Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
OPERATIONAL SIGNIFICANCE OF THE FLASH BLINDNESS PROBLEM Walton L. Jones Bureau of Medicine and Surgery Bureau of Naval Weapons A pilot who has dropped 17 atomic weapons during various tests describes his experiences in this manner: "At the moment of burst I typically am headed directly away from the burst point. When the burst occurs, the horizon disappears and everything seems to be covered by an overwhelming glow. I can distinguish no colors nor can I see any terrain features. It is as if I am ex- periencing the white-out suffered by aviators flying in Arctic zones." That pilot flies at altitude in a large stable aircraft which can be controlled by autopilot as required. His temporary loss of vision is not as serious as if these events were to occur while he was weaving his way around hilltops and through valleys, at as low an altitude as possible, while trying to maintain an accu- rate navigation course to a target he had never seen before. In- asmuch as many Navy missions are flown under just such cir- cumstances, the problem of flash blindness has come to be regarded as quite a serious one. This paper discusses in some detail the operational signifi- cance of the flash blindness problem, and then describes the major areas of effort being undertaken in attempting both to understand this phenomenon and to provide means for minimiz- ing its hazard. The specific devices which have been developed and which are being evaluated at this time, are discussed else- where in this Report. In considering the effects on vision of the light from a nuclear burst, there are, in fact, two major areas of concern: one deals with the permanent damage, or retinal burns, and the other with temporary impairment, or flash blindness. Fig. 1 presents formal 85
EFFECTS OF VISIBLE AND THERMAL RADIATION FROM NUCLEAR BURSTS ON VISION PERMANENT RETINAL BURNS IRREVERSIBLE TISSUE DAMAGE IS CAUSED BY THE ABSORPTION OF EXCESSIVE THERMAL ENERGY IN THE RETINA AND UNDERLYING LAYERS PRINCIPALLY THE CHOROID FLASH BLINDNESS A TEMPORARY LOSS OF VISION PRODUCED BY OVERSTIMUL ATION WITH VISIBLE ENERGY THE PERIOD OF FLASH BLINDNESS IS THAT PERIOD DURING WHICH THE INDIVIDUAL CANNOT PERFORM HIS DUTIES BECAUSE OF LOSS OF VISION FIG. 1. definitions of these problem areas. Obviously, each area is of deep concern to the Navy. Since careful study of these areas has proved that any protection provided against flash blindness will also protect against retinal burns, attention has been focused on the flash blindness problem. Flash blindness is defined as that period during which the individual cannot perform his duties be- cause of loss of vision. From an operational point of view, the concern is actually with loss of performance capability rather than with loss of vision. Thus, if any features within the cockpit can be arranged to provide benefit, such as an automatic increase in cockpit lighting following exposure, the period of flash blind- ness can be decreased, even though the overstimulation of retinal chemicals has in no way been changed. Such a concept is impor- tant since it allows one to consider means of reducing flash blind- ness above and beyond the obvious one of providing direct pro- tection for the eye. As the problem of flash blindness is placed into operational perspective, one can see that in certain respects it poses a more difficult problem than those produced by the blast, shock, and thermal radiations from the weapon. Fig. 2 shows the area within which flash blindness might occur as related to the ther- mal envelope of the weapon. Note that when the burst point is in the central field of vision flash blindness may occur at a distance as far as several hundred miles from the point of detonation. If the burst occurs behind the pilot, however, relatively little flash blindness should occur unless there are clouds or other highly reflective surfaces within his direct field of view. The impor- tant point to note in this figure is that it is necessary to protect
FLASH BLINDNESS HAZARD RELATED TO POSITION FROM BURST POINT FIG. 2. a pilot from flash blindness at much greater distances from the burst than is required for protection from thermal radiation effects on his skin. To achieve a genuine understanding of the operational impor- tance of flash blindness it is necessary to review certain per- formances which are required of Navy aviation personnel. Fig. 3 lists a few of these activities for which the maintenance of ef- fective vision is critical. The first of these, low-level daylight attack, provides an excellent example. Here the pilot is trying to fly the aircraft, maintain an absolutely minimum altitude, search for navigation checkpoints, and occasionally to monitor certain of his panel instruments. The demands placed on his vision are imposing. Research conducted by the Navy indicates that with present equipment for contour flying a pilot looks out- side approximately 90 per cent of the time in order to maintain geographic orientation. It is no wonder that pilots flying these missions estimate that if their vision were lost for as brief a period as 5 to 10 sec they would in all likelihood either crash or become hopelessly disoriented. Another item in the list of activities in this figure relates to night formation flights. There are certain Navy missions in which one aircraft, which has extensive navigation equipment, will lead other aircraft to target areas. Under these circum- stances, it is absolutely essential that the pilot of the aircraft 87
PILOTS DAY OR NIGHT OPERATIONS ⢠LOW LEVEL ATTACK ⢠TERRAIN CLEARANCE ⢠NAVIGATION CHECKS ⢠INFLIGHT REFUELING ⢠FORMATION FLIGHT ⢠DECK OPERATIONS ⢠TAXI ⢠LAUNCH ⢠LAND CARRIER PERSONNEL ⢠NIGHT DECK OPERATIONS NAVAL AVIATION ACTIVITIES REQUIRING FULL USE OF VISION FIG. 3. being led fly a tight wing position on the lead aircraft. To do this he must be able to perceive the outline of the lead aircraft as well as its running lights. To try to follow by reference to the lights alone leads to autokinesis and subsequent severe seizures of vertigo. Thus, the pilot is in a situation in which he must not only be able to see outside the cockpit at all times, he must also main- tain a high level of dark adaptation throughout the flight. Obvi- ously, in this situation a period of flash blindness of even a few seconds would have devastating consequences. As a final item from Fig. 3, consider the night deck operations performed on a carrier. During night launch operations, for example, flight-deck personnel must maneuver jet starting equip- ment around the deck, guide aircraft to the launch position, and aid in preparing the aircraft for launching. While doing these tasks, they must avoid spinning props, jet intakes and exhausts, and moving tractors. The noise level is such that warning shouts are almost useless. And all this takes place on a deck that, on a moonless night, is illuminated only by red flood lights located seven stories up on the island of the carrier. Here, again, any light which produces flash blindness or even destroys dark adap- tation could be quite serious. Many persons have proclaimed that with airborne radar and 88
radar-controlled intercept missions there is no need for con- cern with night vision. It is true that a pilot no longer tries to locate enemy fighters at night by purely visual means. But, as one can see from Fig. 3, there are many activities performed by Navy personnel, both at night and during the day, for which the maintenance of effective vision is of the utmost importance. Realizing the importance of maintaining the visual capability of Navy personnel under combat conditions, the Navy has had, for several years, an active program to investigate flash blind- ness and to develop protective devices and procedures. Fig. 4 shows the major areas of effort within this program. In an at- tempt to make the program as comprehensive as possible, efforts have ranged all the way from laboratory investigations of flash blindness, using high-intensity light sources, to the preparation of training and indoctrination materials for pilots. NAVY PROGRAM CONCERNING FLASH BLINDNESS PROTECTION AREAS OF EFFORT l LABORATORY INVESTIGATIONS OF VISUAL IMPAIRMENT FROM EXPOSURE TO HIGH INTENSITY LIGHT SOURCES 2 COLLECTION OF DATA FOR MODEL ALLOWING PREDICTION OF FLASH BLINDNESS HAZARD IN OPERATIONAL ENVIRONMENTS. 3 DEVELOPMENT AND EVALUATION OF FLASH BLINDNESS PROTECTION DEVICES. 4 PREPARATION OF TRAINING AND INDOCTRINATION MATERIALS FOR PHOTS FIG. 4. Early efforts were aimed primarily at providing protection from the thermal effects of the weapon rather than at protecting against flash blindness. At least in part, some protection from flash blindness has been provided by these efforts. For example, Fig. 5 shows a cockpit enclosure designed to be closed by the pilot, thereby protecting him from thermal effects. In later air- craft, this protective hood will close automatically when energy from the burst is detected. Thus, some measure of protection from flash blindness will be afforded, particularly from those weapons which produce extended fireballs. A number of studies have shown that the recovery of vision after exposure to a high-intensity flash is much more rapid if the visual task is brightly illuminated. It has been found, for 89
READY POSITION f LIGHT BY VISUA- CONTACT CLOSED POSITION FLIGHT BY INSTRUMENTS OPERATION OF COCKPIT THERMAL ENCLOSUR- FIG. 5. example, that a period of flash blindness that would last from 20 to 30 sec under normal lighting can be reduced to approximately 2 sec simply by floodlighting the visual task with 50 foot-candles. Thus, an obvious means of reducing the period of flash blindness suffered by a pilot is to provide for an automatic increase in the intensity of cockpit lighting immediately following exposure to a burst. A system for automatically increasing the intensity has been developed. While thermal shields and automatic high-intensity cockpit lighting systems do offer some protection against the effects of flash blindness, the most obvious means of providing this protec- tion lies with the development of devices designed to limit the visible energy which reaches the eye. Fig. 6 lists areas which offer protection possibilities. As can be seen, these areas range from the use of a simple monocular eye patch to the highly so- phisticated television devices indicated under indirect vision techniques. AREAS OFFERING PROTECTION POSSIBILITIES AGAINST FLASH BLINDNESS 1. MONOCULAR EYE PATCH 2. PARTIALLY-OCCLUDING (FIXED FILTER) GOGGLES 3. FULLY-OCCLUDING (ACTIVE) GOGGLES 4. PHOTOTROPIC GOGGLES 5. PHOTOTROPIC CANOPY 6. INDIRECT VISION TECHNIQUES FIG. 6. 90
As a means of providing a measure of protection while more efficient devices are developed, the Navy has considered the use of monocular eye patches. Figure 7 shows a pilot wearing an eye patch during a recent evaluation at the Navy Aviation Medical Acceleration Laboratory to determine the extent of visual im- pairment caused by light leaks around the patch. For a time there was concern that the afterimage produced in the exposed eye would cause a cortical blurring of the image received from the protected eye after the patch was removed. Recent limited testing, however, indicates that this blurring is not sufficient to cause alarm. Useful vision is regained in the protected eye almost immediately following the flash, while vision through the exposed eye may remain impaired for many seconds. The major problem with this protective device, however, is that since vision is so critical to most missions the maintenance of vision in one eye is generally regarded as only a poor interim solution to the pro- tection problem. Such patches, however, can be carried by pilots as a means of achieving some measure of protection during emer- gency conditions. For several years, the Navy has been investigating the pro- tection potential of fixed-density goggles. In 1958, general pur- pose light-restrictive filter goggles, known as LRFG-58, were issued. Fig. 8 shows these goggles being worn with the aviator's helmet and oxygen mask. Photometric transmission of these goggles was 1 per cent. It was hoped that they would allow a FIG. 7. 91
FIG. 8. pilot to see well enough to fly during daylight by visual contact, and at night by instruments when the instruments were flood- lighted. In spite of the obvious protective benefit of 1 per cent trans- mission goggles, they are not being used by the Navy at this time. Primary reasons are excessive visual fatigue, limited peripheral vision, and an apparent insufficient transmission of light. Recently, however, the Navy and the Air Force have tested gold-covered visors with no peripheral vision limitations for use during daylight missions. The gold-covered visors also have a photometric transmission of 1 per cent. The improved field of vision and the apparent increased brightness in the yellow band seem to reduce problems of visual fatigue. Pilots who examined these materials initially indicated a 1 per cent coating was too dark to visualize other aircraft making the device unacceptable for daylight use. Accordingly, the Navy has changed to a 3 per cent coating which has met with much greater pilot acceptance and seems to involve only a nominal loss of protection. While fixed-filter goggles might provide adequate protection against retinal burns and flash blindness, they do impose a pen- alty of loss of vision because of the low transmittance of the filter. At dusk or at night this is unacceptable. Considerable effort has, therefore, gone into the development of "active" devices which are transparent normally, but which "close" when exposed to 92
intense light. Another paper in this Report will be devoted spe- cifically to the characteristics of these active devices. As has been indicated, there are a number of avenues being pursued in the quest for effective protective devices. It is quite possible that more than one will be accepted as satisfactory for service use. Should this be the case, a specific protective device could be selected to match the particular hazards of a given mis- sion. However, such use of these protective devices cannot be- come a reality until more is known about the nature of the hazard. Thus, the Navy is developing a comprehensive model within which it should be possible to specify a typical or anticipated nuclear environment, and then to evaluate the extent of the flash blindness hazard for aviators flying missions within such an environment. Figure 9 shows the classes of variables which will be used in the development of the model, and, to some extent, the specific vari- ables falling within each class. One advantage underlying the development of the model is that missing parametric terms in the model can be identified which will thus tend to structure re- search efforts toward a meaningful goal. I.TMil IIIHI 1. IllilUi it tlsmn I. IIIMHI H tain nii li li|ll il mini 1. mull il liliutm i. Licitiii il dunlin iitusitf ni fintiii ). liitiici (in lint 4. litmiliiicil cnlituii t. littictiiitr it lirnii FIG. 9. VARIABLES IN ANALYTIC MODEL FOR PREDICTION OF FLASH BLINDNESS HAZARD One final developmental effort being undertaken by the Navy should be noted. Flash blindness protective devices, as they are delivered into the fleet, represent a class of equipment entirely new to pilots. These devices attempt to meet a requirement for which no previous equipment has been provided. Since they are so new and since many of them operate in a strange manner, their introduction into the fleet must be accompanied by a train- ing program designed to illustrate their method of operation and proper use. In addition, the flash blindness phenomenon itself 93
should be so interpreted both psychologically and physiologically in a manner so that pilots understand and appreciate the inherent hazards. Considerable effort is being devoted at this time to the development of a comprehensive training program. Much of this program involves, as might be expected, training manuals and training films. However, at the heart of this program is what is termed a "flash blindness indoctrination and training device." It consists of a high-intensity flash source which simulates the light that would be encountered by a pilot should he be flying within haze conditions or over reflective terrain at the time of the flash. With this source it is possible to produce all of the features of flash blindness; that is, startle, intense afterimages, and visual incapacitation, without the risk of permanent damage to the visual system. By actually experiencing flash blindness, a pilot will better appreciate the need for protective devices and should be more highly motivated to use them correctly. Some of the reasons have been presented here as to why the Navy considers the problem of flash blindness to be quite serious from an operational point of view, and the extent of the programs being pursued to combat this hazard. In spite of the automatic characteristics of the modern weapon systems, the human eye continues to play a dominant role in many if not all military activities. With effective vision of such importance, it must be protected. 94
THE NATURE OF RADIATION FROM NUCLEAR WEAPONS IN RELATION TO FLASH BLINDNESS1 J. H. Hill and Gloria T. Chisum Aviation Medical Acceleration Laboratory, USN Air Development Center The fireball of a nuclear weapon detonated in the atmosphere appears to radiate as a black body. The visible portion of the spectrum of this radiation produces a high-intensity flash that can cause flash blindnessâthe temporary reduction in visual sensitivity following exposure to a high-intensity flash. Although such a weapon flash can also cause retinal burns, this paper is limited to the nature of the radiation in relation to flash blindness. The minimum information about a light source necessary to determine whether or not it will produce flash blindness, and, if so, to design adequate flash blindness protective devices con- sists of the luminance, duration, and visual angle subtended by the source. Estimations of these dimensions of a light flash with a nuclear weapon fireball as a source can be deduced from infor- mation given in Glasstone, 1962. Although the weapon flash pa- rameters thus determined from scaling laws are, at best, esti- mates, it is hoped that the data presented will provide some guidelines for others interested in the problems of flash blind- ness research and development as they have for the scientists at Vision Research, Aviation Medical Acceleration Laboratory (AMAL). The thermal radiation from a nuclear weapon detonated at low altitude amounts to about 35 per cent of the yield of the weapon. The other 65 per cent is in nuclear radiation and mechanical energy as shown in Fig. 1. The ranges within which the latter 1. Opinions and conclusions in this report are those of the authors and do not necessarily reflect the views or the endorsement of the Department of the Navy. 95
V THERMAL lift RADIATION & 35% INITIAL NUCLEAR RADIATION 5% RESIDUAL NUCLEAR RADIATION 10% FIG. 1. Distribution of energy in a typical air burst of a fission weapon in air at an altitude below 100,000 ft. (Glasstone, 1962) two forms of energy are dissipated to safe levels are much shorter than that for the thermal energy. The variation of the ranges with weapon size is shown in Fig. 2. It can be assumed for purposes of this paper that, except for their eyes, personnel can be adequately protected up to the point at which they would receive second-degree burns. On this basis, one can make use of the concept of equal effects of nuclear weap- FIG. 2. Idealized ranges for effects of air burst with the heights of burst optimized to give the maximum range for each individual effect. (Glasstone, 1962) 96
ons regardless of weapon size. This concept is illustrated in Fig. 3. At 0.1 mile from a 1-kiloton (KT) weapon, or 10 miles from a 1-megaton (MT) weapon, or anywhere to the left of the second-degree burn line, the thermal and mechanical damage may well be so excessive that flash blindness is not a problem. The point at which flash blindness can become a problem is to the right of the second-degree burn line. The second-degree burn distance has been arbitrarily defined as the minimum "safe" distance. Figure 4 shows the thermal history of a nuclear detonation at low altitude. The form of this curveâa very rapidly rising and falling pulse followed by a second, longer, slower rising and fall- ing pulseâis the same for all weapons regardless of weapon size. The time at which the minimum occurs varies with the weapon yield according to the scaling law, t(min) = 0.0025 W 1/2, where 0.1 0.1! O.I 07 1 J I 7 10 SLANT RANGE FROM EXPLOSION (MILES) FIG. 3. Ranges for first- and second-degree burns as a function of the total energy yield. (Glasstone, 1962) 97
TIME AFTER EXPLOSION (RELATIVE SCALE) FIG. 4. Emission of thermal radiation in two pulses in an air burst. (Glasstone, 1962) W is the yield of the weapon in kilotons. The rate of emission, on the other hand, is a function of fireball temperature, which is independent of weapon size. Fireball temperature as a function of time is shown in Fig. 5. Because the emission of the fireball approximates that of a full radiator, the luminance of the fireball can be determined from its effective temperature. With the aid of black-body luminance tables (Pivovonsky & Nagel, 1961), the fireball temperature curve was converted to a luminance curve. The effects of weapon yield and viewing distance were taken into account by translating the curve the appropriate distances along the axes. The minimum safe distance for each weapon and an atmospheric transmission of 80 per cent were used in these calculations. These curves are shown in Fig. 6. The luminances of the smaller weapons viewed at the minimum safe distances exceed the luminance of the sun. The significance of these luminances for visual effects is more readily apparent from the integrated curves shown in Fig. 7. The integrated luminances which would be received by the eyes, protected only by the blink reflex, and by some of the protection devices under development, are indicated along the abscissa. The integrated luminance received with only the blink 98
I7 E 1 \ 2 1 - \ i ._. â ... - k i \ 10> \ T , . -. _^ ^ -CAI.CULATtl) ' ' r^ HIIR 1 SH iXK VE xx: 'IV P 4 \ -1 j * _ . . U \ K i | - i - f-1 2 ._ . _ . s V Ul * Q. V S"'4 f" \ X 7 "J \ â¢" ~~~ / \ â - â » ^ : / v 4 "i ' \ \ * vten )"-- 2 BV ' \ â¢> J . S lfl> ' 4 2 '4 , 2 4 HI ' III ' III" Id" 47 247 1 10 TIMF AFTER EXPLOSION (SECONDSl FIG. 5. Variation of apparent fireball surface tem- perature with time in a 20-kiloton explosion. (Glasstone, 1962) Fireball Luminance Atmospheric Transmittance - 80 Percent «â Sun KT Miles 2 O.66 j 6 20 2 2OO 5.5 1000 10 10,000 23 I 2 3 Time - Log Milliseconds FIG. 6. Fireball luminance for five weapon yields at minimum safe distance with 80 per cent atmospheric transmission. 99
-5 -4 -3-2-l0l2 Time - Log Seconds FIG. 7. Integrated fireball luminances for five weapon yields at minimum safe distances with 80 per cent atmospheric transmission. reflex as protection would be about equal to looking at the sun for one full second. The extent of the retina covered by the image of a fireball will depend on the viewing distance as well as the fireball diame- ter, which is a function of yield and time. The relation of fireball diameter to yield and time is shown in Fig. 8. The fireball di- ameter at any specific time after detonation varies directly with weapon size. The visual angle subtended by a fireball when viewed at the minimum safe distance varies inversely with weapon size as shown in Fig. 9. As the fireball diameter increases the retinal image size increases. This increase in image size stimulates 5 ? S 3 & O Q 1 .12 Yield (Kt) 10,000 ' Breokowoy Time FIG. 8. 0I23 Time- Log Milliseconds Fireball diameters for five weapon yields. 100
0 Foveo Meters 62 19 51 10,000 3^014 Blink Electro- Mechanicol Visual Angle Subtended by_ Fireball ot"Minimum Safe Distance Photochromic Elf -2-I 0 I 23 Time-Log Milliseconds FIG. 9. Visual angles subtended by fireball of five weapon yields at minimum safe distances. new areas of the retina. Only the retina at the center of the image receives the full extent of the fireball luminance. In the event that the fireball itself is not imaged in the retina, other surfaces within the visual field will be illuminated by the fireball. For this reason it is important to know how much illumi- nation a fireball can produce. The illumination received at the minimum safe distance can be determined from the fireball luminance and diameter, viewing distance, and atmospheric trans- mission. This relation is shown in Fig. 10. The luminance of a surface with a diffuse reflectance of 10 per cent is shown on the right ordinate scale. Integration of the illumination and resulting luminance is shown in Fig. 11. The integrated luminance received from a surface with 10 per cent diffused reflectance can be over 5 log mL-sec if no protec- tive measures but the blink reflex are used. This luminance is sufficient to cause flash blindness hazardous to a pilot. From these curves it is readily apparent that for low-altitude detona- tions at the minimum safe distances, the eyes can receive more than 5 log mL-sec of visible energy from a 10 per cent reflector before there is time to blink, and the luminance of a fireball viewed directly can be as much as four orders of magnitude greater. Unfortunately, the unclassified material available for high- altitude detonations is severely limited. The dangers of eye damage from high-altitude detonations can be surmised, however, from the qualitative comparison of the rates of emission for high- 101
ILLUMINANCE FROM FIREBALL ATMOSPHERIC TRANSMITTANCE - 80 PER CENT 9 8 7 rvj 6 :iqooo 0123 TIME - LOG MILLISECONOS o 5 ct O _| -1 E _ I 01 3i- 2 UJ LJ LL rf o: K => o n "" "J 0 U. -" O 1- ss FIG. 10. Fireball illuminances for five weapon yields at minimum safe distance and luminances of 10 per cent re- flecting surface at 80 per cent atmospheric transmission. -3-2 Time - Log Seconds FIG. 11. Integrated fireball illuminance for five weapon yields at minimum safe distances and integrated luminance of 10 per cent reflecting surface at 80 per cent atmospheric transmission. 102
-HIGH ALTITUDE SKA-LEVKL ATMOSPHERK TIME FIG. 12. Qualitative comparison of rates of arrival of ther- mal radiation at a given distance from high-altitude and sea-level bursts. (Glasstone, 1962) altitude and low-altitude bursts shown in Fig. 12. The total ther- mal energy for high-altitude bursts is 50 per cent of the weapon size as compared to 35 per cent for low-altitude bursts, and the rate of emission is also many times greater. The predictions of luminances made in this paper should be considered merely as guidelines. Other factors not considered, such as altitude of detonation, terrain, atmospheric and meteoro- logical conditions, are required for prediction of flash luminances in operational situations. With reasonable knowledge of the lumi- nance that personnel may be expected to encounter in operational situations, adequate flash blindness protection will be devised, and the minimum safe distance for personnel will not have to be extended. REFERENCES Glasstone, S. (Ed.) The effects of nuclear weapons. Washington: U.S. Atomic Energy Comm., 1962, Dept. Army Pamphlet No. 39-3. Pivovonsky, M., & Nagel, M. R., Tables of blackbody radiation functions. New York: Macmillan Co., 196T 103
EXPERIMENTAL INVESTIGATIONS OF THE FLASH BLINDNESS PROBLEM John Lott Brown The School of Medicine University of Pennsylvania It has been made clear that military aviators must depend on vision, in many cases unaided vision outside of their aircraft, for satisfactory performance in a variety of situations. High energy flashes from atomic weapons can seriously impair visual function at ranges beyond those at which damage to an aircraft or injury to the pilot may occur. The design of the eye for light detection renders it uniquely susceptible to excesses of light at ranges well beyond those where light energy may affect any other part of the body. There is, thus, an operational requirement for providing protection against visual incapacity in ranges from those within which damage to the aircraft or injury other than to the eyes of the pilot will jeopardize a mission out to a range of complete safety. The first step in the development of appropriate protection is the specification of the physical conditions against which protec- tion must be afforded. The previous paper provides information concerning the temporal and spectral variations in energy from atomic weapons of a number of different yields. The relation of effect to range is also discussed. A second step in the develop- ment of effective protection is the determination of how vision is impaired by exposure to unusually high energy levels in spectral and temporal patterns that are representative of those to be en- countered from atomic weapons. One approach to the problem is to attempt field tests of visual function and visual impairment in conjunction with atomic weapon tests. Field tests may provide valuable information relative to the effectiveness of protective devices and general reactions to weapon detonation, but they can 104
hardly be expected to provide much information relative to the retinal processes involved. Such tests are always on a noninter- ference basis, i.e., secondary to engineering questions and con- siderations. Field test attempts that have been made indicate that the requisite attitude of scientific deliberateness is difficult for either experimenters or subjects to achieve in the awesome proximity of an atomic blast. In one such attempt, no evidence of any flash blindness effect was found. It was later ascertained that all subjects, including the project director, had their eyes tightly closed prior to and during the blast. The count-down was audible to all, and the final syllable produced an involuntary lid closure in all concerned. There are two additional factors which tend to limit the value of attempts to study basic visual processes in conjunction with atomic weapon tests. In the first place, it is never possible to specify beforehand what the precise physical characteristics of a detonation will be, and, in the second place, there is almost no chance of replicating a given set of conditions. These considera- tions all argue for the study of basic visual functions which relate to flash blindness in the laboratory. Field tests, if they are ever again permitted, will afford a knowledge of the physical conditions to be encountered and will permit evaluation of protective devices and techniques, but information on the responses of the eye, from which necessary characteristics for protective devices can be specified, is best obtained in laboratory experiments. It is the purpose of this report to consider flash blindness in relation to laboratory studies of the important parameters of the problem. The effects of these parameters are discussed in the succeeding sections of this paper. Flash blindness in the sense in which it is used in this report is a reversible reduction in visual capability which may incapacitate someone for the per- formance of visual tasks. A case of paramount importance is that of the pilot of a high-performance aircraft. Flash blind- ness may be defined quantitatively in terms of the duration of incapacity, or time for recovery from the effects of a blinding flash. Incapacity and recovery must, of course, be measured in terms of some specific visual task, and will vary in duration with the nature of the task. Such visual-task parameters as luminance and visual-acuity requirements are considered spe- cifically in this report. 105
PARAMETERS Energy of the Flash It is useful to consider the way in which flash blindness, ex- pressed in terms of recovery time, may be expected to vary with the energy of the blinding (adapting) flash. Flash duration is an important variable and a subsequent section is devoted to it, but in this section, flash duration is considered a constant. A hypo- thetical relation between the energy of an adapting flash and recovery time for a specific visual task is illustrated in Fig. 1. Recovery time is on the ordinate; adapting-flash energy is repre- sented on the abscissa (for constant flash duration and constant spectral distribution, adapting-flash energy is proportional to radiant or luminous flux). For very low adapting-flash energies, there will be no effect on visual capability, and recovery time will be minimal. As energy is increased, there will be an increase in recovery time at an increasing rate. The form of this function will depend on the nature of the visual task. As the energy of the adapting flash reaches a level that corresponds to maximum pos- sible bleaching of photosensitive substances of the retina, the rate of increase of recovery time may be expected to reduce. Recovery 100 0 E i- > o o _VISUAL REACTION_TIME_ I I I I I Log Adapting Flash Energy FIG. 1. Hypothetical curve illustrating relation between en- ergy of blinding flash and time required for detection of in- formation in a visual display. Minimum detection time at low flash energy corresponds to visual reaction time. Detection time approaches infinity as flash energy approaches value which will cause irreversible injury. 106
time may actually assume a constant value over some range of adapting-flash energy beyond that at which maximum retinal bleaching has occurred. As energy is further increased into the range where injury may occur, it is to be expected that recovery time will again in- crease. The mechanism of recovery in this range will no longer be the same as that of normal adaptation, and as the extent of the injury increases, recovery time will increase at an increas- ing rate, reaching an infinite value when the injury becomes irreversible. Several experiments have been performed which illustrate various parts of the function presented in Fig. 1. Metcalf and Horn (1958) employed a 0.1-second (sec) adapting flash with a diameter of approximately 4° of visual angle at luminances from 46,000 foot-Lamberts (ft-L) to over 9,000,000 ft-L. The natural pupil was dilated and an artificial pupil was employed. Results were expressed as the time required for two successive detec- tions of a 17-min circular test patch at a luminance of 0.07 ft-L, which flashed on and off at 1-sec intervals. Results for each of four subjects are illustrated in Fig. 2. The independent variable, adapting illumination, is presented as a function of the dependent variable, recovery time. This changes the appearance of the functions somewhat by indicating variability in adapting luminance with increasing recovery time rather than the actual relation. This relation is a somewhat variable one, but can be approxi- mated by a straight line. Thus, for the range of adapting-flash luminances represented, recovery time increases from a few seconds to over 2 min in direct proportion to the logarithm of the adapting-flash energy. Recently, Miller (1964) has obtained similar data for adapting energies of from 5.94 to 7.93 log troland-seconds with flash durations of from 42 microseconds O^sec) to 1.4 milliseconds (msec). Time required after exposure to an adapting flash before a subject could correctly identify a test letter illuminated at 0.066 millilamberts (mL) provided a measure of the duration of flash blindness. Miller presented her data in the form of recovery time versus log adapting-flash energy in troland-seconds. She transformed the data of Metcalf and Horn (Fig. 2) to troland-seconds for comparison, and found the two sets of results to be almost superimposed. Such a comparison is difficult to evaluate, however. Miller employed an artificial pupil of 4 mm^ while Metcalf and Horn employed a 6-mm diame- ter pupil. Under these circumstances for equal troland-seconds, the adapting flash presented by Miller should have been the more 107
7711300 * RECOVCKr TIHC ISCCOMDS/ FIG. 2. Time required for detection of 17-min circular test patch at luminance of 0.07 ft-L as function of lumi- nance of adapting flash. Mean values for each of four observers (Metcalf and Horn, 1958). detrimental, since it was presumably confined to the central re- gion of the pupil and not diminished by the Stiles-Crawford effect. However, Miller employed a recognition criterion, and Metcalf and Horn employed a detection criterion. Under these circum- stances, equal detection times would indicate that the conditions of adaptation employed by Miller were less detrimental. The relation between adapting luminance and recovery time has also been discussed by Whiteside (1960). He has presented data from several experiments (Fig. 3), which, when replotted in terms of recovery time versus log adapting-flash energy, show an approximately straight-line relation between these variables over a wide range. The data from one experiment (Whiteside & Bazarnik, 1952) which extend up to a higher level of adapting- flash energy (greater than 3 x 10^ mL-sec) show a sharp in- crease in recovery time at this level. Whiteside has suggested that this may reflect the beginning of retinal damage. 108
â¢tcovtrr TIMI FIG. 3. Recovery times for detection of criterion targets after exposure to adapting flash as measured by several investigators. BHC curve ob- tained by Crawford with 0.14 ft-L target. US curve represents data of Metcalf and Horn. Other data obtained by Whiteside (Whiteside, 1960). Some of the results of a flash blindness experiment performed by Hill and Chisum (1962) are presented in Fig. 4. Recovery was measured in terms of the time required for detection of the orien- tation of a grating pattern. This pattern consisted of parallel lines and spaces 3 min of arc in width, and hence required the function of cones for its resolution (Brown, 1954). Data were obtained for three grating-pattern luminances and for each of two adapting- flash durations. It is clear that each of these parameters has an important influence on recovery time, and they will be considered in detail in subsequent sections. Of interest here is the fact that for very low adapting-flash energies recovery time is short, and as adapting-flash energy increases, recovery time increases at an increasing rate. There is some indication that recovery time may approach a plateau for adapting-flash energy of greater than 4.5 log mL-sec (data for the longer flash duration). 109
90 80 * 50 P 40 f 30 § 20 cr I0 â¢JHH Adopting flash duration 9.8 m sec Display Luminance Log mL -0.75 -0.75 I 23456 Adopting Flash - Log (mL-sec) FIG. 4. Time required to perceive acuity target as function of energy of adapting flash (log millilambert- seconds). Individual curves represent each of two adapting-flash durations for three acuity target illumi- nations (Hill & Chisum, 1962). In summary, several experiments have been performed that provide a relation between recovery time and the energy of an adapting flash. No single experiment has covered a sufficiently wide range of adapting luminances to illustrate both extremes of the relation presented in Fig. 1. When considered together, the experiments of Hill and Chisum (1962), and Whiteside and Bazar- nik (1952) do support the kind of relation proposed in Fig. 1, however. It is obviously impossible to investigate the upper range of this relation extensively with human observers. Flash Blindness Criteria Comparison of experimental studies of flash blindness is ren- dered difficult by the various criteria of flash blindness which have been employed by different investigators. It is, therefore, of considerable importance to consider further the effects of changing various parameters of the visual task which serves as a criterion of flash blindness. Luminance. Data are presented in Fig. 4 for each of three display luminances: -0.75, 0, and 2.25 log mL. It is clear that an increase in display luminance has a highly significant effect on recovery time, particularly at the high adapting-flash energies where flash blindness may be a problem. The hypothetical rela- tion in Fig. 1 may be illustrated in a more general way by taking 110
the effect of display luminance into account. This is illustrated in Fig. 5. The uppermost curve represents a relatively low dis- play luminance. Increasing display luminances are represented by successively lower curves in the figure. The minimum recov- ery time at low adapting-flash energy for the individual curves is shown to decrease with increase in display luminance. This relation is slightly exaggerated, but visual reaction time will de- crease to a minimum value with increase in the luminance of a visual display (Teichner, 1954). All of the curves show the final increase in recovery time toward an infinite value at the same adapting-flash energy. This characteristic of the curves is at- tributed to retinal injury which will be dependent solely on the characteristics of the adapting flash. It is possible, of course, that the effect of injury on recovery time will vary with the cri- terion task, but for a given task with changes only in display luminance, the relations presented in Fig. 5 would seem to be the most probable. The nature of the relation between recovery time and display luminance can be inferred from Fig. 5. In Fig. 5, each of the vertical dashed lines which cut through the family of curves for 1OO C o u « 10 0 E o u c V. Log Adapting Flash Energy FIG. 5. Hypothetical functions like that in Fig. 1 for various lumi- nances of display. (Display luminance assumed to increase in equal logarithmic steps from low value for top curve to high value for bottom curve.) Vertical dashed lines are construction lines for derivation of functions of kind illustrated in Fig. G. Ill
different display luminances represents a specific adapting-flash energy. It is evident that as display luminance is increased there will be a decrease in recovery time at a decreasing rate down to a minimum value. This relation will be steeper and have a higher final minimum value of recovery time, the higher the adapting- flash energy. Both Metcalf and Horn (1958) and Whiteside (1960) recognized the possible importance of the luminance of the criterion task on flash blindness, but they obtained very little data relevant to this point. In order to clarify this aspect of the problem, experiments were performed in which display luminance was a primary inde- pendent variable (Brown, 1959; 1964). Recovery time was mea- sured as a function of display luminance for each of several adapt- ing-flash energies. Adapting-flash duration was held constant at 0.9 sec. The display consisted of a grating pattern, and observers were required to identify its orientation. A timer was started and the grating was illuminated with presentation of the adapting flash. As soon as grating orientation was detected the observer depressed a switch that turned off the timer and the display. De- tection times were recorded only for correct identification of grating orientation. The relation of detection time to display luminance is illustrated in Fig. 6 for each of two observers and for each of two grating sizes. The families of curves in the upper part of the figure represent results with a grating, the individ- ual lines of which sub-tended a visual angle of 3.8 min of arc. Detection of the orientation of such a grating will depend on cone vision (Brown, 1954). The lower curves were obtained for a grating which represented a visual angle of 12.5 min of arc. Rods may serve in detection of the orientation of the coarser grating (Brown & Woodward, 1957). All of the data in Fig. 6 have been fitted with curves based on a single equation. The fit is good for data of this kind. (cf. Severin, Newton, & Culver, 1962; Severin, Alder, Newton, & Culver, 1962). Data for the two observers are quite similar. As display luminance is reduced, detection time increases at an increasing rate down to a display luminance which represents the absolute threshold for the discrimination of the display pattern. At high display luminances, there is little dif- ference in recovery time for different adapting luminances, although recovery time is slightly longer for the two highest adapting luminances. It is clear that display luminance is an extremely important variable, and that flash blindness effects can be significantly reduced by raising display luminance. 112
SUAL ACUITY⢠2fi FIG. 6. Relations of perception time to log display luminance (foot-lamberts) for each of six adapting-flash luminances. Upper graphs represent grating display that required visual acuity of 0.26; lower graphs represent visual acuity of 0.08. Subjects JB and FS (Brown, 1964). The criterion task. It is clear that a change in the visual acuity required for the visual task may alter the form of functions such as those presented in Fig. 6. Specification of the visual task in terms of acuity requirements is not sufficient to define the function, however, as visual acuity is an arbitrary and incomplete description of any visual display (Brown, Phares, & Fletcher, 1960). The nature of the function relating luminance and visual acuity varies significantly with changes in the kind of test pattern employed (Shlaer, 1937). Nevertheless, it can be concluded that differences in the course of recovery from flash blindness for different criterion visual tasks will be smaller for higher illumi- nation of the tasks and for lower adapting-flash energies. Differ- ences will be most prominant when the illumination provided is near threshold for one of the tasks. 113
Duration of the Adapting Flash The way in which the energy of an adapting flash is distributed in time may be expected to influence its effect. For the extreme case of retinal injury, it has been found that the threshold for a retinal burn is approximately 0.5 to 1.5 calories/centimeter^ (cal/cm^). This energy must be delivered at a rate of at least 0.7 cal/cm^-sec, however, or the rate of heat dissipation in the tissue will be sufficient to prevent elevation of the temperature to a degree where a burn will result (DeMott & Davis, 1959). The distribution of adapting-flash energy in time may also influence its effect on visual threshold. From studies of dark adaptation, there is evidence that for higher adapting luminances of shorter duration the early thresholds during dark adaptation are higher than those found after light adaptation of the same total energy with exposure of a lower luminance for a longer duration (Johannsen, 1934; Wald & Clark, 1937; Winsor & Clark, 1936; Mote & Riopelle, 1953). These experiments have not in- cluded the systematic investigation of very short duration, i.e., less than a second, however. Fry and Alpern (1951) measured changes in extrafoveal visual acuity for a line target illuminated at 0.01 ft-L during dark adaptation. Light-adapting luminances from 0.0137 to 13,660 ft-L for durations of from 0.003 to 3.0 sec were used. The earliest measurements of acuity were made 10 or 15 sec after the termination of light adaptation and showed considerable variability. The form of the function showing re- covery of visual acuity with time in darkness was not affected by duration of light adaptation, but appeared to be determined by the product of the luminance and duration of the light-adapt- ing flash, i.e., the total luminous energy. In the experiment by Miller (1964), the effect of exposure to an adapting flash on time required for recognition of dimly illuminated letters was also found to depend on total luminous energy, independent of flash duration. This was true for durations from 42 usec to 1.4 ms. Miller concluded that a reciprocal relation between luminance and duration of the adapting flash probably holds up to a duration of 100 ms on the basis of a comparison between her data and those of Metcalf and Horn. Differences in experimental proce- dures referred to above may render this comparison unjustified, however. With respect to the purely photochemical aspects of light adaptation, the reciprocity relation between luminance and dura- tion for bleaching of rhodopsin has been demonstrated by direct 114
measurement in recent years. Campbell and Rushton (1955) measured rhodopsin density in the living human eye and found reciprocity to hold for exposure durations of up to 48 sec. For equivalent total amounts of bleaching energy, equal amounts of rhodopsin were bleached during exposures up to this duration. The minimum time investigated was 300 ms. Some evidence obtained with this technique has been presented that would lead to the conclusion that the effects of adapting flashes of very short duration might not be as severe as the effects of longer flashes of the same total energy. Hagins (1956) found it impossible to bleach more than 50 per cent of the rho- dopsin of the rabbit retina with flashes of less than 1 ms duration no matter how high the luminance. If the same amount of energy was distributed between two flashes separated by 1 or 2 sec, it was possible to bleach up to 75 per cent of the rhodopsin. Bowling and Hubbard (1963) have explained this result in terms of under- lying photochemical processes. A portion of certain unstable intermediate products of bleaching are isomerized back into photosensitive forms by light itself. With prolonged exposure, these are again bleached, a lesser portion returns to a photo- sensitive form and so on, until bleaching has reached the maxi- mum possible for the luminance used. Thus, complete bleaching requires both light energy and time. When an adapting flash is of the order of 1 ms duration or less, bleaching cannot be as great as that which will occur for the same or even lesser amounts of light energy spread out more in time. Early thresholds mea- sured after a short adapting flash will, therefore, not be elevated as much as those after a longer duration. Dark adaptation follow- ing a short flash is not equivalent to dark adaptation following longer exposure to a low luminance which results in the same amount of bleaching, however. A part of the dark-adaptation process, hydrolysis of all-trans bleaching products, can occur during exposure to light. Hence, when a longer duration adapting light is extinguished the process of recovery is at a more ad- vanced stage than is the case after exposure to a short flash, even though photosensitive pigment concentrations and initial thresholds are the same. Complete recovery will take longer after the short flash. Brindley (1959) has made some observations of afterimages that probably depend on these photochemical effects. Afterimages induced by flash luminances in excess of 3 x 10^ meter-candles were of the same appearance for all luminances as long as the total flash energy was presented in a short interval. The after- 115
image following a single flash was comparable to the afterimage of two successive flashes of the same luminance if the two were separated by only 250 /isec. If the two flashes were separated by 4 ms or more, however, there was a clear difference in the after- images. Thus, the additional energy of the second flash appar- ently had no effect unless it irradiated the retina at an interval of several milliseconds after the first flash. The findings of Hill and Chisum (1962) presented in Fig. 4 may also be an illustration of these effects. The curves of re- covery time versus adapting-flash energy all indicate a more rapid recovery from a short adapting flash (165 jisec at one-third amplitude) than from a long adapting flash (9.8 ms at one-third amplitude) of the same total energy. When the energy is distrib- uted over a longer time period, it is apparently more effective in the production of flash blindness. The experiment of Hill and Chisum is the only experiment in which extensive data on flash blindness has been obtained for flash durations both below and above a duration of from 1 to 4 ms. This is the range of dura- tions below which Hagins and Brindley found a reduction in the effectiveness of a given amount of stimulus energy for the bleach- ing of rhodopsin and for the production of an afterimage. The flashes employed by Hill and Chisum were produced by gas discharge tubes. Luminance rose sharply to a maximum, and then decayed exponentially. The spectral compositions of the long and short flashes were not identical, and the spectral composition of each may have varied with time during a single flash. In order to avoid difficulties associated with physical measurements, luminance of these flashes was measured in relation to threshold for resolution of a grating pattern with 3-min bars and spaces. Resolution of a grating of this size will depend on cone function (Brown, 1954). Threshold luminance was first measured with 8-ms flashes controlled by a mechanical shutter, and illumination from a uniform source of known lumi- nance attenuated with neutral filters. Luminous energy at thresh- old was calculated by integrating threshold luminance over flash time. Thresholds were then remeasured with the gas-discharge tubes as sources of illumination, again attenuating luminance with neutral filters. It was assumed that at threshold the lumi- nous energy was a constant, independent of changes in its temporal distribution (Brindley, 1952). Thus, the conclusion that the adapt- ing effect of light flashes of a given energy is not constant when Hash duration is reduced depends on the validity of the assump- tion that threshold energy is constant for similar changes in 116
flash duration. Verification and extension of the results of Hill and Chisum with a technique such as that employed by Miller (single source of known luminance; mechanical control of flash duration) will have important implications for the theory of visual function. Spectral Characteristics of Flash and Task Illumination Some studies of flash blindness with various wavelength distri- butions of the adapting flash have been made, but they have been limited to low-flash energy levels by the use of selectively ab- sorbing color filters for control of spectral distribution. The primary purpose of these studies has been to examine the pos- sibility of protecting the eyes from flash blindness with color filters in the form of goggles, or in the form of a tinted aircraft canopy. In the first case, both the flash and the visual task would be viewed through the color filter, and the spectral distribution of both flash and task illumination would be modified accordingly. In the second case, the flash would be viewed through the color filter, but any task related to the instrument panel would be seen with its normal illumination unaltered. Some data which illustrate the first case (Brown, 1959) are presented in Fig. 7. Each of five color filters was used with neutral filters such that the combination transmitted approxi- mately 1 per cent of the visual task illumination (based on the photopic luminosity function). When the visual acuity require- ment is 0.33, there is a regular increase in detection time at low illumination as the wavelength of maximum transmission of the color filter is decreased. No such regular increase is evi- dent in the results for the 0.13 visual-acuity requirement. The explanation of these results probably lies in the fact that the luminance of the flash as seen through the blue filter was some- what higher than the flash luminance seen through the red filter. This follows from the fact that the spectral distribution of energy in the flash was concentrated at shorter wavelengths than the energy of the task illumination, and the matching of photopic transmission of the various color filters was based on the task illumination. At low task luminances, the available illumination was not far above threshold for the finer grating, and its identi- fication was, therefore, quite sensitive to changes in adapting- flash luminance. For the coarser grating pattern (visual acuity equals 0.13), there was no regular increase in detection time with change in the color of filters from red to blue. Detection of the coarser grating depended to some extent on scotopic 117
SuD|ect JU i â¢v MB ^^ - ^v. MO a. 5 cr tf Adapting Ftash Luminance - IOO.OOO Ft Lomberts Display Visual Acuity Req 0l3 033 DISPLAY LUMINANCE-LOG FT LAMBERTS FIG. 7. Time required for perception of orientation of grating displays following exposure to 0.9-sec adapt- ing flash of 100,000 ft-L. Color filters in front of eyes, both during exposure to flash and during viewing of grating. Filters are Corning 2412-red, Bausch & Lomb 600-red interference, Bausch & Lomb 545-green in- terference, Kodak 52-green, and Corning 4305-blue. Neutral filters added to each of color filters to re- duce total photometric transmittance to 1 per cent (Brown, 1959). vision. The desensitizing effect of the higher luminance of the blue flash was partially offset by the higher effectiveness of the task illumination as perceived through a blue filter for stimula- tion of the rods. If colored protective goggles are to be worn continuously, and all visual tasks must be performed with goggles on, an interest- ing problem arises as to the optimum density of the goggles. Clearly, the greater the density the greater the protection af- forded, but if the density is high and the light available for vision is relatively low, very dense goggles may seriously interfere with visual performance. It has been shown that for illumination of an 8-min grating pattern below 1 ft-L, recovery from the effects of a 100,000 ft-L flash will be faster for goggles with 10 per cent transmittance than for goggles with 1 per cent trans- mittance, even though the denser goggles permit only one-tenth as much of the adapting energy to enter the eye (Brown, 1959). If the protective filter attentuates only the adapting flash and not the illumination of the visual task, shorter wavelength trans- mittance is bound to be accompanied by an increase in detection time for a coarse grating at low levels of illumination of the 118
visual task. This may be attributed to the increased desensitiz- ing effects of the shorter wavelength adapting flash on the rods, which, in this case, is not compensated by a more effective spectral distribution of illumination of the visual task. It is difficult to predict the result of investigations in which spectral variations are included. These results depend on the population of receptors required for the visual task, the thresh- old luminance for performance of the task, the spectral charac- ter and the energy of the adapting flash, as well as the spectral character and the luminance of the visual task illumination, and the spectral transmission characteristics of any filters employed. All of these factors must be taken into account. OTHER CONSIDERATIONS There are a variety of additional variables that are of importance. Several of them should be mentioned here. Severin et al. (1963) have demonstrated that pupil size is a relevant variable in deter- mining the amount of the flash blindness effect upon exposure to a given adapting-flash condition. The larger the diameter of the pupil and, hence, the larger the retinal illumination, the greater will be the duration of flash blindness. The effect is approxi- mately proportional to pupil area. Miller (1964) has shown that the area of the adapting-flash image on the retina has little effect as long as the entire fovea is illuminated. The criterion task in this investigation was recog- nition of a 20.3 min of arc test letter illuminated at 0.066 mL. A characteristic of flash blindness experiments is the rela- tively high variability of results from one subject to another (Severin, ^t al., 1962: 1963). When very short adapting flashes are used, a slight deviation of gaze or blinking during presenta- tion of the flash can appreciably reduce the flash blindness effect. There is some evidence that with an increase in subject experi- ence, the results of different subjects become more nearly simi- lar, and, at the same time, the flash blindness effect is somewhat greater than that found with inexperienced observers (Brown, 1964). A number of investigators have discussed the problem of flash blindness in relation to the masking effects of the afterimage of the adapting flash. Actually, an afterimage disappears and re- appears, and its moment-to-moment appearance bears little cor- respondence to the moment-to-moment changes in light-detection threshold during dark adaptation (Aulhorn. 1958). Barlow and Sparrock (1964) have attributed this to the fact that afterimages 119
are stabilized on the retina, and, therefore, must be compared with a physically stabilized retinal image if one wishes to evalu- ate their masking effect in relation to a physical stimulus. The work of Barlow suggests that changes in threshold during dark adaptation may be closely linked to the masking effect of the afterimage of the light-adapting stimulus. THE QUANTITATIVE FORMULATION OF FLASH BLINDNESS EFFECTS From the practical standpoint, it would be useful to be able to express flash blindness effects in the form of an equation. Such an equation would permit the calculation of perception time for any combination of adapt ing-flash energy and display luminance over a wide range. A more useful application might be the cal- culation of required display luminance for the achievement of a perception time within some maximum acceptable time following exposure to a given adapting-flash energy. Several approaches to the development of an appropriate de- scriptive equation have been investigated. The first of these was a consideration of an equation of the same form as one with which Rushton (1961) has been able to fit data of dark adaptation. Al- though techniques of making measurements in flash blindness experiments differ from those which are usually employed in studies of dark adaptation, flash blindness experiments are es- sentially studies of the early dark-adaptation process following exposure to short adapting flashes. Rushton has found that the logarithm of threshold luminance during dark adaptation is an exponential function of elapsed time from the termination of light adaptation. The equation provides a reasonable fit of rod dark adaptation, provided 25 per cent or more of the available rho- dopsin has been bleached. Such an equation did not provide an adequate fit of the data of a representative flash blindness ex- periment (Brown, 1964). Its failure to do so may be attributed to the fact that perception times probably depend on cones rather than rods in most instances in the study of flash blindness, while Rushton's formulation was developed to represent recovery of the rods. It is also possible that the perception of targets within seconds after exposure to an adapting flash may be limited to a greater extent by neural events than by photochemical events. Once the parameters of the experimental procedure have been determined, the time course of dark adaptation in Rushton's formulation is entirely dependent on photochemical events. 120
The second equation considered was one proposed some years ago by Schouten and Ornstein (1939) to relate threshold luminance for the detection of a light to the time interval following exposure to an adapting flash. An equation of this form failed to provide an adequate fit of the data of a flash blindness experiment (Brown, 1964) when the values of various fixed parameters of the experi- ment were substituted appropriately in the equation. Finally, several simple relations similar in form to the curves presented in Fig. 6 were investigated. These included reciprocal and exponential relations between log display luminance and time. In each case a minimum display luminance and minimum percep- tion time were introduced. Minimum display luminance was based on the known minimum threshold luminance in the dark-adapted eye for the gratings employed. Minimum perception time was either a single value based on minimum visual reaction time (0.20 sec), or minimum perception time was permitted to vary with the luminance of the adapting flash. In the latter case, an attempt was made to estimate the minimum perception time by averaging minimum values for all observers for each adapting - flash luminance. Experimental data were transformed to correspond to a straight-line transformation of each of the equations under in- vestigation and replotted. An attempt was then made to fit straight lines to these distributions of transformed data. On the basis of visual inspection, one equation appeared to afford the best fit for all observers. This was an equation of the following form. t - tQ = a + [b/(log L - log LQ)j, (1) where: t = perception time in seconds, tQ = 0.20 sec for all conditions of adaptation, L = display luminance in foot-lamberts, and Lo = minimum luminance at which the display can be per- ceived under optimum conditions. (-1.4 log ft-L for visual acuity of 0.26; -2.3 log ft-L for visual acuity of 0.08.) Perception time t must approach a minimum tQ as display luminance L is increased, and it may safely be assumed that a value L will be reached beyond which there will be no further 121
reduction of t. If this is the case, then the constant, a, must be of the following form: a - -lb/(log L - log L )|, (2) max ' o where L, ,,^ is the luminance at which t is reached. nr«ix u When perception time was plotted as a function of 1/logL - log Lo, the lines of best fit for the various adapt ing-flash lumi- nances tended to converge in the region of t = 0.2 sec. For data representing a visual acuity of 0.26, the corresponding value of l/(log L - log LQ) was approximately 0.25. For the data repre- senting a visual acuity of 0.08, the corresponding value was ap- proximately 0.20. Values of log Lmax can be calculated from these values, since they correspond with the minimum percep- tion time, t = 0.2 sec. Values of log Lmax calculated in this way were 2.6 for the higher acuity, and 2.7 for the lower acuity. These values may be considered equivalent in view of the ap- proximate nature of the entire fitting procedure. Assuming a value of 2.7 for log Lmax, Equations (1) and (2) can be com- bined and rewritten as follows: t = 0.2 + b [(2.7 - log L)/log(L/L )(2.7 - log L ) |. (3) The smooth curves drawn through the data of observers FS and JB in Fig. 6 represent this equation. Values of b were obtained graphically from the straight-line transformations of each ob- server's data. The equation obviously does not afford a perfect fit of the experimental results, but in most cases it provides a reasonable approximation. The value of b is the slope constant of the straight lines of best fit. It varies both with the luminance of the adapting flash and with the visual acuity required by the display. Relations between average values of b determined graphically for JB and FS and the energy. A, of the adapting flash were investigated for each acuity. It was found that the logarithm of b was proportional to the logarithm of A for the data representing both of the two acuity levels. A simple power function, therefore, serves to relate b to A: 0 SR b008 = 0.108A- (4a) b0.26 ' 122
where A represents adapting flash energy in foot-lambert- seconds. Either of these values of b may be substituted in Equation (3) along with the appropriate value of log LQ for the calculation of perception time. Equation (3) thus affords a basis for approximat- ing perception time for conditions of adaptation and display lumi- nances over the ranges employed in the experiment. It must be emphasized that this is an approximate, empirical relation. None- theless, it represents an objective basis for estimating over a fairly wide range of conditions the blinding effects of short adapt- ing flashes in terms of the extension of perception time that may be expected to result. There are some conditions for which these equations do not hold, however. With high target or display illumination, the mini- mum time required for the detection of a given display appears to be somewhat longer when the adapting-flash energy is higher. This is illustrated in Fig. 5. When adapting-flash energy reaches a value somewhat below that which corresponds to the plateau preceding the final increase in detection time in Fig. 1 and Fig. 5, Equation (4) will no longer hold. The results of Hill and Chisum presented in Fig. 4 illustrate the significance of flash duration. Curves for both flash durations were cut by vertical lines at adapting-flash energies of 3.9, 4.2, and 4.6 log mL-sec. The intersections of these vertical lines with the functions for each of the three grating display luminances provided coordinates in display luminance and recovery time with which to check the adequacy of Equation (3) for describing the results of Hill and Chisum. The logarithm of the minimum lumi- nance (Lo) for resolution of a 3-min grating (visual acuity equals 0.33) is approximately -1.2 (Brown, Graham, Leibowitz, & Ran- ken, 1953). In Fig. 8, recovery time is plotted as a function of the reciprocal of log L/LQ for each of the three selected values of adapting-flash energy and for each of the two flash durations. The data can be fitted quite well with straight lines. It is clear, however, that the lines for different flash durations do not all converge at a single point. The values of b for the longer flash duration data are found to fit the equation: bQ 33 = 0.01 A°-8. (4c) The constants in this equation appear reasonable in relation to those in Equation (4b) for a slightly lower visual acuity requirement. 123
100 80 60 40 Z Q S 20 Adopt. Flosh Pur. 1.0 1.5 2.0 Log L - log Lo FIG. 8. Time required for detection of a 0.33 acuity grating as function of l/(log L -log LQ). where L is grating luminance and LQ is minimum (threshold) grating luminance. Functions repre- sent two adapting-flash durations and three adapting-flash ener- gies. Functions derived from data presented in Fig. 4. It is clear that values of b for the shorter flash duration can- not be calculated from Equation (4c). This equation can be used to calculate values of adapting-flash energy at the longer flash duration, which would be equivalent to the shorter duration adapting-flash energy in their effects on recovery time, however. This has been done for each of the three short flash-duration curves in Fig. 8. In each case, the shorter flash duration cor- responds to a reduction in adapting-flash energy at the longer duration by a factor of 2.5. This corresponds to a shift of the short duration curve in Fig. 4 along the abscissa, a distance of 0.4 log unit from the corresponding curves for a longer flash duration. In other words, light-adapting flashes of less than 1-3 ms duration may be only 40 per cent as effective in reducing cone sensitivity as flashes of greater than 3-nis duration. It is evident that recovery from flash blindness is not a simple mechanism. In may involve mechanisms of dark adaptation as this is measured conventionally, it may reflect neurophysiologi- cal effects, the influence of which is not usually seen in studies of dark adaptation, and it may reflect recovery from retinal 124
injury following exposure to very high adapting-flash energies. For a wide range of conditions it can be represented by a fairly simple empirical equation, however. REFERENCES Aulhorn, E. Das Verhalten cler Netzhautempfindlichkeit im Nachbildbe- reich. XVIII Councilium Ophthalinol. (Belgium) 1958. 1605-1609. Barlow. H. B.. & Sparrock, J. M. B.. The role of afterimages in dark adaptation. Science, 1964, 144, 1309-1314. Brindley, G. S. The Bunsen-Roscoe law for the human eye at very short durations. J. Physiol., 1952, 118, 135-139. Brindley, G. S. The discrimination of afterimages. J. Physiol., 1959, 147, 194-203. Brown, J. L. The effect of different preadapting luminances on the reso- lution of visual detail during dark adaptation. J. opt. Soc. Amer., 1954. 44, 48-55. Brown, J. L. The use of the colored filter goggles for protection against flash blindness. USN Air Develpm. Cntr. Rep. ,1959, No. MA 5917. Brown, J. L. Time required for detection of acuity targets following ex- posure to short adapting flashes. J. engrng Psych., in press. Brown, J. L.. Graham. C. II.. Leibowitz, H. W.. & Ranken, H. B. Lumi- nance thresholds for the resolution of visual detail during dark adapta- tion. J. opt. Soc. Amer.. 1953, 43, 197-202. Brown, J. L., Pharos, L.. & Fletcher, D. E. Spectral energy thresholds for the resolution of acuity targets. J. opt. Soc. Amer., I960, 50, 950-9(50. Brown, J. L., & Woodward, L. K. Rod-cone interaction in the dark adapted eye. Optica Acta. 1957. 4. 108-114. Campbell, F. W.. & Rushton, W. A. H. Measurement of the scotopic pig- ment in the living human eye. J. Physiol.. 1955. 130, 131-147. DeMott, D. W., & Davis. T. P. Irradiance thresholds for chorioretinal lesions. AMA Arch. Ophthalmol., 1959. 62, 653-656. Dowling, J. E.. & Hubbard. R. Effects of brilliant flashes on light and dark adaptation. Nature. 1963, 199, 972-975. Fry, G. A., & Alpern, M. Effect of flashes of light on night visual acuity. Wright-Patterson AFB: WADC tech. Rep., November, 1951, No. 52-10, Part I. Hagins, W. A. Flash photolysis of rhodopsin in the retina. Nature. 1956. 177. 989-990. Hill, J. H., & Chisum, G. T. Flash blindness protection. Aerospace Med.. 1962, 33, 958-964. Johannsen, D. E. The duration and intensity of the exposure light as factors in determining the course of subsequent dark adaptation. J. gen. Psycho!.. 1934, 10. 4-41. Mete all'. R. D.. & Horn. R. E. Visual recovery time from high intensity flashes of light. Wright-Patterson AFB: WADC tech. Rep.. 1958. \o. fiS-232. 125
Miller, N. Visual recovery from brief exposures to very high luminance levels. Paper presented at Meeting, Opt. Soc. Amer., spring 1964. Mote, F. A., & Riopelle, A. J. The effect of varying the intensity and the duration of preexposure upon subsequent dark adaptation in the human eye. J. comp. & Physiol. Psychol., 1953, 45, 49-55. Rushton, W. A. H. Dark adaptation and the regeneration of rhodopsin. J. Physiol., 1961, 156, 166-178. Schouten, J. F., & Ornstein, L. S. Measurements on direct and indirect adaptation by means of binocular method. J. opt. Soc. Amer., 1939, 29, 168-182. Severin, S. L., Newton, N. L., & Culver, J. F. A new approach to the study of flash blindness. Arch. Ophthalmol.. 1962, 67, 578-582. Severin, S. L., Alder, A. V., Newton, N. L., & Culver, J. F. Photostress and flash blindness in aerospace operations. Brooks AFB: USAF Sch. aerospace Med. tech. doc. Rep. 1963, No. SAM-TDR-63-67. Shlaer, S. The relation between visual acuity and illumination. J. gen. Physiol., 1937, 21, 165-188. Teichner, W. H. Recent studies of simple reaction time. Psychol. Bull., 1954, 51, 128. Wald, G., & Clark, A. B. Sensory adaptation and chemistry of the rods. J. gen. Physiol., 1937, 21, 93-105. Whiteside, T. C. D. The observation and luminance of a nuclear explo- sion. Air Ministry Flying Personnel Res. Comm. Rep. 1960, No. 1075.1. Whiteside, T. C. D., & Bazarnik, K. The dazzle effect of an atomic ex- plosion at night. Air Ministry Flying Personnel Res. Comm. Rep., 1952, No. 787. Winsor, C. P., & Clark, A. B. Dark adaptation after varying degrees of light adaptation. Proc. nat. Acad. Sci., 1936, 22, 400-404. 126
METHODS OF PREVENTING FLASH BLINDNESS Frederick E. Barstow Edgerton, Germeshausen & Grier, Inc. Bedford, Massachusetts Other papers presented in this Section of the Report have covered the historical and operational aspects of flash blindness, some- thing of the nature of the atomic radiation which causes it, and some of the current laboratory studies on flash blindness. This paper is concerned with methods of eye protection. The first part considers the requirements which must be met by any pro- tective device and some of the many methods that have been in- vestigated. The remainder of the paper deals with a specific methodânamely, the Photochromic Protective System. * The specifications shown in Table 1 are nominal for the average eye protective device and represent a compromise be- tween the requirements for various applications rather than an exact specification for any particular device. Some applications consider open transmission of major importance and demand a TABLE 1. Nominal SpecificationsâFlash Blindness Protective Devices Transmission Density (activated) 3.0 Closure time less than 100 jisec Reopening time 2-10 sec Recycle rate 2-10 sec Life cycles 100 Optical quality Excellent 1. Edgerton, Germeshausen & Grier, Inc. (EG&G) has developed the goggles and periscope shutters described under various government contracts. 127
value higher than 50 per cent. In others, density is considered of greater significance and open transmission can be compro- mised. A density of 3.0 will probably provide burn protection under all conditions, but it is not clear yet whether this density is always adequate to prevent flash blindness. The specified closure time of 100 jisec shown in Table 1 is generally accepted as the desired maximum value, although there are many cases where this is not required. Reopening time depends on the ap- plication, but in the case of a pilot of a modern aircraft the device must clear in a few seconds. The recycle rate is even more arbitrary, but the range of 2 to 10 sec certainly covers most applications. The number of life cycles has been set at 100 since this number of operations would be more than ade- quate to cover any single mission. Finally, a high optical quality is essential if the user's vision is not to be impaired. The search for effective protection methods has been very extensive. Table 2 is only a partial listing of the broad range of protection schemes. With the exception of only one or two, all these have been subjects of government-sponsored research. Perhaps the blink response should not be considered as a passive method, but it is listed under that heading because no auxiliary equipment is required. The use of an eye patch over one eye needs no explanation. It should be noted that a fixed filter having a density of 2.0 gives only marginal protection; in addition, it is of little value at night. Pure mechanical systems are apt to be slow, while those depending on polarization are limited to transmissions of 15 to 30 per cent in the open state. On the other hand, chemical-molecular systems are likely to be fast-closing. TABLE 2. Flash Blindness Protection Methods Passive methods Active methods (directly and indirectly actuated) Mechanical Polarization Chemical-molecular Blink response ELF (explosive) Stressed plate Photochromic or Eye patch Electromechani - Kerr Cell phototropic Fixed Filter cal goggles Pockle Cell chemicals Curtain Photochromic glass Destruction Triple-State enzymes of mirror F -centers surfaces Fast plating Exploding wire Vado materials 128
A brief explanation of some of the methods may be of help. In the ELF device, an explosive charge drives a carbon-black mixture between two optically transparent plates. Although very high densities are reached in 100 to 200 /isec, the lens element must be replaced after each activitation. In the electromechanical goggle, a set of opaque lines or bands on one transparent plate is placed directly behind a similar set of bands on a second transparent plate. A small explosive charge then moves one plate so that the opaque lines on it cover up the spaces between bands on the other plate. The device develops a high density but is limited to less than 50 per cent open trans- mission. Activation time is several hundred microseconds at best. Curtain systems have been investigated as a possibility for goggles and as a method of protecting the entire cockpit. Closing a curtain over the entire canopy is slow, but this method may be used in conjunction with a more rapidly activated goggle device. Methods of destroying mirror surfaces have also been developed, but an exploding wire goggle has never been developed, even though such a device has been used effectively as a camera capping shutter. In the experimental stressed-plate shutter, a thick glass plate placed between two polarizing sheets is momentarily stressed or stretched by means of a stack of piezoelectric crystals so that the system temporarily becomes opaque. Open transmission and closed density depend on the choice of polarizers. The Kerr Cell is more familiar. It is a cell in which nitro- benzene is placed between metal plates and the combination placed between light polarizers. When a high voltage (generally in the range of kilovolts) is impressed across the plates, the bi- refringence of the nitrobenzene in conjunction with the polarizers causes a drastic change in transmission. Though limited to a maximum open transmission of about 30 per cent, closure time can be less than 1 ^sec. The Pockle Cell is approximately the solid-state equivalent of the Kerr Cell. The photochromic system is now described in more detail. First, it should be noted that it is an "indirectly actuated" sys- tem, requiring ultraviolet energy from flash-tubes. This is in contrast to a "directly actuated" system which would use energy from the nuclear explosion to actuate it. The active element in the photochromic system is an organic chemical, normally trans- parent in thin sections, but which becomes colored when exposed to ultraviolet light. The behavior of a typical dye is shown in Fig. 1. The straight line at 90 per cent to 95 per cent shows the 129
100 400 700 WAVELENGTH (MILLIMICRONS) FIG. 1. Behavior of a typical photochromic dye. transmission of the unactivated material. When exposed to in- creasing amounts of ultraviolet energy, an ever deepening ab- sorption band develops in the visible spectrum. Although high densities are possible in this band, light leaks through at the blue and red ends. Consequently, fixed sideband filters are re- quired to block out each end of the spectrum. In this way, high visual or luminous density can be obtained. Fig. 2 is a schematic diagram which shows how the goggle works. The photochromic solution is confined in a layer only 0.010-in. thick between two optically ground and polished quartz wedges. Xenon flash-tubes positioned above and below the quartz wedge system provide the ultraviolet activating energy. Special HROMIC LiQUID O" FILTER AIRCRAFT PQWE FIG. 2. Schematic diagram, photochromic goggle system. 130
filters protect the eye from visible light emitted by the flash- tubes but still transmit the ultraviolet energy to the photochromic material. The wedges distribute the ultraviolet light to the photo- chromic solution in a uniform manner so that the entire goggle area is colored to the same density. Energy for the xenon flash-tubes is furnished by a high-effi- ciency power supply which is triggered by a light-sensitive de- tector circuit, which, in turn, reacts to the very earliest nuclear light. The size of the power supply is a function of the sensitivity of the photochromic material, the goggle area, and the closed density. Fig. 3 shows a photograph of the most recent photochromic protective goggle system, including the power supply and detec- tor-trigger elements. A relatively wide angle of view, good open transmission, fast closure, high closed density, and automatic clearing are achieved in this device. FIG. 3. EG&G photochromic goggle system. Fig. 4 shows the first photochromic shutter designed for a periscope. The two circular apertures provide protection for both optical paths in the binocular instrument. In this case, the portable power supply is battery-operated. Table 3 shows how the goggle and periscope (Fig. 3 & 4) compare with the nominal specifications listed earlier. The values for the goggle differ from those for the periscope since the requirements and limitations imposed on each are different. For example, a wide angle of view is imperative for the goggle, while the same is not required for the periscope shutter. The present goggle with 35 per cent open transmission falls short of the specified 50 per cent but continued research is ex- pected to yield the specified value. On the other hand, the peri- scope device already exceeds 50 per cent, and a new shutter is expected to reach nearly 70 per cent. The closed-density speci- 131
FIG. 4. First photochromic shutter designed for a periscope. TABLE 3. EG&G Photochromic Devices (1964) Specifications Goggle Periscope Open transmission 50% 35% 55% Closed density 3.0 3.5 3.0 Closure time (^ sec) 100 75 50 Reopening rate (sec) 2-10 2 5-10 Recycle rate (sec) 2-10 2 10 Life cycles 100 30-60 30-60 Optical quality All excellent fication is equalled by the periscope device and exceeded by the goggles, while the closure time exceeds the specification in both cases. Reopening and recycling rates fall within the specified range; some adjustment is also possible through choice of photo- chromic chemicals and through power-supply design. The number of life cycles, however, or the number of times the device can be closed before resulting in some small degrading of the other characteristics, is less than desired. Although optical quality is relative, the system permits any degree of optical perfection 132
desired. Thus, it can be seen that the photochromic devices already meet most of the specified characteristics. It is interesting to examine the actual closure of the photo - chromic shutter. The oscillographic traces in Fig. 5 were generated by passing the light from an electronic flash-tube through the periscope shutter onto a photocell. The top trace shows the flash-tube light falling on the photocell when the shutter was not activated. The lower trace is for the same identical conditions except that the very early light from the flash-tube was picked up by the detector, which, in turn, trig- gered the periscope shutter. Note that for the first 8 or 10 ^sec the traces are nearly the same, but by 20 psec the activated shutter blocks out more than 90 per cent of the light, and at 30 to 40 jisec essentially all of the light is blocked out. The stated closure of 50 fisec is, therefore, conservative. -*j (â20 ft sec TIME FLASHTUBE VIEWED THROUGH PERISCOPE SHUTTER FIG. 5. Oscillographic traces: flash-tube viewed through periscope shutter. At this point, it is interesting to give an example relating to closure time. Assume that a pilot flying at sonic velocity is wear- ing a pair of photochromic goggles and encounters an atomic detonation. His goggles would then close in less time than it would take his aircraft to move 1 inch. This paper has stressed the positive aspects of the photo- chromic system. There are, however, still some problems to be solved. As is shown in Table 3, open transmission and life are less than desired. In addition, the photochromic materials are temperature sensitive in that the clearing or reopening time varies with temperature. Another problem is weight, particularly that of the energy storage power supply. However, continued research will undoubtedly lead to improvements in these and other areas. 133
AIR FORCE EFFORTS IN THE FIELD OF FLASH BLINDNESS James F. Culver Ophthalmology Department USAF School of Aerospace Medicine, Brooks Air Force Base It is apparent that interruption of a pilot's vision at a critical time could prove disastrous. The problem, therefore, consists of repeatedly providing adequate vision for aircrew members up to the time of a nuclear flash, shielding their eyes from the flash, and then again providing adequate vision almost immediately. This is not a simple undertaking. The severity of the problem varies considerably, depending on the ambient illumination, i.e., during daylight the problem is minimal unless the observer happens to be looking directly at the detonation; however, night- time operation can offer an extreme hazard during certain flight operations. Flight testing has shown that pilots would prefer to have the protective device placed on the windshield or canopy rather than on the person to obviate interruption of instrument visibility during the closed state. The concept of a goggle-type protective device should not be completely discounted since there will be other operations associated with flying in which they could be most valuable. The problem of flash blindness and that of chorioretinal burns cannot be entirely divorced. Both can occur from a thermonuclear detonation. One happens to be of a temporary nature and the other is permanent. It is felt that a short summary and reference list of Air Force activities in these areas would be of interest to the members of the Armed Forces-NRC Committee on Vision. Ocular hazards were recognized from the beginning of the nuclear era, and protective goggles were provided for observers at the first nuclear device detonation at White Sands, New Mexico. Empirical data from subsequent weapons tests later indicated the extent of the problem of flash blindness and chorioretinal burns 134
(Byrnes, 1953). Researchers at the U.S. Air Force (USAF) School of Aerospace Medicine continued to pursue field investigations of the ocular effects of nuclear flashes until the moratorium on on testing was declared in 1958. The occurrence of chorioretinal burns in rabbits was demonstrated at distances of over 300 nauti- cal miles from a thermonuclear detonation (Glasstone, 1962). During this period, prototype protective devices were also tested by the USAF Aerospace Medical Research Laboratory (Gulley, Metcalf, Wilson & Hirsch, 1960). During the more recent 1962 series of weapons tests, per- sonnel from the USAF School of Aerospace Medicine tested retinal-burn prediction models. This was the most comprehen- sive experiment of its type to date, and provided extensive, well- documented quantitative data which will also be of great value in flash blindness research. Parallel to the experiments carried out in the field of weap- ons tests, a laboratory program to establish the chorioretinal- burn threshold was continuing. Early projects were sponsored by the United States Air Force at the Medical College of Virginia and continued by the Defense Atomic Support Agency. The excel- lent work performed by Dr. Ham, Dr. Geeraets, and their group at the Medical College of Virginia is well known. Their early studies indicate that the macroscopic-burn threshold varied from 1 to 13 calories per square centimeter (cal/cm^), depending on the retinal image size, utilizing a constant input rate (Ham, Wiesinger, Schmidt, Williams, Ruffin, Shaffer, & Guerry, 1958). Subsequent work by this group indicated this threshold could be less than 1 cal/cm^ when the energy is delivered at a very high rate. Research on the effect of high-intensity flashes on the eye at the New York Eye and Ear Infirmary has also been supported by the United States Air Force. Their findings confirmed those of the investigators at the Medical College of Virginia in the areas where the two studies overlapped (Jacobsen, Cooper, & Najac, 1962; Jacobsen, Najac, & Cooper, 1963). Studies were also con- ducted at The Ohio State University where an attempt was made to determine the effectiveness of various spectral bands in the production of chorioretinal burns (Bredemeyer, Wiegmann, Bredemeyer, & Blackwell, 1963). It is apparent that there has been much effort expended in the past on the problem of chorioretinal burns; however, data obtained in these field tests and laboratory studies are of definite impor- tance in establishing eye-protection criteria. Recently, the em- 135
phasis has again shifted primarily to the problem of flash blind- ness. Early field studies conducted in 1952 that were devoted to flash blindness have been supplemented by the initial experiments on the operational aspects of flash blindness conducted by the Air Force Aerospace Medical Research Laboratory (Metcalf & Ham, 1958). Personnel at the USAF School of Aerospace Medicine are presently conducting an in-house flash blindness program to define the operational aspects of this problem and outline a train- ing program. Currently, a research project directed toward a better understanding of the basic mechanism of flash blindness is being sponsored at The Ohio State University (Miller & Fry, 1963). Air Force efforts to provide urgently needed protection are outlined below. In 1955, the Air Force Systems Command initiated a project to conduct the development of eye protection from nuclear flashes. As a first step, feasibility studies of the electro-optical shutter (Kerr Cell) and the electromechanical goggle (Burger & Filer, 1964; Wayne-George Corp., 1959) were conducted. The Kerr Cell approach was found not be feasible for Air Force needs, and de- velopment of the electromechanical goggle was pursued. During this same period, investigation of phototropic, or self-attenuating, filters was also being made. These were parallel developments at the USAF School of Aerospace Medicine and the Aerospace Medical Research Laboratory. The phototropic approach has resulted in at least three eye-protection systems, including the indirectly activated phototropic goggles (Barstow & Lilliott, 1961), the one-way directly activated phototropic goggle (Krekeler, 1963), and the Dynacell directly activated device (Harries, 1963). Only the one-way directly activated goggle has been field tested and deemed to be unacceptable (Parkhurst, 1963). Recent studies have proved the electromechanical goggle to be unacceptable for pilot use. A further study of an electrochemical-type shutter found it also not feasible. This was an investigation of the electro- plating principle in a shutter system (Aitken, 1962). During this same period, the development of a fixed-filter eye- protection system has been under way. This has taken the form of protective visors and goggles. Studies indicate that a 1 per cent transmittance filter (gold-plated goggle) can provide ade- quate vision for flying during daylight. To coordinate research better and to prevent duplication of effort in this field, the Oculo-Thermal Section was established within the Ophthalmology Department at the USAF School of 136
Aerospace Medicine in 1963. It is presently engaged in develop- ing design criteria for eye protection from nuclear flashes, and prototype eye-protective devices. As previously stated, the nuclear-flash eye hazard represents a continuum ranging from temporary flash blindness to permanent retinal-burn effects. Flash blindness is receiving the primary effort. In the study of chorioretinal burns, two criteria for dam- age have been used. One of these is a visually detectable thresh- old burn. This macroscopic lesion, although a gross approach to the problem, may very likely represent minimal significant retinal damage. More recently, with the development of a mathe- matical model for retinal burns, temperature rise in the retina has been utilized as the criterion of retinal damage. Some de- finitive work has been done by Ham, at the Medical College of Virginia, and by Jacobson et al. at the New York Eye and Ear Infirmary, on the temperature rise in the retina. However, this factor remains for the most part rather elusive and in need of considerable investigation. At the present time, an attempt is being made to define the critical temperature rise by comparing laboratory burns with burns received at nuclear weapons tests. It is hoped that this approach will throw some light on the nature of the critical temperature rise. The problem, insofar as the retinal-burn hazard is concerned, consists of developing reliable indicators for threshold retinal burns, then relating this signifi- cant lesion to nuclear weapon output for the various situations. In order to assess properly the flash blindness hazard, it is necessary to define the stimulus. As a result, an in-house proj- ect has been established to develop luminance values of the vari- ous detonation conditions. Weapons effects data from actual weapon tests are being utilized for this effort. A contract is presently being negotiated to develop a mathematical expression for flash blindness. This model will made use of the best avail- able data on recovery times and weapon illuminances, and if suc- cessful, will permit computer programming for development of flash blindness safe separation distance charts. Concurrent with the development of flash blindness separation distance charts will be the development of design criteria for eye-protective devices. In order to fulfill the second requirement of the section's mis- sion, a number of nuclear flash eye-protective devices are being developed. Even though design criteria for eye-protective devices offering a high degree of confidence are not available, estimates can be made of the required protection. This is especially true 137
in the case of retinal burns. An improved filter visor transmitting 2 per cent in the range between 200 and 2,000 millimicrons has been developed. This visor should provide adequate retinal-burn and flash blindness protection during daylight hours. The 2 per cent transmittance allows normal visual response inside and out- side the cockpit in daylight. This item is to be brought into the inventory in early 1965. Other items under continuing development are: 1. Phototropic, or so-called self-attenuating, filters which darken on exposure to ultraviolet and shortwave-length visible radiation; 2. indirectly activated phototropic filters activated by flash- tubes; 3. dynacell phototropic filters which consist of highly sensi- tive phototropic fluids flowing through filter cells; and 4. electronic triggering systems for the indirectly activated filters. It is hoped that this short summary and reference list will be of value. Unfortunately, due to regulations, classified reports cannot be referenced; however, they are available to eligible investigators. REFERENCES Aitken, J. F. Electrochemical light modulator. Wright-Patterson AFB: MRL tech. doc. Rep., 1962, No. 62-29. (Final Report, Philco Research Corp. Contract AF 33(616)-7828). Barstow, F. E.,& Lilliott, C. Development of flash blindness protective goggles. Bedford Mass.: EG&G Inc. Rep. 1961, No. B-2288. Bredemeyer, H. G., Wiegmann, O. A., Bredemeyer, A., Blackwell, H. R. Radiation thresholds for chorioretinal burns. Wright-Patterson AFB: AMRL tech. doc. Rep., 1963, No. 63-71. (DDC AD 416652). Burger, W. R., & Filer, H. C. Electromechanical goggle. Wright-Patter- son AFB: ASD tech. Rep. 1963, No. 63-451. (Final Report, The National Cash Register Co. Contract AF 33(600)-43570). Byrnes, V. A. Flash blindness. Report of Project 4.5, Operation SNAPPER. Brooks AFB (Tex.): Weapons Test Rep., 1953, No. WT-530. Glasstone, S. (Ed.). The effects of nuclear weapons. (Rev. ed.). Wash- ington: US Atomic Energy Comm., 1962. Gulley, W. E., Metcalf, R. D., Wilson, M. R., & Hirsch, J. A. Evaluation of eye protection afforded by electromechanical shutter. Brooks AFB: Weapons Test Rep., 1960, No. WT-1429. (Report of Project 4.2., Operation PLUMBOB). Ham, W. T., Jr., Wiesinger, H., Schmidt, F. H., William, R. C., Ruffin, R. S., Shaffer, M. C., & Guerry, D., III. Flash burns in the rabbit retina. Amer. J. Ophthal., 1958, 46, 700-723. 138
Harries, R. W. The dynacell and focal plane concepts of phototropic ophthalmic nuclear flash protective device. Wright-Patterson AFB: ASD tech. Rep. 1963, No. 63-658. (DDC AD 424140L). Jacobson, J. H., Cooper, Blossom, & Najac, H. W. Effects of thermal energy on retinal function, Part I. Wright-Patterson AFB: AMRL tech. doc. Rep., 1962, No. 62-96. (DDC AD 290808). Jacobson, J. H., Najac, H. W., & Cooper, Blossom. Effects of thermal energy on retinal function, Part II. Wright-Patterson AFB: Final Rep., 1963. (Contract AF 33(616)-7685)(DDC-AD 434726). Krekeler, J. H. Development of an irreversible photo-thermosensitive ophthalmic nuclear flash protective device. Wright-Patterson AFB: ASD tech. Rep., 1963, No. 63-658. (DDC AD 424140L). Metcalf, R. D., & Horn, R. E. Visual recovery times from high intensity flashes of light. Wright-Patterson AFB: WADC tech. Rep., 1958, No. 58-232. (DDC-AD 205543). Miller, N. D., & Fry, G. A. Visual recovery from brief exposure to very high luminance levels. Ohio State University: Interim Report, 1963. (Contract AF 33(657)-9229), (DDC-AD 434729). Parkhurst, D. J. Operational test and evaluation of phototropic goggles. ENTAFB (Col.): ADC Rep., 1963, ADC/73AD/63-26, (DDC AD 428073). Wayne-George Corp. High-speed electromechanical goggle. Wright- Patterson AFB: WADC tech. Rep., 1959, No. 59-114. (DDC-AD-215828). 139
A FLASH BLINDNESS INDOCTRINATION AND TRAINING DEVICE James F. Parker, Jr. BioTechnology, Inc. Relative to the flash blindness problem area, a flash blindness indoctrination and training device has been developed for the Office of Naval Research and the Bureau of Naval Weapons by BioTechnology, Inc. The purpose of this device is to demonstrate to pilots who might operate in nuclear combat zones what would happen to their vision, and consequently, to their mission capa- bility if they unexpectedly encounter the light from a nuclear burst. It is anticipated that this device will serve two training functions. 1. Indoctrination training. The device can be used to demon- strate dramatically the effect of "startle" and temporary loss of vision on the performance of typical flight activities. 2. Proficiency training. The device can be used to demonstrate the protection afforded by various protective systems and pro- cedures, and will allow practice in the use of these systems and procedures. The basic elements of the flash blindness device are: 1. Flasher unit. The light source is a xenon gas discharge electronic flash-tube of the quartz helix type. This unit delivers approximately 400,000 lumen-seconds (sec) of visible energy in 2 milliseconds, which simulates the initial pulse of the weapon. Simultaneously with the initial flash, three 300-watt bulbs are illuminated and gradually extinguished over a 4-sec period. This simulates the dying out of the weapon fireball. 2. Focusing hemisphere. Light from the flash source is bounced off a silvered hemisphere of 4-ft radius and is directed toward the position of the pilot, who is seated at the front of the device. 140
3. Diffuser screen. The screen diffuses the light with approxi- mately a 30 per cent loss in intensity so that it is more repre- sentative of that which might be experienced in a nuclear environ- ment. This screen allows also for the projection of a color film which presents a view of typical terrain as seen by a pilot flying a low level, high speed mission. 4. Pilot's instrument panel. This panel provides tasks for the pilot which are representative of those performed by aviators. It can be used to demonstrate the performance decrement which occurs following exposure. 5. Operator's panel. The instructor uses this panel in con- trolling the over-all operation of the device and in monitoring pilot performance. Fig. 1 shows the major components of the device, the photo- graph being taken at the moment of flash. The operator control panel, pilot's panel, and diffuser screen are shown in this figure. FIG. 1. Major components of flash-blindness device. 141
Fig. 2 shows the reaction of a subject as he receives the initial pulse of light. One of the most important uses of this device is to illustrate the dramatic decrease in visual recovery time which occurs when the illumination of the visual task is increased. A number of subjects have been tested in this device, under somewhat un- controlled conditions, and their recovery times noted. Table 1 presents the range of recovery times for the two conditions of illumination of the visual task. The visual task in this instance consisted simply of reading three digits, which were white against a black background, as soon as possible following the flash. FIG. 2. Subject photographed at moment of flash. TABLE 1. Range of Visual Recovery Times for Two Conditions of Illumination of the Visual Task Illumination Range of recovery times 0.5 foot-candles 20-90 sec 30 foot-candles 4-7 sec 142
VISION PROBLEMS IN LOW-ALTITUDE, HIGH-SPEED FLIGHT James W. Miller, Chairman
