Sequential valve gating and thin walls team up to take notebooks below a millimeter

You know, four years ago I would have told you that you were crazy to do it this thin." Greg Tremblay, processing engineer at GE Plastics, is standing in front of a 550-ton Husky G-Line press at NPE, watching notebook computer covers, all .88 mm thick, roll out in crisp 19-second cycles. The notebook and the mold that makes it are the result of Tremblay's experimentation in an attempt to merge two hot technologies: thin-wall molding and sequential valve gating.

The pressure to mold thinner walls is pervading the industry, but in the computer notebook market the demand is especially keen. Notebook makers want reduced weight, larger computer screens, improved aesthetics, and-if you don't mind-no loss in structural integrity or performance. At the GE Polymer Processing Development Center in Pittsfield, MA, Tremblay's been experimenting with flow ratios and runners and gates and valves and weld lines in an effort to find the best way to quickly mold a thin, strong, attractive notebook cover.

Thin Walls in Notebooks
Wall thickness in the notebook industry is at about the 1.5-mm mark, in some instances dipping toward 1 mm. However, the true measure of wall thickness, says Tremblay, is the ratio of flow length to wall thickness, or L/T ratio. Most notebooks on the market today, he says, have L/T ratios of about 160:1 to 180:1. With a single gate, pushing melt into a 1-mm space over such a great distance requires melt pressures up to 35,000 psi, more than a typical press can provide; those usually fall into the 20,000-psi range.

Since most molders don't have 35,000-psi presses standing around the shop, there is a plan B. By going to multiple gates, the pressure requirement is reduced, often to the 20,000-psi ballpark. But, in exchange you get knit lines, which compromise the strength of the notebook and require painting or texturing to improve aesthetics. This costs money. "You either have to single gate with no knit lines and suffer with the high pressure," Tremblay says, "or go to multiple gates."

The Sequential Solution
To get around this little Catch-22, Tremblay and GE turned to one technology that eliminates knit lines: sequential valve gating. With the help of Kona, and later Husky, Tremblay built a mold to make a notebook cover with a wall thickness of just .88 mm. In the mold, he put five gates-one in the center, and one near each of the four corners.

He then ran a series of tests to compare the speed, performance, and efficiency of molding with one gate at high pressure vs. molding with five sequentially fired gates at a lower pressure. The material was GE's polycarbonate Lexan SP7602. For the single gate test, the cover required 35,000 psi of melt pressure and filled in .45 second, with a total cycle time of about 25 seconds. The part had no weld lines.

The second test, this one of multiple gates, had the center gate fire first. When the melt flow front reaches the four outer gates, they fire and finish filling the part. Like the single-gated part, this one has no weld lines, but required only 19,200 psi of melt pressure. Fill time slowed to 1.08 seconds, but the overall cycle time was reduced to 21 seconds. Because the gates broke up the length of the notebook, the L/T ratio, Tremblay says, was effectively reduced from 180:1 for the single gate to 120:1 with multiple gates. And because there are no knit lines, color can be provided by the resin, and texture is built into the mold. Finally, "your impact strength is higher than it would be if you had a weld line in there," he says.

Tremblay says that working out the timing of the sequential gates was a matter of trial and error. He says he relied heavily on timers to control valve gates. Fire too soon and you have weld lines; fire too late and you might as well go back to a single gate design.

At NPE, the notebook mold, unoptimized and in demonstration mode, was running at pretty high pressure on Husky's G-Line thin-wall press at about 26,000 psi, with a maximum pressure on the press of 29,380-leaving not much wiggle room. Cycles were averaging about 19 seconds; Tremblay says he thinks that number could be reduced to 12 or 14 seconds with improved cooling.

After all his testing, Tremblay came away with some recommendations for molders considering sequential thin-wall molding.

Requirements and Recommendations
Tremblay admits that sequential thin-wall molding can be tricky and more expensive. "It opens up a lot of things that go bump in the night," he says. Take, for instance, the mold. To preserve the aesthetics of the exterior of the notebook cover, the gating is reversed. This puts all of the guts of the mold-runners, pillars, ejector pins-in the same half, making tooling trickier and pricier. Tremblay says reversed gating increases tooling costs 50 percent or more. "You have to go through the manifold with all of your ejector sleeves," he says, "and that can get expensive." A molder would have to balance such additional costs against the money saved on painting and texturing.

Because of the relatively high pressures required for thin-wall molding, Tremblay recommends an H-13 mold steel, or P-20 at the very least. He also packed his mold with preloaded support pillars, large ejector pins, and extra stop buttons behind the ejector plate to prevent pin push back. The high pressures, and the heat created from shear, make cooling a bigger consideration than usual; Tremblay says to look for hot spots around the gates. With proper and effective cooling, he thinks cycle time can be greatly reduced. "Traditional cooling is not enough," he says.

Like many thin-wall applications, hot runners require special attention, particularly pressure drop, residence time, heat history, and shear history. Tremblay recommends that pressure drop not exceed 10 to 15 percent of the available machine injection pressure. To accommodate this, he says, melt channel diameters should be 25 to 40 percent larger when compared to traditional applications. To avoid resin degradation, Tremblay also says to keep residence time in the runners to one shot.

Two hot runner designs were provided for the notebook tests GE performed. Kona provided the hot runners for the initial development with hydraulic valve gating, while Husky provided the system to explore pneumatic valve gate design. Tremblay says the Kona gates used were 2.3 mm in diameter and exhibited less-than-average pressure drop. The Husky gates running at NPE were 4 mm in diameter.

On the press, Tremblay says that to reduce residence time, the barrel should not be too large for the application. He recommends that the shot size should not be less than 40 percent of the total barrel capacity, although for faster cycle times this might be reduced to 20 or 30 percent. The Husky press running the notebook at NPE had a 10-oz barrel 240 mm long with a shot of 57 mm-making the shot 24 percent of the barrel capacity.

Tremblay says he's now in the process of looking at other applications for sequential thin-wall molding. He's considering a series of tests molding network interface devices. Current wall thickness on these devices is .090 inch; he's trying to get it down to .060 inch. Next up for Tremblay is experimentation on parts with thin and thick sections. Although straight thin-wall molding does not handle thin-to-thick sections well, Tremblay expects sequential valve gating should make the process feasible. "The next step is going to be molding more complicated geometry," he says.