ramairJP

iTrader: (3)

java,

the headers help to pull all of the exhaust gas out of the cylinder as possible and try to use the pressure waves to help pull the gas instead of making the piston push it out. During the exhaust stroke, the exhaust valve opens and the piston begins its upstroke. The piston has to do work against the gas in the cylinder to push it out and down the manifold. Furthermore, when the piston gets to TDC, there is still a volume of gas in the combustion chamber that never gets pushed out... it's simply mixed with the new intake charge. The headers try to "scavenge" this gas out and thus make room (combustion chamber are about 38cc.. right?) for more fresh air and fuel and thus more power. The pressure wave helps pull the exhaust out (since a fluid will travel to the area of least pressure) instead of making the piston push it... again saving power. So, instead of making power, a set of headers actually help to get power back that is lost through pumping losses and eliminate residual exhaust gas at the end of the exhaust stroke.

ChrisB is exactly right that a motor can be made to run "around" a part designed for one setup. Head porting, cam selection and timing, and compression can all help make up for, say, the spyder. However, just swapping intakes on the same setup will simply lose you a LOT on the bottom end and give it back on top (if your motor needs that much on top).

One question from your post Chris: In my experience, running higher and higher on the compression ratio has a law of dimishing returns and actually hurts top end "revability" when it's really high. For example, my race bike runs now with about 185 psi of cranking compression, but I have tried combos as high as 220. When I get start getting near 200 lbs, the motor's powerband has a tendency to flatten out up top and pretty much becomes a low end grunt. Of course, the increase in compression helps the bottom end as you stated, but I was wondering if it hurts the top end in these LS1s from your experience. I've been wondering about that....

John

ChrisB

Since you are measuring dynamic compression you are infact dead on - (with the diminshing returns)

The efficiency of the Otto cycle (assuming your motors are 4 stroke - I am not sure how applicable the otto cycle thermodynamics are to 2 strokes, though I imagine it would be similar) depends only on the temperature ratio of the compression process

we can model the efficiency as



so you can see from the general shape of the graph as the temperature delta gets larger the term T1/T2 gets smaller - and the overall function approaches unity - 100% effeciency. Using ideal gasses/adiabatic system, blah blah you would get something like



So this supports your experienc that you get diminshing returns as your dynamic compression goes up.


Just to make things clear for everyone, when we add something like the spyder intake to a stock car at lower rpms we reduce our dynamic compression even though the VE stays the same - so by increasing the static compression you are just bringing yourself up to your normal effeciency range - even if the increase in dynamic compression is constant across the rpm band you will see a bigger power gain at lower rpms - because of the effeciency graph above - higher dynamic compression will tend to flatten out your torque curve.


Chris

[ November 16, 2001: Message edited by: ChrisB ]

[ November 16, 2001: Message edited by: ChrisB ]</p>

java

ok you guys lost me ..

Allan, not anyone can make an LS1 fast ... that is why I came to you!! <img src="images/icons/smile.gif" border="0"> Not giving out any secrets just trying to learn some more.

Tim98TA

iTrader: (7)

Does anyone remember a guy named Tom Bumpus(sp) aka: Iron Man? That guy knew his **** about exhaust. I thought I saved a paper he had written, but I guess I lost it. I'll keep lookin for it. It had a wealth of info on it. Everything you wanted to know about exhaust was on that paper. Down to what wire/rod to weld what type of metal with.

Tim

java

the LS1 runner length is what 14", so is the the optimum lenght for our EFI engine? I have always heard that in a head optimized for swirl (assuming a dry intake) that a long runner is optimum for maximum horsepower output and a shorter runner increases torque. Subsequently the same is tru for the headers.

What I want to do is find the balance between optimum torque without sacrificing the top end horsepower our motors are designed for.

Terry Burger

iTrader: (4)

I don't think it matters much, but FWIW my setup works best with the cutouts 12" past the header collector. We're talking minor difference here.

1BADtruck

Steve,you got it backwards,long runners for torque,short for HP.

Allan

ChrisB

Actually it's not HP or torque, but both. What changes is the RPM at which the magnitude of either are modified.

A shorter runner will tend to favor higher rpm cylinder fill, since runner length and frequency are inversely related

What you generally do is model the tube (header or intake) as a closed tube resonator (valve closed) - since the goal is to generate a standing wave in port. (this is a wave that when reflected off the closed end is exacly in phase with it's "tail" end).

this is modeled by frequency = n*c/(4*L)

where n = the harmonic (unitless whole number)
c = speed of sound
L = length of the resonator

RPM is just a measure of frequency, you just may have to convert units

This is a simplistic approach that leaves out a bunch of other factors, but is the basis on which most of the other corrections/modifications are made.

The strength of the pressure differential will depend on which order you use - with higher orders (up to 1) giving a larger perceived gain.

Not that this tells us nothing about the daimeter of the pipe, but rather just it's length. If you want to get into air volume (and mass) you can treat the system as a simply harmonic oscillator and model it using that. This is the basis for helmoltz tuning - if you do a search for that it will explain it far better than I can. This is just another way of looking at the oscillator system though - and again, the goal is to create a standing wave.


Chris Bennight

[ November 18, 2001: Message edited by: ChrisB ]</p>

java

you're right Allan ..

The LS1 is supposedly a "tuned" intake manifold for the stock cam and timing events (i.e. friction loss minimized & low heat transfer) ... but can we do a btter job with the turns and cylinder head entry angle? or are we limited by the hood angle and engine position and what we have is the best we can do?

One interesting thing I have learned in reading is that what I thought would make a good cylinder head on my stock bottom end is actually what I don't need, LOL. Port velocity, compression and swirl have more affect than flow characteristics especially as we increase cam duration at .050" ... basically my engine used to stall or stumble at low engine speeds but really peak once the engine was up to speed.

McRat

[quote]Originally posted by ramairJP:
<strong>...

To answer the question why the exhaust header runners are so much longer than the intake runner... it's because of heat. The header is of course, much hotter than the intake air, and thus the sound wanves travel much faster (hotter air = faster sound). Therfore, to obtain the same resonant frequency at a given rpm, the runner must be longer ...
</strong><hr></blockquote>


It's been awhile since I built any exhaust systems, but I thought sound moves SLOWER through hot gases.

If I remember right, the reason the intake is so much shorter than the exhaust has to do with which "event" you're trying to syncronize.

The Intake runner "+ pulse" is to timed to the opening of the intake valve to ram air into it.

The Exhaust tube "- pulse" is to be timed to the closing of the exhaust valve to create a partial vacuum in the chamber and assist the intake charge.

Because these two events are way different relative to intake and exhaust centerlines, it creates the need for very different tuned lengths.
Or so my failing memory tells me...

ramairJP

iTrader: (3)

McRat,

I guess if you were thinking from a simple density standpoint then it would be easy to think that sound moves faster in colder, denser air. However, this is not the case. The formula for the speed of sound is C=(square root of (k*R*T)).

C=sqrt(k*R*T)
C= speed of sound
k=ratio of specific heats (Cp/Cv)... varies with temp, but at "general" engine temp (300) is about 1.35
R=individual gas constant, for air it is .287 KJ/(kg*K)
T=temperature measured on the absolute scale (kelvin or rankine)

So, looking at this equation, it is easy to see that the speed of sound is a function of temperature only... the hotter the faster.

For 300K air (rough inlet estimate) sound travels at 347 m/s (about 750 mph). For 1000K air (a little hotter than exhaust) the sound travels at 634 m/s... almost twice as fast.

As far as the duration of events, even if they were timed as you said it seems it would only take place every other cycle, or 720 degree of crank rotation. Of course they are at different times, but there has to be the same number of intake and exhaust events.

Chris, that's an excellent in-depth explanation of what I was saying about the resonance (standing wave). I was afraid of going that in depth and getting way off track.... good job there.

John

[ November 19, 2001: Message edited by: ramairJP ]</p>

McRat

Like I said, it's been awhile. But where in that formula is density? I know for a fact the speed of sound changes dramatically as a function of density. It's about 750mph at sea level and 600mph at 20,000 ft or thereabouts. Hot gases (600° F) are much less dense than room temp gases.

On hot days, it takes more lift for an airplane to become airborne. Because of poor density. Same effect with altitude. If the heat in the gases is trapped in the same volume, I might believe it.

I used to plug exhaust temp into the formula to get tube and cone lengths. Seems it made the tubes longer, but I could be mistaken.

The intake "pulse" event happens when air hits maximum velocity (low pressure) and then the valve closes (high pressure). It is a very short event. The exhaust "pulse" event starts as soon as the valve starts to open, because the gas is under severe pressure. It doesn't finish it's cycle until the valve is shut. So the intake event is about 110° and the exhaust 220°. This is probably where the number 2.5 times comes from.

ramairJP

iTrader: (3)

McRat,

I *believe* that formula is for pressures at or around atmospheric, but I'd have to look in the book to make sure. Remember that air at 20,000 feet is colder as well as less dense, so that too will work to slow down the speed of sound. I know it seems wrong that air travels faster in hotter air, but it's the way the thermo works.

I remeber reading an article from yahama whenever they were tuning one of their v-twin race bikes... it was from a long time ago. The engine builder had a problem with the back cylinder running completely different from the front cylinder despite all of the same inlet and outlet conditions. They finally realized the pipe from the rear cylinder wasn't being cooled like the front pipe was and thus had to be lengthed almost an inch to make up for the difference in temps of the two pipes. I'll have to look to see if I can find some formulas that relate to two stroke pipe building (i have them somewhere....).

I think I see your point on the intake/exhaust events. I suppose that, along with the speed of sound difference, all come into effect in the intake/header designs.

John

ChrisB

Re: the speed of sound and temp.

The problem is that temp, pressure, density, etc. are all inter-related, so it gets difficult (as JPramair points out) to try and simply reason out the relative effect of one another. But we can describe it using physics.

A sound wave is really a moving variation in pressure, where the density "inside" the pulse is greater than the density outside the pulse. If we work this through we come up with Newton's equation for the propigation of acoustic disturbances through an elastic medium (sound through air) - and get

v^2 = d(Pressure) / d(Density)

If we go to the ideal gas law - Press = Dens*R*T - we can rearrange it to get d(Press)/d(Dens) = RT. Plug that in the above equation and we get

v = sqrt(RT)

Seems to easy, right? yep, this is an approximation - it assumes that temperature is constant across the acoustic disturbance - which it isn't. Since T isn't constant you can't differentiate the ideal gas law to the form
dP/d(dens) = RT - thus we can't make the above substitution


Basically (after a long derivation) we will end up with (assuming the pressure wave is adiabatic)

v = sqrt(cp/cv * R * T) - so the above equation multiplied by the squre root of the ratios of the specific heats (cp/cv ~ 1.4).

In physics you can keep getting more and more accurate, and the expense of clarity, and it's really not worth taking it past this point.

So basically the speed of sound in air (the pressure propigation phenomenon we are talking about with intake/exhaust tu ning) varies with the square of the ratio of gas constants time the square of the universal gas constant and the square of temperature - temperature here is the only variable, so as you can see in this derivation it depends soley on temperature.

If you compare air at 20,000 feet and air at 1,000 feet and say they are at the same temperature, then realize that the air at 20,000 feet would have to somehow be *much* more dense then the air at 1,000 feet (which it isn't) - because of Pressure = Density * R * T. You really just can't change one independently, so the dillema you are facing is really false.

As for exhause length vs. intake length, there are a couple of reasons for the difference - first of all (and foremost) - practical considerations - we are realistically constrained to certain proportions which are dictated by the package in which the engine must fit.

The intake also is tuning for multiple reflections of the wave. Our primary cause for depression is the piston's downward stroke. this creates a negative pressure wave which propigates to the plenum and is in a sense reflected, and then travels back down the intake runner as a positive pressure wave, hopefully forcing a bit more air back in the cylinder even which the piston speed is dropping.

The other (and lesser) cause of pressure waves is when the intake vale closes - and momentum left in the colum of air will hit the closed valve. Reflection "inverts" the wave in a sense and we get a negative pulse. This then reflects off the plenum to become a positive pulse - this assist right around the opening of the valve - again, when piston velocity (depression) is low.

Exhaust is much simpler, in that you just tune for the closed end resonator harmonic like we discussed earlier - as there aren't any reflection effects to worry about (to any great extent). You can also achieve a broader power band by tuning for different rpms (but a lower peak) - it all depends on what your goal is.


Chris