Robert F. Dray is president of R. Dray Mfg., which supplies end caps, valves, and feed screws to injection molders. Dray has seen it all. His career spans 37 years in extrusion and eight in injection molding, he has started companies and has helped others get started, and he has invented the most widely used barrier screw design in extrusion. Dray has seen molding machines evolve into the computer-controlled wonders they are today with closed-loop control and SPC helping to keep quality under better control than ever. Still, he believes something is lacking in molding that the extrusion industry licked a long time ago: extrudate quality control.

Dray believes the injection unit is the foundation of the molding process, yet he says it has generally been overlooked, and the process technology involved in the injection unit has been disregarded. "The current screw design technology used is from the 1950s," he says. "In the 1950s, '60s, and '70s, the extrusion industry revolutionized process technology. The motivating force was the advent of measuring and monitoring equipment that could accurately describe product quality."

Though they share some processing similarities, extrusion and molding have their differences, acknowledges Dray. The major difference is extrusion is a continuous process and molding is not. Being continuous, extrudate quality, or melt quality, as it's referred to in injection molding, is easier to measure in extrusion. Improvements in extrusion process technology allow the end product to be continuously measured to within thousandths of an inch or more, so quality problems can be quickly identified and addressed.

However, extrudate quality in molding is usually considered only after the fact, like when obvious part-to-part discrepancies occur or when the part fails in the field. Damage to the quality of the melt often occurs because a molding machine has inadequacies and improper setups are used to compensate for its shortcomings. For example, to overcome a machine's inadequate injection pressure, a setup person may increase backpressure and barrel temperatures to fill out the mold. In doing so, the melt temperature achieved may be too high. The melt may be damaged either by increased shear or by the temperature.

So why hasn't injection molding kept up with extrusion when it comes to extrudate quality control? Dray contends molders have not required their molding machine OEMs to improve the process technology of the injection units on their machines because molders are generally unaware such an advanced level of process technology exists. Therefore, they're unaware of the benefits that could be derived from such improvements. Very few process technicians have made the transition from extrusion to injection molding, Dray says. "The ones that have, rarely have had the opportunity to work in a hands-on production molding facility."

L/D Lessons to be Learned
In a technical paper Dray has authored, "Screw Design In Injection Molding," he details the impact of screw design on extrudate quality and the need for injection molding machines to incorporate systems capable of accurately monitoring extrudate quality and screw performance, systems that have long been standards in the extrusion industry. "In injection molding, the measuring is more difficult, but not impossible," says Dray.

He begins with a comparison of screw L/D in extrusion and in injection molding. In extrusion, L/D is normally 30:1 or greater. In molding, a 20:1 L/D is the norm, and he adds, "The L/D in injection is further reduced as the screw reciprocates. The amount of effective screw length loss is in direct relation to shot size." Therefore, the greater the shot size, the greater the starvation of resin due to the fact the resin inlet is transferred downstream in relation to the first feed flight. Injection screw designs usually have additional turns of feed section to compensate for this starvation (see drawing).

Extrusion processors have realized a number of advantages from using screws with a longer L/D, Dray says, including the following:

In molding, the first two benefits translate into cycle time reductions. Increased rate reduces cycle time if recovery is a limiting factor. Lower melt temperatures reduce mold-close time requirements. Dray says if lower melt temperature causes short shots due to lack of adequate injection pressure or speed or if the mold opens during injection due to inadequate clamp tonnage, either the injection unit was not properly selected or the wrong clamp tonnage was selected.

"The intent is not to run the lowest melt temperature possible, but to run melt temperatures within the manufacturer's recommended specifications," Dray explains, adding his observation that, more often than not, melt temperatures in molding are set well above what is recommended.

The No-Purpose Screw
Dray says screws with a 20:1 L/D, a 40-plus year-old design, are never used in extrusion. Extrusion technicians call this design the "single-stage square pitch design." Molders call them general-purpose screws. Dray calls it the "no-purpose" screw design.

He argues that the idea that g-p designs are more forgiving and capable of running a wider range of resin viscosities is a common misunderstanding. A properly designed mixing or barrier screw has a far greater performance window, Dray says, because of its ability to disperse agglomerates (unmelted resin particles) that enter the metering section. More modern screw designs will provide proper mixing and color dispersion without the rate reductions associated with increasing backpressure.

Agglomerates can produce viscosity variations, molded-in unmelted particles, or short shots. Usually backpressure is applied in molding, but that decreases flow rate and, possibly, pressure stability while increasing melt temperatures. "Backpressure is commonly used and is always a poor substitute for proper screw design," says Dray. He also believes it is incorrect to assume the flow passages downstream of the screw can provide the additional shear energy necessary to complete the melting required to create uniform melt temperature.

The proliferation of mixing sections in injection molding only proves to Dray virtually any type of device placed downstream of the metering section will improve a no-purpose screw design. "This does not intimate all mixing devices are equal," he adds. "Nor does it say the upstream design does not have to be properly modified to enjoy the benefits of effectively designed mixing sections."

In his paper, Dray cites the key historical developments in screw designs, including his own 1970 patent for the barrier screw design that continues to be the most widely used design in extrusion (U.S. Patent #3,650,652). This design also has been used successfully in high-performance/low recovery time injection molding applications, primarily using screws with longer L/Ds.

The key is in the metering section. A screw designed with a longer metering section yields better rates with the same backpressure, and as backpressure is reduced, efficiencies further improve. "No-purpose designs in many cases are not able to run at lower backpressure due to inadequate color mixing or poor extrudate quality," Dray insists. The metering section's function is to develop the required pressure. If it is unable to do so, the pressure development requirements are moved upstream. When this occurs, melting also may be moved upstream, reducing the upstream pressure-developing capabilities of the screw and, thereby, reducing the rate of melting.

In conclusion, Dray says it is not necessary for the injection molding industry to reinvent the wheel. Rather, it would be better served by examining the innovations of the extrusion industry and adapting its field-proven process improvements. "The injection molding revolution in process technology will raise our industry to another level. If we are able to eliminate variables from the injection unit, provide proper extrudate quality, and eliminate the possibility of recovery time increasing cycle time, we can begin to realistically look at mold design to improve flow and eliminate product quality problems associated with poorly designed molds." For a copy of Dray's paper, refer to the contact information listed below.