Table of Contents
Advertisement
User Manual
Appendix B – Torsion Compensation
Torsion Factor:
For all Stanley electric assembly tools, the angle information is based on the rotation of the
resolver which is directly attached to the rotor. This information is used for motor commutation,
and it also serves as an angle encoder. The rotation of the tool output can be determined by
dividing the rotor angle by the total gear ratio for the tool.
All things can deflect when loaded. Just as a long steel bar attached to a socket to produce high
torque deflection, likewise, the gears within an assembly tool will deflect when subjected to torque
loads. In effect, the gears act as a torsion spring between the rotor and the socket, and it is the
deflection of this spring that can give false angle data. In addition to the angular deflection within
the gears of the tool, there can also be deflection of the parts of the joint.
Whenever this deflection is present in the tool or the joint or the tool mounting device, the angle
information derived from the resolver indicates a larger angle than the tool output actually rotates.
This error is directly proportional to the torque level. That is, the deflection at 40 NM will be
twice that at 20 NM.
In a torque versus angle curve of a fastening cycle, at the end when the torque reaches its
maximum value, the angle will also be at its maximum value. After shut-off as the torque falls to
zero, the angle should remain at its maximum value. But in the typical torque versus angle curve,
as the torque falls to zero, the angle also appears to fall some amount. This is not because the
fastener is being loosened. It is actually the resolver indicating that the angular deflection of the
gears is relaxing to the neutral position. In this case, the maximum angle indicated at the
maximum torque was incorrect. The resolver indicated more angle than the tool output actually
rotated.
To correct for this slight error in angle data, the Theta controller has a Stanley-exclusive solution.
The Torsion Factor allows the operator to input a value that compensates for the torsional spring
rate of any part of the fastening system (the gears of the tool, the joint components or the tool
mounting device); this factor is used to correct the angle reading throughout the fastening cycle.
Enter this factor as Degrees per NM; its default value is zero. If the default value is used, there
will be no angular correction. If a value of 0.1 is used, each angle data point (every millisecond)
will be modified by subtracting 0.1 times the torque value. For example, at 15 NM, the controller
subtracts 1.5 degrees from the angle reading for that sample. At 30 NM, the controller subtracts
three degrees for that sample.
The easiest way to determine the correct value for the Torsion Factor is to look at a torque versus
angle trace with Torsion Factor set to zero. The amount of degrees that the socket appears to
loosen after the maximum torque divided by that maximum torque is the Torsion Factor. For
example consider a torque versus angle trace that indicates a maximum torque of 40 NM, the
maximum angle at this torque is 50 degrees. The angle appears to loosen by four degrees as the
torque drops to zero. The Torsion Factor can be determined by dividing four degrees by 40 NM to
arrive at a Torsion Factor of 0.1 degrees per NM. When this value is entered into the Torsion
Factor parameter, each angle reading will be corrected by this factor. When this factor is set
correctly, any torque versus angle trace now indicates no apparent loosening of the fastener as the
torque drops to zero after shut-off (which is exactly as it should be).
Angle Validation:
Now that the angle can be indicated with great precision, the other challenge is to validate these
results against a master torque/angle transducer with monitor. This is not as simple as setting both
the controller and the monitor to the same snug torque and comparing the resulting angle.
It has been found that a tool's torque trace never tracks exactly the same as the external. The
calibration is only the average of a number of readings (generally at a high torque near the
58
Advertisement
Table of Contents
