A torque meter for a helical gear planetary gear box is simple.
Use an annular hydraulic cylinder to measure the helical
gear thrust. Last four jpg's.

Straight cut gears present another problem. Solution: machine a very course
pitch
acme thread on the outer portion of the quasi stationary ring gear and measure
the reaction torque with an annular hydraulic cylinder. Exactly the way they
do
it in a PT6 that has straight cut planetary gears.

All we need is engine torque. With it we can calculate engine HP. If one wants
it electronically use a pressure sensor instead of a cheap pressure gauge on
the instrument panel with a annular hyd cylinder measuring the torque.

Mistral ought to incorporate this in their straight cut planetary gear box.
It would be a big plus for buying their expensive PSRU.

The challenge is come up with a SIMPLER PRACTICAL system cheaper than
this approach that will survive in a hot gear box environment.

My thanks to Doug for stimulating me with his private emails on this subject
:)

Paul Lamar ...No rotor no motor.

Paul,


Maybe I can stimulate or exasperate you some more:)  Mistral and the Bell 47
gears box both have their ring gears securely bolted in place. There is no
quasi movement about them. Bell presses them in as well as bolts them on.
Large torque is expected on a helicopter obviously.

 I absolutely agree that if the rubber torsional absorber could be there
then the problem is elegantly solved. This was the first method I eliminated
years ago.  The second method was with strain gauge to measure the force
between the engine flex plate and the sun gear drive plate.   Sea and Land's
and others company units have done this but  you need to have some
electronics on board the plate affecting balance no doubt as well as a
wireless connection to get the data off the spinning plate.  Now this is
possible but not simple or inexpensive. $5,000. Ditto special torque sensors
sandwiched between the prop and the output of the PSRU. These devices move
just enough to just a strain gauge to stretch and change resistance.  Here
also a slip ring or wireless  method is needed to access the information.



Paul You did mention the following and I would like to share it with
everyone.

"It won't work having one sensor on the front of the engine and one sensor
on the prop flange. Too much slop in all the parts."

 Yes I suspected that, thank you for confirming it. Lots of little torque
reversals happening so an accurate measurement is not possible no doubt.

"You still need a short length of shaft from the source of the torque
to the load not bathed in 225 degree F oil. You don't have it."

Actually I do if you consider 75mm long enough.  There is a 1.25" hardened
steel splined shaft running from the psru sun gear mount plate (hot wet box)
to the sleeved collar attached to the torque dampener on the counter balance
in the dry  bell housing. The Aluminum starter gear wheel and turbo compound
grooved pulley will  be splined cut and slipped on.

In summary as I understand what you are saying, torque  has to be measured
between the engine output = source of torque, and the PSRU/prop input=
source of  main load. So one way to do it is to  measure the twist on the
shaft between those two sources right?

Doug in Japan

I'll say!!!   I am advocating unbolting them!!!!!!

Last question. You are not going to do it with a CD in 200 F hot air.

If you insist on doing it electronically one way is to use two polished
flats on the shaft with laser shinning on them. As the shaft twist the angle
of the mirror causes the beam to be reflected off at a different
angle. You can buy a CCD that in essence is a bunch of closely
spaced photo cells to measure this angle.

http://www.taosinc.com/

The laser and the linear optical sensor can be seven inches away out side
the bell housing in cool air. A computer can read the optical sensor in a
micro second.

7000 RPM is only 166 revs per second. Or one rev every .006 seconds.

One micro second is .000001 seconds therefore any old computer can execute
6000 instruction every rev!!!! That is a very slow computer. 1975 era.

Single chip computers now-a-days are at least 10 times faster so they can execute
60,000!!! instructions in one rev of the engine at 7000 RPM.

Paul Lamar ...No rotor no motor.

The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.