Chrysler's Viper is driving innovation in RTM.

The Viper program shows how RTM can become a 'world-class' technology through strategic partnerships and front-end engineering in tooling and glass preform design.

"We've taken the fear out of RTM. We're through the learning curve."

Buck Hendrickson speaks for more than just the experience of Aero Detroit in Madison Heights, Mich., where he is v.p. of molded composite manufacturing. What his company has learned about resin transfer molding composite body panels for Chrysler Corp.'s Viper sports car teaches important lessons to North American industry at large. Even with only 3000 units planned this year, this high-profile application is helping build confidence in the RTM process.

"We're doing exactly what we said we'd do: build a low-volume niche car with the RTM process," says Nippani R. Rao, program manager for technology development at Chrysler in Detroit. "We said we'd bite the bullet and develop the RTM technology to make Viper a feasible concept for low-volume auto production. I think we've succeeded."

The Viper mandate was to push RTM well beyond its traditional image as a specialty craft, more art than science. RTM's allure, for automotive has been the potential for lower tooling costs in low-volume, niche vehicles. That potential already has been demonstrated in the U.K. by Lotus Engineering, which has made RTM body panels for its own low-volume sports car. The challenge for the Viper team was to demonstrate this composite process as a viable "world class" technology, capable of satisfying the rigorous quality and productivity demands of the North American auto motive industry.

WORLD-CLASS RTM

The Viper project teaches that the principles of world class manufacturing learned in injection molding or other processes apply equally to RTM (see PT, April '91, p. 84). Harnessing RTM requires the same degree of focus on front-end engineering to build tools and preforms. There must be strict measurement of statistical data in order to optimize the entire manufacturing process, rather than a narrower concern over tweaking the quality of any individual part. It also demands regular daily maintenance of mold-clamping systems to achieve automotive-quality parts.

In addition, the Viper experience helps identify those areas of RTM technology that require further clarification. The main concern today is the inability to accurately predict or control resin flow in large parts and complex-shaped structures, which results in pockets of trapped air pockets or dry glass. However, the Viper project afforded an opportunity to try out new computer-aided engineering (CAE) software that may help solve or prevent such problems in the future. This is one of several techniques to emerge that will help in the next generation of RTM technology.

The Viper program also affirms the promise of strategic partnerships between custom molders, end users, and material Jan. '92, p. 51). Besides Chrysler and Aero Detroit, partners in the Viper program include Rockwell International Corp., Troy, Mich.; Nero Plastics Inc., Owosso, Mich.; ETM Enterprises Inc., Grand Ledge, Mich.; materials suppliers ICI Acrylics, St. Louis, and Ashland Chemical Co., Columbus, Ohio; fiberglass mat producer Vetrotex CertainTeed Corp., Valley Forge, Pa.; and paint supplier Akzo Coatings Inc., Pontiac, Mich. Conversations with all parties in this project indicate that despite some occasional conflicts, the good-faith, strategic partnership mechanism has held up well over the past three years.

TOOLING EVOLUTION

Tooling was one of the most critical areas in the Viper program, requiring significant investment in front-end engineering and design. "Soft" epoxy tooling was used during the first two years of the car's development in order to accommodate the multitude of design changes. Epoxy tools also fit the total budget ($50 million) Chrysler dedicated to develop Viper. "Early on, the epoxy tools gave us the flexibility to make design changes," recalls Rao. "But we always knew we would eventually go to hard tooling for production."

By late 1992, Aero Detroit began utilizing nickel-shell tooling for body panels, which were built by PPD Recherche Inc., Waterville, Quebec. Aero Detroit uses a vapor-deposited nickel shell on the mold for a part's "A" or appearance surface; and kirksite, cast aluminum, or billet-cut aluminum for the "B"-surface tool. The molds are water heated to an average operating temperature of 150 F. Hendrickson says all the tools utilize edge gating with a proprietary sprue design. Rao says one major surprise in the development of hard tooling was the ability to attain a surface finish that was twice as good as expected.

The ability to accurately and consistently control resin flow in the mold remains a major challenge confronting RTM processors. John Moore, Aero Detroit research and development manager, recalls a tooling problem involving an air pocket trapped in the center of the deck lid. A computer mold-filling analysis (using the C-Set simulation module for reactive molding in the C-Mold product line from AC Technology, Ithaca, N.Y.) confirmed the resin was "racetracking" around the perimeter of the tool, causing air to be trapped in the center. Initially, the condition required an excessive amount of resin to be vented in order to ensure the elimination of the dry spot. Eventually the problem was managed through changes in tool venting, and a part/tool redesign may be considered at a future date. (One more potential solution is channeling resin flow by creating "dams" in the glass pre-form to slow the "racetracking.")

Thomas F. Lewis, director of the Viper program for Rockwell International, says another key lesson learned from Viper was that the flexibility and relatively low cost associated with RTM tooling is "no excuse" to ignore the importance of front-end engineering work. "The traditional notion of RTM is that tooling changes are cheap and easy to make," he says. "But you still must live with the parts you design and make. The thought was we always could make design changes on the fly with RTM. It turned out not to be that easy."

Lewis points out that although the body panels were improved with every design change from Chrysler, each change generated a new set of processing variables. The predicament of changing multiple epoxy molds while supporting a commercial car assembly operation did cause some concern among the molders, he admits. "Next time, I would push harder to freeze designs earlier in the program and cut the hard tools sooner."

Nero Plastics, serving as a second-tier supplier for Rockwell, has reached the fifth generation of epoxy tooling for the massive, one-piece outer hood and fenders, which has a surface area of about 40 sq ft. Nero makes its own molds for molding the hood/fender outer panel. (The RTM inner hood panel is molded by Tee Jay Industries Inc., Fraser, Mich.)

Bradley Faber, president of Nero, says his firm developed a proprietary manual clamp design, which provides 80 tons of pressure, for its fifth-generation tool. Nero handles the male/female mold halves manually, unlike Aero Detroit, which uses hydraulic presses.

The long-range strategy for processing the Viper hood always was to begin with RTM, then eventually go into SMC, according to Rao and also to Rockwell executives. This plan is on schedule, with SMC outer hood-panel production set to begin this month at Rockwell's Centralia, Ill., facility. The SMC tooling is being provided by Autodie International, Grand Rapids, Mich. Raymond M. Haak, engineering consultant for Rockwell, says the inner hood panel is likely to remain an RTM part.

FOCUS ON GLASS PREFORMS

Glass preforming represents the second critical area for front-end engineering work in the Viper RTM pro gram. Both Aero Detroit and Nero have designed and built their own proprietary systems for cutting, clamping and preforming continuous-strand glass mats. Assisting them in this effort was Vetrotex CertainTeed, which supplies the Unifilo 750 formable E-glass mat and technical support. Nicofibers Co., Shawnee, Ohio, provides the A-glass "veil" for the preform, which helps minimize fiber read-through on the plastic part surface.

Bob Carvalho, CertainTeed applications support laboratory manager in Maumee, Ohio, says a cooperative effort with the molders helped develop correct methods for shaping and cutting the preform, as well as its placement in the mold. The major problem they had to overcome was the improper fit of preforms into sharp radii within the mold. This resulted in resin-rich areas of the part, which inevitably suffered from cosmetic surface cracking.

The solution involved slightly over-sizing the preform in the sharp-radius sections in order to compensate for the sharply angled mold sections. Carvalho says this approach increases the preform area so the glass extends into these fighter angles.

Also necessary for obtaining successful preforms were precision clamping of the glass mat and a specially designed clamp-frame system that allows for proper tension and release along the outer edges of the glass charge as it is pressed into the preform mold. Moore of Aero Detroit notes that a key to success of the overall RTM process is the "ability to uniformly distribute the glass preform throughout the cross section thickness of the laminate in all areas of the molded panel."

The total preform production cycle for body panels at Aero Detroit is about 4 min. Nero's hood/fender preform cycle is about 9 min and requires a two stage, quartz/ceramic electric zone heating oven.

Development of a successful, reliable preform process also is one of the major reasons for Aero Detroit's dramatic reduction in scrap rates: from over 20% in May 1993 to under 6% in December. During the same period, finished-parts quality (a measure of defective parts vs. parts shipped) fell from over 26% to 10%.

CONTROLLING THE PROCESS

Aero Detroit's capital investment in 1993 to ramp up production of Viper parts totaled $5 million. The shopping list included four new 450-ton hydraulic clamps from Eagle White Press Co., Windsor, Out., a seven-axis Ingersoll-Rand CNC waterjet cutter, a custom-built performer and oven, a Novel computer network to collect shop-floor production data, and assorted painting equipment. Aero Detroit collects statistical process control (SPC) data from its seven mold clamping presses. The monitoring and control system, supplied by JHM Technologies of Waterford, Mich., records pressure, temperature, and hold time, and also meters the injected resin components. The weight and dimensions of every molded part also are recorded and charted.

Average cycle time range for processing RTM body panels at Aero Detroit (not including the preforming cycle) is 14-17 min, which includes the time to open the clamp, manually insert the preform, close the clamp, inject resin, hold and cure, open the clamp, manually remove the part, loosen and remove flash with compressed-air guns, and place the part in a cooling fixture. Moore says the next target for cycle times will be under 6 min. Aero plans to introduce a mechanical system for ejecting parts, which will help reduce cycle times.

The company currently has a daily production run of 285 inner and outer body panels, which make up 15 assembly "sets" (19 parts per set). A "roving" process auditor on the shop floor inspects each part for weight, thickness, and dimensions. Aero says it maintains panel thickness within 0.25 mm (9.75 mils) and dimensional tolerances within 1.5 mm (58.5 mils) as specified by Chrysler.

Aero was able to convince Chrysler that a gel coat was no longer needed in the nickel-shell tools, thanks to a proprietary method for obtaining superior surface properties on RTM parts. It involves resin chemistry and subtle, precise differences in the temperatures of the mold halves. He says the A-surface mold must be slightly hotter than the B-surface tool in order for A to cure first and attain the required finish. The secret is knowing the exact temperature ratio between the mold halves and how to consistently maintain that delicate temperature balance.

Nero uses a gel coat on the hood/fender "A" surface with its all-epoxy tools. Nero has two tool sets in production, producing about 20 hoods a day. The firm operates two production shifts, and each molding station has a team of three workers, each trained to perform a specific function. The tools use dual injection ports, with equal pres sure applied independently (not manifolded) at each port. Workers record injection pressure and time and record the weight of vented resin for each cycle, which runs about 30 min. They also custom blend resin to meet Chrysler specifications, and continually monitor viscosity, specific gravity, gel time, and time from gel to peak exotherm. After demolding, hoods are fixtured until they are bonded. There are daily and weekly maintenance checks for the epoxy tools.

Viper body panels are molded of Modar 824 LTS resin, a thermosetting modified acrylic. The resin is now supplied by Ashland Chemical Co., Columbus, Ohio, which recently acquired the Modar product line from ICI Acrylics. Ashland also supplies Viper with its Pliogrip 6600-series polyurethane adhesive, used to bond the inner and outer RTM panels. Through out the Viper program's development curve, Ashland has tailored its Pliogrip adhesive cure rate to comply with the production rates of the body panels.

SPAWNING NEW APPROACHES

Aero Detroit says that as a result of its experience in the Viper program it's now developing a novel RTM processing scheme, which involves use of a proprietary internal release agent, hotter tools (180-200 F range), and a new single-component polyester resin system. Moore explains that the new manufacturing approach, which is slated to be demonstrated later this year, hinges on development of the new internal mold release.

This resin-blended mold release, in concert with the one-component resin system, would eliminate metering and mixing of resin and catalyst and the need to flush the resin pumping devices. Instead, the resin would be preblended with a latent catalyst, which would be activated by the heat of the tool.

The major benefit of this new method would be RTM processing that could be fully automated, with on-line SPC, much like current thermoplastic injection molding. Moore expects that cycle times could be significantly reduced, tooling would remain relatively inexpensive and easy to modify, and RTM parts could duplicate SMC's capability of being able to mold to a more precise, net-shape part.

Another new concept, being researched by Chrysler, is the use of nickel-shell tooling that could be quickly, economically modified to respond to engineering changes. Rao declines to elaborate, but Chrysler is known to be conducting research in this area with an unnamed Canadian tool builder.