American purchase programs, alternatives and reality, 1937-43

[tomo pauk]

Unless the US can figure out how to make lighter fighters they need to build really big engines.

I see this from the opposite direction: make a fighter around as powerful, big & heavy engine as it is available, and presto. If a good small engine appears, up-engine the legacy fighter design.

Now the problem is getting engines that make big power at 20,000ft, not big power at 1-3,000ft.
Next problem is fitting in enough fuel for the really big engines.

A fighter designed around a big radial will be bigger than the P-36 or P-35, meaning that there is a lot of room for the big fuel tanks.
Desire for a lot of power at 20000 ft can be fulfilled with the additional compressor stage, be that stage powered by the exhausts or directly by the engine. Everyone in the business was aware of 2-stage superchargers and the advantages of these by the mid/late 1930s.

Perhaps Americans should have done what has often been suggested in the various what-if threads about the European powers, that is, instead of tripping over themselves with various exotic concepts (Hyper program etc), make a big (~40L?) V-12?

Hear, hear.

 

[Shortround6]

Perhaps Americans should have done what has often been suggested in the various what-if threads about the European powers, that is, instead of tripping over themselves with various exotic concepts (Hyper program etc), make a big (~40L?) V-12?

Starting when (month/year) and with what goal/s in mind?

The Continental IV-1430 and Lycoming O-1230 are obvious choices for cancelation.
The Lycoming was trying to get 1200hp from it's 20.2 liters by running at 3400rpm which was doable considering it's short stroke. Wisdom of doing it might be another story.
There are problems with making larger V-12s.
Packard had been making a 41.6 liter engine in the 1920s but the head designer died in early 30s. It was partially responsible for the move to Air cooled engines in the US in the early/mid 30s . It was also resurrected as the widely built PT boat engine of WW II. It would have made a lousy aircraft engine in WW II. 1200-1350hp at sea level is not the same as 1200-1350hp at 15,000ft.

There were several large V-12s from WW I or from the 1920s but they just showed what not to do.

Problems in 1937-40 for the US (and others) include fuel but just saying that limits boost does not look at the full picture. High octane (PN) fuel is also needed for other things. A cylinder with local hot spots needs to use lower boost than a cylinder with fewer or smaller hot spots that can create 'knock'. Large cylinders tend to run a bit hotter, cooling larger diameter pistons is a bit harder.
Large cylinders restrict rpm. Large cylinders have problems breathing, that is getting the air in and the exhaust gasses out. Maybe only a few percent and maybe it can be solved other ways.
A large engine is easier to get power out of than a small engine but sometimes it does not scale well. You are not only looking for power per cu in (L) but power per LB/KG.
Depending on the state of the art at particular times (very important here) in BOTH materials and knowledge these metrics can shift.

A lot of engine designs start with single cylinder engines. How much power can they get from one cylinder and then how many cylinders can be used together to make the desire power.

The whole fuel thing really confused things. Using the PN scale (performance number) fuel went from 68.3 PN (87 octane) up to 100 PN (100 octane) to 100/130 ( 130 PN rich) from mid 1930s to 1941/42. allowable cylinder head temperatures and boost levels went to unimaginable levels to 1935-36 engine designers.
Bigger is generally better but the Pistons, rods, crankshafts and cylinder blocks/crankcases need to stand up to the strain. And big cylinders impose more strain at a given moment than small cylinders. Better oils and bearings can help. Better con rods can help. Better pistons and/or using oil sprays on the piston bottoms can help.

And then we can argue about when the Air cooled engines started to catch up the liquid cooled engines due to better finning, better cowling and better exhausts.

 

[tomo pauk]

Starting when (month/year) and with what goal/s in mind?

Seeing the RR 'R' trashing competition in the early 1930s should've been a cue.
Making a lot of power is a legit goal, no?

 

[YavorD]

Concerning radial engines, and two-stage superchargers, Pratt & Whitney had a head start and a bit of experience with the Navy programs.
The one and only XP-41 was developed in parallel with the XP-38/39/40/42. It was a modern airframe, powered with a R-1830 engine, two-stage mechanically-driven supercharger and an intercooler. The aircraft was flying during early 1939 but spent most of the time tested in a full-scale NACA wind tunnel. The USAAC was not interested. The powerplant became well known as developed for the Navy to power F4F-3/4 (FM-1).

 

[Shortround6]

Seeing the RR 'R' trashing competition in the early 1930s should've been a cue.
Making a lot of power is a legit goal, no?

Making power is legit.
Some people of the time had not gotten out of the Bomber engine vs fighter engine mentality.

The "R"s predecessor was the Buzzard which used the same bore and stroke as the R and the later Griffon.
It used 77 octane fuel not 87 and maxed out at 2300rpm. It also made max power at 2,000ft and rated power at 5,000ft (?). Weight is in question as figures in Lumsden do not agree with figures in Wiki which is supposed to from Lumsden?
The R was a phenomenal engine, Most (all?) did not run on gasoline. Most ran at sea level. most had short service lives. They could set a goal post. How rivals got there was another thing.
The Buzzard was used in some single engine torpedo bombers and in a few (10-15?) flying boats.
It is a hard path for a company to follow to make high powered engines if it's customers (government for one) wants to buy (pay for) lower powered engines and not the higher power engines. An early Griffon Spitfire would have been pretty spectacular at first glance. Using a giant two blade fixed pitch prop and tall landing gear and giant wing to allow for the required landing speed with the heavy engine would have been a lot less spectacular
2-3 years can make a difference in what the customers what. Unfortunately the customers never figure in the design/development time needed for increasingly complicated engines/propellers and airframes.

Designing large engines and airframes that can do 400mph + in 1939-40 means a lot of people finding out about compressibility at about the same time and not in a good way.

In Britian they were trying to jump from the 1030hp Merlin (and some minor improvements with 87 octane fuel) right over the 1400-1600hp range to the 1800hp Vulture and the 2000hp + Sabre and Centaurus engines.. Going to 100 octane fuel allowed all engines to be improved.
US jumped to improved fuel a bit earlier than the British but only around 1/2 way. US 100 octane was better than 87 octane but not as good as British 100 octane (100/120?)

Unfortunately for the US none of the Air cooled radials were allowed any increase in manifold pressure going from 100 to 100/130 fuel. Rich mixture did a lot less for US engines than for British engines leaving the US engine makers to look for other ways of making power, like high rpm in the liquid cooled engines and/or lots of cylinders.
Chrysler was late to party with the Chrysler IV-2220 16 cylinder but I don't know how far back the design studies went. Government contract was dated June 22th 1940. A rather conflicted engine with 16 smallish cylinders (Allison or Merlin sized) but only two valves in each cylinder but running at 3400rpm. Type of fuel they were planning on using in 1940?
Also planning on using two stage supercharging with a turbo for altitude performance. 2500hp for 2500lbs is not bad but the weight may not include the turbo and the intercooler/s so it is a bit deceptive. And the engine itself was very, very long compared to somewhat pedestrian Griffon and the weight gave some people pause?

Getting back to the Wright R-2600. It was first used in the very large Boeing Clipper flying boats. One of the first land planes was the Curtiss C-20W

and the B-23 bomber.

A-20s moved the air intakes to the top of nacelle.
In 1939/40 the US had 3 different fighters running around with extended prop-shafts and long pointed noses hiding R-1830 engines trying to figure out how to reduce drag on radial engines. The concept did not work even though they learned a lot. But what they learned was not available for production aircraft in 1939/40.

 

[Admiral Beez]

Okay, what to do with the people that can make everything and had no fear of 'land aggression'

Alternative source for the torpedo design and production? Should put some fire under the Rhode Island facility, too.

Buying some military kit from Italy might keep Mussolini out of the war. I'm not sure what the US could possibly want though.

One area of Italian expertise, rather than tangible goods that might be wanted is engine and aerodynamic research. Italian aircraft were cleaner aerodynamically. So, maybe the US buys Italian wind-tunnel tech. As for engines, it would be interesting to see what US firms can do with a Isotta-Fraschini Asso compact V-12. And then there's Italian submarine torpedoes, which may be interesting if the USN knows their own fish are crap. Then there Ford Italia S.A., established in 1923 and churning out some interesting trucks.

 

[z42]

Starting when (month/year) and with what goal/s in mind?

Goal to develop a more powerful follow up to the V-1710 (not necessarily by Allison). And with the storm clouds gathering, not try for anything too exotic.

And yes, there are issues with scaling up a design and you'll likely end up with a lower power per volume than the smaller engine. Still, not insurmountable as can be seen eg with Griffon vs Merlin.

 

[yulzari]

Making power is legit.
Some people of the time had not gotten out of the Bomber engine vs fighter engine mentality.

The "R"s predecessor was the Buzzard which used the same bore and stroke as the R and the later Griffon.
It used 77 octane fuel not 87 and maxed out at 2300rpm. It also made max power at 2,000ft and rated power at 5,000ft (?). Weight is in question as figures in Lumsden do not agree with figures in Wiki which is supposed to from Lumsden?
The R was a phenomenal engine, Most (all?) did not run on gasoline. Most ran at sea level. most had short service lives. They could set a goal post. How rivals got there was another thing.
The Buzzard was used in some single engine torpedo bombers and in a few (10-15?) flying boats.
It is a hard path for a company to follow to make high powered engines if it's customers (government for one) wants to buy (pay for) lower powered engines and not the higher power engines. An early Griffon Spitfire would have been pretty spectacular at first glance. Using a giant two blade fixed pitch prop and tall landing gear and giant wing to allow for the required landing speed with the heavy engine would have been a lot less spectacular
2-3 years can make a difference in what the customers what. Unfortunately the customers never figure in the design/development time needed for increasingly complicated engines/propellers and airframes.

Designing large engines and airframes that can do 400mph + in 1939-40 means a lot of people finding out about compressibility at about the same time and not in a good way.

In Britian they were trying to jump from the 1030hp Merlin (and some minor improvements with 87 octane fuel) right over the 1400-1600hp range to the 1800hp Vulture and the 2000hp + Sabre and Centaurus engines.. Going to 100 octane fuel allowed all engines to be improved.
US jumped to improved fuel a bit earlier than the British but only around 1/2 way. US 100 octane was better than 87 octane but not as good as British 100 octane (100/120?)

Unfortunately for the US none of the Air cooled radials were allowed any increase in manifold pressure going from 100 to 100/130 fuel. Rich mixture did a lot less for US engines than for British engines leaving the US engine makers to look for other ways of making power, like high rpm in the liquid cooled engines and/or lots of cylinders.
Chrysler was late to party with the Chrysler IV-2220 16 cylinder but I don't know how far back the design studies went. Government contract was dated June 22th 1940. A rather conflicted engine with 16 smallish cylinders (Allison or Merlin sized) but only two valves in each cylinder but running at 3400rpm. Type of fuel they were planning on using in 1940?
Also planning on using two stage supercharging with a turbo for altitude performance. 2500hp for 2500lbs is not bad but the weight may not include the turbo and the intercooler/s so it is a bit deceptive. And the engine itself was very, very long compared to somewhat pedestrian Griffon and the weight gave some people pause?

Getting back to the Wright R-2600. It was first used in the very large Boeing Clipper flying boats. One of the first land planes was the Curtiss C-20W
View attachment 866428
and the B-23 bomber.
View attachment 866429
A-20s moved the air intakes to the top of nacelle.
In 1939/40 the US had 3 different fighters running around with extended prop-shafts and long pointed noses hiding R-1830 engines trying to figure out how to reduce drag on radial engines. The concept did not work even though they learned a lot. But what they learned was not available for production aircraft in 1939/40.

The bigger wing area to get the heavier ATL engines off the ground from existing airfields and knock on effect upon performance displays the lateral thinking effect on performance from the massive UK modern hard runway new airfields program which was one of the nation's biggest capital programs. You can see the same when you look at the demand for airfields in NW Europe in 1944/45 which crowded every potential airfield yet still had to have some operations based in the UK. Airfields which were feasible in 1939/40 for the RAF in France ( and some of those were only fit for Gladiators at first) were not appropriate for 1944/45 aeroplanes being too short and too soft for the heavy modern ones.

It is interesting that Chrysler were going down the Halford/Napier road of more but smaller cylinders and running at higher revs.

 

[Shortround6]

The bigger wing area to get the heavier ATL engines off the ground from existing airfields and knock on effect upon performance............

it is not only getting off the ground (power and props can do a lot) but getting back on the ground
Flaps, while advancing fast, are sort of evolving at this time and low wing loading was often considered important to low landing speeds?
Low landing speed also meant smaller brakes or less brake wear and perhaps less nose overs.
Combat tended to change priorities.

 

[Shortround6]

Goal to develop a more powerful follow up to the V-1710 (not necessarily by Allison). And with the storm clouds gathering, not try for anything too exotic.

Historically the AAC thought they had that covered with Continental IV-1430 with a hoped for (but unproven) 1600hp at 25,000ft (with turbo).
Allison may have been offering 1325hp with the engines in P-38F but that was 1-2 years away, what were they offering in 1940? Most engine makers were offering engines 1-2 years before actual delivery dates.

In reality the V-1410 never came close in an actual aircraft and the engines in the P-38F only made 1325hp at 15,000ft and needed better than US 1940 fuel to do even that due to the intercooler mismatch.

The other liquid cooled engines of 1939-40 were 2000+ HP engines aside from the Lycoming which was too small/too late. They needed high rpm/high displacement because they could not use high manifold pressure with the US 1940 100 octane (under 2% aromatic) fuel.

The R-1830s used 9lbs boost or under (48in). The 1600hp R-2600 used about 6lbs or under and the 1700hp version used about 7lbs.
R-2800s used a max of 12lbs and that needed the two-stage supercharger and inter-cooler/s at altitude.

 

[Elan Vital]

On the topic of tanks, there is a lot to unpack in this period.

The decision to favor the front-drive, VVSS suspension-equipped Combat Car M1, Light Tank M2 and extrapolated Medium Tank M2 over the rear-drive, Christie suspension concept of the Combat Car T4 precedes 1937, so the opportunity to fundamentally alter the layout of early war tanks was gone. The necessity to expand tank production and equip both the US and its allies meant that certain fundamental changes were not possible for the early war tanks (Stuart and Sherman families).

1941-1943 coincides with the start of development of more clean-sheet vehicles (Light Tank T7 then Medium M7, T67/70 then M18 GMC, T2X Medium tanks) and here there is more leeway to change those as there is less urgency to keep existing layouts or parts. 1937-1941 still matters in that some important component development can start then and give more options for 1941-43.

In 1936, the US did not renounce all development of the Christie suspension, but the resulting T7 Combat Car was IMO doomed to fail as it didn't try to to implement that suspension in a different way, so was unlikely to be any more appealing than the T4 Combat Car. As in, the the T7's suspension was still internal and still convertible when external or non-convertible Christie suspensions did exist or were though of at the time. Internal meant the space considerations still mattered, convertible meant it was still more expensive, more complex and possibly heavier and bulkier than non-convertible.
Additionally, the US did patent torsion bar suspensions since about 1933 and some officers had pushed since 1934 for it to be tested. IMO, the T7 would have been a much more useful testbed if it had either tested a non-convertible (and maybe external) Christie suspension, or tested a completely new concept like torsion bars. The latter would have been a big boon later on as development of torsion bars for US tanks didn't start until about 1942 historically.

Regarding engines, the matter is complex. The Wright gasoline radials were the baseline as they were powerful, light and more available than other types of high-power gas engines. The US otherwise spent most of the 1937-41 period testing diesel engines:
- the Guiberson radials, which unfortunately probably required a major redesign that Buda/Guiberson wasn't interested to undertake. IMO, the US should either have pulled a Wright/Continental and reassigned military development of the Guibersons to someone else (say Caterpillar which developped diesel radials OTL?), or maybe request development of a dieselized R-975 (or the P&W counterpart), since IIRC that was briefly studied in WW2 and may have the benefit of greater commonality with the gas R-975.

- the GM diesels, first the 6-71 in about 1938-39, then the 4V-223 (a carefully designed unified powerplant) tested in 1940, and from April 1941 onwards the 6V-184 was developped based on lessons of the 4V-223, but on a more modern and efficient 184 cu in cylinder from the 16-184 naval engine. The 6V-184 obtained 450 hp and, as a dedicated Vee tank engine, was more volume and weight-efficient than the twin 6-71 unit developped from August 1941 due to the greater production capacity for 6-71 parts. Short of being bolder with the 6V-184 design start date (the 16-184 was in development since 1938), the compromise path may be to design a 12V-71 (and ideally a 6V or 8V-71 for light tanks) instead to use the more common parts while still using a more efficient layout than the twin 6-71. Twin 12V-71s would still be just as good as the Quad 6-71 for landing crafts.

- a bunch of other commercial diesels for light tanks (twin Hercules, Cummins)

Obtaining a Vee tank engine prior to the Ford V8/V12 would require an earlier anticipation of the tank engine production requirements, or just a desire to obtain a more space-efficient engine than the radials (which is probably what motivated the 6V-184 which sorta precedes the tank engine panic).
All post-1941 radials and non-Ford Vee engines run into the problem of adding too many engine models to the US tank program. They basically need to have the same or better timing than the historical Stuart and Sherman powerplants.

Regarding other automotive aspects, 1934 notes on testing of the T4 Combat Car recommended research on epicyclic automatic gearboxes, fluid couplings, torque-converter transmissions and IIRC Hydramatic-style transmissions. It would be nice to test those prior to 1941 (maybe as part of the T7 Combat Car testbed?).
In 1938-39, the US got to test transfer cases to lower the driveshafts when using radial engines (to either lower the entire tank or free up more room in the fighting compartment), but this was rejected for 1940 production M2 Light Tanks as "they would have delayed production too much". However, for 1941-42 M3 Stuarts or M3/M4 Mediums, maybe they could have been available in time?

Other than that, I'm not sure it was realistic to significantly change either the Stuarts, M3 and M4 Mediums, short of the US making major changes in the 1937-40 period.

The 1941-43 period has somewhat missed opportunities. The T7 Light suffered from scope-creep to medium tank class, but I think it could have been steered into a less efficient but earlier M24 Light if the armor had been kept light and the engines standardized to twin-Cadillacs instead of the R-975.
The M18 GMC arguably suffered from the fundamental doctrinal choices of the Tank Destroyer class, but could possibly have been steered into a decent late-war light tank (with rear drive and an enclosed turret), sorta M24 but with a 76 and more powerful engine, leaning closer to the 1945 lessons that led to the M41 Light Tank.

The T2X Mediums suffered both from the initial preference towards the electric transmission, and the fact they retained too many elements compatible with the M4 so always competed with the latter for production priority. If the tank has to be different enough from the M4 to justify, might as well include the 90mm armament in the gun options immediately in 1942 instead of requesting it in 1943 and design around the 750 hp Ford V12 (greater mobility or armor potential), or the 600 hp 8V-184 diesel.

 

[Shortround6]

On tanks there is a lot the US could have done better,
But the US also, part to a late start, didn't do a lot really wrong either.
The M2A4 was a 12-13 ton tank with a decent 37mm gun and 250hp engine that gave it a 36mph top speed. Armor was thin but was quicky improved in the M3 and M5.
The US was not stuck with numbers of lighter 2 and 3 man tanks with really poor guns and slow speed.
US also had the M2 Medium which was about 20 tons but large in hull volume. Since the US had started with a 340hp/400hp engine upgrading the upper hull, turret, and armament did not call for an entirely new tank. Some of this may have been luck and some of it is good engineers that figured out good up grades. By the time you get to the M4E8 you are dealing with a 37 ton tank with very few common parts to the M2 Medium. But changes did not require tooling up for a completely new tank.
US doctrine was also at fault as tanks, according to doctrine, were not supposed to fight tanks. That was the tank destroyers job.
US was using a lot of it's tanks as infantry support tanks. For this they worked quite well as their short 75mm guns fired a decent HE shell (the US 76mm gun did not) and the 75mm armed tanks carried a very large quantity of ammo. 90-104 rounds depending on model and around 6,000 rounds of MG ammo.
The need for a 90mm gun didn't show up until late and while anticipated not enough emphasis had been used to get a decent quantity into service.
It turned out in Korea that the 76mm armed Shermans could handle the 85mm armed T-34s without much disadvantage. In part to due to better training, and a much faster firing rate.
T-34s were not as well protected as Panther and the 76mm guns could handle the T-34s fairly well.
Yes the M-26 needed a better engine but the M-26 was not needed to win the war.

 

[Reluctant Poster]

- the GM diesels, first the 6-71 in about 1938-39, then the 4V-223 (a carefully designed unified powerplant) tested in 1940, and from April 1941 onwards the 6V-184 was developped based on lessons of the 4V-223, but on a more modern and efficient 184 cu in cylinder from the 16-184 naval engine. The 6V-184 obtained 450 hp and, as a dedicated Vee tank engine, was more volume and weight-efficient than the twin 6-71 unit developped from August 1941 due to the greater production capacity for 6-71 parts. Short of being bolder with the 6V-184 design start date (the 16-184 was in development since 1938), the compromise path may be to design a 12V-71 (and ideally a 6V or 8V-71 for light tanks) instead to use the more common parts while still using a more efficient layout than the twin 6-71. Twin 12V-71s would still be just as good as the Quad 6-71 for landing crafts.

I have always been curious about the unusual diesels developed by GM. Do you have the specs for the 4V-223 and 6V-184?

 

[Elan Vital]

I have always been curious about the unusual diesels developed by GM. Do you have the specs for the 4V-223 and 6V-184?

I have some figures:

4V-223 (223 cu in cylinder displacement, 892 total)
Bore x stroke: 6 3/8 in x 7 in
Maximum gross output: 250 bhp at 1400 rpm
Specific weight: 7 lb/hp
Weight: 1750 lb

6V-184 (1104 cu in total)
Bore x stroke: 6 in x 6 1/2 in
Maximum gross output: 450 bhp at 1800 rpm
Specific weight: 6.7 lb/hp
Weight: 3015 lb

Both of these powerplants included a self-contained oil sump, radiators, gear-driven fans and air filters, and used welded steel construction.

Twin 6-71 (GM 6046) (852 cu in total)
Bore x stroke: 4.25 in x 5 in
Maximum gross output: 410 bhp at 2100 rpm
Weight: 4340 lb powerplant, 4855 total with radiators, exhaust mufflers, fan shrouds, driveshaft, voltage and current regulator and filter panel added

I don't know the relative dimensions of either engine.

 

[Reluctant Poster]

Historically the AAC thought they had that covered with Continental IV-1430 with a hoped for (but unproven) 1600hp at 25,000ft (with turbo).
Allison may have been offering 1325hp with the engines in P-38F but that was 1-2 years away, what were they offering in 1940? Most engine makers were offering engines 1-2 years before actual delivery dates.

In reality the V-1410 never came close in an actual aircraft and the engines in the P-38F only made 1325hp at 15,000ft and needed better than US 1940 fuel to do even that due to the intercooler mismatch.

The other liquid cooled engines of 1939-40 were 2000+ HP engines aside from the Lycoming which was too small/too late. They needed high rpm/high displacement because they could not use high manifold pressure with the US 1940 100 octane (under 2% aromatic) fuel.

The R-1830s used 9lbs boost or under (48in). The 1600hp R-2600 used about 6lbs or under and the 1700hp version used about 7lbs.
R-2800s used a max of 12lbs and that needed the two-stage supercharger and inter-cooler/s at altitude.

An interesting appraisal of the "obsolescent" Merlin by the Office of Production Management. Better to concentrate on the Allison, Continental, Lycoming and Wright Tornado engines. Lots of interesting reading in the attached file on the history of the Packard Merlin.

 

[Shortround6]

Don't send any of these guys to the horse track with any of your money. 1 2nd place out of 4 entries are not good odds.

 

[Reluctant Poster]

An interesting appraisal of the "obsolescent" Merlin by the Office of Production Management. Better to concentrate on the Allison, Continental, Lycoming and Wright Tornado engines. Lots of interesting reading in the attached file on the history of the Packard Merlin.

View attachment 870228