109 e3b against spitfire II

[Viper2000]

Go back to the paper given by Lovesey which I posted earlier.

Engine shaft power is proportional to mass flow rate.

Mass flow rate, W, is given by

W=0.422*Ncylinders*(Pcharge-(1/6)*Pexhaust)/Tcharge

Charge temperature & pressure are measured in the intake manifold. In this equation, W is in lb/minute; piston engines are tiny.

You will note that Lovesey cites a 7% improvement in supercharger pressure ratio from injecting fuel into the eye of the supercharger.

This is very significant given the pretty awful isentropic efficiency of the supercharger.

You can go through the data, and calculate the supercharger work from the efficiency curves in Figure 11.

You can calculate the actual engine air consumption iteratively by assuming that the FAR is about 1/12 at high power. Hence, given the SFC curve and the full throttle power vs altitude curve, you can use the fuel flow to calculated the total rate of charge consumption.

You can use the total rate of charge consumption to calculate the supercharge power consumption as a function of inlet temperature. You must of course add this supercharger power consumption to the brake power in order to calculate the shaft power, because it is the shaft power, not the brake power, which is directly proportional to fuel flow.

This means that you'll need to iterate in order to achieve convergence.

Try this with and without fuel injection into the eye of the supercharger, which may be modelled as a 25 K temperature reduction exchanged for a 1/12 mass flow rate increase.

Because the supercharger work is W*Cp*deltaT, you will find that injecting fuel into the eye of the supercharger results in a considerable reduction in the supercharger power required for any given boost pressure, which naturally improves brake power and brake SFC.

Your argument regarding cylinder temperature is spurious because the engine has a cooling system to maintain CHT, and because the reduction in induction manifold temperature results in a considerable reduction in the charge temperature during the compression stroke because compression through a fixed volume ratio results in a fixed temperature ratio rather than a fixed absolute temperature increment.

This means that there is less compression work.

Peak cycle temperature is essentially fixed by dissociation, and therefore the reduction in charge temperature translates directly into an increase in BMEP.

Alternatively, you could hold constant charge temperature and reduce the size of the aftercooler. Either way, you're still getting a benefit from the latent heat of evaporation of the fuel; but this benefit is greater overall when the fuel is injected into the eye of the supercharger.

Arguments about stoichiometry as less important for an aero-engine at high power than for a car engine because you're not bothered about emissions (at least in this period). Therefore you run the whole thing rich of peak.

Cylinder to cylinder variation in FAR will be small so long as the induction manifold temperature is kept reasonably high; this may be seen from the discussion about lead fouling towards the end of Lovesey's paper; charge distribution is good down to intake manifold temperatures of about 35ºC.

Cylinder to cylinder charge consumption will vary due to intake manifold aerodynamics.

But this actually means that if you go for 1940s DI you'll get a variation in FAR from cylinder to cylinder because the injection system would give each cylinder equal fuel irrespective of its actual air consumption. Modern engines would use an oxygen sensor in the exhaust to maintain stoichiometry. However, this would preclude operations rich of stoichiometric, so the benefit is to BSFC and emissions rather than to absolute power.

I have flown a Citabria with modern after-market engine instrumentation. Obviously the CHTs vary because it's air cooled; you're never going to get the back cylinders as cool as the front ones. Likewise, mixture distribution will always be questionable for a naturally aspirated engine, even on a hot day in South Carolina.

That sort of engine would obviously benefit from DI; and in the modern world, the philosophy is to turbo-normalise if altitude performance is wanted. The convenience of maintaining a fixed engine operating point independent of altitude is considerable. Modern compressors are very much more efficient than those from the 1940s, and therefore there is less compressor work to save in the first place. Additionally, with substantial exhaust energy being dumped out of the waste-gate, there is obviously less motivation to reduce compressor work. So there are a variety of factors driving modern engines towards DI.

This does not mean that single point injection upstream of a supercharger does not have advantages, especially if you have a high degree of supercharge.

Modern piston engines simply have different design goals than the high-powered piston aero-engines of the 1940s.

[Kurfürst]

Originally Posted by Viper2000

Because the supercharger work is W*Cp*deltaT, you will find that injecting fuel into the eye of the supercharger results in a considerable reduction in the supercharger power required for any given boost pressure, which naturally improves brake power and brake SFC.

Pretty redundant if you ask me, if you are already using a variable speed supercharger adjusted already for supercharging needs, as DB did with its barometrically controlled hydraulic clutch.. the supercharger wasn't making less waste, it made almost no waste at all.

Correct me if I am wrong, but injecting fuel into the supercharger eye reduces charge temperature, delaying detonation point. Direct fuel injection does the same (in the combustion chamber), but later, and with better fuel effiency, no risk of backfires, and no negative G problems. High octane fuel is a pretty expensive agent for charge cooling..

[Viper2000]

You're confusing throttling losses, which are avoided by varying supercharger speed, with the aerodynamic losses associated with the supercharger's design, which are not.

The aerodynamic losses increase the temperature rise associated with a given pressure rise, which increases the work required.

Because compression through a given pressure ratio tends to produce a given temperature ratio, you reduce the absolute deltaT by reducing the initial temperature, all other factors remaining constant. This reduces compressor work, which is equivalent to an increase in compressor efficiency.

You obviously get a greater benefit from cooling the working fluid at the start of the compression process.

Eg

Start at 288 K. Cool by 25 K. T = 263 K. Compress through a temperature ratio of 1.5, and then further through a temperature ratio of 2. Temperature = 263*1.5*2 = 789 K. Delta T from start = 501 K.

Compare with:

Start at 288 K. Compress through temperature ratio 1.5, cool by 25 K and then compress through temperature ratio of 2. Final temperature is then 814 K, so delta T is 526 K.

Thus, compressor work differs by 5% or so in this example.

This is the reason for the reduction in the isentropic efficiency associated with a given polytropic efficiency as compressor pressure ratio increases.

[Crumpp]

Direct fuel injection does the same

Injecting into the eye of the supercharger is no more efficient than a carburetor at fuel metering.

Eg

Start at 288 K. Cool by 25 K. T = 263 K. Compress through a temperature ratio of 1.5, and then further through a temperature ratio of 2. Temperature = 263*1.5*2 = 789 K. Delta T from start = 501 K.

Compare with:

Start at 288 K. Compress through temperature ratio 1.5, cool by 25 K and then compress through temperature ratio of 2. Final temperature is then 814 K, so delta T is 526 K.

Thus, compressor work differs by 5% or so in this example.

You are still getting the compression from a supercharger and directly cooling the combustion chamber with direct fuel injection.

I am sorry Viper but it does not change the fact you cannot precisely meter the fuel thru an intake, either.

Modern engine designs are now starting to employ direct-injection.
Here fuel is introduced straight into the cylinder rather than
being premixed outside. This technique has major advantages
as it allows more accurate control of the quantity of fuel used and
the injection timing, leading to an increase in the vehicle’s
horsepower combined with a lower fuel intake.

Direct-injection is far more efficient than traditional
injection techniques, raising performance levels while requiring less fuel.

http://www.onsemi.com/site/pdf/PSDE_0411.pdf

[Viper2000]

Originally Posted by Crumpp

Injecting into the eye of the supercharger is no more efficient than a carburetor at fuel metering.

Of course it's better than a carburettor; you're not reliant upon venturi suction to get the fuel into the airflow; you're positively chucking it in with a pump.This will provide considerably better matching across the flow range.

Because you're injecting fuel under pressure, you can positively atomise it, achieving a considerably lower Sauter mean diameter of fuel droplets than is possible with a simple carburettor, which means that it will evaporate much faster. This means that more of the temperature drop happens earlier in the compression process, which increases the overall efficiency bonus.

Once the fuel has evaporated, the mixture distribution problem goes away.

Originally Posted by Crumpp

You are still getting the compression from a supercharger and directly cooling the combustion chamber with direct fuel injection.

This is inherently less efficient. It's relatively simple thermodynamics.

Originally Posted by Crumpp

I am sorry Viper but it does not change the fact you cannot precisely meter the fuel thru an intake, either.

Fuel metering is not a particular problem. You measure the flow through the intake, and inject fuel in proportion thereto. It's much easier to get this measurement & injection process right at a single point than it is to get it right at 12 points.

Fuel distribution may be a problem at low manifold temperatures where the fuel fails to evaporate fully, but that is a separate problem.

You seem to be mostly hung up on fuel metering issues, which certainly exist for naturally aspirated engines with carburettors, especially away from their design point.

However, fuel metering is not especially important at high power if you don't care about emissions. You run rich of stoichiometric, and fuel flows say +/- 5% won't make a great deal of difference to power output. Obviously the SFC is pretty bad at that point; you can clearly see this on the SFC curve in Lovesey's paper.

Mixture distribution is not a problem at high induction manifold temperature.

The reductions in supercharger work and intercooler size are far more important than the slight increase in fuel mass fraction which you might suffer from the need to keep the leanest cylinder sufficiently rich to avoid detonation. The cost of a single point system is far lower than a multi-point system, and the fuel pressure required is lower than for true direct injection. (Port injection is a pretty horrid compromise which only makes sense if the alternative is a carburettor which would produce bad mixture distribution.)

For a supercharged spark-ignition aero-engine, operating at a fixed non-dimensional power setting, provided that you've got enough induction manifold temperature to avoid condensation, the mixture distribution will be good and the single point system wins.

Multi-point FI is an expensive solution to mixture distribution problems. It is great for naturally aspirated engines, and probably pragmatic for turbo-normalised engines, especially if the engine manufacturer isn't responsible for the turbocharger.

But if you're using a mechanical supercharger and will mostly operate the with reasonably high induction manifold temperatures, then there's no great mixture distribution problem unless your induction manifold is horrible, so multi-point injection offers limited benefit, whilst single point injection into the eye of the supercharger reduces supercharger drive power requirements. So single point injection is a pretty obvious choice.

Now, if you're designing a sports car engine, you might supercharge it to get high power, but most of the time it would operate at very low non dimensional power settings, so mixture distribution would be a major problem with single point injection, and therefore you'd probably go for multi-point FI.

But that's because the sports car engine isn't really designed for high performance. It's designed to make an expensive noise and very occasionally provide bursts of acceleration to impress the girl in the passenger seat. Most of the time it's practically ticking over, and so you're much more bothered about part-load characteristics than would be the case for an aero-engine. You're also trying to meet modern emissions regulations, which means that you're paranoid about stoichiometry so that you don't poison your catalyst. It's a totally different world, with different trades and different drivers.

[TomcatViP]

Originally Posted by Viper2000

But that's because the sports car engine isn't really designed for high performance. It's designed to make an expensive noise and very occasionally provide bursts of acceleration to impress the girl in the passenger seat.

I hve to disagree loudly with the above comment. It's the worst thing to do Having a fat diesel with a body kit is what impress the girl nowadays

[Crumpp]

http://www.slideshare.net/rjperforma...rj-performance

It works in any engine to increase power and performance over any other fuel metering system no matter if the intake is supercharged or not.

http://www.steerbythrottle.com/hccyong/files/DFI.pdf

Advantages of direct fuel injection

Combined with ultra-precise computer management, direct injection allows more accurate control over fuel metering (the amount of fuel injected) and injection timing (exactly when the fuel is introduced into the cylinder). The location of the injector also allows for a more optimal spray pattern that breaks the gasoline up into smaller droplets. The result is more complete combustion -- in other words, more of the gasoline is burned, which translates to more power and less pollution from each drop of gasoline.
Disadvantages of direct fuel injection

The primary disadvantages of direct injection engines are complexity and cost. Direct injection systems are more expensive to build because their components must be more rugged -- they handle fuel at significantly higher pressures than indirect injection systems and the injectors themselves must be able to withstand the heat and pressure of combustion inside the cylinder.
How much more powerful and efficient is direct injection?

Cadillac sells the CTS with both indirect and direct injection versions of its 3.6 liter V6 engine. The indirect engine produces 263 horsepower and 253 lb-ft of torque, while the direct version develops 304 hp and 274 lb-ft. Despite the additional power, EPA fuel economy estimates for the direct injection engine are 1 MPG higher in the city (18 MPG vs 17 MPG) and equal on the highway. Another advantage: Cadillac's direct injection engine runs on regular (87 octane) gasoline. Competing cars from Infiniti and Lexus, which use 300 hp V6 engines with indirect injection, require premium fuel.

http://cars.about.com/od/thingsyoune...tinjection.htm

A good primer article on Direct Injection. It explains very well the difference between the various types of fuel injection and why Direct Injection is the ultimate fuel metering system for power and performance.

http://www.driverside.com/auto-libra..._injection-350

Here is a good article that explains the German Direct injection systems in easily understandable terms:

http://www.motorcycleproject.com/mot...xt/inject.html

The German systems were far from perfect but they certainly did their job and allowed them to level the playing field in terms of aircraft engine performance.

IIRC, the British and United States did a combined effort to develop a Direct Injection engine that was used in a tank at the end of the war.