Ejection dejection
By Jim Anderton, technical editor
Getting molded parts out of the tool and "down the spout" is a critical part of any mass-production molding operation, including the more esoteric processes like rotational molding and hand lay-up. Bu...
Getting molded parts out of the tool and “down the spout” is a critical part of any mass-production molding operation, including the more esoteric processes like rotational molding and hand lay-up. But in injection molding, failure to eject reliably is more than an inconvenience, as I discovered one day on a medium-tonnage press running a small PE automotive part.
The eight-cavity mold produced a dimensionally non-critical (this was before Six Sigma) small cylinder which was ejected conventionally then blown into a hopper which gravity fed into a small conveyor for the “second op” process. Sprues fell by God and Isaac Newton into a barrel where they could be carted away periodically to be frowned upon by the production manager.
The system worked well enough to keep good parts out of the regrind bucket, and I doubt a hot runner system would have lasted long given the housekeeping practices at the time, but every once in a while, mis-ejection would cause a stoppage. “Stoppage” was actually a good thing. That meant that the diminutive operator was close enough to hit the panic button when the mold made that characteristic dull thud that suggested ejector pins that weren’t cycling in a straight line. If she didn’t, it usually meant a crash, bent pins, and lots of expensive downtime.
After re-startup, the machine always checked out O.K. Run the process at or near an economically viable speed, however, and the problems reappeared. Finally, the press was changed over for long enough for me to try a few things to get those needles to retract a little faster.
The first step was to check for binding or roughness manually. They slid like butter. I then loaded the cavity with spray-on mold release and added a little heat to see if there was some kind of buildup or gumming action. Nope. I finally removed the pins and checked that the heads had sufficient clearance to keep them floating behind the retainer plate. No trouble there either. Was it the material? There were no witness marks on the parts, and ejection seemed trouble-free at moderate speeds. I considered a strobe or high-speed camera to watch the action, but there seemed to be no way to adequately see the pins cycle.
The short-term solution was simply to slow the machine down, since the sleuthing consumed production time too. My troubleshooting mistake became clear a little later when the tool went back to the mold shop for a slight modification. After the mild rework, the mold behaved perfectly. It turned out that the bores for the worst offenders were short, very short, at what seemed like not much more that the diameter of the already slender pins. The telltale sign that should have given it away was serious bellmouthing, which I had ignored (“It came like that and the other shop says it made 1.2 million parts with no problem”), to my peril. Rebushed, with new drylubed pins, and the mold worked as advertised.
How could I have missed the simple trouble? Manually sliding those nails back and forth and wiggling side to side failed to produce the lash that I should have noticed because at the extremes of the pin’s motion, the head gave it extra lateral support, which isn’t a realistic operating condition.
Apparently even old pros miss simple things, but when you’re a rookie, they hide better.