AW: Intake manifold construction, intercoolers

Bernd Felsche bernie at innovative.iinet.net.au
Wed Dec 5 02:19:47 GMT 2001


Bruce tapped away at the keyboard with:

> From: "Bernd Felsche" <bernie at innovative.iinet.net.au>
> Subject: Re: AW: Intake manifold construction, intercoolers
> > Rausch, Bernd tapped away at the keyboard with:
> > > the complete intake manifold with plenum and intake runners will
> > > be custom fabricated. So I am open for various suggestions.
> > In that case, having separate throttles is even more attractive.
> 
> Heck, then make it a 7 butterfly system.  Or even 8.

To what end? Idle control? Throttle re-opening response by closing
off the plenum as well as the runners?

> > > I do not understand why integrating the IC in the plenum hurts
> > > throttle response. In a "classic" setup I have the IC before the
> > > throttle body, with extra piping and manifolds on the IC, so
> > > throttle response should become even better.
> > The plenum volume; that between the throttle valve and the intake
> > valve determines how quickly a pressure change can occur to change
> > airflow.

> If you just happen to be ignoring the turbo having to spool up or
> down as the case may be.

Not ignoring anything. The rate at which the turbo spools up is
determined by the exhaust gas flow which depends directly on how
quickly you can fill the cylinders in the first place.

Opening the throttle to a large plenum will require the turbo to
fill that volume as well as that of the runners.

OTOH, if you instead throttle the runners, the 'plenum' is already
filled with high-pressure charge and starts to fill the runners
sooner from the reservoir in the 'plenum'. As the throttles close,
the reservoir accumulates some pressure and a subsequent re-opening
of the throttles can benefit from the store if the over-pressure
hasn't bled out through the turbo.

Which system will have less lag?

> > > I thought individual throttle-bodies are normally used for engines
> > > with long duration cams to avoid cross-flow and get better
> > > idle/part throttle response ? I want to avoid the extra work (and
> > > money) integrating 6 TB´s and linkage.
> > Individual throttles are used where throttle response is required or
> > where there's no common plenum.
> 
> Or when looking for complex answers to simple guestions.

That may be the case. It could also be ricers using them for
brag-value.

A great number of successful competition vehicles do use separate
throttles; typically slide-throttles because they can afford them.
In terms of flow at part-throttle, the slide-throttle offers little
advantage at much greater risk of sticking (or leaking in typical
setups).

> > I envisage a common throttle plate with separate butterflies on a
> > common shaft (per side in a Vee engine). It's simple machining to
> > bore some suitable holes through a sandwich of two long aluminium
> > plates, machine a shaft and to make some butterflies. The throttle
> > cable enters what is the post-IC distribution box and pulls on a
> > lever attached to the butterfly shaft. A spring counters the cable
> > to close the butterflies.

> Till you start worring about the drill wandering and wind up with
> the throttle shaft being offset too much, or worse yet, not at
> all.  You'll want it offset so that if the throttle return spring
> breaks the throttles close rather then remain open.

The spring is mainly required to keep the cable in tension. Cables
aren't very useful for pushing things.

> > If the runners leave at opposite sides of the air distribution box,
> > then a relay link can mirror the motion for an identical setup of
> > throttles on the opposite side. There are only two throttle bodies;
> > each with 3 butterflies on a common shaft.

> then you'll also want a by-pass adjustment for each butterfly to
> really balance out the air flow at idle.

Not required in this application; but simple enough to add by
drilling bypass passages through the inner plate and throttling them
with set-screws accessible from the outside. A PITA to adjust and
probably a liability in this application.

> > A competent machinist can probably make all the throttle parts in
> > less than a day using a lathe and bench drill in addition to some
> > hand tools. A milling machine would be useful but not essential.
> > Especially useful for creating the lip against which the butterflies
> > close on each half of the throttle plate sandwich, and for chasing a
> > groove for the butterfly shaft and its bushes. (Don't skimp on the
> > bushes - the shaft has to be quite thin in places and could be bent
> > by a backfire, causing it to stick subsequently.)
> 
> Now you want thinning of the butterfly shafts too?.

The shafts will have to be thin to minimise obstruction at WOT.
The shaft has to be 'machined' to attach the butterflies.
Given the flow rate and engine capacity, each runner would be
around 70 mm diameter with a similar size of throttle plate.

> Shall we chrome all the bearing surfaces, or run the shaft on ball
> bearings, and incorporate seals into all 12-14-16 throttle shaft
> bearings?.

One could run the shafts on bare aluminium, but that's not very
durable. If the runners are close together, 4 bushes is sufficient.
You'd gain nothing (except profit) by using ball bearings or by
chroming the shaft. More bearings than necessary increases not just
the expense of the bearings, but also aggravates alignment.

No seals are required on the shafts as the shafts are inside the air
distribution box. The only part of the throttle actuation that
requires a seal is the Bowden cable entry.

> > Each throttle plate sandwich, complete with butterflies and spring
> > can be held together with screws through one plate tapped into the
> > other.
> > Don't let the price of commercial units fool you into thinking the
> > parts are hard to make.
> > If you want to, you can attach the spigots for the inlet runners
> > directly to the throttle plates and use the plates as a side of the
> > distribution box.

> Or use *mirrored* runners and have all 6 butterflies in a row.

Que? Wouldn't that make for a very tall arrangement on a V engine?
Suits an inline engine.

> Simplicity is the key to design.

You mean
  "Everything should be as simple as possible; but no simpler?"

> I'll stick with my earlier statements.
> It's much more important to have the plenum so that each cylinder
> feeds freely from it and not having one runner wall at the end of
> a large column of air.  The one good thing about the SDS refered
> to intake fabrication was that they continued the plenum past the
> last cylinder's runner wall.

A runner immediately at a _long_ wall benefits flow-wise from the
Coander effect - the wall "guides" airflow reducing the pressure
difference required for a given flow into the runner. If not all
runners have the same entry geometry, then there will be an
imbalance. This was one of the first flow-effects I encountered in
practical engine design in 1978.

> As soon as the machine shop is done with my manifold I'll get back
> with on what really works in the real world rather then just day
> dreaming or quoting theory

Oh; it really works. It's nothing new; just old stuff applied in a
new way.

But don't let that stop you from doing things your way. If you're
happy with the way your method works, there's no reason for you to
change.

-- 
Bernd Felsche - Innovative Reckoning, Perth, Western Australia
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org



More information about the Diy_efi mailing list