funnel. At the approach end, the lower edge of the beam is 10 degrees above horizontal and the top edge, 45 degrees. The vertical angle decreases 1 degree at each successive light, moving inward along the rows, until the lower edge is horizontal and the upper edge at 35 degrees. The remaining lights are at this same setting.

Use of the wide angle AGA approach lights in the funnel pattern ensures that a pilot flying through a daylight fog which restricts objects visibility to 1/8 of a mile will see at least three of the lights when his plane reaches a point 500 feet from the nearest lights, provided his aircraft is within the funnel and not more than 500 feet above the runway. This approach accuracy is well within the capacity of trained instrument pilots. Fig. 10 and 11 illustrate the penetration of the AGA approach lights when installed in the funnel system. Fig. 10 shows a plane view of the approach end of the funnel, and Fig. 11 is a vertical section along the light row in the same region.

D.RUNWAY MARKER LIGHTING FOR LOW VISIBILITY

Using electronic navigation aids, a plane is guided into the approach light tunnel. The approach lights direct the pilot visually toward the runway threshold. Beyond this point, additional markers along the runway are required to establish the ground plane upon which the pilot must land his aircraft, and to outline the safe area.

The problems of runway marking are essentially the same as those of approach lighting, namely, high candlepower for fog penetration, and sufficient angular distribution to assure visibility throughout the expected vertical and horizontal deviation. Since the plane can be expected to arrive over the threshold at some point within a rectangle 200 feet wide and 150 feet high, the design problem for runway markers is less severe, and the units can be made smaller.

The light currently in use at LAES is the Bartow Type D-1 Runway Marker, Fig. 12. The optical system of this unit consists of a 200-watt prefocussed lamp mounted within a hemispherical metal casing. Over this is mounted a dome-shaped Fresnel-type lens which concentrates the light into a vertical beam approximately 4 degrees high. A large outer dome of pressed glass is fastened over the unit to serve as a cover and also as the horizontal condenser. Vertical flutes on its inner surface redirect the light into a beam approximately 1 1/2 degrees wide. The unit is bi-directional, to mark the runway for landings in either direction, and projects a beam along the runway from either side. The beams are elliptical in cross-section and approximately 4 degrees high by 1 1/2 degrees wide at the 30,000 candlepower isocandle curve, dropping off sharply in candlepower beyond this boundary.

The units are located along each side of the runway at 200-foot intervals. Green filters are installed in the markers at each end of the runway, and yellow filters in those bordering the 1600-foot terminal section of the runway, the remaining units being clear. This color sequence distinguishes the threshold, safe-landing area, and terminal sections of the runway in accordance with the Army-Navy- Standard Color Sequence.

The Bartow Runway markers are aimed so that the axis of the main beam is directed across the runway at an angle of 3 degrees and upward at an angle of 5 degrees. At this setting, the main beams or lights opposite each other intersect above the runway at a distance of approximately 1000 feet.

The narrow beam of the Bartow lights is not fully satisfactory for landing operations during fogs which restrict visibility to 1/8 mile and below . It was designed for use in visiblities of 1/4 mile and above, when the lower candlepower outputs surrounding the main beam are visible at relatively large distances. In moderate to heavy fogs, a pilot will be sure to see enough of the lights to the runway only if his aircraft is within 10 to 15 feet of the runway only centerline and not more than 85 feet above it. In practice, this margin for error is frequently exceeded, so that only a few lights are seen, and perhaps only in one row. Unless several lights are seen on each side, a safe landing cannot be made at many airports, the zero-zero condition rarely or never develops. However, a number of the country's most important airline terminals are frequently wrapped for hours in fogs lying close to the ground, through which nothing can be seen beyond a few hundred feet. At these localities, it is urgent and economically feasible to install FIDO, as the thermal method for dispelling fog has come to be known.

Fog is an aerosol, a suspension in the air of a large number of minute droplets of liquid water. These droplets deflect any light ray striking them from its original path. At a certain distance, in any fog, so much of light proceeding from any reflecting surface is deflected (or attenuated) by the intervening droplets that an observer can no longer distinguish it. For any particular fog, this certain distance is known as its visual range.

To increase visibility, it is necessary to remove all, or a major portion, of the water droplets between object and the observer (or pilot) who is required to see them. Several laboratory means for doing so are known, but the only proven method now available for full scale use is the thermal process. In this process, a suitable fuel, such as gasoline or diesel, is burned around the runway; the heat of combustion raises the air temperature and causes the droplets to evaporate.

The design of a FIDO installation begins with a study of the weather history at a particular airport. In order to select the most advantageous burner locations, it is necessary to know the angle at which winds accompanying fog will cross the instrument runway, the range of wind velocities to be expected at the various angles, and the relative frequencies and durations of the several speed-direction combinations. These relationships dictate the locations and heat outputs of the burners. If fog-laden winds always paralleled the instrument runway, the design of a FIDO installation might be handled by the manufacturers of the Zippo cigarette lighter without overstraining their capacity. However, nature is not fully cooperative in this respect, and it is usually necessary to plan FIDO primarily for the worst weather condition: that is, for the highest expected velocity perpendicular to the runway. If approproximately parallel wind conditions are expected to exist during a significant proportion of the total elapsed adverse weather time, then it may be economical to add a secondary burner system.

In choosing the burner locations and heat outputs, the first factor to be considered is the height of the ceiling required to be produced. For the Integrated Landing System, we feel, at LAES, that the cleared area should be at least 275 feet high at the approach end, 1500 feet out in the approach zone from the runway threshold, and may be permitted to drop off gradually to 50 feet or so at the opposite end of the runway. For commercial airline use, however, somewhat higher ceilings might be provided, in order that a safe margin will be maintained over the average airline pilot's stimulas-response time.

For any desired ceiling, in crosswind positions, there exists an optimum burner location and heat output. Heat is added to the air crossing the burners at the ratio required:

1) To evaporate the liquid water droplets before they reach the runway.

2) To increase the buoyancy of the cleared air relative to the surrounding air masses, by heating, is that it will rise to the desired ceiling height by the time the prevailing wind carries it to the runway.

For any given ceiling, increased crosswind speeds require that higher thermal outputs be provided, and that burners be located farther upwind from the runway. If wide variations in cross wind speeds are expected, or if parallel wind conditions alternate with extreme crosswind speeds, it may be necessary to install additional burners near the runway. Many variations are possible, and each installation must be tailored for its specific location. Local obstructions and terrain conditions may be of considerable importance in the final design.

Having determined the most advantageous locations and the necessary heat outputs, a burner system is laid out to produce the desired heat distribution. Earlier FIDO installations used the low-pressure burners mentioned earlier, which burned vaporized gasoline. These burners served very well during the war, but their expensive construction, wasteful preheating period, and high maintenance requirements and costly gasoline fuel make them unsuitable for commercial use. A new system has been developed by the Babcock and Wilcox Company, for the Navy department, which overcomes all these defects. It employs a different combustion principle, in which the fuel is prepared for combustion by atomization, instead of by vaporization. This method permits heavier petroleum fractions, such as auto diesel fuel, to be burned without smoke as readily as is gasoline.

The essential part of an atomizing burner system are:

(1) A tank for fuel storage.

(2) A gravity head or low pressure pumping system

capable of delivering fuel in large quantities at approximately 90 pounds per square inch.

(3) A multi-stage centrifugal pump capable of delivering the fuel required at pressures as high as 1500 pounds per square inch, (if diesel fuel is to be used).

(4) High pressure supply lines to the burner locations, laid below ground.

(5) Atomizing burner heads connected with riser pipes to the supply lines.

(6) Automatic igniter units.

(7) Remote controls for pumps, supply line valves, and igniters.

At Arcata, the tank farm has a capacity of 240,000 gallons--sufficient for several hours of continuous operation. Four Carter pumps operated by gasoline engines are used to deliver fuel to the high pressure pump intake. This pump is a marine boiler supply unit built by the Byron-Jackson Company of Los Angeles. It is of the centrifugal type with 9 compression stages and discharges 840 gallons per minute at pressures up to 1850 psi. It is operated by an 800 horsepower General Electric induction motor.

Fuel supply lines are of seamless weld pipe, extra strong, capable of withstanding 2000 psi. Their size varies from 8 inches in main lines, to 2 1/2 inches in individual burner sections, depending upon the carrying capacity required.

Along the burner lines 3/4-inch riser pipes are welded to the supply lines; these project 6 inches to a foot above ground level and are cut to suit the terrain contour. At the top of the riser pipes are the burner hoods. These consist of delivery chambers into which are screwed from one to three burner nozzles, depending on whether a single or multiple flame is desired. The burner nozzle consists of a whirling chamber on which a standard Mayflower marine boiler sprayer plate is held in place by a threaded burner tip. The device somewhat resembles a lawn sprinkler and produces a similar cone, except that the spray consists of atomized droplets of a very minute size. Fig. 13 shows a single nozzle burner head equipped with an igniter, and Fig. 14 the flame pattern produced by a triad burner head.

Spacing of nozzles along the burner line depends primarily upon the thermal output required. Heat requirements are generally expressed in terms of therms per yard per hour, the therm being equivalent to 100,000 Btu. Sprayer plates of various sizes, and the use of multiple nozzles, permit the thermal output per hour of a single head to be widely varied. The output and spacing are adjusted to secure the most advantageous compromise amongst such factors as economy in construction most desirable flame height for a given burner location, and the possibility of fog intrusion between burner heads. This latter factor is at present thought to limit the spacing to 20 feet or less, although tests now in progress may indicate the permissibility of wider spacing at localities having low wind velocities. Fig. 15 shows one satisfactory burner pattern, called the crossover, in which opposing pairs of nozzles are spaced on 4-foot centers along the burner line.

Igniting the atomized fuel is accomplished automatically with GE Calrod heating elements. These units are made into helical coils, of approximately 1- inch outside diameter, at the factory; coiling increases the temperature at the center of without a corresponding increase in wattage. The units operate at a central temperature of about 1800 Deg. F, and the remain in operation throughout the burning period. They are located along the burner line at intervals as required by the burner spacing.

Remote control of the system will be obtained by using solenoid valves in sectioning off the various burner lines. This will permit the actual operation to be adjusted as required by a given set of weather conditions.

III TESTS OF THE INTEGRATED LANDING SYSTEM

During the months of September and October, 1946 flight tests were made on the Integrated Landing System at Landing Aids Experimental Station. In 14 test sessions, held at all hours of the day or night as fog occurred, 110 approaches were made in conditions varying from as low as 200-foot visibility up to 1 mile. Of these 110 approaches, only 5 were considered unsuccessful, in that the test plane could not have landed.

Even these excellent results; however, do not reveal the full promise of the Integrated System. Approximately half of the approaches were made at a time when the Approach Light Installation was incomplete, and only the 1300 feet of it located on station property was operative. In addition, suitable monitoring equipment for the ILS radio beams was not available, so that the course could not be checked as accurately as would be desirable. Finally, the FIDO actually employed in most cases consisted of only 600 feet of burner line paralleling the approach end of the runway on either side, rather than the 7,000 feet of burner line, beginning in the approach zone, which is recommended for a full installation. Nevertheless, the pilots were able to use this facsimile of the integrated system with a regularity soon became monotonous. The pilots who flew these tests are doubtless more adequately trained for this type of landing than most airline pilots are at the present time, but there is no reason why such training should not be pushed in order to gain for the airlines the benefits of all-weather schedule reliability.

IV COST OF THE INTEGRATED SYSTEM

A cost estimate has been prepared for an Integrated Landing System contemplated at Mines Field, Los Angeles, California. This estimate is based upon such current material prices as were immediately available with a very liberal margin added for possible unforeseen difficulties. The total comes to roughly a half million dollars for the Electronic Aids, AGA Approach lights, Bartow Controllable Runway Markers, and a High Pressure FIDO installation. This is a large figure, but it represents an addition of only one or two per cent to the cost of a modern airport. San Francisco, for instance, recently voted a $20,000,000 bond issue, not for building, but for improving air facilities at Mills Field. It must also be considered in relation to the cost of airliners, which, for planes now in use, ranges up to $1,500,000 for the planes coming into production. Moreover, a busy airport, such as Mines Field, will land tens of thousands of planes a year. Amortization of the cost of the Integrated Landing System over a few years' operation will, therefore, run little higher than insurance against catastrophe.

Maintenance costs for the system will be largely a matter of salaries for the new personnel required. Experience at LAES indicates that a group of five employees primarily responsible for maintaining and operating

 to Page 4