Avoiding Open Loop Molding on a Closed Loop System

Bernard Stuyvenberg has worked in the injection molding industry for 33 years. He remembers the days when open loop controls were the standard. An operator would choose an injection pressure and speed and let the machine run. This "blind" control could neither assess the process nor correct for error when viscosity changed. As resistance increased, the ram slowed; as resistance decreased, the ram sped up. There was no self-adjustment by the control to react to process conditions.

The advent of the closed loop control was supposed to fix all that. The molder still selects a desired injection speed and pressure, but now, when the ram meets increased or decreased resistance, the control adjusts the pumps and drives on the press to maintain the velocity and pressure.

Stuyvenberg is a consultant and the principal owner of Design and Manufacturing Systems. For much of his 33 years, he was a plant superintendent at Sur-Flo Plastics in Warren, MI. He compares injection molding control to driving a car. In the old days, with open loop systems, it was like driving with the gas pedal in one position. As long as the road is flat, velocity is constant. "As long as the accelerator is held stationary, the vehicle will slow down as it moves up a hill and speed up as it moves down," he says. Cruise control, on the other hand, is the closed loop system of driving. The desired velocity is chosen, and it's up to the car's engine to do whatever it takes to maintain that velocity--up and down hills.

Reality Bites
That, at least, is the theory. In practice, says Stuyvenberg, it's not often the case, and the advantages of using a closed loop system are unrealized. A control, he says, tracks four basic parameters: time, distance, speed, and pressure. Molders tell the control the distance the ram should travel to the cutoff position, the speed with which the ram should travel, and the pressure it should apply on hold.

In open loop molding, the molder knows the mold must be filled during injection. Thus, after choosing an injection speed, he will adjust the feed and then work with the cutoff position until the mold fills. He knows that if he moves the cutoff back a certain distance, the mold will short; move it forward a certain distance, the mold will flash. For example, consider a mold that at .8 inch begins to flash, and at 1 inch starts to short. The molder might believe, therefore, that a logical cutoff is .9 inch--halfway between the two.

The problem, Stuyvenberg contends, is that too many molders are using this open loop strategy to mold with closed loop controls. Because the molder is completely filling the part during injection, there is little margin for error. Any change in viscosity or other resistance during injection could cause the ram to undershoot or overshoot the cutoff, no matter how well-tuned the press is. "The controller is trying to maintain a given velocity regardless of the resistance, and the resistance is the greatest just before the mold begins to flash," he says. "The molder has set up the process so at the very moment cutoff is reached, a very slight inaccuracy in position, or a very slight change in polymer viscosity can result in shot-to-shot inconsistencies."

Again, Stuyvenberg says, there is a car analogy. Suppose you are in a car at the bottom of a hill, at the top of which sits a brick wall. You're to drive your car at 50 mph in cruise control toward the wall. At the last moment you're to slam on the brakes (cutoff) so that your bumper just touches the wall without damaging the wall or your car. If you're late hitting the brakes, your car smacks the wall (flash); hit the brake too soon and you come up short. Complicating this is the fact that you cannot look at the wall. You can use only your odometer as a guide as to when the brakes should be applied. This is, in effect, what happens to molders who attempt to use open loop strategies on a closed loop system. The ram is being told to move at a certain velocity, to constantly adjust to maintain that velocity, and then at precisely .9 inch, cutoff to pack and hold.

So, why do molders do this? "This is something that's just been handed down," Stuyvenberg says. "So much of what you see in the industry is guys out on the floor teaching each other." Like family traditions, the techniques learned using open loop controls are passed on to new employees, despite the fact that closed loop is now the most common form of process control. How closed loop controls should be used dawned on Stuyvenberg while he was a plant superintendent at Sur-Flo, where "if the problem couldn't be solved, it would ultimately fall on my shoulders." He says he realized then that molding is a science.

Coasting to the Wall
Now imagine, says Stuyvenberg, you're heading up the hill toward the wall again. But this time you apply the brake and gradually slow down so that by the time you get to the wall, you're going only a few miles per hour. "Obviously, if you are permitted to slow down first, it will be much easier to control the final impact," he says. "In fact, you could probably do it time and time again with 100 percent repeatability," This is what a closed loop control does best.

Stuyvenberg says his theories were confirmed when he attended his first class on decoupled molding, developed and presented by RJG Technologies (Traverse City, MI). On a simple level, the concept behind decoupled--or scientific--molding is simple: The part is not 100 percent filled during injection. Instead, the part is 90 to 95 percent filled during injection, and then fully packed out during pack and hold. This prevents the melt from smacking the tool and possibly flashing the mold. "Because the final filling of the mold is decoupled from the high-velocity boost, the pressure the mold feels is very much controlled," Stuyvenberg says. Molding is based purely on the material and the mold, not the machine.

Molding this way, based on observable facts and scientific data, is difficult for a lot of molders to swallow. It's a change some "traditional" molders, says Stuyvenberg, cannot or will not make. "Most of the time, once the guys understand it in a usable way, they begin to preach it," he says, "But some hear it, see it, and understand it, but just don't believe it." He also notes that younger engineers and operators seem more willing to embrace scientific molding. The efficiency of the U.S. molding industry may rely on molders' willingness to make this adaptation.