12 October 1943
Burbank, California, USA
Hall had to admit, he was impressed.
When the order came through more than four months ago to send the “Swordfish” P-38 to Nashville so Vultee could use it as a basis for a two-seater variant of the airplane he was expecting an almost exact copy of the tested design. Instead, the engineers had thrown the long-nose Lightning concept out the window and developed an entirely new gondola for a two-seat pilot-trainer version of the airplane.
In the introduction to the design to proposal they stated a number of deficiencies in the Swordfish which would make it poorly suited as a P-38 trainer. Primary among them was that with the pilot moved so far forward the view and feel from the cockpit was altered too much from the standard single-seat P-38. Associated with that was that the Swordfish had a different Center of Gravity, different trim characteristics, different ground-handling—including, most critically, altered take-off and landing behavior.
The design examination continued with an overview of the changes in internal and external structures between the standard P-38 and the Swordfish test-plane. All of the differences meant that only a small percentage of parts and panels could be inter-changed between the two which would add complexity to both manufacture and repair.
Instead, Vultee was proposing a new extended two-seat gondola of their own design using as many existing structures and panels as possible. To accomplish this, they started with keeping the primary pilot in the same location, directly in front of the main spar and placed the second seat between the aft and main spar. The rear, instructor, pilot then sits directly on top of the rear wing structure with a small well cut behind the main spar for his feet and the rudder pedals. This small foot well interferes with the span-wise stiffening corrugations through about two-and-half feet of the upper center wing section so, to compensate, they added diagonal braces which go forward from the rear of the well structure sides to a vertical brace extension from the center of the main spar. This vertical extension then doubles as a bulkhead between the two cockpit sections and serves as a framework from which the rear instrument panel is secured.
The canopy uses the same three sections: forward, rear, and center; but, the center piece is re-framed to remove the sliding mechanism and to hinge from the right side. Between the center section and the rear glass is a plug over the main-spar and vertical extension followed by another right-hinging center piece for the rear cockpit. The rear cockpit canopy then merges with the same contour of the rear glass as the standard airplane.
The new framing on the center Plexiglas extends on the top to change the profile of the glass to better streamline the new junction between the two center canopies. This faring interferes with visibility directly above the pilot’s head, but it is only a minor concern for a training aircraft and a small price to pay to use the existing form for the glass instead of developing a whole a new piece for each section.
The rear of the gondola is then extended about three feet from the standard one-place P-38, accomplished by added a fuselage plug and new skin panels below the rear cockpit. The underside of the gondola is completely unchanged all the way back to the fuselage plug behind the trailing edge of the center wing assembly and terminates in the same tail-cone as the standard airplane, complete with the egress ladder. The plug has several spring-closed hand and foot holds to enable the pilots to get from the ladder—now farther rear—to the wing surface.
Since most of the added weight is planned to be placed between the two main spars, and thus within the Mean Aerodynamic Chord, the only change in aircraft balance that would need to be accommodated is the rear-ward shift of the gondola’s tail cone and the lack of forward armament in the nose. To compensate for these changes, the nose-cone is planned to be replaced with one slightly longer and built with heavy steel and integrated weights. In addition, the hydraulic reservoirs and pumps will be moved from below the radios to in front of the forward bulkhead in the rear of the nose compartment. Any additional shifts in the center of gravity that may be identified from flight-testing can then be added either in the nose-compartment (to shift balance forward) or within the fuselage plug (to shift balance aft).
The end result was a two-seat version of the P-38 which largely duplicated the primary pilot’s experience of the airplane while still accommodating the second pilot in similar comfort. That Vultee achieved this while re-using more than 80% of the existing assemblies was quite impressive, even by Hall Hibbard’s high standards.
The news from Niagara, or more properly from Wheatfield, New York, was just as good. Bell had spent the last three months converting one of the assembly lines to P-38 production and were now ready to start working on their first few test aircraft. In typical fashion, the first handful of planes would be used more to test and tweak the assembly and manufacturing processes than to build real usable aircraft, but it is was an essential step to getting their facility moved into full production. They were not exactly ahead of schedule, but they were running good and getting where they needed to be to get combat-quality aircraft flowing before Christmas—perhaps even as soon as Thanksgiving.
His own designers were also hard at work. Kelly Johnson had sequestered many of Lockheed’s best in a rented circus tent to work on a Secret project ordered by the Air Tactical Service Command (ATSC). Hall, Kelly, and Court Gross were three of only seven people at Lockheed who actually knew what the project was—most of the men working on it were given specific components and sub-assemblies to work on to specifications set forth by Johnson and his core team of four. The order was to deliver a new fighter prototype by November 23rd but the kicker was that this fighter would powered by a British Halford H-1 B centrifugal jet.
A few weeks ago, a representative from the Navy had tried reaching Dick Pulver but was mistakenly transferred directly to Irv Culver in Kelly Johnson’s circus tent—which was unfortunately downwind of a plastics factory. Culver picked up with an informal, “Skonk Works, inside man Culver.” After some initial confusion the call was sent back to Pulver and Hall had later heard all about it. Apparently, Culver and some of the other senior engineers working in the sequestered group thought it was funny to poke fun at the awful smell from the plastics factory by referring to the popular “L’il Abner” comics. Hall agreed, it was funny, but not very professional so he had Johnson deal with it who promptly fired Culver (Culver came back the next day and as far as Hall knew was still working on the project without another word said about being fired).
Progress on the new jet fighter was going well, even though the ATSC still had not sent the official order for the airplane. With access to all of Bell’s work on their P-59 “Airacomet” to go off of as well Kelly Johnson’s previous work on the in-house developed L-133 they were well ahead plan and were expecting to have the completed prototype within a month. Hall was anxious to see the completed product.
Prior to being shunted over to Johnson’s secret team in the “Skonk Works,” Culver had proposed an interesting solution to Compressibility Stalls in the P-38. Although the NACA redesign in the Model 422 had increased the dive limits of the airplane to acceptable combat speeds, the stalls were still occasionally occurring at high altitude in power-on dives and were thus still a problem. Hall heard that most of the new fast fighters were running into the same issue, too, that between Mach 0.78 and Mach 0.8, depending on the plane, the planes would become unstable in some manner with both his P-38 and Republic’s P-47 suffering dives as a result. Culver came up with the idea of fitting 58 inch span by 8 ½ inch chord Dive Recovery Flaps to the mid-chord of outer wings, directly outboard of the engine nacelles. These flaps were to be electrically operated to drop 40 degrees into the airstream on the underside of the wings to change the pressure gradient during high speeds and enable recovery from a compressibility stall.
Ralph Virden had tested a P-38H with these Dive Recovery Flaps installed in a powered dive from 30,000 feet and was able to successfully recover from a Mach 0.83, about 585 mph, dive at 22,000 feet. He reported tail buffeting at those speeds but he was able to maintain vertical control throughout the dive. Further testing revealed that the DRFs provided an overall increase to the average Critical Mach of the airplane by about Mach 0.034, or just over a 4% increase from an unmodified airplane. The Air Force, however, did not see them as essential equipment so they refused permission to produce a retrofit kit for existing airplanes but have given Lockheed the “go-ahead” to add them to future production block so long as it does not interfere with factory output.
In July, General-Electric’s new B-33 turbo was added to P-38H production in what the AAF called the P-38H-15-LO. This new turbo increased the critical altitude of the airplane by several thousand feet and provided a matching performance increase at high altitudes. These Block-15 airplanes were even now starting to arrive in Europe to outfit some of the nascent P-38 Fighter Groups still waiting for airplanes.
On the production line, the final block of 450 P-38’s ordered from the 1942 Budget Year, beginning with AC# 42-103979, were starting to roll off the factory floor as P-38H-18-LO and were almost identical to the Block-15 but had a streamlined landing light installed in the left-wing leading edge instead of the old retractable light the previous airplanes used.
Once that order was complete, they would begin production of Block-20 P-38H’s which were still being finalized and modified according to feedback coming directly to Hall from Tony LeVier, who was stationed with the 78th Fighter Group in England. Some of the requests, such as that for a unified engine control system locking the throttles to the speed and mixture levers, were pretty major and would most likely wait for either a later block or more likely the next major model. Others were more achievable and were being developed by the P-38 Team.
One request was procedural rather than technical and Hall had forwarded it on Milo Burcham and his team to figure out. That was for revised single-engine emergency handling on take-offs. It was a problem which had plagued the P-38 since its introduction but the USAAF had passed on spending resources tackling it with the reasoning that mishandling was a result of pilot error. Now, LeVier had sent word that it was a procedural problem related to the stated actions in the standard Pilot’s Manual and that a better process needed to be developed. From what Milo had relayed to Hall, LeVier and several pilots of the 78th were working on procedures to apply immediately in the field at the Group level but that they wanted review and assistance from the Flight Testing team back in California.
The most recent request, just arriving to Hall the previous week, was related to a rash of engine failures that the 78th and 55th Fighter Groups had started to experience as they were training for high-altitude bomber escort missions and—for the 78th—starting to make their first short range sorties into France and the Dutch Netherlands. The repeated problem seemed to be that the alcohol-based fuel octane booster used in England was vaporizing and causing the humidity in the air to condense and even freeze at high altitude. Hall was not sure there was anything he could do directly about the fuel additives—that would be for the Army to figure out—but LeVier had relayed that the ground crews were recommending come manner of temperature regulation to keep ice from forming in the induction system as well as either insulation or some manner of vapor barrier to keep the fuel lines from icing.
Neither solution sounded likely to Hall. He felt that this was a fuel supply issue rather than an engineering issue and that his groups’ resources would be better spent on other items. The obvious solution to the problems would be to change the octane booster additives in the fuel from alcohol based to Tetraethyl Lead (TEL), which he heard was happening anyway, and that being the case, Hall was inclined to respond that the problem is the Air Force’s rather than Lockheed’s. A recent preliminary report from Col. Kelsey in Ohio, however, had mentioned similar problems with the XP-38J they had been testing.
Hall had been surprised when news came through that the AAF had abandoned the Allison F15 engines in favor of an engine originally intended for the Bell P-63 which had been hastily field adapted to F-Series standards and fit into the XP-38J airplane at Wright Field. This new engine included Allison’s first production Water Injection system which in testing was discovered to cause condensation on the water lines and in the induction system at low boost settings. The chemists explained that because the alcohol vaporizes so quickly it causes a rapid decrease in temperature which in a humid environment can readily fall below the dew point and cause condensation. Under cold and humid conditions, such as at high altitude over the Great Lakes or in Western Europe, that condensation would freeze and cause ice buildup—which was exactly what Tony LeVier was reporting from England with the alcohol-rich fuel.
Since Kelsey indicated that the USAAF would continue pursuing installation of Water Injection he had directly requested Lockheed research solutions to the problem. The issue of condensation on the lines and freezing valves could be easily solved by insulating the water-methanol lines in the Water Injection installation. With the proposed water tank installation location, directly next to the engine nacelles in the first section of the outer-wing leading edge, the water lines will only be a few feet long and could handle the insulation without difficulty. When it came to LeVier’s problem with alcohol in fuel, the insulation was more problematic because of the total length of all the fuel lines in the airplane made this an ill-suited solution. With the plan to move to TEL additives to the fuel, Hall was doubly convinced to ignore the line-condensation problem for the time being.
The problem of induction condensation was both simpler and more complex at the same time. What made it simpler was that all they needed to avoid the condensation was a way to keep the critical surfaces of the intake manifold and induction system above the dew point so the condensation would never form. What made it difficult was managing the temperatures in such a way that it would not increase the charge air temperature to such an extent as to cause detonation.
His engine installation mechanics were now working directly with engineers from Allison on the problem. Allison had determined through testing that the induction condensation was likely caused by uneven heating and fuel-air distribution in the intake manifold, which they were already working to redesign.
Another solution was to find a way to control the minimum temperature of the charge air using the existing carburetor air temperature sensor and existing inter-cooler installation. Although this would not help with the fuel-air distribution problems in the intake manifold it could help keep the induction charge temperature sufficiently high to prevent the condensation problem. Thus, the engineers were working out a way to ensure the air is not over-cooled in cold-air conditions.
The combat groups with the 8th AF were reporting the problem even with the inter-cooler shutters completely closed, that when flying in air colder than -30° F they were discovering that the charge air was not warm enough to prohibit condensation. This meant that they needed to find another way to keep the charge air temperature above a critical point through other means.
A junior mechanic on the engine installation team had the idea of simply covering the inter-cooler inlet with a piece of cardboard, as was commonly done to cover the radiators of automobiles during the cold winters back in his home in Levina, Montana. Lockheed had no way of testing this from Burbank so Hall and joined with the Allison group in sending the recommendation over to LeVier to see if it helps at all. If it does, then Hall will need to divert some resources to developing a more permanent and fully integrated system to enable control of the inter-cooler inlet duct.
The final option, which LeVier was reportedly exploring himself, was to experiment with higher manifold pressure settings using lower engine speed as a way to maintain a sufficiently warm induction charge to avoid condensation while cruising. The Allison representatives had balked at the idea as unsafe, and considering the revealed shortcomings on their current manifold design, Hall was prone to support them, but LeVier had insisted that based on Kelsey’s and Col. Cass Hough’s tests the previous winter on the manifold pressure limits at full power, these new P-38H’s with their F17 engines should not have any trouble running under such conditions.
Hall would just have to wait and see what develops regarding those issues.
Of all the problems the P-38 had experienced during its development and over its first two years of combat, the only one that had not yet been fully addressed was the slow initial roll rate. This was mentioned, repeatedly, in most of the Air Corps and later Air Force assessments and always accompanied by requests to find ways to improve the airplane’s rate of roll; but, it had never been as high a priority as other problems with the airplane. Now, with all of the those other problems solved (for the most part), Hall was able to apply some resources into finding a way to increase the P-38’s roll and reduce the aileron load, especially at high speed.
Previously, the idea had been tossed around to use hydraulics to control the ailerons but in every application they considered they ran into three main problems with the idea: that the pilot would receive no feed-back from the control surfaces and therefore was likely to apply too much force and overtax the ailerons; that there was no way for the system to self-center—that is, to automatically return to a “neutral” position when the control yoke was released; and, that in order for it to work the primary control cables would need to be removed which would prevent emergency control in the event of hydraulic failure. These issues prevented Lockheed from simply installing hydraulic servos to the ailerons connected directly to the yoke.
In July, one of the engineers, Bob Richolt, had dedicated himself to solving the problem by designing a new type of hydraulic servomotor. He finally came up with a design utilizing a pressure valve of his own design which would allow the hydraulic actuator on the ailerons to increase the force applied by the pilot to the yoke rather than simply taking the entire load. This allows an installation which still uses the standard control cables but which multiplies the force on the ailerons from these cables and reduces the force required by the pilot to deflect the surfaces.
Bob had completed his designs and the initial test installation was completed on August 9th. The hydraulic “boosters,” as the flight engineering team were now calling them, were installed on the aft-side of the outer-wing main spar, at approximately mid-span of the ailerons. The installation includes two of the booster servomotors per side, one connected to the “up” control cable, and one to the “down” control cable, by bell cranks which increase the pilot’s force through a push/pull-rod to the aileron.
Milo Burcham himself took the modified plane up on a few test flights in August and September and after some adjustments to the pressure valve settings and changes in the bell crank diameters was able to report back that the control forces required to roll the airplane at all airspeeds were reduced and that at high airspeeds, in excess of 250 IAS, the forces were reduced to less than 20% of the forces required without boosting. At 250 IAS, the initial roll rate of this modified airplane increased from 50 degrees per second to 135 degrees per second; and, at 350 IAS, from 30 degrees per second to an astounding 200 degrees per second.
Bob Richolt was now finalizing his design drawings and specifications so they could be filed for patents and sent out to an appropriate sub-contractor for series fabrication. Meanwhile, a second airplane was fitted with the refined design to be sent off to the Air Technical Service Command for testing and approval by the AAF. Once Hall received the official acceptance from the Army, Lockheed would be able to plan for their integration to assembly in a future P-38 production block.
Another project demanding his attention was the XP-38J, for which he had recently received a revised specification. The biggest change was in the power plant with the move away from the F15R/L engines to a new engine to be developed by Allison based on the E21R with Water-Methanol Injection. This new engine, depending on its final specifications but expected to reach around 2000 bhp, would likely require a purpose-built propeller with a higher specific thrust than that offered even by the three-blade Hamilton-Standard Hydromatic.
Curtiss Electric caught wind of the XP-38J and of the new power projections for the up-rated Allison and had already contacted Lockheed with a proposal to build a four-blade electric high-activity similar to the one used on some P-47’s. Of course, Lockheed was already working with Hamilton Standard as well on a similar Hydromatic, so Hall found himself in the enviable position of being able to ply each contractor off the other. The War Production Board representative at Lockheed had approved Hall to submit formal R.F.P.’s to each company and depending on the tested results hinted that they may approve production from both Curtiss Electric and Hamilton Standard.
Hall Hibbard looked through everything on his desk, amazed at just how many projects his teams were juggling at the moment, and realized that he was over-due for a vacation.