We learn best from our own missteps, but that doesn’t mean we can’t learn from others’ as well. This article is the first in a recurring educational series on design goofs.
In my day plastics technology was not taught at the universities. Like a lot of other graduates, I secured my first job in the rapidly growing plastics industry. It quickly became apparent that we had not been taught to be plastics engineers. This deficiency was overcome by on-the-job training. Trade magazines, books, technical conferences, and seminars all helped. The best of our learning was by trial and error. Our teachers were the industry people who had that valuable wisdom that came with experience.
Fifty years have passed. Looking back, I conclude that some things have not changed. We know a great deal more about plastics technology today than we did in 1957. We have not, unfortunately, done a thorough job of disseminating that knowledge to the people who need to know it.
(Left) The redesigned valve body was used successfully and even ran in a faster cycle using less material. (Right) This valve violated the cardinal rule for designing an injection molded part: Maintain a uniform wall thickness.
Today there is a handful of good universities that are teaching plastics technology. In spite of this improvement, the majority of engineering graduates join the industry with no training in plastics technology. In retrospect, this is as it has always been. What is different today is that employers are no longer willing to provide these bright, young, enthusiastic people with on-the-job training. To make matters worse, many of the experienced industry people, who should be teaching the new graduates, have been downsized out of the industry by mergers, cost cutting, and reorganizations of one type or another. Regrettably, these novice engineers are left on their own to learn by the costly and time-consuming process of trial and error.
My long career in the plastics industry has provided me with the opportunity to make more than my share of memorable mistakes. Those mistakes provided me with an opportunity to learn what not to do. The thought occurs to me that the new people joining our industry can also learn from my mistakes and avoid living through the same costly, time-consuming, and embarrassing experience themselves.
In pursuit of that objective, this is the first in a series of little stories that recount errors in judgment, oversights, or just plain mistakes. I didn’t make all of these mistakes myself, but I was involved with these projects in one way or another. All of these stories are true; however, the names and some of the details will be altered to protect the guilty and discourage the ever-hovering attorneys. In retrospect, some of these mistakes seem trivial. They did, however, seem like major crises at the time and that is what makes them memorable mistakes.
The cardinal rule for the design of an injection molded part is to maintain a uniform wall thickness. Regrettably, this is the most frequently violated part design detail. Our first memorable mistake story involves a three-way valve body with a nonuniform thickness.
Investigating the problem
My first encounter with this project was in arbitrating a dispute between the original equipment manufacturer (OEM) and his custom injection molder. The problem was that the preproduction prototype functioned satisfactorily, but the parts—valves from the four-cavity production mold—leaked. The OEM accused the molder of not fully packing out the part. The molder claimed that the valve body was improperly designed. The before-and-after valve bodies are shown on this page.
The part in the top photo contained sink marks on the thick sections surrounding the small, horizontal, side-cored holes where they meet the large vertical center hole. The vertical hole also forms the seat for the mating valve stem. The sink marks in this location allowed leakage between the various valve ports.
This valve body was molded in nylon, which is a high-mold-shrinkage-factor, crystalline material. The average wall thickness was .148 inch, except in the thick sections around the small horizontal holes. In those locations, the thickness was .226 inch on the left, and .264 inch on the right side.
The size of the gate was increased and attempts were made to eliminate the sink marks by changing the molding conditions. The sink marks were reduced, but the assemblies still leaked.
The decision was then made to redesign the valve body to produce a more uniform wall thickness and to reduce the average wall thickness. The vertical wall between the horizontal ports and the mounting flange was eliminated at the same time.
The redesigned valve body, shown in the lower photo, had a nominal wall thickness of .089 inch, except for the mounting flange, which had a thickness of .113 inch. The areas around the two horizontal side holes were cored out to eliminate the thick section that had been present on the original valve body.
The redesigned valve body required extensive mold modifications. The core pins were reused, but the original cavities had to be replaced. The rebuilt mold was resampled without difficulty. The sample parts were tested and approved. The mold was released for production and the product went on to become a success.
It is interesting to note in passing that the redesigned valve body resulted in a reduction in part cost. The cycle time was reduced by 5.7 seconds and the amount of plastics required was 12.8g less.
During the course of this project it was revealed that the preproduction prototype valve bodies that did not leak were parts machined from nylon bar stock and did not contain sink marks. It was also learned that the machined valve bodies were fabricated using a sketch made by the OEM’s engineers for their own machine shop. As incredible as it may seem, the four-cavity production mold was built using that sketch that was intended for a machined part.
Who is at fault?
In my judgment, the OEM and the molder must share responsibility for this failure. The OEM should have provided the molder with a database or detailed engineering drawing of a valve body with a uniform wall thickness that was properly proportioned for the IM process.
The molder should have recognized that the valve body was not designed for molding. This was an experienced molder who should have known that the thick sections would create sink marks on the critical valve seats. That molder could have suggested redesigning the part before the mold was built and sampled.
About
Glenn Beall Beall owned Glenn Beall Engineering, a plastics product design and development company, from 1968 to 1993, and after retiring from that company, he established and is currently president of Glenn Beall Plastics, Ltd., a plastics consulting business. Learn
more about Glenn Beall
