Oak Ridge collaborates for cheaper carbon fiber

  • 27-Aug-2012 04:16 EDT

Carbon-fiber tow exits a microwave-assisted plasma carbonization unit that processes fiber feedstock.

Improving fuel economy during the next few decades will come down to developing fuel-efficient power plants and lightweight structures. But if automakers are to build vehicle structures that meet increasingly stringent safety and quality standards while shedding significant mass, they will need to use carbon-fiber composite materials, which have a quarter the density of steel but several times its strength and equal stiffness.

Unfortunately, the very stuff of today’s high-performance vehicles—racecars, exotic sports cars, jet fighters, spacecraft, and America’s Cup sailboats—is still too costly for use in tomorrow’s mass-produced vehicles, including range-challenged electric vehicles. Even the lower-quality industrial grades of carbon fiber that would meet the structural needs of future composite cars (tensile strength: 250 ksi, tensile modulus: 25 Msi, 1% ultimate strain) are too expensive at the current price of $10 to $20 per pound.

If the cost of carbon fiber could be halved in the next five years, however, a tenth of all luxury cars and a hundredth of regular cars would use it, according to a market research study conducted recently by Lucintel for the American Composites Manufacturers Association. Worldwide demand for carbon fiber at a price of $5 to $7 per lb could grow to near 305 million lb (138 million kg)—worth some $1.525 billion—annually by 2017, which is about three times existing annual demand for carbon fiber, the study states.

Collaborative research center

Overcoming the challenges to cheaper carbon fiber is exactly what a research team at Oak Ridge National Laboratory (ORNL) is aiming to do at its new Carbon Fiber Technology Facility in Tennessee. ORNL recently received a $35-million award from the U.S. Department of Energy (DOE) to build and operate the lab, which will include a pilot plant capable of producing up to 25 ton (23 t) per year of new carbon-fiber materials, said Tom Rogers, Director of Industrial Partnerships at ORNL.

The industry/government R&D venture is part of the DOE’s Low-Cost Carbon Fiber Initiative in which its Vehicle Technologies Program has invested some $20 million. Industrial consortium members including 3M, BASF, Dow Chemical, Ford, GE, Graftech International, SGL Carbon Fibers, Toho Tenax America, United Technologies, and Volkswagen, as well as other government agencies, have also made significant investments in the project, which was originally established by manager Dave Warren.

Other research efforts to cut carbon-fiber costs are under way as well. Japanese fiber-supplier Toray and Daimler, for example, launched in recent years a joint R&D venture to focus on the manufacture of automotive CFRP (carbon fiber-reinforced plastic) parts, as did SGL Carbon with BMW. Meanwhile, VW took an 8% stake in SGL Carbon, while Toho Tenax partnered with General Motors to co-develop lightweight structural technology.

The Oak Ridge carbon-fiber program is notable because it will evaluate several innovative fiber-processing technologies that have been devised in the lab of principal investigator Felix Paulauskas. This former Bell Labs materials researcher has developed a number of gas-phase, continuous-processing methods such as microwave-assisted plasma techniques and advanced carbonizing methods that the new facility will test once it is running.

Fabricating novel fibers

Current production of carbon fibers is slow and expensive, Paulauskas said. “Costs are high because of the exacting and unforgiving nature of the process,” he noted. “You can’t go back and fix mistakes.”

The current raw material is typically a polyacrylonitrile (PAN) precursor that must be converted to carbon using thermal pyrolysis, a time- and energy-consuming process that must in addition be combined with stretching to align the microstructure to achieve the desired physical properties.

The precursor accounts for 45% to 55% of the production cost of commodity PAN fiber, with most of the remainder related to the processing, which involves steps such as oxidative stabilization, carbonization, graphitization, and surface treatment, Paulauskas said. The first aim, then, is to find cheaper precursor materials that can be processed into good-quality fibers. The plan is to test many types of potential low-cost precursors.

“The new facility, which is the only one of its kind, is designed to be highly flexible,” said Cliff Eberle, Technical Development Manager for composite materials at ORNL. “It will allow us to run all sorts of precursor materials, multiple chemistries, and formats to determine their utility and demo their viability.”

The Oak Ridge team will try out cheap materials such as low-cost polymers, inexpensive textiles made from low-grade, low-quality plant fibers, and renewable natural fibers such as lignin, which is a by-product of paper mills (the last with the help of Pacific Northwest National Laboratory and MeadWestvaco). One test example is a chemically modified textile acrylic PAN with vinyl acetate co-monomer.

Dow Chemical, for example, has received $9 million in DOE funds to create a lower-cost carbon-fiber production process that uses polyolefins—polyethylene, polypropylene—in place of PAN as the feedstock. Prototype fibers have achieved good tensile properties: 200 ksi strength and 20 Msi modulus. Collaborators include Ford, which aims to cut the weight of new cars and trucks by up to 750 lb (340 kg) each by the end of the decade. The industrial partners will contribute some $4 million to the effort. Cost estimates indicate that this method could yield a 50% reduction in carbon-fiber costs.

Another possibility is melt-spin PAN, which would be a single-step alternative to the current multistep process. “If we can avoid many steps in one extrusion process, we can reduce costs,” Paulauskas said. Melt-spinning suitably melt-stable polymers into fiber directly could generate 25% in cost savings, but he warned that this and all the new techniques are still at the feasibility stage.

The test fibers will undergo a thorough parametric evaluation process starting with the filaments, through tows and onto fibers. A lab-scale continuous fiber operation will be able to wind four to six tows into fiber at rates of maybe 10 to 20 in (254 to 508 mm) per minute.

Conversion, post-treatment savings

Beyond the generation of a low-cost precursor, two other key technologies are involved—conversion and post-treatment. Conversion involves processing in which the filament is pretreated, stretched, heated, oxidized, carbonized at both low and high temperatures, and post-treated.

“The surface of the fiber is chemically inert,” Paulauskas explained, “so you need to attach chemical groups that enable the fiber to adhere to the resins.” Following this electrochemical treatment, the fibers need to be carefully washed and dried, he said, adding that “management of the solvents is a large part of the process.”

Paulauskas’ group has demonstrated that microwave heating to carbonize and partially graphitize PAN precursors in a plasma rather than conventional ovens can speed the process by one-third while cutting costs. Tests show that microwave-assisted plasma (MAP) processing can produce useful, uniform properties along fiber tows. The nonthermal plasma process generates ozone and oxidative species, which stabilizes and cross-links the polymer rapidly and efficiently. Recent cost estimates suggest that it could reduce oxidation cost by about half.

Finally, the production process requires a surface treatment, or sizing, to protect the fiber’s properties from humidity, to make it “dry-tacky, handle-able” by workers, as well as to provide partial solubility with and partial cross-linking in the adhesive matrix to produce structural bridges to the fiber surface to improve the bond.

This advanced post-treatment improves short beam shear strength significantly, Paulauskas said. Epoxy composites made with aerospace-quality fiber “have a shear strength of 16 to 19 ksi, whereas automotive-grade fiber is maybe 6 ksi,” he said. “But with our proprietary sizing process, we’ve gotten in the 11- to 15-ksi range. And the process is now ready to go to the next stage.”

HTML for Linking to Page
Page URL
Rate It
4.84 Avg. Rating

Read More Articles On

The technology uses multiple foils with multiple messages and an LED light source. Each specific message is burned onto the holographic film through a photographic process.

Related Items

Training / Education
Training / Education
Technical Paper / Journal Article
Training / Education
Training / Education
Training / Education