Sponsored By

In this recurring column, Glenn Beall of Glenn Beall Plastics Ltd. (Libertyville, IL) shares his special perspective on issues important to design engineers and the molding industry.Putting a design detail on a part drawing does not guarantee that detail will be on the molded part.

Glenn Beall

September 10, 2008

6 Min Read
By Design: Sharp corners = Major liability

In this recurring column, Glenn Beall of Glenn Beall Plastics Ltd. (Libertyville, IL) shares his special perspective on issues important to design engineers and the molding industry.

Putting a design detail on a part drawing does not guarantee that detail will be on the molded part.

I find it difficult to realize, but I have been investigating plastic part failures for more than 40 years. It is even more difficult to understand why the mistakes that were common in the 1960s are still being made 40 years later. Today we know more about part design than we did back then. We now have more design-related books, brochures, and websites than anyone has time enough to access. For a lot of reasons, universities and the plastics industry have failed in impressing that information on the people who are in a position to benefit from it.

I never promoted myself as a failure analysis expert. In spite of that I have spent a lot of time investigating plastic part failures. Reflecting on that work, some mistakes occur more often than others. Not providing a uniform wall thickness was, and still is, the most common part design mistake. This is followed closely by the failure to radius the corners of parts. These two design defects are important, as they affect moldability, cost, and function.

The importance of corner radii

The following story recounts how the lack of a small radius created a large problem. Corner radii are an important design consideration for injection molded parts. There is a thorough review of radii in the December 1999 issue of IMM (immnet.com/articles?article=1074). The radius on corners has two primary functions. It improves the flow of the melt around a corner, which allows a cavity to be filled at lower injection pressure. This, in turn, results in less molded-in stress throughout the part. Corner radii also make molded parts stronger by distributing any stress on the corner over a broader area on the part. In other words, there is more plastic material to absorb the load.

In the northern part of this country, spring is an enjoyable time of year. The birds return and the flowers bloom. On the negative side, giant machines come out of hibernation to tear up the highways in the annual ritual of road repairs. A tiny, but vital, part of one of those ?machines was a cooling fan on a gasoline-engine-powered portable air compressor. This 10.5-inch-diameter, multipurpose, injection molded fan is pictured above.

This fan was mounted on a driveshaft via a slotted hollow boss and a metal clamp. The tolerances on the driveshaft and the hole in the fan did not always provide a snug fit. With a loose fit, the fan vibrated. This vibration was not objectionable, as it was at too high of a frequency for the human ear to hear.

Investigating the failures

My first encounter with this fan was as a consultant. I was hired to investigate the failure of a fan that allegedly resulted in the overheating and seizing up of an air compressor. I had an opportunity to briefly examine the failed fan. The blades and the ring to which they were attached were mostly intact, but the flange between that ring and the boss had broken into small pieces, some of which were missing. The fan was too broken up to allow a meaningful analysis. I did, however, note that the slotted boss was broken into two parts. Neither of these two pieces was attached to the flange.

The plastic material was nylon. This was fortunate, as nylon has a long and successful history as a fan material.

As near as I could tell, the fan had been properly molded. There was no splay, which indicated the material had been thoroughly dried. I found the gate, parting line, and ejection location, which were as to be expected. It was not possible to learn any more about the mold.

I left that inspection with no definite cause for the failure of the fan. While reporting to the customer, I had to admit that I didn’t know the cause of the failure. During the ensuing discussion I asked if there had been any other failures of this fan and there had been some. They produced three fans that had been returned.

Two of these fans were still together as one piece. On the third part, the split boss was intact, but it had broken out of the flange. None of the flanges on these parts had broken into small pieces. All three fans appeared to have been properly molded.

Examination of these fans revealed cracks emanating from the closed end of all six of the slots on the bosses. Based on these observations it was postulated that the cracks were caused by vibration-induced fatigue. Once the cracks progressed through the boss and far enough across the flange, the fan began to wobble. This wobbling caused the fan to break apart. No better theory has come to light.

I was puzzled by the failures starting in the slotted bosses, which were slightly thicker and therefore stronger than the flange. The logical place for a failure would have been at the junction of the flange and the boss.

Discovering the problem

I studied the failed bosses and discovered that there were sharp corners at the closed ends of the slots. All six of the cracks on these three fans emanated from those sharp corners. Once again, the failure to provide radii at corners had resulted in part failures.

A second look revealed there were sharp corners all over the fans. As politely as I could, I told the client its fans were not designed to the state of the art. The failure to provide radii at the closed end of the slots had resulted in the fans failing.

One of the client’s engineers, who attended the wrap-up meeting, was offended by my blunt accusations. He produced a fan part drawing that clearly specified a full radius at the closed end of the slots.

The only conclusion to be reached was that the toolmaker hadn’t put these radii on the mold. The toolmaker is clearly to blame for the subsequent failures. However, the laws in most states are that if the customer pays for a mold, that indicates approval of the mold and the toolmaker has no further responsibility.

My client did not inspect the molded fans to make certain that they were in accordance with the part drawing. He did not demand that the toolmaker correct the situation. As a result, he must share responsibility for the failure of its fans.

The closed ends of the slots were radiused. There have been no additional failures of the fans. The sad part of this story is that there were more than 80,000 of these fans produced with sharp corners on the closed ends of the slots. How many additional product failures or personal injuries can my client expect in the future?

Putting a design detail on a drawing or database does not guarantee that the detail will be on the molded part. There is no substitute for inspecting and testing molded parts from a new mold. This is especially true for molds being built abroad.

Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like