The Roller Skate (Part 1): Handling a Delicate Situation With Data

By Glenn Beall

There’s no excuse for not doing your homework when using a new material—or process.

Being invited to call on a new OEM is always an exciting event. There is the nervousness of meeting a new client and wondering about satisfying their needs. These apprehensions are dispelled by the prospect of turning a profit on what one hopes will be an intellectually stimulating project.

My destination turned out to be a modern building of modest size. There were no material silos out back, but the employee lot was full of late-model cars. The grounds were neat and there was an opening in the visitor’s parking lot near the front door. The day was off to a good start.

Walter, the manager of new product engineering, turned out to be a cheerful fellow with a positive attitude. This was no surprise, as new product development people, at least the good ones, are usually optimistic. Walter and I were to discuss a problem they were having with a new polycarbonate part.

Walter’s office was large, with his desk at the end of a long table. This department obviously spent a lot of time in meetings. Seated at the table were seven people. Our one-on-one meeting had become a conference.

I was introduced but no one smiled or offered to shake hands. They weren’t pleased to have me there to help them and they didn’t want an outsider criticizing their actions and embarrassing them in front of the boss. I promptly forgot all the names, but one of them was from quality assurance and another was a buyer. The rest worked for Walter.

Figure 1. Polycarbonate roller skate undercarriage.

The problem part was the roller skate undercarriage shown in Figure 1. This part bolts to a skater’s shoe and provides a mounting for the wheels and the brake pad. The first problem was cracking of the boss surrounding the molded-in insert at the toe end of the part. The threaded insert was the mounting for the skate’s brake pad. This defect was holding back the introduction of a new line of lightweight skates. Defining the problem and responsibility It took only a glance to recognize the problem and how to eliminate it. Walter and his team had been working on this problem for longer than two months. It would be embarrassing to them to blurt out the answer to their problem after only a cursory look at a failed part. This was

Figure 2. Material shrinkage creates stress cracks around a molded-in metal insert.

that company’s first venture into the mysteries of injection molding. It is easy to understand how they made this memorable mistake. It is not so easy to understand how the supplier who was molding the part allowed this situation to continue for so long. Any experienced injection molder should have been able to solve this problem. Perhaps they were playing the game of “keep your mouth shut and only do what the customer asks for.”

The ability to incorporate inserts into a part is an important attribute of the injection molding process. Molded-in metal inserts allow injection molding to capture applications that could not be satisfied with plastic material alone. Molded-in inserts do, however, require special considerations. During cooling the plastic material shrinks. This shrinkage causes the material to grip core pins tightly. Core pins prevent the plastic from shrinking the usual amount. This creates stress in the material. When the part is ejected off of the core pins, the material shrinks a little more and this reduces the amount of molded-in stress. There will still be a small amount of stress around the circumference of an injection molded hole.

A plastic material doesn’t know the difference between a core pin and a metal insert. Shrinkage around the insert creates a strong mechanical fit between the insert and the plastic material. An insert, like a core pin, prevents the material from shrinking the normal amount. Unlike core pins, inserts remain in the molded part. This doesn’t allow the material to shrink following ejection, and that produces parts with a high level of molded-in hoop stress around the insert.

The wall thickness of bosses containing molded-in inserts must be strong enough to withstand these hoop stresses. If this is not the case, these hoop stresses will relieve themselves with the stress cracks as shown in Figure 2.

As a rule of thumb, the wall thickness of a boss must be equal to the radius of the insert. In this case, the insert had an outside diameter of .750 inch with a radius of .375 inch. The average wall thickness of this boss was only .175 inch, which is less than half of what it should have been.

The insert literature expresses caution
Polycarbonate is the second-largest-volume engineering thermoplastic material. It is strong with many uses, but it is not ideal for insert molding. One material manufacturer’s literature states that “molded-in inserts are not recommended for use with nonreinforced” PC. The literature also recommends removing all sharp corners on inserts, and if knurls are used, they should be minimized and kept away from the part’s edges.

All of these warnings originate from the fact that knurls produce a notch effect and PC is a notch-sensitive material. Referring to Figure 2, it can be seen that this insert has large, sharp, diamond-shaped knurls that start right at the edge of the molded part.

This part was doomed to failure from the beginning. The OEM had done everything wrong. The problem then became how to explain the situation without embarrassing them. Fortunately, I had brought with me a design and a molding manual for PC. I opened the design manual to the section on inserts and suggested that they might like to copy and review pages 30 and 31. In this manner they became aware of the futility of what they were trying to do without my voicing any criticism of their actions. Backed up by the material manufacturer’s recommendations, there were no arguments regarding the changes I proposed.

The thickness of the wall surrounding the insert should have been increased to .375 inch. The width of the undercarriage allowed the walls to be increased to only .270 inch.

A compromise solution
The OEM had purchased a large quantity of inserts from a screw machine company. These should have been replaced but they wanted to try using the current inventory before redesigning the insert. A local metal finisher barrel-tumbled the steel inserts with an abrasive grit. This eliminated the sharp edges on the knurls without damaging the threads. Ultrasonic cleaning removed any trace of lubricant from the screw machining operation (some of these cutting oils are known to stress-crack polycarbonate).

While all of this was happening, the molder built a fixture for heating the inserts. These hot inserts cooled and contracted at the same time that the plastic material was shrinking. Contraction of the insert reduced the amount of stress between the insert and the plastic material. The moral of this story is that anyone working with a new material or process should take the time to review the trade literature in order to avoid embarrassing, costly, time-consuming, and memorable mistakes.

These changes eliminated any further stress cracking around the insert. There were other problems but they are the subject for Part 2 of the roller skate undercarriage story. Just for fun, try anticipating those problems by studying Figure 1.

June, 2007 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.