Out-of-spec parts problems could be solved with … a #2 pencil
By Jim Anderton, technical editor
What's the most important tool your millwright or line technician owns? The obvious answer is the one between his or her ears, but number two on my list is the #2 pencil. Why? Because more important t...
What’s the most important tool your millwright or line technician owns? The obvious answer is the one between his or her ears, but number two on my list is the #2 pencil. Why? Because more important than the ability to quickly and repeatably repair a machine or line is to document the process in a way that reveals the hidden messages that your equipment is sending you about your maintenance practices, line operators and machine capability. An example from my own experience should illustrate what I’m talking about.
A “second op” machine downstream of one of our presses used air cylinders to push as-molded bushings into a small metal housing for an automotive OEM. The simple machine was hand-loaded, but used a computerized feedback system to monitor the applied force necessary to seat the bushings in the delicate, easily distorted housings. Maintenance was straightforward, and like most pneumatic equipment, when fed with clean, dry, lightly lubricated air, rarely went down. The trouble was, it also produced out-of-spec parts for no apparent reason at least twice a day. Some bushings were over-driven, and some not seated at all.
QA immediately swung into action and began the usual process. Was the machine operating normally? According to my tests, it was. Were the parts in spec? All checked out well. Had the molding process for the polypropylene bushings changed? A lot of time was spent at the press, investigating the possibility of a change in molding parameters or lot-to-lot variability in the resin. Power supply, compressor performance, ambient temperature and part contamination were all checked. Operator shifts were inverted to remove the human variable, and still the problem persisted. Engineering even contemplated an expensive part redesign to allow more insertion force.
After a couple of days of fruitless search, and in desperation, I grabbed a pad and a #2 pencil, and instructed the operators to note EXACTLY what happened every time the machine produced a bad part. Every sound, vibration, smell, anything, plus the operator’s initials and time.
The payoff came through a series of odd notations at the end of the “graveyard” shift and about two hours into the afternoon shift. “Machine clunks” and “Funny noise behind wall” were common notations. The breakthrough came when I noticed the correlation between the odd sounds reported by the operators and the time they occurred. It was usually at about 8:00 AM, then again near 6:00 PM. Breakfast and dinner, both peak electrical load periods. But the machine was pneumatic, and line pressures, according to the supply gauges, were consistent.
The machine was air-powered, but the force-sensing feedback circuitry was most definitely electric, and when the power company arrived to switch taps on the main transformers, the problem went away forever. The “clunks” were the pilots on the air cylinders operating erratically in response to confused signals from a voltage-challenged mother board, while the “funny noise behind wall” was the sound of the motors on the compressors in the adjacent room straining to start with inadequate current.
A simple note pad and pencil helped track down a problem that wouldn’t yield to an experienced technician with a $2000 oscilloscope. One warning. If you use this technique in your operation, tie the pencil to your clipboard with a string, or you’ll end up ordering them by the carload.
Jim can be reached at janderton@corporate.southam.ca
