For many years after I began my work at Lockheed, I would attend a Wednesday afternoon seminar conducted by the eminent scientists and engineers resident and visiting at Cal Tech. I also attended classes there, especially those taught by Dr. Clark B. Millikan, then head of the aeronautical department. It was my intention to earn a doctorate, and I completed all classroom work in the proper courses only to discover that there was a requirement then in my field for competency in technical German. There simply was not time for me to embark on that course of study.
When it became evident that we were going to need a wind-tunnel test capability of our own on a continuing basis, not dependent on scheduling of a rented facility, I was able to persuade the company to provide $360,000. The top officers always believed strongly in the need for research and backed our efforts.
From what experience I had in working with other tunnels and from work that had been done by the National Advisory Committee for Aeronautics (NACA), I undertook the aerodynamic design myself and assigned one of my best engineers, E. O. Richter, to draw plans for the structure. We put it out for bid and had the tunnel itself—the bare walls—built for $186,000 The rest of the money went for very expensive instrumentation, other construction work, and the model shop.
The result was a very good subsonic tunnel capable of testing to a simulated speed of 300 miles an hour. The test section housing the models was a rectangle twelve feet long by eight feet wide. The tunnel had a very useful constant-speed propeller system that was unique—and sometimes troublesome. In the Cal Tech tunnel, speed would change and have to be adjusted with each different angle of attack. With a simple electrically controlled drive, our tunnel would hold its speed well throughout the range of model position changes. We sold the design to six other companies for a modest $10,000.
That tunnel actually paid for itself on the first real test we put it to, because on our next big airplane design project, the P-38 fighter, we were to encounter a phenomenon about which very little was known—compressibility.
9
Into the Unknown
IN THE LATE 1930S, THIS COUNTRY WAS AWAKENING to a sense of its own unpreparedness for war, and for several years Lockheed had been at work secretly developing a new fighter for the Army Air Corps. When I got back to Burbank after introducing the Hudsons to service with the RAF, this became my first priority.
Specifications for the new fighter had been very clear—two liquid-cooled engines and a speed of 367 miles per hour. We advised the Air Corps that our design would fly faster than 400 miles per hour, a speed unequaled then. Lockheed received a contract for such a plane in 1937, with construction of the first beginning in July 1938. First flight of the XP-38—X for experimental, P for pursuit—was scheduled for early 1939.
It was considered a radically different design—even funny looking, some said. It wasn’t to me. There was a reason for everything that went into it, a logical evolution. The shape took care of itself. In design, you are forced to develop unusual solutions to unusual problems.
For the new fighter, we were required to use the liquid-cooled Allison engine. This meant that we had to have a Prestone radiator. We had a long engine so we had to use a General Electric turbo supercharger. And we had a landing gear that had to retract into the nacelle. By the time we had strung all of that together we were almost back to where the tail should be. So, we faired it back another five feet and added the tail. It was a twin-engine airplane, and that produced the characteristic twin-tailed airplane that would go through 18 versions in all theaters of action in World War II, set some records, and make some design contributions.
The use of counter-rotating propellers on the P-38 was a new and important feature for fighters. It eliminated the torque effect, or pulling to one side.
With the first plane faster than 400 miles an hour, I knew we would be entering an unknown region of flight and possible trouble. It was the phenomenon of compressibility—the buildup of air ahead of the airplane at high speed. In 1937, in connection with our proposal, I had warned the Air Corps, “… as airplane speeds and altitudes (thinner air) increase, consideration must be given to the effect of compressibility.”
When I first anticipated the trouble with compressibility, I went to the two best experts in the world on the subject, Dr. Von Kármán and Dr. Milliken at Cal Tech. I told them what I proposed to do in design, how we intended to compute performance, and our concern about stability and control.
“We don’t know anything different. Go ahead,” they agreed. Dr. Von Kármán had recently delivered a technical paper in Italy on the expectation of compressibility at higher speeds and predicted that shock waves would form on the wing, but he did not develop resultant effects on the airplane itself.
We encountered compressibility, but not immediately.
Our Air Corps project officer for the XP-38 was a pilot, a young first lieutenant, Benjamin S. Kelsey. He was excellent. In those days a project officer with that rank had more authority than many four-star generals do today. If we asked Ben for a decision, we got it—on the spot.
We had trouble before we got off the ground. We had trucked the airplane, under wraps for secrecy, to the Air Corps’ March Field near Riverside, Calif., and reassembled it for first flight. The brakes had been received just the day before, because they had to be qualified first back at Wright Field in Dayton, Ohio, before we could install them. We had loaded the rudder with a 500-pound pedal force and the usual type of linkage to the brakes.
So, on a bright, sunny morning, Kelsey started up those wonderful-sounding Allison engines. He decided to make a high-speed taxi run. He got up to speed, then stepped on the brakes. No deceleration. He pushed and pushed—in fact he bent those pedals that we had tested to 500 pounds of pressure the night before. Fortunately, he was able to stop the airplane short of the end of the runway.
On disassembly, we found a small residue of grease from the oily rags in which the brake shoes had been packed. That didn’t help the braking power at all. Well, we had located our trouble and were ready to fly. No committee inspections, no review. In those days, when we were ready to fly, we flew. Next day, first flight.
One of the design features of this airplane was the high-lift maneuvering flaps—it was the first fighter to have them. They were a fundamental part of the design and had to work.
Kelsey used a few degrees of flap setting, and the takeoff appeared to be splendid. Then the flap links broke, and all the flaps were sticking above the wing. Kelsey describes the experience:
“We developed wing flutter on takeoff. Looking out, we could see the flaps coming up above the trailing edge, so we retracted the flaps; the flutter stopped, but we had to come in on that first landing without any flaps—which was a little unusual.”
We put a full-scale section of the wing with an actual flap and flap mechanism in our brand-new wind tunnel at Burbank, and that first series of tests proved the value of the tunnel. We found the solution quickly. We had a very fine streamlined aerodynamic gap between the flap and the wing, very sensitive to any small imperfections in air flow. So we just drilled holes in the fairing above the flap to let the basic structure control the air flow; and with air flow stabilized, the flap would not be buffeted. In the 10,000 airplanes that were produced, we never had flap flutter again.
When we finally ran into compressibility later in the airplane’s development, many people thought it was a case of tail flutter because of the unorthodox appearance of the airplane. Kelsey was our staunch supporter in insisting that the problem was, instead, compressibility. Because the airplane was the first to get high enough and go fast enough to reach Mach numbers approaching the speed of sound—Mach 1—it was difficult to convince anyone that we had encountered this phenomenon.
When the first squadron of P-38s was delivered to Selfridge Field in Michigan, it was Col. Signa Gilkey who first encountered it in accelerated service testing, going through combat maneuvers and other performance tests that would subject the ai
rplane to higher stresses than in normal flight. When Col. Cass Hough later reported exceeding Mach 1 in a dive over England, we knew that would have been impossible. The P-38 just could not have withstood that; it had to be an inaccurate airspeed reading because of instrument lag in the rapid change in altitude—speed of sound, of course, being different at different altitudes.
Several pilots were lost in the early days of testing. The aircraft would pitch nose down in an uncontrollable dive from which they could not always recover. It would happen at about Mach .67 to .80, building up rapidly once started. We tried all the usual methods to improve the elevators and made them so powerful that they pulled the tail off; our test pilot Ralph Virden crashed to his death.
Lockheed test pilot Marshall Headle, taking up the first of an initial production model, the YP-38, had proclaimed it the “easiest” plane he ever flew. The counter-rotating propellers produced no torque, or one-sided pull. But other test pilots—including Lockheed’s Milo Burcham and Tony LeVier—encountered the compressibility phenomenon. “A giant … hand … sometimes shook it (the plane) out of the pilot’s control,” was how LeVier described it.
We had to find the solution in the wind tunnel; the aircraft was just too dangerous to fly.
Meeting with a committee at Wright Field in Dayton, I pleaded with NACA to let us put a model in its wind tunnel so that we could measure what forces were at work because there had been no high-speed tests of fighter or any other aircraft at that time. Our own tunnel could not achieve the required high speeds above 300 mph. NACA officials protested that every time they had approached such speeds in the tunnel, the model had thrashed around so violently they feared it would cause damage to their tunnel, a risk they did not want to take.
With a model of the P-38—one of the United States’ most deadly fighter aircraft of the World War II era—engineers W. A. “Dick” Pulver, Johnson, Hall Hibbard, Joe Johnson, and James Gerschler.
Lieutenant Kelsey went to Gen. H. H. “Hap” Arnold, then head of Army Air Forces, with the problem.
NACA got the message from the General.
“Put that airplane in the tunnel and run this test for Kelly.
Find out what’s wrong with my airplane. To hell with the tunnel. If it blows up, call me,” was the gist of it.
We proved that the problem wasn’t flutter, and we began to make some progress on a scientific approach to handling compressibility.
Although never solving the problem of compressibility, we learned how to avoid it. In the NACA tunnel, we learned about pressure distribution on the wings, how effective was the tail, and what was causing the compressibility effect that pitched the nose down and resulted in such extreme buffeting.
On returning to Burbank, we decided that if we could not solve compressibility, we could discover a way to slow the airplane to a speed where the effect no longer was a factor. The answer was external dive flaps, or brakes. Put in the right place, they would cause the nose to come up out of a dive and stop buffeting. That right place was on the front wing spar. It also changed the pressure distribution on the wing’s lower front so that the diving force was thoroughly counteracted.
It worked so well that if a pilot extended this flap and just let the wheel alone the airplane would pull itself out of a dive.
There were those who didn’t agree that we should go to this expense and effort during a war. The airplane had been in service for some time before we confronted compressibility. Kelsey was one of the doubters.
“Let me fly that one with the compressibility flap on it, Kelly,” he said on a visit to the plant. “I want to see what there is to this thing.”
Kelsey certainly knew the XP-38. Shortly after its first flight he had flown it across country to Mitchell Field, Long Island—with stops at Amarillo and Dayton—at 420 miles an hour in seven hours and two minutes. Unfortunately, the engines lost power on the landing approach and in the emergency Kelsey pancaked on a nearby golf course. Kelsey was unhurt but the first XP-38 became a pile of rubble.
Kelsey took up the modified P-38 to prove to himself the need for the new compressibility dive brake. Flying from Burbank airport, he put the plane in a dive and soon encountered compressibility effects so extreme that he could not reach the dive flap switch. When the tail broke off and he was descending to low altitude at very high speed, Kelsey bailed out and broke a leg and sprained a wrist. He then became a believer in the dive flap.
Thousands of P-38s already were in service by that time, of course, so in addition to incorporating the new flap into production design, we turned out dive flaps for the planes already overseas. For the U.S. Eighth Air Force in England, we sent 487 sets, along with aileron boosters and improvements for engine cooling at high altitude, that would make the P-38 the most maneuverable fighter in the world. It could have outclimbed, outrun, outmaneuvered, and outgunned any other airplane by a factor of two.
The modification kits were loaded aboard a military C-54 transport. As the plane was approaching the coast of Ireland, it was sighted and shot down by RAF fighters who mistook it for one of the German four-engine Condors that were threatening our convoys. The dive flaps never were put into service with the Eighth Air Force. But they were used in the Pacific later in the war.
We had another tragic loss although without the irony of being self-inflicted. A convoy with a ship-load of P-38s—more than 400—was headed for Murmansk and the Soviet Union when a submarine sank the ship. Those aircraft would have been our contribution to the Battle of Stalingrad. Both of these losses were at key periods of the war when the superior performance of the aircraft would have made a real difference.
Special highly-streamlined fuel tanks to extend range were introduced on the P-38. With 300-gallon tanks, the P-38 could fly more than 3,000 miles nonstop, unrefueled. Test pilot Milo Burcham said he made the flight on seven candy bars and one sandwich. The tanks were useful in combat during the last stages of the war in Europe and particularly in the Pacific. They also had other uses; for example, with the nose removable they became ambulance planes, capable of carrying a litter with a wounded soldier in an emergency.
The P-38 had acquired a troubled reputation because of compressibility, and some Air Corps pilots were reluctant to fly them. To counteract this, Lockheed test pilot Tony LeVier was assigned by General Doolittle to tour air bases in the US and in England to demonstrate the aircraft’s capabilities. Tony, a daring pilot who knew exactly what he could get out of an airplane, did everything possible, including single-engine performance, in convincing demonstrations for the young military pilots.
Not only a top-notch fighter, the P-38 became very versatile—as camera plane, bomber-fighter, strafer, rocket-carrier. It went through 18 different versions, the last carrying a bomb load greater than the early B-17 Flying Fortress. It had excellent stall characteristics because of wing design and was particularly effective in combat against the Japanese Zero. The P-38 pilot could slow down to near-nothing airspeed, pull back on one engine, cartwheel without stalling, and reverse direction to face his adversary.
After the P-38, Lockheed built the XP-58. This was much larger than the P-38, almost the equivalent in weight and power of a four-engine plane. It carried a 75-mm cannon, with a cannoneer stationed behind the pilot.
Its role was to knock down big bombers, and it certainly could have if it hit the mark with that 75-mm cannon. We got the airspeed up to 450 mph, and the plane flew well. But it was very heavy and expensive. We built only two of them as experimental aircraft.
Because of the importance of compressibility as an aviation industry problem and the wide interest in it, I prepared a technical paper covering our own research and what we thought the solutions might be for presentation to the American Institute of Aeronautical Sciences. It was duly cleared by the War Department, and I presented it at a meeting in January 1943. Naturally there were many requests from other companies for copies, and I supplied them.
Then the paper was recalled and labeled “sec
ret.”
The agency charged with assisting, coordinating, and instituting this nation’s aeronautical development did not want to acknowledge the work as industry-initiated. Later NACA did do some testing on its own but had contributed nothing to solving the problem of compressibility on the P-38 except allowing the use of its wind tunnel. And this only under orders from the Army Air Corps. The successor agency, National Aeronautics and Space Administration (NASA), by contrast, has been very aggressive and eager to assist and work with industry. I am happy to report that I enjoy excellent relations with NASA.
The matter of secrecy on compressibility became a moot point, for when the war was over and we were able to investigate, we found in German industry literature a great deal of information on compressibility, its effects, and how to avoid them. The Germans had handled it primarily with the swept wing, which they had been flying since the beginning of the Battle of Poland. Later claims to invention of the swept wing in this country are without foundation. By the end of 1943, the year I presented my paper, the Germans had all the aircraft types they were going to build, and the compressibility phenomenon was the talk of P-38 pilots everywhere.
In the course of World War II, horsepower for the P-38’s engines had been increased from 1,000 to 1,750 per engine. Yet with all this great gain in power, we had been able to increase the speed only 17 miles an hour because of compressibility, the effects being felt first on the propeller long before they were encountered on the wing.
Kelly Page 8
