by Orr Kelly
Admiral Thomas Hayward, a former test pilot who was later to become chief of naval operations, summed up the doubts harbored by many in the fleet as the navy prepared to move into full-scale development and then production of this new plane:
A number of people like myself, who had seen a number of claimed achievements by the technical people fail again and again in the 1950s and 1960s, were not very enamored of the thought we were going to have all kinds of new things none of us had seen before: Heads-up displays, computers that could be used for air-to-air as well as air-to-ground at a time our air-to-air radars were still having lots of trouble, and our air-to-air fire controls were still not all that good. Our air-to-ground computer system was certainly open to plenty of doubts. How, then, were we to believe the technical people could possibly be right in their claim that we were going to see one airplane do both roles very well?
CHAPTER FOUR
One Plane, One Man
When full-scale development of the F/A-18 began in 1976, almost everyone involved assumed that the fighter and attack versions of the plane would each require a separate set of black boxes to control the radar and other electronic operations. It would be possible to change the boxes and convert an A-18 into an F-18 and vice versa, but it would be a difficult, time-consuming process.
One of the few who were convinced it would be possible to do something far better was Kent Lee. He reasoned that most of the technology needed to produce a true strike-fighter was already flying and performing well. The A-7E, the latest version of the LTV attack plane, had a good computer, a good radar and a good inertial navigation system. Together, these made the plane an extremely accurate bomber.
All that was needed, Lee concluded, was a single radar that could be used for both bombing and air-to-air combat. He described it as a “programmable radar.” This meant that, instead of being wired to do just one job, the radar would be controlled by software—electronic instructions—that would permit it to be changed back and forth between the two missions with the flip of a switch.
To say that this was “all that was needed” was not literally true. A true one-man plane would also require breakthroughs in cockpit design and computerized controls. But it was true that, without a programmable radar, development of an effective strike-fighter would be difficult at best and probably impossible.
To many in the navy, the odds against success seemed so high that they discounted this as just another of Lee’s pipe dreams. No one had ever built such a radar. No one was quite sure whether such a radar could be designed. And there were many who doubted that it could be produced even if it could be designed.
It might seem that combining the two jobs in one radar set would be fairly simple, just a matter of which direction the radar beam is pointed. In practice, however, the solution is far from simple. The demands placed on the radar when it is used for air-to-air combat are almost the opposite of those involved when it is used to drop bombs. When the radar sweeps the sky in search of hostile planes, it covers a large area that is virtually empty. It must be able to find the few moving targets very rapidly. When the radar is used for ground attack, it must sort out vast amounts of confusing information reflected back from the ground, but speed is not nearly as important as it is in air-to-air combat.
McDonnell Douglas took the problem to the nation’s two most experienced radar manufacturers, Westinghouse and Hughes Aircraft. Westinghouse was, like McDonnell Douglas itself, a stolid, conservative organization, content to build carefully on technology that had already been proved in practice. Hughes, like its southern California neighbor Northrop, was more innovative, more inclined to take a chance on new technology that might promise a breakthrough.
Westinghouse adapted a radar it was already working on for the air force’s F-16 fighter and, as might have been expected, came in with a hard-wired radar system with different black boxes for the attack and fighter modes. The more daring Hughes engineers gave life to Lee’s pipe dream. They proposed a programmable radar controlled by changeable software. Hughes, under prodding by McDonnell Douglas, then went on to sign a contract that was almost as daring as the radar. The company agreed to a fixed price for full-scale development and options on the first three years of production—before a single circuit board or component of the radar had been built.
“It was a courageous contract,” says John L. Conklin, manager of the advanced programs staff in Hughes’s F/A-18 program division. The company’s bid was reviewed by its top officials and they knew it was risky. There were two reasons why they took the risk. One was the likelihood that, if their bid was not extremely competitive, they would lose the business. The other was their confidence in their large, competent team of software engineers, and a history, as Conklin puts it, “of records of firsts in almost everything of importance in airborne radar.”
The company had developed its first airborne radar for the air force a quarter of a century before. In 1949, equipped with that radar, the F-94A became the first plane to shoot down a target drone without the pilot seeing the target. Hughes went on to develop the radar for the high-flying YF-12 Blackbird spy plane, as well as radars for the navy’s F-14 Tomcat and the air force F-15 Eagle.
But nothing prepared the Hughes engineers for the challenges they faced when they attempted to put a dual-purpose radar in the F/A-18.
The very first problem was size. Under ideal circumstances, a plane is designed around its radar. In a rough way, it can be said that the performance of an airplane’s radar set depends on the size of its radar antenna: the bigger in diameter the antenna, the farther the radar can see and the smaller the target it can discern. This all translates into a big airplane. The ultimates in antenna sizes are seen on the air force AWACS and the navy E-2C. Each has a huge radome sprouting at the top of a pylon above the fuselage like some giant growth. With such large antennae, these planes serve as airborne battle-management stations, capable of keeping track of everything flying in an area of thousands of square miles.
If the goal is a fighter-interceptor like the F-14, whose radar can see tiny targets more than 100 miles away, keep track of twenty-four of them, and shoot at six targets at the same time, then a big antenna is needed—in the F-14, thirty-six inches in diameter. This means a large fuselage, a big wing and tail, and large tanks to carry enough fuel for long endurance.
But the shape of the new navy plane had already been determined. There was no way Hughes or anyone else was going to put a big radar antenna in the nose of that plane. The original YF-17 had been designed with enough room for an antenna twenty-three inches in diameter, just big enough for a simple little radar set suitable for rudimentary air combat operations. The radar in the F/A-18 would have to be far more capable than that. Not only would it have to be powerful enough for close-quarter dogfighting with guns and the Sidewinder missile, but it would also have to support the medium-range Sparrow missile and an even longer-range missile then on the drawing boards. And, of course, it would also have to do the quite different job of helping the pilot drop his bombs with precise accuracy.
The aircraft engineers managed to expand the diameter of the nose by eleven inches, to thirty-four inches, but that still meant the size of the antenna would be severely limited. Hughes focused on the other three parts of the radar set. In addition to the antenna, the radar consists of a transmitter, a receiver, and a signal processor.
The transmitter sends out bursts of energy that bounce off the target and return to their source. They are captured by the receiver and then run through the processor, which analyzes the signals and determines such things as the location and distance of the target and, if it is another plane, the speed and direction it is flying. If the size of the antenna is limited, as it was in the F/A-18, the performance of the radar can be carried to its ultimate by increasing the size and power of the transmitter so it can hurl its bits of energy farther away, by making the receiver more sensitive so it will be able to capture even faint echoes, and by making the processo
r more sophisticated so it will be able to analyze these faint signals. That is what the Hughes engineers set out to do. But they soon realized that the two obvious solutions to their problem—increasing the size of the radar and the electrical energy fed into it—were out of the question. The radar would have to be smaller, not larger, than anything they had ever built before, and it would have to use less, not more, power.
When they designed the radar for the F-14, they were pleased to compress it into sixteen major components. With the F-15, a few years later, they got it down to nine smaller components. But they would never be able to cram that amount of equipment into the tiny nose of the F/A-18. McDonnell Douglas helped a little by moving the cockpit wall back about three inches, increasing the space available for the radar by half a cubic foot, but a lot of squeezing would still have to be done to accommodate it.
Increasing the electrical power supplied to the radar seemed, on the surface, to be an easy solution. With its two high-performance jet engines, the F/A-18 has an abundance of electrical power. Unfortunately, electrical energy causes heat, and heat is the deadly enemy of electronic circuits.
In the past, radar designers had reached for the best possible performance even at the risk of a buildup of heat. But the radars they produced were notorious for malfunctions. A number of F-4 Phantom pilots were able to brag, in a perverse way, that they had gone through an entire career without ever having a radar that wasn’t malfunctioning in one way or another.
McDonnell Douglas, which was beginning to catch Willoughby’s “reliability and maintainability” religion, adopted a new slogan: “We operate them cooler to last longer.” They borrowed a page from NASA’s space flight book and reduced the current, power, and voltage flowing through the radar. The result was to cut the temperature in the transistors, which switch messages in the electronic system, from an average of more than 100 degrees centigrade to seventy degrees, thus dramatically increasing the lifetime of the components in the radar. The standards were made even more strict than those for space flight when it was realized that the thermal shock suffered by a radar set when a plane zooms from the desert floor into the stratosphere and back again is greater than that suffered by the equipment in a spaceship.
The designers saved a little room by replacing the heavy hydraulic system used to move the antenna in older planes with a lightweight, virtually fool-proof electrical drive system. This gave them enough room for a special liquid cooling system for the transmitter, the heaviest user of heat-producing electrical energy. Lightweight corrugated and honeycomb panels were developed to separate parts of the radar and help keep the heat from building up.
This still left the radar too big and too hot. Denied more space and more electrical power, the Hughes designers focused on constructing their radar with the solid-state components that were becoming available in the mid-1970s. This would not only squeeze more performance into a smaller package, but it would also mean fewer moving parts and less buildup of heat.
Up to that time, scientists had found ways to make little computer units, known as circuit boards, in which as many as eight layers of circuits were stacked one on top of the other. If they could produce boards with fourteen layers of circuits, the Hughes designers calculated, they could save a cubic foot or more of space—room for greater power and more cooling capacity.
While Hughes focused on the problem of squeezing the radar into the plane’s nose, McDonnell Douglas and the navy were at work on another problem that would vastly complicate the job of the radar designers.
In the early days of the development of the plane, the navy went through one of its perennial arguments over whether the Hornet should carry a gun. One admiral, since retired, who worked on the early designs says he put the gun in, and the top brass removed it three times before he finally made it stick. The argument in favor of leaving the gun off was a powerful one. With missiles of increasing accuracy, there were those who insisted that the gun would never be needed—despite the experience in Vietnam. An even more persuasive argument was that the gun and ammunition and the fuel required to carry it around added about 5,000 pounds to the weight of the plane. That was a lot of weight for a weapon that might not be needed.
The fliers who had learned the hard way in Vietnam, when they went into combat with missiles but no gun, prevailed. The F/A-18 would have a gun. But where should it be placed? It might be hung on a pod under the plane. That would be fine for strafing, but it would be a sure loser in a dogfight. Similarly, if it were placed in the wing, it would suffer in accuracy and would degrade the performance of the plane.
The only really suitable place for the gun was right on the center line, directly in front of the pilot. Hughes was given the bad news: their radar would essentially be in the same compartment as the gun. The gun mechanism would be between the pilot and the radar, and the barrels of the gun would extend out through the nose, only an inch and a half from the top of the radar.
Because a pilot in a dogfight may have his opponent in his sights for only a few seconds at a time, the gun must be capable of putting out as many bullets as possible in those brief moments. In World War II, many fighters carried six machine guns in their wings, with the guns aimed so the bullets would converge a short distance in front of the plane. During the Vietnam War, when serious attention was once again focused on the gun, the experts searched for a way to pump as many bullets as possible from a single gun aligned with the centerline of the plane. That way, it would not be necessary to worry about focusing the converging lines of bullets from a number of guns. It proved impossible, however, to fire enough bullets through the single barrel of a conventional machine gun without quickly burning out the barrel.
Radar-Gun Configuration in Hornet Nose
The gun, with its six barrels capable of firing 100 rounds a second, nestles on top of the delicate radar set.
The search for a solution took the scientists to the Army Ordnance Museum at Aberdeen Proving Ground, north of Baltimore. There, they examined a Gatling gun of the type developed by Richard Gatling in 1862, during the Civil War. A predecessor of the machine gun, the Gatling gun consisted of a cluster of barrels around a central axis. As a soldier turned a crank, each barrel rotated into position in turn, and a bullet was fired through it.
The guns carried by modern fighter planes employ the same principle. A motor turns a cradle holding the six barrels, and the gun spews out big 20-mm bullets at the rate of 6,000 rounds a minute, although it holds only enough bullets for about six seconds of sustained fire.
When the gun goes off, it sets up a tremendous noise and vibration, heats the surrounding air to 700 or 800 degrees Fahrenheit, and belches out clouds of noxious gases. A more unlikely marriage than that between the delicate radar and the loud, smelly gun is hard to imagine.
The problem faced by the Hughes designers was to insulate the radar against the heat of the gun, seal out the smoke and gas of the burning powder, and cushion the radar so it would not be damaged or thrown out of alignment by the mechanical and acoustical vibration of the gun.
The cooling system already in place was beefed up to solve the heat problem, and the entire radar system was tightly sealed to keep out the gases. Instead of surrounding the radar with access doors, the entire set was mounted on rails. When a mechanic wants to work on the set, he opens one door—the hinged nose of the plane—and the set slides forward on its rails. When he is done, he slides the radar back in place and closes the nose, sealing the radar away from the outside world in its own little cocoon. The most difficult problem was to deal with the vibration. The solution was to identify the frequencies of all the vibrations coming from the gun, create baffles to dampen them, and then tie all the parts of the radar together into one package and balance it delicately so that it acted as a single vibration isolation system.
By the time the Hughes designers finished their work they were confident they had solved the problem of shielding their radar from its noisy, smelly neighbor, but they still w
aited with crossed fingers for the first tests in actual flight. They could also boast that they had cut the components in the radar from nine in the F-15 to five, and reduced the number of individual parts by 8,000 from the radar in the older F-4. The radar, not including the antenna, took up less than four and a half cubic feet and weighed only 338 pounds.
It was a magnificent breakthrough—in the design laboratory. But it didn’t travel well. In the short trip to the factory floor, something went very badly wrong. Even the highly skilled production workers, laboring in a spotless, dust-free area that was more like a scientific laboratory than a factory, were unable to conform the circuit boards to the extremely demanding tolerances required. They even had trouble attaching one layer of the circuit board to the next. More than half of the boards went directly from the production line into the garbage. And, what’s worse, Hughes learned the penalties of being a pioneer. There was no one else they could turn to who knew how to make such circuit boards any better than they did.
Looking back on the experience, it was clear to people in the navy that they should have paid far more attention in the early design of the plane to the problem of producing it efficiently and economically. For Hughes, especially, this proved to be far more difficult than anyone had imagined.
It took Hughes several years to work out the problems of building the components of the radar efficiently, and in the meantime, under its fixed-price contract, the company lost a good deal of money. Just how large the loss was is not known because the company was, at that time, a private firm, not required to publish financial data. Eventually, the company turned the situation around, and the F/A-18 radar became one of its most profitable operations. The solutions involved moving the design lab onto the same “campus” as the factory, just south of Los Angeles International Airport, creating a “producibility lab” where the designers and those who would assemble a new product worked together, and designing an automatic machine capable of making circuit boards with as many as eighteen layers.
