Automakers have a lot of incentive to improve fuel efficiency.
As fuel costs continue to rise and commutes get longer, customers want better miles per gallon (mpg) and often make gas mileage a key factor when making car purchases.
In addition, current U.S. regulations mandate that by 2025 the average fuel economy standard must meet 54.5 miles per gallon, a 60 percent improvement over the 35.5 mpg required of vehicles now.
One potential path to creating more fuel-efficient vehicles is to create vehicles that are more lightweight. Reducing the weight of a vehicle by 10 percent yields a 6 to 8 percent increase in fuel economy, according to the U.S. Department of Energy (DOE).
One of the most promising lightweight material systems to replace heavy steel on automobiles is carbon fiber composites.
In addition to being lighter, carbon fiber composites also give engineers more design freedom than conventional materials, allowing them to produce different shapes and structures.
However, carbon fiber composites do not come without challenges.
The properties of the material are more complicated to model than metal, as carbon fiber composites depend on complex features such as the fiber loading as well as fiber length distribution and fiber orientation that occur in the material due to the manufacturing process.
Testing is also a challenge, said Leonard Fifield, PhD, a materials scientist and research leader at Pacific Northwest National Laboratory (PNNL).
“Carbon fiber composites are very different than metals, in that instead of having the same material throughout the part they are made up of a mix of different materials,” said Fifield. “One challenge to conventionally testing carbon fiber composites is the extrapolation of test sample results. If you are making a steel part, maybe you can test a small steel test specimen and you can know what to expect for the whole part. But when you have a carbon-fiber reinforced composite part you might have different properties at different places in the part. So if you just take a test sample, it may not represent the properties throughout the part.”
Because of this challenge, individual parts have to be molded, produced and tested to understand performance.
A team of experts from industry and academia, led at PNNL by Fifield and others, is working to solve these issues, developing predictive engineering tools for designing new, economical and lightweight automotive composites.
The research group—which includes representatives from Toyota, tier one part producer Magna, long carbon fiber material and technology supplier PlastiComp, modeling software provider Autodesk, and research partners from the University of Illinois, Purdue University, and Virginia Tech— was funded by DOE’s Office of Vehicle Technologies Lightweight Materials Program. Together, the team created software tools that successfully predict the fiber orientation and length distribution of complex carbon fiber thermoplastic parts.
“The DOE was looking for a demonstration of predictive tools for a certain class of carbon fiber composite,” said Fifield “Our goal was to demonstrate and validate the existing tools. We put together different pieces of the process and demonstrated them for a long carbon fiber composite. The tools, like the software, were improved in the process.”
The software is designed to be commercially available to automotive companies looking to better understand the performance of these materials.
Using the engineering software, manufactures will be able to “see” what the structural characteristics of proposed carbon fiber composites designs would be like before it’s molded. The tools allow manufactures and auto part designers to experiment and explore new ideas at a much faster rate.
To predict fiber orientation and fiber length distribution in molded components, long carbon fiber components were molded and the fibers extracted for measurement.
Researchers then compared the predicted properties from the simulation software to the test results of the molded fibers to validate the accuracy of the software and models. They found the software tool successfully predicted fiber length distribution in all cases and fiber orientation in 88 percent of cases.
They also analyzed the performance gains and costs of long carbon fiber components versus standard steel and fiberglass composites. PNNL found that the carbon fiber reinforced polymer composite technology studied could reduce the weight of automobile body systems by over 20 percent.
However, they discovered that production costs of carbon fiber components are currently up to 10 times higher than those of steel.
“The costs are associated with the material itself—the carbon fiber—but also with the processing,” explained Fifield. “The cycle time really dominates the cost of producing the parts.”
One way to reduce the cost would be to decrease the production time, he said.
“For high-performance carbon fiber composites, they take time to cure, so if it takes five minutes to make a part, that is years in car production time. They want to be able to make a part in a minute or less. That is really a barrier to the penetration of carbon fiber composite parts in the vehicle market. If you could make many more parts with the same equipment in the same amount of time then each part is going to be cheaper.”
Research focuses for the DOE have now shifted from reducing the cost of carbon fiber to reducing cycle time, which should reduce production costs, said Fifield.
“That is an active area of research right now,” he said. “Some ways you might do that are develop active resins that cure faster or have more ways to control heat in the molding process.”
Better optimization of processes and structures using predictive tools could also significantly reduce production costs, paving the way for greater use of carbon fiber in automobiles.
The next step is to improve the predictive tools, said Fifield.
The team was able to use the software to predict the result of the production process of a part and then predict the stiffness of that part. However, while many components on a vehicle are stiffness driven, many more parts are strength-driven, said Fifield. Not understanding the strength of the carbon fiber composites can lead to overdesign and safety limitations, he said.
“Predicting strength is much more complicated and the tools don’t yet exist to do this,” said Fifield. “That would be a next step for this research, to develop tools to predict strength. We need to develop the tools to predict other properties like strength and to optimize the production process. Existing research needs include understanding the thermal properties of the process, controlling them, minimizing the time required and maximizing the performance.”