Small Aircraft Carriers

[Howard Gibson]

Point is that there were solutions to the high altitude interceptor problem without two stage superchargers and 100/130 fuel.
Two stage supercharges and 100/130 fuel allowed for more general purpose aircraft (both low and high altitude) and requiring a bit lighter power plant means more weight for fuel for longer range. The two stage supercharger and 100/130 gave the Allies more options. It did not mean the Germans (or Japanese) could not build more specialize aircraft to deal with high altitude bombers at all.

No, there weren't. There is only so much you can do with a centrifugal blower. The original Merlin supercharger had a 10.5in impeller rotating at 28000rpm. That has to be close to the speed at which the impeller would structurally rip itself apart. The air coming off the blower was supersonic. If you want more pressure, you need two blowers in series.

High octane fuel only works below the critical altitudes of the superchargers. For most two-stage Merlins, this was around 26000ft. High octane fuel, and methanol/water injection allow you to run at higher boost pressures.

The Thunderbolts, Mustangs and Spitfires had superior performance at 30000ft in 1943 and 1944.

 

[tomo pauk]

No, there weren't. There is only so much you can do with a centrifugal blower. The original Merlin supercharger had a 10.5in impeller rotating at 28000rpm. That has to be close to the speed at which the impeller would structurally rip itself apart. The air coming off the blower was supersonic. If you want more pressure, you need two blowers in series.

Perhaps there is slight misunderstanding between two of you?

SR6 was talking about all the engines, not just about the Merlin, and Merlin was not even available for the Germans. They have had the much bigger engines than the Merlin, meaning that Merlin needed the help of the additional S/C stage to compete (and it did). A DB 605 with a big single-stage S/C, like the As or D versions were, was about as good as the 2-stage Merlin, and without the need to figure out the space for the intercooler radiator, and also not paying the drag penalty for it.
As noted by SR6, such the DB 605s were too late, being introduced after the Merlin Mustang and it's teammates helped trashing the LW in the 1st 6 months of 1944. If the DB 605AS or D was introduced in late 1943, then there is a much more level playing field for the Bf 109 drivers to be had at the altitudes above 7 km.
Similar with the BMW 801 - if a good 1-stage S/C was introduced for it by second half of 1943 instead of late 1944, it would've gave the Fw 190 pilots much better chances between the late 1943 and mid-1944.

Yes, both of the 109 and 190 would've also been well served if the tidying out of the airframe was done. Here is where a 2-stage S/C for the German engines would've mattered, meaning that even imperfect airframes are still performing very good due to the excess power.

Granted, having both (engine with a better S/C + ironed-out airframes) would've been the best for the LW. Installing the big V12s on the Fw 190s as early as late 1943 would've also helped them a lot.

High octane fuel only works below the critical altitudes of the superchargers. For most two-stage Merlins, this was around 26000ft. High octane fuel, and methanol/water injection allow you to run at higher boost pressures.

Ironically, LW have had a much better access to the MW and high octane fuel than it was their supply of advanced superchargers.
Engines being big covered a lot of ground wrt. the power at lower altitudes.

 

[EwenS]

If Seafires flew the mission, it wasn't long range.

You asked how "deep" they went. As you were quick to point out, Japan is an island chain. You can't get much deeper than the opposite side of the island from which you are flying.

Edit : As for the range of a Seafire L.III fitted with a 89/90gal drop tank, as used by the Implacable Seafire squadrons in July / Aug 1945, this is from David Brown's book

"Drag did increase by as much as 10%, but when all aircraft in the Wing had been modified to carry the 89-gallon tank, it meant they could carry out offensive sweeps or strike escort duties to a useful radius of 185 miles - 225 miles was achieved by one 'Ramrod…" - The Seafire, David Brown

Its operational endurance was increased to 3.5 hours from 2.5 with the previous 45 gal slipper tank ( Makes accommodation for 5 minute take-off climb to 5000ft, 15 minutes of combat at 5000ft, 20 minutes of holding pattern at 2000ft.)

Indefatigable's squadrons used 90 gal slipper tanks.
Not as good as Hellcats or Corsairs but better than available previously.

 

[tomo pauk]

The original Merlin supercharger had a 10.5in impeller rotating at 28000rpm.

About this tidbit. The original S/C on the Merlin have had in impeller of 10.25in. The Merlin 46 & 47 were with a 10.85in impellers, and that one was of a more refined type.
A good deal of the Merlin's potential power/thrust was lost via the bad carbs before 1944, and via the uninspired exhausts on the 1-stage engines. The Merlin 46/47 were unfortunately with only 1-speed drive, that reduced their abilities under 15000 ft.

 

[tomo pauk]

There were also the ways to improve the BMW 801 installation on the Fw 190.
Going with the external air intakes was tested already by 1943 (if not in 1942?), but as-is these were pretty draggy, so the gain in the higher altitudes was paid with the losses under 6 km. The 801F was to be outfitted with the less pronounced external air intakes, situated roughly as it was the case with the La-7 or the F8F.
(one does wonder how well the gasses from the guns' firing so close will ... agree with that installation, though)
Another thing was the abandonment of the closely-coupled oil tank and cooler, so the oil tank was of the 'ordinary' shape and relocated next to the firewall. The oil cooler was shaped like the coolant radiator on the Jumo 213, for example, in a more Meredith-y way. The reduction gearing was also of the more streamlined shape to help out.
The inward-opening doors for the cooling air to 'escape' were also implemented. HMGs - know to be a draggy item - were to be getting the covers.

pictures here

The exhaust stacks do look draggy, though.

 

[Shortround6]

To further some of Tomo's comments,

There are a lot of things going on in engines. The Germans (and Soviets) swapped large displacement for high boost. This means that the DB605 AS with big blower was using 1.42Ata at 26,200ft. Just about 6lbs of boost. The Germans also used higher compression in the cylinders than the allies did so that limited the amount of boost they could use for a given grade of fuel. A real unknown is the heat of intake charge. This depends on the efficiency of the supercharger and amount of compression being done. At 26,000ft the Germans were getting just about a 4 to compression ratio out of the supercharger which is pretty darn good for a single stage compressor (about what British were getting out of their jet engine compressers near the end of the war). The Merlin in the P-51B&C was compressing the air over 5 times (2.2Ata) to get to the 66in (Now with a two stage compressor that means you need less power and heat the air less for the same level of compression. But the Merlin is compressing the air more, I don't know where the crossover point was. Germans were using the direct fuel injection which didn't suffer from knock or pre-ignition as bad. Germans could (but didn't) use an intercooler on the DB 605. They did on some of the Jumo 211s used in JU 88s and He 111s.
The bigger German and Soviet engines were heavier than the smaller Merlin and Allison but they turned a bit less rpm. Once you start adding two stages and intercoolers the total weight starts to even out.

The Germans had choices. How they used those choices is subject to debate. But the choices existed and with the shorter range needed by the Germans vs an escort fighter they were viable even if not the best possible choice.

 

[tomo pauk]

The Germans had choices. How they used those choices is subject to debate. But the choices existed and with the shorter range needed by the Germans vs an escort fighter they were viable even if not the best possible choice.

A lot of times, Germans made a bad choice, not just wrt. their engine program and what engine goes into what airframe.

For example, as early as October of 1941 (!), the 'roadmap' issued by DB notes the DB 601 with big S/C (the 601Q and R) as well as the DB 605 with the big S/c (605A, E and F). The 601Q on 2700 rpm was to have a tad more HP above 5.7km than the 605A on 2600 rpm, while the 605D/E/F were supposed to have 1400 HP at 6 km - ie. as good as the BMW 801D. (Un)fortunately, the 601Q/R never materialized, while the 605D/E/F became available 3 full years after the 'roadmap'.
Granted, the huge problems with the 605A put a lot of the brakes on the development of that family of engines. The 'big S/C' engines were to be using the C3 fuel, however by the time these were available the B4 was also allowed.

Again humping on the BMW 801 - by the time the G&R 14R4/5 engines were tested and disassembled, BMW should've start working 24/7 in order to adopt the S/C from that engine for the 801.
Both companies used inordinate amounts of time, money, raw materials and manhours in chasing the white elephants, instead of further developing what they already had.

 

[don4331]

To further some of Tomo's comments,

There are a lot of things going on in engines. The Germans (and Soviets) swapped large displacement for high boost. This means that the DB605 AS with big blower was using 1.42Ata at 26,200ft. Just about 6lbs of boost. The Germans also used higher compression in the cylinders than the allies did so that limited the amount of boost they could use for a given grade of fuel. A real unknown is the heat of intake charge. This depends on the efficiency of the supercharger and amount of compression being done. At 26,000ft the Germans were getting just about a 4 to compression ratio out of the supercharger which is pretty darn good for a single stage compressor (about what British were getting out of their jet engine compressers near the end of the war). The Merlin in the P-51B&C was compressing the air over 5 times (2.2Ata) to get to the 66in (Now with a two stage compressor that means you need less power and heat the air less for the same level of compression. But the Merlin is compressing the air more, I don't know where the crossover point was. Germans were using the direct fuel injection which didn't suffer from knock or pre-ignition as bad. Germans could (but didn't) use an intercooler on the DB 605. They did on some of the Jumo 211s used in JU 88s and He 111s.
The bigger German and Soviet engines were heavier than the smaller Merlin and Allison but they turned a bit less rpm. Once you start adding two stages and intercoolers the total weight starts to even out.

To nitpik SR6's comments:

It doesn't matter how many stages you use to achieve the same compression ratio - if the compression ratio and the efficiency are the same, they use the same amount of power.

However, it is easier to design a more efficient compressor to achieve a lower compression ratio​

For the Merlin with 2 stages, each stage only compresses ~2.24:1 to achieve 5:1 total at 26k'. (2.24*2.24=5.02)​

By having more efficient supercharger stages less heat is generated and that heat generated is the additional power lost.​

 

[Howard Gibson]

Perhaps there is slight misunderstanding between two of you?

SR6 was talking about all the engines, not just about the Merlin, and Merlin was not even available for the Germans. They have had the much bigger engines than the Merlin, meaning that Merlin needed the help of the additional S/C stage to compete (and it did). A DB 605 with a big single-stage S/C, like the As or D versions were, was about as good as the 2-stage Merlin, and without the need to figure out the space for the intercooler radiator, and also not paying the drag penalty for it.
As noted by SR6, such the DB 605s were too late, being introduced after the Merlin Mustang and it's teammates helped trashing the LW in the 1st 6 months of 1944. If the DB 605AS or D was introduced in late 1943, then there is a much more level playing field for the Bf 109 drivers to be had at the altitudes above 7 km.

Atmospheric pressure is 14.7psi at sea level, 6.67psi at 20,000ft, 5.46psi at 25,000ft, and 4.37psi at 30,000ft. On a Merlin Mustang or Spitfire on 130 octane fuel, you need an absolute manifold pressure of 33psi. The critical altitude for the Merlin 66 was 26,000ft, or approximately 25,000ft. To get full power at 25,000ft, the Merlin's supercharger must increase pressure by over 27psi. The Merlin's two-stage supercharger was up to the job. The critical altitude on the Merlin_114s for the Mosquito_35 was 30,000ft. On a turbocharged and supercharged P-47, the critical altitude also was 30,000ft. For the single-stage supercharged German engines, the critical altitude was 20,000ft, or 6km. At 20,000ft, an Fw190 was slightly faster than a P-47. At 30,000ft, the P-47 was 50mph faster. The Germans could not match allied performance at high altitude until the two-stage supercharged Jumo 213E appeared right at the end of the war. The DB605Ls were two-stage supercharged. Did any of these reach service?

The intercoolers are needed because gases increase temperature when they are compressed. The Merlin's intercooler was in the right place, at the output of second compressor. The P-47's was located between compressors, which is not as good. You want the air entering the cylinders to be as cool as possible.

In the Battle of Britain, the DB601 and the Merlin put out about the same power. The BMEPs were close. The 27_litre Merlin ran at 3,000rpm. The 34_litre DB601 ran at 2,500rpm. The rpms made up for the Merlin's smaller displacement. The DB605 was slightly bigger at 36_litres, but by that time, the Merlins had better superchargers and higher octane fuel.

The book Mosquito, by C. Martin Sharp & Michael J. F. Bowyer has a speed/altitude chart for some Mosquitos at 3000rpm and 18psi boost. The Mosquito_35 with two-stage Merlin_114s did 425mph at 30,000ft. The Mosquito_33 with single-stage Merlin_25s and ejector exhausts did 387mph at 13,000ft. Below 10,000ft, it was substantially faster than the mark_35. This must be one of the reasons they kept single-stage Mosquitos in production throughout the war.

 

[tomo pauk]

Atmospheric pressure is 14.7psi at sea level, 6.67psi at 20,000ft, 5.46psi at 25,000ft, and 4.37psi at 30,000ft. On a Merlin Mustang or Spitfire on 130 octane fuel, you need an absolute manifold pressure of 33psi. The critical altitude for the Merlin 66 was 26,000ft, or approximately 25,000ft. To get full power at 25,000ft, the Merlin's supercharger must increase pressure by over 27psi. The Merlin's two-stage supercharger was up to the job. The critical altitude on the Merlin_114s for the Mosquito_35 was 30,000ft.

Nobody is denying that Merlin's 2-stage S/C was up to the job.
Saying that 'critical altitude was at that or this altitude' is only half of the story. Another half is what actual power (or, even better, the thrust) was kinda completes that. Merlin 66 was making some 1600 HP at 16000 ft (~4900m), without the ram effect. The DB 605ASM was doing 1600 PS at about 6 km.
Major caveats here:
- the the 605ASM was a much later engine than the 2-stage Merlin of any kind, even the Mk.66 was more than a year earlier - that is a major thing
- Merlin 66 was the 'low altitude' 2-stage Merlin (even though these power/altitudes were a dream for many engines) ; the Mk.63 and 70 series were set for the greater altitudes, and so was the V-1650-3; eg. the Mk.63 was good for 1520 HP at 6400m

tl;dr - a big S/C on the DB 605 in theory cancels all the advantages the wartime 2-stage Merlins had, but in practice it was too late

On a turbocharged and supercharged P-47, the critical altitude also was 30,000ft. For the single-stage supercharged German engines, the critical altitude was 20,000ft, or 6km. At 20,000ft, an Fw190 was slightly faster than a P-47. At 30,000ft, the P-47 was 50mph faster. The Germans could not match allied performance at high altitude until the two-stage supercharged Jumo 213E appeared right at the end of the war. The DB605Ls were two-stage supercharged. Did any of these reach service?

Probably only the 213E reached some service of all the 2-stage S/ced German engines.
The DB-603A- or Jumo 213A-powered Fw 190 would've cancelled a lot of the P-47 advantage. Luckily, Germans were slow to jump on this opportunity, favoring to power the Ju 188 with the Jumo 213 and the indifferent Me 410 with the DB 603.
The outdated S/C on the BMW 801 was a brake for the performance potential on the Fw 190A fighters.

Note that the turboes required the good intercoolers, they were benefiting less wrt. the ram effect (that added 4000-5000 ft on the late-war aircraft, but just about 1500 on the turboed engines), and there was next to no exhaust thrust to help out (unlike with the non-turbo engines, where the gain at 25000 ft was some 12%, give or take, and depending on the exhausts).
Turbo, ducting and interoller pushing the size, drag and weight of the aircraft was also a thing.

The intercoolers are needed because gases increase temperature when they are compressed. The Merlin's intercooler was in the right place, at the output of second compressor. The P-47's was located between compressors, which is not as good. You want the air entering the cylinders to be as cool as possible.

I know that the intercoolers are needed when people are running the high-boost machines, the high boost being needed for the small engines so they can compete. A big engine can do with lower boost to make the similar power, so the intercooler is not as mandatory as it will be for a small engine/high boost combination. The water-alcohol system combined with the lower compression ratio can also work well.

 

[don4331]

Note that the turbos required the good intercoolers, they were benefiting less wrt. the ram effect (that added 4000-5000 ft on the late-war aircraft, but just about 1500 on the turbo'd engines), and there was next to no exhaust thrust to help out (unlike with the non-turbo engines, where the gain at 25000 ft was some 12%, give or take, and depending on the exhausts).
Turbo, ducting and intercooler pushing the size, drag and weight of the aircraft was also a thing.

The turbos required the good intercooler because GE made an atrocious compressor (it made BMW's compressor look good)...fortunately for them, there was lots of exhaust energy, so when combined with intercooling it didn't really hurt them.

Note: An air to air intercooler (P-47, Corsair) is more efficient than air to water, then water to air (P-51)​

 

[tomo pauk]

The turbos required the good intercooler because GE made an atrocious compressor (it made BMW's compressor look good)...fortunately for them, there was lots of exhaust energy, so when combined with intercooling it didn't really hurt them.

There was no turbos to help out the on the BMW 801s (and their compressors) that powered the Fw 190s or Ju 882/188s.

P&W-made 1-stage superchargers for their engines were even less capable than what the BMW did; it was these S/Cs that the GE-made turbo helped out make good/great power above 20000 ft. Impellers used on the P&W engines from mid-1930s, lacking the curved or parabolic guide vanes, were of worse shape than what their Hornet have had in the ealry 1930s; that was rectified some time in 1944 for the C series R-2800s.
Before late 1944, both P&W and BMW radials were also with the squished intake elbows, further messing with the airflow, and thus robbing the power that would've otherwise went to the prop. P&W also used the too small impellers (same problem as with the V-1710) - not good if the altitude power is asked and there is no turbo.

P&W's 2-stage S/C were better, something that the BMW never gotten in series production.

BTW - any source on the claim that "GE made an atrocious compressor"?

Note: An air to air intercooler (P-47, Corsair) is more efficient than air to water, then water to air (P-51)

Is it?

 

[Howard Gibson]

I know that the intercoolers are needed when people are running the high-boost machines, the high boost being needed for the small engines so they can compete. A big engine can do with lower boost to make the similar power, so the intercooler is not as mandatory as it will be for a small engine/high boost combination. The water-alcohol system combined with the lower compression ratio can also work well.

The Bf109K4 did 450mph at 20,000ft. The performance dropped off as it climbed, but it looks to me like it would equal the speed of a two-stage Mosquito at 30,000ft. The Luftwaffe pilot could keep the Mosquito in sight, and shake his fist at it. Prior to fall 1944, the Luftwaffe had no piston engined aircraft in service that could hit 400mph at 30,000ft.

The intercooler works when the air heats up due to compression or anything else. The methanol water injection also reduces the air temperature. The cooler the air is at the intake manifold, the more boost pressure you can run at. The Rolls Royce Griffon had about the same displacement as the DB605, and two-stage superchargers in the SpitfireXIVs.

Centrifugal compressors get you a certain pressure increase. If that is not enough, you need two of them, or some other technology, like a Roots blower.

 

[tomo pauk]

The Bf109K4 did 450mph at 20,000ft.

Tidying up of the airframe and having a good S/C on a mature engine paid off. Granted, that was some 12 months too late.

The intercooler works when the air heats up due to compression or anything else. The methanol water injection also reduces the air temperature. The cooler the air is at the intake manifold, the more boost pressure you can run at. The Rolls Royce Griffon had about the same displacement as the DB605, and two-stage superchargers in the SpitfireXIVs.

All good.

Centrifugal compressors get you a certain pressure increase. If that is not enough, you need two of them, or some other technology, like a Roots blower.

Roots blower was tested in the inter-war period, and was never judged as better than the centrifugal type for this role.
Smaller the engine, more it is depending on having 'perfect' supercharging.

 

[Howard Gibson]

Tidying up of the airframe and having a good S/C on a mature engine paid off. Granted, that was some 12 months too late.

Roots blower was tested in the inter-war period, and was never judged as better than the centrifugal type for this role.
Smaller the engine, more it is depending on having 'perfect' supercharging.

The Bf109K14 had the better supercharger. I don't think it reached service.

If you scale blowers down a lot, their pressure performance drops. Positive displacement Roots blowers work on cars, with displacements around an order of magnitude less than that of WWII aircraft. I would expect a fighter plane centrifugal blower to be working somewhere near its maximum/optimal capacity. If you need more pressure, you need two blowers in series.

Let's put it another way. If I select a blower that is 50% bigger than the one I have, my mass flow at zero restriction increases. My maximum pressure will not improve much. I now have a much bigger bump on the side of my fuselage, or a longer fuselage, depending on where the blower is located.

 

[tomo pauk]

The Bf109K14 had the better supercharger. I don't think it reached service.

The 109K4 was with a better S/C than it was the case with the, for example, G6.
K14 was with an even better S/C.

If you scale blowers down a lot, their pressure performance drops. Positive displacement Roots blowers work on cars, with displacements around an order of magnitude less than that of WWII aircraft. I would expect a fighter plane centrifugal blower to be working somewhere near its maximum/optimal capacity. If you need more pressure, you need two blowers in series.

Let's put it another way. If I select a blower that is 50% bigger than the one I have, my mass flow at zero restriction increases. My maximum pressure will not improve much. I now have a much bigger bump on the side of my fuselage, or a longer fuselage, depending on where the blower is located.

Nobody here is making assumptions that 2-stage S/Cs don't work.

 

[MIflyer]

Note: An air to air intercooler (P-47, Corsair) is more efficient than air to water, then water to air (P-51)

Not the point. An air to air intercooler is:

1. A lot more draggy when it comes to airflow, getting it in and then getting it back out.

2. Larger than liquid cooled, VERY important when stuffing it under a Mustang or Spitfire cowl..

2. Much, much more difficult to regulate the temperature. Hence the serious problems the P-38J and L had when the chin type intercooler overcooled the air; they finally added some cowl flaps and the recon models added locally devised fixed blockage in the intercooler exhaust.

 

[don4331]

BTW - any source on the claim that "GE made an atrocious compressor"?

Back calculation from the numbers on engine intake temperature/intercooler effectiveness/outside air temperature/boost level, say the GE compressor was <50% efficient versus the ~70% for the RR compressors.

The numbers in R-2800 P&W Dependable Masterpiece, have the air to air intercooler effectiveness at ~45% at 20k', vs those of Mustang at ~61% for the engine air to liquid and 65% for the fuselage liquid to air for a combined effectiveness of ~40%.

As we discussed ~6 months ago, there's a lot of variables in intercooler effectiveness - bigger is better, but comes a cost of restriction/weight.​

USAAF had a design requirement of no more than 1.5" Hg restriction for the air coming from the 2nd stage compressor to 1st stage; not sure if RR was required to meet that or not.

Not the point. An air to air intercooler is:

1. A lot more draggy when it comes to airflow, getting it in and then getting it back out.

2. Larger than liquid cooled, VERY important when stuffing it under a Mustang or Spitfire cowl..

2. Much, much more difficult to regulate the temperature. Hence the serious problems the P-38J and L had when the chin type intercooler overcooled the air; they finally added some cowl flaps and the recon models added locally devised fixed blockage in the intercooler exhaust.

1. & 2. The wing air to air intercoolers of the P-38D-G intercoolers added no additional drag while meeting the USAAF requirements for airflow restriction. By using leading edge intakes/belly outlets in the Corsair, Vought used otherwise unused space for their installation while adding next to no drag. Republic got a little carried away IMHO, but part of that is result of the GE compressor effectiveness and part of that is the engine makes almost 2X power of Merlin, so it has to be larger and part is timing .
I'm not entirely convinced that North American couldn't have made a fuselage mounting the turbo behind the cockpit and a chin type intercooler that would have allowed the Allison engine to produce results equal or better to the historic Mustang. (Allison is sufficiently more fuel efficient when turbocharged than the RR Merlin to not need the fuselage tank...per Dan Whitney in his Allison engine tome).

​

3. Not aware of chin intercoolers having issue - would be surprising as 2nd part of USAAF's requirement was for air at engine to be ~90*F.
The wing intercooler has issues with overcooled air when operating at high altitude and doesn't have capacity to handle airflow of the 1,400+ hp Allison V-1710-89 (91)s but that is more about Johnson's crystal ball not telling him he needed to plan for 40% increase, not just 20% from the original V-1710-C7 (C9) engines. And that he wasn't designing just an interceptor (albeit a long range one), but an aircraft that would spend hours cruising at altitude.

Kelly should have polished his crystal ball a little more. ​

 

[tomo pauk]

Back calculation from the numbers on engine intake temperature/intercooler effectiveness/outside air temperature/boost level, say the GE compressor was <50% efficient versus the ~70% for the RR compressors.

Whose calculations are in question, and is that math easily accessible?

The numbers in R-2800 P&W Dependable Masterpiece, have the air to air intercooler effectiveness at ~45% at 20k', vs those of Mustang at ~61% for the engine air to liquid and 65% for the fuselage liquid to air for a combined effectiveness of ~40%.

Thank you.
Are the % numbers for the Mustang also in that book?

I'm not entirely convinced that North American couldn't have made a fuselage mounting the turbo behind the cockpit and a chin type intercooler that would have allowed the Allison engine to produce results equal or better to the historic Mustang. (Allison is sufficiently more fuel efficient when turbocharged than the RR Merlin to not need the fuselage tank...per Dan Whitney in his Allison engine tome).

The V-1650-3 was using 165 US gals per hr at 26400 ft (no ram) for 1300 HP (+ perhaps another 150 HP worth of exhaust thrust there?).
The V-1710-89/-91 were using 162 US gals per hr at 25000 ft (no ram) for 1425 HP (+ barely no exhaust thrust worth speaking about?).

187 US gal per HR was the consumption of the V-1710-98/-91, for 1600 HP at 25000 ft (no ram).
189 US gals per HR was used by V-1650-3 for 1410 HP (+ how much worth of exhaust thrust) at 23750 ft (no ram).

At max cruise, -89/-91 used 72 gal/hr for 825 HP, the -3 used 68 gal/hr for 860 HP. The V-1710 needed to use auto-rich there, while the V-1650 was doing that with auto lean.
On auto-lean, the -89/-91 used 57 gal/hr for 737 HP at 'desired cruise'.

At full ram (ie. high speed), the critical altitude went up by 5000-6000 ft on the P-51B, and just by 1500-2000 ft for the P-38J (such was the nature of turboed engines)

tl;dr - even if the turbo can be boxed in the P-51, it offers no advantage over the 2-stage Merlin, and it cost the fuel tankage. That is before we account for the volume needed for the ducting required for the setup to work.

 

[Shortround6]

The numbers in R-2800 P&W Dependable Masterpiece, have the air to air intercooler effectiveness at ~45% at 20k', vs those of Mustang at ~61% for the engine air to liquid and 65% for the fuselage liquid to air for a combined effectiveness of ~40%.
As we discussed ~6 months ago, there's a lot of variables in intercooler effectiveness - bigger is better, but comes a cost of restriction/weight.
USAAF had a design requirement of no more than 1.5" Hg restriction for the air coming from the 2nd stage compressor to 1st stage; not sure if RR was required to meet that or not.

We have to be careful we are looking at the same things. They often rated the "intercooler effectiveness" as the temperature drop per 100 degree F, temp difference --degrees F.
So a 45% effectiveness would reduce the temperature of the intake charge air by 45% per 100 degrees ?
Now when trying to compare from one aircraft to another we run into a number of problems. In an air cooled system ( a decent one) the amount of cooling is very dependent on the ratio of cooling air mass to the mass of the intake charge air. You can get around 45% using about a 1.25 ratio. A 1 to 1 ratio gives about 40% and a 2 to 1 ratio gives just over 60%. Then things got to pot. A 3 to 1 ratio only gives about 70%. But you need a truly huge intercooler to handle 3 times the amount of cooling air compared to the intake air.

The whole concept of the USAAF turbo charger design was that the engine would "think" it was at sea level. Not only was there a restriction on the amount of pressure loss going through the intercooler (higher pressure loss meant less power which kind of defeats the whole thing), but there was supposed to be a limit on the inlet temperature, goal was around 100 degrees F (rarely, if ever, meet). So we have a temperature roller coaster. At 20,000ft our plane takes in -12 degree air, the turbo compresses it back to sea level pressure but, depending on how good the supercharger is, adds around 175 degrees of temperature and then the intercooler tries to get it back down to 100 degrees or a little more (instead of the 59 degree of a standard day) and then the engine supercharger boosts the pressure from 30in (roughly) to 42-52inches (depending on engine/plane) with the resulting temperature rise.

I have no idea what RR was trying to do temperature wise. Basically they were using an after cooler. All the compressing was done in the two stages with very little cooling (if any?) done between stages and the vast majority of the cooling was done between the supercharger assembly and the intake manifold. For RR it was what it was, The RAF didn't tell RR how to design their engines. For the USAAF, it was a general standard/goal for all turbo charged engines, P & W (R-1830s and R-2800s), Wright or Allison.
The USAAF also had the 'goal' that the exhaust back pressure was supposed to stay the same from sea level to 25,000ft or above. On a normal engine the back pressure drops due to the decreasing air pressure. How close the USAAF got varied somewhat but it went to pot with WEP when they used the turbo to provide the extra boost. You have to close the waste gate's early to get more pressure from the turbo to go into the engine supercharger.