•
June,
1974.
EF'F'ECIS OF SONIC :sa:M ON AU'ItM:BILE-DRIVER BEHAVIOOR
by
Orest Volodyrnyr Na.vakiwsky
VLIEG TU.G30U\\ i~U,'JûE EH':'UDn'l~::J( Kluyvcrwe~ 1 - DELFT
6 SEP.
\974
.,
.
EFFECI'S OF
sauc
:BXM CNAUTCMOBlLE-DRIVER BEHAVIOUR
by
Orest Vo1odymyr NCMakiwsky
Subrnitted May 1974
Acknowledgerrent
The author wishes to express his sincere thanks to hls supervisors, Dr. I. I. Glass and Dr. L. D. Reid, for suggesting the prd:>lem and providing the opportuni ty to conduct this research at UTIAS. Their patience, advice and guid-ance are gratefully acknowledged.
Special acknowledgem:mt is due ~.r. Reinhard Gnoyke, the staff technician, those neticulous work and hundreds of hours of assistance WëtS essential for the p:reparation and execution of the ~iIrent. His tine and technical assistance are
very rruch appreciated.
Thanks are due to Mr. A. Perrin and Mr. J. Unger for their design and canstruction of the electronic equiprrent. Mr. B. Leigh 's help in developing the delayjtrigger circuits is greatly appreciated.
Cooperation recei ved fran Mr. E. S. Annis, Director of the Uni vers i ty
Facili ties at York Uni versi ty in providing a location for the test track, as weIl
as the assistance of Mr. D. W. Farren Director of Systems Research at the Ministry of Transportation and Carnrunicatians in providing one hundred plastic traffic cones for the experirrent is acknowledged with thanks.
To the twelve subjects who volunteered their tirre special thanks are due for their efforts which made the dri ving experinent possible.
The financial assistance recei ved fran The Transportation Developrrent Agency, Ministry of Transport and the National Research Council of Canada is acknowledged wi th thanks.
"
Test results are presented on the response and behaviour of autarobile drivers subjected to sonic boem disturbances under actual driving conditions. A description is gi ven of the design and developrrent of a portable sonic boom siml-lator, auxiliary equiprrent and experimental techniques used to study the nature and severity of the disturbance effects .
The sonic baan simulator, consisting basically of loudspeakers and a function generator was m:A.U1ted inside a test vehicle. It was able to produce sonic boans that were very similar to what drivers would experience follcMing SST overflights. The simulated boans had overpressures of 3 psf, rise tirres of about one millisecond and durations of 100 milliseconds.
'!Wo aspects of driving were investigated; the Tracking Manoeuver and the Stopping Task. Results from both tests indicated that driver behaviour was not affected by the simulated boans, even though s~ drivers considered i t annoying or disturbing. It may therefore he concluded frem present limited
statistical tests that current commercial supersonic aircraft under normal flight conditions (without superboans) would not produce adverse effects on a driver's stopping distance or hls abili ty to follCM a particular course.
1.
2. 3.4.
5. 6. 7. 8. 9. Table of Contents INTRCDUCTICN BACKGROUNDrunc
BCX:M PRESSURE WAVEFORM rnSIDE AN AUTCM)BILE DEVELOPMENr OF A PORI'ABLE SCNIC :sca.1 SIMULA'roR EXPERIMENTAL PRCCEDUREmUIPMENT
6 .1 The Vehicle Track Recording System
6.2 The Carrier Oscillator,Amplifier
&
Course Inductors 6 .3 Recei ving Inductors6.4 Receiver Unit 6 .5 Test Track 6.6 Photo Cells
6.7 TiIre Delay Circuits 6 • 8 Balloon System 6.9 Other Ei:Juiprrent
SUBJECTS AND TRAINIffi
TEST PROCEDURE 8.1 Tracking Manoeuver 8.2 stopping Task CONCLUDrnG REMARKS Page 1 1 2 4 6 7 8 8 9 9 9 9 10 10 10 10 10 11
14
15 REFERENCES 17APPENDIX A: Statistical Analysis APPENDIX B: Questiormaire
TABLES FlGURES
1. INTRODUcrICN
The age of the supersonie transport aireraft (SST) has begun. Not only are the prototypes of supersonie transports being tested but the first production mcrlels are rolling off the asserrbly lines. The SST brings with i t new tedmolo-gies and shorter flight tirres, as weIl as prcblerns associated wi th possible pollu-tion of the upper atmosphere and the possible adverse effects of sonie
boom
on hurnans, anirnals and structures at ground level. There is therefore, an urgent need to assess the startIe , ecological, and structural effects of the SST in order to provide goveITIITEnts and regulatory bodies wi th concrete background data that would provide for optimum operating polieies for the SST.Concern has been expressed regarding the possible inerease in aeeident-proneness of autorrnbile drivers subjected to sonie booms. As aresuIt an initial analog study was undertaken by Lips (Ref. 1) at tn'IAS eoncerning the effects of sonie boem disturbances on an individual 's corrpensatory tracking performance for an unstable system, which in certain respects simulated autorrobile dri ving. It was found that oost individuals were disturbed by the sonie boem and recovered in varying degrees. It is the purpose of this thesis to extend Lips I preliminary
investigation by developing a portable sonie boom simulator and experirrental teclmiques to he used in evaluating the nature and s eve ri ty of a sonie boom disturbance on aetual automobile-driver behaviour.
2. BACKGROUND
To answer the question of what, if any effect do sonie booms have on the behaviour of autonnbile drivers, an extensiveli terature survey was conducted into the folleMing related fields i physics of the sonie boem, hmnan reactions to sonie booms, simulation tedmiques for produeing sonie booms and behaviourial analysis of automobile drivers.
Sonie boom is a phenC'llTeI1on associated wi th supersonie flight. It is basieally an aeoustieal disturbanee eharaeterized by physieal parameters such
as overpressure, risetirre and duration (Fig. 1). 'Ihe sonie boan disturbanee emanates from an aireraft flying supersonieally having a shock wave pattern eon-sisting of two cones generated at the bCM and tail of the SST (Fig. 2). The inter-cept of the trailing cones wi th the ground produces a horseshoe corridor in which the sonie boem is experieneed. By the time the sonie boom disturbance reaches ground level it has eoaleseed roughly to a pressure waveform having ideally the shape of the letter IN I (Fig. 3) charaeterized by a sharp rise in air pressure at
the beM wave, a sleM drop in pressure to beleM arrbient and a sudden inerease in
pressure back to arrbient pressure at the tail wave. The pressure signature of the sonie boom varies eonsiderably as a result of such diverse influences as ground reflection , aireraft Mach nurrber, altitude, weight, attitude, size and shape, atmospherie variations (winds, terrperature, humidi ty, and turbulenee), lateral distance fran the flight path, aireraft manoeuvers (turns, acceleration and de-eeleration) and whether the boem is heard indoors or outdoors. 'Ihus, the sonie boem is influenced by many variables which in turn result in different degrees of reaction from the subjeets experieneing the boem. Through field-test observa-tions it has been established that for current oormereial supersonie transports under cruise conditions the sonie boom (Fig. 1) would be characterized by over-pressures from 2 to 3 psf, risetirres fran 0.1 to 10 rrsee. and durations fran 200 to 300 msee (Ref. 2). HeMever , turbulenees and manoeuvers ean generate superbooms that have overpressures several tirres greater than these naninal va lues .
The human response to a sonie boem is exceedingly complex, being simu
-taneously a physiological, psychologieal, and sociologieal .reaction. It is
influ-enced not only by the physieal stimulus but also by the im:rediate envirornrent,
arrbient noise, and various bther factors not related direetly to the stimulus. The
ITOSt significant hurran reaetion to the sonie boom is that of start1e (Refs. 3,4)
and is a funetion of the risetime and overpressure of the sonie boom (Ref. 5) . In
general the shorter the risetirre, the greater is the startIe reaction, but this
again depends upon sum factors as age, health, fatigue level, errotional state,
and activity at the ti.rre of the boom. No direct physieal injmy ean be expeeted
frem the sonie boom (Ref. 5). Hcwever it is the startle reaction that presents
the greatest threat of injury due to IDvoluntary rroverrents.
Work has already been done by LlpS (Ref. 1) on one-dlIrensional tracking
and by Thaekray et al (Ref. 6) on t"wo-dirrensional tracking tasks. These tasks
rnay be eonsidered to simulate autorrobile-driving behaviour. The results of these
experirrents were inconclusive and on the surface seerrro contradictory . Thackray
"suggests that the simulated booms rnay have elicited ITOre of an orienting or
alerting response than a startle reflex" while Llps coneluded that "sonic boom
disturbances of the type generated by the Concorde SST resul t in a rreasurable
startle effect" (Ref. I, pg .12). The discrepancy in the conelusions rnay be due
in part to the difficulty in simulating the sonie boom under controlled experirrental
condition. Aecordlng to 'Ihackray the simulated sonie booms had longer rise tirres
and smaller overpressures than w:mld be expeeted from an SST. Therefore the booms
rnay have served as an acoustie stimulus rather than a startle trigger, whereas
the booms used by Lips did simulate the rlse times reasonably weIl and the
over-pressures very weIl.
3. SCNIC BCXl1 PRESSURE WAVEFORM INSIDE AN AUIDMOBlLE
The purpose of this p!.oject was to study the effects of sonie boom on
the behaviour of automobile drivers. For the resul ts to be eredible i t was deeided
to simulate driver behaviour using a real vehiele. First, it was necessary to
determine what a driver hears inside a vehiele during an SST overflight. ThlS was
aecomplished by positiomng a Ford Pinto sedan inside the hom of the UTIAS sonie
boom simulator (Figs. 4,5) and subJeeting the vehiele to a range of sonie booms
and
reeording their pressure waveforID inside the vehiele. The Ford Pinto was selee-ted
for' this project since it was the only vehiele that rret all of the experirnentalcri teria, requiring the vehicle to:
(a) be able to fit into the sonie boom simulator
(b) be available Wlth an automatic transIlUSSlon in order to
facili tate subJect selection
(c) be easily obtdlnable for rent or lease
(d) be in relati vel y widespread use by the public.
The automobile was inserted into the simulator and centered along
the
longitudinal axis of the hom with the front end facing the apex (Fig. 6). 'Ihe
driver's head was thus located at the 70 foot station inside the hom. The total
volurre of the passenger ccnpartment was estimated to be 65 cubic feet. 'IWo
con-denser-type microphones were used to rroni tor the sonie boom pressure signatures.
A Bruel and Kjaer (B
&
K) one inch diameter 4145 ffilcrophone and a B & K 2631pre-anplifier carrier system was located inside the Pinto at the driver's head
position. The other mierophone, a half-inch diameter B & K 4147 together wlth a
B
&
K 2631 carrier systemwas placed outslde the vehiele adjacent to the driver'sdoor and at the same height as the microphone inside the vehicle. 'Ihe sonic boom signatures detected by the microphones were displayed on a Tektronix 5103 N
storage oscilloscope and a Polaroid camera was used to make permanent records of the two pressure waveforms. In all, fifty-four pictures were taken. The car was subject to simulated sonic booms ranging fram 2 to 8 psf overpressures and durations fram 50 te 300 msec. The autorrobile interior pressure signatures were recorded for the follaving cases;
(a) both windavs closed
(b) left hand windON closedjright hand windON open (c) left hand WindON openjright hand windav closed (d) both windavs open.
These resultsare presented in Figs. 7 to 15. In each case the upper trace des-cribes the overpressure signature rreasured at the driver' s head posi tion while the laver trace represents the incident senic boom rreasured beside the car. 'lhe microphones were calibrated using a B & K 4220 pistonphone so that the one inch microphone was exactly four tirres as sens i ti ve as the half inch microphone. 'lhey were adjusted to read:
one-inch microphone half-inch microphone
1 volt = 1 psf overpressure 1 volt
=
4 psf overpressureFigure 7 (a,b ,cl shON three particular cases when the calibration of the two micro-phones was checked before, during and af ter the experirrent respectively. It will be noticed that the upper and laver traces are identical, as they should be. Figure 7b contains lot of 'hash' or superirrposed jet noise which results when the fiberglass jet-noise filter was blown out half way thraugh the testing pro-cedure. The filter was repaired and replaced and the tests were concluded.
The sonic booms prcrluced in the hom simulator were not ideal, having longer rise tirres, and rrore superirrposed jet noise than actual SST booms.
Therefore all the results must be viewed within these limitations . In the case TNhere both windavs were closed (Figs. 8,9) the interior waveform may generally be described as a danped sinusoidal wave ha ving a longer rise tirre, and on the
average, an overpressure 39% less than that of the exterior senic boom. For the case mere only the right hand windav was open (Figs. 10,11) or only the left hand windav was open (Figs. 12,13), again longer rise tirres and danped sinusoidal carrpo-nents are seen in the interior response wavefonrs. Havever , the ini tial over-pressures on the average were 35% and 33% greater, respectively, than their corres-panding exterior exci tation waves. In the case where both windavs were open (Figs. 14,15) it is seen that the interior pressure-tirre history has a very close res~ blance with the exterior sonic boom signature, the only exception being that the overpressure is 27% greater inside than outside the car. Finally, in all cases i t can be ooserved that the arrount of superinposed jet noise is considerably less inside than outside the vehicle, irrplying that the car acts as a ION pass filter by absorbing the higher frequency conpenents of the jet noise. Also i t seerns that the passenger ccnpa.rt::nEnt together with one open windON acts as a Helmholtz resonator. 'lhe resonator pressure fluctuation has the appearance of a darrped sine wave, persists for a longer period of time, and can have greater arrplitudes than the
ini ti al excitatien wave. The He lmho 1 tz resonator approach was used by Lin (Ref. 7) and Vaidya (Ref. 8) to predict analytically the tirre history of the sonic boom
pressure wave as i t propagates into rooms through open windows. Observations made by Vaidya (Fig. 16) were very similar to the present results, in particular those
shawn in Fig. 10.
4. DEVElOPMENT OF A PORI'ABLE SCNIC :BOeM SIMULATOR
'Ihe first of two major prc:blems that had to be overcare in order that the study be perforrred, invol ved the developrrent of sarre portable device to siml-late the sonic boom in arroving vehicle. Three alternatives labelled (a) shock tubes, (b) headphones, and (c) speakers were proposed for consideration wi th respect to:
(1) the cest of the equiprrent involved
(2) tirre required to assembIe, debug, and calibrate equipment (3) credibility of the results dbtained using that equipment.
'!he shock tube system would have utilized several shock tube charnbers connected to a wooden horn located either on the roof or inside the trunk of the car. Breaking any of the shock tube diaphragms would generate a very short duration sonic boom. A similar system has been develeped at LTV Research Center
(Ref. 9). This systern was rejected due to the uncertain credibility that could be assigned to the resul ts because the boorrs had rnicrosecond rise times and very short durations in the order of ten rnilliseconds.
'Ihe headphone system would have reqw.red the driver to wear a set of headphones which would prcrluce the required sonic boom signal. The feasibi li ty of using headphones was investigated considerably before i t was concluded that the prcblems associated wi th sealing the headphones in order to reproduce ION frequencies, together wi th the unnatural environment created for the driver, carpletely outweighed and cancelled the relatively sirrpler approach that the head-phones seerred to offer.
It was finally decided to go ahead with the speaker system, which was in principle similar te the Loudspeaker-Driven Booth SiImllator (Ref. 10) in that the passenger corrpart::rrent of the Pinto had to be rrade air-tight and loudspeakers were used to produce the required sonic boom waveform.
To nount the speakers inside the car a 1-3/4 inch laminated plywood wall was rrounted behind the two front seats (Figs. 17,18,19). 'Ihe top of the wall was attached by stainless steel brackets to the shoulder belt anchor balts,
and the bottom was bolted directly te the chasis of the car. The wall was lami-nated to increase its rigidity as weU as to contain a built-in rear-view plastic windON and a one inch wide channel around its perirreter. An inflatable rubber
se al made from bicycle inner tubes was instalIed inside the channel. 'Iherefore, the front passenger compartment could be sealed perfectly fram the rest of the vehicle, just by inflating the rubber inner tube until i t had expanded into and filled all the gaps and crevices between the wall and the body of the car. Wi th the wall in place the volurre of the passenger corrpartrrent was estimated to be 50 cubic feet.
To generate a sonic boom, it was decided to use 18-inch-diarreter Goc:x:lrPan loudspeakers (Fig. 18), which have a good frequency response ranging from a few cycles per second, where they act as pistons, all the way to up 3000 Hz. 'Iheir overall rroving diameter is only fifteen inches and under full paver they seem to
have a 1/4 inch thro.v or di sp lacerrent . 'Iherefore, the maximum pressure change brought about in the passenger corrpartnent due to one of these speakers can be
calculated as follavs:
speaker diameter, D
=
15 inches speaker area, A=
1.24 sq.ft.'!he vol'L1Ire (V) displaced by the speaker in a 1/4 indl thrav WJuld be: V
=
0.03 cubic feetSince we are dealing wi th srnall, acoustic disturbances we rnay ass'L1Ire perfect gas relations and use Boyle's Law: (PI) (VI)
=
(P2) (V2) where PI
=
14.7 psi=
2117 psfVI
=
50.00 cubic feet V2
=
VI-V=
49.97 cubic feet Therefore P2
=
(PI) (Vl)/(VI-V)=
2118 psf.Therefore the change in pressure in the passenger corrpartIrent resul ting fran one speaker displaoement should be one psf.
In order to prcrluce overpressures of three psf, four GcxXlrnan 18-inch speakers were rrounted on the wall (Fig. 18). In this rrode the wall acted as an infinite baffle. Each of the four speakers was driven separately by an RCA HC2000 dc arrplifier having 100 watts rms maximum paver output. '!hus the system had a capacity to supply the speakers with a total of 400 watts of electrical
pa.ver. 'Ihe simulated sonic boom was triggered by a five segrrent function generator that prcrluced a voltage signal, which in turn was anplified and fed into the speakers that generated a pressure signature inside the passenger corrpartnent. Both the
function generator and the amplifiers required a portable 38 volt dc power supply which <x>nsisted of six lead-acid 12 volt rrotorcycle batteries . It supplied
sufficient electrical paver for about 200 sonic boans, rrore than enough for one day of tests and was easily rec:harged overnight.
Initially i t was decided to simulate the type of sonic boc:ro that the driver \\Duld hear when his windows were closed. The reason being that the over-pressure and rise time requirerrents were less severe than in the other cases. 'Ihe type of waveform we wanted to prcrluce inside the passenger corrpartrrent is ShCM1 in the upper trace of Fig. 8. '!he function generator was prograITlIEd to prcrluce an exact electronic analogue of this pressure signature . 'Ihe upper trace in Fig. 20 shavs the electronic signal which was played through the speakers, while the l~r trace shavs the pressure waveform resulting inside the passenger corrpartIrent.
Mlat we wanted but did not get was a pressure signature sllnilar in shape to that of the electrical signa1. '!he main reason that the pressure did not follav the speaker wavefonn was due to air leaks in the passenger corrpartrrent as seen fran the laver trace in Fig. 20, where as certain pressure was reached but could not be rnaintained.
As aresuIt another effort was made to seal the leaks. All the ventilation ducts were sealed off, the heater, fan, and hand brake were rerroved and the remaining holes in the bulk head were plugged up. '!he only leaks rernaining occurred around the doors and windavs arrl these were minirnized to the best of our abili ty by using silicon sealants. The pressure waveform was repeatedly c:hecked, but the pressure
could not be :rraintained for the desired length of ti:rre.
It was then decided to atterrpt to simulate the open window case, mich had shorter rise ti:rres and higher overpressures but did not require holding the pressure at a particular level for any length of tiIre. One conclusion that was drawn from the above exercise was that it would never be possible to obtain a pressure signal similar in shape to the electrical input, since the speakers and the passenger conpartIrent distort the acoustic wave. 'Iherefore, oor electrical input had to be predistorted to cbtain the reqw.red re sult . A trial and error approach was used (Fig. 21) to obtain areasonabIe looking N-wave with an over-pressure of 3 psf, duration of 100 milliseconds and rise time of the order of one rni11isecond. The signature of the simulated sonic boom that was eventually used in the automc:bile driving experi:rrent is shown together with its electrical analogue in Figs. 22 and 23. This simulated sonic boem pressure waveform, ha ving
3 psf overpressure , corresponds very closely to mat V-Duld in fact be ooserved inside the passenger comparbnent if the Pinto, wi th both windows open, was exposed to an external sonic b-om wi th an overpressure of 2.5 psf. 'Ihis conclusion is based on the results of the tests conducted inside the UTIAS hom simulator indi-cating that for the case mere both windows were open, the interlor pressure wave was sirnilar in shape to the exterior pressure signature except that the interior waveform had a 27% greater averpressure . It should also be observed that the dura-tion of the simulated boan is only 100 rrsec. long, about one-half of the duration expected from a cc:mrercial SST. 'Ihis difference however, does not affect the re-sults that :rray be obtained men studying hu:rran response since work by Rcbinson and Jahnson (Ref. 11) had indicated no variation in subjective loudness with sonic booms of different durations, ranging from 100 to 500 rrsec.
5. EXPERIMENTAL PROCEDURE
Having developed a portable sonic boom simulator, the second major prob-lem to be overcorre was that of developing an experi:rrental procedure which would indicate mether or not sonic boorns would make drivers accident-prone. Many studies had been conducted to determine what pararreters are invol ved in a driver' s behaviour and his interaction wi th the vehicle and the road. Bloek diagrarrs and mathematical m:dels have been constructed, but the dri ver-vehicle-highway corrplex has so many variables, any one of which may dorninate at any given instant that the probability of cbtaining a high oorrelation with ct single factor or group of factors related to accidents is very rerrote. Presently there exist no criteria by which a driver, or mmoeuver can be classified wi th respect to accident-proneness. It was there-fore decided, only to look for visibIe effects that sonic boorrs might have on the behaviour of autorrobile-drivers such as: a rnoItEntary loss of control, deviation
from a prescribed path or inability to react to an unexpected event. Consequently two aspects of driving were investigated, the tracking mmoeuver and the stopping task.
Twelve subjects were asked to drive the instrumented Pinto (Fig. 24)
on a test track located in a large paved parking lot. The track was 2500 feet long and had both slalom and straight sections defined by plastic pylons and a longi-tudinal copper cable down the centre. The cable carried a 2000 Hz signal which was detected by solenoid inductors rrounted on both sides of the car. 'Ihe rrodulated
signal fran the solenoids was recorded on tape and. provided a continuous record of the position of the vehicle with respect to the cabIe. A photo-electric circuit was set up across one of the straight sections. Breaking the circui t caused two balloons (on random runs) located 150 feet further down the track to inflate rapldly,
'll1e tracking rranoeuver invol ved dri ving along the prescribed course and recording the accuracy achieved by noting the nurrber of plastic cones that had been disturbed as weIl as recording on tape the vehicle's deviation fram the centre-line. The stopping task consisted of rreasuring the distance required to stop the vehicle when its path was suddenly blocked by rapidly inflated balloons
(Fig. 25). During this driving sequence, the subjects were exposed to the simu-lated sonic bOOIIS which were generated inside the vehicle by the four 18 inch loudspeakers and function generator described earlier .
6. N;dUlPMENT
'n1e test vehicle used in the driving experiIrent was a regular 1973 Ford Pinto, 2-door sedan, equipped with a3-speed autaratic transmission , 2300-cc 4 cylinder engine, manual steering , manual brakes (front discs) and an electric rear winda.v defroster. 'll1e weight of the vehicle wi th the inst.runent package and driver was approxirrately 3000 lbs. The principle diIrensions of the vehicle were: overall length overall width overall height wheelbase track: front rear 169.0 inches 69.4 inches 50.5 inches 94.2 inches 55.0 inches 55.8 inches
Wi th the ' speaker wall' in place and spare tire rernoved the volurres were: front passenger oarnpartment
rear passenger carpart:rrEnt trunk
50 cubic feet 15 cubic feet 7 cubic feet
Since the experirrents involved driving an autorrobile around a track it is best to differentiate between the rrebile e::}Uiprrent rrounted on and inside the Pinto and the stationary Equiprrent located at the control si te. A flo.v chart of the equiprrent is sha.vn in Fig. 26 . All the rrobile equiprrent is sha.vn located belOtJ the dotted line in the fla.v chart. l t consisted of: FM antenna, FM radio receiver, function generator, dc anplifier, loudspeakers, solenoids, tracking receiver, FM tape recorder, pCMTer supplies and De-AC inverter . Most of the above equiprrent was located in the trunk. of the Pinto secured by brackets and cushioned with polyurethane foam as sha.vn in Fig. 27.
'll1e FM antenna (rrodel FM(26) consisted of crossed folded dipoles to enable i t to recei ve signals fram all directions . l t was rrounted on the roof of the car using rubber suction cups and secured with nylon guidelines. lts purpose was to reeei ve the trigger signal that would set off the sonic boom in the passenger
oompartrnent. '!he antenna was connected to a srrall battery pCMTered FM receiver which was tuned to 92.8 MHz, the transmi tting frequency. 'll1e signal from the receiver was fed into a pulse width noise discriminator which alla.ved only genuine trigger pulses to activate the function generator. Onee activated the five segrrent function generator prcrluced a predistorted electronic analogue signalof the
sonic boom (Fig. 22) which was fed simultaneously into the four channel amplifier and into one of the channels of an Arrpex FM tape recorder. 'Ihis signal was then anplified and played through the four loudspeakers whic:h generated the required sonic boom.
The Arrpex FM tape recorder had two channels thraugh which simultaneous recordings could be rrade. One channel was used to record the signal from the function generator thus providing a record of hav rnany and when the boorrs were heard. '!he other channel was used to record the vehicle' s tracking performance. The paver supply for the tape recorder consisted of a 12 volt car battery and a De-AC inverter providing the 110 VAC 60 Hz required electrical paver. The tape recorder used seven-inch reels and Scotch '207' brand magnetic tape. All record-ings were made at a speed of 3-3/4" per second, which permitted a maximum of three hours of data to be recorded on one reel of tape.
A 38 volt dc pcwer supply which was made up of six 12 volt notor-cycle batteries supplied the required electrical power to the four channel amplifier, function generator and the vehicle tracking system and was located on the floer in front of the right hand passenger seat in order to maintain a balanced weight distribution •
6.1 The Vehicle Track Recording Systern
A driver' s performance throughout the experirrent was rreasured both in a traditional and novel way. The traditional ITEthcd consisted of using plastic traffic cones to provide an outline of the course and a record was kept of the nurrber of cones that were disturbed on any particular run. 'Ibis was fairly crude and detected only large deviations frem the set course. The novel rrethod consisted of using an electronic system recently described by Grant (Ref. 12) which detected and continuously recorded the path taken by the vehicle with respect te a fixed cabIe. 'Ihe vehicle track recording system shavn schernatically in Fig. 28 depends on the mutual inducti ve coupling between the course inducti ve cabIe, energized by an audio frequency carrier and two recei ving inducters.
The system consisted of an oscillator/amplifier which generated the carrier frequency and a receiver unit to process the signals induced in the receiv-ing inductors. 'Ihe receiver unit provided a polarized dc signal with the frequency of the vehicle' s deviations fran the radiating cabIe. As the centre of the car crossed the cable the polari ty of the signal would be reversed. Also the signal amplitude varied as a function of the vehicle's deviation from the cable according to the calibration curve shavn in Fig. 30. 'Ihis system was limi ted by the induc-tor spacing te detect a deviation of 36 inches from the centre of the track. At this position one of the inducters would be directly over the cable and the signal amplitude would be in the nonlinear region of the calibration curve. Hcwever, since the inductors were separated by six feet and the track width was only ten feet, any deviation from the centre-line of nore than two feet would result in hitting the plastic oones which defined the borders of the track. 'Iberefore the function of this systern was to rroni tor the driver' s performance wi thin the linear region of the deviation/'response curve (Fig.30).
6.2 'Ibe Carrier Oscillator, Arrplifier & Course Inductor
'!he osclllator used a Wien bridge circuit tuned to 2000 Hz in a feed back loop of an IC amplifier connected to an RCA HC 1000 pcwer amplifier. The course inductor was the amplifier's load and had a resistance of 1.6 ohms per 1000 feet. The total length of the cable was 2500 feet, thus providing the ampli-fier with four ohms resistive lead.
0/
6 .3 . "Réceiving Inductors
These were six-indl long solenoid inductors ha ving 2000 turns of single copper wire around an iron core. They were secured inside two adjustable black plywood boxes located six feet apart on a wooden beam bolted to the chasis of the car just belcw the front bmrper (Fig. 36). The centres of the inductors were eleven inches alx>ve the grc:und. The inductors were also inclined sixteen degrees to.vards the car in order to be tangent to the circular field being radiated by the wire. Using this arrangerrent the deviaticn/response curve was calibrated as shavn in Fig. 30.
6.4 Receiver Unit
The schernatic diagram of this unit is shONIl in Fig. 29. l t consisted of two identical arrplifier and rectifier channels. Each channel received induced signals from its cwn separate solenoid. 'Ihe carrier frequency was filtered out using RC-networds leaving enly the lo.v frequency tracking signa1. '!his unit was located in the trunk of the car and i ts output was connected to the second channel of the FM tape recorder.
6 .5 Test Track
The track was laid out in parking lot 'F' at York University which was 500 feet long and 250 feet wide and contained no physical rostructions except for a single grass covered rredian running approximately davn its centre. lt was oriented roughly in a north/south directien having a gentle do.vnward slope of about ene foot per one hundred feet to.vard the sOlith. '!he course consisted of three sectionsi first a 1000 foot large slalcm course followed by a 500 foot shallo.v slalcm and then two 500 feet straight sections . The large slalcm was corrposed of seven semi-circular arcs having a 32 ft. radius labelled A, B, C, D, E, F, and G in Fig. 31. '!he shallo.v slalom was constructed around a sinusoidal wave with a 6 foot anplitude and 80 foot wavelength and contained eight turns
stretdling from corners H to J. The centre of the track was defined by asolid
'!'WH #12 insulated 'Flameseal' copper wire that was attached to the asphalt sur-face of the parking lot by 1-1/4 indl concrete nails and ethyl cellulose cable clarrps spaced every twëlve inches. The track was ten feet wide. lts borders were defined by one hundred 18-inch high orange plastic traffic cones located at
strategic pa3i tions around the course. The control si te was placed near the centre of the straight section of the course running parallel to the edge of the parking lot. Here rrost of the follcwing stationary equiprrent was located: photo-cells, counter, trigger-circui t, balloons and transmitter as shavn in Fig. 32.
6 .6 Photo-Cells
'!he photo-electric circuits separated by ten feet were positioned across the track at the control si te. Each circui t consisted of a light source and photo-cell placed en ene side of the track and a ene square foot adjustable mirror on the other side, thus avoiding prcblems wi th wires crossing the track. The two photo circuits were connected to a 5325A Hewlett Packard universal digital counter which determined the tirre i t took the vehicle to break both circui ts . Kncwing
the tirre and the separation of the photo-cells, the speed of the vehicle could be calculated. The first photo-cell was also connected to two tirre delay circuits.
6 .7 ' Tirre Delay Circui ts
TWo time delay circuits were developed to provide an accurate and auto-rnated method for studying the driver I s performance during the stopping task. One circuit delayed for one second the inflation of two balloons located 150 feet fram the first photo-electric circuit, af ter it had been interrupted. 'Ihe other delay circuit automatically (if desired) triggered a sonic boom within the car at a preselected time wi th respect to the balloon appearance. The purpose of this arrangement was to study any difference in performance that rnay have resul ted if the driver was boomed prior to, rather than during or af ter seeing the balloons inflate. By turning a dial, the second delay circuit could be set to trigger a sonic boamwith a delay of 0.0,0.5,1.0, 1.5,2.0, and 2.5 seconds af ter the first photo-electric circuit was braken.
6.8 Balloon System
'Ihe purpose of using balloons was to have both a cheap and easily re-placeable system that would be able to block suddenly the driver' s path and at the sarre time not dama.ge the vehicle in case of a collision. Latex '7 - 28 t
balloons were used. 'Ihey were located 150 feet dcwn the track from the first photo-cell. They were inflated by helium to a height of 28 inches in just 250 msec. by a two stage solenoid valve. Helium, fram a high pressure cylinder ,was used to cbtain a faster inflation rate than was possible with either nitrogen or air, due to its higher speed of sound. It also provided some buoyancy which helped to stabilize the balloons in a vertical position.
6.9 other Equipment
In addition to the time delay circuits there was also a manual over-ride button vi1ich could be used to trigger a boom at any time, anywhere on the test track. The trigger pulse was sent from the control site to the vehicle by an FM transmitter operating at 92.8 MHz. The transmitter, pcwered by a 12 volt dc supply, had 500 milliwatts of pcwer into its final stage, resulting in the radiation of about 250 milliwatts of pcMer. 'Ihe only remaining piece of equiprrent was a stop watch used to clock the driver's time required to ccmplete one lap of
the course.
7 • SUBJECTS AID TRAINING
Volunteer subjects were selected from the staff and student body at the Ins ti b.lte for Aerospace Studies, uni versi ty of Toronto, under the provisions regu-lating the use of volunteers in experirrents. The purpose and general procedure was exp1ained, a time and date for the test was selected and a consent form was signed byeach of the subjects. In all, twe1ve volunteers took part in the acb.lal testing. Most were experienced drivers ranging in age from 23 to 38 years. Tab Ie 1 provides additional background info:rrration about the subjects .
Upon arriving at the test site the subjects were asked to familiarize themse1ves with the vehicle and track by driving twice around the course at a relaxed speed. They were then ins tructed about the tes t procedure.
8. 'lEST PROCEDURE
The test involved canpleting four runs fo11CMed by a rest period during which a questionnaire (App.B) was answered. The subjects then returned and
completed four nore runs. Each run consisted of doing two laps around the course, unless the track. was blocked by two inflated balloons, in which case the driver was required to stop the vehicle in as short a distance as possible trying not to hitthe balloons. Before eadl run the drivers were told that they mayor may not experience a sonic boem and/or be stopped by balloons. They were also told that their lap time would be rronitored and were urged to drive as quick.ly and carefully around the course as possible while avoiding hl tting cones, since each cone dis-turbed would add a fi ve second penalty to their lap tirre. They were then asked to check their seat-belts, close their windcw and to start the run when they heard the trunk close. The test operator, af ter instructing the driver, checked and turned on the equiprrent in the trunk noting the pos i tion of the recording tape. The trunk was closed arrl the test run began. A stop watdl was activated simul-taneously with the start of each run. The operator then reset the counter and preset the automatic time-delay trigger system if the stopping distance was to he studied. If not, the operator would manually trigger the sonic boom on corners F and/or J, depending on the run and observe the nunber and location of any cones that were disturbed. The driver's lap tine was recorded at the end of the first lap. Eadl run would end af ter two laps were corrpleted with the vehicle back. at the control si te. If havever , the driver was stopped by balloons, then the stopp-ing distance ITEasured from the tip of the front burrper to the balloons and the vehicle 's stopping speed, as determined by the counter, were recorded. 'lhe equip-rrent in the trunk was turned off and the driver asked to drive back to the start-ing position. All drivers were given the sarre instructions and the test runs were conducted in the sarre sequence as described helcw:
Training
run
wi th no boc:ms, conpletedtwo laps. Run#
1. Stopped wi th no boomRun #2. Stopped with boom occurring 0.5 seconds prior to balloon inflation. Run #3. Booned on corner F during first lap, and boc::med on corner J during
second lap.
Run #4. Stopped with baan occurring at time of balloon inflation. REST PERICO
Run #5. I30cned on corner J during first lap and stopped with boom occurring 0.5 seconds af ter balloon inflation.
Run #6. BooIred on corner F during the second lap.
\
Run #7. Stopped with boom occurring 1.0 seconds af ter balloon inflation. Run
#
8. stopped wi th no boom.8.1 Track.ing Manoeuver
'lhe track.ing manoeuver was one aspect of dri ving behaviour to he studied. Its purpcse was to determine whether or not a driver would m::::rrentarily lose control or deviate from a prescribed path when subjected to a sonic boom. For this investi-gation the track. was designed to contain two slalan sections and several different corners requiring the subjects ' full concentration and ccrrplete physical control over the vehicle at all tirres, in order for the course to he successfully negotiated.
This was to prevent driver conditioning and to eliminate his anticipation of the sonic boorn occurrence. In spi te of i ts demanding dlaracter the course did
simu-late actual dri ving behaviour that rnight be encotmtered under conditions that
restriet the driver to folloving a particular path as in fast ITDving rrulti-lane heavy traffic or single lane narrav highways. 'lhis study involved recording the
driver's accuracy in follawing the course (Figs. 33,34,35) by observing the nurnber
and location of the plastic cones that had been disturbed as weIl as recording on
tape the vehicle' s deviation from the centre-line. Special attention was devoted
to both corners F and J where on two separate occasions each driver was subjected
to
a sonic boem. In this study corners F and J were each negotiated a total of24 tirres wi th booms and 96 times without booms and the folloving results were
Observed for all drivers:
Cones were dis turbed on 2 occasions when the driver was boared on corner F. Cones were dis turbed on 1 occasion when the driver was boaned on corner J. Cones were disturbed on 1 occasion when the driver was not boared on corner F.
Cones were never disturbed when the driver was not b()()IIEd on corner J.
Therefore when drivers were booIred on corners F and J, cones were disturbed in 8% and 4% of the cases respectively. Sirnilarly when drivers were not b()()ll'Ed on corner F and J, cones were disturbed in 1% and 0% of the cases
res-pectively.
The infonnation frorn the vehicle track recording system confirrred
our visual observations • Tracking records of three different runs are shavn in Fig. 33. Each record consists of an alphabetically labelled sinusoidal trace rep-resenting the driver' s performance on the track. The letters identify the vehicle' s
location and corresporrl to the corners indicated in Fig. 31. At the bottom of
each record is a straight line wi th one or two vertical narks indicating the tirre
at which the driver was subjected to a sonic bcx:m. Both of these traces were
recorded simultaneouslyon tape and then reprcrluced by a two channel chart recorder, thus preserving their exact tirre relation. The horizontal tirre scale in all
records is one second per rnillirreter; one millirreter being equivalent to one small
grid space. The vertical scales differ and are indicated belaw each tracking record. All traces exhibita kind of sinusoidal appearance indicating that drivers contin-uously deVlated from the centre of the track as they negotiated the course. Accord-ing to the calibration curve (Fig. 30) when either the right or left hand receiving inductor was cIos er to the radiating cable a positive or negative signal was pro-duced respectively. Therefore the tracking records indicate that all drivers had a tendency to take all corners on the outside of the track deviating between 1.5 and 2.0 feet from i
ts
centre. other characteristic shapes appearing in thesetraces include particularly the snall oscillations fotmd superimposed on the peak anpli tudes of the larger sinusoidal waveforms as weIl as the large, very rapid
oscillations having a spike appearances. The forrrer waveform resulted when one of the receiving inductors had travelled same distance imrnediately abave or close
to the radiating cabIe. This is ooserved in Fig. 33 (a) on corners, A,E,F,G,A,
B, and E, as weIl as in Fig. 33{b) on corners E,G,A,D,G, and K. The other spike
waveform resulted when the vehicle corrpletely crossed the cab Ie and the tracking signal was forced into the nonlinear region of the calibration curve (Fig. 30).
This run cx:msisted of two laps with a boom occurring on oorner F on the second lap. In this run the driver disturbed a cone on corner F on the first lap when there was no lxx:xn, but managed to avoid the sarre cone on the second lap when being
sub-jected to a boem. '!he track record indicates that his average speed on the large slalan, between corners A and H, was fifteen miles per hour and that on both laps on oorner F, he deviated approximately two feet from the centre of the course. Similarly Fig. 33 (b) shows the tracking record of driver #6 performing the sarre run as driver #1 described in Fig. 33 (a). Here the vertical scale is different from the previous case, being a factor of 4.4 greater. Again the tracking record indicates that the driver' s maximLrrn deviation fran the centre of the track. on
corner F was similar on both laps, but the driver did hit a cone on corner F during hls seoond lap when subjected to a boom. His average speed on the large slalom was 17 miles per hour. Figure 33 (c) describes the track.ing record of run #3 perforrred by driver #7. The run consisted of two laps with booms occurring on corner F and J on the first and second lap respecti vely . The driver' saverage speed on the large slalom was 19 miles per hour. No cones were disturbed on this run since he deviated only 1.5 to 2.0 feet fran the centre of the course. Sonic boans which occurred on oorners F and J did not result in any cbvious change in his driving performance.
The vehicle track. reoording system was used to provide a continuous record of the path taken by the vehicle with respect to the fixed cabIe. It was hoped that the output of this system could be corrpared to the analogue results cbtained by Lips (Ref. 1). In hls case subjects perforrred a one-dimensional cern-pensatory tracking task which required them to follow a random function signal by rreans of keeping an error bar centered on an oscilloscope display unit. The error signal which the subjects had to minimize was equal to the sum of the simulated vehicle dynamics output and the random function input. The sensitivity and
stabili ty of this system was varied and subjects' response to sonic boans from the UTIAS horn simulator was studied. Continuous graphic records were made of the input, error, wheel output, sonic boom, and scoring signaIs. Lips claims that the exposure to sonic boom disturbances resulted in a variety of responses ranging from negligible change in perfornance to a totalloss of control. The more significant response involved an "initial jerk or holding action" occurring
less than half a second follcwing the boom, and was called an "ini tial" startIe. Following this a 'normal' or 'classic' startIe appeared. This response was charac-terized by both amplitude and frequency change wi th respect to pre-baan conditions as sha-m in Fig. 33 (d), and required an average of 15 seoonds for subjects to re-gain normal oontrol over the track.ing task. Also the peak amplitude response af ter the baan was a factor of two greater than the pre-boom arrpli tude envelope.
The track.ing records fran our actual dri ving experiment did not indicate any similar driver response cbserved by Lips. 'Ihere was no apparent increase in the amplitude of the vehicle 's deviation from the centre of the track, nor any
increase in the frequency to correct the tracking error following a boom occurrence. The tracking records did establish oonclusively that drivers negotiated the course wi thin the boundaries of the track, manoeuvering the vehicle to wi thin inches of
the cones. t'Ti th this in mind the few cases where cones were disturbed during the experiment assume little significance.
The discrepancy between our results and those of Lips may be explained by the fact that his task was rrentally intensive while ours was physically inten-si ve . In hls case subjects never SêM the randan signal that they were track.ing
and only reacted to the tracking error post-factum. 'Ihis required continous and intensive concentration as weIl as anticipation of the randan signal. Any
'initial jerk' or 'holding action ' that resulted from sonic boom startle would then significantly rnagnify the tracking error requiring the subjects to increase the amplitude and frequency of their response in order to correct the increased error.
In our real dri ving experi.rrEnts drivers had a weIl defined track to follc:w which did not require as Imlch rrental concentration as i t did physical control of the vehicle. 'Ihe track also provided the driver wi th several seconds of lead time to prepare for a particular rnanoeuver. Therefore a rnanoeuver during which the boom occurred was not affected since i t was already predeterrnined and only had to be physically executed. Any' ini tial jerk' or 'holding action ' that may have resul ted from the sonic boom startIe would be too short to affect the vehicle's position on the course and thus would not require large amplitude or high frequency correcti ve rnanoeuvers. In addi tion to lead times, drivers also had real-tirre, visual, kinesthetic , static , and audi tory feedback about the per-fonnance of the vehicle, while Lips' subjects had only a visual perception of the tracking error. All of these factors probably contributed to rnaking the vehicle dynarnics and driver systern less sensitive to a sonic boom than had been indicated by Lips.
Several problems were encountered with the vehicle track recording systern. A slc:w but continuous drop in the pover supply voltage had occurred throughout the experirrent affecting the recording levels of the tape recorder. 'Ihis problern was overoome in part by recording calibration pulses before runs began. Ha..-ever the voltage drop did cause the zero or centreline calibration to drift, particularly during runs later in the day. Also the drivers' norrnal per-formance on the track resulting in large deviatian made it difficult to identify any srnall perturbations that rnay have resulted from any 'initial jerk' or 'holding action' that may have been caused by a sonic boom. In general ,the vehicle track recording systern had provided valuable inforrnation about the drivers' nonnal
driving behaviour that would otherwise not have been kncwn. It had also established conclusively that our drivers' perforrning a rea I driving task, did not exhibit the
type of response observed by Lips. 8.2 Stopping Task
'Ihe stopping task was another aspect of dri ving to be studied. lts pur-pose was to determine hc:w a sonic boom would affect a driver 's performance during an errergency situation, in particular it effeGtson the stopping distance. This study was conducted on the straight section of the track near the control site involving the photo-cells, counter, delay trigger circuits and balloons. It con-sisted in rreasuring the stopping speed and distance of the vehicle when i ts path was suddenly blocked by rapidly inflated balloons. For our purpose the stopping
speed was defined as the speed at which the vehicle was rroving when i t passed through the photo-cell trigger circuits, one second prior to balloon inflation. The stop-ping distance was comprised of several different components; namely the distance travelled in one second from the location of the first photo-cell before balloons began to inflate, plus the distance covered during the driver's perception and reaction to the balloons as weIl as the actual breaking distance. 'Ihe photo-cell circuit and the inflated balloons were separated by exactly one hundred and fifty feet. For our convenience, during the tests, the stopping distanee was detennined by rreasuring the distance between the tip of the front bunper and balloons (Fig. 36). '!his data was then converted to actual total stopping distances . '!he driver' s norrnal stopping distance without boorrs was determined at the beginning and end of the testing sequence. During the test the drivers were subjected to sonic booms
prior to, during and af ter balloon inflation . The raw data is presented in Table 11 indicating the different speeds at which drivers were travelling and their corresponding stopping distances . Since the stopping speed at which the drivers were travelling varied from twen-ty-three to thirty-eight miles per hour i t was decided to canvert all stopping distances to correspond to a thirty miles per hour stopping speed using a straight line proportional method with a propor-tionality factor of unity. Table 111 presents the converted stopping distances for a stopping speed of 30 miles per hour as weIl as the rrean and standard devia-tion for each case. This data is presented graphically in Fig. 37, where the srrall circles indicate the average stopping distance and the vertical bars indicate
the standard deviation of the data points.
Upon examining this data, it is noticed that the first time drivers were required to step but were not boorred, their average stopping distance was l3l feet. Next time when they were boorred just prior to being stopped, their distance in-creased to l32.45 feet. In cases that follo.ved no rna.tter when the boom occurred, the stopping distance irnproved from case to case wi th the best average stopping distance beihg recorded on the last run when no boom occurred. The data points also appear to lie on a curve suggesting that rraybe a learning process had taken place from case to case with drivers acquiring or developing better stopping SkilIs. Alternatively, the results may be interpreted to irnply increasing driver antici-pation of being stopped on successi ve runs, thus reducing his reaction tirre which in turn reduced the stopping distance. r~ost probably, elerrents of both learning and anticipation were involved in shaping the outcorre of the resul ts • 'Ihe hypo-thesis that there was essentially no difference between the first two and the last two stopping cases was tested at a 1% level of significance using the Student 's t Test (Appendix A). 'lhe results indicated that there was no statis-tically significant difference in the stopping distances between the first two and last two cases. Hc:wever, when the tests of the same hypothesis was applied to the first and last case when no boom occurred, the hypothesis was rejected, indicating statistically that it is very unlikely that the first and last case beloog to the sarre distribution . But, both of these cases were conducted under
the sarre corrlitions, except that one occurred first and the other last in the testing sequence. Therefore the 't' test would seem to support the above assertion that sorre process other than sonic bcx::ms was invol ved in shaping the outcane of the results. HCMever, if sonic booms did adversel y affect driver be-haviour, then a noticable anomaly would be d:>served in the results. Since the spread between the best and worst average stopping distances was only 15 feet and the standard deviation in each case overlapped each other, i t is difficult to attach special significance to any particular case.
9 • CON:::LUDING REMARKS
The sonic baan pressure wavefonn inside an autonobile was investigated. A portable sonic boom sirrnllator together with other equiprrent necessary to rronitor dri ver behaviour was designed and developed. Two aspects of dri ving were studied. Both the Tracking Manoeuver and Stopping TaSk indicated that driver behaviour was not affected by the simulated sonic boans with 1 msec rise tirres and 3 psf over-pressures. Therefore on the basis of the limited statistical results d:>tained, i t may be concluded that present commercial supersonic aircraft under normal flight conditioos (wi thout superbcx::ms) would not have adverse effects on a driver' s stopp-ing distance or hls abili ty to follo.v a particular course. 'Ihis conclusion may not be surprising in view of the fact that drivers are subjected to far greater
startling aJndi tions during a thunderstorm when overpressures from thurrler-type sonic boem and i ts consequent startle effects may be manyfold greater than those used during the present experirrents. There does not appear to be much evidence in the literature that severe thunder booms have led to autorrobile accidents*.
* Inquiries regarding accidents re sul ting from thunder were addressed to the Collision Data Centre in the Ministry of Transportation in Ontario
and Quebec as well as the Depart:n'ent of Motor Vehicles in Albany, New York. In each case there were no data available to indicate any direct cause-and-effect relation between thunder and automobile accidents.
1. Lips, K.
w.
2. Frabose, M. Parrrentier, G. Mathieu, G. Rigaud, P. Franke, R. Evcard, G. 3. Lanclis, C. Hunt, W. A. 4. Hoffrnan, S. Searle, L. 5. Carothers, R. 6 • Thackray, R. I. Touchstone, R. M. Jones, K. N. 7. Lin, Sui. 8. Vaidya, P. G. 9. DahJke, H. E. Kantarges, G. T. Siddon, T. C. Van Houten, J. J. 10. FaJkiewicz, A. 11. Jolmson, D. R. Rcbinson, D. ~1. 12. Grant, A. C. REFEREt-rnSAn Unstable Steering Task with a Sonic-Boem Disturbance. UTIAS Tech .Note # 179 , 1972 .
Bang sonique de Concorde 001. Rapport Technique
Rl' 3/72 Ins ti tut Franco-Allemand de Recnerches de Saint-lDuis, May 1971.
"'I'he Startle Pattem ", New York: Farrar & Rinehart Inc,
1939.
"Acoustic and Terrporal Factors in the Evocation of StartIe" , Acoust. Sec. Arrer. 43, 269, 1968.
lnitial Calibration and Physiological Response Data for the Travelling-Wave Sonie-Boom Simulator, UTIAS Tech Note #180, 1972.
The Effects of Simulated Sonic Booms on Tracking Per-fornance and Autonomic Response. Federal Aviation Administration Report No. FAA-AM-71-29, 1971.
Sonic Boom Analogues for Investigating Indoor Waves and Structural Response. UTIAS Tech. Note # 158, 1970. The Transmission of Sonie Boom Signals into Rooms
Through
epen
WindCMS . Journal of Sound and Vibration (1972) 25 (4), pp. 533-559.The Shock Expansion Tube and lts Application as a Sonie Boom Simulator LW Research Center, Anaheim, California, Tech. Report #0-71200/7TR-125.
Developrrent of a Loudspeaker-Driver Simulator for Sonie Booms and other Transient Sounds. M.A.Sc. Thesis 1972,
UTIAS.
"Procedure for Calculating the lDudness of Sonic Bangs" , Acoustica Vol. 21, pp.307~317, 1969.
Australian Electronics Connrunications, lREE Proceedings, Sept, 1970, pp.333-335, Australia.
APPENDIX A: ' STATISTICAL ANALYSIS
The purpose of this Appendix is to provide a quick referenoe to the relevant details of statistical decision theory that was used to evaluate the data from the stopping task. Procedures which enable us to decide whether to acoept or reject hypotheses or to determine whether obse:rved sanples differ significantly, ei ther from other sanples or fran expected results, are called tests of hypotheses or test of significance. In testing a gi ven hypothesis , the rnaxirrn.nn prooabili ty wi th which we ~uld be willing to risk rejecting a hypothesis when i t should be accepted is called the level of significance of the test. In
practioe, a level of significance of 0.05 or 0.01 is custanary, although other values are used. If for exarrple a 0.01 or 1% level of significanoe is chosen in testing a hypothesis , then there is one chanoe in one hundred that we would reject the hypothesis when i t should be accepted, i.e., we are about 99% con-fident that the hypothesis is correct.
Since our sarrple size was srrall, we rrust errploy the Student' s t dis-tribution in testing any of our resul ts . Upon examining the average stopping distances for the different cases ShONIl in Fig. 37, i t was decided to test the follCMing hypothesis that there is essentially no difference in the rrean stopping distance beween the cases. Under this hypothesis :
X
l
- X
2
t =-11
+
1
NI N2
where t is the Student's distribution with N
l
+
N2- 2 degrees of freedom.have:
therefore Xl' X
2 are the sarrple rrean values of case 1 and 2 respectively. SI' 8
2 are the sarrple standard deviations of case 1 and 2 respectively.
NI' N
2 are the sarrple size of case 1 and 2 respectively.
~ is the population standard deviation.
If we apply the above hypothesis to the first tw'O stopping cases we
Xl = 131.0 X2
=
132.5 t=
-0.5598inoe this case has 21 degrees of freedom, then on the basis of the wo tailed test at a 0.01 level of significance we would reject the hypothesis if the value of 't' was outside the range of -2.83 to 2.83. Since the 't' value was within the required range, the hypothesis was not rejected arrl we rray conclude that there is essentially no significant difference beween the stopping distanoes of the first two cases. Similar calculations were perfonred for other cases and the results presented in the thesis.
APPENDIX B : QUESTIONNAIRE
'Ihe purpose of the questionnaire was to provide addi tional information which lM:>uld help in the interpretation of the test results. Five questions were asked and the replies to thern are gi ven be lCM .
1.
2.
Rate the effect of the sonic boem as: (a) very startlingi (b) quite startlingi (c) startlingi (d) mildly startlingi (e) not startling.
8% rated the effect of the sonic boem as startling.
50% rated the effect of the sonic boem as mildly startling. 42% rated the effect of the sonic boom as not startling.
Were the booms rrore startling in sare si tuations than in others? 33% answered YES
67% answered NO
The follcwing cOIlllEIlts were also made:
3.
The boem caused confusion during the sequence when the balloons cane up. My reaction to brake for the balloons was delayed. Depends on the severity of the situation and the nunber of prior exposure.
YES, when recovering from a turn.
Turning a sharp corner, the booms had the rrost startling effect. No, since they were expected.
Did the booms affect your driving performance? 42% answered YES
58% answered NO
The follcwing corrrrents were made:
4.
Yes, only in CCilPlex situations
Yes, during a spin, a boom is sorrewhat disconcerting.
Yes, caused nervousness and hence missed part of the track in a lap. Affected stopping before balloons were inflated.
when I heard the boem.
Wanted to brake
For the instant when the boem happens i t tends to distract one' s conoentration ëMay from setting up the proper drift for the next corner.
In your oplillon, would sonic baars affect your dri ving performance under actual dri ving conditions?
42% answered YES 58% answered NO
The follaving ex:mments were made:
5.
Yes, probably only under stress situations or on long uncomplicated driving.
No, i t might wake rre up.
Yes, if Ild never heard a boom before and was in a trouble situation. No, traffic noise masks some of the peak and background level and could no doubt dull the senses as well. Other cars are easier to guide by especially when they are rroving.
I would think it might depend on onels present rrental state, Le., if one was upset over sClIt'Bthing else, a sonic boom may aggravate an already tense situation.
This is an artificial situation demanding full concentration, which rreans i t would take an enorrrous disturbance to distract rre .
Yes, since they would happen unexpectedly, my first reaction would be what part of the car fell off, and i t would tend to be sornewhat disturbing.
No, not very loud.
Yes, in a way, while doing this test, I knew I was going to receive a boom, and was subconciously awaiting it. In actual conditions , the absence of this "subconscious preparednessll
will increase the startle effect.
RON well does a course sirrnllate actual difficult driving situations (e.g., avoiding an accident, errergency stopping, etc)?
Qui te well, al though cornerning, etc., is expected and so easier and the emergency stop does not require constant
vigilance because of only one point possible. RONever , except for anticipation, quite geod.
Concentration required here is prcbably greater than in actual case. You must keep i t up langer here.
Qui te well, except that there is no need to avoid other ITOving traffic in your lane as well as in turns, etc.
Maybe the driver is concentrating continuously for too long. In actual driving, ITOrrents of severe concentration occur only intermittantly.
'!he energency stopping is gcx:d and very realistic . The turns
are very sharp and under realistic conditions (normal speeds) , would prcbably be lltpossible to negotiate.
Course is difficult to see. Asolid line would be easier to follCM.
Not at all. In actual dri ving other cansiderations have to be
made. You are usually restricted to swerving to the right side, for exanple.
