Subject: Turbo compound update.
From: ACRE
Date: 9/6/2004, 7:45 PM


Variable speed compressors are not new.
Some old messages on the subject.

______________________________________________________________________
Subject:         Variable ratio turbines and blowers.
Date:         Mon, 10 Aug 1998 08:14:33 -0700
From:         Paul <rotaryeng@earthlink.net>
To:         bit@any

I went to a car show on Main Street in El Segundo last Saturday
and saw an old hot rod with dual Paxton Centrifugal superchargers.

I contacted a TWA mechanic co-worker of Robin's whom I knew
to be an expert on old hot rods. Turned out his father was a TWA station
manager and my expert had flown at the age of 12 on a twenty hour
non-stop flight to Paris from LA on one of the 1649 R3350TC Connies.

He also knew all about the Connies and the R3350TC.

He filled me in on the history of the Paxton supercharger. It was
originally designed by McColluck(s?) back in the thirties or forties to
supercharge trucks in high altitude mining operations in South America.
This is the same company that made the four cylinder horizontal opposed
two cycle drone engine. The VR version of the blower had a variable
ratio, ball bearing speed up  drive instead of gears.

No need to explain the altitude and turbo compounding significance of
this I take it.

The automotive portion of Paxton Products company was recently bought by
one of the sons of Andy Granatelli... J.R and David Adams Jr. and is
located at 1250 Calle Suerte Camarillo CA. (805) 987 8660. They still
overhaul the VR version.

I intend going out there and taking pictures. This product was way ahead
of its time and is now making a come back with the addition of
intercoolers.

Paul

The Aircraft Rotary Engine Newsletter.              Powered by Linux.

________________________________________________________________________
Subject:         Variable speed blowers and turbines
Date:         Tue, 11 Aug 1998 21:44:18 -0700
From:         Paul <rotaryeng@earthlink.net>
To:         bit@any




George Moore wrote:

Paul, If I'm reading you correctly, the Paxton not only controls
turbine speed in relation to manifold pressure, it as also variable by
the planetary ball effect up to a 4.5 to 1 increase. Is their any
operating point, where the 2 variable mechanisms could counteract each
other ?  I remember Ford using the paxton and the Studebaker Avanti's
were Paxton Supercharged, but I thought people like ATI had sorta pushed
them aside. Obviously not !


The ball planetary is a fixed ratio device. Paxton was a little slow
to embrace intercooling. IMHO if they had recognized that earlier they
would have been much more successful.

Paul


The Aircraft Rotary Engine Newsletter.              Powered by Linux.
__________________________________________________________________________

Subject:         Variable speed blowers and turbines
Date:         Tue, 11 Aug 1998 22:50:55 -0700
From:         Paul <rotaryeng@earthlink.net>
To:         bit@any




This is an excerpt from the book called Street Supercharging by Pat
Ganahl 1984 and
1992 ISBN 0-931472-17-3 S-A Design Books North Branch Mn

There is much more about Paxton in the book and it is well worth the
$20.

Paul


The Aircraft Rotary Engine Newsletter.              Powered by Linux.



THE PAXTON SUPERCHARGER

The Paxton centrifugal blower is almost as venerable as
the GMC roots. As we mentioned in the section on swap meet blowers,
Robert Paxton McCulloch developed his first centrifugal supercharger
for flathead Fords well before WW II. This early McCulloch blower
was entirely different from the current model. During the war,
McCulloch also developed some Roots-type blowers on government
contract for use in landing craft and other special vehicles. One or
two of these blowers appeared on race cars, but none was ever
produced for aftermarket automotive sale. The current
Paxton/McCulloch centrifugal blower was developed by the Paxton
Division of the McCulloch Corporation in 1952, and the first kits
were made for flathead Fords in 1953.

Although this blower has undergone a few changes in belt drive
mechanisms over the years, the basic design especially in terms of
impeller and scroll housing shape, and the unique planetary ball
drive system has remained virtually the same to the present The
Paxton blower uses a plain radial blade impeller and a volute type
scroll housing. Actually only a small portion of the Paxton blower
unit is comprised of the impeller and scroll. Most of it houses the
impeller drive system, which is accomplished by five large diameter
steel balls which ride against two opposing "races." One race is
driven by the blower pulley (which is driven, via a belt, by the
crankshaft). A spring loaded clutch pack, much like that in the
engine's pressure plate, applies pressure to the back of the driven
race, which is directly connected to the impeller drive shaft Thus
the steel balls are sandwiched between the two races, and the clutch
pack applies enough force to maintain a friction drive from one
race, through the balls, to the other race. However, the rolling
action of the balls between the two races acts like the planetary
gears in an automatic transmission to step up impeller speed at
a 4.4 to 1 ratio over the blower pulley speed.


This unique design feature of the Paxton blower serves a few
important purposes. First of all, it is a consider- ably quieter
drive system than spur gears would be. Second, the ball bearing
drive system absorbs less energy to operate than step up spur gears
would. But most important, this planetary ball drive design
eliminates the major failure of previous centrifugal blowers for
high performance automotive engines the fact that rapid acceleration
or deceleration of the engine would strip the teeth off of
conventional drive gears in the blower housing, due to the extremely
high rotational velocity of the impeller and its consequent inertia.

In other words, a major asset of the Paxton blower is its nearly
fool proof drive mechanism. Another unique feature of the
Paxton/McCulloch blower introduced in 1953 was a variable speed
drive pulley. These blowers were known as the "VS" series (for
variable speed). The idea was to compensate for the "peaky" boost
curve inherent to the centrifugal blower design by driving the
impeller at a higher speed ratio at lower engine rpm, and then
"shifting" the blower drive to a lower ratio at higher engine
speeds. This feature would not only produce more blower boost at
lower engine rpm, or flatten the boost curve over the engine
operating range, but it would also limit maximum blower rotational
speed (to about 35,000 rpm) to safeguard against internal blower
damage.

The VS blower shifted gears by means of a special V-belt
pulley, which had one fixed flange and one movable flange. The
blower was driven from the crank with a single large, wide V-belt At
low engine speeds, the blower pulley would be spread apart,
allowing the V-belt to drop down between the flanges to an effective
pulley diameter of 3.5 inches, giving a blower overdrive ratio of
1.61 to 1 (in most cases). As engine speed increased, so would
manifold boost An electric pressure switch inserted in the intake
manifold would trigger a solenoid in the blower tension of the idler
spring, the tension of the spring between the diaphragm and the
movable pulley flange, and the pressure in the manifold were all
balanced to each other so that the blower speed would vary with
engine speed and boost, rather than just shifting abruptly from
one speed to another at a given rpm.

The reason we are telling you all this is because there are lots of
these early McCulloch/Paxton VS blowers out there, and (as we
mentioned earlier) they can often turn out to be your best swap
meet blower buys. After the VS blower kit was introduced for
flathead Fords in 1953, similar kits were developed for nearly all
the engines of the mid 5O's, and one source states that "tens of
thousands were sold year after year." Paxton Products will still
make the VS-57 blower (on special order), and they can rebuild any
units going back to 1952. According to Paxton's Rick Gattennio,
"Unless a guy drops it and breaks it in half, we can fix it" Typical
rebuilding of a Paxton/McCulloch would consist of checking the
impeller for balance, replacing the ball drive element, replacing
bearings and seals, and a thorough cleaning and repainting. At the
time this was written, a complete rebuild cost about $450
(including parts cost for the ball drive element, of about $300).
___________________________________________________________


Paul Lamar

Paul,

Somehow I failed to read the WHOLE message. Now understanding takes place. I
should have known.  Disc are rolling along perpendicular to the races.
And you want to use this gearbox instead of making a CVT.  Ah so. Nice:)
So it [has] been in your database all along.

Doug in Japan

In six years of running this newsletter letter there is a lot of stuff in my
data base
I have forgotten about. I spoke to Ed Cole's (ex GM CEO) son almost 30 years
ago about
turbo compounding the rotary at the Detroit Auto show. He was a ME professor
at U of M
at the time. Don Sherman introduced me to professor Cole and was a witness
to the conversation.  Apparently I was 50 years ahead of my time :)

Paul Lamar


Yes you were more than a little ahead of your time. Especially since you are
only 49 years old:) But now is the 'time' to make the unit though. And now
that I got a handle on the 'steel balls' and how they work it it seems
doable for a quick test 'as is'.  We can use the 4.4 ratio for the ball
planetary and a fixed ratio pulley system ratio to arrive at 4300 rpm on the
e-shaft with 35K at the turbo.

The big question is?  What is the maximum speed the gear box can handle?  We
know application design is  35K. I doubt Mazda's turbine can put out much
torque at that speed but if would be worth a try for a first test unless we
find a bigger turbine  before test time.

I will start planning a simple  exhaust manifold for using two turbos. I
would appreciate any thoughts for an inlet valves. I have lots of stainless
pipe both round and square in 3" and 4" sizes. I also have scraps of 2mm
thick  titanium plate if needed to make several  butterfly or slotted slide
valves. This test is by no means state of the art but I am betting it will
be enough to prove the turbo-compound concept.  At least in the world of R&D
in Japan the project looks more believable if the initial results are
modest, and steady step-by-step progress (through great suffering under
severe and near impossible conditions) can be demonstrated and documented.
Heck since this research is being done under the auspices of our officially
registered  non profit institute, Mazda may even support the later
development of the work. Of course you will have to be an consultant for the
project:)

Paul logistically I don't wish to buy a PSRU and C/S prop just yet, so for
dyno work I want to just load the engine with a water pump.  Since that load
can be recorded as  water pressure, engine MAP and rpm, I should be able to
extrapolate from that data to ballpark the corresponding C/S prop settings.
In other words if I want to simulate a take run and climb then I will want
to know how how much to load the engine.

Paul on your engine run test stand,  how did you measure the engine torque?
Inline transducer?

Included are a few pictures of the shop from whence my engine came. What a
mess!!! Parts on top of parts:(    Some rare, others scrap. And most parts
covered with  a healthy layer of dust and oil. No rust though. In the colder
months I and one of my staff are going to spend a few days assembling racks,
bagging rotary parts with silicon gel packs inside, taking pictures and
cataloging everything he has.  The cut off saw on the right is what we used
to trim off my 13B from this engine:)

Regards

Doug in Japan


I have not gotten around to making a decision on the torque measuring device.
Soon I will get to that. 

That guy is sitting on a gold mine :)

Paul Lamar

 
The Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2004 All world wide rights reserved.