What Are the Best Intake Manifold Runner Designs for Maximum Power?

[peter.blais]

Hello, I am working on building some intake manifolds for an I4 2.0L street car project. The engine is being built for maximum power output, using a larger turbocharger.

My basic plan was a log style, tapered plenum manifold, as you can see here:

http://www.xmission.com/~eblais/intakemani/20vintakemani.jpg
http://www.xmission.com/~eblais/intakemani/20vintakemani2.jpg
http://www.xmission.com/~eblais/intakemani/20vintakemani3.jpg

This uses very short runners (4" in that version) which weld to a european "shorty" manifold cast aluminum piece which has a runner length of approximatly 3" and incorperates the fuel injector seats and mounts etc.

Because I'm using that european shorty piece, the ID of the bottom of the runners is already set at approximatly 40mm, it could change slightly but pretty much defined by the casting.

My question is this... where can i find information on various inlet bell or runner designs? What you see there is a slightly tapered runner with a 180 degree curve lip on the top. However i have seen some runners which have a much more gradual curve as seen here...

http://www.dtafast.co.uk/jenvey/drawings/AH_Al.gif

I think I'm probably splitting hairs here but am trying to find out what the absolute ideal design would be for maximum power. A broad powerband is not a huge concern as the engine will be drag raced and kept in a very narrow powerband between 6000 and 8000 rpm.

Also really trying to figure out some guidelines to determine the shape... I see a lot of talk of diameter and length odviously, but am having trouble finding this.

 

[faust9]

You will need access to an engine dyno to determine the optiman runner length/shape and optimal plenum shape. Intakes and exhaust systems are voodoo black magic throw the chicken bones pieces of an engine and need to be designed as such. You could do Helmholtz http://www.phys.unsw.edu.au/~jw/Helmholtz.html calculations to match the runner length to your desired operating band but in the end without a dyno to test various runner lengths you're just shooting in the dark at a target even the automakers can't theoretically hit. Yes, even the automakers use dyno's to determine the optimal designs for their intakes.

Now you can get close (Ricardo http://www.mscsoftware.com.au/products/software/msc/easy5/easy5_engine.htm ) to help you predict a good design but require reams of data to be even 20% close to an optimal design.

For a force-fed application, runner design is less of a factor than for a N/A design. Plenum volume is the key---the rule of thumb is plenum volume >= engine volume when force-fed. A large plenum has a few draw backs though so it is not the magic bullit either. A large plenum breathing through a single butterfly will have a lag between throttle opening and engine response.

Basically, there is no easy recipe for intake design. There are some rules of thumb---long thin runners=low end torque, short fat runners=high RPM power plenum volume=engine volume for superchared applications but there is no perfect solution nor is ther an easy formula (helmholtz calculations can get pretty scary when one takes valve timing exhaust pulse velocities, intake pulse velocities exhaust length and desired operating conditions into consideration---even then there are 12937492137 not even accounted for so you are nowhere near optimal).

Trial and error.
Good luck.

[edit] Also, I would not dream of designing a non-symmetric intake without access to a flow bench. I see you are attempting a side draft design and that means you have to verify the flow through each runner to make sure the runners flow about the same CFM. If they don't the you'll end up with a cylinder that can make more or less power than they others---not a good situation. Too much more or less and you'll be scratching your head wondering why your crankshaft failed or why your block decided to turn itsef into to two cylinder engines.

Also, you probably don't want those raised lips on your venturis' because you'll end up producing a turbulent zone as the air splits to either side of the line of stagnation. Those big humps will produce a lot of turbulance at the port inlet which will reduce the flow through the port at least that's my---if I were a betting man. You want to try and funnel the air in---don't let force it to split like that.

 

[Cliff_J]

...so it is not the magic bullit either...

Mr. McQueen would be proud. And now LeMans is on DVD finally - whoo hoo!

Regarding the intake - I watched the documentary about Ford's 2005 Racing Mustang development (instead of a flower like rosebud or petunia they called it the "Boy Racer Mustang") and when testing the intake designs they used SLA models on the dyno. It became very obvious when at one point the engine stopped while in the middle of a pull and it was because a couple sections of the intake had become unglued. Oddest thing to me was that they had straight runners in the SLA part on the dyno that looked like a 60s era tunnel ram and a more conventional folded design like on the production mustang in the engine delivered to the builder (but could have been simply because the prototype intake was intended to be delivered later).

 

[faust9]

Mr. McQueen would be proud. And now LeMans is on DVD finally - whoo hoo!

Regarding the intake - I watched the documentary about Ford's 2005 Racing Mustang development (instead of a flower like rosebud or petunia they called it the "Boy Racer Mustang") and when testing the intake designs they used SLA models on the dyno. It became very obvious when at one point the engine stopped while in the middle of a pull and it was because a couple sections of the intake had become unglued. Oddest thing to me was that they had straight runners in the SLA part on the dyno that looked like a 60s era tunnel ram and a more conventional folded design like on the production mustang in the engine delivered to the builder (but could have been simply because the prototype intake was intended to be delivered later).

I was hoping someone would catch that

 

[peter.blais]

haha lemans DVD, i always threaten my girlfriend with that... If I have to watch the lion king, you have to watch lemans. Its worked so far

The intake manifold is side entry because it is required by the transverse mounted engine. What i'll probably do is rig up a home (ghetto) flow tester and fiddle with various stuff until its close, and then take it to a real flow bench to verify.

Plenum volume there is well over engine displacement, I haven't calculated it again for the latest version but it is over 2x. I know this hurts throttle responce but the car will be launched from full throttle / anti lag system and then flat shifted anyways. I would also reason that the bigger plenum volume will help even out flow between the 4 cylinders, but who knows...

This next version i lowered the stacks all the way down to the floor, kinda want to keep them there though because they are the "trick parts" in vw land lol regardless if they work or not. I also decided to ditch the OEM lower section piece because it was driving me nuts, and its a european, junkyard part, so getting a good supply is kinda a joke.

http://www.xmission.com/~eblais/intakemani/smallportintake.jpg
http://www.xmission.com/~eblais/intakemani/smallportintake2.jpg
http://www.xmission.com/~eblais/intakemani/smallportintake3.jpg

I'm thinking about going even bigger on the plenum... Runner length is just a tad shorter and fatter then the OEM piece, this is to reflect the new powerband, because the stock turbo has full boost by 2000 and torque is falling by 4000.

I have a feeling there will be many versions of this piece haha before i get something near ideal... I've been kind of cheating a bit and spying on what works for the mitsubishi evo guys because they have a very similar displacement / head flow / setups (cams turbo etc)

 

[jay14u]

Hello all iam to looking at making my own manifold, question i ask is can i make my design out of M.D.F board and put it on a flow bench to test design. Or does it need to be made from metal to give accurate figure. Cheers jason.

 

[HowlerMonkey]

How much horsepower are we talking about?

Here's a good thread to read.

http://forums.hybridz.org/index.php/topic/61549-turbo-intake-plenum/

 

[jay14u]

thanku Howlermonkey for the thread, power level iam chasing is 600rwhp. the engine will be N/A v8 434 Cubes, fuel injected running twin 4 barrel throttle bodys. So hopefully a well designed engine package can reach my goal

 

[mender]

Or copy this:
http://www.mastmotorsports.com/2010/news-view.php?id=25

 

[jay14u]

Thank you Mender, I should have mentioned engine in more detail. It will be a ford engine 434 cubes n/a, dart/ford engine block, bore=4.155, storke=4.0, heads kaase C-400 flow between 340-400cfm cnc ported, mech roller cam, 11.47 compression. Tranmission is 4spd toploader, 9inch diff with 3.60 ratio. Car is a mustang 70 mach1 and weights 3571pounds or 1619kg. Redline will be 7500 and run a haltech or motec computer.

 

[mender]

I think you'd be farther ahead having them do up a single 4 bbl TB EFI intake manifold to match the heads for you; no need for two TBs unless you're after the look.

But if you still want to do your own, just about any material will work for flow bench work as long as the walls reasonably smooth. MDF would be a good choice.

Keep in mind that the real test is the dyno, not the flow bench.

 

[jay14u]

Thanks mender for the info, I was speaking with a guy who use to do the heads for Dick Johnson Racing back in the 80,s. He said the samething, single 4 barrel throttle body and a off the shelf manifold would be just as good. He talked me out of trying to get 600rwhp because of traction issue, And there would be no tyre in street form that could handle that amount of power.