Sponsored By

September 1, 2001

8 Min Read
WEB EXCLUSIVE:Scientific Molding, Part 2: Implementation

Editor's note: In August's issue (p. 77) we chronicled the machine-to-machine portability of molds when using scientific molding. This month Part 2 looks at how one molder learned of and adopted scientific molding. This series concludes with Part 3 next month. 


Figure 1. Odometer wheel SPC before optimization
Targeting Six Sigma quality levels, odometer wheels were consistently inconsistent. Flash, sinks, and other product flaws led to postmold errors.

It's enough to make you scream. There is trouble in the process and you are trying to shoot it. Bad parts are made. Bad parts are scrapped. Customer shipments are delayed. Money is lost. Where to begin? 

This was the hole in which injection molder Hobbs Corp. (Springfield, IL) found itself in early 2000. The problem was an odometer wheel that, postmold, went through a printing process that produced flawed wheels. Said wheels then had to be sorted, reworked, or scrapped. Sorting and reworking defective wheels necessitated costly overnight shipments. 

The Investigation 
Molding Manager Gary Chastain, Manufacturing Engineer Bob Markum, and Six Sigma Blackbelt Dan Olson were charged with determining the cause of the problem and devising a remedy. This team, called the Six Sigma group, evaluated the entire production process, interviewing operators, collecting data, and analyzing variables. Chastain and his colleagues determined quickly that molding issues, if any, couldn't be addressed until the printing process was stabilized. 

The problem at the printer was relatively apparent: After molding, the wheels are delivered from a vibratory feeder to a track that places the wheels in a nest. A pick and place unit clamps onto a wheel, over which runs a print head, applying the desired graphics. Finished wheels are ejected through a chute. The printer did not always produce properly printed wheels. 

Failure analysis of the printers revealed four main causes: Wheels jammed or did not enter the track properly; wheels did not orient properly; wheels could not be picked up; and operators did not know when or how long printers were down. 

Several printer improvements were made. Among them, alarms were installed, shock absorbers and stops were replaced, a vacuum regulator was added, track burrs were removed, the air/oil cylinder was replaced, and gauges were designed to speed printer setup and changeover. 






Fill time



















Table 1. Velocity setpoint vs. actual, before and after check ring replacement

Pointing the Finger 
With all of the printer improvements in place, problems with the molded wheels themselves became apparent. The list included flash, sinks, mismatched printing surfaces from mold to mold, blisters, splay, and silver streaks. After solving some materials issues with the purchase of a mixer, fine separator, moisture analyzer, and pyrometer, and after making minor changes to the tool, Chastain and Hobbs got to the heart of the matter: the molding process. 

Frustrated by an inability to stabilize the mold, Chastain enlisted the services of Mike Van Duine, a technical service rep for General Polymers and a scientific molding aficionado. After spending some time with the mold and press, what Van Duine and Chastain discovered answered just about every question Hobbs had about the flawed odometer wheels. 


Figure 2. Odometer wheel SPC after optimization
After fixing the mold temperature control, replacing the check ring, and optimizing the process using scientific molding, odometer wheel quality improved vastly.

To start, the entire process was pressure limited; this meant that, shot-to-shot, almost all of the pressure made available by the press was being consumed, leaving no spare pressure to accommodate changes in viscosity that naturally occur in molding. Exacerbating this was the discovery that the setpoint and actual velocity differed by 65 to 75 percent. The check ring was worn and the machine couldn't hold a cushion. Also, pressure loss tests showed unusually high resistance through the nozzle, which puts an atypical strain on the injection system. The coup de grace occurred when the mold cooling lines were disconnected and Chastain discovered no water flow through the mold. Statistics reflected the poor part quality (Figure 1). 

Van Duine introduced scientific molding concepts to Hobbs and Chastain, teaching the primary tenet: The molding process should be based on plastic material variables, not machine variables. This led to the establishment of a new, stable process that produces good parts. Pressures were dialed back, the check ring was replaced (see Table 1 for velocity setpoint), the mold temperature controller was repaired and cleaned, and a cycle that fills 95 percent on injection was established. Cycle time was reduced from 26 to 18.8 seconds—a 27 percent decrease. Six Sigma metrics showed that part quality and consistency improved dramatically (Figure 2). 

Sell, Sell, Sell 
This could have been the end of the story. In fact, there are likely many molders who have had a taste of scientific molding only to revert to old, comfortable methods. Reasons for such regression are numerous: lack of interest on the floor, no management buy-in of the process, inadequate training, and lack of an in-house champion. 

In Chastain, Hobbs was lucky to have the latter. Once he tasted the sanity, stability, and logic offered by scientific molding, he became the driver and is becoming an expert. He had to convince his managers that embracing scientific molding was the only way the firm could attain the process and part consistency desired. Today scientific molding is standard. "Scientific molding has proved we're headed in the right direction," Chastain says. "But we still have a long way to go." Long-term visions include networking RJG's plantwide monitoring equipment to track and contain parts based on real-time data. 

For molders sitting on the scientific molding fence, hesitant about a full buy-in, Chastain says he knows what's holding them back. "I can tell you why more molders aren't using scientific molding. They're not getting management buy-in to spend the money," he claims. "Scientific molding, to succeed in a company, needs a driver, someone who will yell and scream to get it done." Beyond that, Hobbs is backing up the commitment with solid implementation and an ROI strategy. This includes ongoing work with John Bozzelli, General Polymers, and the RJG team. 

To its credit, the Hobbs management team has had the foresight to realize the benefits scientific molding provides. It's a competitive advantage Hobbs intends to exploit. 


The history of scientific molding implementation at Hobbs is a classic one, as expressed in this chronology Chastain prepared for IMM. The story starts with a chronically troublesome mold that's installed in a new press: 

January 2000. A process technician comes in to set up a new machine [Arburg] and teach us new controls. Reduction in cycle time of 38 percent. Two weeks later we are back to same old parts and cycle time. We took notes on everything he did. Either took wrong notes or don't understand. 

February 2000. Arburg sends another process man, although different mold in machine. He establishes new process and improves cycle time. He suggests that RJG would be a good place to learn this type of processing. We lose process again, cannot understand. Management wants to see improvements. I notice that we are constantly turning knobs on the machines and seem to have a different setup. 

March 2000. I attend Arburg training and watch the instructor build a process on the chalkboard. This seems like a good sales pitch although it would not work in real life. After he finishes we install the mold and use the parameters he set with scientific molding. I can't believe it—good parts on the first shot. I wish I understood more. 

May 2000. Dorothy Lovelace [General Polymers] brings technical service rep Mike Van Duine in to meet us and promises that he can help solve our problems. 

June 2000. Need help, need improvement. Parts [odometer wheel, see story above] always have sink and dimensional stability is not there. Mike comes in and evaluates process. We are pressure limited and filling and packing part on first stage. How can we be pressure limited when the part is not filled? Mike says we cannot hold a cushion. Mike suggests there is a problem with the mold cooling. We remove the hoses; no water is coming out. 

Mike returns after we repair check ring and mold temperature control. We establish process again. Gate is sealing too soon. Also, actual velocity does not match setpoint. 

July 2000. Runner modified to improve balance. Mike returns again to establish a process. We finally mold a product with no sink that is dimensionally correct. 

We revisit the bell curve in our Six Sigma project and notice it is centered and much narrower, validating scientific molding. We monitor part weight variation and notice how quality parallels. We track process through July 10 and it remains stable. 

August 2000. Ran the numbers on cost savings. Saved 23 percent on cycle time. Scrap has been reduced and no one is turning knobs. Got management's attention. Had meeting with president, directors, upper level managers, and General Polymers to explain importance of buy-in and support needed to succeed with scientific molding. Buy-in includes training, equipment. I attend RJG training on decoupled (scientific) molding; immersed in process and methodology. 

September 2000. Gave Six Sigma presentation to upper management showing what we accomplished and the success of scientific molding. 

October 2000. Have now started training all setup employees using STK (share the knowledge) developed by GE Polymerland as a basic understanding of polymers. We will use RJG equipment and methods and General Polymers to teach actual processing. 

Someone told me the other day that we were traditional molders. I hope we can change that very quickly. 

Contact information
Hobbs Corp.
Springfield, IL
Gary Chastain
(217) 753-7672
[email protected]



Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like