Despite Progress, Composite Repair Will Remain a Major Barrier to CFRP Adoption

In Lux’s recent overviews of the landscape of carbon fiber recycling technology and of the outlook for carbon fiber in emerging rail and marine applications, we highlighted the need for scalable composite repair technology as a key limiting factor in future growth in continuous fiber composite adoption. Repair has also been a consistent concern for automotive carbon fiber.

Minor cosmetic damage can be repaired with adhesive and coatings, but for structural damage, historically the most common approach was to discard and replace the damaged composite part. While expensive, such an approach is fast and reliable, and in high-end applications like aerospace and racing cars where reliable performance is key, generally acceptable. This was especially true when non-destructive composite testing and inspection technologies such as ultrasound and thermography were less available, portable, and reliable than today. For less extensive damage or damage that does not involve structurally critical regions, the most common approach in aerospace, high-end automotive, and bicycle frame CFRP repair is to remove the damaged portion by cutting, milling, sanding, or laser ablation, and then to produce a custom patch to fill the damage. The patch is molded from equivalent material to the original part, and bonded to the part with adhesive. Traditionally this was done by hand by highly skilled experts. For example, from 2008 to at least as recently as 2012, Lamborghini maintained a team of Boeing-trained composite repair specialists who would fly around the world to repair car parts. The process typically takes one day to three days, but sometimes up to 14 days.

The trend in aerospace has been to move away from hand-layup and design for patches and towards automated processes. Between 1999 and 2005, machine tool company American GFM began working with a consortium of aerospace companies to develop the “Inspection and Repair Preparation Cell” (IRPC), designed to automate and integrate the various steps of the composite repair process. Prior to full integration and automation, however, large companies such as Airbus and start-ups such as Oxford Performance Materials have developed methods to use 3D imaging and scanning technologies, including stereoscopic cameras and high resolution laser scanners, to determine the precise shape of the damaged region of a part. With portable handheld cameras and scanners, this can be done quickly in the field. Using computer aided design (CAD) software and automated prepreg layup, engineers can efficiently design and produce custom patches as reliably as repairs done by hand, with more opportunity for use of modeling software to predict and confirm the performance of the repaired part. Companies like GKN Aerospace have claimed that this kind of automation can reduce composite repair costs up to 60%.When BMW launched the all-CFRP i3 electric vehicle, reparability of a large fleet of composite vehicles was a major challenge. To address it, BMW designed the vehicles so that large composite parts could be easily cut at pre-defined points with a custom cutting tool. That way, the damaged section can be removed and replaced without patching and without replacing the whole part. BMW’s own blog post on the topic suggests that, based on insurance premiums, they do not expect i3 repairs to be significantly more expensive than other vehicle repairs, in part because the automated cutting process is faster and less labor intensive than metal repair.

Aside from the costs, repaired part performance remains a significant concern for CFRPs. Unlike in metals, for continuous fiber composites the local mechanical integrity of a part depends on the global as well as the local structure, orientation, and bonding of the carbon fiber network. Patched parts no longer have continuous fiber at the bond between the patch and the original part. Non-destructive testing methods such as ultrasound, radiography, holography, thermography, and x-ray testing can verify the quality of the repaired structure, but cannot confirm performance parity to the original. Inspection tools that can verify the performance of the repaired part are a major unmet need in this space.

Moving beyond both the repair-by-replacement and patch-based approaches is a daunting prospect, but there are a number of possible paths forward. All of these paths, however, require innovation in the composite materials themselves. For example, start-up Mallinda developed a polyimine matrix polymer that exhibits what is called dynamic covalent chemistry. When heated, the crosslinked thermoset becomes malleable and can be re-molded, but then forms new crosslinks so that the re-molded part is as mechanically strong as the original. This approach is still susceptible to damage to the carbon fiber itself, but highly resilient to damage to the matrix material. Many companies in the industry have explored using thermoplastic composite matrix materials for the same reason. In principle, switching to anisotropic composite structures (for applications that can tolerate them) could allow a more fundamental but impactful shift. These include chopped fiber composites, as well as more exotic materials such as carbon nanotubes (CNTs) and carbon nanofibers (CNF). In principle, CNTs and CNF should be able to achieve comparable mechanical performance to CFRPs despite much shorter lengths. For example, Oak Ridge National Laboratory (ORNL) developed a 3D printed CNF composite (10% loading in an acrylonitrile butadiene styrene (ABS) thermoplastic matrix) with claimed mechanical performance comparable to molded aluminum 6061. For similar reasons, Zyvex Technologies claims its Epovex CNT-containing epoxy adhesive produces stronger repaired composite parts; essentially the CNTs act as a bridge linking the fiber in the original part to one another or to fiber in a patch, instead of leaving them connected only by resin. Despite these advantages CNTs and CNF are expensive, unfamiliar, and less available than carbon fiber. Unfortunately, no one has demonstrated comparable mechanical performance with chopped carbon fiber, so to date that material has been limited to lower end non-structural parts. While chopped fiber composites will likely never rival the best continuous fiber CFRPs on performance, enough improvement may be possible through advances in sizing and surface treatment, resin selection, and molding processes to make them competitive for applications where reparability is important.  A more experimental approach, such as the ultrasonically aligned chopped fiber composites developed at the University of Bristol (for 3D printed glass fiber, but the principle applies more broadly), could boost performance further.

Ultimately, there is no clear winner today or in the pipeline that can make composite repair as straightforward and reliable as metal repair. However, advances in materials development, process automation, and part design are gradually making repairs more feasible and reliable, while reducing the level of expertise necessary to make them. CFRP repair will continue to make incremental progress that will make it easier to deploy CFRP in established applications, but it will remain a key bottleneck for emerging applications for the foreseeable future.

By: Anthony Vicari