|
Aircraft parts |
|
Home | Software | Knowledge Base | Customer service | Products |
|
|
Home> Maintaining your Continental and Lycoming engine Torque Wrench Accuracy and Moment of Inertia
Torque accuracy achieved during calibration is not what you will get in actual use
Not only does torque accuracy depends on how the operator uses the measurement instrument - the human factor - but it also depends upon how accurately the calibration lab duplicates real world usage. I call the latter reality errors. Ideally, calibration of any measurement device should try to duplicate actual conditions under which the instrument will be used. There is an implied trust between the calibration agency and the user that the measurements taken during calibration will relate to the results obtained during use. If the measurements a calibration lab obtains do not relate to the results obtained during usage then the calibration lab's measurements have no meaning. Sources for reality Errors The reality error occurs when the calibration lab tests the wrench in
artificial conditions that do not duplicate real world usage of the wrench. Under carefully controlled conditions a calibration
lab can achieve very accurate and repeatable measurements. The error occurs in that these measurements
do not duplicate real world conditions. For example, many torque wrench calibration labs use mechanical pullers.
These pullers:
Using a name brand 3/8" torque wrench we will apply pressure to the wrench handle just as you might in normal usage:
Using the first example, in normal usage you should expect to see this wrench to click anywhere from 790 to 810 inch pounds. A calibration lab using a mechanical puller would only detect one reading - probably the 800 reading. What is the correct result? They are all correct in that the wrench gave the same results under the same conditions. However, when the lab reports that this wrench, when set to 800 inch pounds, will snap at 800 inch pounds, they are not correct. This wrench, under normal expected usage conditions, will snap anywhere from 790 to 810 inch pounds. In this case the calibration lab introduced the "reality error". Lets now say that the calibration lab advertises that their equipment can calibrate with an "accuracy" of 1/2 of 1% Should we be impressed? This means that we can be very confident that the 800 number that the calibration lab reported is indeed 800. This is nice. But when you use the wrench you may use the wrench horizontally, vertically; you may apply the force fast or slow; or have a slight inward or outward pull. Your and will also not apply the force at the same spot that the calibration lab did. We know how "accurate" the lab's equipment is; We don't know how "accurate" the wrench is! Isn't that what we want to know? If I had a wrench that always snapped at 800 inch pounds regardless of how the wrench was oriented, how the force was applied, and how I applied the force, wouldn't that wrench be more accurate than the wrench used in the above example. A calibration lab using a mechanical puller would report both wrenches as being the same when obviously they are not - at least not in the real world! Using a mechanical puller is not the only source for reality errors. How many of you take a torque wrench out of the box where it has been stored and take some "warm-up" snaps. Often, the first snap of a wrench that has been in storage is not accurate. The reality error gives the user a false sense of accuracy - not only
in the calibration lab's accuracy, but also in the accuracy that you will
achieve when you use the torque wrench. Engineers who specify torque values need
to allow for unavoidable errors when
applying a torque.
Snap style torque wrenches exhibit a variance in torque due to:
Moment of Inertia A rotating body has the same tendency to maintain its state of rotational motion that a body moving in a straight line has to maintain its linear motion. The moment of inertia is a measure of a body's resistance to changes in rotation rate, analogous to mass as the measure of resistance to changes in transitional motion. Specifically, torque T and angular acceleration (designated by the Greek lower-case letter alpha) are related through the moment of inertia I by the equation T = I(alpha) , just as force f and acceleration a are related through the mass m by the equation f = ma. The moment of inertia depends not only on the mass of the body, but also on the distribution of mass relative to the axis. This distribution accounts for the fact that objects of various shapes with the same masses and diameters (such as sphere, solid cylinder, hollow cylinder, or wheel and axle) will not take the same time to roll down an inclined plane. Objects whose mass is concentrated near the axis have the smallest moment of inertia and thus, reach the bottom of the plane sooner than the others. If you apply the torque to your wrench quickly the mass of the
wrench resists acceleration because of its moment of inertia. This resistance
to turning reduces the turning force on the fastener. You apply 40 ft-lbs
of force and the wrench snaps at 40 ft-lbs but 5 lbs is used up in overcoming
the wrench's moment of inertia and never gets to the fastener. There is
probably a better (more accurate) explanation of how the Moment of Inertia
can introduce inaccuracies to our torque values -- if you have one please
contribute to our understanding.
Where you apply the force to the torque wrench makes a difference in how much torque is applied at a give setting. For example, today, just for fun, I was calibrating a Snap-On click style wrench and thought I'd record some readings for you. With the wrench set to 100 in-lbs, I applied the force on the wrench handle at the very end of the knurled portion of the handle. The wrench clicked at 106 in-lbs., six percent high! Next, I applied the force at the inside portion of the knurled knob. The wrench clicked at 98 in-lbs. six percent high or 2 percent low! Download our Torque Wrench Extension Calculator from The Mechanic's Toolbox this small windows program computes what wrench setting to use when using an extension to your torque wrench
|
Copyright 2003 by
Sacramento Sky Ranch Inc. All rights reserved. Prices subject to
change without notice. Not responsible for typographical or misprint. |
Broken
cylinder hold-down stud on Continental IO-520 shows evidence of a loose joint:
fretting (rubbing) damage, smooth break to stud (1/2 is missing). Loose joint
can be caused by improper torque, OR joint loosening, or loss of preload. Smooth
break is typical of fatigue failure. Fatigue requires cyclic stress which in
turn requires a loose joint. For