custom intake manifold

so i've been researching what goes into making a custom intake manifold, and i've found all these rediculously long equations to come up with runner length. so i'd be very grateful if you guys could answer a few of questions for me... (this is for a naturally aspirated engine)

PLENUM...
1) should the plenum be a 0.065 wall exhaust pipe, in whatever diameter i need?
2) is plenum volume supposed to be 100% of engine displacement? if so, what should i do if i can't get EXACT engine displacement size?

RUNNERS...
1) is runner volume supposed to be 100 % of cylinder displacement? if so, what should i do if i can't get EXACT runner size
2) if you know what volume your runners need to be, how do you find the lengths they need to be?
3) can i divide the runner lengths by factors of 2 to accomodate engine compartment limitations, and long as i make my runners wider? (there isn't much room under the hood...
4) should i have a 10% taper on my runners? how do i factor that into my equation?
5) how does the velocity stack work? what degree of curve should i use?
6) are straight runners "the shiznit"?
7) can i have the manifold angled down if i don't have room to have it coming straight off the head?

THROTTLE BODY...
1) should i go bigger?
2) can i use the stock accessories on a custom big body?


this is being considered for a 2000 toyota celica gt-s, and then for a gt. basically, i'm trying to get all my efficiencies as close to 100% as possible without forced induction.

I suspect you know more about the subject than me, (please feel free to correct my comments here) but from what I know...... the variables that you mention, runner length and diameter and plenum volume affect the power curve at various rpm level.
Therefore, you can tailor the maximum potential power production for a specific rpm.

So what, you say?

Well, this allows you to coordinate the maximum power production to your choice of camshaft, header and valve size.

Generally speaking, each time an intake valve opens and suck air, it sends a resonance or vibration up the runner. At a specific RPM, ideally, the opening of the intake valve corresponds to the point at which the resonance pulse from the previous valve opening reaches the plenum. This produces a 'ram' effect that helps stuff the air into the cylinder at better than 100% efficiency, but only at a narrow rpm range.

From what I can recall, shorter, wider runners favour higher rpm efficiency, longer, narrower runners favour lower rpm. Larger plenums will lower this rpm range as well.

A tapered runner, like a velocity stack widen the effective rpm range of this ram effect, but I suspect they may lower the overall efficiency, (I am not sure)

I would suggest you choose the RPM range that you want to make power. Choose a camshaft to match this range (aftermarket cam manufacturers usually supply such information) and make sure it will work properly with the cylinder head you have (ie stock or ported) THEN design a runner length and diameter which will complement the camshaft.

Finally, if you are going to use exhaust tubing, (which is fine) consider using a variable length runner, with a telescoping runner (one pipe inside another) so you can change the runner length and 'tune' the intake design.

The most difficult part of this exercise will be to package a suitable design within the limited space under the hood. Unless you run at extremely high rpm, you will find the ideal runners are fairly long, 2 to 3 feet. They can curve, so long as all the curves have a uniform pipe diameter, which may limit yourself to mandrel-bent tubing.

so i've been researching what goes into making a custom intake manifold, and i've found all these rediculously long equations to come up with runner length. so i'd be very grateful if you guys could answer a few of questions for me... (this is for a naturally aspirated engine)

PLENUM...
1) should the plenum be a 0.065 wall exhaust pipe, in whatever diameter i need?
2) is plenum volume supposed to be 100% of engine displacement? if so, what should i do if i can't get EXACT engine displacement size?

Using steel pipe is a great start. Its easy to weld and it also rejects heat slightly better than aluminum. Consider using the thinnest possible walls for several reasons. 1) thinner walls don't transfer heat as well. I know it sounds counter-intuitive, but it has to do with the mass to specific heat capacity of the metal. 2) Its lighter, 3) easier to bend in tight spots accurately, and 4) unless you're going to supply it with excessive boost, the intake manifold sees very little "stress" so you need not make it harder on yourself.

The size of plenums is a cause for debate, especially with EFI (which I'll go into more later... or maybe you already know :)) but you can use some calculus and trig to get the dimensions you need. Once its built, you can fill it with a known volume of water or oil to check it, and chances are it will be off a little. If its too small, you can add on to the length of the TB boss when you build it. If its too big you can weld in a steel bar somewhere that doesn't affect flow. I've also seen racers take a hammer to the end cap of their plenum and smack it until its the right size. Not pretty, but it works and since very little of the velocity is built in the plenum, it doesn't affect flow.

RUNNERS...
1) is runner volume supposed to be 100 % of cylinder displacement? if so, what should i do if i can't get EXACT runner size
2) if you know what volume your runners need to be, how do you find the lengths they need to be?
3) can i divide the runner lengths by factors of 2 to accomodate engine compartment limitations, and long as i make my runners wider? (there isn't much room under the hood...
4) should i have a 10% taper on my runners? how do i factor that into my equation?
5) how does the velocity stack work? what degree of curve should i use?
6) are straight runners "the shiznit"?
7) can i have the manifold angled down if i don't have room to have it coming straight off the head?

1) Again, cause for debate with EFI (and I'll explain now). The ramming effect that you are trying to produce here that gets you closer to the 100% VE is based on velocities and masses of air. With a carbed car, you are moving air/fuel mixture in both the plenum and runners. With Multi port EFI, you're only moving air which has an incredibly different mass. This is someone else's math that I haven't checked, but if you tuned a manifold for 95% VE with a carb, then switched to EFI, you would have to almost triple the length of the runner to keep the same inertia.

If you can't get exact runner size in your space constraints, it is quite acceptable to extend the runner INSIDE the plenum. If you need 11 inches and you only have 8, extend the runner 3" into the plenum. This of course provides its own packaging problems, since you've now a) potentially created flow inconsistencies in the plenum, and b) removed some effective volume of the plenum.

2) Once you find the volume, the size/length is a matter of where you want the VE to peak. Shorter fatter runners will peak much higher than long skinny ones. I'm sure there is a formula for it but I don't know it.

3) yes you can alter the length of the runner, but like in #2 you will alter the RPM of the peak VE. For all-out tuning purposes, there is no substitute for the proper configuration. In reality, engines like the LT1 achieve peak VE at remarkably low RPMs despite their short fat runners and very large plenum. It goes back to the EFI thing. It matters less since there is less mass in the air alone.

4) in most naturally aspirated engines, torque can be boosted with the tapered runner. It is one of those things that I know works, but since I haven't figured it out yet I don't know why. How's that for honesty :) The taper serves for an increase in velocity, which can help. From my remedial grasp of the taper, but good grasp on the theory, my edumacated guess is that the taper might serve as a way of incorporating both of two features; velocity and space. If you have space constraints, a tapered runner could be used to maintain the proper volume, but the net velocity would still end up higher. You would have the benefits of proper volume in a small space without the higher-rpm peak associated with short fat runners. Someone else will offer the right answer now :)

5) in general, a velocity stack is a method by which the greatest increase in velocity is achieved with the least amount of turbulence and restriction. The commonly accepted method is to use a bell cuve that has a radius that is equal to the radius of the tubing. Basically, if you're using a 2" tube, the radius of the flare should be one inch. The flare should span the entire 90*, so basically it should start parallel to the plenum wall and radius 1" until its parallel with the runner.

6) since you don't have the advantage of millions of dollars and sixty prototypes to test, yes straight runners are beneficial. Certainly not shiznit territory, but when you're designing your own manifold, keeping them straight will remove that variable from their design. You could flow test each individual runner, but then they might change after they're installed in the plenum. Having said that, in your case as long as all of the curves are identical, I don't think that some gentle curves would show a measurable difference in VE.

7) you can angle the runners and plenum any way you want. The further down you go, the more the vacuum will have to work to get the air to counteract gravity. Although air is "weightless" inside itself, moving air is different since it now has a greater velocity than standing air.



THROTTLE BODY...
1) should i go bigger?
2) can i use the stock accessories on a custom big body?


1) yup. With EFI, you have a much wider area to play with. With exceptions, bigger is in fact better on EFI TBs. Some have observed a decrease in power or elapsed times with excessively large TBs. My guess is that when you use a huge TB, the incoming air is slow and then the runners have the job of accelerating that air. Matching a TB to the flow prevents that problem. This is also why small plenum intakes tend to make more torque; the incoming air never has the space to slow down and stays fast, but then lacks the physical space to really support high-hp flow. With carbs, I usually say calculate flow, then round down. With EFI, I say calculate flow and round up. Another way to help velocity (and VE) is to achieve that flow using two or three smaller butterflies than one huge one. Its a subtle difference, but we are talking about maximizing efficiency.

2) depends. You can have your stock TB bored out some and replaced with a bigger butterfly. I don't know if it will be enough for your plans, but worth a try. I'm also not sure how easy that ECM is to hack, but its often common practice to use the TPI/LT1 style TB on several different cars. They just reprogram the ECM to recognize the GM signal from the TPS and other sensors. That would not only net you up to 850 CFMs, it would also get you the twin butterflies.

.wider runners favour higher rpm efficiency, longer, narrower runners favour lower rpm

Magic... I didn't know you were Canadian? Excellent post by the way.

Magic... I didn't know you were Canadian? Excellent post by the way.
Thank you........ originally I'm from Britain by way of Canada......that 'u' does seem to find it's way into a lot of words, doesn't it?

BTW I learned some interesting things from your very informative post, too. Well done.

If you talk to Dupox on the forums I'm sure he can give you some tips, and check out is 351C fuel injection intake hes building, theres pictures of it in the fabrication forum.

Generally speaking, each time an intake valve opens and suck air, it sends a resonance or vibration up the runner. At a specific RPM, ideally, the opening of the intake valve corresponds to the point at which the resonance pulse from the previous valve opening reaches the plenum. This produces a 'ram' effect that helps stuff the air into the cylinder at better than 100% efficiency, but only at a narrow rpm range.


My understanding was that the resonance pulse was sent back up the runner by the intake valve slamming shut and the air hitting a wall, then "bouncing" back up the runner. Then, if the runner is the proper length, the pulse finds it's way back to the intake valve just as it's opening up again, producing the "ram" effect you describe.


From what I can recall, shorter, wider runners favour higher rpm efficiency, longer, narrower runners favour lower rpm. Larger plenums will lower this rpm range as well.


How do the larger plenums effect target RPM range?


A tapered runner, like a velocity stack widen the effective rpm range of this ram effect, but I suspect they may lower the overall efficiency, (I am not sure)


I don't understand how it would lower the efficiency if it's widening the range of the "ram" effect.


Finally, if you are going to use exhaust tubing, (which is fine) consider using a variable length runner, with a telescoping runner (one pipe inside another) so you can change the runner length and 'tune' the intake design.


Where can I get more info on this? I've never heard of it before.



2)I'm also not sure how easy that ECM is to hack, but its often common practice to use the TPI/LT1 style TB on several different cars. They just reprogram the ECM to recognize the GM signal from the TPS and other sensors. That would not only net you up to 850 CFMs, it would also get you the twin butterflies.

A nice idea, but this ECM is picky about intakes. It only consistantly has accepted one, which has (stock replica) MAFS fins incorporated into the design.

Thanks a lot for the help on this stuff guys. I really appreciate it.

One of the best advantages of fabricating the intake maifold is to make it 'tunable'. Using telescoping tubes to tune the overall length makes logical sense to me and saves you the fabrication time with trying different lengths. This would be easier if the runners were straight, as opposed to curved.
Back in the early '60's, Chrysler made their high performance 413 wedge (not hemi) V8 motor with a 'cross ram' intake manifold with LONG runners that were routed up and over the engine.
There were actually two different designs, the 'short ram' (with runners of 28 inches, I think) and a 'long ram' manifold, with (I think) 34 inch runners. The short runners were a high rev, almost 'race only' piece, while the long runners were more suitable for the street.
This article http://www.camaro-untoldsecrets.com/articles/crossram.htm
has a diagram of such a piece.

You can also tune the plenum volume, by fabricating a bolt-on spacer. Many V8 type manifolds, especially those intended for carbs or TBI have provision for 'carb spacers' which effectively increase plenum volume. (I am not sure about the flow dynamics which makes a larger plenum behave differently.

My understanding was that the resonance pulse was sent back up the runner by the intake valve slamming shut and the air hitting a wall, then "bouncing" back up the runner. Then, if the runner is the proper length, the pulse finds it's way back to the intake valve just as it's opening up again, producing the "ram" effect you describe.

Yep. I think you and MagicRat were talking the same language, just different approaches. He is also right, though that the intake tuning is only experienced at a very narrow range. If done properly it will increase peak VE, but does very little for VE across the rest of the band.

How do the larger plenums effect target RPM range?

Larger plenums tend to shift the target range up, but again with EFI it matters slightly less.

I don't understand how it would lower the efficiency if it's widening the range of the "ram" effect.

It could lower overall average efficiency in the pursuit of raising the peak efficiency. Basically lifting the VE curve at the peak while dropping it elsewhere. This is a finer point that I can't speak more about since I've not experienced it first hand.

Where can I get more info on this? (variable length intake runner) I've never heard of it before.

Its a not-very commonly used technique since its still pretty tough to engineer for even the automakers... but the ones I've seen have been two tubes that slide one inside the other joined by a rubber bellows like an air spring. A vacuum operated or computer operated servo moves the plenum in and out effectively changing the length of the tube. It is brilliant. The problem with that design is the life expectancy of a constantly flexing rubber tube exposed to hydrocarbons. Other manufacturers have experimented with using two runners per cylinder and having one isolated by little throttles. The second throttle will open at higher RPMs letting it breathe easier.