Can we make a faster better performing Wildcat in 1942?

[Admiral Beez]

Return to the lighter non-folding wings while using the final, most powerful engine ever fitted to the Wildcat, with a Fw 190 type cowl and cooling fan. That's a start. And if we're looking for a speedy variant that's staying on land: ditch the hook, lighten the undercarriage and reduce the internal fuel weight.

 

[RCAFson]

Unfortunately some of these time tables require engines that either were never built or were at least contemplated and either didn't work or never made it through testing.

P&W had been working on their two stage supercharger since 1938, They skipped right over the 2 speed supercharger until sometime in 1940.
The R-2000 supercharger was sort of short of breath. How much was common to the R-1830 I don't know. But the R-1830s were good for 1200hp from sea level to 4900ft in low gear.
The R-2000s were good for 1350hp up to 2,000ft. The R-2000 weighed just about as much as an R-1830 with the 2 speed supercharger and at over 18-19,000ft the two stage engine would give better performance. Putting a similar supercharger on an R-2000 is going to add at least 100lb. Problem is that is just for the dry engine, Trying to put more power through tooth pick props doesn't work well.

Some of the Avenger problems were because of the Navy, why it took 6 months to order the two prototypes after the 7 month design selection process I don't know.
Flight trials began on 7th Aug 1941, Getting the navy to actually order the Prototypes even 3 months earlier might have meant more than 6 Avengers at Midway (still would not have fixed the torpedoes)
Navy also wanted two stage superchargers. What they were expecting to torpedo at 30,000ft I have no idea. Now keep this in mind, Wright announced, but only delivered about 2 two stage R-2600 engines. Which did not work very well and soon disappeared never to be heard from again.

Wright R-2600 engines,
1600hp for take-off. Available in large numbers in 1940-41.
1700hp for take-off, none in 1940, 443 built in 1941, 206 of them in Dec, Production in 1942 is rather sketchy in early part of the year, it takes until June to beat the Dec total, In the worst month of 1942 (Feb) they only built 60 engines between two plants.
1900hp for take-off. Does not show up in double digit numbers until July of 1943.

Now for the reality check.
The 1700hp T-O rating R-2600 was good for 1450hp at 14,100ft no ram.
The engine in the F6F-3 was good for 1800hp at 15,500ft but with ram could hold to 18,700ft
In a British test they got 371mph in low gear at 18,700ft.

I doubt that at 16,000-17,000ft the R-2600 powered Hellcat was going to make 360mph.

What the two stage supercharged R-2600 might have done I don't know, they yanked that engine out of the F6F-1 prototype in about one month.

There was a prototype F6F-4 with a single stage R-2800-27:

Wayback Machine

 

[Shortround6]

The 'now' part is covered by the historical F4F. The perfected part is what the people in Grumman will be working instead of on the F5F and P-50.

Grumman was, as shown by the list, working on a lot of other things than the F5F/XP-50.
They were already working on the F6F in fall of 1940.

The F5F/XP-50 was not taking up a big share of their time.

Grumman was working on 3 production airplanes aside from the F4F while they were working on the F5F/XP-50 those tended to overlap quite bit.
Besides, the XP-50 was part of the "turbo" program for the improved F4F.

I am just not seeing 30mph worth of aerodynamic "tweaks" to the F4F.

It is more the area of 2-3mph per tweak, not 5 mph.

I mentioned the difference in speed.

Some of these 'tweaks" are also a different percentage of the total drag.

A hanging tail wheel might have identical drag all on it's own but on a streamlined airplane it may be 0.033% of the drag and on a high drag fighter it might be 0.024% of the total drag at the same speed.

 

[Shortround6]

Wow so the solution may just be get FM-2 into production earlier. That could have been very helpful for the Marines and the FAA as well.

Shortround's point about the 'toothpick propellers' is also interesting. Can you put a broader bladed propeller on an F4F-3 or FM-2?

The prop on the FM-2 went 394lb.
The prop on the F4F-3/F4F-4 went about 315lbs, individual props could vary a few pounds.
You need enough prop to handle the power, too much prop just adds weight and drag.

I feel like I am repeating things a lot.

Grumman or Eastern Aircraft CANNOT get the FM-2 into production until Wright is making the 1300hp version of the R-1820 engine in quantity. like 200-300 a month.
Having a couple of tool room samples for testing does not mean you are ready to go into production.
Wright needed to figure out what they needed to get 1300hp out of the R-1820 and THEN they needed to figure out how to make the engine AND get the new tooling and factory space for it.

Once again, the 1300hp version (the -56) used

Different cylinder heads than the older versions,
Different cylinders with a totally new way to make the fins
A different crankcase
Different crankshaft.

assorted other changes. Next to NO parts on the 1200hp version would work on the 1300hp version.
You need a lot of new jigs and fixtures and you need some new types of tools.

 

[Shortround6]

There was a prototype F6F-4 with a single stage R-2800-27:

Wayback Machine

Thank you,
The F6F-4 had about as much power running "normal"/max continuous as the 1700hp R-2600 did running at military power (5 min).
So the 3rd column might be closer to the performance of an R-2600 powered plane.
The F6F-4 was rebuilt out of the XF6F-1 after it belly landed in Aug 1942. It was later used as the prototype for the four 20mm gun installation.

I would also like to call attention to the much lower ceiling and times to climb to 20,000ft of this machine.

IN fact this R-2600 powered F6F would be within a few mph and 30 seconds of time to 20,000ft of a P-40E.

 

[don4331]

Navy was actually very nervous about P&Ws two stage engines. 3rd & 4th airframes were set aside for Wright R-1820 installations - designation XF4F-5 and another airframe set aside for R-1830-90 (two speed single stage) as XF4F-6, but becomes the F4F-3A in production. P&W does deliver on the the two stage though.
Maybe the R-2800 experimental aftercooler? https://www.enginehistory.org/Piston/P&W/Recips/R-2800_4.jpg

R-2000-9 gets its own supercharger (higher altitude) performance.

P&W R-2000-9 1,565 lbs. vs R-1830-90 1,495 lbs. vs R-1830-86 1,560 lbs. vs Wright R-1820-G205A 1,320 lbs. All weights dry.
Aside: G200 supercharger is ~65% efficient.

 

[don4331]

Few comments on superchargers.


You may very well be right about 37% but that seems awfully low. Efficiency of the supercharger is measured by the power it take to compress the air the desired amount by the power it takes to drive the compressor.

As in it takes 65hp to get the desired airflow and pressure but it takes 100hp to drive the compressor you have compressor operating at 65% efficiency. The problem is where does the other 35hp go? Only 2% or less goes into friction/drag in the supercharger gears. The rest is converted to heat in the intake mixture.

A 75% efficient supercharger would take about 87hp to get the same airflow/pressure (65hp required to do the work) as the less efficient supercharger leaving the engine with 13 more HP going to the crankshaft and a lot power turning into heat, just over 1/3 less. Now this is the extra heat over and above what is going to happen from a strict mathematical formula for the heat of compression.

This does not tell us anything about the airflow or pressure ratio of the supercharger, except that trying to get high boost from a poor supercharger is going to take a lot of power and very hot hot air.

The supercharger on the Kestrel V gave 1.5lbs boost at 14,000 but could give 6lbs at sea level or higher, 6lbs seems to be detonation limit the British were not going to exceed using their 87 octane fuel. The Kestrel V at altitude would run at 2900rpm and used an 8.8 gear ratio on the supercharger, however I have no information (others may have a lot of information) on the impeller size, number of blades, size/shape on the inlet and diffuser vanes/housing.
At any rate Kestrel V at 14,000ft was compressing the air about 1.88 times the standard atmospheric pressure (17.57 in hg/8.63lb/sq/in)
The Kestrel XV at 14,500ft running at 3000rpm and using a 9.4 supercharger gear could give 6lbs and was compressing the air about 2.45 times.
The Merlin III supercharger at 16,250ft had a pressure ratio of about 2.62 (no RAM as are the others). In the late 30s this was as good as it got.
The engine in the F4F-4 was limited to 7.66lbs of boost at 19,000ft.
The F4F may have been limited by it's cooling fins but basically the two stage R-1830 with intercooler roughly equaled the Merlin X/XII single stage supercharger.







Intercooler design is another ballgame. The needed airflow is highly dependent on the amount of cooling desired.
One simple chart says that an equal amount of cooling air to intake air to get a 40 degree drop in charge temp, you need 1.5 to 1 ratio to get just over 50 degree drop and a 2 to 1 ratio gets you just over a 60 degree drop.
The R-2800 used higher supercharger pressures but these were intercooled so we don't know what the pressures and temps were between the stages, We do know that the R-2800s were moving a lot more air through the intake system.

Initial centrifugal compressors were added to radials, because that nasty phenomenon - gravity - made the fuel fall to the lower cylinders making them run rich, while upper cylinders ran lean. Using a "mixer" ensured that all cylinders got equal share of air and fuel....and a little heat helped vaporize the fuel, so no one saw the drawback of the poor efficiency. But when you start wanting to beat the other guy by increasing the density ratio, efficiency matters.

I've actual done compressor efficiency testing in my career.

A couple assumptions - air is an ideal gas (pretty close actually), and adiabatic compression at some efficiency level, std pressure and temp at altitude e.g 14k' 17.57 in Hg, 468.8 °R (10.1 °F) (we need to work from absolute zero, so Rankin, not Fahrenheit)

Based on the 1.88 compression ratio (CR) a bunch of boring math says an ideal (adiabatic) supercharger would heat the air 91 °R [to 559.8 °R/101.1 °F). But the pre-improved supercharger was heating the air to 714.7 °R (255 °F).

The difference says the supercharger is 37% (I'm using that number from the RRHT book by Rubbra - a designer remembers).

We will note that while the CR was 1.88; the density ratio is only 1.23 (and it is the density ratio which matters)

After Rubbra and Ellor had increased compressor (supercharger efficiency) to 65%, the air was coming out over 100 °R cooler (608.8 °R) and density ratio improves to 1.44.

Note in both cases, the same CFM and pressure is coming out of the supercharger, so same power was going in. Just in the improved case, ~17.5% more fuel and air are going in as the mixture is cooler = 17.5% more power (and a lot less heat)

Aside: Kenneth Campbell improved the supercharger for Wright from the pathetic values GE was supplying in their compressors to 65% for the G200 series and 75% before he was finished - right inline with what Hooker did at RR.

 

[wuzak]

Grumman was, as shown by the list, working on a lot of other things than the F5F/XP-50.
They were already working on the F6F in fall of 1940.

Does spending time, money and effort on improving the F4F, beyond what they did historically, impact the development of the F6F?

 

[mack8]

Navy also wanted two stage superchargers. What they were expecting to torpedo at 30,000ft I have no idea

That made me laugh, that's so US Navy of the day lol.

Unfortunately some of these time tables require engines that either were never built or were at least contemplated and either didn't work or never made it through testing.

P&W had been working on their two stage supercharger since 1938, They skipped right over the 2 speed supercharger until sometime in 1940.
The R-2000 supercharger was sort of short of breath. How much was common to the R-1830 I don't know. But the R-1830s were good for 1200hp from sea level to 4900ft in low gear.
The R-2000s were good for 1350hp up to 2,000ft. The R-2000 weighed just about as much as an R-1830 with the 2 speed supercharger and at over 18-19,000ft the two stage engine would give better performance. Putting a similar supercharger on an R-2000 is going to add at least 100lb. Problem is that is just for the dry engine, Trying to put more power through tooth pick props doesn't work well.

Some of the Avenger problems were because of the Navy, why it took 6 months to order the two prototypes after the 7 month design selection process I don't know.
Flight trials began on 7th Aug 1941, Getting the navy to actually order the Prototypes even 3 months earlier might have meant more than 6 Avengers at Midway (still would not have fixed the torpedoes)
Navy also wanted two stage superchargers. What they were expecting to torpedo at 30,000ft I have no idea. Now keep this in mind, Wright announced, but only delivered about 2 two stage R-2600 engines. Which did not work very well and soon disappeared never to be heard from again.

Wright R-2600 engines,
1600hp for take-off. Available in large numbers in 1940-41.
1700hp for take-off, none in 1940, 443 built in 1941, 206 of them in Dec, Production in 1942 is rather sketchy in early part of the year, it takes until June to beat the Dec total, In the worst month of 1942 (Feb) they only built 60 engines between two plants.
1900hp for take-off. Does not show up in double digit numbers until July of 1943.

Now for the reality check.
The 1700hp T-O rating R-2600 was good for 1450hp at 14,100ft no ram.
The engine in the F6F-3 was good for 1800hp at 15,500ft but with ram could hold to 18,700ft
In a British test they got 371mph in low gear at 18,700ft.

I doubt that at 16,000-17,000ft the R-2600 powered Hellcat was going to make 360mph.

What the two stage supercharged R-2600 might have done I don't know, they yanked that engine out of the F6F-1 prototype in about one month.

Thanks for the details. Was looking into the XF4F-8, any info as to when work started on it? Weight was down about 500 lb, what/how did they lightened it? Obviously the engine is not ready in 1942, but for this TL the light airframe must be ready sooner, and powered by the regular 1830-86 by necessity. The folding F4F-4 flew in spring 1941 while the light airframe in late 1942, so to get the light airframe ready earlier something has to give, OTL they were working on the rather useless F4F-7, so perhaps instead of that they focus on weight reduction since the F4F-4 is so overweight. Perhaps it can be done in stages so as not to disrupt production, let's say they manage to pare down say 200-250 lb by late 1941, so this will be an F4F-4A, then by summer another 200-250 lb, so this will be an F4F-4B. coupled with small improvements such as jet exhausts, this would at least restore the speed /climb to the F4F-3 level, a welcome boost.

Another way is to order the F4F-4 with the non-folding wings, and meanwhile embark in a top priority program to get the weight down as above. I think the US carriers SHOULD still be able to carry 27 non-folding F4Fs at Midway (due to use of deckparks etc.- though i'd welcome comments on the matter) but can't carry 36 like at Guadalcanal, so the lightened folding wing version must be ready by then.

A third way, how about the light airframe coupled with the non-folding wing? This would offer the highest performance, probably FM-2 performance in 1942, but again, the carriers can't carry more than 27 (and Wasp probably less), or something else has to give (a few less SBDs, TBFs?)

 

[EwenS]

While the prototype XF4F-4 flew on 15 April 1941 and was handed over to the USN on 14 May 1941, there were delays productionising it. The earliest folding wing sets went to aircraft destined to become RN Martlet II produced from about Aug 1941 (RN acceptances in the USA from beginning of Oct). The RN delayed delivery of 90 aircraft just to get the folding wing.

The first F4F-4 destined for the USN flew on 7 Nov 1941. (Francillon's "Grumman Aircraft since 1929").

As for the XF4F-8 it was ordered in July 1942 "after Grumman had studied ways to tailor the Wildcat for operations from the small deck of escort carriers". Francillon's again. First flight 8 Nov 1942.

The USN AVG-1/ACV-1/CVE-1 Long Island was commissioned on 2 June 1941 at which point she had a 362ft flight deck suited to her intended air group of Curtiss SBC Helldiver scout-bombers and SOC Seagull floatplanes (VS-201 was formed to operate from her). There is a photo of her with a couple of F2A Buffalo fighters on her flight deck around July. The short flight deck proved unsuccessful and was extended in Oct 1941.

At that point the USN was looking to the British escort carrier experience and in Oct 1941 the US naval attaché in Britain reported on the experience of HMS Audacity on the Gibraltar run with an air group of 6 fixed wing Martlet I. (Audacity only completed her conversion in June, landed her first Martlet on 10 July and escorted her first convoy in Sept thereby proving the concept of the escort carrier with a fighter complement as being far superior to the CAM ships).

This period coincides with the redesign of BAVG-6 that became HMS Tracker that served as the prototype of the Bogue class. The 20 Bogue class were not ordered until Jan 1942 with the early ships being conversions of merchant ships already under construction.

So if Francillon is correct with his comment, the driver for Grumman to begin looking at what was to become the XF4F-8, with its reduced weight, designed to operate off escort carriers doesn't arise until the Nov 1941-Jan 1942 time period.

Yet another example of how the design of carriers and their aircraft interact.

 

[mack8]

That's very interesting, so perhaps there is scope for the USN to ask say summer 1941 "can you find ways to reduce the weight of the F4F-4 since speed, handling and climb dropped significantly". so then Grumman looks into it and then builds a prototype. Late 1941 or early 1942 maybe? Then production in spring/summer? (roughly looking at the timeline of F4F-4 production as well as folding Martlets)

 

[Geoffrey Sinclair]

The two XF4F-8 were allocated the final two BuNos from contract 99340 dated 16 May 1942 originally for 575 F4F-4, both XF4F-8 accepted December 1942 "change No. G. Development of F4F-4 with improved slotted flaps" and Wright R-1820-56 versus the XF4F-5 R-1820-40.

The folding Wing Martlet II/F4F-4 Split was
Jun-41 1 / 0
Jul-41 0 / 0
Aug-41 2 / 0
Sep-41 3 / 0
Oct-41 6 / 0
Nov-41 14 / 1
Dec-41 24 / 4
Jan-42 0 / 62
Feb-42 1 / 79
Mar-42 33 / 56
Apr-42 6 / 92

The initial batch of F4F-3 production ended in September 1941, apart from the above the period October 1941 to April 1942 saw the official acceptance of 1 F4F-4B in February and 1 F4F-7 each in January and March 1942.

 

[tomo pauk]

I am just not seeing 30mph worth of aerodynamic "tweaks" to the F4F.

It is more the area of 2-3mph per tweak, not 5 mph.

We can see here the NACA report from late 1940 (tests were made in Spring of 1940) where the XP-41 - firmly in class of the F4F speed-wise - gained far more than 2-3 mph with better exhaust stacks fited. Granted, this is a propulsion 'tweak', not an aerodynamic one, but still.
2-stage engine on the XP-41. Gain of 18 mph at 20000 ft:

 

[don4331]

Was looking into the XF4F-8, any info as to when work started on it? Weight was down about 500 lb, what/how did they lightened it?

Note: The 500 lbs. was gross weight, not empty.

230 lbs. is the weight reduction of the experimental Wright R-1820 compared to the std P&W R-1830-86.
20 lbs. is the moving/consolidation the wing to cowl oil cooler, and the redesign of cowl.
130 lbs. is the removal of outer 2 guns (back to 4 gun wing)
The rest is the removal of ammunition for said guns. (inner guns retain the 240 rpg of the F4F-4, not the 450 rpg of F4F-3))

Other than rebalancing the airframe to account for the lighter engine, Grumman didn't do much.* (The slotted flap wings weren't a success and were replaced with regular split ones).

* The XR-1830-56 in addition to being lighter made 150hp more. Grumman increased fin and rudder size to account for this.

 

[Wild_Bill_Kelso]

The prop on the FM-2 went 394lb.
The prop on the F4F-3/F4F-4 went about 315lbs, individual props could vary a few pounds.
You need enough prop to handle the power, too much prop just adds weight and drag.

So fine tune the prop a bit. Certainly you see much skinnier props in the early war than later.

I feel like I am repeating things a lot.

Grumman or Eastern Aircraft CANNOT get the FM-2 into production until Wright is making the 1300hp version of the R-1820 engine in quantity. like 200-300 a month.
Having a couple of tool room samples for testing does not mean you are ready to go into production.
Wright needed to figure out what they needed to get 1300hp out of the R-1820 and THEN they needed to figure out how to make the engine AND get the new tooling and factory space for it.

Once again, the 1300hp version (the -56) used

Different cylinder heads than the older versions,
Different cylinders with a totally new way to make the fins
A different crankcase
Different crankshaft.

assorted other changes. Next to NO parts on the 1200hp version would work on the 1300hp version.
You need a lot of new jigs and fixtures and you need some new types of tools.

Ok I get that, but if they had a working prototype that early, with more money, effort, more companies involved, maybe with a bit of prioritization you could in fact have some FM-2s available a year or more earlier. I know then getting them out to the units would take some doing. And some luck, probably.

But considering the quite small number of fighters involved in some very key battles, I think they could have made quite a difference. Even if you had say, 100 aircraft.

With their good low altitude performance, FM-2s would be particularly good at operations against islands etc. In fact based on the stats RCAFSon posted, it looks like FM-2 has much better performance (especially climb) all the way up to ~ 20,000 ft unless I'm missing something. So it would be much better for CAP as well compared to an F4F-4.

Some other firms which made planes for the US, which may have had some capacity, for reference

Bell (P-39s, P-63)
GM - FM Wildcat, TBM
Goodyear (Corsairs / FG-1)
Ford (B-24s)
Vultee (Valiant, Vengeance)
Northrop (Black widow)
Martin (B-26 and B-29)

I'm not going to even include Brewster because they were having so many problems. I figure Boeing, Lockheed, Consolidated, Douglas, Curtiss-Wright, North American, Republic and Grumman have their hands full.

 

[Wild_Bill_Kelso]

Thank you,
The F6F-4 had about as much power running "normal"/max continuous as the 1700hp R-2600 did running at military power (5 min).
So the 3rd column might be closer to the performance of an R-2600 powered plane.
The F6F-4 was rebuilt out of the XF6F-1 after it belly landed in Aug 1942. It was later used as the prototype for the four 20mm gun installation.

I would also like to call attention to the much lower ceiling and times to climb to 20,000ft of this machine.

IN fact this R-2600 powered F6F would be within a few mph and 30 seconds of time to 20,000ft of a P-40E.

The early R-2800 powered F6F also had a fairly low climb rate, in fact it was worse

 

[Wild_Bill_Kelso]

Does spending time, money and effort on improving the F4F, beyond what they did historically, impact the development of the F6F?

Part of what Tomo and Shortround were debating was could they have taken resources / time from failed projects like F5F and put it toward improved F4F instead

 

[Wild_Bill_Kelso]

What is the "normal" time delay between prototype and factory production, if we can say there is such a thing? Or what is the normal range

 

[mack8]

Care to post a picture of that aircraft?

This is what i have on the XF4F-1, from here:

Grumman F4F Wildcat

PS: Looking at the fuselage, looks like Grumman went backward, at least the rear fuselage and tail are awful on the XF4F-2 and the early XF4F-3. It lost it's bubble canopy too.

 

[mack8]

Note: The 500 lbs. was gross weight, not empty.

230 lbs. is the weight reduction of the experimental Wright R-1820 compared to the std P&W R-1830-86.
20 lbs. is the moving/consolidation the wing to cowl oil cooler, and the redesign of cowl.
130 lbs. is the removal of outer 2 guns (back to 4 gun wing)
The rest is the removal of ammunition for said guns. (inner guns retain the 240 rpg of the F4F-4, not the 450 rpg of F4F-3))

Other than rebalancing the airframe to account for the lighter engine, Grumman didn't do much.* (The slotted flap wings weren't a success and were replaced with regular split ones).

* The XR-1830-56 in addition to being lighter made 150hp more. Grumman increased fin and rudder size to account for this.

Thanks for the details, didn't realized there were no structural mods like lightened fuselage or lightened folding wings.

So what remains are the 4 gun wing, and maybe jet exhausts, and maybe a slightly refined cowling if at all. I've also read that the FM-2 carried a bit less fuel, don't know if that would be an option (would be compensated by carrying slipper/drop tanks when required).

Back to getting the R-2000 earlier then, for a bit more power, between the about 200 lb weight reduction possible, but also the 300 lb heavier engine, the end result would be a tad heavier than the OTL F4F-4, but with better power to weight ratio.