Skip to main content


Fall 1989 | Volume 5 |  Issue 2

There was never supposed to be a Sidewinder missile. The Navy didn’t ask for it; the Air Force didn’t want it. Yet a small, dedicated team of Navy scientists and engineers at a desert laboratory produced it against all odds. Bill McLean and his coworkers succeeded where larger teams with much more money and support failed, in developing a simple, inexpensive, immensely successful air weapon, and in so doing they showed how technical creativity can soar in the right environment.

Thirty-three years after it entered service, the Sidewinder heat-seeking air-toair missile remains one of our nation’s most reliable and economical weapons. It represents one man’s vision; that vision was nurtured and brought to reality in an unusual government-research environment especially designed to foster creativity- the Naval Ordnance Test Station, at China Lake, in the Mojave Desert 150 miles northeast of Los Angeles. The man was William Burdette McLean.

McLean, born in 1914, was drawn to mechanical things at an early age. His father, a Protestant clergyman, believed in self-reliance and practical skills; his mother taught him to knit, crochet, and use a sewing machine before he entered kindergarten. As a boy McLean built surfboards, canoes, rafts, and his own photographic enlarger, and when he was eleven he helped his family build a new house. A quiet, serious young man, he was good at science and mathematics in high school and poor at English and history. His technical abilities got him admitted to Caltech, where he took three degrees in physics, finishing with a doctorate in 1939. One professor, Charles Lauritsen, observed that McLean sometimes seemed more interested in the laboratory equipment than in the measurements it produced.

How a small and independent Navy research team built a simple, inexpensive guided missile that is still indispensable more than thirty years later

World War II brought McLean to the National Bureau of Standards (NBS)1 where his mechanical abilities were quickly recognized, and he became head of the Mechanical Design Section of the Ordnance Division, designing the arming devices for proximity fuzes. These were complicated and dangerous devices; a fuze absolutely must function when it gets to the target and not at any other time. Achieving designs that could be safely and reliably manufactured was an awesome challenge. He also worked, while at NBS, on a guidance system for the Bat missile, a large glide bomb. Jacob Rabinow, an inventor who worked under him at NBS and now holds 225 U.S. patents, remembers McLean as the “best engineer I have ever known.”

At war’s end McLean went to the Naval Ordnance Test Station (NOTS) on a visit that he expected to last two months. He stayed there for twenty-three years. In 1954 he was named technical director, assuming a major role in shaping programs and policies at what had become one of the Navy’s most important laboratories. NOTS—today the Naval Weapons Center—was a product of lessons learned during the war about what science and technology could accomplish given sufficient leadership, money, and staff. It was set up in 1943 as a “unique experiment in civilian-military partnership,” to be run by the military under naval commanding officers but with civilian technical directors supervising the technical work. NOTS was to be a “full-spectrum” laboratory, with all the personnel and equipment needed to take ideas from the first rough sketch through full development and even initial production.

At thirty-one McLean was already recognized as a technical leader. His quiet demeanor and half-smile accompanied a lightning mind. He was not charismatic, but he had a presence that commanded both attention and respect. As head of NOTS’s Aviation Ordnance Division, in 1946, he found himself working on aiming systems for air-to-air rockets. His analysis of these systems required consideration of many factors, including altitude, speed, and maneuvering acceleration for both planes. This work convinced him that the most precise possible aiming system for an unguided weapon could never enable it to destroy an unpredictably accelerating target. The weapon would have to be able to guide itself after being fired. Radar systems in those days were quite bulky—too large, McLean felt, to fit into the nose of an air-to-air rocket. But infrared detectors able to sense the heat emissions of aircraft targets were about the size of a dime. So he concluded that a guidance unit sensitive to infrared radiation might be compact enough to do the job.

The small but growing guided-missile community had arrived at a different conclusion. Infrared was thought useless for air-to-air missiles because it wouldn’t work through clouds; only an allweather missile was considered worth pursuing. Since radar homing would work through clouds, both the Navy and the Air Force wrote requirements specifying allweather missiles.

McLean had come to distrust requirements during his stay at the National Bureau of Standards. These official guidelines for new technology to be produced often asked for the impossible. Worse, they sometimes failed to ask for what was feasible and wise. McLean felt that requirements were best written after a device was developed, so instead of thinking in terms of a requirement, he concentrated on the actual needs of a combat pilot.

He began to survey the guided-missile scene and found that indeed, many people expected the all-weather missiles to be unsatisfactory. The radar missiles under development were called pilotless interceptors; they were small airplanes flown by robotic mechanisms instead of pilots. They were going to be bulky, complicated, expensive, likely to break down, and hard to store and maintain aboard ships at sea. McLean decided that a successful missile must address all these faults. He set to work to create a simple, cheap, easy-toready, easy-to-use missile.

To begin with, he concluded that infrared would be good enough. He knew that most bomber attacks would be flown at above twenty thousand feet, where rain is rare and clouds are few. Indeed, almost no military missions of any kind were flown through clouds or fog. And a guidance system that worked 90 percent of the time would be far better than a radar system that was too complicated ever to work reliably and too expensive to be produced in quantity.

In 1947 he began to evolve a heat-homing guidance system. The key idea was a spinning and hence gyroscopically stabilized magnet that would reflect the emissions from a heat source onto a sensitive photocell. When the source moved off the axis of the gyro, the photocell would generate a signal in a coil surrounding the magnet, forcing it to turn its axis toward the moving heat source. This meant it could continuously track a distant aircraft target. A servomechanism (a feedback device for automatic control) could use the same signal to make the system’s control fins accelerate the missile toward the target.

In 1949 McLean submitted a proposal to NOTS management for the development of a “heat-homing rocket.” The last word is significant: McLean thought of the missile not as an unmanned aircraft but as a piece of ordnance, a rocket . Aircraft are carefully and continually maintained for repeated use; ordnance is simply stored and then fired. If a missile were to be treated as ordnance, it would have to be extremely simple and sturdy. McLean also felt it should cost only a few hundred dollars. Others at the time expected guided missiles to cost hundreds of thousands of dollars apiece.

McLean began testing his concept. His team started small—since there was no Navy requirement for a heathoming rocket, there was no major funding. But McLean preferred a small team. For the first two years, he relied on internal NOTS discretionary funds plus Navy money for fuze development. The effort became “Local Fuze Project 602,” with the rationale that the seeker head was actually a fuze attempting to get the missile closer to the target before detonation. Later the project became “Feasibility Study 567"—another obfuscation.

Navy funds for guided missiles were being pruned, and NOTS had just failed to produce an authorized missile, the NOTS AAM (air-toair missile). Meanwhile the Navy’s Bureau of Aeronautics was developing the Sparrow missile and made it known that all guided missiles should be under its direction, as aircraft with robot pilots. It looked as if McLean had begun his project at a bad time and in the wrong place. But he was able to get volunteer help from many of NOTS’s scientists and technicians. His rocket began to take shape.

He set his sights carefully. He aimed at simplicity not only for manufacture but also for shipboard assembly and use in combat. He felt that a good servomechanism for steering would be crucial —with the right servo “you could fly a barn door"—so the wings and control fins could even be rectangular. This reasoning nearly drove his aerodynamicist, Lee Jagiello, to distraction. The first model of Sidewinder did in fact have rectangular fins and wings, and they flew fairly well. But eventually Jagiello prevailed, and the final version had swept-back leading edges on its wings and fins.

The problem of preventing the rocket from rolling too rapidly was solved by a technician named Sidney Crockett. He suggested that solid gyroscope wheels be mounted on flaps on each of the four rear wings. These “rollerons” had notches cut into their outer rims so that the airstream flowing past the missile would make them rotate; should the missile start to roll, the rollerons would automatically respond by forcing the flaps out into the airstream to oppose the roll.

Another key element was the infrared seeker. McLean, wanting to leave options open whenever possible, encouraged parallel work on five different seeker arrangements, labeled A through E . The A seeker won out in the end because it had a gyroscope that functioned independently of the missile’s roll. The gyroscope kept track of the motion of the target, and the fins then turned the missile to move it toward a collision course. The missile avoided a lot of additional circuitry by never worrying about where it itself was; it kept track only of how the target was moving, and it used what was called proportional navigation to reach the target.

Another remarkable innovation was the Sidewinder’s torque-balance servo. Because air density diminishes as altitude increases, a steering system can’t work by simply commanding the fins to turn a given angle for a given turn; that would produce a sharper turn both at higher speeds and at lower altitudes. McLean’s servo solved this complicated problem by generating a torque on the fin. At high speeds or low altitudes the torque would turn the fin a relatively small angle; at low speeds or high altitudes the commanded torque would turn it much farther. This “torquebalance” system automatically compensated for differences in speed and altitude. It enabled McLean’s team to boast that “our missile is independent of aerodynamics.”

Even though the functions the missile performed were complex, its hardware—seven vacuum tubes and about two dozen moving parts— made it about as complicated as a portable radio combined with a washing machine. Since top-of-the-line radio monitoring equipment was expensive and fragile, much of the telemetry on the missile had to be done with kits made for electronics amateurs. An engineer named Rod McClung taught a group of women at NOTS to build the Heathkits; after they were skilled at that, he had them assemble seeker heads.

The Navy was requiring radar-homing missiles; McLean decided infrared would do much better.

As the design progressed, more money was required, and groups from various Washington agencies started showing up to see how the project was coming along. The team developed a series of strategies for briefing and impressing such groups. One particularly successful ploy was to mount the seeker from the missile head on the antenna of a surplus radar van, wired so that the seeker’s output could be used to point the radar antenna. The whole system would then be made to follow an aircraft flying overhead or a distant person holding a lighted cigarette. This proved an endless source of fascination and a valuable tool for converting reluctant bureaucrats and admirais. Another trick was to demonstrate the rollerons, which were both ingenious and easy to understand. But the team was kept on its toes trying to predict just what objections the next group of visitors might bring with them.

In 1951 Adm. William S. (“Deak”) Parsons, deputy chief of the Bureau of Ordnance, visited NOTS; McLean briefed Parsons on his progress to date, and Parsons acted decisively. “We will pursue this weapon system,” he announced, and he decreed that the project would be directed primarily from within NOTS rather than from Washington. Shortly thereafter, NOTS received the first installment of three million dollars in funding. The missile had already been designated Sidewinder, a name suggested by McLean’s colleague Gilbert Plain, who knew that the sidewinder snake senses its prey by detecting its heat emissions. The project, already well along, was now also official.

To get the job done, each part of the missile had to be designed almost simultaneously. A number of small groups set to work, some with help from outside contractors. McLean devoted himself mainly to designing the servo system, but other components presented serious problems as well. The energy to operate the servo and electronics was to be supplied not by batteries but by the hot gas from a small grain of chemical fuel. Getting the fuel to burn properly proved an especially difficult job.

Another focus of attention was the canard fins, which turned the missile in flight. McLean put those control surfaces up front so that the missile could be easily taken apart for shipboard storage. If they had been in the rear, as was common with missiles at the time, electrical connectors would have been needed to link them to the seeker. That would have made assembly more difficult. Also, McLean distrusted connectors because they often became worn out from repeated use. By putting seeker and fins in one package, he avoided the problem.

As the airframe and the seeker were completed, they were tested first separately and then together. Ground firings were followed by air firings, then by air firings against drones and target rockets, and finally by air firings with live warheads against drones. The first Sidewinder test pilot was a young Navy lieutenant, Walter A. (“Wally”) Schirra, who later became world-famous as a moon-mission astronaut.

Late hours at little or no extra pay became common for everyone on the project. Howard A. Wilcox, a member of the Sidewinder design team as well as coauthor of this article, remembers returning to NOTS’s Michelson Laboratory almost every evening in the early fifties. He would invariably see a small knot of cars in the parking lot—the cars of the Sidewinder group. The laboratory lights were always on, and the doors always open, and McLean himself was usually among those at work, often staying until 2:00 A.M. Sometimes a group found itself out on the rocket ranges in the middle of the night. People were willing to search on their hands and knees in the dark for pieces of rockets when they knew that Bill McLean was searching on his hands and knees right alongside them.

The physicist Warren Legier recalls work proceeding “not day to day but hour by hour.” McLean disapproved of delay, and a suggestion from him was usually taken as an indication of a need. “What happened with that component we were talking about yesterday?” he might ask; people felt decidedly uncomfortable when they couldn’t answer such a question.

McLean encouraged the development of instrumentation that could give quick answers to important questions. An engineer named Bob Hummer put together a device that provided an easily read photograph of twenty-four channels of radioed information about what went on in the missile while in flight. The “Hummergram” was far from state-of-theart, but it made data available less than an hour after a test flight. Sometimes a new missile component was designed and built in one day and its performance was checked by Hummergram the next.

McLean kept the group so focused on its objectives that discussion of the Sidewinder took place not only in the laboratories and offices but also over back fences on weekends and at cocktail parties, where one might see a few people with McLean in a corner talking rocket motors. The isolation of the Mojave Desert provided unusual freedom from distractions, and the base became an intellectual pressure cooker.

There was a sense of excitement born of working on a project that everyone felt was important and obviously a good idea, and McLean was always ready to listen to anyone’s ideas and see them tried out. He firmly believed that test results, not argument or calculation, should be the ultimate arbiter. As one physicist later put it, “You could do anything you were big enough to do.”

Some members of the group nervously compared themselves with much bigger operations at work on other missiles—Lee Jagiello once asked, “How can we compete? Over at Hughes they have hundreds of aerodynamicists working on the Falcon. Here we have only one. How can we possibly hope to accomplish anything?” Howard Wilcox replied, “All those people will just get in one another’s way. Besides, we have a secret weapon: Bill McLean.”

The team’s faith in McLean was crucial, for the problems were constant. The seeker gyroscope began to wobble when the signal got too strong, until a damper was designed for it. The tiny infrared detector behaved strangely under some conditions. The gas grain burned properly on the bench but not in flight. For a while vibration from an unknown source—it turned out to be unbalanced rollerons—overwhelmed the tracking signal. One after another such hurdles appeared and were overcome.

By 1953 more than two hundred people at NOTS were involved in the project, along with a like number at other government laboratories and industrial contractors. Flexibility was vanishing, and the design was becoming stabilized. But the missile still couldn’t be gotten to work. By September there had been more than a dozen unsuccessful shots. People began to worry that time was running out. WiIcox ordered careful preparation of two identical missiles. One of them was fired, and a Hummergram revealed that it had suddenly lost its target signal when the rocket motor burned out. This suggested that the gyroscope cager’s retaining pins had broken during acceleration, allowing the gyro to recage. So the second round was fitted with new, stronger retaining pins. It was fired on September 11, 1953, and passed within six inches of a target drone, and Howard Wilcox declared it a virtual hit in a telegram to the Bureau of Ordnance, in Washington.

Emotions ran high both East and West that night. McLean and his wife threw an impromptu beer blast to celebrate. At 2:00 A.M. a call was placed from McLea.n’s home to Adm. P. D. Stroop, former NOTS commanding officer—everyone felt he would want to be among the first to hear the good news. But in the hullabaloo he couldn’t make out what had happened. He demanded to speak to someone dependable: “Get Polly Nicol. She’s got a lot of sense; she’ll tell me what’s going on.” He was told: “I can’t. She’s up on the roof.”

Further shots hit drones. One particularly persuasive photograph, used with effect in Washington, showed an old QB-17 bomber-drone going down in flames after a direct hit. Some “experts” had said the Sidewinder’s proportional-navigation homing system would create instability near the end of the missile’s flight. They were proved wrong. Five-inch-diameter rockets were designed as targets and were usually hit. Small thermite heat sources were mounted on drones’ wing tips; they were usually knocked off with no further damage to the expensive drones. WiIcox estimated the average miss distance of the Sidewinder, against a point target, as an inch or two.

Seeking maximum simplicity, McLean even suggested rectangular wings and fins.

The Sidewinder team now knew it had a working missile, so plans had to be made to engineer it for use in the fleet. Every effort was made to make its assembly and storage as simple as possible. The missiles became so easy to put together that eventually two seamen could assemble them at the rate of one per minute.

Meanwhile the Air Force remained essentially unaware of the Sidewinder. Howard Wilcox, by now head of the project, knew that the Air Force needed the missiles even more than the Navy did, to defend the continental United States against attack by enemy bombers. He tried unsuccessfully to convince the Air Force of the importance of the new missile and finally gave up. Then one day he got a 6:00 A.M. phone call from McLean, who wanted him to join a breakfast meeting three hours hence with Professor Charles Lauritsen of Caltech and Assistant Secretary of the Air Force Trevor Gardner. As the four ate breakfast, in Pasadena, McLean and Wilcox presented the case for the Sidewinder. After two hours of hard listening and pointed questions, Gardner was persuaded. He said he would commission an Air Force study and be sure to tell the analysts what conclusion was needed. But there would have to be a “shoot-off between the Sidewinder and the Air Force’s Falcon missile.

Within a few months the study duly appeared, bearing the promised conclusion that the Air Force needed the Sidewinder and that first it must be tested against drone targets. The shoot-off would take place at Holloman Air Force Base, in New Mexico.

On June 12, 1955, the Sidewinder group arrived at Holloman with two jet fighters, a transport plane, and a telemetry truck. Four Sidewinder rounds had been mounted on the two jets. Air Force eyebrows were raised when the Sidewinders were readied for action in four hours and left on the planes overnight, a procedure that would have been unthinkable with the delicate Falcons. Eyebrows lifted even higher when the visitors requested only a flashlight and a small multimeter for test equipment. The Air Force colonel in charge took the group to see the Falcon test equipment: a battery of dials, gauges, and meters along a wall forty feet long. The Falcons, each loaded with dozens of radio tubes, moved along on a trolley in front of the equipment and were checked at a series of stations. The extent of the equipment suggested that the Falcon might be a very temperamental missile. It was.

When the first Sidewinder was fired it went straight up the exhaust pipe of the drone, blowing it apart even though the missile didn’t carry a warhead. There was cheering from the crowd. The Air Force then announced that the next Sidewinder must fly in downward and track a drone against the infrared glare of the hot desert sand. The Sidewinder again destroyed the drone. The pilot, Lt. Glenn Tierney, was so exultant that he pulled a diving vietory roll that broke the sound barrier. Since sonic booms were against military regulations, he and the project engineer were officially cautioned. But the point had been made.

However, Air Force officials remained reluctant. They wanted to see the Sidewinder prove itself at above fifty thousand feet. No drones could operate that high, so NOTS said it would target ground-fired rockets. Air Force eyebrows rose again at the confidence this showed, but the tests were scheduled. Six rockets were sent skyward, and six planes fired Sidewinders at them from fifty thousand feet. When the pilots returned, they reported success; all the missiles had homed on the rockets. So the Air Force finally accepted the Sidewinder.

Ironically, when photographic data became available several months later, it turned out that all the shots had gone astray. The missiles had been fired with their rollerons removed, making them unstable at high altitude. But by then the Air Force had already written and issued formal requirements for the Sidewinder missile system.

On the first working day of 1956, the Sidewinder was released for naval fleet use; it had its first use in combat two years later, after NOTS implemented a crash program to provide Sidewinders to the Chinese Nationalist Air Force, of Taiwan. On September 24, 1958, a squadron of Nationalist F-86D Sabre jets attacked a much larger contingent of Communist Chinese MIGs. The Communist pilots, surprised by the new weapon, found themselves being hunted down. The Sidewinders accounted for four kills. The Communists were so impressed that they declined all further aerial engagements with the Nationalists.

In Vietnam the Sidewinder and other air-to-air missiles produced a number of allmissile combat aces, and the Sidewinder emerged as the missile of choice among combat pilots. Official figures remain classified, but it has been publicly estimated that later Sidewinders had a success rate of more than 80 percent in Vietnam, much higher than any other air-toair missile. In the 1980s, when AIM-9L Sidewinders were used by the British over the Falklands, they achieved a single-shot kill probability of 82 percent. In both 1981 and 1989 Libyan fighters attacking American planes over the Mediterranean were shot down with Navy Sidewinder missiles.

Other nations have, developed their own versions of the Sidewinder. Using the same principles and sometimes virtually the same design, several hundred thousand U.S. and foreign copies of the Sidewinder have been produced, by far the most of any air-to-air missile. Later models are more sophisticated and more expensive, with longer ranges and finertuned tracking, and costs have escalated. Still, compared with the cost of its targets, Sidewinder is a bargain. Indeed, it is today the least costly U.S. air-to-air missile, with a price in the mid-tens of thousands of dollars apiece while other missiles range into six or even seven figures. And by all accounts it remains the best.

The Sidewinder’s simplicity has given it amazing flexibility over the years. It can be mounted on virtually any airplane, and although it was designed as an air-to-air missile, it has spawned ground-to-air versions (Chaparral) and airto-ground versions (Focus). It has even been adapted for ships (Sea Chaparral) too small to handle larger antiaircraft missiles. Because the missile is self-contained, very little apparatus has to be mounted on its launching platform. The Red-Eye and Stinger missiles are both close relatives. McLean’s emphasis on simplicity has paid off richly.

In 1954, when he was elevated to the post of NOTS technical director, McLean stopped managing the Sidewinder project, leaving it to his number-two man, Howard Wilcox. McLean’s attention then turned elsewhere. He became involved in the development of space reconnaissance and weapons systems. In 1958 NOTS launched six satellites that used orientation systems devised and patented by McLean. Similar systems have since been used by many other satellites.

As a result of his missile and satellite work, McLean came to believe that in the future no ship on the sea surface would be safe from missiles—surface ships were too visible and too slow and couldn’t hide by submerging. He began working on the Moray two-person submarine, on “mother” submarines, and on other submersibles. In 1967 he became director of the Naval Undersea Center (now the Naval Ocean Systems Center), in San Diego.

As he diverged into these larger naval- and space-warfare areas, McLean’s ideas met more resistance. The Sidewinder had meant a switch from radar to infrared, but it was still a guided missile and fitted into the scheme of things; its buyers were simply choosing one missile over another. In suggesting new naval or space tactics, McLean was a technical man invading the province of admirals. His position was not helped by his bluntness about how vulnerable favored systems, such as giant aircraft carriers, might be. With regret he watched his influence wane and the system of official requirements he had fought become even more strict. He complained, shortly before his death in 1976, that innovation had been easier in the old days at NOTS.

In a crucial test for the Air Force, the missile went straight up its target’s exhaust pipe.

Sidewinder and other projects like it could have been a warning that the military development and procurement system was becoming too rigid and bureaucratic. But the kind of creative budgeting that gave birth to the Sidewinder might now land its practitioners in jail. Less discretion is allowed at military laboratories, and intrusive “micromanagement” of government projects has thrived. As bureaucratic controls have multiplied, successes have grown fewer.

What should have been learned is that a system’s success simply cannot be calculated in advance. McLean felt that a specification should never be written for an experimental system but only for production in quantity. And while experimenters must keep the needs of their clients in mind, they must also be free to follow their instincts and findings. The instincts of managers lead toward prediction and control, but creativity must be encouraged and cultivated. The Sidewinder was the result of cultivation. Today, when so many new weapons work badly and cost too much, it is worth remembering how NOTS supported Bill McLean and his band of inventors.

We hope you enjoyed this essay.

Please support America's only magazine of the history of engineering and innovation, and the volunteers that sustain it with a donation to Invention & Technology.


Stay informed - subscribe to our newsletter.
The subscriber's email address.