gas speed analysis

[Tim Walsh]

I'm working through my engine rebuild and I was playing around with jluetjen's equations and I've got a few questions. Does the gas speed equations account for gas speeds throughout the whole power band or just at the optimal power RPM? I was wondering since it would seem that you would want your gas speeds to on the upper end of the optimal gas speed range (70m/s to 100 m/s).

My thinking for this is that if you have gas speeds closer to 100 m/s at your max power RPM then you will still have optimal gas speeds much farther down the rev range.. IE if you have gas speeds of 100 m/s at 6800RPMS (S cams) with 33.56mm ports then the optimal gas speed range starts at 4800 RPMS. BUT if you have ports of 38mm your optimal rev range would be between 6200 and 6800.


The ultimate question is for optimal gas speeds should the gas speeds at your maximum RPM be near 100 m/s or is anywhere in between 70 and 100 m/s just fine? and how does gas speed vary across the rpm range? constant? linear? log?

[Type961]

Hi Tim,
Moving the 110 m/s (Mach 0.5) RPM too high will give you a poor low RPM tork and you won't really win extra power at higher revs. This is because your port and valve design won't let you get a good efficiency at high revs. This is why 911 often have their stock max power at about 6200 rpm and it's hard to maintain good tork higher.
To increase you're 911 peformances, I suggest you to obtain 110 m/s at your actual best power RPM (ie 6200 RPM) and let the tork fall for higher RPMs. If you want to get really more tork higher, you have to change valves and/or cams....

regards,
Luc.

[Tim Walsh]

Luc,
I've worked my engine out at this point. It will be a hi compression E engine with 32 or 33mm ports. That will leave the gas speeds right at 100 m/s. I was wondering more about race engines than what I'm building. I have a friend building a 2.4L engine with 40mm ports and probably GE60's with those #'s his gas speed at 6200RPMS is about 75 m/s. That's on the low side of the optimal range and I was trying to figure out if that would be too much porting.

[Type961]

Did you calculate 75 m/s at valve ? It's impossible at this rev.
Since the valve has a 55% average lift out of 260° you can't have the same stream at the valve.
So I think having 75 m/s @ 6200 RPM at port while the stream speed is probably 120 m/s at the valve is non-sens ! You will never have more than something like 80 m/s at port.... wasted inertia

Keep a good inertia and get a good global 80 to 110 m/s from plenum (or carb) to valve at your 6200 RPM. You will obtain a smooth, powerful engine !

[Tim Walsh]

I know that all of those measurements are at the port. I'd think that the gas would have to speed up a fair amount at the valve.

John did some calculations and found out that basicly between 100 m/s is the limit of the gas speeds at the port, at least that's what the factory did.

[jluetjen]

Hi Tim;
After I did the (fairly simple) analysis to come up with the 100 m/s number, I did some more work and I think refined the concept a little more. To come up with the 100 m/s number I just calculated the gas speed at peak HP and peak torque for every carb'd or MFI'd configuration that I could get me hands on. Low and behold, virtually none of them had a speed faster then 100 m/s. Hmmm.

Since then I've been able to get the torque charts for a number of configurations (both street and race) including carb's and MFI again. I noticed that in almost every case the peak torque happens at about 80 m/s (+/- 5 m/s) (at the ports). Above that the torque drops but the HP continues to increase up to the peak HP speed which as I pointed out earlier is almost never above 100 m/s.

Hmmm....

When I've checked the gas speed at the valve, I've generally found that 911's have very generous valve sizes. BTW - this is not uncommon with hemi-headed engines. So this is why I have not spent a lot of time worrying about the valve size. Now if you're trying to build a 2.7 engine with early (small valve) 2.0 heads, it might be worth while to check the gas speed at the valve and potentially install larger valves. But for most of the 911 engines, it valve size doesn't seem to be an issue.

That's what I read from the data. Others may have drawn their own conclusions.

[dtw]

So what does that say about gas speed at different RPMs? I've run your calculation for my heads, and at 40mm intake port and peak HP RPM I'll be running ~70m/s. Does this mean when my revs are lower, the engine will be running ineffeciently (ie, below 70m/s)? I can back into 100m/s with a 34mm port. Would this be more optimal? If so I can probably put throttle bodies on that are 34mm, effectively restricting flow at the t-body.

[jluetjen]

So what does that say about gas speed at different RPMs?

The function is linear with a fixed slope. So if you turn the engine twice as fast, the gas speed will double.

at 40mm intake port and peak HP RPM I'll be running ~70m/s. Does this mean when my revs are lower, the engine will be running ineffeciently (ie, below 70m/s)?

The key point seems to be the gas speed at the peak torque engine speed, not the peak HP engine speed.

In general engines are most efficient at creating energy from air (~14 times as much air as fuel by volume, so we'll just say air as a simplification) at the peak torque engine speed. Above that the increased rev's offsets the decrease in torque and the HP continues to increase until the peak HP engine speed. After that point the torque drops off faster then the engine speed increases, so the HP drops.

Ultimately for a 911 I believe the trick is to get the ports flowing adequately for the torque spread and engine speeds defined by the cam. Unlike Detroit V8's, 911 ports seem to be pretty straight forward. Either way, most of my data is from stock configuration ports, so I'm sure that there are some secrets out there that I'm not aware of, but if you keep the stock port proportions, I'd expect these assumptions to work.