WI: NACA Modified P-38

[tomo pauk]

The other point to mention is that by using a Turbocharger for the first stage, rather than a variable speed supercharger, you get to run at full open throttle at a lower altitude (beginning at the critical altitude of the engine stage supercharger) and the wastegate then controls just how much boost you're getting.

I'd disagree with the notion of ability to use full throttle with a turboed ww2 aero engine, even at low levels.
edit: It was possible with very high octane fuel, like the 150 grade, and/or when combined with water-alcohol injection.

 

[Draconis]

The other point to mention is that by using a Turbocharger for the first stage, rather than a variable speed supercharger, you get to run at full open throttle at a lower altitude (beginning at the critical altitude of the engine stage supercharger) and the wastegate then controls just how much boost you're getting. This improves engine breathing and efficiency since it does not have the throttle restriction.

I know they developed an automatic turbo overspeed protector. But I thought the pilot still had to monitor the boost and thereby adjust the throttle if required. Especially important at lower altitudes. But I could be overlooking something here.

Here is a chart from one of Greg's Youtube videos on the P-38 comparing the difference between supercharged and turbocharged engines as they climb into the thinner air. It illustrates the smooth and almost continuous power output from sea level to critical altitude the turbocharged engine provides. Take note of when the ram air is opened. The orange line. Are turbochargers worth the extra weight and complexity? I think so.

 

[tomo pauk]

Here is a chart from one of Greg's Youtube videos on the P-38 comparing the difference between supercharged and turbocharged engines as they climb into the thinner air. It illustrates the smooth and almost continuous power output from sea level to critical altitude the turbocharged engine provides. Worth the extra weight and complexity? I think so.

V-1710 on the P-38J/L with 60 in Hg was in WER power setting, 100/130 grade fuel. The V-1650-3 and -5 in WER setting (= 67 in Hg on 100/130 grade fuel) were making much better power than at 60-61in Hg, eg. see the chart here. Greg, in his infinite wisdom, does not acknowledge the effects of the exhaust thrust (worth some 12% extra power at ~15000 ft, and some 15% at ~20000ft) - that amounts to another 200+ HP at high altitudes, negating any book HP advantage the best V-1710s has. Another thing that Greg does not note is the ram effect, worth some 2000 ft to the book rated altitude for the turboed engine, but some 4000-5000 ft to the engines with good 2-stage engines (that 2-stage Merlin certainly was). The big difference with ram and without it can be noted at the chart I've linked previously.

Net effect of all this was that a 2-stage Merlin was every bit as good a high altitude engine as the turboed V-1710 power-wise (if not better), with far easier packaging (= can be retrofitted easily), requiring less bulk to be installed, and not requiring a fully debugged turbocharger.

 

[Draconis]

@tomo pauk I've having trouble reading that chart you put up. Do you have something a little clearer to read?

"Net effect of all this was that a 2-stage Merlin was every bit as good a high altitude engine as the turboed V-1710 power-wise (if not better), with far easier packaging (= can be retrofitted easily), requiring less bulk to be installed, and not requiring a fully debugged turbocharger."

The 2 stage Merlin was fine but pilots had to be cautious about being caught on the downside of the first step at the in between altitude range just before the second stage kicks in. It did the job though.

 

[tomo pauk]

@tomo pauk I've having trouble reading that chart you put up. Do you have something a little clearer to read?

This one is much easier to read. It does not note the boost; the power is too high for the 61 in Hg (military power), as well as too low for 67 in Hg (WER) - perhaps the chart is preliminary?

The Merlin 65 is equivalent of the V-1650-7, that were mostly installed on the P-51D/K. Via change of S/C gearing, the hi-alt power was sacrificed (some 150 HP less vs. the -3), in order to gain the lower altitude power (some 200 HP more).

The 2 stage Merlin was fine but pilots had to be cautious about being caught on the downside of the first step at the in between altitude range just before the second stage kicks in. It did the job though.

Both stages were always on, being on the same shaft there was no possibility to engage it or disengage them separately. It were the speeds (ie. S/C drive ratios) that were changed. Only the P&W supercharger stages on their 2-stage engines' versions offered the possibility of separately engaging and disengaging the 1st stage (the 2nd stage was always turning when the engine was running)
Note that even at 21-23 kft, the -3 will still be making 1300+ BHP plus ~200 HP due to exhaust thrust = 1500+ HP total (without the ram effect).
The -7 will be doing another 100 BHP there.

2 stage Merlin did the job d@mned well. Any aircraft that was up-engined with it was a far superior aircraft than the previous version of the respective type, and, what is even more important, it made a difference against the Axis assets.

 

[EverKing]

I'd disagree with the notion of ability to use full throttle with a turboed ww2 aero engine, even at low levels.
edit: It was possible with very high octane fuel, like the 150 grade, and/or when combined with water-alcohol injection.

To be clear, I'm not talking full forward on the throttle lever (in aircraft with automatic wastegate control) but rather a fully open throttle butterfly on the engine. This was the way USAAF turbo-charged aircraft were set up, once above the critical altitude of the engine itself (factoring in the integral supercharger which most had), the engine power/manifold pressure was regulated entirely by the turbo wastegate with the engine throttle left full open. In fact, in aircraft without an automatic wastegate control (early model aircraft for the most part) they used a separate lever to manually control the wastegate and the operating manual specified that to attain desired Manifold Absolute Pressure at altitude the throttles should always be left in the fully open position and MAP should be managed by adjusting the wastegates. For example, the B-17 operating instructions:

EDIT: And the P-47 Handbook [here they say "supercharger" but they are talking about the Turbo-Supercharger controls]:

This contrasts to a two-stage supercharged engine where the engine throttle remains the primary means of controlling MAP. You could almost look at it as a Supercharged engine uses restricted airflow to reduce MAP, while a Turbo-Supercharged engine uses increased compression to increase MAP.

Obviously, this doesn't apply if you're below critical altitude and demanding less power than the integral supercharger can provide.

 

[pdf27]

It doesn't become like a supercharger, it becomes a supercharger; a centrifugal supercharger which is no longer turned by exhaust gas.

The reason I'm using "like" is that while at a component level it's a supercharger (centrifugal impeller turned by mechanical power at a multiple of engine speed), at a system level it isn't. Key point there is that a supercharger pulls power from the engine, this would inject power into the engine.

The V-1710-127 (E22) which you reference was originally intended for the P-63 but never got past the concept / test stage. In that engine the exhaust gas turbine was not a Pressure Turbine, as used on Turbo-Superchargers, but a Blowdown Turbine. So, not only did that turbine not turn a compressor (power recovered from it was instead delivered directly back to the crankshaft by means of a fluid coupling) but it operated on a different principle, relying on pulsed exhaust gas jet velocity instead of a collected steady pressure differential between inlet and outlet.

And to be clear I'm not proposing to use a blowdown turbine - the physics wasn't well enough understood at the time and the additional piping complexity for an inline engine would have been too hard a sell. There is still a significant advantage to the proposal even with a normal pressure turbine however.

Removing the primary supercharger which was built into the engine would require a full redesign of the induction system and rear engine housing. It was integral to the engine operation and the carburetor fed into it with the fuel being injected directly on the impeller so that the feed to the manifold was already a compressed fuel-air mixture--which is different that we're used to now days where the compressor is only handling air and the fuel is mixed after compression.

That's what I'm proposing and I think the knowledge of the time would have supported, although it's worth noting that the V-1710 was designed to be modular with the Supercharger being a separate module rather than an integral part of the engine design - that makes swapping out modules vastly easier since they have a standard (and well controlled) interface. So the air flow for the engine would be:

1. Air enters what had previously been the turbocharger compressor through a Szydlowski-type swirl throttle. Like any other throttle, this controls the air mass flow / pressure downstream of the compressor - the benefit is reduced pumping work and so air temperature increase.
2. The turbine-compressor unit is geared to the engine at a fixed speed ratio. This is partially for mechanical simplicity/lightness, partially to optimise the operation of the power recovery turbine by keeping the velocity triangles in it roughly constant.
3. There is a glycol-to-air intercooler downstream of the compressor to minimise charge temperature.
4. Fit a Bendix-type pressure carburettor downstream of the intercooler - this is to maximise charge cooling from evaporation by applying it after the intercooler. Injecting into the compressor eye gives you a certain amount of cooling plus better distribution so is preferable if you aren't intercooling, but once you are you want to move it. Again, standard components, layout and control system.
5. The engine section is essentially the same - potentially slightly lower air inlet temperatures allowing for higher maximum boost, but that depends what intercooler was fitted before.
6. The key difference is in what happens to the exhaust gases: previously some of them are used for pumping work on the turbocharger, and the rest is dumped overboard through a fairly primitive venturi which can only be tuned for a particular flight condition. This gives additional jet thrust from the exhaust which is significant in supercharged engines at high altitude. HOWEVER, propulsive efficiency is pretty low at WW2 fighter speeds: running the same gas through a power recovery turbine and sending the energy to a propeller will give significantly more power.

EDIT: Actually, on the -93 with the Aux. Supercharger, the carb and fuel fed into that stage rather than the engine stage. So, with that in mind, what you're saying would just create a single-stage, multi-speed supercharger which I think would still fall short of a multi-stage variable speed system. Of course, I'm not well versed on the Sydlowski throttle so maybe that would help make up the difference.

Not quite - I'm proposing a single-stage, single speed system plus power recovery turbine. The Szydlowski throttle partially mitigates some of the problems with single speed/single stage, and together with the power recovery turbine **should** mean you come out ahead of early-war US turbo-supercharged systems while using the same technology bricks as they do.

Greg, in his infinite wisdom, does not acknowledge the effects of the exhaust thrust (worth some 12% extra power at ~15000 ft, and some 15% at ~20000ft) - that amounts to another 200+ HP at high altitudes, negating any book HP advantage the best V-1710s has. Another thing that Greg does not note is the ram effect, worth some 2000 ft to the book rated altitude for the turboed engine, but some 4000-5000 ft to the engines with good 2-stage engines (that 2-stage Merlin certainly was). The big difference with ram and without it can be noted at the chart I've linked previously.

Net effect of all this was that a 2-stage Merlin was every bit as good a high altitude engine as the turboed V-1710 power-wise (if not better), with far easier packaging (= can be retrofitted easily), requiring less bulk to be installed, and not requiring a fully debugged turbocharger.

This is a big part of the reason I'm interested in turbo-compounding: there is no way the USAAF leadership is going to abandon turbocharging to switch to a Merlin-style 2 stage supercharger, no matter how good it is given the decades of work they've put into it at this point. However, gearing the turbocharger to the shaft has some clear advantages - increased power recovery and elimination of waste-gate issues being particularly important.
Key point is illustrated by the graph below - using a power recovery turbine on the jet exhaust shifts you from a Harrier-style disk loading to something more like a tilt-rotor, giving you up to 3x the thrust (less at high speed) from the same exhaust energy.

 

[tomo pauk]

This is a big part of the reason I'm interested in turbo-compounding: there is no way the USAAF leadership is going to abandon turbocharging to switch to a Merlin-style 2 stage supercharger, no matter how good it is given the decades of work they've put into it at this point.

Already by the mid ww2, USAAF was more than eager to incorporate the 2-stage gear-driven superchargers on their fighter aircraft. Merlin Mustangs, P-61 (used the same 2-stage engine as the F6F-3/-5), P-63 (used the 2-stage V-1710), with P-82 in pipeline together with lightweight Mustangs (no turbo, but 2-stage Packard Merlins or V-1710s). Even the XB-42 bomber, that was supposed to use 2-stage V-1710s. There was no new, bespoke fighter with a turboed engine in the pipeline by mid-/late-war, even the P-72 was to be using a 2-stage supercharged engine, without the turbo.

The only place for turboes were the big bombers.

Turbo-compound V-1710 S/C system used the engine-stage S/C plus the auxiliary S/C (= in effect a 2-stage supercharging), with the power recovery turbine (talk half of the turbo from the P-47) and water injection to reach 100 in Hg and 3000 HP down low; reduction of compression ratio from 6.65:1 to 6:1 helped out with the great boost levels there, too.

However, gearing the turbocharger to the shaft has some clear advantages - increased power recovery and elimination of waste-gate issues being particularly important.
Key point is illustrated by the graph below - using a power recovery turbine on the jet exhaust shifts you from a Harrier-style disk loading to something more like a tilt-rotor, giving you up to 3x the thrust (less at high speed) from the same exhaust energy.

Note that your graph notes hovering, ie. airspeed is zero. The faster we go, the greater effect of the exhaust thrust is to be obtained, while propellers' efficiency goes down. Main reason why the Harrier will never be that good in hovering as the Osprey, while the Osprey will never be as fast as the Harrier.
That way of 'upgrading' the power will be just fine for transport and attack aircraft (since the increase of thrust at low altitudes and low speeds there matters a lot), but it will be lacking when we want a ww2 fighter to really perform at 25000 ft and up. Especially with the small engine in question, like the V-1710 (or Merlin).
If the turbine is directly powering the engine, there is less power left to power the supercharger on the same shaft - the turbines on the turbocomopound engines were all without the compressor.

Some disadvantages of having a turbine run by hot exhaust gasses still remain - tricky retrofit on the existing fighters designed without it in the mind (P-39, P-40, P-51) still being the big one, plus the need that someone (GE, in reality) actually make the turbine and the shaft. The loss of exhaust thrust is still a thing, so eg. the theoretical 25-30% of increase in BHP is decreased by ~15% at 20000 ft. Installing a good, big S/C, preferably powered via the hydraulic drive (as it was the case with aux stage on the V-1710s with it) on the V-1710 will net us easily 15-20% power increase at high altitude, but with a far easier way to retrofit that engine on the existing fighters.
(similar thing happens when we compare the Merlin XX and 45 vs. the Mk.III, but here the increase was a result of the better intake elbow + better impeller design + higher rotational speed of the impeller)
The 2-stage S/C on the V-1710 will net us even more.

 

[EverKing]

@pdf27 I see, I misunderstood what you were saying and thought you were saying the engine should drive the turbine rather than the reverse.

Luckily, we don't need to guess or rely on conjecture for how these different systems compare. NACA did a full (post-war) study, published in November 1948 as NACA Technical Note 1735, "Performance of Exhaust-Gas Blowdown Turbine and Various Engine Systems Using a 12-Cylinder Liquid-Cooled Engine." This report used a V-1710-93 core engine (or at least one with the same specification and 9.5" primary stage S/C at an 8.1:1 drive ratio) with five different configurations:

1. Full Turbo-Compound (Blowdown recovery turbine followed by a steady-flow turbocharger)
2. Compound driven by a steady-flow only (@pdf27 's engine, less the swirl throttle)
3. Compound driven by blowdown only
4. Standard Turbocharged (like a P-38)
5. All geared S/C and exhaust stack jet thrust.

There are a lot of comparisons in the report but the key ones are here:

Thrust and Fuel consumption at altitude (best power):

Thrust and fuel consumption at altitude (best economy):

The S/C only engine (E) produces superior power to the Steady-Flow Compound engine (B) up to about 8000 feet or about 17000 when set for best econ; although the Steady-Flow Compound has superior (reduced) fuel-consumption at all altitudes in either condition. Full Turbo-Compound, (Blowdown Turbine with Power Recover to crank, followed by a Steady-Flow/Pressure turbocharger, feeding compressed air to an engine-driven supercharger) is superior to all others beginning at low/mid altitudes but at the cost of complexity and (history has shown) reliability.

This also show us that, excluding compounding, we can compare (D) and (E) (turbo vs. supercharged) and see that S/C is superior at lower altitudes but turbos blow past between 15000 and 20000 (in this case). Of course, this is comparing an Allison with Aux.Stage S/C to one with Turbocharging and is not comparing the superior S/C'ing of a late mark Merlin to a Turbo Allison.

 

[pdf27]

That's fantastic, thanks. What I have in mind is that the two relevant scenarios in the early war are (B) and (D): the USAAF aren't going to move away from turbochargers in the early war even if 2-stage superchargers plus an intercooler are the best technical solution to the problem.
I think we're left with my proposal of gearing the turbocharger to the engine through the existing supercharger drive, or the OTL choice of straight turbocharging. Mine is always going to be a hard sell, but I don't think it's impossible - you're essentially bolting an existing turbocharger onto an existing gearbox, welding the waste-gate shut and adding a novel throttle. If someone (presumably it would have to be at GE) had built one in say 1940, generated your graphs and showed them to NACA & the USAAF I think it has a chance.
Performance is promising: under best power conditions you achieve 10% more equivalent horsepower at altitude than the equivalent turbocharged system. At the same time fuel burn is reduced by 10%. You're going to have some challenges to get the turbine stage close to the engine for good packaging (meaning you have to reduce speed or actively cool it to keep reliability acceptable), but I don't think they're insurmountable.

A proper turbo-compound is always going to perform a lot better, but my view is that they weren't feasible early in the war and by the time they were the effort was better spent on jets.

 

[EverKing]

I agree that your proposal is a simple solution with good gains but with two little changes needed.

I think we're left with my proposal of gearing the turbocharger to the engine through the existing supercharger drive, or the OTL choice of straight turbocharging. Mine is always going to be a hard sell, but I don't think it's impossible - you're essentially bolting an existing turbocharger onto an existing gearbox, welding the waste-gate shut and adding a novel throttle.

[stress added] There's the issue. Because you'll have variable impeller speed and variable power output from the recovery turbine you cannot use a directly geared Supercharger drive for the power recovery. Doing so will mean you only gain anything at one specific power setting, at one specific altitude, for each engine speed. At all other settings and speeds the engine will be fighting the turbine or vise versa. This is why they used variable fluid coupling gearboxes in the real world compound engines, it allows the power from the turbine to transfer only to the engine regardless of the relative turning speeds and prevents the engine from loosing power back through the drivetrain to turn the turbine when there is a mismatch. Not a difficult problem to overcome, but it certainly prevents using a fixed gear transmission.

I also wouldn't weld the wastegate closed (or delete it, etc) because there will be lower-power conditions at low pressure altitude where the pressure delta between the turbine inlet (engine exhaust outlet) and turbine outlet (essentially the static atmosphere) will be too low to overcome the turbine friction and inertia to produce appreciable gains in recovery. In these conditions, you may discover that the turbine restricts the exhaust flow too much, produces appreciable back pressure, and reduces the engine breathing efficiency more than than the slight power you may be able to recover from the turbine. So, by still having a wastegate you can bypass the turbine at these low alt/low power conditions to avoid these additional losses and inefficiencies.

 

[tomo pauk]

To be clear, I'm not talking full forward on the throttle lever (in aircraft with automatic wastegate control) but rather a fully open throttle butterfly on the engine. This was the way USAAF turbo-charged aircraft were set up, once above the critical altitude of the engine itself (factoring in the integral supercharger which most had), the engine power/manifold pressure was regulated entirely by the turbo wastegate with the engine throttle left full open. In fact, in aircraft without an automatic wastegate control (early model aircraft for the most part) they used a separate lever to manually control the wastegate and the operating manual specified that to attain desired Manifold Absolute Pressure at altitude the throttles should always be left in the fully open position and MAP should be managed by adjusting the wastegates.

The engines with gear-driven compressors were also able to fly above the rated altitude with the throttle fully open, hence the term 'full throttle height' as the equivalent for the term 'rated altitude' as well as the 'critical altitude'. Eg. the Merlin 65 from this graph will have it's throttle open between 5500 ft to 10000 ft, and then again from 16000 ft upwards. Merlin III on 100 oct fuel can have it's throttle open from 9000 ft upwards; Mercury XV on 100 oct can have the throttle fully open above 9000 ft, too.

This contrasts to a two-stage supercharged engine where the engine throttle remains the primary means of controlling MAP. You could almost look at it as a Supercharged engine uses restricted airflow to reduce MAP, while a Turbo-Supercharged engine uses increased compression to increase MAP.

That should be 'increased pressure', not 'increased compression'?
Increasing the boost via turbo was indeed a thing, since there are now two superchargers working in series (talking about the ww2 engines); increse of boost served the purpose of making the required power at high altitudes despite the thin air there.

The two-stage engines behaved in similar fashion, working in series to provide the major increase of pressure ratio vs. a 1-stage engine (talk 5:1 vs. ~2.5-3:1, respectively) and thus providing the required boost. Throttle was used in the altitudes under the rated altitude, however the Allison system also used the hydraulic drive to keep the boost in the required range, while the P&W system used clutching and de-clutching system to control the aux stage (settings being off, low and high). A lot of power that was nominally 'lost' due to the need to turn the impellers ever faster as the altitude increased was gained via the much increased thrust of the exhaust gasses; at the end providing the very similar total thrust as the turboed engine's versions but with more compact packaging.

RR went balls to the wall, due to the Merlin (and later Griffon) being well suited for greater boost levels, that was also helped by having the intercooler (unlike the 2-stage V-1710).

 

[Draconis]

Developing some method of turbo-compounding early enough to be used in WW2 is an old topic of interest. Considering the performance improvements gained compared to mechanical supercharging and turbocharging, turbo-compounding offered some clear advantages. The thrust gained from the engine exhaust stacks is helpful but this must be weighed against the hundreds of horsepower consumed by the engine driven 2 stage superchargers at altitude. These were propeller engines operating at a speed range where all possible energy should be directed toward driving the propellers.

Here a link to an article that explores in some detail the energy available in the engine exhaust and a few other things as well. It includes a graphic illustrating the proportions of where the engines' energy is used or lost. https://www.enginehistory.org/members/articles/WasteHeat/WasteHeat.shtml

I explored turbo-compounding in an earlier posting in this forum. Interesting discussions. I was much younger then.

WW2 Earlier Turbo-compound aircraft engines

Turbo-compound aircraft engines were a significantly more efficient design. What if somebody had thought of this improvement in 1939 instead of 1944 too late to be used in WW2? The technology had existed since the early 1930s at least. All that was needed was somebody thinking of connecting an...

www.alternatehistory.com

 

[Andras]

I'm gonna repost two videos by Greg. One is turbo vs Supercharging, and the other is power recovery turbines.

At 36 minutes is a direct comparison chart between the systems on a R2800.

turbocompounding

I lied, a third video on the best supercharging system on a Corsair

 

[pdf27]

I agree that your proposal is a simple solution with good gains but with two little changes needed.

[stress added] There's the issue. Because you'll have variable impeller speed and variable power output from the recovery turbine you cannot use a directly geared Supercharger drive for the power recovery. Doing so will mean you only gain anything at one specific power setting, at one specific altitude, for each engine speed. At all other settings and speeds the engine will be fighting the turbine or vise versa. This is why they used variable fluid coupling gearboxes in the real world compound engines, it allows the power from the turbine to transfer only to the engine regardless of the relative turning speeds and prevents the engine from loosing power back through the drivetrain to turn the turbine when there is a mismatch. Not a difficult problem to overcome, but it certainly prevents using a fixed gear transmission.

I also wouldn't weld the wastegate closed (or delete it, etc) because there will be lower-power conditions at low pressure altitude where the pressure delta between the turbine inlet (engine exhaust outlet) and turbine outlet (essentially the static atmosphere) will be too low to overcome the turbine friction and inertia to produce appreciable gains in recovery. In these conditions, you may discover that the turbine restricts the exhaust flow too much, produces appreciable back pressure, and reduces the engine breathing efficiency more than than the slight power you may be able to recover from the turbine. So, by still having a wastegate you can bypass the turbine at these low alt/low power conditions to avoid these additional losses and inefficiencies.

I'm not 100% sure on that - currently digging through that NACA report to see how they were planning on gearing the turbine (not very clear) and the maths so far sort of fits with what I was expecting/remembering. Most helpful point is where they state that the gearbox efficiency assumed in their calculations between turbine and crankshaft is 95% - way too high for any sort of continuously variable gearbox even today.

With regard to your specific points:

1. Turbine efficiency should be a function of the ratio of gas inlet velocity to the turbine wheel (i.e. downstream of the diffuser) to wheel tip speed. At wide open throttle your mass throughput of combustion gases will be a function of RPM and boost pressure, and the volumetric throughput should be independent of boost pressure. There are some effects from the back-pressure or lack of it on the turbine exhaust, but I **think** that these probably won't affect the efficiency map much.
2. Not convinced about the wastegate given that low altitude/low power isn't really an operating regime you care very much about - the turbine is turning no matter what because it's geared to the engine, and you can always open the throttle a bit more if you're down on power. Leaving it in is certainly easy to do, but my suspicion is that it'll be simpler and lighter to delete it.