If a search engine brings you directly to this web page, then for more
written information go to our
Axle Assembly page.
To print this web site page try Internet Explorer 5.5 or 6.0. However, these and
others may
produce strange results at times.
|
A Pinion Preload Adjust Cycle Plot vs Time.
|
PA1 --- Pinion Preload Drag Torque vs Time and Corresponding Nut Torque vs
Time Plot
of a Trio Pinion Preload Adjustment Machine Cycle with an Aluminum Carrier, 1982.
This plot is from a Trio machine with Betatronics gaging equipment controlling the
adjustment cycle. This is the first Trio machine that combined the two measurements
and was installed in production in early 1982.
In 1972 Betatronics provided the first electronic control of pinion adjustment
on a Trio machine installed at Pontiac Motor Div. of GM, and therefore probably any
pinion preload adjust machine. Prior to that a force balance air switch was Trio's
method. This line ran close to one part every 15 seconds. Two Trio machines were used
to meet this thruput. Our initial experiments on line (this was a retrofit to existing
machines) produced a lot of rejects and required 4 to 6 operators with impact wrenches
to fix our rejects as we worked on parameter development. This equipment used analog
circuits to amplify the straingage signal to the 1 volt level, but beyond this all
processing was digital, including the threshold limits. This was without a microprocessor
because it predated the availability of microprocessors.
The Trio machines support the pinion on a pair of centers, the carrier is free floating
on it's bearings and restrained only by a force transducer used to measure the
drag torque. This means there are no residual torque errors other than some very
small errors in the transducer mechanism. There is an air motor (sometimes hydraulic) that is used
to pre-rundown the nut to bring
all components in contact with the collapsible spacer. This is done while the
pinion is stopped. After pre-rundown the pinion is rotated and two
clutches, fast and slow, are
used to rotate, thru gearing from the same drive motor, the pinion nut
relative to the pinion. Built into the nut driver shaft is a nut torque
transducer. The Betatronics gage controls the pre-rundown time, the spindle
motor speed, the clutches to achieve the desired pinion drag torque, and
finally to test the drag torque and nut torque. Both drag torque and
nut torque are compared with preset
limits at the end of the cycle to determine if the part is a good part.
Note, nut torque can not be controlled in a collapsible spacer application
and it will be whatever it will be.
Also nut torque is a function of nut angular velocity.
The plot shown above was made 3 Feb 1982 and had no seal because it was
easier to assemble and disassemble for testing without the seal. The residual
drag torque, about 1/2 #-in, before bearing contact is due to the carrier
weight on the inner bearing. This increases just slightly before bearing
contact because the outer bearing starts to make contact. With a seal
installed this residual drag torque typically would be 3 to 4 #-in
higher and more noisy.
In this application the bearings and flange are pre-pressed to contact the
collapsible spacer, and the nut is finger started about one thread. The carrier
is then loaded into the machine and the cycle started. The carrier is raised and
the centers are engaged. A "regulating unit" is brought into the tube bore. The
"regulating unit" is the torque transducer, and restrains rotation of the
carrier.
The driver that engages the flange or yoke is locked and does not rotate.
The clutches are disabled and an air motor rotates the nut thru a one-way clutch
until stall. In this particular cycle stall occurs at 4 seconds. The prevailing
torque of the crimped nut is about 30 #-ft. You see a gradual buildup to
the full engagement of the crimp.
The peak torque at stall is about 140 #-ft and partly due to inertia. The
steady state stall is about 120 #-ft. It is important that this stall
torque be below the torque to collapse the spacer. Note, if the collapsible
spacer is missing, then the unit will lockup from pre-rundown torque.
In a collapsible spacer product the design is such that there is guaranteed
endplay at the start of the cycle. This is typically 1/16".
In this application pre-rundown was allocated a fixed 7 seconds, later
applications are
dynamic. After pre-rundown the pinion driver brake is released. Thus,
you see the drag torque go from zero to about 1/2 #-in when the drive motor starts.
After a 1 second test for excessive drag torque, someone may have left the
collapsible spacer out, fast motor fast clutch is enabled. Here this occurs
at 8.1 seconds.
This cycle the nut torque rapidly builds to about 145 #-ft where the spacer
starts to collapse. Different individual spacers may have quite different
characteristics even though from the same manufacturing batch. This torque
gradually builds to about 230 #-ft and then remains approximately constant.
During this time the drag torque does not change until bearing contact. After
bearing contact drag torque rises at about 12 #-in/sec. This is fast motor
fast clutch, 90 rpm for the motor, and about 6 rpm for the nut relative to the
pinion.
At a preset threshold the fast clutch turns off and the slow clutch is
engaged. Then somewhat higher the fast motor changes to slow motor. Final
adjustment is slow motor slow clutch, and here drag torque rises about
3 #-in/sec. When final drag torque threshold is reached both clutches are turned off.
A dwell is initiated, then drag torque is checked. Note that this part has almost
3 #-in peak-to-peak variation during the test period. This is primarily roller
noise. The part should be tested for drag torque later, just before case is installed,
because after banging around, being rotated, and other variations the torque may
change, especially with an open cup situation.
This plot is very much an ideal build. A lot of parts do not build this well.
There are many different shapes to the nut torque. The drag torque has a
lot of noise during adjustment, some that is due to torsional vibration.
During the preload test period at the end of the cycle we check the
peak to peak value against a limit, typically 6 #-in. This is to catch
very bad bearings, bent stone shields, etc. The peak to peak magnitude is
very much affected by the measuring system bandwidth. With a bad oil and a
hand torque wrench I have seen peak to peak greater than 50 #-in on an
axle with 20 to 25 #-in average torque. A twice per revolution variation
may result from ovality, most likely bores. A once per revolution is a
parallelism problem. And a variation slower than pinion velocity is most
likely a cage problem. See photo P1 on AXLE PHOTOS page.
Total cycle time here is about 38 seconds excluding part loading.
Today, meaning late 90's and on, there is much more nut torque variation
from nut to nut because cadmium plated nuts are banned.
|
Note 1:
If a search engine brings you to this site and you do not know
why, then save this page to a file on your computer (under FILE
you would use SaveAs) and after saving, then open the file with
a word processor and search for the words individually that
you used in your search. For example if
the words --- drag torque --- bring up this
site and this page of this site, then you would first search for
drag or torque and look around those locations.
Pick If_search to return to beginning.
.
Copyright © 2003, 2004, 2005 Gordon A. Roberts
All rights reserved.
050128-1130
|
|