This section was edited by Associate Editor Alan S. Brown.

MATERIALS AND ASSEMBLY

LONG FIBERS, LARGE STRUCTURES

AFTER SEVEN YEARS OF DEVELOPMENT, CANADA’S CAMOPLAST INC. HAS UNVEILED WHAT IT SAYS IS A LESS EXPENSIVE WAY TO MAKE LARGE COMPOSITE PARTS. The company has developed an automated, closed-mold technique to make long-fiber-reinforced structures. Camoplast’s engineering director, Yves Carbonneau, said the new technology cuts costs by 20 percent.

Camoplast said it is using the technique at Magog, Quebec, to make 12-foot by 3-foot reinforced polyurethane boat hulls. The hulls weigh upwards of 200 pounds, yet are 25 percent lighter than the polyester design they replace with no dropoff in strength.

The switch is a step change for long fiber polyurethane composites. Camoplast and two development partners, Bayer MaterialScience and KraussMaffei, began working on the project in 2002. According to Harry George, new applications development manager at Bayer MaterialScience, the largest long fiber-reinforced polyurethane parts that he was aware of at the time were 2- and 3-pound automotive door panels.

Originally, Camoplast made hulls from long fiber-reinforced polyester in a labor-intensive process, Carbonneau said. First, workers would apply a gel coat to a mold to give the hulls a glossy appearance. After a robot sprayed a glass-polyester layer, workers attached foam cores for ribs to strengthen the hull and applied metal inserts to secure the engine. Then the robot sprayed a second glass-polyester layer to lock the attachments in place and to increase strength.

Camoplast is using a newly developed Bayer resin system to make 200-pound boat hulls from long fiber-reinforced polyurethanes in closed molds. The technique, one of the largest long-fiber applications ever developed, slashes labor costs and reduces emissions, according to the company.

After the gel coat and two glass-polyester applications, workers would push rollers over the layer to remove voids and prevent the layers from pulling apart when processed. They heated each layer until it partly solidified, then heated the entire hull until it fully hardened. The operation took 60 workers.

“We knew we could lower costs and weight if we could figure out how to make very large parts in closed molds,” Carbonneau said. Closed molds had the potential to reduce processing steps and labor, but Camoplast had several hurdles to surmount.

First, there was the resin. Polyurethanes are lighter than polyester and expand when heated. In a closed mold, the expansion would create all the pressure needed to squeeze out voids or delaminations as the resin hardened. Unfortunately, polyurethane liquids turned solid within seconds of application, too fast to spray up an entire hull. Bayer developed a new resin with a 60-second work time, long enough for the glass-resin mixture to flow into all the mold’s tight spaces.

The second problem involved spraying speed. The robotic system chops the long fibers, then mixes them with polyurethane liquids before spraying. Long fibers—up to 2 inches—are valued because greater length provides more strength and stiffness, but they are also hard to mix without breaking. When Camoplast’s robotic spray system provider, KraussMaffei, began working on the problem, its highest output was 180 grams a second. It needed to reengineer the entire system to boost it to 300 grams a second.

The resulting closed-mold system is highly automated. It reduces the number of workers needed to make hulls to four, from 60. The hulls weigh 25 percent less than the polyester units they replace, which improves their acceleration and handling.

Equally important, polyurethanes contain only a negligible amount of styrene, a volatile organic compound, compared with 30 to 40 percent styrene in polyester, so there are fewer environmentally sensitive emissions.

Camoplast said it working on new applications for the process, but did not provide details.

ROBOT FILLS IN THE DONUT

SCARA ROBOTS ARE THE COMMON ROBOTIC ARMS USED TO MOVE AND ASSEMBLE PARTS IN FACTORIES. Yet SCARA robots share one big weakness: their work area resembles a donut, with nothing in the middle but dead space. Now Epson Robots says it has developed a new SCARA robot that can reach the middle of the donut. As a result, its new RS3 robot has the same workspace as a robot twice its size.

To appreciate how Epson did this, consider a conventional SCARA robot. SCARA stands for “selective compliance assembly robot arm.” In practice, it consists of a two-part jointed arm that moves easily back and forth along the x and y axes, but remains rigid in the z direction.

The upper arm contains a ball screw spline that lowers a tool or gripper to the workpiece. If that arm tries to spin all the way back towards its pedestal, the spline will collide with the base or the lower arm. As a result, it cannot get to the center of the workspace.

Even when a conventional robot hangs from the ceiling, it cannot reach into the center of its workspace without colliding with its lower arm.

The RS3 sets this on its head. First, it is designed to hang from the ceiling, so there is no pedestal in the workspace. What is the upper arm in a conventional installation becomes the lower arm in the ceiling-mounted configuration. “In the RS-3, the entire spline unit is encased beneath the lower arm so it can move freely in the space,” said Epson applications engineer John Yett.

As a result, the robot can move freely throughout the entire workspace. This enables an RS3 with a 350-millimeter arm to work on pallets as large as 494 millimeters by 494 millimeters, a size that currently requires a 700 millimeter robot. Smaller robots cost less than larger models, but the real payoff comes from more efficient use of factory floor space, Epson says. Moreover, the robot’s 360-degree workspace enables more flexible configurations along the line.

Epson’s new RS3 SCARA design has twice the workspace of conventional SCARA robots. It can reduce cycle times for some movement cycles by up to 25 percent.

Equally important, since the RS3 can move through previously dead space instead of going around it, it can reduce cycle times for some movement applications by 20 to 25 percent.

Like all SCARA robots, the RS3 provides more flexibility than linear gantry-mounted Cartesian robots. It is also simpler and less costly than more flexible multiaxis robots, and especially when it comes to control. The RS3 comes with Epson’s software and RC180 controller, which can be customized to support vision guidance, .Net, and Ethernet, Profibus, and DeviceNet communications.

The company sees potential applications in laboratory automation and other uses that require movement of large quantities of parts to process or testing stations.

RENEWABLE-SOURCE PLASTICS

For more than a decade, DuPont has sought to source more high-end engineering polymers from corn, castor beans, and other renewable resources. Now, the results of this effort have begun to reach the market.

Consider, for example, Salomon SAS’s latest freerider ski boots. Salomon makes the collar of its new Ghost boot from DuPont Hytrel RS, a polyurethane resin that contains molecules derived from both corn and petrochemicals.

Polyurethanes are thermosets, a class of resins that form when two or more chemicals react in the mold with one another, much like a two-part epoxy glue. After mixing, the liquid components react to form a solid material. While epoxies turn hard, polyurethanes remain pliant.

DuPont's Hytrel RS takes one of its chemical components from a renewable source, the starch in corn kernels. After cooking and grinding the kernels, workers feed them to enzymes that convert the starch into glucose sugar. The glucose then goes into a three-story vat of genetically engineered organisms that transform sugar into 1,3-propanediol (PDO), a chemical used to make polyurethane.

Hytrel RS contains from 20 to 60 percent by weight of this material, to which DuPont has given the brand name Bio-PDO. The rest of the resin comes from petrochemical-based materials. DuPont claims its Hytrel RS is similar to petroleum-based Hytrel in both properties and composition.

The Ghost ski boot collar uses 27 percent renewable-source material. The resulting polyurethane is tough enough to resist scuffs, and strong and flexible enough to lock a skier's foot into place while smoothly transferring motion from the leg to the ski at temperatures as low as minus 40 °F. Salomon tested the boot this past winter with skiers that it sponsors and will begin commercial sales next winter.

DuPont also makes renewably sourced Zytel 610 nylon resin. Nylon is a thermoplastic. Unlike thermosets, which form resins by mixing chemicals in a mold, thermoplastics are made into polymers inside a chemical reactor. Processors then melt the polymers and pour them into molds where they solidify as finished parts when they cool. Zytel 610 consists of 60 percent petrochemicals and 40 percent renewable castor beans.

DuPont developed the material with Denso Corp., which is using it on radiator end components. When production begins this spring, Zytel 610 will become the first DuPont renewably sourced plastic used under the hood of an automobile, where heat and chemicals degrade many materials.

In addition to corn and castor beans, DuPont makes renewable resins from soybeans, sugar cane, and wheat. The company's researchers are looking for ways to use plant waste—stalks, husks, and chaff—rather than edible starch and sugars to make chemicals.

SOUND LEVITATES SOLAR CELLS

IMAGINE PICKING UP GLASS WAFERS THAT ARE 300 MILLIMETERS IN DIAMETER BUT ONLY 0.6 TO 0.8 MILLIMETER THICK. One false step and the glass will chip or fracture.

That is the challenge facing companies that seek to automate solar cell processing. Solar cell wafers are made of silicon or other equally fragile ceramic compounds. While most range from 0.6 to 0.8 millimeter thick, some are as thin as 0.1 millimeter. A slight variation in pressure is enough to destroy a lot of valuable production in a very short amount of time.

The problem is even more acute for semiconductor electronics. Even in cleanrooms, robotic grippers can introduce particles that are large enough to short circuit the minute electronics on a finished wafer.

The answer, according to Germany’s Schunk GmbH & Co., is not to touch the wafer at all. Instead of using mechanical or vacuum grippers, Schunk levitates the wafer ultrasonically. The technique was developed by German startup Zimmermann & Schilp Handhabungstechnik GmbH.

Schunk has installed the levitators on a robot that moves wafers 200 millimeters in diameter from a depositing unit onto a linear transfer system. The wafers are only 0.16 millimeter thick.

Schunk has developed a gripper that uses a combination of vacuum and piezoelectric vibration to levitate and manipulate thin solar cells with risking fracture.

The ultrasonic gripper is based on a technique called “near-field levitation.” It rapidly cycles an ultrasonic generator—a flat, circular piezoelectric material—on and off. This compresses and decompresses the air, creating a cushion on which the workpiece can ride.

To transform this concept into a gripper that can actually pick up a wafer without contact, Schunk drilled holes in the circular face of the ultrasonic generator. Schunk draws a vacuum through the holes while the generator is vibrating. The suction pulls the wafer towards the ultrasonic generator, while the cushion of air created by the generator prevents contact.

There is no direct mechanical contact. According to Schunk, the device holds the workpiece securely between 0.05 and 0.5 millimeter from the ultrasonic gripper. It neither produces friction nor uses compressed air, two sources of contamination, so it is suitable for cleanroom conditions.

The ultrasonic gripper self-centers smaller components. Self-centering, combined with the solid grip, allows the robot to accelerate rapidly or rotate the gripper. Larger parts require lateral stops to keep the workpiece from slipping during acceleration, but Schunk says the ultrasonic gripper minimizes mechanical clamping.

This process is generally suitable for all reverberant metals, plastics, and ceramics. Schunk hopes to use it to build universal grippers and transfer tracks, and to stabilize workpieces with unstable shapes. The devices can be integrated into existing machines, and work with a variety of sensors.