A Small Plastic Part Creates a Big Problem

By Glenn Beall

Related links: Plastics Strength

There are times when working fast results in big delays.

International trade has resulted in an increase in the number of companies competing for the same markets. As these multinational companies strive for market share, they have propagated the concept of low-cost, high-quality products. During the past decade the producers of durable products have made unprecedented improvements in quality and manufacturing efficiency. The cost and delivery of molds and injection molded parts have been reduced to levels that were considered impossible just a few years ago.

These improvements are impressive, but they still aren't good enough. America needs to further reduce its costs in order to compete with the rest of the world. Lean manufacturing, parts consolidation, design for moldability, replacing metal with plastic, and change from one plastic material to a lower-cost resin all are being vigorously pursued.

The large increases in material costs over the past few years have encouraged manufacturers to change from one material to lower-cost alternatives. It is not uncommon today to encounter a product that started out in polycarbonate but was changed to PPO and then to ABS and finally to a filled or reinforced polypropylene. This is a perfectly logical cost reduction technique, if the substituted material meets the marketplace requirements.

Generally speaking, the least-costly plastics have lower physical properties than more expensive materials. Plastic is not as strong as steel but polycarbonate is stronger than polypropylene. In all too many cases, the desire to reduce cost is so urgent that one material is substituted for another without taking the time or going to the expense of thoroughly testing the functionality of the substitute material.

A Metal-to-Plastic Conversion

This month's story is about a cost reduction conversion from metal to plastic. The change was a success, but not before a simple little plastic part created a big problem. The star in this story is an injection molded, glass-fiber-reinforced timing sprocket.

	In the left-hand part, sharp corners contributed to stress cracks in this copier machine sprocket, but changing the hole shape from square to more rounded was not possible because of the number of machines already in the field. Adding short, rounded extensions to the four corners as shown at right enabled sprocket to fit on the square drive shaft while removing the stress concentrators.

This project involved a large, commercial photocopying machine that was being upgraded to incorporate new duplicating technology. That redesign provided an opportunity to reduce manufacturing costs. The sprocket was a minor component, but its function was critical to the operation of the copier. Compared to the other larger and more complex components, the sprocket was too simple to command serious attention.

The sprocket was injection molded in an eight-cavity, three-plate, cold-runner, ejector-plate-type mold. The cavities were built using the metal sprocket part drawing. The new mold was sampled, the parts were approved, and production commenced.

After three or four months, depending on the frequency of use, the sprocket began to fail. The failures took the form of cracks that emanated from the corners on the square hole through the sprocket, and propagated to the root of the nearest gear tooth. At some time shortly thereafter a second crack developed in the opposite corner. At that point the gear fell off of the square drive shaft and the copier ceased to function. By the time the first failures were reported there were more than 1200 of these copiers in the market. That represented a lot of replacement sprockets and many expensive service calls.

Misdiagnosis

It was at this time that I was called into the project. I was told that the sprockets were failing due to a chemical attack by vaporized photocopy chemicals. I was asked to recommend a more chemically resistant material for the sprockets. This did not seem logical, as nylon is a chemically resistant material. Examination of the fracture surfaces under magnification did not reveal any of the telltale signs of chemical attack. The cause of the failures was probably something else.

Visual examination indicated that the four corners on the hole in the sprocket were sharp with no radius. I asked why these corners weren't radiused, and was told that (1) the metal sprocket had square corners, and (2) putting radiuses on the inside corners of the plastic sprocket would have incurred an extra cost in providing radiuses on the four corners of the steel drive shaft.

The second rule of plastic part design is to radius all corners. This is especially true of nonround holes with sharp corners. The hole in the sprocket was formed by four flat surfaces that defined a square hole. During the cooling and shrinking part of the molding cycle, each of these flat surfaces shrunk almost independently of the other three. Each flat surface attempted to shrink, but was prevented from doing so by the core pin. This attempted shrinkage caused the material to grip the core tightly, putting stress in the plastic material around the core pin. These molded-in stresses were the highest at the four sharp corners, which were stress concentrators. This is especially true for nylon, which is a notch-sensitive material.

The more I studied the broken sprockets, the more convinced I became that the failures were caused by the high stress on the corners of the holes. I concluded that getting rid of the sharp corners would eliminate the failures.

Radiusing the four corners on the holes was not an option. Any modified sprockets would have to be retrofitted to the square-cornered drive shafts on more than 1200 copiers already in the field. Radiuses and fillets are commonly used to eliminate sharp corners, but there are other ways.

An Alternate Solution

To solve the problem, short extensions with rounded contours were provided at the four corners of the hole. The sprocket still fit on a square drive shaft, but the stress-concentrating square corners were eliminated.

Sprockets from the modified mold were tested with an accelerated, rapid start-stop mode in the presence of concentrated photocopy chemicals. After just a few days of testing with no failures, the new sprockets were retrofitted to the copiers already in the field and used on all new machines. In the ensuing years there have been other plastic copier part failures, but there have been no problems with the sprockets.

The project manager who was responsible for the plastic copier parts was an experienced and knowledgeable plastics engineer. I had worked with and learned many things from him over several years. Long after this problem was resolved I asked him how he could have overlooked such a basic design defect. His only justification was that he was so busy with all of the other more demanding plastic parts that he just didn't think about the square corners on the sprockets.

With today's rush to be the first to market, it is difficult to find time enough to think about all of the little details that combine to make a good part. In this regard it helps to remember that you may think that you don't have time enough to do it right the first time, but you will have however long it takes to fix the problem when parts fail, the whole project is on hold, and everyone is staring at you.

February, 2008 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.