Turbo header design

[Kevin Yachimec]

I missed the part of this post that stated the intended use of the vehicle
in question, so the following information may not apply.
If the header/manifold is for a street turbo application there is no
measurable gain by using a tuned the length header unless you using a pulse
type turbine and housing designed for receiving exhaust pulses. Most turbos
you find on production vehicles have the best results if the manifold
smoothes out all the pulses into a constant flow with heat lost being
minimized. I have a promotional paper from Allied Signal outlining manifold
design for their turbos. It covers both their new pulse type turbos (like
the ones you see on all those 10 second Honda's with equal length headers)
and their standard turbos, if I can find it I'll post it 'incoming'.

>Greg Hermann wrote:
>
>
> FOR THIS PARTICULAR V-8> (IT HAS A 90 degree crank)
> >                Firing order
> >          1  8  4  3  6  5  7  2
>
> With this firing order, take every other cylinder in the firing order if
> you want  180 degree (crank rotation) spaced pulses into a pair of 4 into
1
> collectors:
>
> 1, 4, 6, & 7 go into one collector (two end cylinders from one bank, and
> the two center cylinders from the other bank.
>
> 8, 3, 5, & 2 go into the other collector. (the two end cylinders on the
> other side, and the two middles from the first side.
>
> If you want PROPERLY done tri-wyes, go 1/2 way around the firing order, so
> that the pulses in the pairs of tubes which combine at the first wye are
> spaced evenly, 360 (crank) degrees apart:
>
> 1 & 6; 8 &5; 4 & 7; 3 & 2. As you can see, each pair takes one tube from
> each bank of the motor.
>
> Then, at the second wye, pair 1-6 with 4-7 and pair 8-5 with 3-2 so as to
> get evenly spaced pulses at the second wye in each header.
>
> If you wanna go with REALLY long tri-wye branches, (surprising how
> fantastically good this is for street, high torque, & 4x4 type apps.)
start
> out with a regular set of 4 into one headers (of the smallest tube size
you
> can find), cut them off a bit before the collectors, and do the snaking
> back and forth under the bell housing/tranny to get the correct first wye
> pairings. The second branches want to be the same length as the first
ones,
> but there are no crossovers if you plan ahead. After the four second
> branches combine into two (tertiary) pipes, you want a length of pipe of
> length equal to all the first and second branches. At the end of that
pipe,
> you need a gap (no diameter change of offset in the pipe) in the pipe
about
> 3/4 inch long. Put an empty plenum chamber around that gap. The volume of
> the plenum needs to be maybe double the internal volume of the last
> (single) length of pipe feeding it. Just run a full size tail pipe and low
> restriction muffler behind the plenum (Not much vehicle length left after
> this, anyway!)
> The headers will effectively see the plenum as an open ended (to
> atmosphere) pipe. DO NOT omit the plenum, it is worth quite a bit of HP
and
> response!
> For street stuff, usually, primary tubes 1 trade size SMALLER than the
size
> which can be swedged square to fit into a flange and match the port
> properly work out to be the correct size. (If a 1-7/8" tube can be squared
> at one end and fitted into a flange so that the inside of that tube
matches
> the port shape, then run the primary tubes with 1-3/4 " tubing.) You want
> the internal AREA of your primary tubes to be about equal to the
> cros-sectional AREA of the exhaust port. The above is what usually works
> out right. (Especially if the engine designer did his homework!) The extra
> work involved in getting down to the smaller tube size is why most off the
> shelf headers use too big a tube size!  Figure the length of the primary
> tubes so that each of their internal volumes is about 140% of the
> displacement of an individual cylinder. (For instance, for a 350 cid V-8,
> using 1-3/4" tubes (which are about 1-5/8" inside diameter in 16 gauge
> tube, this would mean you want (at least) 24 inch long primary tubes.)
(And
> if you are using lighter than 16 gauge tubing, don't bother!)
> If the primaries are 1-3/4", then 2" is usually about right for the
> secondaries, and 2-1/4" for the next ones. (A quarter inch increase in
tube
> size at each successive wye is usually about right.
>
> 1-3/4" diameter primary tube size was only an example. It is prolly BIG
for
> a street 350, 1-5/8" or 1-1/2" is more likely to be right (but with more
> tube length so as to get to the same internal tube volume!)
>
> If you build a set of tri-wyes this way, they will sound and run like
> nothing else! The throttle response will be astounding! But it is a ROYAL
> pain in the #@$% to do it!
>
> All my experience says that the biggest gain from headers is from giving
as
> much of the exhaust gas as possible somewhere to go freely during the
> "blowdown" part of the exhaust stroke--before the piston starts back up
> significantly. Yes, it's nice to have a low pressure pulse in the exhaust
> port at overlap, and that pulse helps to scavenge clearance gasses if you
> have it there at the right time, but the longer you can keep the blowdown
> stage flow sonic, the less work the engine will have to waste pumping
> exhaust gas out of its cylinders, and the more sonic energy there will be
> available for creation of a scavenging pulse!
>
> Regards, Greg
>

Kevin Yachimec
Cybertech Automation Inc.
(780)483-6222 Ext. 118
keviny at cybertech.ca