By Tom Kirkland, tributek.biz
A recent SPE TECH CHAIN post mentioned a molding problem that began with a resin dryer which was not functioning properly. This resulted in charring of the material in a corner of the mold because of outgassing. The molding technician then turned down the barrel heat, which resulted in both excessive shear damage (not noticed) and short shots (noticed), so the pack pressure was increased. The end result was badly damaged material hidden in a part that otherwise looked good.
“Thankfully, we caught it,” the story concluded, “at the sonic [sic] welder.”
Anyone who has spent any time at all in the ultrasonic welder or ultrasonic tooling business will tell you this is a story often repeated. The ultrasonic welding process has historically been sensitive enough to throw a fit if the parts were not molded correctly, functioning in some sense as the canary in the manufacturing process coal mine.
This should alarm us for a few reasons. If, in fact, the ultrasonic welding process is this sensitive, can it be trusted to consistently produce good assemblies? What does this say about our ability to detect parts made by a non-optimal process if the parts do not go through an ultrasonic welder?
Most significantly, given that the ultrasonic welding process has itself recently become significantly more robust because of controls improvements and most especially servo welding, could substandard parts get through the ultrasonic welding process and into the field where their concealed weaknesses will ultimately be revealed?
Taking the most logical course through this maze of questions, the most important answer to all of this is that the ultrasonic welder should not be your canary at all. If there is a problem with molded parts, the answer is most definitely not to change settings until the problem appears to go away. A much more scientific approach to molding process control is needed, and documentation of any process changes – along with part weight and melt index testing after a running change – is the least we should expect from a well-run molding shop making parts with critical physical properties and/or critical functions. An increase in melt flow index signals a drop in average molecular weight and, with it, material toughness. An increase in part weight signals a possible increase in nominal wall, internal stresses and higher material cost per part.
While we are on the subject of materials testing, any critical molding process should begin with initial, final and periodic melt flow checks on the incoming material at the very least with each new lot of material and every quarter- to half-ton (or less) thereafter. You cannot make a silk purse from a sow’s ear, though oft has it been tried.
Second, a deep understanding of ultrasonic welding process control is vital to developing a welder setup that will be not only tolerant of part variations, but alert you to changing part conditions that may affect performance of the assembly in the field.
For instance, if energy, peak power, trigger distance or any other dependent variables in the welding process are trending or show a step when graphed over the length of a shift or a run, troubles could be hiding in those parts. If the graphs of power, distance, force, velocity, etc. during a cycle look much different on the last ten cycles than they did on the first ten cycles, an investigation should begin.
Every ultrasonic welding process is different, so there is no one-size-fits-all solution, but today’s best ultrasonic machines collect and deliver the information needed to spot problems not only periodically through the run, but to sound an alarm or even stop the process given an anomalous cycle. However, they do not do this on their own. They must be intelligently programmed for this to happen, and the data needs to be interpreted with understanding.
The process of manufacturing an ultrasonically welded part does not begin at the welder. It begins at the plant that manufactures monomers and weaves them into polymers; proceeds through pelletizing, drying, compounding and re-pelletizing; continues with more drying, storage, transportation, more storage, more drying, blending, coloring, plasticating, injecting, cooling, ejecting, more cooling, more packaging, more storage and – finally – welding. Each step in the process contributes to the final result.
That’s a lot to put on the shoulders of a canary.
For more information, email Tom Kirkland at email@example.com or go to www.tributek.biz.