Steam Catapult Development

As part of the British development and qualification of the steam catapult, a platform containing the hardware was added on top of the flight deck of HMS Perseus.

After catapulting deadloads at dockside in mid-1951 in initial development tests, the carrier proceeded to at-sea trials in the outer Firth of Forth, beginning with deadloads and then the catapulting of six surplus, unpiloted Seafire 47s which had the wings removed at the fold joint and just enough fuel for start, warm-up, and the launch. One reportedly took umbrage at being sacrificed in even a worthy cause and managed to climb and turn back toward the ship before crashing into the sea short of the carrier. A video survives of the testing (the Seafires were not, as stated, radio controlled):

https://www.britishpathe.com/asset/51657/

Perseus was subsequently sent to the U.S. for demonstrations of the steam catapult in February 1952, first dockside at the Philadelphia Navy yard and then at sea. The steam catapult allowed the successful launch of a Douglas F3D Skyknight with a 10-knot tailwind. The existing Navy hydraulic catapult required a 30-knot headwind for the same airplane weight. The Navy immediately began planning to require steam catapults in its new aircraft carriers and retrofit existing carriers that had enough service life remaining to justify the conversion. As a result, bigger airplanes with higher performance could now be carrier-based.

Westinghouse: From Hero to Zero

Who would have thought that a company that didn’t even know what a jet engine was (Westinghouse) and one that had never developed a carrier-based airplane (McDonnell) could have succeeded in producing a fully operational carrier-based jet fighter on the first try during World War II? But the more experience engine and airplane manufacturers were too involved in production contracts and more conventional development programs, so the Navy gave them the assignment.

The Army got the plans for the Whittle engine, which was actually running in England, and gave them to General Electric. The Army then contracted with Bell Aircraft, an experienced airplane manufacturer, to design a land-based jet fighter using two of those engines. The resulting P-59 was, by all accounts, a dog.

Westinghouse created its engine from scratch with no outside assistance due to the secrecy imposed. It had an axial-flow compressor, which was the future, instead of the centrifugal compressor used in the Whittle engine, a dead end from the standpoint of increasing thrust significantly. The original Yankee engine, only 19 inches in diameter, worked well enough to demonstrate that the configuration was sound. One was flown under an FG-1 Corsair at NATC Patuxent River beginning in January 1944.

After development and qualification, two flight-rated versions of the Yankee powered the Navy's first jet fighter, the McDonnell XFD-1 Phantom. It became the J30.

The Phantom incorporated all the features necessary for carrier-basing: strong landing gear, folding wings, catapult/holdback hooks, and the tail hook. It was armed with four .50 caliber machine guns. One squadron operated it briefly, including carrier qualification, but the FD lacked range, cockpit pressurization (necessary for comfortable operation at the altitudes jet airplanes operated most efficiently), and an ejection seat. This required a bigger, heavier airplane requiring a more powerful engine.

So Westinghouse scaled up the 19-inch diameter engine to 24 inches for more thrust.  It was subsequently designated the J34. It powered the McDonnell F2D Banshee and the Douglas F3D Skyknight (which necessitated the change of the McDonnell designation from D to H; the original Phantom became the FH and the Banshee, F2H). And also the Vought F7U-1 Cutlass, with the addition of a government-furnished afterburner that was subsequently provided by Westinghouse.

The Navy needed still bigger, more capable jet fighters so it held a competition for an even more powerful engine. Picking Westinghouse to scale its engine up again as the J40 must have seemed like a no-brainer. Originally, the J40-WE-6 was to power the A3D Skywarrior and the J40-WE-8, which was essentially the -6 with a Westinghouse-developed afterburner added, was to provide the thrust for the Douglas F4D Skyray, the McDonnell F3H Demon, and the Grumman F10F Jaguar. See Paul Christiansen's comment below for the subsequent changes in dash number.

For much, much more on the J40 program, see Paul Christiansen's excellent monograph:

It's available on Amazon.

 Fair warning: it's detailed, technical, and comprehensive even by my standards, a blow-by-blow description of the engine and its development. There is relatively limited discussion of the airplanes it powered and the difficulties it caused in their development. For that however, you can read my much less detailed and technical books, Naval Air Superiority and Strike from the Sea.

Unfortunately, Westinghouse, which had been so successful up until then, failed miserably with the new engine as well as the electronic fuel control for the engine and afterburners. Development problems delayed the availability of the Navy's new fighters and long-range jet bomber, first by putting Westinghouse well being behind schedule on deliveries of engines for flight test and then by it not being able to qualify the engine for production at the necessary thrust, requiring Douglas and McDonnell to substitute other engines for the J40.

Westinghouse was not fully responsible for the Cutlass program problems. Vought was in part at fault for the cancellation of F7U-1 production because it wound up significantly overweight; however, availability of a satisfactory afterburner for the proven J34 contributed to program delays and a lack of enthusiasm for the airplane. Vought and the Navy attempted to address the weight problem with the J46 that was based on the J34 and developed concurrently with the J40. It was almost as disappointing but the F7U-3 was produced and deployed, albeit with a J46 of lower thrust and higher fuel consumption than planned, insuring that the F7U-3 Cutlass would forever be known as "gutless" and be replaced by other fighters as soon as possible.

Although production and support of the J34 continued, Westinghouse eventually exited the aircraft engine business.

Where did Westinghouse go wrong? One theory is the early successes did not result in a problem-solving culture within the engine division. Another is that the company did not continue to invest in technology and innovation like P&W and General Electric did. For example, to improve compression ratio for better thrust to weight and lower specific fuel consumption without incurring compressor stalls, P&W developed the two-spool engine and General Electric, the variable inlet guide vane concept. Westinghouse just kept scaling up the basic design, which eventually proved inadequate to the task.

Transition to Martin-Baker Ejection Seats

Early on, even before there were ejection seats in jets, the U.S. Navy established a relationship with Martin-Baker, a company located in England and dedicated to the development of the ejection seat. Two of its officers went there in 1945 to witness ejection seat testing. Although the Navy continued to rely on its airframe contractors to supply ejection seats when required, its specifications favored the Martin-Baker design philosophy of face-curtain activation of the seat.While it didn't buy Martin-Baker seats for various reasons at the time, it did purchase a trainer/test rig from M-B and required their airplane contractors to furnish seats that were fired with the M-B type face curtain (the Air Force used triggers located on the seat's arm rests).

The man in the fedora is James Martin, later knighted as Sir James, visiting the Philadelphia Navy Yard in August 1946 where the Martin-Baker test rig had been installed.

The early ejection seats were simply bailout assists. Even after features like automatic seat separation were added, a survivable ejection had to be initiated at least 500 feet above the ground, much more if the airplane had a high sink rate. Ejection during takeoff or final approach was not an option.

Martin-Baker was dedicated to improving the seat capability and in the mid-fifties succeeded in qualifying a seat system that resulted from a survivable ejection on the runway at speeds above 100 knots. In 1956, BuAer contracted with Grumman to install the new Mk4 M-B seat in an F9F-8T for a demonstration. Flying Officer Sidney Hughes, RAF, successfully ejected from it at Patuxent River in August 1957 while on the runway at 120 kts.

This successful demonstration resulted in a Navy contract with Martin-Baker for the Mk5 seat, which was a strengthened version of the Mk 4 from a crashworthiness standpoint. (An actuator loop was also added at the front of the seat pan to expedite ejection if required, such as an emergency during catapult launch, or when g-levels made reaching the face-curtain loops difficult.) Most in-service Navy fighters with ejection seats were changed over to the new seat, either in production or as a retrofit. It took a few years because each installation had to be developed and certified by Martin-Baker. The seats were somewhat tailored to each aircraft type and therefore identified by a prefix letter, e.g. H5 for the F4H, F5 for the F8U, and P5 for the F4D (shown here).

The changeover appears to have begun with the single and two-seat F9F Cougars in the training command.  The first F9F-8T ejection on Martin-Baker seats was in September 1958. The first Cougar ejection was in November.

Martin-Baker seats were installed in F3H BuNos 146709-146740 at the factory. The first ejections using a Martin-Baker seat from the F3H reportedly occurred in March 1958 with the last in a McDonnell seat in November 1960.

The F-4 Phantom first flew with a McDonnell seat but it was quickly supplanted by the M-B Mk 5 seat in production, probably in 1960 with the second block of production airplanes. (For more detail on the change to Martin-Baker seats in general and the F4H in particular, see http://phantomphacts.blogspot.fr/2013/10/f4h-phantom-ii-ejection-seat-history.html.)

Vought was reluctant to admit that the M-B seat was better than its own (although they did install a M-B seat in the third F8U-3 which flew in late 1958). Its seat was a modification of the Douglas Escapac seat which also provided survivable ejection on takeoff and approach. However resistance was futile. M-B Mk-F5s were installed in F8U-2s toward the end of their production run and retrofitted to surviving F8Us.  (Contrary to some reports, a cockpit console width change was not required, only different rails and catapult fittings plus stiffening of the bulkhead;  M-B assembled the F5 with a narrower seat pan.)

LTjg John T. Kryway cut it a bit fine with the M-B seat but survived after his hard landing on FDR in October 1961 necessitated jettisoning his F8U-1:

The North American FJ-4s appear to have begun to be switched over in early 1961 at the first or second major overhaul after late 1960. The first reported ejection using the Martin-Baker seat was in September 1961. If you don't have a photo of the specific aircraft being modeled, the best bet is the original seat. The earliest example that I found of a MB seat in the FJ-4 is VA-144's 3rd deployment, November 1961 to May 1962. However, there is a picture of a reserve FJ-4B dated July 1963 with the original seat.

No M-B seats appear to have been installed in F4Ds during production at Douglas, but most survivors were eventually converted to the Mk-P5 during an overhaul.

One notable exception to the changeout to M-Bs was the F11F Tiger. A quantity of 201 seats was procured for the Tigers but they were not installed, with the exception of two F11Fs subsequently pulled out of long-term storage for an in-flight thrust reverser program.

Martin-Baker demonstrated its zero-zero seat in 1961, with the Navy procuring it in 1965 as the Mk7. The performance increase was accomplished by the addition of a rocket.

The Mk5 and Mk7 seats are easily differentiated by the parachute housing. On the Mk5, the
parachute was enclosed in a horseshoe-shaped fabric casing that was housed in a black-painted metal shell on the upper seat back. On the Mk7, the parachute was enclosed in a green composite horseshoe-shaped shell mounted on the upper seat back.
 

There were however, many detail differences in the headrests, ejection-initiation handles, seat cushions, straps, among the seats installed in different aircraft, services, and countries. For example, the Martin-Baker seat in a U.S. Navy F-4 was not identical to the one in a U.S. Air Force F-4. The one in the British Phantoms (the right one in the illustration above) was a third configuration.

The changeover to the Mk 7 seat in U.S. Navy fighters began in late 1967 or early 1968 with Mk 5 seats being modified to the Mk 7 configuration. The first F-4Js were produced with Mk 5 seats (deliveries with Mk7s reportedly began in December 1967) and then retrofitted, as were F-4Bs; the changeover was reportedly complete by 1970. (The last RF-4Bs were reportedly delivered with the Mk 7.) The F-8J conversions from F-8Es included the installation of the F7 seat. The Hs (rebuilt F-8Ds) after mid 1968, and all Ks (rebuilt F-8Cs) and Ls (rebuilt F-8Bs) models were also delivered with the F7 seat.

The Conception of the F8F Bearcat

The interweb would have you believe that the F8F Bearcat resulted from a Grumman evaluation of the Focke-Wulf Fw 190 accomplished in early 1943, possibly even at Grumman's facility at Bethpage on Long Island, New York.

There is a report of a captured FW 190 arriving at Wright Field, Ohio in August 1943. However, it seems very unlikely that it would have passed through Bethpage first for an evaluation by a Navy contractor, although I'm sure that it would have been a closely held secret if it did.

As best I can determine, the F8F originated with a memo from Roy Grumman to Chief Engineer Bill Schwendler dated 28 July 1943 requesting a predesign of a small fighter built around the most powerful R-2800 engine available and providing some additional guidelines. It was reportedly the result of previous discussions between those two dating back to at least late 1942.

A predesign drawing by Dick Hutto dated 20 August 1943 indicates that the basic size and shape, including a bubble canopy, of the Grumman G-58 were well established by then.

The story that seems more credible (and supported by contemporaneous documentation) is that Grumman's Bob Hall and Bud Gilles went to England in September 1943 to fly a captured Fw 190. I haven't seen Hall's report, but they were undoubtedly impressed by its speed and maneuverability. It seems likely that they would have returned to Bethpage with the intent to match, if not exceed, its  performance and handling qualities with the new Grumman fighter that they had already envisaged.

It therefore seems almost certain that the basic philosophy that shaped the F8F was not the September FW 190 flight evaluation but the result of 1942 combat experience in the Pacific vis-a-vis the Mitsubishi Zero, the need for what amounted to a fast-climbing interceptor, and the requirement for a fighter the size of a Wildcat to operate from the small decks of the newly created escort carriers.

The Navy ordered prototypes of the G-58 in November 1943 and designated it the F8F. The first one flew only nine months later, in August 1944. Deliveries of the first production aircraft were made in February 1945.

However, the first air group equipped with F8Fs arrived in the Pacific just days too late to participate in the war. (In fact, it was first Navy carrier-based aircraft initiated after Pearl Harbor to get that far; most were canceled before reaching fleet squadrons.)  Within a few years, it was supplanted by jet fighters and relegated to a training role.

The F-111B versus the F-14A, One More Time

I recently had the incentive to revisit this diatribe in the process of responding to a request from another author about the F-111B program. As before, this assessment was made using the respective Standard Aircraft Characteristics charts, the F-111B’s dated 1 July 1967 and the F-14A’s, dated April 1977. While the argument can be made that the F-111B SAC did not reflect its final weight, I believe that the same argument can be made concerning the one used for the F-14A, so it’s at least a pretty close apples-to-apples comparison. Note: In the following discussion, the F-14A is penalized with the weight of an internal gun and ammunition whereas the F-111B is penalized with the weight of the original Airborne Missile Control System (AMCS) design, roughly the same.

Much has been made of how terribly overweight the F-111B turned out. And it was, compared to a totally unrealistic specification. Many think that the F-14A was far lighter than the F-111B, primarily because most comparisons neglect to do so using the F-111B’s design mission for both aircraft. The F-14A is still lighter, of course, because the Navy changed its requirements so that it would be. Deleted were the escape capsule, bomb bay, and swiveling wing pylon stations among other things. The Hughes Airborne Missile Control System, given a few more years of development, was lighter. The structure was designed for 6.5 gs at 49,548 lbs, about 10,000 pounds less than the F-111B’s design gross weight at that g level. In other words, the six Phoenixes and 3,800 lbs of fuel were treated as an overload for the design of the F-14A structure. At combat weight (13,800 lbs fuel and six Phoenix missiles) the F-111B therefore had a load limit of 5.8 g and the F-14A (12,000 lbs of fuel and six Phoenix missiles), a lower (but not particularly constraining) 5.2 g. The result, however, is a somewhat lower structural weight for the F-14A.

According to the F-111B SAC, when it was loaded with full internal fuel and six Phoenixes, it weighed 77,566 lbs and required 11 knots wind-over-deck on a tropical day for launch; the F-14A, not surprisingly, weighed almost 7,000 lbs less but, surprisingly, required 16 knots wind-over-deck, five knots more than the F-111B. Moreover, at its takeoff gross weight the F-111B was carrying 3,000 lbs more fuel than the F-14, making the difference in takeoff gross weight for the same amount of fuel and weapons carried only 3,866 lbs, or 5%, not exactly the amount or percentage difference that most would have guessed given all the negative publicity garnered by the “Sea Pig.” With that additional fuel, the F-111B could loiter on station for 1.5 hours with the combat fuel allowance assuming an acceleration to 1.5 Mach; the F-14A with the two external tanks of overload fuel, and with the same combat Mach number (one has to read the SACs very closely), could only loiter for 1.1 hours.

As for landing, they were both heavy. In fact, the maximum arrested-landing weight limit of the F-14A precluded it from landing back aboard with all six Phoenixes (strictly speaking, the F-14A could land on a carrier with all six Phoenixes, but that was by approaching with little more fuel than required for one attempt, which meant a bolter or wave-off which meant hitting a tanker immediately thereafter rather than having enough fuel for two or three more). The F-111B had a 5,000 lb margin, all fuel, between its maximum landing weight and the landing weight with the standard landing fuel load of 2,417 lbs of fuel and six Phoenix (56,980 lbs). One does not need to be a Naval Aviator to appreciate being able to land with three times the required fuel. On a tropical day at the standard weight, the F-111B needed 15 knots wind-over-deck for landing; the F-14A needed 17 knots wind-over-deck at its maximum landing weight of 51,830 lbs. The F-111B was also less of an handful following an engine failure since its engines were not as widely separated as the F-14A’s.

The F-14 was not notably smaller than the F-111B either. An airplane that has to takeoff with six big missiles (the Phoenix was nicknamed Buffalo), a very capable missile control system, two guys, and enough fuel for an hour or so on station is not going to be the size of an F-4, even though that's what the Navy insisted on when the TFX program began.

 (Full disclosure: the F-111B's nose is folded upward as it was designed to be for parking on a carrier deck.)

For a little more on the F-111B at-sea trials often used to condemn the F-111B's carrier compatibility, see http://thanlont.blogspot.com/2009/03/f-111b-carrier-trials.html

This is not to say that the Navy didn’t do the right thing in getting the F-111B program cancelled and replacing it with the more versatile F-14, particularly since the Hughes AMCS wasn’t ready for prime time (the Navy's Phoenix program was about two years behind schedule and production F-111Bs were about to be delivered). However, with respect to its Fleet Air Defense design mission, it got an airplane that could not loiter as long or land with its full complement of missiles, had a higher stall speed, required more wind-over-deck for takeoffs and landings, and was more difficult to bring aboard with two engines running, not to mention with one inoperative.

So which is the real "Sea Pig" then?

My answer is neither of the above. The F-111B could do, pretty much, the Phoenix-based Fleet Air Defense (FAD) mission that it was intended to do while weighed down with Air Force low-level supersonic mission and other requirements. The F-14 could not do the FAD mission quite as well—but well enough if needs be—and it could also accomplish the carrier Navy’s other, equally important, fighter missions.

Updates to Prior Posts

11 January 2011: Expanded discussion of nuclear weapon delivery technique, "Not Doing It Right", posted 5 June 2008.

The Gift That Keeps On Giving - Final Chapter

The Supreme Court has heard arguments on the dispute over who owes who how much in the settlement of the 1988 contract between the Navy and its contractors, General Dynamics and McDonnell Douglas (now Boeing), to develop the A-12 Avenger. See Here for the New York Times report. The decision is expected this summer.

For my prior blogs on this long running, multimillion-dollar gift to the legal profession over a multibillion-dollar claim and counterclaim, see here, here, here, and here.