firing Pay-load Gross Thrust Estim. Estim.
Example Propellant of Steps Factor Ratio Time Mass Mass First Step Length Body Dia.
No. Combination (N) M (X) (Sec.) (Lbs.) (1,000 Lbs.) (1,000 Lbs.) (Ft.) (Ft.)
1 Acid-aniline 5 0.25 0.104 200 10 823.0 3,000 130 13.0
2 Acid-aniline 10 0.25 0.386 400 10 136.0 350 70 7.0
3 Acid-aniline 00 0.25 1.000 400 10 37.0
4 Acid-aniline 10 0.20 0.425 400 10 52.0 136 51 5.1
5 Oxygen-ethanol 5 0.25 0.155 200 10 112.0 450 71 7.1
6 Oxygen-ethanol 10 0.25 0.432 400 10 44.0 118 52 5.2
7 Oxygen-ethanol 5 0.20 0.207 200 10 26.0 104 44 4.4
8 Oxygen-hydrogen 5 0.33 0.260 200 10 8.3 40 40 4.0
9 Oxygen-hydrogen 5 0.33 0.260 200 100 83.0 400 86 8.6
10 Oxygen-hydrogen 10 0.33 0.532 400 100 55.1 160 74 7.4

diameter of b o d y .= 13 ft. T h e characteristics of t h e By inserting S = 6.15 and e = 0.25, the values used in
rocket are summarized in Table 3. the ten-step rocket, the resulting value of 3,700 is ob-v
Similar calculations m a y be carried o u t for a second tained for G«>. Therefore, in this hypothetical limiting
example, a ten-step acid-aniline escape rocket carrying case, the proposed 10-lb. pay load can be projected away
t h e same 10-lb. p a y load. Although t h e pay-load ratio from the earth by a 37,000-lb. acid-aniline rocket having
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

in this case m u s t be somewhat larger t h a n in t h e five- an infinite number of steps. It can be seen that the
step example, t h e value of e is still assumed to be 0.25. greatest mass reduction was accomplished in going from
T h e firing time of each step is assumed to be 40 s e c ,
five steps (823,000 lbs.) to ten steps (136,000 lbs.) and
making a total period of firing of 400 sec. The-results
that a further increase in the number of steps would not
of t h e computation are t h a t t h e overall mass of t h e re-
be as beneficial.
quired rocket is 136,000 lbs. and t h a t the initial t h r u s t The nitric acid-aniline combination is representative
is 350,000 lbs. These requirements are considerably of the so-called standard propellants. It is interesting
less t h a n those of t h e five-step rocket. T h e weighing of to consider multiple-step rockets that utilize a high-
the advantages of a smaller rocket against the disadvan- performance propellant, such as the liquid oxygen-
tages of t h e greater complexity of an increased n u m b e r liquid hydrogen combination.
of steps is a m a t t e r for t h e exercise of practical engineer- Oxygen-hydrogen rockets possess the distinct advan-
ing judgment. tage of requiring a much lower mass rate of propellant
consumption for a given thrust than acid-aniline. On
This simplified m e t h o d of analysis becomes less satis-
the other hand, the lower average density and the need
factory as t h e n u m b e r of steps increases. T h e calculated
for insulation compel the use of proportionally greater
mass ratio G is extremely sensitive to t h e assumed struc-
"dead" mass of structure, tanks, and propellant feed
tural factor e, a fact t h a t can b e verified b y comparing
system, with the result that the advantage in propellant
Figs. 7, 8, a n d 9. Present experience indicates t h a t e
consumption is reduced. For the oxygen-hydrogen
is dependent on t h e scale of t h e rocket and on t h e b u r n -
case, the value of e is assumed to be 0.33 to account
ing time for which it is designed. I t is considered un-
for this increase in dead mass.
likely t h a t rockets having a mass of less t h a n 500 lbs. or
As an example, a five-step rocket may be calculated.
a firing time of less t h a n 20 sec. can be constructed with
The average exhaust velocity is 12,000 ft. per sec, and
e as low as 0.25. As t h e chosen n u m b e r of steps in-
a total firing time of 200 sec. is assumed. Then 5 —
creases, the masses of t h e latter steps in the series be-
3.40, X = 0.26, and G = 832. If the same 10-lb. pay
come smaller and t h e individual burning times decrease.
load described above is considered, the mass of the
Therefore, although t h e simplified analysis indicates
entire rocket becomes 8,320 lbs., and the initial thrust
t h a t t h e mass ratio continually decreases as the n u m b e r
would be about 40,000 lbs. This reduction in mass as
of steps is increased, this principle m a y b e expected t o
compared with the acid-aniline five-step rocket empha-
break down a t some point. A more correct approach •
sizes the importance of developing practical high-per-
would be to establish, if possible, analytical relations ex-
formance rocket systems. The sharp reduction is a
pressing t h e dependence of e on tp, Mo, and X, to be used
result of the exponential dependence of the mass on the
in conjunction with E q s . (22) and (23). T h e derivation
reciprocal of the exhaust velocity, as expressed in Eq.
of such relations is outside the scope of this paper.
(25).
I t is nevertheless of interest to examine analytically
Obviously the required mass of the rocket assembly
t h e results of increasing t h e number of steps indefinitely
would not be unreasonable even if the pay load were in-
while retaining t h e same values of 5 a n d e. B y elimi-
creased tenfold. A 100-lb. pay load could include a
nating X between E q s . (22) a n d (23) a n d letting N be-
more powerful radar beacon to provide stronger and
come infinite in t h e resulting expression for G, t h e limit-
more accurate signals and a greater number of measur^
ing value of G is obtained:
ing instruments to provide more scientific data from,
G > e[S/a~e)] as N > oo (25) the single flight. The required mass would be 83,200
 480 JOURNAL OF THE A E R O N A U T I C A L S C I E N C E S - A U G U S T , 1947

lbs. and the initial thrust would be about 400,000 lbs. Thus, although many problems of navigation, flight
Possible dimensions for this rocket are given in Table 3, control, instrumentation, etc., remain to be investi-
on the basis of an estimated average density for oxygen- gated, the basic technique is available by which a body
hydrogen rockets of 17 lbs. per cu.ft. can be propelled with the necessary velocity to escape
Other cases of multiple-step escape rockets are sum- from the earth.
marized in Table 3 for comparison with the cases that
have been discussed in the preceding paragraphs. Addi- REFERENCES

tional examples of rockets capable of escaping from the 1

Ziolkowsky, K. E., Rockets in Cosmic Space, published in
earth, as well as examples of orbital rockets and rockets 1903 and reprinted in 1924 (in Russian).
2
capable of escaping from the solar system, are presented Goddard, R. H., A Method of Reaching Extreme Altitudes,
Smithsonian Miscellaneous Collections, Vol. 71, No. 2, 1919.
in reference 14. 3
Lorenz, H., The Possibility of Space Travel, V. D . I. Zeitschrift,
Vol. 71, pp. 651-654, 1927.
4
6. CONCLUSION Senftleben, H. A., On the Mechanics of Space Rockets, Z.F.M.,
Vol. 19, pp. 319-323, 1928.
5
It can be concluded that it is possible from the engi- Oberth, H., Methods of Space Travel, 3rd Ed.; R. Olden-
neering point of view to construct rockets that are ca- bourg, Munich, 1929.
6
Rynin, N . A., Astro-Navigation, Vol. 9; Leningrad, 1932.
pable of escaping from the earth's gravitational field. 7
Campbell, J. W., Rocket Flight to the Moon, Phil. Mag., Vol.
Downloaded by 123.2.138.201 on July 19, 2019 | http://arc.aiaa.org | DOI: 10.2514/8.1417

One of several systems can be selected for the purpose. 31, p. 24, 1941.
8
A multiple-step acid-aniline rocket can be used, but the Ley, W., Rockets, Viking Press, New York, 1944.
9
size of rocket required to carry even a 10-lb. pay load is Ackeret, J., On the Theory of Rockets, Helvetica Physica Acta,
Vol. 19, p . 103, 1946.
larger than any rocket heretofore constructed or de- 10
Malina, F . J., Is the Sky the Limit? Army Ordnance, Vol. 31,
signed. A more reasonable multiple-step rocket can No. 157, p . 45, July-August, 1946.
be designed utilizing oxygen and hydrogen, and in this 11
Malina, F . J., and Smith, A. M . O., Flight Analysis of the
case a pay load of the order of 100 lbs. may be conveni- Sounding.Rocket Journal of the Aeronautical Sciences, Vol. 5, No.
ently carried. Larger pay loads can be carried with 5, pp. 199-202, March, 1938.
12
Tsien, H . S., and Malina, F . J., Flight Analysis of a Sounding
proportionally larger rockets.
Rocket with Special Reference to Propulsion by Successive Impulses,
In the case of single-step rockets, it is only by the Journal of the Aeronautical Sciences, Vol. 6, No. 2, pp. 50-58,
introduction of power plants operating on nuclear December, 1938.
13
energy that the velocity of escape may be achieved. Brombacher, W. G., Altitude Pressure Table Based on the
U.S. Standard Atmosphere, N.A.C.A. Report No. 538, 1935.
Single-step rockets operating on the most energetic 14
Seifert, H. S., Mills, M . M., and Summerfield, M., The
chemical propellants known would fall far short of the Physics of Rockets, American Jour. Physics, January-February,
required velocity. March-April, May-June, 1947.

Errata
Several errors appeared in "Systematic Analysis of Thermal Turbojet Propulsion," by Hans Reissner (JOURNAL
OF THE AERONAUTICAL SCIENCES, Vol. 14, No. 4, p. 197, April, 1947). They are corrected below:
CD Page 198, right column. Line 7: Dh=z twice area/circumference Line 9: £v2/2Dh
Jh_
(2) Page 201, right column. Eq. (5c) : Eq. (5d) : + W2Dn ,
" 2Dh"
(3) Page 203, right column, 1st line of text following Eq. (17b): From Eq. (15b) •
(4) Page 206, right column, 3rd line of text under main head and subhead: of head (£<p2/2Dh) per unit .. . .
(5) Page 207, left column. Text following Eq. (7e): . . . . . s == £l/2Dh.
Equations between Eqs. (5f) and (5g): CM == ju/(/x — X), C\ = X/(,u — X)

In Eqs. (5g), (26), and (27) note that the mu should be inferior to c (cM) and t h a t i and 5 in Eq. (5g) denote
limits of the integral. In Eq. (28), 1 and 0 denote limits of the integral, and fit are inferior to the c (c/«).