The linemen who work on high-voltage power lines may be crazy, but they are not stupid. They take their protective gear seriously, especially the leather and latex gloves that insulate them from the last shock they'll ever get. Before even looking at a live wire, linemen inspect these gloves carefully. They even blow up the latex liners like a balloon and listen for leaks that would indicate a puncture. "The gloves are all that stand between them and high voltage lines. Their life depends on the integrity of these gloves," says Bruce Bier, who runs electrical equipment supplier Richards Manufacturing Inc. (Irvington, NJ).
While latex gloves have protected many linemen over the years, they don't work nearly so well from an engineering standpoint: The hand-dipping process for making the gloves requires lots of labor and lots of harmful solvents. And yields for the hand-dipped gloves are as low as 70 to 80%, Bier estimates. These manufacturing difficulties have, in turn, led to supply snags. "There are only two domestic glove manufacturers left," says Larry Carmichael, program manager for power distribution at the Electric Power Research Institute, or "EPRI" (Palo Alto, CA). He reports that many utilities now have to wait a year to get new gloves, and not having the gloves on hand can be a big deal. "The linemen can't work 'hot' without them," says Carmichael. "And if they can't work 'hot', they can't fix power problems without shutting down distribution lines, which takes time and costs money."
In an effort to ease these supply problems, EPRI ten years ago launched a $3 million project to produce gloves using modern manufacturing methods. Injection molded thermoplastic elastomers held out the most promise in terms of yields, production volumes, and manufacturing cost. "We believe injection molding will halve the cost of the gloves, but a much larger benefit will be increased productivity for the utilities, " Carmichael notes.
On the downside, elastomer gloves represent a shockingly difficult thinwall molding application, with all the associated tooling hurdles. The gloves also have stringent mechanical and electrical requirements—including a test that tries to burn through them with a 30,000-volt jolt. Developing a moldable thermoplastic vulcanizate with the right balance of mechanical and electrical properties took years (For more information on the materials development process, visitwww.designnews.comand click on current issue). "When we told our elastomer suppliers what we wanted to mold, they told us we were crazy," says Mel Goldenberg, president of Polymer Technologies Inc. (Clifton, NJ), the contract manufacturer working with EPRI and Richards to develop the new glove. "It's one of the most difficult molding jobs we've been involved with." And PTI, with its roots in military, medical, and aerospace parts, doesn't usually focus on the easy stuff.
Hand molding. To see why this job has been so difficult, start by holding up your hand and forearm. Now imagine the tooling you'd need to injection mold a 0.055-inch-thick skin of thermoplastic skin over them. This unhappy combination of complex surfaces, thin walls, and long flow lengths is what PTI faced when it took over the glove manufacturing and tooling design four years ago. Goldenberg points out that the injection mold would have been much easier to make—and less costly too—if the glove could have been molded flat. In fact, a flat glove had initially been the intention, but the project switched gears to an anatomically correct glove with an opposing thumb and the slightly curled fingers of a relaxed hand. "Otherwise, the gloves would not have been comfortable for the linemen to use since they would always be working against the elastomer," Goldenberg says, noting that PTI and EPRI engineers designed the glove as a solid CAD model using anthropomorphic data.
The ergonomic shape of the glove may ultimately reduce hand fatigue, but it also complicates the tooling. "There isn't a flat surface in the entire mold cavity," says Goldenberg. What's more the mold is all thumbs when it comes to parts removal. "We were faced with the problem of getting the part clear of that thumb," Goldenberg says, adding that the glove also constricts near the wrist. One early iteration of the tool even had a removable thumb, one that detached from the core with the part and re-attached before the next shot. This approach wouldn't have been practical from a production standpoint.
The size of the part created its share of mechanical difficulties too. The 14-inch hand-and-forearm-shaped core weighs 350 lbs and extends horizontally into the mold, where it has to hold up to injection pressures in excess of 15,000 psi. These forces put substantial bending loads on the core, yet the core has to resist deflection in order to avoid holes or thin spots in the thinwalled glove. "We couldn't simply thicken the glove because that would have changed how they feel to linemen," Goldenberg adds.
PTI managed to solve all these tooling issues over the course of two years. To create the mold's contoured surfaces and complex shutoffs, they made extensive use of CAM. "You wouldn't be able to even make this mold cavity without CAM," Goldenberg says. As for the demolding difficulties, PTI engineers developed a proprietary surfactant that allows the glove to slide off the core—with a bit of help from the elastomer's natural stretch to get clear of the thumb and wrist.
Fighting core shift required more drastic measures. The tool has a system of hydraulic pins that physically support the core during injection. Actuated on a timer, these pins retract just before the flow front reaches them. The mold also makes use of an elaborate gating set-up with four valve gates that open and close in a carefully controlled sequence. Aside from its primary role controlling part filling and controlling weld lines, this sequential valve gating system also keeps the core from shifting. Goldenberg explains that valve gates at the far end of the core, near the fingers, open first so that the pressure from the molten plastic can help immobilize the core.
An invisible obstacle. Sometimes the worst molding problems are the ones you can't see, which turned out to be the case with the glove. According to Goldenberg, PTI engineers have spent the bulk of their time dealing with persistent problems with flow front convergences, or "weld lines," and entrapped gases in the mold cavity. Normally, weld lines don't have to be a big deal for parts, like the gloves, that lack cosmetic requirements. Anyway, sequential valve gating systems can usually go a long way in managing the convergence of flow fronts. Entrapped gases, meanwhile, can be eliminated by the expensive yet effective practice of adding vents in the mold cavity and running the tool under vacuum.
Yet even after adopting these tooling technologies and fine-tuning its molding conditions, PTI engineers found that the gloves still had a problem passing the voltage breakdown tests. This ASTM test involves submerging a fluid-filled glove in a bath of liquid, creating a potential differential, and seeing how much voltage the glove can take before it burns through. Goldenberg knew it wasn't a material problem, since flat samples of the material had no problem withstanding up to 34 kV. The electrical "weak spots" appeared only on the molded gloves—and then only in the regions with converging flow fronts. Goldenberg surmised that trapped gases in the molding cavity were creating micro-porosity where the flow fronts come together. "These holes are microscopic," he says. "You definitely cannot see them." Though tiny, the holes still diminished the overall electrical properties of the gloves. "The dielectric strength fell dramatically at the weld lines," Goldenberg says.
PTI engineers have spent the better part of two years trying to solve this problem, running many mold-filling simulations to identify the optimal gate and vent locations as well as to optimize the gate timing. At first, the simulations allowed them to make limited progress since they had access only to a "2½D" version of Moldflow simulation software. This software handily flagged larger pockets of trapped gases but did not show the small-scale entrapments. "They would show up only during dielectric testing," Goldenberg says. About a year ago, PTI switched to the Moldflow's current 3D simulation software. "That was a real eye opener," says Goldenberg, who says that the subtle entrapments not visible in 2½D suddenly appeared on the screen. PTI and ERPI are currently about 3 kV short of reaching the 30 kV mark for the molded glove, but the Moldflow simulations have helped identify some tooling, and molding changes should take care of the problem. For example, Goldenberg believes PTI can eliminate the remaining micro-porosity by adding some gates to the end of glove fingers, changing the part's flow leaders, and moving some vents. "The last bit is always the hardest," he says. "But we're almost there."
For more information on Santoprene from AES, enter 551 at www.designnews.com/info.