Subject: Flex Plate Failure - Improvements
From: Rotary Engine
Date: 7/15/2011, 8:55 PM
To: AAA Put this in the To box
CC: Tracy Crook <tcrook@rotaryaviation.com>, Tracy Crook gmail <rwstracy@gmail.com>


I finally got to the hangar and measured out my S5 turbo flywheel and the
remains of my S5 flexplate.  The results are very encouraging, even as good
as the possibly could be to make the swap, but I only had ten min at the
hangar before I had to leave to pick up my daughter from daycare, so they
are one try and slightly sloppy... I make no guarantees.  The height
measurements of the flex plate were particularly suspect because, well, mine
is all warped and broken.

*Ring Gear Diameter* - Perfect fit.  As we suspected, same size, same
tooth spacing.

*Ring Gear Fore-Aft Placement* - As measured from the underside of the big
nut. (for all further reference, aft means aft in the CAR).  The flex plate
puts the aft side of the ring gear 1.290 inches from the front of the big
nut. The flywheel puts it 0.913.  So the flex plate puts the ring gear
(appx) 0.377 in. further aft.  This is actually fortuitous because it just
so happens that the ring gear thickness is .365 ...    So all we have to do
to fix the discrepancy is weld and extra ring gears onto to the flywheel
ring gear and there will be teeth that the starter can reach:(bottom of the
page.. the cost $80)
http://mazdatrix.com/flywheel.htm
It may also be possible to fix the discrepancy by changing the starter...
will look into that.

*Damper Plate Placement* - On the flywheel, the mounting surface is proud
(or aft) of the ring gear.  On the flex plate, the mounting surface is
recessed (or forward) from the ring gear. And it all pretty much works out.
That position on the flywheel will vary somewhat depending on which flywheel
you have and how much wear you have on the clutch slip surface. But on my (I
assume well-worn) flywheel the placement was 0.965 while the flex plate was
0.950.  Pretty damn close and given measurement error it might me meant to
be right on.  Or, any difference could be made up by trimming down the
spline on the damper plate or removing or adding spacers between the two.

So in my mind, a flywheel is the way to go.  Stout, easy to get, pick your
weight, only mods seem to be drilling the mounting holes and welding on a
second gear ring.

--
David Leonard

Turbo Rotary RV-6 N4VY
http://N4VY.RotaryRoster.net
http://RotaryRoster.net****


I'm still in analysis mode on this myself.

So far the possibilities are:

1. The flex plate may have been warped when installed  ( I bought several
that were)  I think the standard should be close to zero runout or wobble on
the flexplate in view of this failure.

2.  The flexplate could have had cracks when installed  (I also bought
several that had cracks and they were not visible unless you looked closely)

3.  The dang things (RX-7 flexplates) are not strong enough for boosted 13Bs
or 20Bs.

I do NOT think the problem is fore & aft travel of the input shaft.  The
RD-1x drives have had thrust bearings from day one.  There is no thrust load
at all on the damper, flexplate or crankshaft imposed by the drive.   This
was not the case with the Ross drive I used at first.

I have over 1000 hours on the same flexplate on my NA 13B.  I have only 55
hours on the 20B.   But I'll be checking both.  I did inspect both
flexplates VERY carefully before installing.

If I found that a change were necessary, I would consider two possibilities.


1.  Use a modified aluminum racing flywheel used with an auto
counterweight.  (Or possibly a lightweight steel one if there is such a
thing)

2.  Use an RX-8 flexplate with an adapter plate made of aluminum or steel
sheet to go between the flexplate and damper.  The RX-8 version looks
stronger than the RX-7. I briefly studied this when good RX-7 flexplates
were getting hard to find.   I did not develop a formal procedure for doing
this but it looked like it would be pretty easy.  Anybody with spare time
out there want to take a closer look?  Hint:   The damper bolt pattern would
be offset 45 degrees from the RX-8 torque converter bolt pattern.

Tracy Crook,   RWS

Mark Steitle's is cracked too.

What do you think of this Tracy?

I think just one piece of .090" 4130 would be lighter. We don't need
the clutch wear surface. 12 inch diameter disk of steel
would be:

A = pi x r squared.
r = 6 inches.
r squared = 36
Times pi = 113
Times .090 thick = 10 cubic inch
steel weighs .3 pounds per cubic inch
so it is 3 pounds. Less ring gear.
You could go up to .125 thick and still not weigh the same.
14 cubic inch times .3 is 4.2 pounds.

Not counting the large hole in the center or the 7/8th inch high aluminum
ring.

I say make it.
It is the holes that are starting the cracks.
The ring gear is available separately from Mazdatrix.


Paul Lamar

-- 
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Youtube key word PaulLamar2
Copyright 1998-2011 All world wide rights reserved.