16 replies to this topic

[anbeck]

Hi there,

I was just wondering if there is a theoretical maximum to produce power with a given discplacement.

Say, we have a displacement of 3 litres and the same fuel as today (that is, the same amout of energy stored in 1 kg). You would think that there is only so much fuel you are able to burn in this engine per minute, all depending on rpms, I guess. So today's engines are just trying to maximise efficiency to reduce the loss of energy per kg of fuel (anybody knows, how much % today's F1 engines get out of the fuel's energy?)

Is this ok so far (I don't know much about chemistry and physics)? Are the main limiting factors the rpms (how much fuel can be burned in a given time) and efficiency (how much of the energy is used to push the car)?

Could one say that there is a limit a 3litre engine can possibly produce?

Thanks for anybody having the patience to enlighten me on this one

a.

[TDIMeister]

Quite simply, power is proportional to not only engine RPM and displacement, but also the effective work that each cycle delivers per unit of displacement. In engine terms, this is referred to as the mean effective pressure (MEP). The MEP is a function of the fuel conversion efficiency of the engine, the volumetric efficiency and the heating value of the fuel mass that is drawn into the combustion chamber to be burned in each working cycle.

If I haven't already lost you, read on. If you use the same fuel and impose the same displacement, power output can only increase by either an increase of the RPM, as F1 dramatically demonstrates, or increasing the MEP and its dependent variables. For naturally-aspirated engines, the MEP is limited by the physics and flow dynamics of cramming as much atmospheric air as possible. The benchmark engines have brake MEPs of around 16 bar at peak torque and somewhat lower at peak power (many contemporary F1 engines fall in this range), while some have claimed (like in Super Touring and NHRA Pro Stock classes as much as 17.5 bar). This limit has stayed with us since at least the 1960s and will unfortunately not be breached by a significant degree barring significant innovation in reducing friction losses in the engine and increasing the thermal efficiency (e.g. via direct injection, increased compression ratios and fully variable valvetrains).

Forced induction is a different story; here the limit is only the onset of knock for spark ignition engines as well as thermal- and mechanical stress limits dictated by the engine design. Highly boosted gasoline street engines will be pushing 27 bar BMEP in the near future. Supercharged drag racing and Diesel tractor pull classes demonstrate extreme cases of forced induction giving sky high BMEPs exceeding 100 bars, but this is not achieved in engines that require any semblance of endurance.

[Canuck]

Originally posted by TDIMeister
Highly boosted gasoline street engines will be pushing 27 bar BMEP in the near future.

HA! Had to read that twice to conciously process the "BMEP" portion. Had me all agog for minute, not that 27 bar BMEP is anything to be less than keenly interested in.

[Greg Locock]

If we stick to NA engines (otherwise the thing just turns into the combustion chamber for a gas turbine), what is the highest volumetric efficiency that can be achieved?

[Bill Sherwood]

Originally posted by Greg Locock
If we stick to NA engines (otherwise the thing just turns into the combustion chamber for a gas turbine), what is the highest volumetric efficiency that can be achieved?

I'd guestimate about 130%, maybe a little more but certainly over a relatively narrow range of revs without variable length inlet/exhaust, etc.

[Greg Locock]

Ok, and the best fuel for a given amount of oxygen is hydrogen, assuming that we aren't going to 'burn' explosives, based on MJ/kg.

Or is it?

[TDIMeister]

For the measure of engine performance, it's not the fuel heating value itself but the mixture heating value that is of interest. That's because the volumetric efficiency is fixed by other geometric parameters (although charge cooling will certainly influence VE). Hydrogen has a high LHV, but when delivered to the engine through the intake manifold, its displacement of air means that for a given VE, the heating value of the inducted charge is low. That's one of the reasons why H2 ICEs like the BMW Hydrogen 7, has a low horsepower figure, even being a supercharged V12. Direct injection of H2 changes everything, however, and raises the mixture heating value per kg of trapped air significantly.

The mixture heating value is less a function of the fuel heating value than it is of the stoichiometric air-fuel ratio. For example, nitromethane has only a little more than a quarter of the MJ/kg of gasoline, but the stoic. AFR is 1.7:1 compared to 14.7:1 for gasoline, so that a stoichiometric mixture of nitro in a given kg of air is far more energetic than a stoichiometric mixture of gasoline in the equal kg of air. Fuels with low stoic. AFRs tend to yield higher mixture heating values because more fuel -- the stuff that actually supplies the heating energy -- can be burned in a given mass of air, and the drop off in LHV is outpaced by the increased fuel mass.

OTOH, many teams in the past have played with fuel compositions to get the optimum combination of octane rating, heating value, density, etc., to maximize their advantage. Density is particularly interesting because all the talk about mixture heating values are in mass terms, but how much fuel you can actually carry on board is on a volume basis.

[J. Edlund]

Net heating per kg of air (typical values)

Hydrogen: 3.53 MJ
Methanol: 3.08 MJ
Ethanol: 2.98 MJ
Gasoline (regular): 2.89 MJ
Gasoline (premium): 2.96 MJ
Toluene: 3.03 MJ
Diesel: 2.93 MJ
Nitromethane: 6.65 MJ
Methane: 2.91 MJ
DME: 3.20 MJ

[Greg Locock]

OK, so which of those burn fast? Are we going to exclude nitro and similar fuels on the basis that we could just use an explosive or some other self oxidising fuel if we were being silly?

[J. Edlund]

Originally posted by Greg Locock
OK, so which of those burn fast? Are we going to exclude nitro and similar fuels on the basis that we could just use an explosive or some other self oxidising fuel if we were being silly?

Hydrogen burn really fast, it's flame speed is about ten times that of gasoline. Nitromethane also burn a bit faster than gasoline, but generally, the differences between the fuels mentioned above are not that large.

The problem with hydrogen is that a high output would require direct injection, partly to go around the issue with reduced VE and partly to avoid the danger of backfire.

[Joe Bosworth]

Engines are nothing but air pumps. Air contains 21% oxygen. Oxygen combines with the combustables in fuel to create energy. Energy is converted to HP.

The amount of air that can go though and engine is only a function of capacity, RPM and volumetric efficiency. (Unless you are blowing one way or another and then you have to add the blowing pressure and subtract the added blower driving energy).

If you disregard adding oxygen to the air by chemical means you can only convert any fuel at the rate of about 100 BTU of heat per 1 cubic foot of air. Use 1 hp is made for every 42.4 BTU/minute.

If you are going to have oxygen in the fuel then you have to nominate what the fuel is or at least how much oxygen it contains and then add that to the oxygen input.

Then you have to multiply by the net efficiency of the engine after taking into account friction, heat and pumping losses to get flywheel HP.

It really is quite easy. What do you think is the theoretical maximum capacity, RPM, volumetric efficiency and theoretical minimum internal losses?? You tell me that and I will pass my slide rule over the rest to tell you what the Theoretical Maximum HP is.

At least the original question pegged volume at 3 liters.

If the question was based on what the practical max HP was it might be answerable but the theoretical max has no answer short of approaching 100% VE and 0% internal losses at approaching infinite RPM.

All assuming being able to use theoretically infinitely expensive materials.

Regards

[Joe Bosworth]

Seeing as how nobody has bit on my post above, I will give you an example:

Say you have a normally aspirated 3 liter engine capable of 20,000 RPM you could theoretically make just short of 2500 HP on pump fuel containing no added oxygen.

But we "know" that the last of the 3 L F1 engines were putting out something less than 1000 HP.

This would put their efficiency at about 40% or put another way 60% is lost in various forms. Interestingly, this is about the same efficiency developed by NASCAR engines! MotoGP engines are also very close to the same efficiency.

Lot's of room for more if someone can figure out how to increase VE, revs or decrease internal losses.

Regards

[Greg Locock]

The 'obvious' one to improve is the thermodynamic efficiency. Good luck with that. Limits are metallurgical (burning up pistons and exhaust valves) for the hot end, and are limited by the expansion ratio at the cold end, unless you go to an overexpanded engine, which has a direct hit on your power output.

Improving VE at one speed above 130% may be possible with a high Q intake system, I suspect your Q is limited by the friction seen in the flow through the valve.

Increasing the rpm is just a materials problem, again good luck with that.

[Canuck]

I thought I'd read here somewhere that at X RPM, the airflow became super-sonic and gains were not to be had with increased engine speed. I could be right out to lunch about that mind you, but I thought I'd read it here once upon a time.

If we took all the FIA-imposed material rules and tossed them, we might find considerably more power via engine speed and reduced component weights. AlBeMet was only beginning to be utilized in the bottom end assy. before it was turfed. There are likely some other seriously $$$ materials that might have similar advantages (high strength, low density), perhaps some with better thermodynamic profiles (but I don't know of any).

[Greg Locock]

You are absolutely right, compressibility (speed of sound effects) will be an issue. It'll show up in the VE on the intake side, and I suppose if the exhaust chokes the VE will drop as well, as it won't empty properly.

[Joe Bosworth]

Port air velocities are just one of the things where you move between theoretical and practical.

To keep velocities down as you add revs you just theoretically make the ports larger and keep getting VE and therefore HP.

The practical part comes in that the very big ports chop off lower speed torque and therefore driveability.

But the subject of the thread was what was theoretically possible.

Keeping to the practical I will repeat a fact that I hit on several decades ago and which keeps popping up to this day. That is that max HP for practical engines almost always comes up where the avarage velocity through the inlet valve is 80 meters per second. There have been a few examples where this velocity has been highly exceeded but not many. The "normal" range is within plus 10%.

Regards

[J. Edlund]

Originally posted by Canuck
I thought I'd read here somewhere that at X RPM, the airflow became super-sonic and gains were not to be had with increased engine speed. I could be right out to lunch about that mind you, but I thought I'd read it here once upon a time.

This is generally more related to piston mean velocity than engine speed alone.

You also won't reach mach 1 in the inlet of a practical application (except during reverse flow), but you can have peaks in the inlet ports of mach ~0.6 during intake and average inlet velocities around mach 0.4 (125 m/s). This at around 25 m/s piston mean velocity. Going faster than that would not affect mach numbers significantly, but the volumetric efficiency will suffer.

Larger ports have generally no significant negative effect on low speed torque, but tend to increase power at engine speeds the engine is tuned for, and decrease power at speeds the engine isn't tuned for. If designed for a certain output, the engine with larger ports will generally have a better torque curve (as that allow milder valve timing).