Dialing in composites performance via dynamic digital twins

Source (All Images) | Sport Dynamics Lab

Composites continue to play a huge role in the world of sport. New materials are yielding lighter, stronger, more durable and customized skis, bikes, bobsleds, surfboards, bats, rackets, golf clubs, paddles, poles, hockey sticks, helmets and shoes. Meanwhile, the pressure to improve sustainability in sports is increasing, but for manufacturers this must be balanced with performance. And that performance is not solely based on the equipment, but how the athlete interacts with it.

Sport Dynamics Lab (Andorra) brings a new approach to evaluating performance. It moves beyond standardized static tests to measure dynamic response of equipment, couples that with athlete telemetry and other sensor datasets, and applies AI to provide correlations and actionable insights. The company also creates a digital twin calibrated with this data to validate equipment performance predictions, which also speeds prototype development and evaluation.

Featured Content

What is the future for Owens Corning Glass Reinforcements? READ MORE

Advancing mold release technology for aerospace composites manufacturing READ MORE

Episode 55: Dave Noblin, Greene Tweed READ MORE

Sport Dynamics Lab founder Alex Hunger has spent more than a decade developing and advancing this technology with brands like Mavic, Salomon and the Nidecker Group, as well as with elite professional teams.

“We help teams and manufacturers make performance decisions with evidence, combining dynamic testing, field telemetry and modeling,” he explains. “For manufacturers, we enable understanding of what is actually happening in these complex systems as well as objective comparisons that help improve product design and optimize materials. Through our patented Flexdynamics testing and ‘Empirical Digital Twin Loop’ workflow, we are turning assumptions and ‘feeling’ into accurate data that is reshaping how performance is measured and understood.” He also sees potential to expand this approach beyond sports into applications like drones, including propeller blades and wings, where dynamic response can also play a critical role in performance and durability.

Three-point bending is not enough

In the world of snowboards and skis, says Hunger, almost all manufacturers base performance evaluation largely on static stiffness and compliance. “How do I know if a snowboard is good quality and won’t break? It’s tested in three-point bending, and if the values are in a certain range, then it’s okay,” he explains. “But we’ve tested skis that have the same stiffness and when they are tested on the slopes, the athlete says, ‘This one is really good, but that one is bad.’ What they are ‘feeling’ is not a static property but dynamic behavior. We have seen that many skis with roughly the same stiffness react very differently in damping with torsion and bending.”

Sport Dynamics Lab turns dynamic behavior into data via lab test results combined with in-use telemetry from precision GPS, accelerometers and other sensors, and then uses AI tools to gain insights.

Damping is the reduction of oscillations in a system over time. In the case of sports equipment, these oscillations are caused by an input of kinetic energy and dissipated by structural material damping, although during use, other factors such as friction or aerodynamic drag may also be involved. “When a skier or snowboarder talks about responsiveness, this is actually torsional damping,” says Hunger. “We could see this clearly in a large correlation study we did with the Nidecker Group.”

He also notes a study completed with bicycle wheel and rim manufacturer Mavic. “In our testing, when comparing an aluminum wheel to a full carbon fiber composite wheel with the same design and dimensions, lateral behavior correlated more strongly with damping than with stiffness.” The lateral behavior of a wheel is critical for performance — affecting stability and handling as well as the efficient transfer of power from the rider to the bike.

“In sports, we have these beliefs that systems work in a certain way, but very few measure what these systems are actually doing in terms of physics,” says Hunger. “I'm trying to put a light on what is actually happening by collecting and analyzing data to establish correlations and enable better decisions.”

Developing the solution

Hunger has 20 years of experience in R&D. “My first jobs were as an industrial designer and engineer, developing interiors for Rolls-Royce and Aston Martin, lung capacity measurements for the medical industry and lighting with Philips. And in all this work, I was always building my own machines. For one project, I needed to do thermoforming, so I built my own vacuum forming machine and small injection molding machine.”

“Trying to develop more sustainable materials for surfboards is where I started to see the impact of vibration damping,” he continues, “and I realized that I needed to have some test methods. So, I built something like a three-point bending machine, but with an arm that flexes, so that when released quickly, the board would bounce, and I could see damping. I then developed software and was able to achieve good quality data and repeatability. That was in 2018.”

Hunger patented that technology, which is called Flexdynamics. “But it’s based on data from a real physical phenomenon during use,” he explains. “The goal is to get as close as possible to how the product is used but in a way that is accurately measurable and repeatable.”

The hardest part came next, which was understanding the physics behind the data, “especially in terms of the mathematics. As I developed the company, we also built our ability for telemetry. We have centimeter-level GNSS [global navigation satellite system, umbrella under which GPS sits] with real-time corrections, and equipment that is standard in training athletes. This data from real use is also important. For example, accelerometers on the skis or bike frame capture vibration signatures under different terrains and speeds, which show up as different frequency content. We then cross that data with what we get from the lab tests, and that enables a better understanding.”

Hunger still needed to make the whole approach work together to provide value. “We had data that was precise and helped us to understand the physics, but it had to be analyzed and visualized in a way that gives meaning.” Sport Dynamics Lab now provides a range of services, including R&D, testing and interpreting the results. “I have customers from South Africa, France and Switzerland to Asia, including sports equipment brands and OEMs to teams and individual athletes.”

How it works

A ski being tested in the Flexdynamics machine.

Hunger gives an example of evaluating a ski. “I establish where the contact points are when it’s in use and those will be the two bases that support the ski in the machine. At those grip points, it cannot move up and down, but you can move it freely otherwise. I then move the loading arm to the point where the ski binding would be located. With software, I set the arm to displace down from 2 to 30 millimeters. After it reaches this setting, it stops for less than 1 second, records how much force is applied and then releases. The ski will oscillate in response.”

Flexdynamics testing of a snowboard in torsion and test data in plots comparing torsional response (left) and other dynamic properties (right).

A sensor mounted on the machine records the oscillations at ≈240 samples/second until they stop. “The software will repeat the test until it reaches 10 trials within our set tolerance of ±1 millimeter for the initial displacement,” he explains. “After this, we typically test the tip and tail of the ski. We will then lift the arm and do a torsion test. We first record the angle with the contact point and the torsion force applied and then perform the same damping test with 10 repetitions, but with torsion applied. All the data is recorded and then analyzed using AI tools to filter the data, identify patterns and give objective feedback. My goal is to have the AI learn from the data.”  

Why torsion? “It changes the damping behavior of a ski or snowboard,” notes Hunger. “And our correlation studies show this is what the athlete feels. Flexdynamics testing without torsion showed very little correlation with the athlete’s assessments. But with our complete set of damping and torsion data, we can understand how a change in thickness, design or materials changes the performance of the equipment in use.”

First response, energy dissipation, complex systems

The products that Sport Dynamics Lab tests typically behave as underdamped systems — meaning they oscillate after disturbance and eventually settle. “In these systems the most informative features often come at the start of the response,” says Hunger. “The first rebound peak gives a simple rebound (overshoot) ratio — for example, if you displace a ski downward by 10 millimeters and it rebounds upward by 5 millimeters after release, that’s a 50% rebound ratio; if it rebounds 7 millimeters, that’s 70%.”

Flexdynamics moves beyond three-point bending to provide not only stiffness, but a complete set of damping and torsion data – including first response and damping coefficient – a kind of dynamic fingerprint for comparing materials, designs and systems.

“That first peak also tells how quickly the structure snaps back. Two skis can have the same static stiffness, but one rebounds faster. That ‘snap-back’ timing is closely related to what athletes describe as ‘pop’ or responsiveness — important, for example, for snowboard and skateboard maneuvers that require quick vertical lift and rotation.”

“Beyond the first peak, we also quantify how the oscillations decay and how much energy the system dissipates,” says Hunger. “In real equipment, this damping is often ‘effective’ damping — not only material damping, but also losses from interfaces, friction, assemblies and, for some products, the tire or binding system. To make it actionable, we extract metrics that reflect energy absorption and control, which are key properties in components like bicycle handlebars.”

Handlebars are not simple systems, comprising multiple tubes and other components, but bicycle wheels are much more complex. Hunger explains: “You have spoke tension, different materials in spokes versus rims, the rim cross-section, the tire, tire pressure and casing thickness — there are many interacting variables. Each supplier tries to isolate their part, but the rider experiences the system-level dynamic response.”

“Our approach is built for that reality,” he continues. “We combine controlled lab tests with telemetry and sensors on both the equipment and the athlete to create a profile — how excitation enters the system from the road or slope, and also from the athlete — so we can interpret the dynamics that matter during actual use.”

Simulation, end-to-end solution

FEA and simulation are used to validate performance predictions and speed prototype development and evaluation.

However, a key part of being able to predict and understand performance is augmenting Flexdynamics testing with modeling and simulation. “We are combining FEA simulations with testing to replicate the same loading conditions and boundary conditions we have in reality using a variety of simulation software programs. In static FEA, we’ve achieved ±3% agreement in controlled static cases. This means we are very close in what we model and measure. Manufacturers can bring three or 20 different constructions, and we run Flexdynamics tests and the corresponding FEA so that we can accurately characterize the material, supported by automated analysis tools. This enables what we call the Empirical Digital Twin Loop, where we can not only assess behavior but feed in changes to predict and validate new performance.”

Examples of insights provided:

• Performance tracking of new constructions, such as a program with the Nidecker Group to understand how to best match customers’ styles and meet their needs.
• Parametric studies to identify critical design features — e.g., 0.5 millimeter change in ski tip thickness significantly affects it damping (see graph below) — or effects of increasing top or bottom layers of sandwich constructions.
• Viscoelastic studies, including hysteresis of rubber used in tires as well as lateral response and behavior of foams, cellular architectures and new, non-Newtonian materials used in shoes.

Sport Dynamics Lab then visualizes this data in ways that athletes, teams and manufacturers can access online, including maps showing speed, athlete kinematics and the vibration and damping in the system. “They can then understand the performance of the product,” says Hunger, “but also how to improve in cornering, for example. These insights can also be linked to a calibrated virtual twin, so multiple metrics can be interpreted together. This approach enables comparisons as well, because you can see bicycle A versus bicycle B, and how each performs in different scenarios.”

“Thus, we are not just measuring but also modeling and integrating both into a development workflow,” he adds. “Our solution is transversal — it runs from end to end. I don’t want to give just answers from Flexdynamics testing, but also to cross with the statistical models and insights from the AI analysis.”

Evaluating bio-based and recycled materials

“This is a field that I love,” says Hunger. “I was sponsored by Entropy Resins during my work with surfboards in 2017.” Founded in 2010, Entropy was an early pioneer in bioepoxies. It was acquired in 2018 by Gougeon Brothers which produces West System and Pro-Set Epoxy. Hunger has also worked with Bcomp flax fiber reinforcements.

Data, modeling and AI-assisted analysis enable comparison of different materials in applications like the surfboards shown here, which helps ensure performance is maintained or even improved, for example, with new, more sustainable materials and processes.

“Many groups want to move from traditional composites to more eco-friendly and sustainable products,” says Hunger, noting that 5-6 years ago, adoption of bio-based epoxies was still limited. “Now, this has changed, which is really good, but I think the industry still needs time and data to build confidence in how these biocomposites perform in real parts. Data-driven testing, modeling and AI-assisted analysis can help accelerate that learning cycle and reduce trial and error.”

Hunger observes one issue, where brands or manufacturers want to improve their sustainability but don’t want to change the design. “We see that many new materials can be used, but the board may need to be a bit thicker or thinner to provide the same kind of flex. My approach is to stop trial and error in the field and first go to the lab. Let’s start with the ski or hockey stick you are already producing and establish a baseline of static and dynamic characteristics. Then we can play with different materials in prototypes. By the time you have downselected what you want to trial in the industry, you will also have data to show if they behave the same or differently in damping and torsion, and the testing then becomes final validation. This approach lowers the risk and makes it more economical to try these new materials.”

Performance metrics can be objectively compared for different brands and designs, like the snowboards shown here.

Future applications

Sport Dynamics Lab is working toward ISO-aligned procedures and certification as it envisions wider applications, including outside of sports. “My approach is closer to aerospace practices than the traditional approach in sports,” says Hunger, “but adjusted for smaller budgets and shorter timelines. I’m trying to use an approach that is affordable but still provides the data and actionable insights, and which is also flexible, because you can’t run multiyear R&D cycles on a bicycle wheel.”

The Flexdynamics machine has been adapted to test a variety of products, materials and behaviors.

“We are using the Flexdynamics machine to measure a wider range of products, materials and behaviors than ever. Previously, we focused on stiffness, rebound and damping metrics, but we have now created many different jigs, for example, to measure tires, both in free-response testing and under progressive load and pressure. This has opened the door to measuring a range of elastomeric materials, including foams and cellular architectures. We can also incorporate different machines.”

Possible future applications for Sport Dynamics Lab include structures that experience high vibration loads in drones and motorsports.  Source | Getty Images

“I would love to start working with drones, because their use is rapidly expanding, and we have to know what happens with propeller blades, wings and supports as they withstand all the dynamic loads and excitation in the system,” Hunger continues. “We also see applications in motorsport aeroelasticity — for example, passive elements that move primarily in response to aerodynamic loads rather than direct actuation. We can also measure this kind of behavior, but we are a small company, and so we advance step by step.”

“We now have experience with a wide range of products, geometries and materials, and we can build simulations that help us predict how the structure and system respond dynamically. We are trying to make sure the data we provide is as accurate as possible and genuinely useful — helping companies and teams make critical decisions. And we’re applying this methodology across cycling, snow sports and composites — turning feeling and assumed knowledge into measurable physics and objective data. Composites, by definition, are a system of components. And the equipment we optimize represents another scale of systems — where you must optimize not only the composites, but also the overall design. To succeed, you have to be able to orchestrate the system. And to do that, you need reliable data.”