Structural composites take to the water

Chesapeake, VA -When engineers at Volvo Penta set out to create the company's first entirely new stern drive in more than a decade, they charted course that may be the engineering equivalent of Magellan's journey to the end of a supposedly flat earth. Rejecting forty years of conventional wisdom about marine drive construction, they steered clear of bolted aluminum and instead built the drive's structural components from adhesive-bonded composites. "Stern drives have always been made from die-cast aluminum and have barely changed since they first appeared in 1958," says Jeff Hardesty, the technical project leader who led the drive's design team. "But all of our external components, the parts that come in contact with the water, are composite."

Loose lips may sink ships, but it's salt-water corrosion and loose joints that can scuttle marine drives. With an inherent corrosion resistance and the elimination of bolted joints, adhesive- bonded composites target both problems. The new Xtreme drive, as it's called, also exhibits the more usual hallmarks of a smart materials substitution: It weighs 20% less and has 30% fewer parts than the company's aluminum stern drives, reports Christopher Savoie, one of the drive's design engineers. This new approach to drive design also resulted in an overall performance boost, a reduction in manufacturing costs, and new mechanical features.

The revamped stern drive sports two different types of structural composite components. The most visible one, a cowling that used to be made from die-cast aluminum, houses the propeller shaft and all the related drivetrain components. It also defines the hydrodynamic shape of the four-foot-long drive leg that dips into the water to power the boat. The less obvious part, a 20-inch gimbal ring once made from sandcast aluminum, fits in an opening in the back of the boat and supports the 100-lb drive leg. For both components, the design team picked ESC 8700, a 63% glass-filled vinyl ester from Quantum Composites Inc. (Midland, MI). The material resembles a traditional sheet-molding compound, but its chemistry has been modified to add toughness and to accommodate the high glass loadings, according to Mike Kiesel, Quantum's technical service manager. The drive retains a bit of aluminum only inside the drive leg. This one-piece aluminum "endoskeleton" performs a structural role and delineates the interior passages for exhaust, oil lines, cabling, the propeller shaft, and other power transmission components. This aluminum piece has been coated to protect it against exhaust gases and corrosion.

Hardly a hit-and-run engineering project, the Xtreme got its start five years ago with a challenge from an enlightened executive. "It was an engineer's dream," Hardesty recalls. "Our company president told us he wanted a new stern drive that wasn't just 'more of the same.'" Still, leaving behind the familiarity of aluminum for the unknowns of vinyl ester required a big leap of faith. "We hadn't even heard of the stuff when we started," Hardesty remembers. "We were aluminum experts thinking in the box of traditional marine transmissions." But Volvo Penta's engineers didn't make a blind leap. Their rethinking of stern drive included three full years of basic materials research, tooling development, and optimization of the compression molding process. "We spent many hours around the molding press," says Savoie.

Finding the strength. As much as pure part design counts, this application exhibited a particularly close relationship between molding conditions and strength, according to Hardesty. And these composite parts had to be strong to replace aluminum. Hardesty points out that these stern drives-which are commonly paired with 250 hp engines and propel boats up to 25 ft long-have to contend with heavy tensile, flexural, and impact loads. Thrust from the propellers, for example, applies a 2,023-lb (9-kN) load through the drive leg and gimbal ring, while sideloads from steering come to 1,000 ft-lb (1400 Nm). The biggest strength challenge of all comes from a test that gives new meaning to the phrase "shiver me timbers" as the company crashes the drives into floating log at speeds up to 40 mph. The resulting impact force exceeds 20,232 lb (90 kN), he says.

All of the team's hands-on experience helped avoid pitfalls that could have reduced the strength of its composite parts. At first, guided by composites orthodoxy, they molded in the large gimbal ring holes that serve as attachment points for the drive leg and the hydraulic trim cylinder. "When we started, we were told all we could do was mold the composite and trim the flash," says Matthew Wilkins, another of the stern drive's designers. Experience on the shop floor, though, taught the design team members to machine the holes instead. Hardly a trivial change, machining reduced the number of knit lines and improved the gimbal ring's strength. "For maximum strength, you really don't want to interrupt the resin flow when you don't have to," says Savoie.

To understand the close relationship between flow, fiber orientation, and strength, the team engaged in its share of old-school flow analysis. "We used the actual material for flow analysis," says Savoie. The engineers tested hundreds of molded samples for strength and then burned the resin out of them, leaving just the glass fibers. By studying this tangle of fibers, the design team identified which fiber orientations resulted in the strongest parts. They likewise optimized the resin injection by injecting different colored resins within the same molding cycle and observing which flow patterns produced the strongest parts.

And rather than just picking a material grade and hoping it would hold up, Volvo Penta engineers let their part design and composite choice evolve in lockstep. Even as prototypes started flying off the molding press, Quantum fine-tuned the materials with different catalysts and glass lengths. "At one point, we found that we needed a slower cure to reduce residual stresses in the part's thickest sections," says Savoie. For each new material variation, the design team had a testing regimen to determine the "real" mechanical properties. "They differed substantially from the properties you'd get from a test plaque or data sheet," explains Wilkins.

As for the initial design work, it took place without a hitch. Hardesty reports that the cowling and gimbal ring required only three major iterations, in part because the design team relied heavily on computer-based design and analysis tools from PTC (Needham, MA).

Using the composite materials did have one downside: lack of a track record in this kind application. As Wilkins puts it, "We know a lot about the fatigue life of aluminum, but we knew very little about composites and fatigue." The design team worked around this sketchy history with composites by subjecting the new stern drive to more aggressive life-cycle testing than comparable aluminum components. For example, though the measured loads from propeller thrust don't exceed 1,574 lb (7 kN), Volvo Penta conducted its fatigue tests with 2,922-lb (13 kN) loads. "We applied almost double the load that the drive would see in a typical working environment," says Hardesty. To make it even tougher on the composites, the company accelerated the fatigue tests, applying more loading cycles in a given time interval.

Adhesive issues. All these efforts to maximize the composite's strength would have come to nothing without the right adhesive to join the cowling's two main composite pieces, which come together like a clamshell to form a single component. Hardesty points out that all the impact from the log test falls first on the glued joint that forms the cowling's leading edge. To hold the cowling together under these conditions, the design team picked a two-part epoxy from Dexter Corp. The adhesive does double duty in that it also bonds the cowling pieces to the aluminum superstructure.

In addition to picking the right glue, lots of engineering effort also went into the joint design. "We tried many traditional joint designs," recalls Wilkins "But they weren't strong enough." After many cycles of destructive prototype testing, the design team came up with a keystone joint whose geometry, overlap, and clearances for the glue bead had been optimized to hold up to the log test.

The use of adhesive-bonded composites drastically simplified the drive housing. The bulk of the 30% reduction in parts count came from the elimination of fasteners, according to Savoie. To take the most important example, the composite construction does away with the bolted and gasketed horizontal joints that normally divide aluminum drives into upper and lower sections. Savoie notes that these bolted joints would have to come apart for gearbox servicing, leaving a potential leak path when the gearbox goes back together. "This joint is where the first seeds of corrosions are sown," he says. The composite cowling, which features an access panel for servicing, does not come apart at all, eliminating this leak path and making it easier to service the drive. "We consider the entire cowling to be a single part once it leaves manufacturing, " Savoie notes.

Better drive, less money. Corrosion resistance and reduced parts count alone might seem like good reasons to switch from aluminum to composites. But Volvo Penta stands to realize cost and performance benefits too.

The composite stern drive supports automated assembly, including robotic adhesive application. And its design could accommodates either one- or two-propeller drivetrains. By contrast, previous models needed dedicated gearboxes for each type of drivetrain, a more costly approach that essentially doubled the number of large stern drive parts the company had to produce and stock. Citing these two advantages, Hardesty makes a strong argument that composite materials, though more expensive than aluminum up front, will lower manufacturing costs substantially. "We'd be in trouble if we designed one that costs more to make," he quips.

Composites also fostered an important performance enhancement in the form of a tuned exhaust system. According to Wilkins, exhaust passageways in a single-piece diecasting run into limitations imposed by the reach tool's cores and slides. Molding and bonding composite pieces offered far more design freedom. "Composites allowed us to take a whole new approach with the exhaust system. We let the function dictate the exhaust system geometry, not the diecasting tool," he says.

While corrosion resistance may be the single most important gain from this materials subsitution, cost and performance enhancement amount to more than a bonus. Instead, they helped define the design team's starting point when they decided to improve on existing drive designs. "We never set out to make a composite stern drive," says Hardesty. "We set out to make a superior stern drive. "

For more information

Go to www.designnews.com/info or enter the number on the Reader Service Card:

Composites from Quantum Composites Inc.: Enter 545

Adhesives from Dexter Corp.: Enter 546