Better Molding By Design (and Experiment)

Before Suhas Kulkarni started his Design of Experiments (DOE) workshop, he talked about two different times he participated in such events but as a student.

Founder and president of injection molding consultancy, FimmTech, Carlsbad, Calif., Kulkarni was in Garden Grove Calif. at injection molder and moldmaker Comar LLC where he lead a half-day DOE workshop that took place immediately prior to Plastics Technology’s Molding 2018 Conference (Feb. 27-March 1; Long Beach, Calif.)

The instruction combined elements of DOE and Scientific Injection Molding—two topics Kulkarni said he wasn’t acquainted with until his second job after graduating with an M.S. in Plastics Engineering from UMass Lowell. At that job, two years removed from Lowell’s labs, he learned about both subjects in seminars lead by Hewlett-Packard and GE Plastics (now SABIC). Kulkarni noted that HP knew DOE while the former GE Plastics was more edifying on the topic scientific injection molding.

When he founded FimmTech in 2004 a central tenet of the business became the combination of DOE and Scientific Injection Molding, best embodied by the software he created—Nautilus—and his seminars. Kulkarni explained to participants of the event, which was split between a classroom lecture and time at the press, that variation in anything, but particularly injection molding, is 100% to be expected. “You can’t eliminate variation,” he said, before talking about three different kinds of consistency that injection molders should seek—cavity to cavity; shot to shot; and run to run, noting that the last consistency type might be the most elusive.

Why a DOE?

So if process variation is inevitable and it’s impossible to chase every process variable; where should a molder focus? Running a DOE, informed by scientific molding principles, can show you what is the most impactful variable to your process, Kulkarni said. Once a molder sees what levers they can pull to keep consistent, they can achieve what Kulkarni called Scientific Processing. To illustrate he offered an analogy: processing in cruise control.  If injection molding is the car, will it be easier to maintain a consistent speed along a cliff’s edge or on a 6-lane freeway? To get your process off the cliff and onto the Interstate: undertake a DOE and see what kind of process window you’ll be driving in.