A system for quick change

Modern Machine Shop, May, 1995 by Tom Beard

The larger issue was the production equipment. It was not so much a matter of inaccuracy in the motion of the lathes - in fact, even though some of the machines are more than ten years old, they still perform surprisingly well in terms of repeatability - as it was the variability in receiver position from station to station on the turrets. That's not such an easy thing to fix, so rather than trying to, they decided to quantify the variances, and simply factor them into the initial tool position calculations. Thus, they embarked on a rather painstaking process of indicating and recording every "zero" position of every station on every turret in the shop. Once the data was gathered, it was all entered into batch files in the presetter control. When the crib attendant measures a tool, the presetter automatically applies the appropriate compensation factor for the destination tool station.

That doesn't mean they've been able to totally dispense with conservative first-off cuts, but they have gotten there in the majority of their work. Says Mr. Minnick, "If the tolerance is at least plus or minus a thousandth, we go right at it," meaning that all offsets are set at exactly the readings from the tool presetter. It's only when the tolerances are tighter, he says, that they "back off a bit."

Faith And Science

Maybe this all sounds overly fussy if you're accustomed to machining fairly expendable parts. Maybe you expect to scrap the setup part anyway. At Danville Metal, however, workpieces can easily be worth thousands of dollars, so over the years they've developed a strong disdain for scrap of any kind. As such, Mr. Minnick knew it wouldn't be easy to convince operators to let go of the practice of trial cutting on new setups. But he also knew that unless they did exactly that, the shop would not be able to achieve their setup reduction goals.

Since he was confident the system worked, he viewed the challenge as primarily one of education. But rather than trying to deal with it through instruction, he felt it would be better to let the operators see it for themselves. As an experiment, five operators were selected and given complete access to the tool setting system over a three week period. The operators preset their own tools, loaded and ran them, and returned them to the crib. Most important, they personally monitored the predictive accuracy of the presetting system. And they became convinced. Thereafter, all other operators in the shop were rotated through two-week cycles in the crib, so everyone could observe first hand the total workings of the tool management system.

Interestingly, they've applied essentially the same tool setting methodology on a row of manual VTLs the shop still employs on a range of work. The lathes are all now fit both with KM tooling and DROs. Obviously, you can't enter offsets into a manual machine, but they can apply the offset data with digital position display, and consequently, they have been able to eliminate the majority of the trial cutting here as well.