Port size/ low end torque correlation

[69911e]

Why is there a inverse correlation beween intake port size, and low end torque when MFI is used?
My understanding (probably wrong) is the low velocity of the intake charge causes the fuel to be improperly atomized or pool up on the walls. This makes sense with a carborated syetem, or even EFI where the injectors spray before the ports. With MFI, the injectors are far into the port. By the time the fuel disperses, the charge is at the valve.
From what I have seen in 911s, the port size near the valve is the same in a "T" or an "S" head.
Any insight into the loss of low end torque?

[Tim Walsh]

Inquiring minds want to know..

I'm building a 2.4 on the ragged edge of port sizes, 32mm and peak HP@ 6300rpms

[911s]

Velocity of air moving through the port. In order to move a measured volume of air through a smaller port means that it has to move through there faster than it would through a larger port. ie: 450 cc cylinder draws air in. Air has to pass though port. 450 cc of air has to move more quickly through 32 mm port than 36 mm port, in order to fill the cylinder in a set amount of time. This is a very simplified explanation, but I'm sure you get the point. Basically, increase air velocity into engine to increase low end torque, increase air volume to increase top end horsepower. Trick is to increase volume moving through port without decreasing velocity (or perhaps even increasing it).

[Bobboloo]

Ever noticed how it takes very little effort to take a sip of a drink with a small straw as compared to a big straw. Conversely when your real thirsty and want a gulp you need a big straw.

[beepbeep]

Ehh...it actually has to do with Volumetric Efficiency [VE]...high speed flow "rams" air into the cylinder better at lower RPM's. When flow is slow, it will lazily sip into the cylinders and won't fill them properly untill you start to rev and it starts gulping more air.

With turbocharged cars it's all diferent. You have a "force feeder" and all obstacles are only in the way.

[jluetjen]

I've come to the conclusions that in the case of the 911 (which has pretty efficient ports to start with) that there are at least 3 major factors in the port and intake design that affect engine performance. If you stick with me to the end, you'll see my answer to 69911e's question.

1) Is the port big enough without being too big. I've come to the conclusion that there is a fairly wide "sweetspot" in port dimensions which will give essentially optimal performance. You can get a sense of this range by seeing how much air an engine is pulling at peak torque and peak HP. In the case of non-CIS NA 911's, the range at peak torque is 55 to 75 meters/second. Note that the range is 65 m/s +/- 15%. There are successful engines at the low end of this range ('71 ST) and successful engines at the high end of this range (73E and the 2.7RS). There are also successful examples outside this range (Kurt's 3.2 at 79.6 m/s). The diameter also seems to affect the flow at higher valve lifts which is an important factor if you are using an S cam or something more radical. Smaller ports just don't flow at higher lifts.

2) Port Shape: By increasing or decreasing the cross section of the port, the gas speed in that area can be increased or decreased. Which can be helpful in negotiating the turn, fitting past the valve guide or going past the valve head. The shape over the whole length also affects the accoustics and harmonics of the intake system which can come into play at certain revs. It also affects the turbulance and swirl inside the cylinder which has a big impact on combustion efficiency.

3) Port volume, especially in the bowl area. Keep in mind that a 4 cycle engine does not ingest air at a constant rate like a turbine, but ingests air in "gulps". Out of the 720 degrees in a complete cycle, the valves are closed at least 2/3 of the time, even in an S. In general, the intake path's volume is greater then a cylinder's volume, so in each cycle only a portion of the air in the intake port and manifold is ingested. So the volume of air close to the valve (ie in the bowl) I suspect has a big impact on the flow into the cylinder. Once the intake valve closes, the rest of the intake system can take the other 2/3 of the time to catch up and balance itself which goes back to the accoustics of the system.

Going back to '69911e's question, with carbs, the fuel is introduced into the port fairly high in the system. As a result everything downstream of the carb is "wet" and has to handle the flow of a fuel air mixture which has significantly more mass then just the straight air in an MFI'd car. When the velocity of the mixture drops, the fuel has a tendency to fall-out and settle on the bottom of the port. The extra mass of the mixture also tends to make it not want to change directions or velocity as fast as comparatively less massive pure air.

In a sentance, it would seem that injection systems where the injectors are relatively close to the intake valves will be less sensitive to inefficiencies in the port design. If you add in the fact that MFI systems only introduce the fuel around the time that the intake valve is open and the superior atomization resulting from the MFI's high injection pressure, and you wind up with an engine that can withstand slow gas speeds much better then a carb'd engine. Since it doesn't have any venturi and the fact that it can withstand oversized port dimensions means that an MFI system will outperform carbs at peak rev's too.

When I've compared the torque curves of MFI'd engines with non-MFI'd engines, this is what I saw. While the peak torque will be about the same, the MFI'd engine will generate more torque at the two extremes of the RPM range. The extra torque in low rev's increases the flexibility of the engine while the increased torque at high RPM's results in more HP.

[beepbeep]

Originally posted by jluetjen
If you add in the fact that MFI systems only introduce the fuel around the time that the intake valve is open and the superior atomization resulting from the MFI's high injection pressure, and you wind up with an engine that can withstand slow gas speeds much better then a carb'd engine.

I beg to differ. EFI dyno pulls with batch and sequential fired injectors on otherwise same engine showed no difference in torque/power whatsovever, just slighlty better emissions at lower revs.

[304065]

Goran, I don't understand your point. John points out that carbureted engines suffer from fuel falling out of suspension in the intake tract, and that MFI doesn't suffer from that due to introducing the fuel closer to the intake valve.

What does EFI have to do with it? Are you suggesting that EFI systems that locate the injector further from the intake valve have the same droplet size and consistency as a carburetor, and that it doesn't make a difference to torque and horsepower across the engine's operating range compared to MFI?

[jluetjen]

I think Goran got hung up on my suggestion that the injection timing is critical. That wasn't my point. In that regard I agree with Goran, that purely from the perspective of introducing fuel into the intake, CIS is virtually as good as MFI. My point was only that fuel can't puddle in the intake if there isn't any fuel there. But I agree that this is an insignificant benefit as proven in numerous situations.

Now running an injection system (any injection system) with only injectors high in the stacks at low RPM will most likely get you no improvement over a carb for the reasons that I stated. This is why many of the top shelf injection systems have injectors close to the valves (for low RPMs) and high in the stacks (for high RPMs).

[69911e]

John: Do you think the small amount of loss in low end torque on a motor with big ports may be that with big ports people normally put in big stacks?
I realize that the stacks are tunned for high RPM, not low RPM. As the stacks and throttle bodies are made bigger the air contatined within gets larger, and the velocity gets proportionally smaller. The energy contained within the moving intake charge is E~MV^2. Therefore a smaller amount of energy available to charge the intake when big stacks are used.

[jluetjen]

Ed;
I'm sure that stack dimensions has something to do with it. Generally, the longer the stack, the LOWER in the rev range and the stronger the accoustic affect. I was recently reading "The Design and Tuning of Competition Engines" by Philip Smith and he had a dyno chart from Jag E-Type motor with different stack lengths on it. The longest stacks had the tallest torque curve at the lowest engine speed. As the stacks got shorter, the torque curve got flatter and lower to the point when no stacks were used, the torque curve was almost a straight line.

This makes sense if you consider a flute. You can tune a flute go down in pitch (lower frequency) a few ways:

Make the flute longer
Make the diameter bigger
Change the flute's shape conical as opposed to a closed cylinder.

I've posted previously exploring harmonics and accoustics in relation to stack lengths. Don't forget that there are also harmonics so that a flute (or intake) tuned to a certain pitch can also produce tones at multiples of the frequency.

Anyhow, you start to have a chain affect if you want to run a high lift camshaft like an S cam. To get the full benefit you need to increase the port diameter so that it can flow across the full range of valve lift. This makes the "tone" of the intake lower which may be a good thing since at low RPM's as you point out the flow in the intake will be comparatively slow. Since the flow slows down, it has less energy at the valve face. And then keep in mind that it's not clear to me that the sound energy in the intake follows exactly the same rules as motion energy.

At the end of the day, I have no means to compare the trade-offs. Maybe someone else with academic connections will be able to quantify it.

[69911e]

John: Of course the longer the stack the lower the resonant frequency. I did not realize an increase in diameter also lowered the resonant frequency(I will need the think about why). I think I was not clear on my idea.

I was simply stating that as the volume of the stack goes up, the velocity goes down by the same proportional amount. Energy contained within the renonant pulse is proportional to mass, and Energy contained within the renonant pulse is proportional to velocity squared. Therefore as the stack is bigger in diameter, the energy contained in the resonant pulse (all else being constant) will be reduced.
i.e. If the stack has a cross sectional area increased by 50%, the velocity decreased by 1/1.5. Then the Energy contained in the pulse will be 1.5/((1/1.5)^2)=.66. A reduction of 34%. I think it is the energy contained in the pulse will translate to a proportional increase in pressure at the valve (to a first order approximation)
Then again I could be way off, I don't know much about this engine power stuff.