MOST OVERRATED AIRCRAFT OF WWII (2 Viewers)

[Shortround6]

Every B series engine produced before 1943 probably ended up on the scrap heap before ever seeing combat, along with their early variant aircraft.

You do of course have some sort of proof of that or reference?

P & W built 4,300 B series engines (single stage) in 1942. Ford built 5,711 in 1942 so you are saying that over 10,000 engines ended up being scrapped along with their aircraft before seeing combat in the middle of WW II and nobody knows about it ???

WOW!!!!

I wonder if the B-26 crews that flew at Midway knew their planes had been scrapped before they attacked the Japanese fleet?
Likewise the B-26 crews in the Aleutian Islands in 1942 or the RAF crews in No 14 Squadron or the crews of the 22nd and 38th bomb groups operating out of Australia in early 1942.
A lot of crews flying around in imaginary aircraft if you are correct.

Not mention what happened to hundreds of P-47Cs? Just what were those P-47 Groups in England flying in the Spring of 1943?
In Feb of 1943 Republic delivered it's 602 and last P-47C and rolled out it's first P-47D-1. But I guess all the R-2800s built in 1942 were scrapped so those fighters were re-engined with 1943 production delivered by submarine to the cargo ships transporting the fighters to Europe with crews working under tarps by flashlight so the merchant seamen wouldn't find out?

This is a look at the severe problems associated with R-2800 development.
http://www.enginehistory.org/NoShortDays/Development of the R-2800 Crankshaft.pdf

http://www.enginehistory.org/NoShortDays/Development of the R-2800 Crankshaft.pdf

An interesting document but doesn't seem to make any mention of mass scraping of engines???
It also makes very little reference to the practical problems they were trying to solve. Like time between failures or time between overhauls. New Crankshaft and dampers that allow 5-600hrs between overhauls vs 300-350 hours???
I don't know the actual numbers but a lot of that document is reference to the later C and CA engines and post war commercial engines.
Maybe I missed it. please give a page number or section in that document that says the Early R-2800s were failing at a high rate (aside from the Ford built ones failing due to being improperly cleaned of casting sand)

Comparing Packard building a developed design versus Ford building a design in development is not a straightforward look at numbers of engines built. Additionally the 18 cylinder R-2800 was a very complex design.

The R-2800 was complex and Ford did do a tremendous job but so far you have failed to show they could have cut months off the delivery time over Packard. Ford would have had to start with a bare plot of land just like they did with the R-2800 factory, they would have had to redo all the drawings/blueprints just like Packard did. They would have had to design the factory and tooling pretty much from scratch like Packard did instead of sending 15-20 engineers to the P & W factory for several weeks to study how P & W was doing things and then pretty much coping the P & W factory. And P & W was only 3-4 hours away by slow airplane or overnight by train.

and as for "The two stage/turbo versions of the R2800 were not combat ready until 1943 nor were their associated fighters, mainly due to engine problems. Consequently, Ford could not ramp up to full production as the assembly line was constantly being altered."

Ford never built a 2 stage mechanically supercharged R-2800 engine. Ford also never built a "C" series engine.

Ford Built 11 different models of the R-2800 and only 6 of those models were before Dec of 1942.

the -5s of 1850 hp (A series) went into B-26s and few experimental aircraft ( B-23 and XC-46)
the -21 of 2000hp that went into a variety of P-47s.
the -51 of 2000hp that went into variety C-46s
the -43 of 2000hp that went into C-46s and B-26s

and that brings us to Nov of 1942 when Ford started building 2 new models although production of some of the old ones kept going.
They built 262 of the -27 model and that went into a variety of experimental aircraft and some Douglas A-26s.
The -31 engine (built to a total of 6088 engines) for Lockheed Venturas and Ventura variants was added.

Yeah, I can see how confused and discombobulated Ford must have been dealing with single speed and two speed superchargers during their first year of production.

Ford did reach full production about as fast as any other factory reached full production It took Buick and Chevrolet over a year each from first production engine to get close to full production building R-1830s, It took Nash about 16 months for the R-2800 it took Studebaker over a year to get production of the Wright R-1820 anywhere near up to speed. Studebaker built 4 engines in Feb of 1942 and finished 1942 with 6091 engines. they hit 2000 engines a month in May of 1943 and never fell below 2000 engines a month until some point in 1945.
Sorry, nothing points to Packard either being slow or to Ford being super fast. All of these companies did a tremendous job but claiming that company "A" could have beaten company "B" into production by months (like by 25-33%) takes some real proof.

 

[GrauGeist]

Actually they were pretty much a clean piece of paper design. Aside from being a V-12 and having an overhead cam there wasn't a much in common between the Liberty and the Packard. Packard had built Liberty engines in WW I and the Packards were pretty good Liberty engines, quality varied tremendously between different suppliers. But Packard figured they could do better in the 1920s and built about 250 of the 1500 cu in Packard engine and about 330-335 of the 2500 cu in aircraft engines.

Please note that the weight given for the PT boat engines may include the clutch/gearbox.

Right, the last of the evolved Liberty engines would have been the 3M.
The 4M was from a fresh piece of paper but, it's design was still based on what did and did not work on the late Liberties and was quite successful.

 

[gjs238]

Right, the last of the evolved Liberty engines would have been the 3M.
The 4M was from a fresh piece of paper but, it's design was still based on what did and did not work on the late Liberties and was quite successful.

What's the story with the Nuffield Liberty's?
Where they a similar evolution like the Packard engines?

 

[GrauGeist]

What's the story with the Nuffield Liberty's?
Where they a similar evolution like the Packard engines?

With the war coming on, Britain was ramping up it's military assets and the Nuffield Liberty engine was one such project. They bought the license to produce the Liberty L-12 for their tanks and while it didn't see quite the evolution as Packard had done over the interwar years, Nuffield did make about a half dozen variants.

If memory serves right, it had pretty much the same horsepower as the L-12 engines, about 350 or so - if I get more time, I'll look up the specs, but they were underpowered for their applications and prone to such problems as overheating, which I suspect was more of a problem with the tank's engine compartment design and cooling system than the engine itself.

 

[Shortround6]

Here is a list of all Packard aircraft engines:
http://www.enginehistory.org/Packard/StatsAllPackardAero.pdf

The PT boat engines were modified/developed from the 2500 series aircraft engines.
Different bore than a Liberty, different stroke, different number of valves per cylinder, different angle between the cylinder banks.
Given the number of different engines Packard built attributing the PT Boat engines to the Liberty seems a bit of a stretch.
The details of the Packard built Liberty engines are in the list.

 

[GrauGeist]

Here is a list of all Packard aircraft engines:
http://www.enginehistory.org/Packard/StatsAllPackardAero.pdf

The PT boat engines were modified/developed from the 2500 series aircraft engines.
Different bore than a Liberty, different stroke, different number of valves per cylinder, different angle between the cylinder banks.
Given the number of different engines Packard built attributing the PT Boat engines to the Liberty seems a bit of a stretch.
The details of the Packard built Liberty engines are in the list.

As has been mentioned before, the Marine engines (X)M-XXXX series were derived from the Aircraft (X)A-XXXX series.

Because of Packard's evolving marine engines, especially the 1M-2500 used to set speed records in the 1930's, that Packard was approached by the Navy for building and powering, the PT boats. The 3M-2500 was the last in the long line of highly evolved Liberties. The 4M-2500 was a fresh start and incorporated 2 spark plugs and 4 valves per cylinder, compression ratio of 6.4:1, a centrifugal supercharger and a Holley 1685F aircraft carb.

Packard even hosted a "university" at their plant for instructing Navy techs proper maintenance on the engine and drive coupling systems.

So the point being that Packard had extensive knowledge and experience with large displacement engines, particularly V-12s. So when they accepted the Merlin contract, they were in a good position to get the engine into production without a lengthy delay while producing their 4M-2500 engines without interruption.

 

[RCAFson]

Right, the last of the evolved Liberty engines would have been the 3M.
The 4M was from a fresh piece of paper but, it's design was still based on what did and did not work on the late Liberties and was quite successful.

You do of course have some sort of proof of that or reference?

P & W built 4,300 B series engines (single stage) in 1942. Ford built 5,711 in 1942 so you are saying that over 10,000 engines ended up being scrapped along with their aircraft before seeing combat in the middle of WW II and nobody knows about it ???

WOW!!!!

I wonder if the B-26 crews that flew at Midway knew their planes had been scrapped before they attacked the Japanese fleet?
Likewise the B-26 crews in the Aleutian Islands in 1942 or the RAF crews in No 14 Squadron or the crews of the 22nd and 38th bomb groups operating out of Australia in early 1942.
A lot of crews flying around in imaginary aircraft if you are correct.

Not mention what happened to hundreds of P-47Cs? Just what were those P-47 Groups in England flying in the Spring of 1943?
In Feb of 1943 Republic delivered it's 602 and last P-47C and rolled out it's first P-47D-1. But I guess all the R-2800s built in 1942 were scrapped so those fighters were re-engined with 1943 production delivered by submarine to the cargo ships transporting the fighters to Europe with crews working under tarps by flashlight so the merchant seamen wouldn't find out?


An interesting document but doesn't seem to make any mention of mass scraping of engines???
It also makes very little reference to the practical problems they were trying to solve. Like time between failures or time between overhauls. New Crankshaft and dampers that allow 5-600hrs between overhauls vs 300-350 hours???
I don't know the actual numbers but a lot of that document is reference to the later C and CA engines and post war commercial engines.
Maybe I missed it. please give a page number or section in that document that says the Early R-2800s were failing at a high rate (aside from the Ford built ones failing due to being improperly cleaned of casting sand)



The R-2800 was complex and Ford did do a tremendous job but so far you have failed to show they could have cut months off the delivery time over Packard. Ford would have had to start with a bare plot of land just like they did with the R-2800 factory, they would have had to redo all the drawings/blueprints just like Packard did. They would have had to design the factory and tooling pretty much from scratch like Packard did instead of sending 15-20 engineers to the P & W factory for several weeks to study how P & W was doing things and then pretty much coping the P & W factory. And P & W was only 3-4 hours away by slow airplane or overnight by train.

and as for "The two stage/turbo versions of the R2800 were not combat ready until 1943 nor were their associated fighters, mainly due to engine problems. Consequently, Ford could not ramp up to full production as the assembly line was constantly being altered."

Ford never built a 2 stage mechanically supercharged R-2800 engine. Ford also never built a "C" series engine.

Ford Built 11 different models of the R-2800 and only 6 of those models were before Dec of 1942.

the -5s of 1850 hp (A series) went into B-26s and few experimental aircraft ( B-23 and XC-46)
the -21 of 2000hp that went into a variety of P-47s.
the -51 of 2000hp that went into variety C-46s
the -43 of 2000hp that went into C-46s and B-26s

and that brings us to Nov of 1942 when Ford started building 2 new models although production of some of the old ones kept going.
They built 262 of the -27 model and that went into a variety of experimental aircraft and some Douglas A-26s.
The -31 engine (built to a total of 6088 engines) for Lockheed Venturas and Ventura variants was added.

Yeah, I can see how confused and discombobulated Ford must have been dealing with single speed and two speed superchargers during their first year of production.

Ford did reach full production about as fast as any other factory reached full production It took Buick and Chevrolet over a year each from first production engine to get close to full production building R-1830s, It took Nash about 16 months for the R-2800 it took Studebaker over a year to get production of the Wright R-1820 anywhere near up to speed. Studebaker built 4 engines in Feb of 1942 and finished 1942 with 6091 engines. they hit 2000 engines a month in May of 1943 and never fell below 2000 engines a month until some point in 1945.
Sorry, nothing points to Packard either being slow or to Ford being super fast. All of these companies did a tremendous job but claiming that company "A" could have beaten company "B" into production by months (like by 25-33%) takes some real proof.

Production of R2800 powered aircraft to Dec31 1942:

P47 532
F4U 178
F6F 12

B26 ~200? (edit = ~600)
C-46 ~200?
----------
~1150 (Edit = ~1500 with ~2300 engine installations)

There was a handful of other prototype aircraft built as well. It's easy to see that there was little pressure to built the R-2800 in large numbers before 1943 and the large numbers of engines actually produced must have waited for suitable airframes, or been put aside in favour of later built R-2800s that had fewer problems.

 

[Shortround6]

Once again, please document the problems with the R-2800, or at least the ones that led to the wholesale scrapping of engines that you claim while the British struggled with the Napier Sabre?

There were over 1000 B-26s built before the end of 1942. in the last few months of 1942 B-26 production was running close to well over 100 planes a month.
BTW they had completed 210 B_26 bombers by some point in Oct of 1941 at which point production switched to the B-26A. Over 1100 B-26s were on order at this point.

At one point in the middle of WW II England had almost 3000 Merlins in store. Did they scrap them or put them aside for later models?
Or did airframe construction catch up with engine production.
Source for the 3000 Merlins is "Planning in Wartime" by Sir Alec Cairncross.

 

[RCAFson]

Once again, please document the problems with the R-2800, or at least the ones that led to the wholesale scrapping of engines that you claim while the British struggled with the Napier Sabre?

There were over 1000 B-26s built before the end of 1942. in the last few months of 1942 B-26 production was running close to well over 100 planes a month.
BTW they had completed 210 B_26 bombers by some point in Oct of 1941 at which point production switched to the B-26A. Over 1100 B-26s were on order at this point.

At one point in the middle of WW II England had almost 3000 Merlins in store. Did they scrap them or put them aside for later models?
Or did airframe construction catch up with engine production.
Source for the 3000 Merlins is "Planning in Wartime" by Sir Alec Cairncross.

B-26 Production:

B-26 PRODUCTION SUMMARY
________________________________________________________________________

variant built notes
________________________________________________________________________

B-26 201 201 built for test, evaluation, & training. (first build Feb 1940 first flight = Nov 1940!)
B-26A 139 52 provided to UK as "Marauder I". (first build = Oct 41)
B-26B 1,883 19 provided to UK as "Marauder IA". (first build = April 1942 Last build = ~Jan 1944)
B-26C 1,210 100 provided to UK as "Marauder II". (first build = ~Jan 1943)
XB-26D - 1 B-26 converted for de-icing test.
B-26E - Lightweight version, not built.
B-26F 300 200 provided to UK as "Marauder III".
B-26G 893 150 provided to UK as "Marauder III".

TB-26B (AT-23A) 208 Trainer / target tug.
TB-26C (AT-23B) 375 Trainer / target tug, 225 to USN as "JM-1".
JM-1P - JM-1s converted to reconnaissance spec.
TB-26G 57 Trainer / target tug, 47 to USN as "JM-2".
XB-26H - B-26G converted for bicycle landing gear tests.
________________________________________________________________________

SUM 5,266 (spread over 6 years) (My comments italicized in brackets source =The Martin B-26 Marauder and others )

I underestimated B-26 production but total production to Dec 31 1942 was probably about 600-700.

 

[Shortround6]

Actual B-26 production was a bit over 4 years. From Feb 1941 (get to this later) to April of 1945. That is at Martin Baltimore.
Martin Omaha started delivering B-26s in Aug 1942 with 3 planes completed. Martin Omaha finished 86 B-26s in 1942 but stopped production in April of 1944 when the Martin Omaha plant was converted to B-29 production.

Trying to estimate monthly (or yearly) production based on averaging total production over a number of years fails to take into account the learning curve in production and multiple sources (which rarely had the same production capacity as the Home factory, in some cases it was greater).

It took Martin 3 times longer to go from the 5th production B-26 to the 500th than it did to go from the 501st to the 1000th. This "learning curve or ramp up" was pretty typical during WW II with some early aircraft taking much longer due to prewar small production batch orders. Exception the other way was the P-40 but then the P-40 took over P-36 production lines instead of starting from scratch.

Getting back to the Feb 1940 "start" of production. That is a misprint. With the First B-26 flying for the first time in Nov 1940. Depending on source the 2nd or 2-4 Aircraft were delivered to the Army in Feb 1941.

You can estimate all you want. The Statistical Control Office, Air Technical Service Command, Wright Field thought that 1000 B-26s had been delivered before the end of 1942. Before meaning several weeks.

 

[Peter Gunn]

A lot of other moving parts were running along the A-36 and P-51A designs that were crucial to the P-51B. The most important was migrating to a true mass production line when the funding for the A-36/P-51A materialized in April-June 1942. The second most important derivative of the A-36 was the external bomb rack/fuel tank design to augment internal fuel.

In parallel with those two initiatives, were the efforts to stimulate the development of ferry tanks then combat tanks. Independent of the P-51B, the pressurized 75 gallon tank and the systems in the airframe to manage the pressurization were in the formative stage and I don't see that problem being solved any sooner. The 85 gallon tank initiative started after February 20 meeting driven by Arnold to solve the Long Range Escort 'problem'.. so the XP-51B had been flying 80 days and the first P-51B-1 was near starting on the production line. The first flight was mid July in a highly modified P-51B-1 and the production drawings completed by first of September. The first Base Depot mod was in early November. (That timeline doesn't change so the March, 1944 Berlin mission wasn't happening any sooner.)

So the answer to the question is "yes, with a caveat". Given the introduction of the Merlin based on BPA making the Merlin decision in 1940 - then the answer is Yes. That said, the P-51B would have been deployed in time to match the 353rd, 352nd, 355th, 356th as well as 9th AF 354th FG's in parallel with their operational deployment.

The limit of escort however would have been limited to Hamburg, Brunswick, and Ulm - Additionally, the 8th AF really didn't have the airplanes and crews to execute the Combined Bomber Directive to Joint Chiefs satisfaction in June-August 1943. Nor could they have prevented the Schweinfurt/Regensburg, Munster thumpings. To be sure the 8th AF losses would have been less and the LW operational losses greater.

Much appreciated, it's a "what if" that I've mulled over for many years. I didn't think there was a way to really get the Mustang into combat much earlier than they did but didn't know why. It was always easy for me to look at the issue and be confused as to why you couldn't get the -B into combat in 1942. I thought in terms too much like the old hot rodder that I am, hey just stick the bigger engine in and let's go.

What I've learned here in the last few months has made me realize not only was that not going to happen, but if it was, it would have taken the hand of God to make it a reality. It has also made me realize how much I DON'T KNOW about these issues, but be patient...I'm learning.

Thanks again.

 

[Peter Gunn]

Still not seeing how Ford could have done better than Packard in this scenario, in fact my opinion is they might have not done as good.

Ford was good at producing cheap affordable cars for the everyman, Packard built luxury quality cars and V-12 Boat/Aircraft engines. It's no coincidence Gar Wood used Packard -12's, especially as the first one to do two miles a minute over open water.

The design and engineering staff at Packard were second to none and SR6's post #437 is a good summation of the situation IMO.

 

[gjs238]

The military seemed fine using Ford's GAA, GAF, GAN as well as the GAC engines.

Ford GAA engine - Wikipedia, the free encyclopedia

 

[Shortround6]

Well, limit the RPM to 81-87% (unless you believe the 3600rpm redline in which case the tank engine was limited to 72% of the rpm of the aircraft engine) rip off the supercharger a thus lower combustion pressure dramatically and throw the planned fuel injection in the trash bin and use dual carbs and the stress on the internal parts was much reduced. It also simplified development considerably, tank engines rarely having to operate upside down

I would note that Chrysler Started work on their V-16 in May of 1940 and got a contract offer in June of 1940 (Hmmm, coincidence??) and managed to get the engine airborne for the first time 26 July 1945. Not sure what the Problem was as aircraft engines are really simple to design/build.

As a comparison P & W started work on the R-4360 in Nov 1940. First engine ran April 28th 1941 but used the front cylinders from an R-2800, the rear cylinders from an R-2800, connecting rods from the R-2180 (limited production engine around 30 built) and the salvaged reduction gear from an experimental H 24 cylinder sleeve valve engine. It also used a cast crankshaft.
By the end of 1941 P & W had completed 500 hours of full scale operation (four 28 cylinder engines), 176 hours on a single row test rig and 1412 hours of single cylinder testing. First flight in a test hack was May 25th 1942 and in 1943 ten engines were built for installation in experimental prototype aircraft. ( a number of development engines had been built before this) Several of which flew during 1944 but use in a production aircraft was only started in the summer of 1945. Perhaps in part due to lack of demand, Shoehorning this thing into existing Aircraft like the F4U and P-47 was possible but obviously only certain aircraft were going to be able to take it (and a few of the early installations left something to be desired. On one aircraft the cowl had to be extended 6 inches).

Not sure how Ford was supposed to get their V-12 in production as an aircraft engine in time to do much during the war.
Large staff can only do so much, It's a bit like the old joke about making a baby. It takes one woman 9 months but nine women cannot do the job in one month.

 

[wuzak]

I would note that Chrysler Started work on their V-16 in May of 1940 and got a contract offer in June of 1940 (Hmmm, coincidence??) and managed to get the engine airborne for the first time 26 July 1945. Not sure what the Problem was as aircraft engines are really simple to design/build.

And they had a starting point in the USAAC's "hyper" cylinder, developed in the late 1920s/early 1930s. The V-16 shared many of the "hyper" cylinder's design features, such as a hemispherical combustion chamber with two overhead valves per cylinder.

Though similar in power and capacity, the IV-2220 was about half as long again as a two stage Griffon 60-series and slightly heavier. And that doesn't take into account the weight of the turbo system required for altitude performance.

 

[Shortround6]

Here is a concept drawing of the Ford engine.

While the basic engine may have been of sound design Ford was planning on direct fuel injection like the German engines.
The concept was well known but the execution may have been a problem. Other American engine builders relied on separate/specialized Carburetor manufacturers.
Ford apparently was depending on using a single stage turbo supercharger with NO mechanical supercharger. and was planning on developing their own compressor and turbine. Exhaust gases from turbo were to be vented directly to the rear of the engine Which makes another snippet of information from one source a little questionable. The source claims it was the Navy that was interested (Army was still fooling with Continental and Lycoming engines) but doesn't state what aircraft it was intended for. While that turbo location and exhaust vent doesn't for sure mean it can't be used in a single engine aircraft it does make one wonder.
Whole developing several subsystems in addition to the main power section can mean a dramatic improvement over existing engines it also means the whole project falls apart if one of the subsystems doesn't work.
A quick google search turns up no photographs of the Ford V-12 aircraft engine although a few photos of wooden mock up can be found and at least one photo of a 2 cylinder test rig.
If someone can find pictures of the aircraft engine and not the V-12 tank engine (or recent hot rod versions of the tank engine) I would be grateful.

 

[gjs238]

Still not seeing how Ford could have done better than Packard in this scenario, in fact my opinion is they might have not done as good.

Ford was good at producing cheap affordable cars for the everyman, Packard built luxury quality cars and V-12 Boat/Aircraft engines. It's no coincidence Gar Wood used Packard -12's, especially as the first one to do two miles a minute over open water.

The design and engineering staff at Packard were second to none and SR6's post #437 is a good summation of the situation IMO.

I wonder why the Packard engines didn't go on to power US tanks and other armored vehicles.

 

[Peter Gunn]

Considering the Ford engines didn't power anything past the M26 Pershing in the heavy/medium tank line.

 

[Shortround6]

The US had a large number of engine makers in the 1930s and not all engines are the best choice for different jobs. Chrysler used FIVE six cylinder Plymouth car engines ganged together for one of the alternative M3/M4 medium tank engines. The Australians used THREE Cadillac V-8 car engines to power the Sentinel tank which used a modified M3 transmission and M3 running gear/suspension.
another M3/M4 engine combo was side by side 6-71 Detroit diesel straight 6 engines.
M-24 "light" tank used a pair of Cadillac V-8s of 110hp each but as they were 349 cu in engines they had a lot more torque than an 85 hp Ford V-8
Tanks cannot use car engines, even large car engines without resorting to multiple units.
The US Army decided to go with air cooled engines pre war for power and compactness but was stopped by the need for even low powered (250hp and up) air cooled aircraft engines for trainers and transports. Which resulted in the substitute standard engines listed above.
Post war they moved for air-cooled engines again and Continental responded with a 1790 cu in air cooled V-12;
Continental AV1790 - Wikipedia, the free encyclopedia
A flat six version using the same cylinders was used in the M-41 "light" tank and it's family of fighting vehicles (Self propelled guns,etc)

A pre-war list of American stock, industrial and Marine engines (includes bus and truck) engines runs to 41 companies and 483 engines. This does not include aircraft engines or regular car engines unless modified for non-car use.
The Need for Packard to build tank engines didn't exist considering what else they were doing for the war effort.

 

[gjs238]

I assume that it's easier to use engines already available than develop new ones.

This excerpt from the Wikipedia article on Jacob L. Devers is interesting:
Jacob L. Devers - Wikipedia, the free encyclopedia

Often against the views of his superiors, Devers lobbied for a still more heavily armored and better armed medium tank, the M4 Sherman. He played an important role in the M4's design, development and manufacturing, particularly its engine and armament. Throughout his tour as Chief of the Armored Force he worked closely with the Ordnance Department, manufacturers and the Armored Force Board at Fort Knox on the research and testing of tanks, guns, armored vehicles and ammunition. The biggest obstacle was engines. Those of pre-war tanks were rated at 250 horsepower (190 kW), which was insufficient for a 35 short tons (32 t) medium tank. Devers wanted an 800 horsepower (600 kW) engine, but this was beyond the ability of the American automotive industry to produce. Extraordinary efforts resulted in the development of a number of 400 horsepower (300 kW) engines. He controversially rejected the General Motors 6046 diesel engine in favor of gasoline engines. Battlefield experience would demonstrate that the diesel engine was superior. The quest for a better engine eventually settled on the Ford GAA engine,[59][60] but there was a persistent shortage of tank engines.[61] Some 49,234 of the reliable, versatile, low-cost M4 Sherman and its variants would be produced.[62]

I find this interesting because we were recently discussing the Packard engines used in the PT boats.
Packard 1A-2500 - Wikipedia, the free encyclopedia
These engines were developed through the 20's and 30's and seemingly could have been what Devers was looking for.
Of course, the Wiki article could be wrong - something we keep running into.