The Edwards Diaries

Test pilot Glen Edwards kept book on the Flying Wing. Now we know what he thought of the airplane he died in.

John K. Northrop and Glen Edwards never met, but their paths nearly crossed in February 1946, at Wright Field in Dayton, Ohio. Captain Edwards, a 50-mission veteran of World War II, had earned a coveted slot as an Army Air Forces test pilot. But this was the peacetime Army, with low pay, slow promotion, and, for test pilots, a death rate rivalling that in wartime. When Northrop visited Wright Field to discuss his B-35 Flying Wing bomber, Edwards got wind of it: "Mr. Northrop is here today," he wrote in his diary on February 20. "Hope to talk to him tomorrow concerning employment. . .. If such is possible, [I] do believe I'd leave the Army." But before Edwards could finagle an interview Northrop was gone. "Blast the luck," Edwards wrote the next day.

The Army would also need pilots for the XB-35 acceptance tests, and in March, the Bomber Test Branch sent Edwards to California's Muroc Army Air Base, a place of scorching loneliness. He was one of several test pilots--Robert Cardenas and Daniel Forbes were others--to qualify in the N-9M, a pint-size version of the big bomber. "The first takeoff is an experience not soon to be forgotten," he reported to the Air Technical Service Command. "The plane comes off the ground of its own accord between 70 and 75 mph and immediately assumes a steep nose-high attitude." Its directional stability was poor, and when turbulence disturbed it, the little Wing would take four or five oscillations before Edwards could return to the compass heading. He stalled the Wing once ("with great caution") and it recovered normally. "The plane flew surprisingly well," he concluded, ". . .far better than most would expect."

But in his diary he expressed more skepticism. "Boy, that was quite an experience," he wrote that evening. "Quite different from flying anything else. It would take a good bit of practice to get really good at flying the little beauty."

Most of Jack Northrop's airplanes were beauties. He believed that "if something is efficient and beautiful, it is right." During stints at Douglas and Lockheed, he was always dreaming of aircraft sleeker and more efficient than the conventional craft the manufacturers had to build in order to survive. Early on, he met a Czech-born barnstormer and shop foreman named Anthony Stadlman, who told him about tail-less, swept-wing aircraft that had been flown in Europe. From about 1919 until 1927 the two men worked together at Lockheed (when it was spelled "Loughead"), then Douglas, then Lockheed again, and in their spare time they actually built an all-wing glider. Sometime toward the end of this period a rift developed between them, and Stadlman would later claim that a model flying wing he had built and shown to Northrop was the basis for Northrop's subsequent designs. Their parting was bitter.

Northrop produced his first attempt at a powered flying wing in 1928. His Experimental No. 1, which does look suspiciously like the one Stadlman is holding in an old photograph, flew in 1929. Northrop financed the project with profits from the Vega, a sleek, high-wing, conventional airplane he had designed during his second stint with Lockheed.

In September 1939, just as Germany invaded Poland, Northrop opened his own plant. It was superlative timing, and he soon had orders from Norway and Britain as well as Boeing and Consolidated. His vision, though, was still bent on the perfect airplane, the Flying Wing. He retained Theodor von Karman, who taught aerodynamics at the California Institute of Technology and who recruited one of his brightest students, William Sears. They worked like this: Northrop sketched a Flying Wing, von Karman wrote long equations on the blackboard, and Sears inked their thoughts on paper.

The airplane Sears drew became the N-1M, for Northrop First Mockup. It was a true Flying Wing, with all its control surfaces--including rudders--in the trailing edge. It was underpowered, though, and on its maiden flight in July 1940, it couldn't climb out of "ground effect" and simply flew along at low altitude, gaining lift from the cushion of air between its wing and the ground. "It looks like we have an airplane with a 20-foot ceiling," Northrop said, watching the pretty craft skim the Muroc lake bed.

With an improved airfoil and bigger engines, the N-1M finally got up to a respectable altitude. Its best feature was that its wingtips were adjustable, and both the droop and the sweep of the wing could be modified between flights. The winning configuration proved to be a straight wing, one without any droop at the tips and with the sweep set back as far as it would go. Sweep, it seemed, was the key to tail-less flight: Positioned behind the geometric center of the wing's lift, the wingtips became tails. (The restored N-1M Wing can be seen at the National Air and Space Museum's Garber Facility in Suitland, Maryland, although without the inserts that increased the wing's sweep, which was the N-1M's most significant contribution.)

By 1941 German troops occupied most of Europe's capitals and appeared ready to capture London and Moscow next. Fearing it might have to fight a transatlantic war, the U.S. Army wanted a super-bomber that could carry 10,000 pounds of bombs to Germany from North America. Northrop built the B-35, a Flying Wing, while Consolidated Aircraft developed the B-36, conventional in every respect but its size (see "B-36: Bomber at the Crossroads," Apr./May 1996).

Even the Wing was huge by the standards of the day: It spanned 172 feet, weighed 100 tons, and developed 12,000 horsepower from four huge engines. It was such a giant leap from the N-1M that the Army also agreed to fund an intermediate version. One-third the size of the big bomber, the N-9M was supposed to be aerodynamically identical. So it was a bad omen when the first one was delivered late, cost more than budgeted, rattled and shook, failed to deliver the promised range, and then crashed, killing its pilot.

Lacking a fuselage and tail, a Flying Wing can rotate easily around its lateral axis--the imaginary line running spanwise through the wings--and even Jack Northrop fretted that his design might be inherently unstable. While he was experimenting with moving the N-9M's center of gravity forward and aft, the test pilot apparently let the nose point so high that the wing stalled.

Now, if you stall a Cessna, you can release back pressure on the control stick and the airplane will drop its nose naturally and resume flying. A Flying Wing is not so polite. In the N-9M and other all-wing aircraft of the time, the airflow separation at the stall rendered the trailing-edge control surfaces ineffective (see "Go With the Flow," June/July 1995). Or worse, the forces on the controls reversed, slamming the control stick into the pilot. The Wing might go into a tail slide, flip over backwards, or fall off to one side in a spin.

In time, the team of Northrop, von Karman, and Sears tamed the N-9M, but the Army was losing patience with the development problems at Northrop and Consolidated. Development of the B-36 was behind schedule, and the XB-35, which was even more of an engineering challenge, had yet to fly. So the production contract was canceled, leaving only two X (experimental) and 13 Y (service test) B-35s on order.

The XB-35 finally took to the air in June 1946, almost a year after Japan had surrendered. Company pilot Max Stanley flew it to Muroc Army Air Base. "No trouble," he reported. The same couldn't be said of the XB-35 thereafter: Its engines overheated, its propeller shafts vibrated, its propeller gearbox broke down, and its auxiliary power unit (a gasoline-powered electrical generator) failed. In the end, the X models would manage to fly for a total of 36 hours, for an amortized cost of $1.8 million per hour.

It didn't matter, because the Army had already decided to adapt its super-bombers to jet propulsion. Consolidated (now called Convair) would hang four turbojets on its B-36, outboard of its piston engines. Northrop's Flying Wing would get an even more radical makeover, its engines replaced by eight turbojets, along with a new designation: the B-49. Because the airframe had proved airworthy, it could go straight into flight test, and with little delay: The first YB-49 was rolled out in October 1947. "Spewing a twin trail of black smoke, the sky monster swept into the air before the awed thousands gathered to witness the historic take-off," gushed a company press release. Max Stanley was again at the controls, and again it was a delivery flight to Muroc--now the property of the newly independent United States Air Force.

In theory, company pilots performed the initial flight tests on new airplanes to prove their airworthiness, then military pilots put them through acceptance tests. In fact, the Air Force's Bomber Test Branch was involved from the first day. When the second YB-49 was ready, Major Robert Cardenas flew it from the factory to Muroc. Cardenas was also the pilot on February 23, 1948, when the nosewheel door blew off the "Two" (the test pilots coined a shorthand to refer to the first two airplanes--the One and the Two). With jet engines, the Wing simply took off too fast. Thereafter, Cardenas lifted the nose as soon as the wheels left the ground, bleeding off speed until the wheels retracted. "Then you leveled off," he recalls, "and you'd sit there rocking in your seat, back and forth, in unison with the slosh of the fuel." The B-49 had no fuel-cell baffles.

Perhaps more serious, the turbojet lacked the stabilizing effect of the B-35's propellers and prop-shaft fairings. To compensate, the airplane had been equipped with four tall vertical fins, which were carried forward as "fences" to prevent the air from flowing toward the wingtips, a tendency that marred the performance of early swept-wing aircraft. So in some respects, the YB-49 really was a new airplane and could not be counted upon to behave exactly like the XB-35, never mind the relatively docile N-9M.

On April 26, a Northrop crew kept the Two aloft for nine and a half hours, and the flight set a record for jet endurance at altitude, though they had to shut down one engine, and the electrical generator failed. Nobody seemed concerned that the distance was only 3,500 miles--not much better than hundreds of Boeing B-29s had done in the Pacific. Even more astonishing, the old B-29 could carry a larger bomb load, including the large and heavy atomic bomb. Rarely mentioned in the Flying Wing saga is the fact that the YB-49 bomb bays were too small for "Fat Man," the plutonium bomb dropped on Nagasaki and still the U.S. standard in 1948.

Shortly after this modest success, the Air Force sent Cardenas to the University of Southern California to get an engineering degree. Back at Wright Field, Glen Edwards wrote in his diary: "Then this evening all heck broke loose. Seems like I'm bound for Muroc tomorrow by fastest means possible. Plan to run stability [tests] on the YB-49. . .what fun! Sounds like I'll be there for awhile. Packing like mad tonight."

The Air Force had two series of acceptance tests, one for performance measures like speed, altitude, and rate of climb, the other for stability and control--the qualities that make an airplane either easy, comfortable, and safe to fly or tiring, uncomfortable, and risky. Accordingly, the two YB-49s each had a different set of instruments. The first was meant to test stability and control; this was Glen Edwards' specialty, so the One would be assigned to him.

After a hectic flight to Muroc with an Air Force civilian engineer named Charles LaFountain, Edwards met the One on Thursday, May 20. As his check-out pilot, Cardenas occupied the raised pilot's seat under a bubble canopy that was bolted in place. Edwards was in the copilot's seat, at deck level, looking through plexiglass panels in the leading edge of the wing--a view so restricted that he could neither take off nor land. The next day they changed seats. "I flew the airplane," Edwards noted in his diary, "and must confess it is somewhat of an experience. Stability is poor all around--landing is peculiar. Has a great tendency to float." (That was ground effect, especially pronounced on an all-wing aircraft.)

Cardenas left Muroc that Friday. Though Danny Forbes outranked him, Edwards was the more experienced and better-trained test pilot, so he was given overall command of the YB-49 program. It was slow going.

"Got two flights off today with doubtful success," he wrote on the following Thursday. "Darnedest airplane I ever tried to do anything with. Quite uncontrollable at times. Hope to be more favorably impressed as time goes on."

That day, Northrop formally turned the Two over to the Air Force. This was the YB-49 equipped to measure performance, and it would be commanded by Danny Forbes. The two men flew seven hours that day. When they flew the Two, Forbes was aircraft commander and Edwards his copilot; in the One, they swapped seats and responsibilities.

First a rainstorm drenched Muroc Dry Lake and put a halt to the program. Then Colonel Albert Boyd, the famously ferocious chief of the Flight Test Division, turned up with a contingent from Wright Field. What with one thing and another, it was Thursday, June 3, before they were able to resume the original pace. "What a wonderful day this has been!" Edwards wrote that evening, in what would be the last entry in his diary. "Got off two flights on the YB-49, a lovely flight on the DC-6, one on the C-74--and I'm bushed!. . . Col. Boyd flew the YB-49 for the first time today and wasn't too impressed. We all share the same views. A passable airplane in ideal conditions."

Edwards had flown the One in various configurations for nearly four hours that Thursday, stalling it with the landing gear down, with the landing gear up, and with flaps at all possible settings. Then Boyd flew it for another hour, after which the airplane was turned back to the manufacturer for a new round of tests. As Gary Pape summarized these tests in his book, Northrop Flying Wings, they were to probe "stability, control, and stalling characteristics that the AF pilots. . .felt were unacceptable."

On Friday, therefore, Edwards was back in the right seat as copilot in the Two. Over the course of four hours, Forbes and he made a high-power climb, ran speed-power tests at 10,000 feet, and calibrated the airspeed indicator. Then the two pilots telephoned Colonel Richard Horner in Dayton. Both pilots complained about the data they'd collected on aileron response, though it isn't clear whether they were finding fault with the airplane or alluding to trouble with the data collection systems. In any case, both the Dayton and Muroc contingents agreed to make an extra flight next day with "an augmented crew of test engineers."

So at 6:44 a.m. on Saturday, June 5, the Two took off on its final flight. Forbes was in the left seat, as aircraft commander, with Edwards flying as copilot on the right. Lieutenant Edward Swindell, who had the crucial task of balancing the airplane by feeding fuel to the engine from tanks that were forward and aft of the center of gravity, was behind them in the rear-facing flight engineer's seat. Two civilian engineers, Claire Leser and Chuck LaFountain, made up the "augmented crew."

Because it was Saturday, no chase plane was available, so the only first-hand reports came from Forbes himself. Just after 7 a.m., he radioed that he was over Bakersfield and climbing; half an hour later, that he was "over north end of Antelope Valley, 15,000 feet, and descending." As Northrop test pilot Fred Bretcher recalled the mission, it should have involved a climb to 40,000 feet and a series of performance tests--too many tests for Forbes and Edwards to have completed in 30 minutes. Probably the auxiliary power unit had pooped out before the Two reached its service ceiling, so Forbes had returned to an altitude where the APU could generate the power he needed to operate the electrical systems.

Their next chore was a series of stall tests at 15,000 feet, where the air was smoother. According to technical files from this series of tests that are preserved in the National Air and Space Museum archives, an Air Force civilian, Robert Coleman, said that the plan for this flight was to keep increasing engine power and angle of attack as the tests proceeded. "If the airplane proved to handle cleanly during [low-power] stalls," Coleman recalled, "stalls with higher power settings were to be obtained. . .. It is known," he added, "that the pilot was reluctant to attempt the higher power stalls."

If Forbes was reluctant, he had reason. In addition to repeated stalls as copilot to Edwards, he'd had a terrifying experience in February with Major Cardenas. That particular flight, Cardenas recalls, "resulted in a gyration that was abnormal. And I wouldn't stall the airplane again. Forbes knew that." In Test Flying at Old Wright Field, Cardenas gave more details: After stalling the One, he found himself at the controls of an airplane that was pointing almost straight up; refusing to respond to the controls, it was falling tail-first at 5,000 feet per minute. "The aircraft then tumbled over backwards," he wrote. After landing, he urged that no one perform intentional stalls in the YB-49. Forbes, he added, "heartily agreed" with that recommendation.

Wind tunnel tests had predicted just such a possibility, but Jack Northrop had dismissed it. "A vertical tail slide is hardly a maneuver to be courted [in] a 100-ton bomber," he'd told the Royal Aeronautical Society in 1947.

True enough, but a test pilot does what a test pilot must do. Forbes probably experienced his second and last tail slide on June 5, after which the YB-49 went over backwards or sideways into a high-speed, perhaps supersonic spin. Its outer wings tore off, probably when Forbes tried to level out and regain control. At some point the airplane exceeded its structural limit, later calculated to be 4.8 times the force of gravity. At 7:30 a.m., Major Russell Schleeh happened to be driving along the highway north of Muroc Dry Lake. A flash caught his eye--sunlight on aluminum--and he saw the Wing tumbling in at least two pieces. They crashed into the desert northwest of Muroc, the outer panels three miles from the crew compartment, which was consumed by fire.

The crash killed the Wing, as surely as it killed the men on board. After a company pilot showed that the One could indeed be stalled safely--at high altitudes--the Air Force took it back for bombing trials that did not help its cause. The airplane was not a steady platform and needed about four minutes to stabilize into straight-and-level flight. And with the bomb bay doors open, air turbulence tossed the bombs around as they were released. The airplane's supporters pointed out that its faults were easy to remedy.

In January 1949 Cardenas took the YB-49 on a high-speed exhibition run to Washington, D.C. It was also used in a secret project that may have involved its stealth capabilities, which Max Stanley had noticed while flying over an air defense radar station near San Francisco. The radar operators couldn't see the YB-49 until they stepped outside and looked with their own eyes.

Whatever it was about, the project never went anywhere, and the Air Force just wasn't interested in stealth--not then. The gargantuan B-36 carried a bigger load over twice the distance and flew so high that enemy fighters couldn't touch it. The Wing was a medium bomber, in some respects inferior to the B-29 of World War II and no match at all for Boeing's swept-wing B-47 Stratofortress. On the day the Air Force sent its new jets to Washington to impress the politicians there, Russ Schleeh crossed the continent at 603 mph in an XB-47, compared with the 511 mph Cardenas achieved flying the YB-49.

Nor could the gap ever be closed. "Northrop had insisted that the crew, fuel, and everything else had to go into the Wing," explained Theodor von Karman in The Wind and Beyond, his 1967 autobiography. "This load made the Wing thick," with the result that at high speeds the airflow separated and "the plane began to shake and lose stability." Boeing's jet bomber could fly faster, higher, and almost as far--and it could carry Fat Man. "The B-49 had gear problems," Cardenas says. "It had engine problems, it had fuel cell problems, it had all kinds of problems. . .. It was not an operational bird. The cockpit layout was miserable. The crew could not escape if anything happened."

The Air Force had begun to name its bases after Air Force heroes, favoring native sons where possible. The name of Glen Edwards--reared in Lincoln, California--was attached to the lonely facility in the Mojave Desert, which in time became the home of the Air Force's Flight Test Center. Muroc is now known to all the world as "Edwards," an icon so familiar that one word is enough to identify it.

In 1952, broken by the failure of his beloved Wing, Jack Northrop turned his company over to new management under Oliver Echols, a retired Air Force general. The firm prospered during the cold war, and in time it produced the ultimate cold war weapon: the B-2 stealth bomber, of which only 20 are to be built. Because the B-2 has no tail, comes from Northrop, and spans 172 feet, it's sometimes described as a modern B-49--and Jack Northrop's vindication. But even Northrop engineers have said that its shape evolved from stealth research--they started with a clean sheet of paper (see "The Invisible Men," Apr./May 1997).

The B-2 is flown by computers, not by its crew. Jack Northrop built his Flying Wing a generation before there was a proper means to control it, and the design was pushed beyond its capabilities. It couldn't carry an atomic bomb, nor could it reach targets in the Soviet Union--and if it had, it probably would have missed.

Get the latest stories in your inbox every weekday.