Insert Here

Ever try to attach a lid to a plastic box while it speeds down an assembly line? “Leaks”, “Cracked Housing” and “Distorted Part” are common, and will drive your QA personnel crazy. I once encountered such a problem with a small molded automotive part. The part was a polypropylene box, the open end of which was covered by a small metal plate fastened with #2 Type “B” sheet-metal screws. The box was designed with bosses at the corners, each with a small pilot hole. The assembly line attached the covers with pneumatic screwdrivers, each equipped with an expensive torque limiter. Naturally, nothing worked.

The best case failure was a crack, but the worst was a stripped thread, leaving an assembly that was cosmetically sound, but ready to shed a few screws anywhere between the inspection station and the loading dock.

The solution was small aluminum threaded inserts, which allowed fast line speeds (i.e. clumsy installation) without breaking the fragile resin parts. Getting the inserts into the parts, however, posed a problem. Pulling the mold for rework to take the inserts was a non-starter, because the tool would be replaced with a different design at the model changeover, and the customer wouldn’t spend the money anyway.

To sell the concept, a pilot run of a little more than one thousand pieces was demanded by QA. A temporary “second op” line was set up and two operators were equipped with modified soldering irons. Each station was equipped with a coffee can full of the Tic-Tac sized inserts and a shallow steel pan. Operators poured a small quantity of inserts into the pan and shook vigorously for a few seconds. About a quarter of the inserts came to rest the right side up, and when “poked” with the end of the hot iron, would cling to the tip long enough to allow the operator to insert the (now hot) insert into the bosses’ pilot holes until the insert’s shoulder bottomed. Repeated three more times, the part was done. The process required a little skill, but speed came quickly, and few parts were rejected.

In retrospect, we were lucky. Temperature control was non-existent, but the operators quickly learned just how much heat soak to allow for each insert on the iron tip, and how fast to plunge into the bosses. It was an amazing performance, and the set-up probably cost less than $100, including the coffee cans.

It was a non-starter from a production perspective, although not because of productivity issues. I figured (with my customary chisel tip marker on a cardboard gaylord) that a vibrational feeder sending correctly oriented inserts down the middle of the assembly table could feed six operators, easily matching the injection molding machine’s output. The problem was control and documentation. “Good hands” just wouldn’t wash in a capability study. The final configuration used a vacuum loader to position the boxes into a fixture, which used a “hot plate” to plunge all four inserts simultaneously. It was ugly, but it was controllable.

Occasionally I wonder, however, just what is the best way to get a metallic insert into a thermoplastic part? Ultrasonics? Hot nitrogen-gas jets? Friction-welding? For under a quarter of a million parts, I’d put my money on “good hands” and an arbor press.