Content area
Practical plastic car bodies came a step closer to reality when an automotive front-end section built from glass-fiber-reinforced polymer composites passed a key 35-mile-per-hour barrier crash test. The tests were conducted in early 1996 by the Automotive Composites Consortium (ACC), a research partnership established by the Big Three automakers - Chrysler, Ford, and General Motors - and their suppliers to integrate advanced composite materials into car structures. ACC engineers installed the experimental composite assembly in a steel Ford Escort, fitted it with sensors, and set the test vehicle crashing into a wall as high-speed cameras recorded the impact in detail. ACC researchers are working with several national laboratories in the SuperComputing Automotive Applications Partnership to develop a predictive computer tool that will aid in modeling crash performance with composites. The project will eventually result in a design tool for composite auto structures.
Full text
Cost-effective polymer composite structures that exhibit "crush control" in crash tests are being developed to compete with mass-produced steel car bodies.
PRACTICAL PLASTIC CAR bodies came a step closer to reality when an automotive front-end section built from glassfiber-reinforced polymer composites passed a key 35-mile-per-hour barrier crash test. The tests were conducted earlier this year by the Automotive Composites Consortium (ACC), a precompetitive research partnership established by the Big Three American automakers-Chrysler, Ford, and General Motors-and their suppliers to integrate advanced composite materials into car structures. ACC engineers installed the experimental composite assembly in a steel Ford Escort, fitted it with sensors, and sent the test vehicle crashing into a wall as highspeed cameras recorded the impact in detail.
"This was the first demonstration that a composite front-end structure, designed for mass-production manufacturing, could display outstanding energy-management performance," said Alan Taub, ACC board director and manager of the Materials Science Department at the Ford Research Laboratory in Dearborn, Mich. In the past, Taub said, handlaid-up composite front-end units had displayed this capability, but they were only technical feasibility demonstrations.
The crash tests, the culmination of the ACC's focal project 1, showed that composites can manage the energy of vehicle crashes as safely as steel, according to John Fillion, an ACC board director and manager of organic materials engineering at Chrysler Corp. in Auburn Hills, Mich. "There is no safety trade-off when you replace steel with a correctly designed composite part."
When a car crashes, Fillion explained, the goal is for the structure to fail in a relatively gradual, predictable way that absorbs much of the impact energy, keeping it away from the occupants. The trick in crash-energy management is to create what's called a controlled crush.
"Car designers know how to produce a controlled crush with steel because they've been using it for so long," Fillion said. But polymer composites are different. "Where a steel component plastically deforms rather slowly on impact, absorbing crash energy as it folds up, composite parts have to be specially designed to fracture in a manner that uses up the impact energy."
When an appropriately designed composite part-typically a stiff, hollow tube-is hit on the end, it tends to tear down its length in several places around the tube's perimeter in an effect called flowering. In flowering, fracture structures that look like flower petals form and spread out from the tube's central axis. This kind of tube cracking absorbs much more impact energy than does a clean break into a few pieces, as composite parts tend to do.
The successful crash-test result is the most significant accomplishment of the ACC, which was founded in 1988 to foster precompetitive research cooperation that would accelerate the development of low-cost, reliable composite car and truck structures suitable for large-scale production.
Cost-effective mass production of large, complex composite components will require the use of low-cost, high-reliability materials; new high-speed processing techniques; and new structural design approaches tailored for fiber-reinforced polymer materials. ACC engineers are focusing on liquid-molding techniques including resin transfer molding (RTM) and structural-reaction injection molding (SRIM). The effort's primary material systems are vinyl esters and polyurethanes, reinforced with inexpensive chopped-glass rovings. Automated glass-fiber preforming processes and high-rate molding procedures are being studied in an effort to reduce cycle times and production costs substantially.
Though current RTM and SRIM processes are somewhat similar, RTM generally involves a slower injection rate and slower polymer chemistry, as well as vinyl ester, polyester, or epoxy resins, according to Doug Denton, senior materials specialist at Chrysler. SRIM typically uses polyurethane resins, he said.
"We're taking a rather narrow view of composites, having decided to focus on liquid-molding techniques with glass-fiber reinforcement because press-molding processes such as sheet molding compound (SMC) and thermoplastic sheet processes were already being developed elsewhere," Fillion said.
In addition, composites researchers believe that liquidmolding processes hold greater potential for structural applications than the other, more-mature composite-forming technologies. "Liquid-molding processing-as compared to, say, SMC-allows you to put the glass- or carbon-fiber reinforcement exactly where you want it. This results in a truly optimized structure," Ford's Taub said. "While SMC is more amenable to lowercost production, it doesn't allow full optimization." To be tough and stiff enough for the task, the composite parts will have a glass-fiber content ranging from 40 to 50 percent by weight.
FOCAL PROJECTS
The ACC created four work groups to conduct generic materials evaluation, process development, crash-energy management studies, and joining technology research as well as two focal projects (FPs) to fabricate prototype composite structures. FP 1 resulted in the successful front-end section for crash testing; FP 2, which is still in its early stages, is concentrating on manufacturing-process development for producing an entire pickup-truck box from composites. Both components families have been designed to bear dynamic loads, contribute to crash-energy management, and provide a significant mass reduction (up to 30 percent).
In FP 1, which began in 1990, ACC engineers built a vehicle front-end (side-apron) section made of left and right top and bottom rails plus a shaped connecting panel. They chose to produce the front-end assembly from a vinyl ester resin reinforced with E-glass preforms in a resintransfer-molding process. The RTM fabrication process was conducted in epoxy-based prototype tooling from a previous proprietary Ford development program. The 42-pound component is about 50 inches long, and its height varies from 18 inches at the wheel to 30 inches at the engine. Its wall thickness ranges from 0.1 to 0.3 inch. Different structural reinforcements were used for the various parts; triaxially braided glass fiber over polyurethane foam cores make up the rails, while the connecting panel is built on two preform shells made from either directed glass fiber or thermoformed continuous-strand mat in a vinyl ester matrix. Directed fiber processes use pressurized air to deposit fibers onto perforated preform substrates (screens). Preforms of thermoformed mat are made by pressing masses of long glass fibers with binder additives under heat.
ACC researchers are working with several national laboratories in the SuperComputing Automotive Applications Partnership to develop a predictive computer tool that will aid in modeling crash performance with composites. This project will eventually result in a design tool for composite auto structures.
Although composite structural design can be used to consolidate parts and tailor fiber orientation to save cost and weight, composite parts still cost more than steel, which hinders their use in the auto industry. Past calculations indicate that the weight savings, lower capital investment, and reduced assembly requirements expected from the use of composites will not cut costs sufficiently to compensate for their higher production costs. The need is for more cost-effective parts-processing methods. What remains to be determined is how processing could be accomplished in the much shorter cycle timespreferably 1 minute per part "button to button"-that the auto industry expects. For the FP 2 effort, Fillion said, the goal was to learn how to process large, complex composite parts at a much faster rate.
A static crush test of a hollow composite rail element produces the desired flowering effect, a controlled crush fracture in which petallike pieces peel back along the tube's long axis.
The ACC was awarded a three-year, $6.9 million cooperative agreement to fund the processing demonstration technology from the National Institute of Standards and Technology's Advanced Technology Program. Additional money was made available through a cooperative agreement with the U.S. Department of Energy. Consortium members are matching the federal funding in a costsharing arrangement.
Choosing which part to build was not a simple exercise. Given that the project was a collaboration among competitors, each car company had its own agenda based on proprietary needs. The final choice was a pickup box, which has a relatively simple shape and is big enough for the purpose. "All three automakers had experience with prototype composite pickup boxes, so that's why it was chosen," Taub said. The box design includes a large rectangular container (open at the rear end), several support ribs underneath, and a tail gate.
"The pickup box is a bolt-on part, chosen because it requires no substantial investment in time to integrate the composite component into a steel structure," said Denton, who is also the FP 2 chairman. "This will be the first demonstration of the ability to manufacture such a large liquid-molded part at a sufficient rate for high-volume production [50,000 per year]."
To cut production costs, which is crucial to making structural composites commercially viable, faster processing speed is key, Fillion said. "We already have the weight savings, lower capital investment, and fewer assembly steps in the bag," he said. "We're hoping we can make the fabrication costcompetitive, so the cost of the entire process is a wash compared to a steel box."
A steel auto body costs about $650, which includes a material cost of 40 cents a pound, said Tom Moore, Chrysler's general manager for liberty and technical affairs, in a report to the ACC last fall. A comparable composite car body, using the lowest-cost composite (SMC, at 80 cents per pound plus greater processing time and labor costs), would be 24 percent lighter, but the total cost would be more than twice as high. "To be competitive, a composite material cost of 50 cents [per pound] should be the objective," Moore said. With current SRIM costs estimated to be $1.20 per pound, there is much room for improvement.
"We need to make a preform, transfer it into a mold, mix resin and catalyst, and infiltrate the resin without leaving dry glass and resin-rich areas," said Elio Eusebi, ACC board director and head of the General Motors Polymer Department in Detroit. "We have to time everything so we will fill the mold before the polymer's viscosity starts to rise." If it's too slow, the resin will cure prematurely, ruining the part. He said that special computer controls and optimized urethane chemistry are two ways to speed up the processing.
MOLDING CYCLE TIME
The current state-of-the-art cycle time for a typical liquidcomposite-molding process is 8 to 10 minutes, according to Fillion. "Our goal is a 4-minute cycle, which is a pretty aggressive but feasible target. The car makers would all really like a 1- or 2-minute cycle time, however," he said. "If we get it down to 2 minutes, you'll see a lot of composite application in vehicles. If we can cut cycle time to 1 minute, there'll be major shifts in the auto industry." "Molding cycle time is a function of the complexity and the size of the part, and the curing kinetics of the resin," Taub said. "Resin has to flow through the mold, so the intricacy and length of the flow path is key."
The ACC's molding-technology partner for the pickup box is Textron Automotive Co. in Americus, Ga. Engineers from MascoTech Inc. in Auburn Hills are managing the project details for the box processing. Magna International's Tacoma (Wash.) Division is the molder partner for the tailgate effort. Bayer Corp. in Auburn Hills is providing the resin technology. After studying the details of the liquid composite processing, ACC engineers decided that another way to achieve higher speed was to make the glass-fiber preforms more rapidly and more efficiently. To this end, they selected several preform fabrication processes with potential for high-speed, reliable quality, and low costs. The robotized or programmable powdered preform process (P-4) from Owens-Corning Fiberglas Corp. in Toledo, Ohio, will be used to build the pickup-box preforms. Tailgate reinforcement technologies under study include a slurry preforming process from the Budd Company Technical Center in Auburn Hills; the CompForm engineered fabrics process from American GFM Corp. in Chesapeake, Va.; and a rigid polyurethane core approach from American Sunroof Co. in Southgate, Mich. The smaller tailgate allows ACC engineers more flexibility in evaluating these preform processes. Fillion reported that preforming and molding machinery for P-4 is now being installed in a test production facility in Kettering, Ohio. The Edison Materials Technology Center is working with the ACC to establish test facility. The full-size demonstration equipment should be ready by the first quarter of 1997, and the group is targeting the first part production for September 1997, he said. A demonstration-size version of a refined P-4 processing unit has been developed, though the ACC has revealed little about the exact nature of the improvements.
PREFORM PROCESSES
Development of P-4 began in the mid-1980s, and it was introduced in 1993. "We recognized the need to cut the cost of making glass preforms by making them faster and more consistent, compared to continuous strand mat, for example," said Bill Mellian, transportation marketing specialist at Owens-Corning. "At this point, we've taken the technology as far as it can go on the lab scale."
"The P-4 process is a roboticized preforming system that allows greater design freedom by controlling the orientation and positioning of the fibers," said Glenn Sandgren, a leader at the Owens Corning Composites Innovation Laboratory at the company's Science and Technology Center in Granville, Ohio. "The process was designed to overcome drawbacks of previous processes such as slow cycle time, manual spray-up, difficult demolding, poor reinforcement placement consistency, waste, and energy consumption."
Owens-Corning's high-volume P-4 process uses a robot to spray inexpensive chopped-glass rovings and thermoplastic binders onto screens. The oriented fibers are then warm-pressed to shape.
The preformer, which was developed and first installed at the Owens-Corning facility in Battice, Belgium, consists of a robot with several glass-delivery systems, an air ducting and heating system, and a mold carrier with preforming screens. The glass-delivery systems send fiber through a nozzle that sprays the screen at specified locations and thicknesses. The delivery system can change glass length on the fly, Mellian said. One system can spray 3 kilograms of fiber per minute.
The process works as follows. First, a surface veil is applied to the screen. The veil is composed of thinner glass filaments that help produce a smooth class A surface appearance and keep binder from going through the screen. These fibers are sucked onto the screen and can be placed in any desired orientation. "With our system, 90 percent of the chopped fibers can be placed in any desired direction," Sandgren said. The powdered binder is applied at the same time as the fibers. The last part of the spray-up process is an additional coating of surface veil. When spraying is complete, the screen table is moved to the consolidation press where hot air is forced through the preform, melting the binder. Once cooled, the preform is demolded. According to Mellian, the resulting preform has good structural integrity, which makes it readily shippable.
"This technology gives users many options for design. It allows a wide range of glass or fiber contents and orientations to be achieved," Sandgren said. "One of the best features is that one can obtain preforms with net shape and virtually no waste."
The economics of P-4 are attractive, according to Sandgren. "P-4 has a lower raw material cost by virtue of a lower waste rate-2 to 5 percent versus a 12- to 50percent waste rate in thermoforming. The process also uses less binder because very little escapes from the preform." In addition, energy consumption, labor, and scrap rates are reduced.
"P-4 has a lot of potential to produce fiber preforms in high volume, with a high degree of control in fiber placement and lower variability in mechanical properties," said Chrysler's Denton. "It also allows you to customize the fiber to the part. In addition, P-4 uses a low-cost input material, has low wastage, doesn't use much energy, and is fairly environmentally friendly."
The Owens-Corning process requires a moderate capital investment, according to Mellian, including a $100,000 robot, a consolidation press, and an air-handling apparatus.
The capability to handle large production runs of more than 15,000 parts makes the initial investment less of an issue, he added.
According to ACC engineers, there is growing interest in using carbon-fiber reinforcement, which is widely used in the 80-miles-per-gallon Supercar designs being developed by the industry/government Partnership for a New Generation of Vehicles, a program associated with the ACC effort. "Because carbon fiber is electrically conductive, it would play havoc with the current P-4 equipment, so it would need to be modified to handle carbon fiber," Mellian said.
BUDD'S SLURRY PROCESS
For parts designed to be fabricated over randomly oriented fiber preforms, the Budd slurry process may be appropriate. According to Bruce Greve, senior staff engineer for Budd, the slurry preform process offers low cost, short cycle times, minimal scrap, and minimal preform loft (a term that describes relative density, which ensures permeability to resin).
The slurry process is based on a water-filled tank with a hydraulic cylinder mounted beneath it. The top of the cylinder is attached to a cradle that holds a full-diameter platen forming a seal with the tank wall. A perforated screen, shaped to the geometry of the final, is located at a cut-out in the platen.
The process cycle begins when the hydraulic cylinder is lowered to the bottom of the tank, which positions the preform screen there. Reinforcing fibers are added to the water together with some thermoplastic binder fiber. Uniform dispersion of the fibers and binder is produced through agitation. The cylinder then rapidly moves the screen up the through the water, capturing all the fibers.
As the screen is raised, a reduction in atmospheric pressure developed below it draws water quickly through the screen, speeding the entire process. The screen and wet fiber preform is then dried, and the binder is set by heated forced air. A perforated drying aid placed against the preform keeps the fiber in place until the binder sets.
Another high-volume fiber-preforming process under evaluation by the ACC is American GFM's CompForm technology. The CompForm process can fabricate complex liquid-molding preforms with anisotropic properties from almost any reinforcing material, said Dan Buckley, manager of CompForm R&D. The method uses high-speed ultrasonic cutting machines and ultraviolet-cured thermoset binders to rapidly fabricate fiber preforms from rolls of engineering fabric.
Buckley explained that a typical CompForm preform production system would include a three-axis computernumerical-control-type ultrasonic cutting machine that uses a blade vibrating at 20 kilohertz to cut roll goods at speeds as high as 2,400 inches per minute, a forming station, automated material-handling equipment, an ultraviolet curing system, and a fiveaxis ultrasonic cutting machine to trim the preforms to net shape prior to molding operations.
After the roll fabric is cut to shape, the binder is applied and irradiated with ultraviolet light, which typically solidifies it in several seconds, thus stabilizing and rigidizing the threedimensional shape of the preform. To save processing time, the ultraviolet rigidizing step can be accomplished while the tooling is moving from one station to another. The thermoset binder can also be used to bond separately fabricated preforms into a single unit, a procedure Buckley called "energetic stitching."
In one part of the ACC tailgate project, triaxial glass roving reinforcing fabric supplied by Brunswick Technologies Inc. in Brunswick, Maine, will be made into molding preforms. In the process, the formed triaxial fabric will be attached to foam cores prior to the liquid-molding step. Buckley said the CompForm process would also be used to attach the Budd slurry preforms together before molding.
ACC director Fillion noted that the consortium's first two focal projects had different goals-the front-end effort addressed crash energy management, while the pickupbox project looked at processing cost issues. "Though the details have not yet been determined, we'll probably roll the results of FP 1 and FP 2 into focal project 3," Fillion said. The outcome is expected to be an advanced design that integrates good crash-energy management and lowcost processing.
"The pickup box was built by design, not by accident," Fillion joked, adding that "the hardest design issue was ensuring predictability." Given the ACC's progress so far in addressing that issue and the issue of cost, it may not be too long before the engineers start to predict success in the quest for the composite car.
Copyright American Society of Mechanical Engineers Dec 1996