Scientific molding slowly making progress in plastics processing

By: 
November 04, 2014


John Beaumont, President of Beaumont Technologies Inc. (Erie, PA), was recently elected to the Plastics Pioneers Association. His contributions to plastics processing are many, and he's dedicated to the science of molding. PlasticsToday's Clare Goldsberry sat down with Beaumont recently to discuss why the industry has been so slow to adopt scientific principles when polymer developments and machine technologies have advanced so far over the past two decades.

John Beaumont
John Beaumont, President,
Beaumont Technologies.

"To understand our industry—to understand where we are in our industry—we have to reflect on our history," said Beaumont. "The industry grew in an age when there wasn't much science connected to injection molding. This is despite the fact that injection molding is one of the most complex part formation processes on the planet. We start by injecting a hot, molten, non-Newtonian fluid, often consisting of blends and including numerous additives, into a cold mold where it is simultaneously flowing and freezing. The complex interaction of the polymer, mold design, part design, and process all combine to control the mechanical properties, shrinkage, residual stresses, and warpage of the final molded part.

"The final part properties are virtually impossible to predict and are not known until we have built a mold and molded the part. The science applied to plastic materials grew up separately from the industry, which has resulted in the injection molding process and tooling practices evolving as an art form. Molders are typically more like creative firefighters, not scientists."

The injection molding process and tooling design practices have developed over the decades through trial and error, with much of the success in getting conforming plastic parts dependent on the process technician's intuition, expertise, stubbornness, creativity, and years of experience. It involved more "knob turning"—constantly adjusting temperatures, pressures, fill speeds, etc.—until it was "just right," kind of like Goldilocks' porridge. "Without a clear scientific foundation, the approach to problem solving is typically symptomatic and superficial rather than root cause," said Beaumont.

Interestingly, Beaumont doesn't even recall ever hearing the word "rheology" while he was studying plastics in college. "We have this bootstrap industry evolving, devoid of real science," he commented. "The result is much of current practices are art-based techniques passed down from father to son. Oh, there has been science evolving in academia and larger polymer development companies; however, this is mostly focused on the development of new polymers and additives rather than on how to convert them into a molded product."

So how has Beaumont and his company managed to change all of that? He was working in the plastics industry in the mid1980s at Moldflow when he became heavily involved with injection molding simulation, which, at the time, was an emerging new technology. "To develop and apply this new technology, we needed a deeper understanding of what was happening during molding in order to know what to model," he explained. "Additionally we needed to develop an understanding as to how to use the simulation results. There was a lot of good brainstorming and head scratching going on."

In 1989 he left Moldflow to help start up the plastics program at Penn State Erie, where he brought with him a deeper understanding of material behavior. "We were in the early stages of how polymers behave, and just beginning to understand how to develop computer models and how to apply this new information. Coming out of the molding industry, I began to set the groundwork for where my head was and started the Plastics CAE Center at Penn State."

There were many different simulation companies emerging and competing with each other. The university provided Beaumont with a neutral environment where he could work with the various programs without commercial pressure. While at the Plastics CAE Center, Beaumont ran into a phenomenon: the simulation programs weren't predicting mold filling imbalances that were resulting from a shear-induced melt variation. Beaumont explained that this discovery challenged one of industry's most established art forms: "the naturally balanced runner."

"If variation and imbalances did occur, it couldn't be due to something called a naturally balanced runner," he said of the conventional wisdom of that time. "But a deeper understanding of the rheology of the runner showed that it could be. It was just that none of the simulation programs could pick up this phenomenon. The understanding wasn't there to model it. Having unveiled a major flaw in industry's state-of-the-art practice of the day, I then developed a solution, which was then patented. I bought the patent and started Beaumont Technologies."

Beaumont's beginning came with challenging the most tightly held "art-based" beliefs about plastic injection molding and mold design: Runners had been treated by the industry as simple hydraulic channels. "The entire industry missed this phenomenon," he stated. "If the runners are balanced, you shouldn't have to repeatedly rework every new multicavity mold to achieve a balance. When people observed the imbalance, they always had other explanations—but it could not be the runner. If they were seeing a variation, it was attributed to things such as cooling, venting, and mold deflection. Those are all wrong. But it was the art talking; everyone speculating without applying science. I couldn't even rely on simulation because of this phenomenon."

Beaumont stated flatly that "shear-induced imbalance is fundamental and is founded in rheological science. We typically see a 30% variation and this occurs in every mold that's ever been built. But people don't see it. The average Joe is still using the naturally balanced runner method."

Another example of the lack of science in the industry that was being taught across the country was relative viscosity versus relative shear rate. "The conventional methodology was to inject the mold from very slow to very fast and create a curve," Beaumont explained. "Then you pick a point on the curve, and that is the optimum processing point. This was taught by everyone. We looked at it and said, ‘this doesn't make sense.' As we looked deeper, we started playing with the curve. We found we could duplicate every curve ever developed on a molding machine without even running the machine. If we could duplicate the curve without running a molding machine, how could the method have any actual value? But this is the standard method for setting up a process—based on nothing."

That's the how the industry evolved, noted Beaumont. "We keep doing the same things over and over and expecting different results. Very seldom do we get to the root cause of a problem," he said. "Here at Beaumont we've put a significant amount of our income back into R&D trying to break the continuation of doing the same things over and over again without getting effective results."

Beaumont's latest work is in characterizing the injection moldability of plastic materials. "Essentially this maps how a material will mold through a wide range of wall thicknesses, fill rates, and temperatures," he explained. "This is in contrast to a melt flow rate (MFR) test, which is an isothermal test with one flow rate and one die geometry—nothing like injection molding."

Beaumont is a strong advocate for simulation; however, he believes that what it can do is oversimplified. "The idea that a simulation program will automatically give you all the answers and take the thought process out of designing a new product, tool, or process is absurd. It's a tool to use, much like a paint brush is to an artist," he stated emphatically. "We use Autodesk Moldflow and we can get good results but it takes understanding the strengths and weaknesses of the programs and how to deal with their weaknesses. Don't take the results blindly—use your experience and expertise not only with programs but with plastics, mold design, part design, and the injection molding process. Every analyst should be a molder in order to make practical decisions."

Beaumont's mantra for bringing the science into injection molding is, "Question everything," because everything must be verified by scientific means. This applies not only to today's mold design and molding practices but also the use of simulation. Don't take what the program tells you as the absolute solution.

"Using simulation effectively is about learning to think beyond the colored pictures provided by the program," he said. "There's still a lot to learn in this industry. There's a lot of change going on and a lot to discover. It's almost a rebirth when you begin to see this."

Comments (0)

Please log in or to post comments.
  • Oldest First
  • Newest First
Loading Comments...