Freedom's Forge (46 page)

For their pressurized cabin, Wells and his team worked out a three-“bubble” system instead of trying to pressurize the entire plane—much less complicated and much safer, since sudden loss of pressure in one “bubble” wouldn’t mean loss of pressure in the rest. The first bubble
was the cockpit area, where the pilot, co-pilot, engineer, and radio operator would sit. The second was in the midsection where the gunners were, and the third was for the gunner in the rear. He would be effectively locked in for the duration of the flight, sometimes for eighteen hours—doomed to be the loneliest man in the Army Air Forces.
²⁴

The guns were a problem. Conventional turrets like those on a Flying Fortress or Liberator couldn’t be pressurized, and areodynamics expert George Schairer pointed out they would also add exterior drag on a plane that could tolerate very little drag. So Schairer proposed leaving them out altogether. After all, the B-29 was designed to fly higher and faster than fighters could reach, anyway. Why worry about protection against a theoretically nonexistent threat?

The Army thought about this. Then, a few days after Pearl Harbor, General Kenneth B. Wolfe, the man who would eventually head the B-29 program, sent his assistant Jake Harmon out from Wright Field along with tech chief Roger Williams. They read Wells and Schairer the riot act. There not only would be gun turrets on the B-29, he told them, but retractable ones, both below and above the fuselage. The Boeing men retorted that this would make pressurizing the interior cabin impossible. They explained the unacceptable drag and other technical problems that would arise. Harmon was sympathetic but adamant.

“The general will bust us both to second lieutenants,” he told Wells, if he and Williams didn’t come back with a pledge from Seattle to install those gun turrets.
²⁵

There was some silence in the room. Then Roger Williams happened to mention something he had seen demonstrated at Wright Field by the General Electric Company, an electronic device for aiming and firing the machine guns of fast-flying aircraft: the first onboard airplane computer.

GE’s little machine was like the fire-control systems that had been on Navy warships for years—but they were bulkier and slower and far less precise than a fighter pilot would need. This one could not only aim every gun on a warplane but fire them as well, while correcting errors in direction and angle of deflection simultaneously. Its “brain” was a tiny black box connected to a motor called a selsyn, which was
able to compute in fractions of a second the speed and direction of an incoming plane, including variables like wind speed and exterior temperature, then could either aim weapons for firing separately at a fighter making a sweeping pass, or have the guns all converge on a single “aiming point”—all at the touch of a button.
²⁶

Wells made a call to General Electric, and working with Sperry Gyroscope, Boeing and GE were able to create a remote fire-control system for the superbomber—the first “smart” automated weapons system and ancestor of today’s precision-guided munitions and “smart bombs.”

That left what to do about the wings. The B-29 would be twice the size of the B-17, but had to have the same drag in order to get the high-altitude performance the Army was demanding. Aerodynamics expert Schairer designed the wing with sleek narrow lines and new lighter metals for the engine nacelles and supercharger system for the Wright 2200-horsepower R-3350 engines. But it was still not enough.

The result was Fowler flaps, developed by aviation engineer Harlan Fowler. They were in effect airfoil spoilers that—unlike most flaps—didn’t just hinge down from the wing. Boeing built them so that they actually slid out from inside the wing and then rotated down, creating a visible slot between the flaps and the wings. The device actually increased the wing area by 20 percent, in addition to increasing the wing’s lift. More wing area, noted George Schairer with satisfaction, also meant more range and a bigger load capacity—not just more bombs but more armored protection for the crews.

Flaps like those on the B-29 are now standard for big jet propulsion planes. But in 1941 they were a breathtaking revolution, one that expanded the envelope of the standard four-engine bomber. Together with a heavy aluminum beam called the spar chord inserted through the heart of each wing, it enabled Boeing to guarantee that this plane would carry a high wing loading of sixty-nine pounds per square foot.
²⁷

Impossible, exploded the Air Corps engineers. No plane could sustain flight under that kind of wing load. They dragged Wells, Schairer, and their boss, Wellwood Beall, out to Wright Field in March to show them the error of their ways by running the numbers on a hypothetical airplane called “Design X.” Wells patiently explained why they thought
a real B-29 would do better than Design X. The Army engineers listened, and backed off. The Boeing men returned to Seattle.
²⁸

On May 17, 1941, Boeing’s president got a letter placing a provisional order for 250 B-29s, with a production goal of 25 B-29s per month by February 1943. Ten million dollars would be advanced for development, with $3.5 million for expanded plant facilities.
²⁹
The XB-29 was now the YB-29, and its first three prototypes would be the templates for a warplane whose orders would rapidly expand when war came in December.

Considering that Boeing was already working flat-out to produce its B-17s, the fact that the first prototype rolled onto the runway in early August 1942 was a considerable achievement. It was a stunning sight. Ninety-nine feet long and weighing fifty-eight tons fully loaded, it had a 141-foot wingspan—almost half a football field. Balanced on its tricycle landing gear, its long olive-drab fuselage stood nearly 30 feet high. A B-17 sitting on the ground was only 19 feet high. Yet somehow the YB-29’s four Wright Cyclone R-3350 engines would give this gargantuan beast a cruising speed of 357 miles per hour—70 miles an hour faster than the Flying Fortress—and a ceiling of 31,000 feet—plus an unheard-of range of 5,330 miles, enough to go from San Francisco to New York and back in one trip.

Two more prototypes were finished by September. Still, Boeing had warned the Army that it would take at least two hundred hours of engine tests before any of them were ready for flight. Almost nine hundred different engineering changes had been made by September 9, 1942, when the engines were revved up for the first time and the plane was given its first taxi test.
³⁰
On the fifteenth, the engines got still more tests and the plane was put through a series of “hops” fifteen feet off the runway, to test the landing gear.

Then on September 21, 1942, Boeing’s test pilot Eddie Allen climbed into the cockpit of XB-29 41-002 and at 3:40
P.M
. was airborne. One hour and fifteen minutes later, the plane flashed over the field, flared out, and dropped her Fowler flaps, then her wheels touched down with a screech. Allen climbed out and was surrounded by a crowd of engineers, designers, and mechanics. “Well, she flew,” he said, and broke into a broad smile.
³¹

By that date, some 1,664 Superfortresses were on the order book. Designing and building the prototype had been the easy part. Manufacturing them would be another matter.

First problem was where. Boeing plants were slammed building B-17s, a production program that had spilled over to Vega and Douglas. Those two firms, Boeing’s erstwhile competitors, would produce the so-called BVD Flying Fortresses, more than eleven hundred of them.
³²
The only solution seemed to be for Boeing and its subcontractors to set up entirely new production plants to build the B-29 airframes—which also meant training an entirely new workforce—while other aircraft companies and
their
subcontractors handled the parts and equipment needed for final assembly. To top it all, everything would have to be done under the veil of official secrecy.

No government agency, not even the War Production Board, was prepared to handle this sort of challenge. So in the end, five principal companies—Boeing, North American, Bell Aircraft, Wright Aeronautics, and GM’s Fisher Body—got together with the Army Air Forces to work out a comprehensive production plan.

They agreed Boeing would build most of the B-29s at a brand-new plant out in Wichita, Kansas. It would produce twenty-five B-29s a month by May 1943, they decided—not bad for a facility that didn’t yet exist. North American would in turn convert its B-25 facility in Kansas City to make B-29s, and Bell would build a plant in Marietta, Georgia, to make still more. Meanwhile, Wright would churn out the R-3350 engines the planes would need as fast as they could at their main plant in Paterson, New Jersey, and General Motors would dedicate a new Fisher Body plant in Cleveland to producing the last B-29s needed to fill the Army’s order.

Even this arrangement didn’t last long. It was soon decided to let North American continue to make B-25s in Kansas City and take up the B-29s at a new plant in Omaha. Fisher Body never did make entire B-29s, although they supplied the bulk of wing assemblies and engine nacelles for Wichita, Marietta, and Omaha—as well as for a fourth principal assembly plant in Renton, outside Seattle.
³³
The production
of R-3350s ended up being farmed out, as well, with new plants coming on line in 1943 at Woodbridge, New Jersey, and outside Chicago.

All in all, the B-29 was turning out to be the most massive project in the history of aeronautics. It was also, in the words of historian Tom Collison, “the most
organizational
airplane ever built.”
³⁴
American business had never before been asked to undertake an industrial project of this size or cost or complexity. Even the Manhattan Project turned out to be cheaper. Boeing and its partners set up a Liaison Committee to supervise the entire effort, which included representatives of the biggest of Boeing’s one hundred major subcontractors: Chrysler, Goodyear, Hudson Motors, McDonnell of St. Louis, and Republic Aviation. Major government agencies agreed to stay away. Production and development of the B-29 was left to American business and the Army. In the process, a new working relationship would be forged that would last long after World War II.

And at the center of the entire project were four principal plants in four different cities: Wichita, Marietta, Renton, and Omaha. All four would produce B-29s for Boeing in staggering numbers; all would end up employing tens of thousands of men as well as women; and all would transform the economy of their localities.

But the first, and most important, was Wichita.

“One continuous landing field.” That was one newspaper’s description of Kansas in 1942, and Wichita in particular. Fog was rare in Wichita, and the winters were mild and clear. Boeing already owned a plant there, producing civilian aircraft under the name of Stearman. Two other Wichita companies, Cessna and Beech, had been making small Army and Navy trainers and “puddle jumpers” for the past twenty years.
³⁵

Altogether the three companies employed some fifteen hundred workers. When Wichita’s city fathers found out that the new Boeing plant would employ ten times that number, they realized an economic tornado was about to hit their city. Sixty years later, it would still be making its impact felt as Wichita changed from a rather sleepy former cow town into a major industrial center.

Even as construction of the Wichita plant got started, Boeing production
managers realized they were facing a massive problem: No one, not even Ford at Willow Run, had come up with an assembly-line layout that could handle the gargantuan size and staggering complexity of the Superfortress, with 40,540 different parts and a
million
rivets. The solution Boeing came up with was something not even Sorensen or Knudsen or the other denizens of mass production had ever contemplated. Boeing production engineer Oliver West had developed it with his counterparts at Douglas and Lockheed for building the BVD Flying Fortress, and dubbed it “multi-lining.”

This replaced the classic one assembly line with six, all funneling together around three short final assembly lines at the threshold of the plant’s main doors. The workers on one line worked on the nose and forward fuselage sections; those in the second, on the center section and bomb bays; those in the third, on the tail section; and so on, with the center wing, engine nacelle, and the outboard and leading edge wing, including the all-important Fowler flaps, all getting their separate assembly lines, while pushcarts and forklifts and conveyor belts kept the parts flowing to each separate line.
³⁶

The plan was to keep all the B-29’s preassembly sections as small as possible until the final stage, when cranes hoisted each section into place in the giant final assembly bay, where four B-29s took shape at a time. In the final assembly, workers clambered around and through the fuselages and under the wings, bolting wings together with the whir and thud of a hundred rivet guns, stringing and fixing miles of electrical wiring and control systems, attaching the sixteen-foot props to the engines, and then lowering the landing gear before rolling the gleaming aluminum airplane out of the door to ready it for flight test.

The multi-assembly-line method marked yet another revolution in American manufacturing. First tried at Wichita, it became standard at all the Superfortress plants. Compared to Willow Run, with its long, winding L-shaped construction, a B-29 plant could be built as a square or rectangle: a huge cost saving both for Boeing and the government. And it meant parts didn’t have to travel as far on the plant floor, a saving of time and man-hours impossible to achieve at a plant like Willow Run. It also required fewer workstations than the standard auto single
assembly line, and had the flexibility to introduce new engineering modifications almost as part of the flow of production, instead of forcing everything to come to a halt while changes were made.
³⁷

That turned out to be important, because no airplane ever required more modifications, both on and off line, than the B-29. It was not only the most expensive machine ever produced, but the most complex. From nose to tail, a B-29 consisted of more than 40,000 different parts—compared to a measly 25,000 for a B-24 Liberator. Building one also required Boeing production managers to keep track of fourteen hundred subcontractors, both large (like GE and Bendix, who made the automated gun turrets, and DuPont, who made the Plexiglas observation blisters) and small, who were responsible for everything from the letters of the gauges on the instrument panels to machine tools for pressing and cutting the aluminum for the wings. A single subcontractor slowdown could throw production schedules into a tailspin, while nearly every inspection, every preassembly test and check, turned up another glitch, another problem in a part that perhaps had never been made before, which had to be engineered out before production proceeded. The engines alone required over nine hundred separate engineering changes from the time the first prototype rolled out until the first flight.