Human Tolerance to Multistage Rocket Acceleration Curves

BY H. PRESTON-THOMAS, PH D., CAPTAIN ROBERT EDELBERG, USAF, JAMES P. HENRY, M.D., J. MILLER, M.D., FIRST LIEUTENANTS EDWIN W. SALZMAN and GEORGE D. ZUIDEMA,USAF (MC)

N THE course of its flight, the oc- booster rockets, a rocket designed to cupants of a passenger-carrying leave the earth's surface and take up a I rocket are subjected to accelera- circular orbit outside the atmosphere tions differing substantially from that will probably present the various ac- normally experienced. T,he use of a celeration stresses in their most acute human centri.fuge capable of high peak form. From this it follows that accel- acceleration and rate of change of ac- eration programs that simulate the per- celeration allows us to duplicate suffi- formance of such escape rockets pro- ciently closely the acceleration curves vide useful criteria for test purposes. of a rocket and to determine their ef- Engineering limitations of a funda- fect on human beings in advance of ac- mental nature require an escape rocket tual flight. If rocket performance and to consist of several steps, each being the limitations of human tolerance are fired successively. Such a rocket will closely matched, it would probably be undergo subjective acceleration that advisable to make individual tests for approximates to a series of hyperbolic each rocket design. Alternatively, if it in,creases with time separated by dis- can be shown that normal tolerance is continuous decreases/ in excess of that required 'by any fore- seeable rocket design, the design prob- THEORETICAL PREDICTIONS lems themselves and those of passen- In predicting the acceleration curves ger selection are greatly simplified. In of an escape rocket it will considerably addition, the possi, bilities of emergency simplify the analysis if we can show corrective action being undertaken dur- that to a rough approximation : ing flight can be investigated. The 1. The characteristic velocities (i.e., present paper shows that a very large the velocities the rockets would be ca- proportion of feasible rocket designs pable of attaining in a vacuum and in impose no undue strain on the ,passen- the a,bsence of a gravitational field) o,f gers. all rockets considered are equal. Except for the high, short duration, 2. The constituent steps contribute accelerations that can be imposed by equally to the overall characteristic ve- From the Division of Applied Physics, locity of the rocket. National Research Council, Ottawa, Cana- 3. The initial accelerations of cor- da, and the Aero Medical Laboratory, Wright-Patterson Air Force Base, Ohio. responding steps of various rockets are Presented on March 21. 1955 at the 26th equal. annual meeting of the Aero Medical Asso- ciation, Washington, D. C. 4. The effective exhaust velocity of 390 AVIATION MEDICINE HUMAN TOLERANCE--PRESTON-TttOMAS ET .AL a rocket varies, if at all, only slightly propellant, both without payload) we and in a similar manner for all .rockets. can show that under certain conditions Although the above considerations that .bear some resemblance to the ac- hold to a fair degree of accuracy, two tual conditions the ma~s rati.os (mass

TABLE I. DATA FOR ENTERING VARIOUS CIRCUMTERRESTRIAL CIRCULAR ORBITS VIA THE SYNERGY CURVE FROM THE EARTH'S SURFACE.

Height Period of Velocity Initial Final Total of Orbit Orbit in Orbit Thrust Thrust Thrust Transit Time 0 km. 1 hr. 25.5 min. 7.95 km./sec. 7.95 km./see. 7.95 km./see. 0 224 km. 1 ~ hr. 7.82 km./sec. 8.02 km./see. 0.067 km./sec. 8.09 km./see. 44 min. 1,620 km. 2 hr. 7.10 km./sec. 8.39 km./sec. 0.412 km./see. 8.80 km./sec. 51 min. 4,100 km. 3 hr. 6.20 km./sec. 8.84 km./sec. 0.807 km./sec. 9.65 km./sec. 1 hr. 5 rain. further considerations that complicate with propellant divided by mass with- matters are: (1) the number of steps out propellant, both with payload) of of the rocket is not, unfortunately, the individual step.s should be equal fixed but may be three, four or five, for maximum overall efficiency of the with a strong probability of three ; and rocket. If the exhaust velocity is con- (2) the exhaust velocity is not fixed stant, it then follows that the charac- but may vary between approximately teristic velocities of all the steps are 2.5 and 4 km. sec. equal. They are, of course, given by We shall now discuss these six con- V.n~-Ve/N where N is the number of ditions very briefly and fllen try and steps. show how they affect t.he result. The initial subjective acceleration of There are adequate reasons for sup- any rocket at takeoff will be of the or- posing that a rocket will be required to der of two g. At v~ilues much higher enter a circular orbit round the earth than this the decrease in gravitational instead of acting as a "one shot" es- losses, which are obviously infinite at cape rocket. Although the power re- a subjective acceleration of one g and quirements, and the ratio of the initial decrease progressively for higher accel- to the final velocity increments, vary erations, is more than offset by the in- with the orbit selected, for practical crease in structural and motor mass, orbits the variation is not large, as is and vice versa for values much lower shown in Table I. After making the than two g. The initial acceleration of rather large allowance necessary to off- sttbsequent stages will probably be of set the gravitational and atmospheric the same order or a little less, although losses we can say with some confidence the rigorous lower limit of one g does that we require a rocket having a char- not apply in these cases; their lower acteristic velocity in the range from 10 acceleration limit is fixed by the re- to 11 km./sec. quirement that they expand fuel at as If we make the simplifying assump- low an altitude as possible, because a tion that all steps of the rocket have the minimum potential energy should be same structural factor (mass without imparted to the fuel for high efficiency, propellant divided by the mass witch but increase velocity fast enough to OCTOBER, 1955 391 HUMAN TOLERANCE--PRESTON-THOMAS ET.AL

prevent their falling into the atmos- our point of view we are concerned phere. Once again the optimum initial only with its effect on the acceleration acceleration lies within rather narrow curve; the higher the exhaust velocity limits. the smaller the variation in a:ccelera- If the same propellant mixture is tion. "Dense" propellants, which in- used for all steps of the rocket, the volve relatively low structural mass, exhaust velocity will be constant except will have exhaust velocities within the in the initial portion of the orbit. Here limits of 2.5 and a little over 3 km./ the external air pressure will reduce sec. ; "light" propellants may have ex- the exhaust velocity by about 20 per haust velocities as high as 4 lcm./sec. cent. We assume here that a constant We may take specimen exhaust veloc- expansion ratio and pressure are used ities of 3 and 4 km./sec as being rep- in the motors and also that it is not resentative of the two classes. worthwhile equipping the final steps We will assume from the ~bove con- with a higher performance but more sideration that the rocket we are con- expensive propellant mixture. The re- cerned with will have a characteristic duction of the overall characteristic velocity of 11 km./sec, will consist of velocity of the rocket owing to the ini- three or four steps, and will have an tial low exhaust velocity is allowed for exhaust velocity of 3 or 4 km.flsec, under atmospheric losses. this value being reduced by 20 per cent Limitations of a fundamental char- for the initial part of the orbit. We acter suggest that in the foreseeable will also assume an initial subjective future orbital rockets will be chemical- acceleration of two g for the first stage ly propelled step-rockets having an ex- and 1.7/7 for the remaining steps. h'aust velocity of 4 km./sec, or less. If M is the ifiitial mass of the step For a given structural ratio and ex- under consideration plus its payload, haust velocity there is a minimum pos- and a mass m of propellant is being sible number of steps for a rocket that burnt away every second, the a,cceler- is required to give some specific char- ation at time t .seconds after the step acteristic velocity. Neglecting the pos- begins to fire is given by: m ve sibility of entirely revolutionary en- a=--( ) gineering techn.iques, this lower limit M 1--rnt will be three for the present case. Al- M though more steps mean more efficien- If the exhaust velocity (re) iseonstant, cy, the rapid increase of complexity the acceleration curve is therefore a that accompanies an increase in the hyperbola. number of steps makes it probable Because we have already fixed the that three will be the number used, value of the initial acceIeration (a,) with a rapidly decreasing possibility of four and five. The last figure can safe- m and v~ is known, the value .o:f -- is ly be taken to be a maximum. M Although the value of the exhaust determined from the equation: velocity is of fundamental importance in ~1 ~ -- Ve in the overall design of a rocket, from M

392 AVIATION MEDICINE HUMAN TOLERANCE--PRESTON-THOMAS ET AL

The ratio of the initial (al) to the teristic velocity of 11 km./sec. It is ob- final (a2) acceleration of a step is de- vious that the variation in acceleration termined by the mass ratio of the step decreases rapidly as the exhaust veloc-

i..= V c = II Km/sec

_o I- flr

,~ N=3

N-5

2

I I I 2 25 3 3.5

EXHAUST VELOGITY (Vii IN K m /ser

Fig. 1. The effect o~ changes in exhaust velocity on the mass ratios of the individual steps of three-four- and five-step rockets capable of taking up a stable orbit round the earth.

(r.) and the ratio, of'ten unity, of the exhaust velocities : vr .~ Km/s.c N' 3 a2 Ve~

-- rn al Vel and from rn 'we also obtain ~, the total time of firing: N'3 M 1 o:'~z ve,4 K~/lo r = -- (1----) Ill rn rn itself is obtained from the kno,wn characteris'tic velocity of the complete w N,4v, 3 K~/.* rocket (vc) and the known number of steps (N) : > Ve Vc o

r~ = e or ------ve loge ro N,4 Nv. N :~ vl =4 KmI=*r Figure 1 shows the variation of the individual step mass ratios (rn), and sb ,6o igo e6o 2go 3o'o s~o 4oo ,~o hence of the over-all variation of accel- TIME IN SECONDS eration (see third equa,tion) with vary- Fig. 2. Subjective acceleration against ing exhaust velocity for three, four, time for four multi-stage escape rockets. These acceleration curves formed the basis and five-step rockets having a charac- of the present series of tests.

OCTOBER, 1955 393 HUMAN TOLERANCE--PRESTON-THOMAS ET AL ity and/or the number o.f steps in- METHOD creases. The subject assumed the supine T.his is further exemplified in Figure position on the floor of the centrifuge 2, which shows the acceleration curves cab with his back raised at a 15 ~ angle for the four cases we have chosen as with the horizontal. Unlike the earlier being most likely to correspond with tests conducted by Ballinger 1 his legs actual conditions. These curves are were raised at a 60 ~ angle so that he calculated from the first equation ex- was roughly in the conventional seated cept for the first step accelerations, posture with respect to the line of which have been reduced by an amount flight. In a practical vehicle such a varying linearly with time from 20 per position would prc:bably be necessitated cent at t=0to0at t~-~. It is this by the limitations of cockpit space. correction that accounts for the steep Suspended above and slightly for- rise of the first step curves. ward of his head were two illuminated galvanometer dials, placed side by side HUMAN CENTRIFUGE STUDIES and oriented so that one needle moved Earlier calculations by Gauer and in the apparent vertical plane, the Haber ~ had indicated the very high other in the horizontal. A short and prolonged accelerations needed to "joy" stick designed for wrist controI attain the necessary velocity. In 1952 was located on the cab floor to the Ballinger 1 described human tolerance right of the bed and was connected to to such continuous accelerations and two potentiometer shafts in such a way found, up to eight g, no serious com- that vertical or horizontal control plaints from a group of nine subjects. movements of the stick were reflected He used Gauer and Haber's 'table of in corresponding movements of the constant accelerations necessary to at- galvanometer needles. An oscillograph tain escape velocity. They ranged was wired in series with each so that from three g for ten minutes to ten g the movements of the pointer about the for approximately two minutes. In zero mark could be recorded (Figure practice, due to the decreasing weight 3). The output of a low frequency of the vehicle as the fuel is burned, oscillator set at 0.083 e.p.s, was fed to the acceleration would not be uniform2 each galvanometer pair to produce The question aro.se whether the con- asynchronous, spontaneous deviations. stantly changing accelerations would The subject was directed to maintain cause symptoms or at least impair per- the needles in the zero position by com- formance. It was known they would pensatory movements of his control not prove intolerable if applied trans- stick~ He was given a five minute verse to the body but there was no familiarization period at 1 g, then a evidence whether a man would, for ex- 4 g ride and finally a ten .second 6 g ample, be able to perform a tracking ride prior to his test run. The test run task adequately during the rocket consisted of the first three-stage ac- thrust and whether he could be ex- celeration curve of Figure 1 wi.th peaks pected to respond to a sudden emer- of 8, 5.8, and 5.8 g over a period of gen.ey with an appropriate sequence of six minutes. The three-stage runs were actions. preceded by a five minute practice 394 AVIATION MEDICINE HUMAN TOLERANCE--PRESTON-THOMAS ET AL period and a fifteen second control re- viation produced by the oscillators was cording and were followed by a the same and a fair comparison of per- twenty second recording after the run. formance could be made.

A B

SECONOS Fig. 3. Comparison of performance on two subjects, A and B, during three-stage rocket acceleration. The upper trace of each pair is the recording of the horizontal deviation, the lower one of vertical deviation.

ANALYSIS RESULTS For comparison of performance the Sub]ective.--The nine subjects re- total area above and below the zero ported that the three-stage run was reference line, circumscribed by the subjectively tolerable in the "legs- galvanometer tracing, was determined raised" position. There were about planimetrically. Because the ampli- the same relative numcber of complaints tude of the recorded deviation is pro- of fullness in the chest and difficulty in portional to the angular deviation of breathing as in the former studies on the dial pointer, the area is a measure extended supine position. No difficulty of cumulative error. Although per- was experienced by any subject in formance in this test situation cannot making the wrist movements necessary be interpreted in terms of actual con- for controlling the stick, nor was vision trol of a rocket, the use of a control a serious problem, except during the sti.ck and the two indicators would later part of the deceleration phase. suggest that this is a test of ability to At this time vertical nystagmus usual- hold a given orientation. ly occurred, which varied in severity The duration of a single cycle of from subject to subject. Vertigo was oscillation (12 seconds) was used as usually mild or ~bsent. The hydro- the length of selected samples. Thus static effects of the legs-raised position for each sample the amount of de- will be considered elsewhere. OCTOBER, 1955 395 HUMAN TOLERANCE--PRESTON-THOMAS ET AL

Performance.--As expected, there produce the most serious effects on was a great degree of individual varia- performance (Fig. 4). It should be tion in the effect of acceleration on noted that for four of these subjects,

TABLE II. PERFORMANCE OF NINE SUBJECTS DURING VARIOUS PERIODS OF THE THREE-STAGE ROCKET ACCELERATION. Values are in arbitrary units and refer to the areas of deviation from the zero reference trace. Letters A to H refer to the specific regions of the curve shown in Figure 4. Uncompensated areas were 45 vertical, 49 horizoutal.

STAGE A B C D E F G H Control To 8g From 8 g To 5.8 g From 5.8 g To 5.8 g From 5.8 g Control 1 Vertical 28 26 27 23 29 25 20 Horizontal 21 24 30 16 21 20 19 2 Vertical 22 30 28 20 25 16 15 Horizontal 20 30 22 20 21 19 15 3 Vertical 12 15 16 16 15 15 14 Horizontal 13 19 26 19 19 18 18 4 Vertical 6 15 24 14 17 17 17 Horizontal 25 33 29 23 23 20 25 5 Vertical 13 19 18 15 17 16 16 Horizontal 14 22 13 18 17 22 14 6 Vertical 19 29 27 19 18 19 23 Horizontal 15 19 20 18 24 2O 20 7 Vertical 13 25 33 24 28 28 29 Horizontal 22 33 29 27 23 2O 16 8 Vertical 18 41 46 25 49 9 17 Horizontal 17 27 36 20 33 31 34 9 Vertical 27 37 60 35 54 (No Record)- Horizontal 34 56 57 38 58 performance. Inexperienced subjects (Numbers 2, 5, 8 and 9) this was the showed a greater tendency toward de- first exposure to a three-stage, 8 g, terioration under g than did those ac- run, and for seven out of nine it was customed to the centrifuge. In general, the first try at performance under performance deteriorated somewhat such accelerations. Nevertheless de- during acceleration, in a few cases viation for many samples was less than rather negligibly (Fig. 3, A) and in a the average control deviations of 17.6 few seriously (Fig. 3, B). From an vertical and 20.1 horizontal. Further- examination of the over-all results, more, while some subjects showed there is good reason to believe that noticeable deterioration throughout the manual control can be depended on run, two subjects, Numbers 3 and 5, during three-stage accelerations of the maintained their deviations at very type studied. Table II shows the areas close to average control levels in all of deviation for the nine subjects, cases where acceleration was increas- lower values indicating better per- ing. The deterioration of these sub- formance. Samples were taken during jects during the descent from each static conditions and during the run at peak acceleration was prObably related points where severe and rapidly chang- to angular deceleration and its asso- ing conditions would be expected to ciated oculogyral effects. This was

396 AVIATION MEDICINE HUMAN TOLERANCE---PRESTON-THOMAS ET AL especially true of sample G where the when the subject was distracted from angular deceleration was more severe the tracking task by a transient but than for samples C and E due to ap- disturbing abdominal discomfort. On

~C

~ . ~ G

G 4

/

, ~ ~ J l i , I ,00 2o0 ' 30o 40o

SECONDS Fig. 4. Acceleration curve for three-stage rocket acceleration showing sampling points. plicati,on of brakes in this last stage. further runs when the pain had van- Because in rocket take-off there will be ished, deterioration of performance no such angular decelerations, it is not during the acceleration was negligible. anticipated that this noticeable decre- We believe that the use of two sep- ment in performance will occur during arate indicator dials makes this test a burnout. more valid measure of performance

DISCUSSION than would a combined dial. The task of proper division of attention between This preliminary study indicates that the two indicators then becomes as im- manual control of a three-stage rocket, portant as is the proper coordination vehicle is within the realm of pos- of the motor response. It was found sibility, especially if some small loss that the subject, during moments of in control accuracy is accepta~ble. distraction, will continue to control one Throughout indoctrination on the cen- indicator while allowing the other to trifuge might be of value from the wander. As a test of the handicap im- standpoint of familiarization with the posed by divided attention, a subject sensations involved and the changes in was told to control only the horizontal control characteristics and to allay ap- dial and to ignore the other for a prehension. We believe that the cause twenty second period. When told to of some of the poor showings was as control both meters, the deviation of much a matter of alarm as of lack of his horizontal tracing was increased by practice under g. With two separate 42 per cent. It is likewise interesting dials to monitor simultaneously, the to note that while Subject 4 (Table subject must be able to concentrate on II) did exceptionally well with his his task. Thus, in one case a marked vertical pointer on the control and the deterioration in performance occurred 8 g sample, he performed poorly with Oc'roBm, 1955 397 HUMAN TOLERANCE--PRESTON-THOMAS ET AL the horlzon.tal pointer, probably be- velocity necessary for establishment in cause of unevenly divided attention. a practical orbit round the earth. The extent to whic:h the present pre- A preliminary study has evaluated liminary studies duplicate the demands the capacity of nine subjects to per- which would be put upon the operator form a dual pursuit task while under- of an intercontinental ballistic vehicle going a typical series of curves. or orbiting rocket, both of which must Evidence is presented to indicate very closely approach escape velocity that select crewmen can be expected to to attain their objective, may be de- assist in the control of such a vehicle bated. The tests do establish, however, during the critical acceleration phases t'hat a trained occupant of such a ve- of the flight. hicle could be expected to do more than merely endure the accelerations REFERENCES while computers took over the task of 1. BALLINGER, E. R. : Human experiments adjusting it accurately to a prede- in subgravity and prolonged accelera- termined course. tion. J. Av. Med., 23:319, 1952. 2. GAUER, O. H., and HABER, H.: Man under gravity free conditions. In SUMMARY German Aviation Medicine, World War II. Washington: Department of Hyperbolic acceleration curves are the Air Force, 1950, Vol. 1, p. 641. 3. SIEFERT, I-I. S., MILLS, M. M., and derived for three or four stage rockets SUMMERFIELD, M.: The physics o.f which could attain the 10 to 11 km./sec rockets, A. Y. Phys. 15:1, 1947.

Helicopter Ambulance Service Kenmore Mercy Hospital in Kenmore, New York, a suburb of Buffalo, is one of the two hospitals in the United States that now has helicopter emergency ambulance service. A temporary heliport, a 20-foot square wooden platform, has rbeen constructed a ~few feet from the emergency en- trance to the ,hospital. and a helicopter is on call any time of the day or night .... Physicians in this highly industrialized area of approximately a million people estimate that the helicopter will be used 'for at least eight to ten patients annually--perhaps a stricken sailor aboard a lake freighter miles from port or an expectant mother in a remote, snovebound farm- house, or a wounded hunter along a wooded trail .... Helicopter ambulance service was first established in the United States at a hospital in Santa Monica, Calif. Here a rooftop platform serves as a landing area, and helicopters are made availat)le by rotorcraft firms in the Los Angeles area. Arrangements are now being made for the construction of a heliport in the Dallas-Fort Worth area.--Public Health Reports, 7:432, April 1955.

398 AVIATION MEDICINE