The F3H Demon that could have been

taralon

ACCESS: Restricted
Joined
3 August 2025
Messages
4
Reaction score
2
On August 7, 1951 after excruciating delays due to engine development the first XF3-1 took flight in the hands of Robert M Edholm. Powered by a non-reheated J40-WE-6 the test aircraft was dangerously underpowered, and during the landing approach actually suffered a catastrophic engine failure.

Armed with the failure, and a history of delay after delay, the McDonnell team approached the Navy and advised them that the only way that the F3H was going to be able to continue was with an engine change. The Navy was faced with a harsh choice, but McDonnell had a series of design changes to propose.

The more basic ones were a 'simple' engine change. The P&W J57 and the Allison J71 were currently in development and McDonnell had the foresight to approach both manufacturers and get the expected specifications of both engines. Of the two the J57 was expected to be heavier, but also had higher thrust and was the better funded and more mature program per McDonnell's own engineers. The J57 was already flying in a test installation, while the J71 was languishing on the test stand for want of funding and prospective aircraft designs. McDonnell had also gotten verbal promises from P&W that a reheated version of the engine could be made available in Late 1952/Early 1953 with a non-reheated version available for test flight in Early 1952. These designs would require 'modest' reworking of the fuselage and wings, of which some work had already been done by McDonnell engineers on paper based upon their experience with the F3H as well as the XP-88 and it's follow on designs for the Air Force.

Also in the mix was a dual engine redesign of the fuselage with several prospective engine choices, the most powerful of which was a reheated version Wright J65 already selected by the Navy for the powerplant of what would become the FJ-3 Fury, and a version powered by the in production Allison J35. Unfortunately this redesign was going to be expensive as additional modeling and wind tunnel testing would be required for the development of the dual engine fuselage, and it was going to carry a significant weight penalty which would require an entire wing redesign to meet carrier landing requirements.

The Navy after deliberation signed a Letter of Intent to amend the F3H production powerplant to the J57, mainly due to budget issues. It was, of all of the choices put forth, the cheapest choice by far, and at this point the Navy was funding multiple airframe developments and had little to no budget left for expensive redesigns. What was more telling, to the Navy, was that it's backup of three other deeply troubled projects (the F4D, the F10F, and the F7U) it was, despite the issues with the engine either most mature machine, or the one with the least number of troubling flight characteristics. The move to the J57 was what the Navy saw as the 'safe' choice at the time.

McDonnell immediately began reworking the two XF3-1 prototypes for the newly chosen J57 engine, receiving from P&W an empty, ballasted engine shell complete with accessories and a mock afterburner. At the same time they began prototyping a new, larger, wing for the heavier aircraft using test data from their use of the NACA 65A00x series on the XF-88 Voodoo prototype. The wing was broadened in chord at the root by forty inches to increase wing area although span was kept the same resulting in a . The prototype wings were built on a NACA 65A007 plan at the root, thinning to NACA 65A006 at the tip with a 45 degree designed sweep at the quarter line. This design was expected to have less parasitic drag than the NACA 0006 series originally chosen for the wing, and as a redesign to a larger wing area was needed anyways, and no retooling costs were involved, it was seen as a perfect time to test it.

By April 1952 prototype 1 with the original 442 sqft NACA 0006 series wing and protype 2 with the newly designed 519 sqft NACA 65A00x wing were awaiting delivery of the evaluation J57 engine. By May 1952 McDonnell had delivery of two pre-production evaluation J57 engines rated at 9000 lbs of thrust dry. Initial test flights of both prototype aircraft by Robert E Edholm (flown back to back on the same day) were incredibly promising. In a stark contrast to his somber mood after the first test flight of the XF3-1 he was all but ecstatic in his praise of the new aircraft as it had ~30% more thrust with nearly the same dry weight. The test series did show deficiencies with the installation, prospective series production, and range of the new engine. Additionally, the new wing showed issues during high angles of attack where the airflow became detached from the wing, leading to loss of roll control and almost the loss of the aircraft. Fences were installed just inboard of the ailerons in an attempt to correct this issue, but the final correction was found in wind tunnel testing to be a 'dogtooth' transition of the leading edge of the wing. Carrier suitability testing was carried out with Prototype 2 during September of 1952 after the new dogtoothed outer wing panels were built and extensively tested.=. Two new protoypes are ordered by the Navy for delivery in Q1 1953 with the required production changes to the fuselage. These were most notable a change to the engine intakes as they moved six inches aft so that they could be enlarged to provide better airflow, a movement of the wing from 'hip' to 'belly' mounting to simplify the changed needed to mount the larger engine and carry the required strength of the fuselage to wing mounting structure, a move of the fuselage from oval in total form to being straight sided from the midline down to accommodate the required fuel increase, the cockpit and nose were canted down at five degrees to give better visibility, and a shorter, enlarged nose was designed to house the newly specified APD-50 radar.

October 1952 McDonnell received the first Afterburning variants of the J57. Rated at 10000 lbs dry and 13500 lbs wet and featured a crude but workable 'eyelid' nozzle to control engine turbine pressure for dry and reheat operation. Testing of the two prototypes showed that the changed wing design was as fast on the same power settings and experienced significantly less transonic buffeting as velocities reached, and in the case of of prototype 2 exceeded the speed of sound in a shallow dive. December 3 1952 the XF3-1 prototype 2 airframe became the first Navy prototype, and first US prototype of a production jet to exceed the speed of sound.

The US Navy immediately ordered two additional pre-production prototypes with the fuselage design changes and the redesigned wing for delivery in 1953, with plans to move the plane in full production by fourth quarter 1953 as the Navy needed a new production jet immediately given delays and failures in other products as well as the outstanding success that McDonnell had seen in the new jet. Additionally, the US Navy contracted P&W for delivery of the necessary P&W J57 afterburning engines (expected 15K lb+ wet thrust and with a new nozzle) starting no later than Q2 1953.

Additional issues were found with the XF3-2 prototypes when they first flew in May of 1953 mostly with compressor stalls with the engine during rapid throttle increases at high angles of attack and/or speed, and the necessity of wing spoilers to provide aileron input at high speeds, but the US Navy considered those issues minor given the performance of the super sonic jet, and ordered delivery of the 150 planes contracted in March 1951 (prior to the first flight of the prototype!) along with an additional 100 planes.

December of 1953 the first operational squadrons of the F3H-1 Demon were deployed, beginning it's long and storied history which stretched 2 decades in US service and 5 decades in foreign service. Eventually a Dual Cockpit version (F3H-2 pilot/RIO) became the US Navy standard when it transitioned to Missile Armament, providing a template that was used for the F4 Super Demon when it became operational in 1959 powered by two J57-P-8 engines. The single cockpit design of the F3H was eventually embraced by the US Airforce, who faced immense issues with the own development of a super sonic fighter, the F100 being outright canceled due to a rash of fatal accidents, and the F101 Voodoo, a McDonnell product itself, being deemed as 'insufficiently capable' by the McDonnell when compared to the F3H in front of congress. Eventually the F4 Super Demon was also accepted by the Airforce albeit in a single cockpit variant, where it was well loved by the service and the pilots who flew it.
 
Eventually the F4 Super Demon was also accepted by the Airforce albeit in a single cockpit variant, where it was well loved by the service and the pilots who flew it.
Say what now?

USAF Phantoms were all two seaters.
 
Yes, but in a scenario where the F3H was J57 powered in testing in 1952, and operational in 1953, and where the F100 failed the Airforce would have likely purchased some number of F3Hs to fill the role that the F100 did. Given that, and the fact that the designs that progressed to the F4 included multiple cockpit and nose variants. If the f101 was axed in favor of a F4 design I can see McDonnell catering to the Airforce preference for a single seat aircraft. The Airforce never really went all in on the Pilot/RIO concept and from everything I've read it seems that in most F4 missions in Vietnam the second seater in Airforce F4s was basically a junior pilot and second set of eyes, where the front seater managed all the weapon systems. The only ones where the RIO was actually used as a weapons officer in F4s during Vietnam was when they were flying Wild Wiesel missions.
 
I don’t think that the scenario presented here is realistic. My biggest reservations concern the supersonic performance of the J57-powered Demon. First, while it appears that switching from the J71 to the J57 should be feasible, at least based on the dimensions of both engines, I think you overestimate the potential gains. The J71-A-2 was rated at 10200 lbs military power and 14500 lbs with afterburner, compared to 10200 lbs and 16000 lbs, respectively, for the J57-P-4 (the F-8A Crusader engine). If we take a look at the power/speed curves of the F11F-1 as an example and consider that the F3H-2’s starting point is below 1 M, then it is hard to imagine that those extra 1500 lbs would push Demon into any respectable supersonic performance. At best, I would expect something like 1.1-1.15 M, F11-F -like performance, rather than 1.5+ M of the F5D or F8U. And it is hardly surprising, considering the size of Demon’s airframe. It is easy to forget, but F3H-2 was quite a big plane, with 519 sq ft wing area (F-4 Phantom was 530 sq ft), and 21287 lbs empty weight. For comparison, other early J57-powered planes were sitting at:
- F-100A: 18 185 lbs empty, J57-P-9 with 14800 lbs of thrust, 1.28 M
- F-8A: 15 513 lbs empty, J57-P-4 with 16000 lbs of thrust, 1.55 M
- F5D Skylancer: 16 737 lbs empty, J57-P-8 with 16000 lbs of thrust, 1.33 M (SAC value, during testing, they pushed it to 1.63 M with some tweaks per Steve Ginter)
- F4D Skyray: 15 225 lbs empty, J57-P-2 with 14800 lbs of thrust, 0.98 M
In fact, the aircraft that compares best with the Demon in terms of dimensions is the F8U-3 Super Crusader, which weighed 21869 lbs empty, with 450 sq. ft. wing area. However, it had a J75 engine with 24500 lbs of thrust, 10000 lbs more than the F3H-2. It is clear that the Demon was greatly overweight, a consequence of the constantly changing requirements – it started as a light interceptor, then the Navy added the requirement of 30%+ additional fuel to make it suitable for all-weather fighter missions, then J40 was replaced by J71, pushing the empty weight from 18691 lbs for the F3H-1 to the aforementioned 21287 lbs for the F3H-2. Switching to a new engine would not cure this issue, as, frankly speaking, F3H was a rather mediocre airframe compared to its contemporaries. In fact, one could argue that the F7U Cutlass had more unrealized potential due to the engine issues than the Demon. To push Demon into respectability it would require 20 000 lbs+ engine thrust, just to achieve 1:1 ratio thrust-to-empty weight power ratio, which was more or less a bare minimum at the time – so we are speaking about J75 class engine and a completely new airframe. Not to take anything from the F3H-2 – even if limited, it turned out to be by far the most successful fighter among its contemporaries, with far more deployments than F4D and F7U combined, and it was an important stopgap for the Navy before the F-4 became available. Still, it is hard to envision it becoming anything more than that, and most certainly being picked up by the Air Force, as performance-wise even the F-100 largely outclassed the F3H-2 (1.28 M vs 0.97 M, 24k ft/s RoC vs 14k ft/s), not to mention F-101, F-104 or F-106. We can further discuss timelines, which I think are widely optimistic, but still the main issue for me is that F3H just didn’t have enough untapped potential to become something more than it was.

F11F.png
 
The problem is that F-101A had an empty weight of 25 374 lbs, combat weight of 39 475 lbs and 30 000 lbs of thrust, for a T:W ratio at combat conditions of 0.76:1. The F3H-2, even being generous and assuming it would not gain additional weight by replacing J71 with J57, was 21 287 lbs empty and 31 145 lbs at combat weight, what gives T:W for J57-P-4 equal to 0.51:1. To push the F3H-2 to T:W equal to F-101A, you would need 23 700 lbs of thrust - basically J75. As Admiral Connolly would say, "there isn't enough power in all Christendom to make that airplane what we want".
 
It's also worth noting that the J57 engine was in extremely high demand for other aircraft, so much so that when the Air Force adopted a modified A-3 Skywarrior as the B-66 Destroyer they replaced that plane's J57 engines with J71s specifically due to a shortage of J57s.
 
I don’t think that the scenario presented here is realistic. My biggest reservations concern the supersonic performance of the J57-powered Demon. First, while it appears that switching from the J71 to the J57 should be feasible, at least based on the dimensions of both engines, I think you overestimate the potential gains. The J71-A-2 was rated at 10200 lbs military power and 14500 lbs with afterburner, compared to 10200 lbs and 16000 lbs, respectively, for the J57-P-4 (the F-8A Crusader engine). If we take a look at the power/speed curves of the F11F-1 as an example and consider that the F3H-2’s starting point is below 1 M, then it is hard to imagine that those extra 1500 lbs would push Demon into any respectable supersonic performance. At best, I would expect something like 1.1-1.15 M, F11-F -like performance, rather than 1.5+ M of the F5D or F8U. And it is hardly surprising, considering the size of Demon’s airframe. It is easy to forget, but F3H-2 was quite a big plane, with 519 sq ft wing area (F-4 Phantom was 530 sq ft), and 21287 lbs empty weight. For comparison, other early J57-powered planes were sitting at:
- F-100A: 18 185 lbs empty, J57-P-9 with 14800 lbs of thrust, 1.28 M
- F-8A: 15 513 lbs empty, J57-P-4 with 16000 lbs of thrust, 1.55 M
- F5D Skylancer: 16 737 lbs empty, J57-P-8 with 16000 lbs of thrust, 1.33 M (SAC value, during testing, they pushed it to 1.63 M with some tweaks per Steve Ginter)
- F4D Skyray: 15 225 lbs empty, J57-P-2 with 14800 lbs of thrust, 0.98 M


You can't just look at thrust/weight ratios and determine a jet's top speed capability, otherwise the F4D skyray would be faster than the F100a and the F8A.

The big change (other than the engine) is in the wing design. Moving from the NACA 0006 to a NACA 65A00x wing as described reduces parasitic wing drag in the Mach .8 range by 20-25% even accounting for the increase in wing area, and drag in the Mach 1.2 range by 40-45%. At mach 1.2 the thrust requirements for the redesigned wing are ~5000 lbf lower than with the original designed wing, and the 'super sonic' claim is for the xf3h prototype which was 3 feet shorter (another 10-15% reduction in wetted area drag) and 7K pounds lighter than the original timeline F3H-2. The XF3H prototype was ~18K lbs empty and ~26K lbs max take off, and didn't have the 3 foot fuselage plug, or the fuselage deepening that the original timeline F3H-1n received (due to needing more fuel and install space for the larger and thirstier J40 which was specced for it, and whose fuselage was carried over into the J71 powered F3H-2N) which puts it in the realm of the F100a/XF100a in a thrust/weight ratio, and with the F3H wing design changes as stated has nearly identical wing parasitic drag with the XF3H needing roughly 10-15% more power to overcome increase fuselage drag for the same speed. I would expect that the prototype would be able to achieve 1.02 - 1.03 clean and level and 1.2-1.28 in a dive which, under the testing protocols and chest thumping propoganda of the day would make it 'super sonic' .
 
It's also worth noting that the J57 engine was in extremely high demand for other aircraft, so much so that when the Air Force adopted a modified A-3 Skywarrior as the B-66 Destroyer they replaced that plane's J57 engines with J71s specifically due to a shortage of J57s.
A 1951 departure means that the Airforce only has the engine planned/in planning for the B52. The F100 doesn't become an ordered project until January 1952, 25 ordered in February and 250 more in August of 1952. It merely moves the first ordered production jet with a J57 to the Navy rather than the Air Force, with maybe 100 engines produced by P&W in 1953, and steadily accelerating to meet the pace of orders as the timeline unfolds. The Navy likely doesn't pull more than 200-300 engines per year out of production and by 1956-57, that's a drop in the P&W production bucket.
 
Last edited:
What I'd like to see is an AH where the Air Force goes with the single seat F3H-G instead of the Navy's F-4.
 
Back
Top Bottom