Taking it Slow to Stop Burning Rubber

Our company had about a dozen similar injection molding machines, but only one would burn the rubber as it flowed to the die after breaks and lunch. Since a cycle was 20 to 40 minutes for 2 or 4 parts, depending upon the die, it was a big hit to productivity. The workers tried speeding up the feed so rubber would speed past whatever hot spot was causing the burn, but the condition persisted.

They topped out on the feed mechanism and asked for my help. They needed a feed mechanism that could go faster and was incrementally controllable, not just the 4 or 5 clickpoints of the current one.

I devised a system using compressed air that could be adjusted with a knob on the supply line for convenient side-of-the-machine adjustment--previously operators had to climb a ladder behind the machine to adjust the feeder. For moving the hot rubber along I chose off-the-shelf straight gear teeth. A 6-inch gear cost about $120, and an 8-inch gear cost about $700.

By my calculations the 6-inch gear would go a little faster. My cheapness got the better of me--perhaps a little bit was enough. If not, I could clean the pair up and return them for credit towards the $700 pair.

The mechanism was built and installed. When doing the calculations I counted on the stated force of the air motor, which is often rated at 100 or 120 psi while shop air usually hovers around 80 to 90 psi.

After it was installed, I was out on assignment for a few days. When I returned, I got word that it was too slow. I went to the machine to fetch my feeder for the gear swap--but the operator wouldn't let me take it.

"Haven't had any burned parts for three days," he said. "No scrap." The supervisor said he still can't explain it, but slowing down the feed solved the burning problem.

In all production problems an engineer should always consider whether it is the tool and not the production method at fault. At the same company one engineer worked late into the second shift to try out a new mold which had taken two months to machine and polish. The molds are supposed to be heated up prior to the first run, but the thermocouple read 40 degrees shy of the necessary temp. The engineer turned up the heat, then turned it up more.

After two hours of this, he could feel the heat radiating from the machine, yet still cranked it up a bit more. By the time it dawned on him that his thermocouple might be defective, he'd reached 1400 degrees F. He turned it down and removed his die, which had started to warp like a Salvadore Dali painting.

It was unsalvageable. Even worse was that product wouldn't make the spring lineup in a very seasonal business.

Tell us your experience in solving a knotty engineering problem. Send stories to Chris Wiltz for Sherlock Ohms, and read more cases here.