The Lockheed F-104 Starfighter looked more like a rocket than an airplane. Out in front was a sharply pointed nose with a long pitot tube. The airplane’s straight, stubby wings were canted downward, and they were so thin and small, like fins, that you wondered how it could fly. Lockheed press releases even described the airplane as “the missile with a man in it.” For pilots, its tiny cross-section made it the kind of aircraft you put on like a glove. The cockpit was small but comfortable, and the pilot sat reclined with legs extended, the way you sit in a sports car.
From This Story
Early versions were designed with an ejection seat that fired downward, and to prevent injuries the pilot wore metal spurs attached to his flight boots, cowboy style. The spurs were connected to cables that would automatically pull his feet against the ejection seat during an ejection. Later, the seat was redesigned to fire upward, but the spurs stayed. Most pilots put their spurs on just before they boarded and took them off immediately after deplaning; others wore them around to show off. When I was a second lieutenant attending flying school, I saw an Air Force colonel wearing an orange flying suit and a dress military hat with “scrambled eggs” on the visor. His spurs were clinking and clanking as he walked. Then and there I knew I wanted to fly the Starfighter.
I got my chance in December 1963, when I was selected to attend the U.S. Air Force Test Pilot School at Edwards Air Force Base in California. At the time, the grand old man of supersonic flight, Colonel Charles E. “Chuck” Yeager, was the commandant of the school, and he was guiding the Air Force toward the new frontier of spaceflight.
Our class had 10 Air Force pilots, two Navy pilots, two NASA pilots, and one pilot each from Canada and the Netherlands. We all wanted to be part of the Space Age even though our very presence here put us in competition with NASA. The Air Force had initiated its own manned space program with the Boeing X-20 Dyna-Soar, a single-seat space vehicle scheduled to make its first flight in 1966, just three years away.
All X-20 pilots would be graduates of Yeager’s school and actually fly their spacecraft from liftoff to an unpowered landing on Edwards’ Rogers Dry Lake. NASA astronauts, on the other hand, returned to Earth in a capsule suspended from a parachute and landed in the ocean.
Yeager was instrumental in changing the curriculum of the test pilot school to include spaceflight training. The name of the school was also changed to Aerospace Research Pilot School, though it was commonly referred to as Yeager’s Charm School. He still had the golden touch: Yeager seemed to have a credit card enabling him to tap into the Air Force budget, and there seemed to be no limit to what he could spend. His motto appeared to be “Follow me. I will put the Air Force in space.”
To give his students a real taste of space, Yeager contracted with Lockheed to modify three production F-104s for high-altitude flight. Designated NF-104s, they were inexpensive trainers that would expose students to altitudes above 100,000 feet. Like the X-15, the NF-104s had small directional thrusters in the nose and wingtips for attitude control up where normal controls had no effect.
Each NF-104 was equipped with a Rocketdyne liquid-fuel rocket engine that used JP-4 fuel and hydrogen peroxide as an oxidizer to produce 6,000 pounds of thrust. With the reaction control system, a student could control the NF-104 on a zero-G trajectory through the thin atmosphere at the edge of space for about 80 seconds. The pilot wore a pressure suit; without engine power at that altitude there was no cockpit pressurization.
It was widely understood that whoever first pushed the NF-104 to its maximum performance was certain to set a world record for altitude achieved by an aircraft taking off under its own power. In 1961 the Soviets had set a record of 113,890 feet with the E-66A, a rocket-powered variant of the
MiG-21 fighter. Some U.S. X-planes had flown higher, but they had to be carried aloft by a Boeing B-52 (see “Mother,” June/July 2001).
In 1963, Lockheed began shakedown flights on the NF-104 with company test pilot Jack Woodman. After a few months the program was turned over to Major Robert W. “Smitty” Smith at the Air Force Flight Test Center (AFFTC), flying out of the Fighter Branch of Test Operations. A year later, when I was assigned to the fighter branch, I did a little off-the-record dogfighting against Smitty. By disabling the safety system that prevented loss of control at high angles of attack and high Gs, he could fly the F-104 near its aerodynamic limits. You couldn’t beat Smitty in an F-104.
To reach maximum altitude, the pilot accelerated the NF-104 at full power to maximum speed, then pulled up into a “zoom climb.” In a zoom, the more energy you could build up during acceleration—and the more precisely you could maintain the optimal climb angle—the higher the airplane would climb when it coasted to the top of the zoom. Smitty reached 120,800 feet on one zoom—not an official world record because it was a test flight and the official monitors were not in place. Optimum climb angle for the aircraft turned out to be between 65 and 70 degrees, which, added to a 14-degree seat cant and a five-degree angle of attack, left the pilot reclined at an angle of about 85 degrees. You couldn’t see the ground from that position, so all zoom maneuvers were made on instruments. On one flight, Smitty tried an angle of 85 degrees, but he lost control and tumbled, going over the top upside down. The aircraft entered a spin but he recovered. Smitty was fearless.
Yeager had taken the NF-104 up three times to get a feel for it, and on December 10, 1963, he was scheduled to fly two zoom flights in preparation for an all-out record attempt the next day. During the morning flight he reached 108,700 feet, but Yeager felt the Starfighter could be taken much higher.
On the afternoon flight, Yeager’s test profile called for him to accelerate to Mach 1.7 at 37,000 feet, light the rocket engine to accelerate to Mach 2.2 at 40,000 feet, and then climb at 70 degrees. As the aircraft passed through 70,000 feet, ground control informed Yeager that he had less than the desired angle of climb. He applied the reaction controls to get back on the flight path, a technique he had used before. But on this flight he was at a lower altitude (101,595 feet) and the reaction controls were not yet effective. There was a higher dynamic pressure on the control surfaces, meaning the horizontal tail would have been more effective. Then, when he attempted to lower the nose at the peak of his climb, he found that neither the aerodynamic controls nor the reaction controls could reduce the angle of attack enough to prevent a spin. Soon he was gyrating in all directions, and nothing would stop it. A mile above the desert and falling like a manhole cover, he ejected.
As his parachute opened, he was struck in the face by the base of his rocket seat. His helmet’s visor broke and burning residue from the rocket entered the helmet. Pure oxygen for breathing was flowing to the helmet, igniting a flame that started to fry his neck and face. As he descended, Yeager removed a glove and used his bare hand to try to put out the fire around his nose and mouth, charring two fingers and a thumb. The aircraft hit the ground in a flat attitude, and Yeager landed a short distance from the wreckage. Within a few minutes a helicopter and flight surgeon arrived. Yeager had second-degree burns on the left side of his face and neck and on his left hand, and a cut on one eyelid.
The loss of an NF-104 was not the only bad news that day: Secretary of Defense Robert S. McNamara announced the cancellation of the X-20. The Air Force lost a manned space program, Yeager was injured and wrapped in bandages, and the Air Force had put a hold on his spending.
The two surviving NF-104s were grounded pending an investigation, so I wouldn’t get to fly one. But the standard Starfighter was still the hottest airplane in the Air Force inventory, and I wanted to get into it. As a new student, I got my first flight in the back seat of an F-104 with an instructor, Major Frank E. Liethen, as he conducted a functional check flight, or FCF. Regulations called for an FCF any time major maintenance had been performed. The FCF pilot would fly the repaired aircraft at the limits of its envelope to determine that it was safe for student pilots to fly. Only the most experienced pilots were asked to fly these potentially hazardous flights.
Liethen had been the outstanding student in his class at test pilot school. After a year as a project test pilot at Nellis Air Force Base in Nevada, he returned to Edwards to attend the new space school. After graduation, he became an instructor in the school. He applied to become a NASA astronaut, but he was turned down—too tall. Just as he graduated from space school, the Dyna-Soar program was canceled. His only chance for a spaceflight was the Air Force program called the Manned Orbiting Laboratory, or MOL. Unfortunately, the MOL (canceled in June 1969; see “First Up?” Aug./Sept. 2000.) was on the drawing board at the time, and crew selection was years away.
Before attending the school, I became proficient in flying FCFs in the McDonnell F-101B Voodoo at Hamilton Air Force Base in California. The F-101B and F-104 were both designed in the 1950s as supersonic interceptors. The F-101B was a twin-engine, two-seat aircraft with a radar intercept officer. The F-104 had a pilot, one General Electric J-79 jet engine with afterburner, and a short-range air-to-air radar. It could fire a heat-seeking AIM-9 Sidewinder missile. Both had high wing loading (total weight carried per square foot of wing area), a T-tail, and pitch-up characteristics (see “Now Departing: T-Tails and Other Killers,” p. 70). Both also had electronic systems to prevent a pilot from entering the pitch-up region.
The F-101 had a horn that sounded in the pilot’s helmet as it neared the pitch-up boundary. If the pilot continued to fly the F-101 to an even greater angle of attack or G-force, a mechanical pusher moved the control stick forward. This very complex system required the FCF pilot to adjust the boundaries during flight. The F-104’s instrument panel had an angle-of-attack gauge. To warn the pilot that he was approaching pitch-up, a needle would move into a red area on the gauge. If the pilot continued to increase angle of attack or G-force, a stick shaker system caused the control stick to shake in the pilot’s hand and emitted a sound similar to a rattlesnake’s.
The Starfighter could be a handful and had a terrible safety record; many pilots had been killed flying it. Only a few years earlier, Iven Kincheloe, who had set a world altitude record in the Bell X-2, was killed in a Starfighter when the engine failed just after takeoff. So as Liethen performed maneuvers in the F-104, tickling the pitch-up boundary, I held the control stick ever so lightly in my hand. He talked on the intercom as he flew, but I watched him like a hawk.
As a student, my zoom flight would be the high point of the 12-month course and my last flight. I’d take the F-104 (not the rocket-powered NF but a standard -104) to the rarefied atmosphere above 80,000 feet.
On the day of the flight, I was sweating profusely, having spent an hour and a half in a full pressure suit. Wearing the helmet and faceplate was like looking at the world from inside a fishbowl. And the helmet was almost as wide as the canopy. I could move my head only a few inches from side to side before the helmet bumped against the plexiglass.
As I sat cooking in the Mojave Desert sun, I felt confident. I’d logged thousands of hours in Air Force fighters, from the F-86 Sabrejet to the F-101B Voodoo. But I’d never flown a Starfighter to 80,000 feet—“Angels 80,” military pilots call it. I’d flown the F-104 often in the previous months to get the feel of it. But you always have little doubts when you’re trying something that you’ve never done before.
If I overcorrected at the top of the zoom, I’d be uncontrollable in seconds. Lieutenant Patrick “Pat” Henry, a Navy pilot in the class just ahead of mine, lost control at the top of the zoom, entered a spin, and eventually ejected. If I were not precise in my planning and control, I’d share his fate. If the engine failed to restart as I was coming down, I’d be committed to a flameout pattern.
The tower’s call interrupted my thoughts. “Zoom 5, you’re cleared onto Runway 04 to hold.”
Sweat was dripping into my eyes, but it would be cool up where I was headed. A quick glance to my left confirmed that my chase aircraft, an F-104 with the call sign “Zoom Chase,” was in position and ready for takeoff. He’d chase me until the pull-up point and then, as I descended through about 30,000 feet, he’d rejoin in formation in order to accompany me through the traffic pattern. He’d check the airplane’s exterior, be ready to offer any assistance I might need, and help keep me clear of other airborne traffic, since I’d be focusing most of my attention on the instrument readings.
The J-79 gave its characteristic howl and roar as I eased the throttle full forward and back again to idle.
“Zoom 5, winds are calm, you’re cleared for takeoff,” the controller said.
No time for other thoughts now. I got a good afterburner light, then pushed the throttle up to maximum afterburner. The acceleration pressed me against my parachute. Control stick aft at 100 knots (115 mph), nose wheel raised at 150, airborne at 175. Landing gear up before 250 knots or I’d rip the gear doors off. Then flaps up. Passing 400 knots, I raised the nose slightly to start my climb and throttled back out of afterburner. Then I started a turn to the east and climbed at 450 knots, waiting for the Mach to build to 0.85.
The chase pilot radioed that my Starfighter looked fit to continue. Climbing toward the morning sun, I had only a few seconds to enjoy flying this beautiful aircraft. It was no time to daydream; I had to focus on the test mission. Climbing at 0.85 Mach, I leveled off at 20,000 feet, passing abeam the Three Sisters Dry Lake. It was time to dump cockpit pressurization and inflate my pressure suit. If my pressure suit failed at this low altitude, I would have plenty of time to repressurize the cockpit, abort the mission, and return to Edwards. Slowly the suit inflated. I felt like a fat man in a telephone booth.
On the way to 35,000 feet, I could see Baker’s Dry Lake in front of me. The lake bed was about 100 nautical miles east of Edwards, and my turning point for the run back in the supersonic corridor—airspace where speeds over Mach 1 were legal. I made a gradual 180-degree turn to the left, glancing over my right shoulder to confirm that my chase was still in position.
Rolling out, I pointed the nose toward the town of Tehachapi. Moving the throttle forward, I selected maximum afterburner, easing the control stick forward ever so slightly to unload the one G of level flight and help the Starfighter ease through the transonic zone. The airplane passed Mach 1.0 with no physical sensation. The Mach needle was really climbing fast now: 1.3…1.4…
I tried pushing the throttle harder against the forward stop, hoping to get every last pound of thrust from the engine. Mach 1.7
…1.8. The F-104 was at its design speed now, and the Mach number was climbing fast. At an indicated airspeed of 675 knots, I started a gradual climb to 38,000 feet. What a tremendous feeling to be going faster and faster. The chase aircraft was miles behind me now. Mach 2.1…2.15… I let the Starfighter accelerate as long as I dared—I wanted every bit of energy I could get. The more speed I built up, the more altitude I’d get over the top.
One last glance at the checklist. I had penciled a reminder for myself when I reached this point: “Check gloves.” Just before he started his pull-up, my classmate, Captain Jerry G. Tonini, had the thumb of one of his gloves start to balloon. Fortunately, he caught it in time. Had the glove popped open, he would have lost all suit pressure. If that had happened, he would have lost consciousness in a few seconds and crashed.
The compressor inlet temperature was approaching the limit: 155 degrees Celsius (311 Fahrenheit). A last check on fuel showed just under 1,200 pounds, the minimum before starting the zoom in order to recover with a safe reserve at Edwards. Go for it, I thought. Pull up. At that moment the image of Yeager wrapped in bandages flashed before my eyes.
I pulled back on the stick gently, entering the climb at a rate of 1 G per second. When the G meter reached 3.5, I kept the pressure constant, and I focused on the attitude indicator in the center of the instrument panel. As I reached 40 degrees of pitch, I began slowly easing off the backstick pressure and held 45 degrees. I monitored the exhaust gas temperature (EGT)—I didn’t want to overtemp the engine.
Quickly I glanced at the altimeter. The needles were spinning too fast to read. I’d passed 60,000 feet; EGT was at maximum: 615 degrees Celsius. I began to retard the throttle to hold EGT constant. Passing 67,000 feet, I brought the throttle back into idle cutoff. The engine shut down and started to unwind; at this altitude, if I left it running, even at idle, it would overtemp.
I held the 45-degree climb angle until the angle of attack reached eight degrees, then pushed forward on the stick. Minimum indicated airspeed over the top was 120 knots, the lowest speed at which there was still enough air flowing over the horizontal tail to ensure the tail would be effective. I felt weightlessness coming on. Even though my shoulder harness was firmly tightened and locked on the ground, I felt my pressure suit lift off the ejection seat and my helmet touch the canopy.
Just approaching the peak of the climb, I treated myself to a sweeping view of Earth. Most of the flight so far had been “head in the cockpit, fly the gauges.” The sky was very dark blue—almost black. I could see the Pacific Ocean in front of me, although still a hundred miles away. There was smog in the Los Angeles basin down to the left, and at my right I saw the San Francisco Bay area. Sightseeing was over; I had to return to business. I’d topped out at Angels 80. It was so quiet I thought I could hear my heartbeat.
I held zero G until the Starfighter had pitched over into a steep dive. I put the speed brakes out, and airspeed started to build up fast as the light brown Mojave Desert came back into view. I was now diving straight down, with Rogers Dry Lake directly below me. Passing 35,000 feet, I restarted the engine.
The EGT started to rise—I had a good light. With the engine running, I started a turn back to the Edwards runway when I was startled by a silver flash on my faceplate. Then I realized it was a drop of sweat.
I passed my landing reference point at 25,000 feet directly above Edwards’ Runway 04, where I had started the flight about a half-hour before. I’d be landing out of the same dead-stick pattern that the X-15 used: 300 knots indicated airspeed and in a 20-degree dive. Base leg altitude was 15,000 feet, but I had flown the pattern many times before and felt quite comfortable. Rolling out on high final at 6,000 feet, I had the 15,000-foot runway directly in front of me. I started the stick coming back for the flare and lowered the landing gear at 250 knots. I checked to ensure the gear was down and locked just before touchdown at 190 knots.
The tires squealed as they burned rubber on the painted white line that crossed the runway at the 10,000-feet-remaining marker. As I lowered the nose gently onto the runway and pulled the drag chute handle, my chase sped past me in a low approach.
With sweat dripping into my eyes, I looked up at the contrail my zoom had etched against the blue desert sky. I had returned safely from the edge of space.
How the F-104 Starfighter Was Born
During the Korean War, the U.S. Air Force became concerned about the advantage the Soviet MiG-15 had over the Lockheed F-80 Shooting Star and the North American F-86 Sabrejet. The MiGs were lighter and had a greater thrust-to-weight ratio, so they could climb faster and reach higher altitudes than the U.S. fighters. The Air Force was downing 12 aircraft for every one it lost, but that was believed to be due to superior pilot proficiency. The Air Force wanted a jet fighter that would exceed the MiG’s performance in every category.
Chief designers Kelly Johnson of Lockheed and Lee Atwood of North American Aviation were invited to visit South Korea to talk with Air Force combat pilots. The two men learned that the pilots wanted greater speed, power, and maneuverability.
New aircraft were already in design or under construction: North American’s F-100 Super Sabre and McDonnell’s F-101 Voodoo, both of which used the Pratt & Whitney J-57 engine. Johnson decided he had to use a more advanced engine. He considered several but chose the General Electric J-79. It was unproven and using it carried some risk, but it had higher thrust than the J-57; if the pilots wanted speed, Johnson would give it to them.
The revolutionary new jet would have Mach 2 speed, be unequaled in time to climb, operate at over 60,000 feet, and combine the attributes pilots wanted. Johnson had to keep airframe weight and drag low. A very thin, straight wing offered excellent performance at high speed. A delta wing has less drag per square foot at transonic speed, but its lift during takeoff and landing is reduced. To compensate, designers had to double its area, so the total drag of a delta wing was greater. The high speed regime, combined with a high thrust-to-weight ratio, pointed to a low-aspect-ratio (in a word, stubby) wing, because it would produce less drag. Johnson proposed a wing so thin and sharp—the leading edge had a radius of only 0.0016 inch—that the edges had to have covers to prevent nicks and keep people from cutting themselves. Ice would not build up on the edge, so there was no need for heavy de-icing equipment. The wings were located nearly two-thirds of the way back on the fuselage, and the tips were squared off, permitting the installation of jettisonable fuel tanks or the Starfighter's primary armament, a pair of AIM-9 Sidewinder heat-seeking air-to-air missiles.
The vertical tail extended about as high as the wing extended sideways, so the vertical tail would contribute a large dihedral effect; dihedral tended to restore the fighter to straight-and-level flight. To moderate the overall dihedral, the wings had 10 degrees of negative dihedral. They drooped a little.
The Starfighter had a “flying tail”—the entire horizontal surface moved—placed high above the engine exhaust, so it could be made of aluminum instead of heat-resistant but heavier stainless steel. Even at Mach 1.5 the flying tail was very effective, allowing the pilot to pull five Gs in a turn at 35,000 feet.
Johnson’s fighter never got a chance to tangle with any MiGs, but if it had, it would have left them in its contrails.
The Home Front
When I was selected for test pilot school at Edwards Air Force Base, my wife Jan was delighted. But she was concerned about the number of pilots killed during my previous assignment and wondered if test flying would be even more dangerous.
Our second child was expected to be born about two months into my year of test pilot training. We already had a four-year-old son, and Jan did not want her children to grow up without a father. Asked about the risk, I explained that I had had the best training in the world, the test aircraft were maintained to a higher level, and that we flew during the daytime and in clear weather. I wasn't sure how much of that was accurate, but she seemed to accept my explanation.
All but three of the pilots in our class were married, and most had children, so Jan and I weren’t the only couple having such discussions. The school may have known this: They planned an open house—an opportunity for the families to visit the school.
On the appointed day, we gathered in the auditorium for Colonel Charles Yeager to make his entrance. When he arrived, he had on rows of ribbons for combat in World War II and for flight test accomplishments. But he also wore a large white bandage around his neck, and his left arm was in a sling. If the premier test pilot in the Air Force was this banged up, it seemed clear to Jan that flight test could be a very dangerous business.
The tour of the hangar held another surprise. Yeager was bandaged up because he’d recently punched out of an NF-104, the wreckage of which was spread out on the hangar floor for an investigation. No piece of his crashed aircraft was larger than a refrigerator, and everything was covered in gray ash. The sight of wreckage was familiar to me, but most civilians, and certainly Jan, had never viewed such a shocking sight.
Jan’s concerns would prove to be well founded. Over the next 25 years, 32 test pilots—friends of mine—would be killed in aircraft.
Now Departing: T-Tails and Other Killers
The T-tail on the McDonnell F-101 Voodoo and the Lockheed F-104 Starfighter could create major problems. At high angles of attack, the outer wing sections stalled before the inner wing sections did, and that tended to move the center of lift forward. At the same time, the downwash from the wing began to impinge upon the horizontal tail, changing the angle of airflow over it and reducing its effectiveness. The combined effect caused the aircraft to pitch up.
Most aircraft pitch down when they stall. The nose drops, the aircraft picks up speed, and it returns to controlled flight. In a pitch-up, the angle of attack increases even with the control stick full forward. The aircraft goes out of control and may end up in a spin from which, in both the F-101 and the F-104, it was sometimes impossible to recover. If a pilot could recover by deploying a small drag chute attached to the tail, the ensuing dive recovery could take up to 10,000 feet. A pitch-up below 10,000 feet resulted in an automatic ejection. Pilots were directed never to intentionally pitch up or spin either the F-101 or the F-104; the pilot’s flight manual called those prohibited maneuvers.
Lurking in the background was another serious phenomenon. Beginning with the North American F-100 Super Sabre, “Century Series” fighters (those with numerical designations from -100 up) like the F-101 and F-104 were designed with a high concentration of mass along the fuselage. This led to a dynamic characteristic known as inertial coupling, a phenomenon that can best be explained by considering a rapid rolling maneuver. Picture the aircraft at a positive angle of attack. It begins a rapid roll around its longitudinal axis, which is displaced from the direction the aircraft is moving in by the angle of attack. As the high mass along the fuselage begins rotating at an angle to the flight path, it tends to diverge from that path, increasing its displacement in pitch and yaw the longer the roll continues. So in addition to needing a big vertical tail for directional stability, an even larger tail was needed to prevent inertial divergence.