ITB Question: What is the ideal length from intake valve to butterfly?
 

In some distant future I'll be attempting an ITB project. I'm toying with the idea of modifying a set of triumph tripple ITBs.

I would simply weld on an "intake runner" and intake flange on each ITB, effectively extending the triumph ITB and merging it onto the intake port shape of the porsche.

My question is, how long should I make this weld-on runner, distance from intake valve to butterfly?

Mine ended up being really short. The reason was the clearances required to fit it under the intercooler and plenums. Proper clearance for the AC compressor etc. You'll pretty much need to mock everything up together and figure that out.

This depends on engine configuration to optimize. In reality, it does not really matter unless you have to have nth degree performance.

A few guidelines I follow:

Low overlap cams can tolerate butterflies closer to the intake valve because there is limited reversion pulse.
The more cam and more reversion due to overlap the farther you want to move the butterfly as this increases intake tract volume to buffer the reversion pulse.
The longer the effective runer path the batter for low-end torque.
The shorter the path the higher the overall resonance frequency pushing to higher rpm.
Shorter the path improves throttle response.
Shorter the path the less time for fuel mixing at atomization at high rpms.

In reality its a complicated thought process if you really want to optimize.

In a street turbo application, you will never be using a cam with enough overlap to create reversion problems.

I know street turbos running 3.5 mm's of overlap.

Good (inferred) point: EFI is far more forgiving, so you can get away with more overlap running EFI and an IACV.

I tend to think of overlap in terms of degrees. What does 3.5 mm of overlap translate into?

For example, here are the calculations I have for the following common cams in terms of degrees overlap:

• Stock 930 - 23
• SC - 33
• 964 - 42
• 993SS (in my car) - 43
• GT2 - 52

Nice resource: http://members.uia.net/pkelley2/Overlap.html

I too have 993SS's. I want the reversion (I love a "lopey" sounding cam, I'm 'Murican, ok?!) when I eventually go ITB's, but know it reduces this.

Question is, will I still "hear" my 993SS's with ITB's?

I sure hope so! given my experience with ITBs and MFI, you don't lost anything in the character department.

I love the V8 drag cam lope of my 993SS cams! Best part! Who cares if you need to idle them at 1000 RPM. And on a Zork...heaven.

You can only get an improvement at one narrow range with a fixed system.

Many would say that they would like to bump up torque in the 1200 - 1500 range. The desired length can be closely estimated using the organ pipe or Helmholtz resonator equations (Google it). If you do, you will find that it is ridiculously long (like 30 inches). That is why Porsche had to wrap the runners around like they did with the Varioram tuned intake.

Plus, if you tune for low end torque, you hurt the top end. That problem forced Porsche to put the "vario" in the Varioram. It shortens the runners with a clever rotating valve to tune at the top end.

Given all of this, most would probably just go with something that fits well,

I 'd go with as short as possible myself, assuming you have EFI. With a big intercooler there is enough volume in the system as is.

Tippy: you will maybe hear some of those 993SS cams' lope with a good exhaust, but it will mostly be minimized.

Part of the beauty of ITBs is they make running really big NA-style cams possible on a turbo engine without having the reversion on one cylinder mess up another cylinder (as happens on a single throttle manifold/plenum).

GT2 cams are suddenly small when running ITBs, as are SS cams (basically mechanical version of 993RS hydraulic cams).

3.5mm overlap? Awww, that's so cute... Kidding aside, you can get away with larger than that even and still have a civilized idle at or below 1,000rpm. Our 4.0L Twin Turbo has some big bad Edelweiß cams and she starts and idles with no drama like a plain 964/993 3.6L.

Back to your question jsveb: you'll probably be more limited by limited real estate beneath the intercooler, like Ken indicated. I'd almost start from the top and work back down -- space below the IC for volume for plenums and their flow & plumbing, less the minimum space for base adapters and the t-bodies themselves... then inlets to t-bodies... you run out of space real fast.

Also, at some point you are splitting hairs on a street car. General rule of thumb on 911s since the 1970s was high mounted throttles & injectors for maximum mixture and velocity and peak power (picture high butterfly RSRs) vs. low mounted (993 3.8RSR). Even if you built two systems and tested them back-to-back, I doubt you'd see any significant difference.

Start & idle with really big cams, 4.0L Twin Turbo engine with ITBs:

https://youtu.be/lkB7v5nxDaw

No drama.

Got throttle response?

https://www.youtube.com/watch?v=hA3dLkO4B8o

That car isn't boosting in that video - those are open stacks I am seeing right?

Right, we have the plenum tops off, exposing the tops of the trumpets. Mike just did a final synchronization on them.

Revs the same with plenums and intercooler on.
Will post that video later, plus dyno.

<a href="http://s128.photobucket.com/user/MarkNH/media/misc/anim/clapping.gif.html" target="_blank"><img src="http://i128.photobucket.com/albums/p170/MarkNH/misc/anim/clapping.gif" border="0" alt=" photo clapping.gif"/></a>

Thank you all for your inputs, a good tech discussion.

Sounds like it will be a compromise no matter what, as always. My take-away is that there is no really big nono. Short for throttle response, but unfortunately also giving less low-end TQ.

How about a short-runner ITB setup with timing (and twin plug) to back it up along with a relatively big turbo for good upper end Powerband?

Chris, I realize your accomplishments have received plenty of superlatives from me previously, but that RWB is simply stunning. It is RWB crazy, yet super tight with the what appears as a masterpiece of a powertrain. Congrats!!!

I hope it will be put to good use.

Unless the design of the ITB intake is really poor, the cylinder filling will be much more efficient than on a 930, 3.2L, or 3.6L manifold. IOW, you're going to see gains everywhere.

Pretty much what we're doing on this job, a stroker-3.5L (3.45L 76.4mmx98mm)
Really good twin plug heads, really good cams, take out the backpressure.

http://forums.pelicanparts.com/uploa...1461355737.jpg

http://forums.pelicanparts.com/uploa...1461356047.jpg


I have no doubt it will be! It will have heat and A/C, the owner plans to put a lot of miles on it.

100% agree here. With ITBs you have equal and efficient cylinder filling. No divided airflow like a common plenum will have. Increased efficiency gives you more power overall.

What really made me want ITB's is when Chris posted a build that went from common to ITB's and it picked up 50whp (or torque?) in midrange with no other changes plus additional power across the board.

The question "What is the ideal lenght from intake valve to butterfly" is not the right question if looking for tourqe and power.

The BUTTERFLY itself is only an air restriction when it comes to tourqe and power, not a tuning instrument, whereas the "empty" runner/pipe design is a true tuning instrument.

The question instead should always be what is the optimal length, diameter and shape of the total "runner design" from intake valve to the end of trumpets in the plenum based on a certain engine specs. This is where the tuning and design is made in regards to tourqe and power. The distance from intake valve to butterfly should be as short as possible for optimal throttle respons without damaging the air distribution. NA F1 engines had the butterfly build into the cyllinderheads for that reason.

For optimal runner lenght and shape design we most times have to make compromises due to the available space for the inlet runners length and shape. However if space is not a restriction then you can do a lot of interesting tuning. There is a US based NA calculation tool called PIPEMAX (cheap actually), it is one out of many tools to calculate runner lenght for an certain powerband with full engine spec and output included. NA specs works fine on forced induction engines as long as you know how to handle backpressure. Many old school professionals engine builders use this tool and it is damn close to reality, gives good direction when choosing cam profiles, valve sizes, bore/stroke, compresion ratio and building headers or inlet manifolds...

One more thing to remember is that the plenum on top of the ITB's is no different to a plenum in a casted intake manifolds. Only difference is that with ITB's you get to design everything from buttom up and hence make your own optimal air distritution not limited to any pancake, unequal, small/large etc. runner design to begin with. With ITB's and 6 x throttles full open you are back to a "Carrera manifold" just build with another style of runner and plenum design.

Actually a correct designed tigwelded or casted inlet manifold with runner setup just like an optimal build ITB setup has same advantages when it comes to tourqe/power on full throttle, even possible more advantage from lower air restriction out of the 1 x large butterfly instead of 6 x small butterflies in narrow runners. Same reason why the high output turbo drag racers are using big plenums + 1 x big butterfly as throttle respons is not part of the scope. This goes also for the few Porsche 1/4 drag racers as well. But idling (based on agreesive duration/overlab cam profiles) and throttlerespons is lost. I always choise ITB's for that reason for non-drag race cars. I love throttlerespons over everything else!

Whether my question is the right question or not, is in fact irrelevant. It is merely a starting point for a good tech discussion where I (and hopefully others) gain valuable knowledge.

What you are saying is that my question is one third of the "right" question.

Most often, in order to ask the "right," or rather, the correct question one must know the answer - I don't.
Maybe by the end of this I'll be able to ask that question. Thank you for your input - lots of good information.