11/02/2003

Mac under the controls section talks about dual throttle plates.  The
main between the compressor and the intake manifold, and another in
front of the turbo - with a mechanical gearing such that it does not
block the turbo completely.  Key here is it maintains positive atmo
pressure against the oil seals so that air "leaks" into the oil instead
of vice versa.  ( Side note - I think my spelling is deteriorating to
dyslexic )

Many systems use blow off valves.  When the throttle is suddenly closed
or reduced, the pressure build upstream very quickly.  No place for the
air to go? - it seriously slows the compressor down.  Then it has to be
wound back up if this was a shift.  A BOV vents some of the air - for
response best is probably upstream of the turbine in the exhaust - for
ricers - its a horn.

But, if as you close the primary butterfly ( throttle ), you close
partially a valve in front of the turbo, the air flow thru the turbo
drops radically.  You don't seal it - which will cause a vacuum - you
just seriously hinder it.

With a BOV - you are pumping the same amount of air but dumping it
overboard.  With a MacGinnes throttle setup, you are pumping seriously
less air, and since that requires a lot less power than a BOV dumping,
the compressor stays at a very high rpm and when the throttle open -
like after a shift - the second throttle also opens - to a very high rpm
compressor and instant air flow response occurs.

Pull out your trust Mac or similar book showing turbo maps.  I'll use
the compressor map for a Rajay turbo on page 9.

There is the surge line on the left of the curves, the vertical is
pressure ratio and the horizontal is inlet air flow.


Using 200 cfm for example, draw a vertical line.  The compressor can run
from 40,000 rpm to 65,000 rpm and provide this air flow - with a
pressure ratio of 1.2 to 1.6.

Using 300 cfm for example, draw a vertical line.  The compressor can run
from 40,000 rpm to 85,000 rpm and provide this air flow - with a
pressure ratio of 1.1 to 2.1.

Using 400 cfm for example, draw a vertical line.  The compressor can run
from 55,000 rpm to 105,000 rpm and provide this air flow - with a
pressure ratio of 1.2 to 2.7.

If we live within the island, we get about a 1.4 200 to 1.9 400 ( just
outside the island ) pressure ratio.

Thus the experts would advise this to be gospel and life goes on.

Perfection is the enemy of good enough. The small high efficiency island
is above the .70 and below .75 ok?  Its about 1/6 the size of the .65
island - so a minor increase in heating greatly expands the usable
island and being willing to go to .60 will allow you a 3.2 pressure
ratio at 600 cfm.

Read them and weep.  Trying to live within an island makes Corksheet
Bellony a bunch of money.  It does seriously increase in the designers
mind heat control - but this turbo could be pumping at 45 psi.

But, now for tweested sister darkside thinking.  Think not of the
compressor as a compressor - think of it as a flywheel.  Now surge is
only a problem with a compressor banging against a serious resistance.
Mac, among many others, states that surge tends to go away with the
throttle in front of the turbo and also that turbo overspeed is NEVER a
problem when the compressor is in a vacuum.  

Say we keep the compressor in a partial "vacuum - enough restriction so
that the air flow is say 100 cfm - and a closed throttle - so there is
little flow but at a higher than normal pressure ratio - and say the rpm
is 70,000 at the time.

Snap the throttles - 70k is 1.7 pressure at 200 cfm, about 1.67 at 300
cfm, about 1.62 at 400 cfm, and by 500 cfm about 1.5.

Notice - that as fast as the air flow increases, the rpm's are already
up and air is being pumped.  The key is the flywheel effect - speed it
up by creating an equivalent to a vacuum and when you slam it, you are
trying to slow it down.

An exercise for the student, but what's the exponent for changing speed
of a rotating object and how much force is needed?  Brain pharte
tonight.

Now you could use an active system and use the exhaust as a combustor to
drive the turbo to any desired speed.  All you need is a methodology to
determine the turbo speed.  

Lets take a few minutes, drink a beer and see how to estimate the
turbine speed.

Times up - answers ready to e-mail - ding ding ding - the winner is Doc
and Dave W.  Yupper - use a microphone plumbed into the compressor
output.  Dates back to the air raid siren day - a whirling blade making
noise.  

Pick a boost, any boost, measure its frequency of whine, record a bunch
of them, and add power in the exhaust as needed to the boost you wish to
be.  Dang - this could almost use the power on a stamp.

Or you could complain about "turbo lag", spend lots of money and still
be first loser.

-------------------------------

> Date: Mon, 3 Nov 2003 16:21:59 -0500
> >From: "Bruce Plecan" 
> >Subject: Re: MacGinnes Dreaming.
> >
> >Have you seen this?.
> >I've seen a turbine wheel, fracture and explode, and that was from heat
> >(lots of it).   Then a compressor wheel that exploded, after being in a vise
> >(don't ask).
> >I've heard/seen of cavitation, but usually using a restrictor, on the oil
> >feed, was the answer.   And that was only on a couple mis-ques on Indy
> >motors.
> >
> >It's just something so often quoted, yet just it just really doesn't make
> >sense to me.
> >
> >Bruce

It has always bothered me.  Not a problem ever on a suck thru carb'd
turbo.  Not on aircraft - Northrup never mentions it in training
manuals, not a problem on Corvairs, Buicks, Olds etc. 

MacGinnes himself says its not a problem with a throttle in front of the
blade.  Really would like to see some real material non internet myth
about compressor overspeed in a vacuum.

A BOV slows the turbine down - it still pumps air - but not as much as
no BOV.

The use of two blades was beginning to get popular Formula 1 until
turbos were outlawed.  It deliberately keep the blades at extreme rpm so
that the throttle response was instant at opening.

A linear tps relationship has not proved satisfactory because the first
blade can never close completely and must lag by some value the second
blade.

If you were looking for an optimum sensor to determine relationship, you
would again look at a compressor map and notice that as the air flow
decreases and you maintain the rpm, the pressure ratio increases.  By
tweaking the first blade to restrict the air flow to the amount that
causes the highest mid stage pressure, you would have the right value.
This would keep the compressor at high rpm at all times and virtually
eliminate perceived lag at throttle opening.

Is it worth it?  If I chose to do it, I will not report it.  Everybody
already disagrees with MacGinnes about everything he wrote about anyway.
Who da heck was he - just some engineer for a major turbo manufacturer
with a few patents on turbo design to his name.  He was not down with
the scene on imports and definitely not a photo journalist who can put
his own sound system into his ride. 

------------------------------------

Macs plan with multiple throttles is to unload the compressor so its
spinning in a near vacuum.  Taking the least energy and keeping the
compressor at a high rpm.  Remember the turbine is driven by the exhaust
gases and with a closing throttle, the turbine is fast losing its driver
power.  

Venting the output air back to the inlet means the compressor as its
spinning down must continue to pump the same amount of air but with much
less power and the compressor slows down radically.

Mac's way, as the power comes off, the compressors is pumping much less
air and while not in a vacuum, is not needed near as much to turn as if
the normal spin down air flow was allowed.