Custom Carbon Fiber Automotive Wing (NB Miata) - September 2025

Overview

This project focused on the design and near-completion of a custom carbon fiber rear wing for my NB Miata, targeting autocross, time attack, and track-day use. My primary goal was to outperform commercially available aftermarket wings in aerodynamic efficiency, mass, and cost, while maintaining an aggressive, motorsport-inspired aesthetic. Because no formal constraints were imposed, all design requirements were self-defined and benchmarked against existing market offerings, particularly against high-end Miata wings. The project has spanned approximately one academic semester and included concept development, CAD, CFD analysis, structural load estimation, material selection, and fabrication of the main wing structure. At the time of writing, the wing assembly is nearing completion, with pylons and endplates pending fabrication and no on-car testing yet performed.

Project Objectives

I had the following objectives in mind throughout this project:

• Design a visually aggressive, functional wing consistent with modern motorsport aesthetics
• Achieve a higher downforce and downforce-to-drag ratio than comparable aftermarket Miata wings
• Offer a competitive total mass relative to aluminum and carbon competitors
• Maintaining a lower target cost than premium carbon fiber alternatives

To me, success was defined as achieving at least two of the above objectives when compared to existing products.

Commercially Available Wings

My design decisions for this project were centered around competing with other popular wings. More specifically, I had long been interested in purchasing a Miata wing from 9livesracing.com, but its price of over $1,000 for just the base model made it hard to justify. However, the Nine Lives Racing (NLR) wing is generally considered the cream of the crop when it comes to wings. Thus, I set out to design and build a similar or better wing for cheaper.

To make a better or cheaper wing, I first had to pull benchmark values from the NLR website. NLR offers several customizations for the wing, and I've listed 3 popular, realistic builds below.

Because the NLR wing and my wing will both have heavy adjustability, and because race cars operate at a plethora of speeds, I've decided to benchmark testing at 100MPH and an Angle of Attack (AoA) of 10 degrees. I calculated weight, downforce, and drag numbers for each NLR build:

NLR does not share much info on their website, so I used several supplementary sources and general rules to make downforce, drag, and DF/D ratio calculations, including:

• Base Kit Total Weight: A 64" aluminum wing with endplates and bottom mounts weighs 14.6 lbs.
• Carbon Fiber Weight Savings: The carbon element is reported to be ~4 lbs lighter than its aluminum counterpart for certain applications.
• Aluminum Chord Dimensions: Standard aluminum wings have a 9.5" chord.
• Carbon Fiber Chord Dimensions: The carbon fiber wing features a 12" chord, making it 28% larger than the aluminum version.
• Lift-to-Drag (L/D) Ratio: The "Big Wang" aluminum airfoil is engineered for a maximum efficiency of 15.2:1.
• Swan Mount Performance: Swan-style mounts provide a 6% to 10% increase in downforce by cleaning the air on the wing's high-pressure underside, with a 1% increase in drag.
• MEGGA Endplate Performance: These plates increase downforce by 10% compared to basic endplates with no reported increase in drag.
• CFD V2 Endplate Performance: These plates increase downforce by 10% while simultaneously reducing total drag by 7% to 7.5%.

Using these general rules, and the data that Nine Lives Racing does publish, I calculated the 3 wing options to have downforce/drag ratios of 9.95 for the basic build, 11.92 for the mid level, and 12.96 for the exotic build. I estimated the wing builds to weigh 14-15lbs, 19-23lbs, and 16-19.5lbs respectively. At 100MPH, the 3 builds make 199, ~240.8, and ~308.5 pounds of downforce. These benchmark values and build prices established realistic performance and cost goals/standards, and ensured the project remained focused around real constraints and objectives.

Wing Specs and Geometry

I used the same airfoil shape that NLR uses, as their Benzing 123-125 style airfoil is specifically tailored to motorsports, and regarded as one of the best in its field. (per Occam's Racer) The 9.5" chord they use proved to be a visually appealing and high performing size, something that I kept for my wing. An NB Miata is just over 66" wide, so NLR's wing width of 64" leaves ample clearance, and thus, I'll be carrying that forward into my wing also. While mostly aesthetic oriented, swan neck mounted wings (wing mounts on top vs below) do offer up to 10% more downforce than bottom mounted wings. With cool factor and performance in mind, I opted for swan neck mounting. In my eyes, the NLR mounting fins (mounting fins that protrude from wing body) are bulky and ugly. Thus, I machined low-profile, sleek dual mounting fins. The pylons were designed with several modern cars in mind (more below), and offer a smooth, satisfying rake, and aesthetically pleasing mounting neck. Unlike most aftermarket wings that mount to the trunk lid, which can cause heavy deflection and gasket wear over time,(via Dsport.com) my wing will mount to the trunk's rain sills, directly connected to the chassis. No rubber, no flex, just downforce. Most wings that mount to the Miata chassis block the trunk lid from opening, and thus greatly diminish the car's practicality and daily drivability. As my daily driver, this was a big deal, so I designed the pylon-wing interface to use clevis pins at the front mounts, allowing the user to quickly disconnect and swing the entire wing body downwards, creating more than enough room for the trunk to open.

I wanted the wing to be adjustable, and thus added several mounting holes, offering a total of 12.5 degrees of adjustment, in 2.5 degree increments, all while keeping the mounting solution low profile. Endplates work the best when they're at a certain angle relative to the ground and wing, and thus, I added adjustment slots, allowing for up to 30 degrees of endplate angle adjustment.

This geometry and design were specially developed to prioritize aerodynamic efficiency and stability while remaining compatible with the Miata’s compact rear profile.

Structural Design and Load Analysis

At 100MPH, my wing makes 220.42lbs of downforce, and 21.13lbs of drag. This gives my wing a downforce/drag ratio of 10.43. All downforce and drag values were measured via Simscale simulations at 45M/S. An air density of 1.196kg/m^3 was used instead of the “default” 1.225kg/m^3, as it presents a more realistic, everyday scenario for the wing. (higher ambient temperature)

Structural loads were derived directly from CFD-generated downforce values. To determine my wing's maximum working load, I ran a wing CFD at roughly 145MPH, a realistic top speed for my Miata. At 145MPH, the wing creates 966.99lbs of downforce, which is almost half a ton. (That's a whole extra grand piano of traction!) I multiplied this by 1.5 to determine the ultimate limit, 1,450lbs, to account for uncertainties in loading, material properties, and manufacturing variability. (Remember that this load is evenly distributed between each pylon, and thus load calculations call for 50% of wing force) I ran stress simulations on the pylon, both at maximum working load, and ultimate limit to determine whether they scaled the same. Because they scaled the same, we can infer that the pylon does not yield before the ultimate limit. The maximum allowable strain for typical carbon fiber components is ~0.8–1.5%. (Daniel & Ishai – Engineering Mechanics of Composite Materials) At 100MPH, strain measured just 0.2353%. At a top speed of 145MPH, strain measured only 0.3530%. These measurements are not only linear, but far within allowable values. I used these values to calculate the margin of safety, which, at 1.266 is very safe, and the maximum load, a whopping 1643.3lbs per pylon, or 3286.6lbs total! These measurements and calculations indicate that the pylons can safely hold the wing up to, and past any downforce the wing will realistically generate. All calculations are displayed at the end of this page.

Structural features include:

• Carbon fiber main structure using quad layer 2×2 twill weave shell and wire cut XPS foam internals, all sandwich bonded together using 3M DP420.
• Internal 6061 aluminum ribs at each wing edge and pylon interface. Edge ribs are drilled/tapped for endplate mounting.
• Custom-machined aluminum inserts prevent localized carbon damage at all interface points, including pylon-chassis interfaces, pylon-wing interfaces, and wing-endplate interfaces.
• Isolating resin beads strategically placed between all aluminum and carbon fiber bonds, preventing material contact and galvanic corrosion.
• 1/2" Chopped carbon tow (“forged carbon”) pylons and endplates.

Based on estimated aerodynamic loads, a four-layer carbon laminate was selected for the main wing structure. While full laminate optimization and destructive testing were beyond the project scope, conservative assumptions were applied to ensure adequate stiffness and strength.

Materials and Manufacturing

I knew from the start that I would want carbon fiber components, as they are regarded one of the most exotic automotive materials due to the complexity and cost of manufacturing. I ended up using US Composites' 5.7oz Twill '1st Quality' Carbon Fiber Fabric for the wing body, and EasyComposites' 12mm Virgin Carbon Fiber Chopped Tow for all forged parts.

For the wing body, I opted to use a 3D printed mold. I used CU Boulder's 3D printers in the ITLL to print 8.5" mold sections, and bonded them together using JB Weld. After some sanding, PVA, and mold release prep, I laid 4 sheets of carbon fiber with precise amounts of resin in between, for an optimal 60/40 carbon/resin ratio by volume. Initial resin distribution and application were not critical, as I used vacuum bagging to ensure even resin distribution and minimal air bubbles. Vacuum bagging was fairly straightforward, and I didn't run into any hiccups. I opted to use a vacuum pump from Harbor Freight, as commercial, composite-specific options were prohibitively expensive. I did this twice overall, once for the top half and once for the bottom.

Wing internals consist of 4 aluminum ribs, and Extruded Polystyrene (XPS) foam. I used resin beads to prevent direct carbon-aluminum contact, and used DP420 epoxy for all aluminum-carbon, foam-carbon, and carbon-carbon bonds. I bonded one interior rib at each end of the top shell, and two mounting ribs centered, 39" apart. The foam was also bonded, and the bottom half was bonded to this assembly and left to cure.

Pylons and endplates used a more primitive carbon fiber build technique called forged carbon fiber. This process consisted of 3D printing male and female molds that were specially designed to bolt together, prepping and loading each female mold with 40/60 resin and chopped carbon tow by volume, and pressing the male mold on, sandwiching the carbon and resin. This assembly was clamped using vises and C-clamps and left to cure. This process was used for both pylons and endplates. Aluminum inserts were bonded in at all bolt interfaces once cured.

All carbon fiber parts were sanded and prepped, and sprayed with a coat of UV-resistant 2K clearcoat for maximum durability and longevity.

The design avoids overly complex molds, keeping fabrication achievable while still delivering high-end performance and appearance.

Cost Considerations

The wing was designed with a target cost of approximately $1,000, with the explicit goal of undercutting premium carbon Miata wings while delivering comparable or superior performance. Spoiler alert: goal reached. The total cost of my wing build comes out to $1089.28. This includes all materials, filament to 3D print my own molds, and a brand new vacuum pump from Harbor Freight. However, if you can get your hands on a used vacuum pump, or even borrow one from a friend, you can save on costs. Additionally, if an institution near you lets you 3D print for free, (like CU Boulder does) you save 3D printing costs. If you can source a pump and print for free, the total build price drops down to $523.46, or around 50% of the cost of the cheapest NLR build.

I achieved cost reductions through:

• Single-element aerodynamic design
• Simplified geometry
• Strategic material use
• Minimizing part count

Aesthetic and Design Inspiration

Although I was far from starting the project, I already knew that I wanted to develop a custom wing for my car when I was still studying abroad. I was lucky enough to travel to the Porsche museum in Germany, where I took special notice to rear wings, and how they were mounted. I attribute much of the design of my wing's uprights (pylons) to the Porsche 911 GT3 (992)'s sweeping, smooth swan neck. I worked to encompass its aggressive but purposeful proportions, and expose details and exotic finishes. This presented an excellent link between form and function that I sought to reproduce. Another Porsche I took inspiration from is the 911 GT3 R Hybrid (819), which blends functionality and aesthetics in its aero components.

Results and Discussion

Geometry and Design

Although looks and beauty are subjective, I can use general trends to determine if my wing wins the design category. My wing's endplates and pylons are constructed of forged carbon fiber, which is widely considered an exotic, premium material. It is often used in high-performance cars and luxury accessories. On the other hand, most comparable wings, and every NLR wing build uses powdercoated aluminum endplates and pylons. These are bulkier, heavier, and one-dimensional. Due to its multidirectional construction, forged carbon fiber creates a multi-dimensional texture that's not just ultra high-end, but also a piece of art. This is a testament to high-tier engineering, and shows that I'm not just buying off-the-shelf aluminum, but actually prioritizing advanced composite materials. Additionally, NLR uses pylons with large-radius curves and slab design. This results in dated, unrefined looking pylons. At the same time, my pylons use sharper curves, defined chamfers, and a skeletonized design, creating an aggressive machined look, consistent with modern race car styling. My skeletonized pylons utilize relief cuts for weight savings and visual transparency. Being able to see through the pylon makes the entire rear of the car look more intricate and engineered, and relief cuts show my commitment to world-class racing: If material doesn't need to be there for structural rigidity, remove it. I carried features that work well on the NLR wing forward, like their Benzing style airfoil shape, their wing dimensions, and their chassis-mounting feature. Overall, my wing shares features that work well on the NLR wing, and it improves in areas where possible. Considering my wing's more advanced design and aesthetically elevated construction, I have met my goal of designing a visually aggressive, functional wing consistent with modern motorsport aesthetics.

Downforce and Load Performance

At my benchmark of 100MPH, my wing makes 220.4lbs of downforce, and 21.1lbs of drag. This outperforms the NLR basic wing build, which makes only 199lbs of downforce and 20lbs drag. My wing makes 91.5% of the downforce of the mid-level build, and 71.5% of the downforce of the high-end exotic build. At this same benchmark, my wing has a Downforce/Drag (DF/D) ratio of 10.4, while the three NLR builds have ratios of 9.95, 11.92, and 12.96 respectively. This makes comparison to my wing a little difficult. My wing outperforms the basic NLR wing, making 10.8% more downforce at the benchmark. While it doesn't make more downforce than the other two builds, it is more efficient, demonstrated by its higher DF/D ratio. Ultimately, the only NLR wing build in the same universe of pricing is the basic build, which my wing outperforms. While the other builds make more downforce, they also cost 80-282% more than my wing, while being less efficient. Considering just the basic NLR build, and the efficiency of the other two, I would consider my goal of achieving a higher downforce and downforce-to-drag ratio than comparable aftermarket Miata wings completed.

While I didn't have any specific goals for structural rigidity or load, it was important to ensure that the wing pylons would be strong enough to function, and wouldn't fall apart or break. At maximum speed of 145MPH, the wing would make approximately 1/2 ton of downforce. Split between each pylon, this results in only 0.235% strain, which is well within allowable maximums of 0.8-1.5%. At the ultimate load, strain reached just 0.353%. At this load, the margin of safety was +1.266, an extremely safe value. To reach a conservative strain maximum of 0.8%, the pylons would each need a load of 1643.3lbs, for a total wing load of 3286.6lbs, more than the entire weight of the car.

Materials and Weight

I deduced that the Nine Lives Racing wing builds weigh 14-15lbs for the basic, 19-23lbs for the mid-range, and 16-19.5lbs for the exotic build. At a total of only 6.76lbs, my wing weighs only 29.4-48.3% of the weight of an NLR wing. I attribute this large difference to my wing's more sophisticated and lighter pylons, endplates, and body. Considering this difference, my goal of offering a competitive total mass relative to aluminum and carbon competitors has absolutely been met.

Total Cost

I configured 3 Nine Lives Racing wings that I would benchmark against, each build with its own strengths and weaknesses. The cheapest build was fully aluminum, very basic, and used regular bottom mounting. It comes out to $995.00. The next build still uses an aluminum wing, but features more advanced endplates, and swan neck mounting, priced at $1,789.67. The highest end build most closely resembles my wing, featuring a carbon fiber body, developed endplates, and swan neck mounting. It costs a whopping $3,804.33. My wing has a build price of $1,089.28. Unfortunately, this is 9.5% more expensive than the cheapest NLR build. However, considering my wing's much more advanced construction and features, its value is also much higher than the basic NLR wing. Additionally, my wing can be built for only $523.46, (47% cheaper than basic NLR) if the builder can borrow a vacuum pump, and 3D print or borrow a mold for free. Finally, if you take features and build into account, my wing outperforms even the most expensive NLR build, at an incredible 71.4% cheaper. That said, I have met my goal of maintain a lower target cost than premium carbon fiber alternatives.

Current Status & Next Steps

Main wing structure: Fabricated
Pylons and endplates: In progress
Vehicle testing: Pending completion of mounting hardware

Planned future work includes on-car testing to validate CFD predictions, evaluate handling impact, and refine endplate geometry. Planned related work includes carbon fiber coupon manufacture and testing to determine optimal design, weave, structure, and ratios.

What I Learned

• Performing carbon fiber layup using vacuum bagging techniques
• Creating forged carbon fiber elements using 3D printed molds
• Designing load-bearing structures using lightweight composites
• Performing simulations to determine forces and stresses
• Using CNC machining to create precise aluminum mounts
• Tuning wing angle for optimal aerodynamic performance
• Designing components to maximize downforce and aerodynamic efficiency
• Optimizing carbon fiber to aluminum and aluminum to steel connections to limit corrosion