Properly sized and maintained, the torque scatter of a good (though much less expensive) clutch tool will not be substantially different than a programmable tool, if the latter is simply programmed to run at the same free speed and shut off at the same torque as the “dumb” tool. You should only make the additional investment for the opportunity to improve joint quality.

Many engineers don’t take full advantage of that opportunity and fail to maximize their investment or their product quality. While there are several reasons for this, one big culprit is uncertainty as to how to determine the parameters for programming a good tightening strategy.

I’ll address the most common situation: Target torque has been established by product engineering and the plant engineer must develop the tightening strategy. We’ll also assume that the joint doesn’t have unusual characteristics or requirements, so the most widely used strategy of torque control with angle monitoring can apply. Factors such as equipment, rundown, tightening and process limits should be examined.

Equipment. This factor is applicable for any tool that has the programming capability to implement it. Though not directly tied to programming, one requirement for effective tightening strategy development and process control is the ability to plot torque vs. angle traces.

Note that plotting torque against time is not the equivalent, as changes in spindle speed affect the slope of the trace. This, in turn, suggests joint behavior that is not actually present. Simultaneously plotting both torque and spindle speed against angle is the ideal means for understanding joint behavior during tightening.

Rundown. This step covers advancing the fastener from the point of socket or bit engagement until shortly after clamp load is being generated. The primary question in this step is how fast the spindle should turn. If the fastener is free-running (no resistance from thread-forming or a locking feature), maximum speed can be used if socket engagement can be maintained.

Thread interference may require reduced speed, particularly when the fastener or nut member is a material other than steel. This can be determined by comparative testing. Rundown, and sometimes tightening, is more consistent when all adapters are eliminated and the distance between the fastener and tool drive is minimized. Take advantage of soft-start capability when the fastener’s drive features are difficult to engage.

Tightening. This step starts at completion of rundown and ends with tool shut-off at the program’s conclusion. In most cases, the transition to tightening involves reducing spindle speed. At what torque this transition should occur, how speed is reduced and to what value, are decisions that should be made after consideration of each joint. This knowledge is nearly impossible without access to torque-angle-speed traces.

The tightening step should start after the joint components are fully aligned, indicated by the torque-angle trace’s slope. This slope is a measure of resistance, which is more useful to think of as joint stiffness. Stiffness increases as component contact area increases, and then stays constant until clamp load is great enough to yield components.

How close the transition point should be to the start of this linear portion of the trace is largely a function of how much the spindle turns before target torque is reached and how speed will be reduced. Earlier transitions tend to reduce scatter at an increase in cycle time. However, in most cases, rotation is so short the time penalty is trivial.

Another speed-related phenomenon that should be avoided is stick-slip, a condition where friction fluctuates, causing “noise” in the trace as torque oscillates up and down. Premature shut-off is often the result. Experimenting with changes in speed should be the first response. Both these conditions usually produce a squealing noise, which should also be a flag for investigation.

Final tightening speed should be related to fastener diameter, as friction can be influenced by the relative velocity of mating surfaces. Common ranges for clamp load-critical joints are 40 to 80 rpm for M8 (5/16 inch) bolts down to 10 to 20 rpm for M16 (5/8 inch) bolts.                                                                                 

*Process Limits. While establishing nominal tightening strategy without studying joint behavior is common, not taking advantage of programmed torque and angle windows of acceptance is endemic. Establishing these limits can detect cross-threading, bad threads, improper locking features, wrong bolt length, changes in friction and joint yield.