Subject: Traction Drive for turbo compound
From: Rotary Engine
Date: 8/4/2008, 7:56 PM
To: AAA Put this in the To box


Paul,


While researching turbo-compounding, I came across pictures of your
traction drive on your site, and was wondering if you
had made any progress on it since the date on the pictures (5-22-06). Does
it
work?

Also, I can find very little information about traction drives in
general, but I'm interested in reducing a 100,000 to 150,000 rpm
turbine down to manageable speeds for use on a conventional
reciprocating engine or rotary.  It looks a little like a planetary
differential without gears. Is it a continuously variable
transmission? Can you describe its operation? Would you share the
Solidworks files or drawings used to make it?

I'm fascinated by your description of the possible efficiencies in
turbo-compounding a rotary engine, and I'm very excited about the
implications of Mazda's new 1.6L all-aluminum direct-injection rotary
in homebuilt use.

Thanks,
-Chris

Traction drives do work. They have been widely used for years.
Doug on here has a Rotrex supercharger that uses a traction drive.
Paxton/McCulloch made truck super chargers starting in the late 1930's.

Yes I can send you some step files which work the best in Solid Works.

Paul Lamar

Paul,

That's fascinating. Seems like a very reasonable way to develop high
reduction ratios with shafts of otherwise unreasonably high speeds.

I have a million questions:

What exactly is meant by "high contact surface pressure", ie how does
one calculate the roller preload necessary for it to work? Does wear
substantially decrease the interference fit (or preload) over time?
What type/viscosity of oil is used? Is the ideal roller surface finish
very smooth?

Step files would be much appreciated.

Thanks,
-Chris

You need the torque out of the turbine. You can work backwards by estimating
the torque knowing the HP and the RPM of the turbine.

My guess on a full chat 200 HP rotary would be 50 HP out of the turbine.

HP eqauls torque times RPM divided by 5252.

My wild ass guess is about 1 foot pound at 100,000 RPM.

Then you need the coef. of friction for steel on steel in the presence
of traction fluid. There are several auto tranny traction drives on the
market
for cars so that is a good source of traction fluid. My guess on the subject
is .3 coef. Next you know the geometery and you know the contact pressure
you can calculate the torque capacity. FEA will give you the contact
pressure
of a steel ring deflected into a triangular shape with three internal
rollers.

Here is a pdf on traction fluid.

You insert the ring in a large lathe 3 jaw chuck and insert the rollers and
release the chuck.

Doug may have done static torque measurements on the Rotrex.
I basically copied the Rotrex.

Paul Lamar


That should be no trouble for me to assess and tweak. I've been
second-guessing the simplicity of the drive -- it really is
straightforward.

Do Nissan/Audi/etc CVTs use a traction fluid? Figuring accurately for
coefficient of friction and whether or not to consider it linear
(constant with increasing contact pressure), and how critical roller
finish is are really the only issues I'm uncertain about. If .3 is
what you guess it to be, that gives me a good place to start.

Thanks,
-Chris

Yes they do Chris. It should be easy to find out what brand it is  as
that info is probably in the owner's manuals. I suspect the finish should be mirror quality.

BTW There is a golden opportunity out there to get a huge grant from
the DOE to study the rotary turbo compound. Now is the time to do that.

On the other hand we have had over 5000 hits on our youtube P-port
RX8 engine and only about 2500 total for all  three of our turbo compound videos. Obviously the youtube audiences is more interested in more RX8 power than they are in better RX8 MPG I am sad to say.

Paul Lamar


The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Youtube key word UTUBPLEASE
Copyright 1998-2008 All world wide rights reserved.