CompositesWorld’s annual Carbon Fiber conference was held in early December 2014 in La Jolla, CA, US. Editors Jeff Sloan and Sara Black were there and offer this report on the proceedings.
Composites industry consultant Tony Roberts, principal of AJR Consultant (Orange, CA, US), presented to the conference his estimates of carbon fiber supply and demand over the next several years. He foresees carbon fiber demand more than doubling in the next decade, reaching 170,000 MT by 2025 (see break out figures in chart at left).
Leading the effort to meet this demand will be Toray Industries (Tokyo, Japan), followed by Toho-Tenax and Mitsubishi Rayon Co. Ltd., all headquartered in Tokyo, Japan, with production facilities there and in North America and Europe. Supply of carbon fiber out of China, a new and growing phenomenon, is the most difficult to forecast, he said, noting that precursor quality and supply problems in that country have made consistent production of quality carbon fiber difficult to maintain. Roberts predicted that the industrial carbon fiber market (large (24-48K), dominated by automotive and wind energy, will continue its growth throughout this decade and into the next. His overall message regarding carbon fiber supply and demand was positive: “I feel the industry has really moved forward now, and we’re well-set for the future.”
One of the conference highlights was a panel assembled on the second day to tackle questions generated by CW in regard to automation. The panelists were Gary Lownsdale, consultant to Globe Machine Mfg. Co. (Tacoma, WA, US) Dan Ott of Web Industries (Marlborough, MA, US), Jim Huddy from CGTech (Irvine, CA, US) and Barrett Milenski of ATK Aerospace Structures (Clearfield, UT, US). Although panelists gernerally agreed with Ott added that the growth of automation is has enabled greater consistency, repeatability quality in composite products, and reduced variability, Huddy said he believes we’re behind the metals industry in terms of using/adopting automation, but we’re learning. Milenski stated that it’s possible not to get the desired results with some automation, not that is it has to be the right process.
To the question, “What is the greatest benefit of automation, and why?” Lownsdale responded, “Cost reduction is the obvious one, but it’s more complex than that. For instance, look at an autoclave process: you could spend years optimizing the parts of the process, like automated layup, or better bagging, etc., but in the end, what we did,” he noted, referring to development work on Globe’s fully automated RapidClave exterior automotive body part molding/curing system (for more information, click on “Sub-8-minute cycle times on carbon/epoxy prepreg” under "Editor's Picks”), “was eliminate the autoclave itself, which wasn’t necessarily obvious at the start.” Ott’s answer was that the benefit is a better connection between the designer and the manufacturing floor, for improved manufacturability in a supply chain effort. Huddy pointed out that a huge fuselage part is virtually impossible to make manually, due to the labor cost/time restraints. In Milenski’s opinion, return on investment (ROI) is key. Given the huge expense of some of these machines, it is possible that, under some circumstances, a manual process could still be the better option.
When panelists were asked if composites manufacturing should be considered more art than science, and whether automation might change that perception, Huddy firmly stated that it’s not an art, it’s science and math, and very precise. “Automation is really helping the composites industry be less art and more science,” he affirmed.
Lownsdale agreed, and got a big laugh when he related a story from his experience as a co-op student at a brake-manufacturing facility. “They walked me around, and showed me the process, and there was a guy in the corner mixing up the ‘recipe’ for the molding process. My tour guide said that they had a great way of ensuring process success, because the guy doing the mixing would spit twice into the vat! Now that’s art.”
When asked, “Is automation a necessity for the composites industry to advance to a more equal footing with metals?” all panelists answered in the affirmative. “We must automate,” Ott insisted. “We need to have the continuous data capture and feedback on the process, to climb the curve, to have a chance of keeping up with metals.” Huddy and Milenski reiterated that when expected production rates and material out-times are considered, together, it’s simply not possible anymore to make large parts manually.
Low-cost carbon fiber research
At recent conferences, alternative precursors that might enable production of less-expensive carbon fiber have been on the agenda. The 2014 meeting was no exception. Connie Jackson, manager of Oak Ridge National Laboratory’s (ORNL, Oak Ridge, TN, US) carbon fiber facility, reviewed ORNL’s progress in the area of creating carbon fiber from unmodified textile PAN material supplied by Kaltex (Naucalpan, Mexico). The ultra-large tow (300K to 610K) fiber has a kidney/dog-bone shape with “a lot of surface area” and “modest” properties: Since issues with fused filaments were addressed, said Jackson “we’ve achieved approximately 500 ksi in fiber testing.” The group is tailoring the fiber, which can be split to lower tow counts, to specific applications that don’t require aerospace-grade fiber. In addition, three other textile PAN sources will be evaluated in FY2015, she reported, along with methods to reduce precursor crimp. In the face of some pointed questions, Jackson defended the group’s work, pointing out that ORNL is trying to lower market entry barriers for groups interested in making fiber.
Hailing from Australia’s Future Fibres Research and Innovation Centre (AFFRIC) located at Deakin University (Geelong, VIC, Australia), Shaun Smith described work there on improving carbon fiber. The group’s Carbon Nexus fiber-manufacturing facility, a pilot line that is open to groups engaged in advanced-fiber research, aims to dispel the “mystique” surrounding carbon production, which has stymied innovation by current priducers and presented a high barrier to entry for startups. AFFRIC has been investing in the application of RAFT (reversible addition fragmentation-chain transfer) monomer technology, which reportedly has the ability to create longer, more consistent carbon chains during fiber production. Scale-up is in progress, reported Smith, and he aims to answer the question “is RAFT PAN precursor better than traditional PAN?”
Mike Canario, VP and GM Americas at Hexcel (Stamford, CT, US), addressed the opportunities and challenges associated with his company’s role as a carbon fiber a supplier to the composites industry. He noted, first, that since the mid-1980s, carbon fiber demand in the aerospace industry has been fueled, in part, by several important but relatively low-volume military programs, including the B-2, V-22, F-22 and F-35. In commercial aircraft, carbon fiber had seen only limited use on a variety of Boeing and Airbus planes for many years until the Boeing 787 and the Airbus A350 XWB were developed and firmly established its place in the aerospace industry. These two planes, taken together at full-rate production, said Canario, will consume more carbon fiber in one year than the F-35 program will during its entire life.
Canario went on to say that the industry should expect 40,000 MT of additional carbon fiber demand in the next five years, adding that construction and commissioning of a new carbon fiber plant can take 12-18 months, which prolongs return on capital expansion investment. Given the current economic model, fiber producers must minimizing risk and cover their capital costs, and Canario contended that a carbon fiber manufacturer will expand capacity only if a sustainable rate of return can be guaranteed. That means the emphasis will be on long-term contracts — like that which Toray has with Boeing for the 787 and Hexcel has with Airbus for the A350 XWB.
Reality behind the headlines
One of the headliners on day one was John Byrne, VP aircraft materials and structures at Boeing Commercial Airplanes (Seattle, WA, US). Byrne was optimistic to start, offering several data points designed to highlight the health of passenger air travel. Boeing itself is also doing well; his company is currently building 63.3 planes a month (all models) and has a record backlog that is mixed geographically and categorically in terms of aircraft model. The company is building 10 of its composites-intensive 787s a month, and plans to increase that to 14 by 2018.
Yet, when it comes to composites and the 787, Byrne was less sanguine, offering his take on several problems: The composites industry supply chain (compared to that for metals) is relatively immature. Composites are not fully understood by Boeing designers and engineers and, thus, are not always employed optimally. As a result, the same basic combination of fiber reinforcement and resin matrix were applied on all parts and structures on the 787. The crux of the problem, then, said Byrne, is that composites industry immaturity has made Boeing’s capital outlay unusually expensive. This might be tolerable at the right volume, but the 787 build rate has incrementally increased more than anticipated, which has driven material prices higher than Boeing expected. Byrne went as far as to state that if Boeing knew then what it knows now, "material decisions might have been very different on the 787."
Byrne noted specifically that materials standardization in metals allows Boeing to avoid sole sourcing, which keeps costs low and the supply chain moving. He suggested that carbon fiber suppliers to the commercial aerospace industry should, like aluminum suppliers, provide a product or products that meets a given and established set of mechanical specifications, regardless of source. Looking further ahead at the next big commercial aerospace programs — replacements for the A320 and 737 — there appears to be in the carbon fiber community resignation to the fact that the fuselages of these craft will likely be aluminum, even if the wings are composite. But such programs, noted Byrne, are probably more than a decade away, which Byrne said gives composites materials time to mature and “once again earn their way onto aircraft.”
Bill Regan, VP engineering and technology at Entec Composite Machines Inc. (Salt Lake City, UT, US), offered an update on some new developments in towpreg manufacturing technology, particularly for filament winding. Towpreg, he noted, allows faster fiber payout (up to 5.7 m/sec on Entec’s new FW750 filament winder), better bandwidth variance control (±0.5 mm) and better resin variance control (±3%) compared to wet winding. Making good speed with towpreg, however, require good fiber tension control, which Regan says Entec has worked to optimize. He presented several case histories that illustrated the speed and efficiencies offered by a towpreg-based filament winding operation, including a study comparing a steel flywheel (for energy storage and discharge) to a composite flywheel. The composite wheel, which included carbon fiber from Zoltek (St. Louis, MO, US) and Toray (Tokyo, Japan), provided 45% more energy than the steel version (5.8 kWh vs. 4 kWh), had a rotor more than twice as long as the steel rotor (732 mm vs. 267 mm) and featured a total assembled weight of 220 kg, about one-sixth that of the 1,270-kg steel version.
Carbon Fiber 2015 returns to Knoxville, TN, US, this December. Watch for updates on the CW website: www.compositesworld.com.