It’s every manufacturer’s nightmare: the customer call complaining about a brand new product. “I just got this disc player, and there’s something rattling around inside! I think there’s a loose screw.” And, since our customers have to follow strict industry regulations: “I can’t install this product.”
You can imagine the concern on our end when we got this message years ago, just as we were starting out manufacturing operation. We wondered if a technician was not following procedure. Nope. Audits showed that procedure was followed well, even on the handful of units that later had this rattling problem. Was the customer mistaken? Nope. The units returned to us clearly had loose or unattached screws rattling inside.
That’s when we turned to Highly Accelerated Life Testing. We used HALT to vibrate our product, simulating lifetimes of use, and see what happened. What we saw was clear: after a few hours of vibe, certain screws started backing out. Even though they had been installed correctly, vibration would loosen them and eventually make them fall out, rattle around in the product, and scare our customers.
Now we had the answer. We just needed to design a solution. And that’s the second place HALT came in handy. If you’re in manufacturing, you know that there are dozens of ways to affect how tightly a screw is seated. Would our problem be solved with increased torque? washers? locking screws? glue?
Before HALT, we’d have to guess at the best solution, and wait to see if any products came back. But with HALT, we could try each solution and simultaneously stress them all. We set up tests for units where the screws used higher torque, lower torque, different profiles, various washers, and a variety of epoxies. Through HALT we observed which failed first, and which would keep screws seated tight even under dramatic stress.
That’s how we found that ChemPatch screws were the best solution for our product. And it’s why we use HALT all the time, now, to diagnose our problems and evaluate possible solutions. (Of course, the story has a happy ending: since we diagnosed the problem and implemented the solution, our annual loose screw number has gone from an average of 5 all the way down to zero!)