Investigation of Carbon Fiber Composite Surface Preparation Processes for Making Unshakeable Bonds

By Heather Wilson, Technical Writer, and Shawn Small, Owner and Engineer, Ruckus Composites

James Bond may like his martinis shaken, not stirred, but here at Ruckus Composites we prefer our bonds to be unshakeable. Repair of carbon fiber composite bicycles requires removing the damaged composite and replacing it with new composite lay-up. When it comes to safety, a repair is only as good as the adhesive bonds that hold it together. The major ‘holding’ force provided by the epoxy adhesive used in composite repair is chemical and not mechanical, making employment of surface preparation techniques, which produce favorable chemical bonding conditions, critical. An improperly prepared surface leaves the repair prone to failure and creates a serious safety hazard. In the eponymous 1962 film, a strong Bond thwarted Dr. No’s plan to bring about a failure to launch at Cape Canaveral, and failure is absolutely not a word cyclists want to ponder while rocketing down a hill on two wheels. Hence we take surface preparation for successful bonds very seriously at Ruckus Composites as part of our ‘’Vow of Quality Assurance.” We are continually striving to improve our repair processes, an aspiration which prompted us to question whether our surface preparation processes could be improved upon. 

There are many different ways to prepare a composite surface for bonding, and it is difficult to determine when a prepared surface is in an optimal condition in a production environment. In the past we have explored many different multi-step surface preparation methods based on information from our solvent and adhesives suppliers. However, we realized that the information we had to work with was insufficient for our quality control needs, so we turned to the Big Sticky Book of Science to develop a method and apparatus to evaluate our surface preparation process. We wanted to collect data that would guide us in formulating an improved process that: 

• Is highly effective in the shop environment (as opposed to laboratory conditions)
• Has universal application within the shop (works equally well for repair and paint prep)
• Minimizes inventory and complexity
• Is easy to train
• Is highly repeatable for all staff members to execute

The free surface energy of a prepared surface plays a major role in the success of adhesive bonding. Not a concept you’re familiar with? Read on for an introduction to surface energy, why it matters, how it’s measured. If you’re already best buds with surface energy, you can skip down to the section where we serve up the experimental apparatus we cooked up in-house to test the effect of a range of surface prep processes on surface energy, along with our results and data analysis.

Surface Free Energy

Bonding of two surfaces is created by formation of molecular bonds between the surface of the material and the surface of the adhesive. The strength of the bond is determined by how many molecular bonds can form, which is dependent on surface free energy. Surface free energy, or simply surface energy, is the excess energy present on the surface of a material as opposed to its interior bulk. Inside a material, the atoms comprising it are generally stable and have a balanced set of bonds with other atoms. However, at the surface of a material there is an unbalanced set of bonds, creating unrealized bonding potential, or energy. A surface with low surface energy is not capable of forming enough molecular bonds for strong adhesion. A surface with high surface energy allows for more bonding between the adhesive and surface. Here comes a slightly confusing bit of terminology: Cohesion forces are those that hold an adhesive together internally and adhesion forces are those that bond an adhesive to a substrate. Understanding the difference will be important in a minute when we come to cohesive forces within a water droplet.

Surface energy cannot be directly measured, but it can be assessed through a wetting test. Wetting is the ability of a liquid to maintain contact with a surface due to interactions between molecules in the liquid and molecules on the surface. The amount of spread of a liquid is due to a balance between cohesive forces that hold the liquid together and adhesive forces that attract it to a surface. A wetting test quantifies the amount of spread a water droplet has on a surface through measurement of the angle where the edge of the droplet intersects the surface. The amount of droplet spread is influenced by two main factors: surface energy and surface tension of the droplet. 

Surface tension is a phenomenon generated by cohesive forces in a droplet. Molecules on the interior of the water droplet are completely surrounded by other molecules and do not experience a net force as the attractive forces from neighboring molecules cancel each other out. On the surface of a droplet, the molecules are subject to a net inward attractive force because there is no attractive force from above, thus creating surface tension.

There are three interfaces that are involved in wetting: surface to liquid, liquid to gas (air), and surface to gas. Interfacial tension is a measure of adhesive force between the liquid phase of one substance and the liquid, solid, or gas state of another substance. Surface energy is described mathematically with Young’s equation, which states that γ_sg = γ_sl  + γ_lg cosθ, where γ_sg is surface free energy, γ_sl  is the interfacial tension between the liquid and the solid, γ_lg is the surface tension of the liquid (a.k.a. interfacial tension between liquid and gas), and θ is the contact angle of the liquid with the solid.

Math nerds may now be smuggly proud of your understanding! However, for those of us for whom this is all Greek, here’s a distillation. A high energy surface has greater attractive forces than a lower energy surface, producing a greater amount of interfacial tension between the surface and a liquid. The greater the interfacial tension, the greater the difference is between interfacial tension and surface tension of the liquid, leading to a greater amount of wetting and production of a lower contact angle.

Clear as mud? Just remember that a surface with higher free energy will cause more wetting and hence a lower droplet contact angle relative to a surface with lower free energy. 

Wettability can be used as a proxy for how well a surface has been prepared for adhesive bonding because adhesive bond strength is also dependent on the amount of surface free energy. This is why we chose to do a wettability study to determine which carbon fiber composite surface preparation processes would lead to the lowest droplet contact angles, and by association, the highest surface free energy. 

Wettability Study

To measure water droplet contact angles, an experimental set-up was built in-house. 80/20 T-slot aluminum was used as a base for mounting a 1080P USB camera and an adjustable height platform. The platform supported a custom 3D-printed holder for a Hamilton 10µL syringe, loaded with distilled water, at a fixed height over the platform. A diffused LED light source with a monochromatic gel was used as a backdrop to allow for greater resolution of the droplet margin in photographs for more accurate angle measurement.

Samples of standard modulus twill weave 6k carbon fiber composite from Protech Composites were used. These samples were not finished, cleaned, or treated by the manufacturer after the casting/infusion process. An untreated sample was used as a control and the remaining samples were subjected to one of 12 different preparation processes. Preparation variables tested, both alone and in combination, were type of cleaner (Simple Green^© or PPG SX330 ACRYLI-CLEAN^© wax and grease remover), sanding grit size (80, 320, or 800) or use of a Scotch Brite^© pad, solvent used in wiping (isopropyl alcohol, acetone, or denatured alcohol), and use of flaming. Each sample was placed on the platform and three equal-volume droplets of distilled water were successively dripped using the mounted syringe spaced apart in a row and photographed. ImageJ (Image Processing and Analysis in Java) open source software was used for image postprocessing and angle measurement. 

Results

Photographs of water droplets with contact angle measurement for each of the surface preparation treatments are shown in the gallery below (click on the thumbnail for a larger image) and the means are visualized in a bar chart.

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

Analysis

An analysis of variance (ANOVA) was performed to test the null hypothesis that the mean contact angles are the same between treatments. ANOVA uses the ratio of between group variation to within group variation to determine whether there is a statistically significant difference between the groups. For this analysis, we treated all the steps in a treatment as a single variable. For example, use of Simple Green^©  is a variable as is use of Simple Green^©  along with 80 grit sanding. The analysis showed that there was a significant difference in contact angle [F(12,26)=100.98, p=1.20E-18] between the treatments. This indicates that there is a difference between one or more pairs of treatments. A post hoc Tukey’s HSD (honestly significant difference) test was conducted to identify those treatment contact angle means that are significantly different from each other. This is a statistical test that compares the mean of each treatment to the mean of every other treatment and identifies any difference between two means that is greater than the expected standard error. The Tukey’s HSD test results are visually summarized in the table below.

Comparison of contact angle means between the two cleaners showed that there was no statistically significant difference between Simple Green^© and PPG SX330 ACRYLI-CLEAN^© wax and grease remover. Comparison of contact angles between pairs of the three solvents showed that use of isopropyl alcohol produces a significantly lower contact angle than denatured alcohol, or acetone. Comparison of the Simple Green^© plus one of four sanding treatments (80 grit, 320 grit, 800 gr, or Scotch Brite^© pad) showed that the SG + 320GR treatment yielded a significantly lower contact angle than use of the other grits or the Scotch Brite^© pad. The full Ruckus Process (which is proprietary) produced a lower mean contact angle than SG + 320GR + Iso + Fl, although the post hoc Tukey’s test did not show a statistically significant difference between the two treatments (p=0.29). However, the small number of replicates (n=3) in this experiment may be skewing results. Regardless, we can be confident that the full Ruckus surface prep process is highly effective at reducing contact angles, indicating that the surface free energy produced is high and will provide for strong adhesive bonding.

Carbon Bike Repair Failure Modes

An adhesive bond is only as good as the preparation process that preceded it. So what happens when surface preparation best practices are not followed? Bad, bad things can happen and leave riders in a sticky situation. The following situations are illustrative of the quagmire in which real riders have found themselves after a bond has failed. The faint of heart may want to read no further. Please note, these failures did not occur on bicycles that were repaired at Ruckus Composites. We will not dish the dirt by naming and shaming the responsible parties because everybody should have the dignity of learning from their mistakes without being thrust into the spotlight.

• A DIY repair patch job rarely includes proper surface prep and could easily peel off, thereby not provide a safe repair. Plus, your bike will be embarrassed to be seen in public. Please, just don’t do it. 

• A bonded and over-wrapped joint can become debonded. Debonded joints sound extremely unpleasant, whether they be on bicycles or the bodies that were riding them when the failure occurred.

• Bottom brackets can become debonded. The bottom bracket connects the crankset to the frame and allows it to rotate freely. Bike rides instantly become Type III Fun when you can’t pedal. Not familiar with categorizing fun? Then you really need to check this out: https://www.tetongravity.com/story/culture/the-three-and-a-half-types-of-fun-explained
• Dropouts can become debonded and fail. The dropouts are where the rear wheel hub rests. This type of failure can lead to an undesirable rapid transition from bicyclist to unicyclist. Nothing against unicycling, it’s just that folks normally get to choose this mode of transportation instead of having it foisted upon them at very inconvenient times.
• Front derailleur ‘braze-on’ mounts fall off. You didn’t really want to shift to the smaller chainring to ride up that monster hill, now did you? Didn’t think so.

• Cable stops can fall off. Brakes. Brakes are really important. Enough said.
• Tubeless rim tape will slide off or endlessly leak causing you to curse the gods of air pressure. The gods of air pressure can only be appeased through the type of sacrifice you weren’t prepared to make in the back of beyond (and it’s probably not legal in most places anyway).
• The paint and clearcoat can peel off a bike too easily when masking tape is removed–the stuff of bike painter’s nightmares. 

The Future of Ruckus Bicycle Science

The wheels of bicycle science will continue to roll at Ruckus Composites. Safety and quality of repairs is an imperative and we will continue to use science to inform our best practices. This study of the effect of preparation processes on carbon fiber composite wettability has helped us develop an evidence-based preparation protocol that will allow us to provide the best possible service. As a company with science-informed practices, we are continually thinking about ways to improve facets of our repair processes and what kind of testing would lead to informative data. We mentioned in the failure modes above that low surface free energy can make it too easy to accidentally peel off paint or clearcoat when removing masking tape. Which raises the question of how “easy” or “hard” to peel can be tested and quantified in order to compare different paint and coating types or preparation techniques. We love our bike painter and don’t want him to have nightmares, so tune in for our next edition of Bike Science Friday to see how we use science to make masks less frightening.

Acknowledgements

We would like to thank Protech Composites in Vancouver, WA, for their generous supply of test samples. https://protechcomposites.com