Virtual Wind Tunnel with Andrew Buckrell and Michael Liberzon | EP#47
The Virtual Wind Tunnel (VWT) is an exciting new innovation by Canadian startup STAC Performance, co-founded by Andrew Buckrell. It allows triathletes to get a complete, accurate aerodynamic analysis done in an affordable and convenient way, just by taking a few iPad scans and sending them to STAC for computational analysis. Joining Andrew and myself in the conversation is VWT-provider Michael Liberzon, who discusses the importance of combining aerodynamics with a good bike fit.
In this Episode you'll learn about:
- How the Virtual Wind Tunnel (VWT) works, and what happens from the second you step into a bike shop for an analysis to when you have your full report in your hands
- How you can benefit from using the VWT by comparing different positions, equipment, and even comparing your aerodynamic drag with that of other riders in STAC's database
- The importance of combining aerodynamics with a great bike fit
- How accurate the Virtual Wind Tunnel is, how it compares to actual wind tunnels, and potential limitations
- Case studies of pros and amateurs gaining big aerodynamic advantages from using the VWT
About Andrew Buckrell
- I am currently still a PhD student but my interest in aerodynamics and bikes and training is what really got me to where I am right now.
- Through experimentation and working with a business partner, we got this idea about measuring aerodynamics for cyclists virtually, because wind tunnels are often inaccessible and it's difficult for a lot of people to be able to afford them.
About Michael Liberzon
- My triathlon career started around 2013 as an athlete.
- My background is in engineering. I’m a mechanical engineer so fluid dynamics and aerodynamics are not new to me.
- When I started coaching in 2014, my approach to the sport was always scientific.
- I’m always on the lookout for new and useful technologies and ideas to make my athletes perform better.
- Aerodynamics is definitely one of those things that is very important to us on the bike because you can get free speed by this without getting fitter.
How does the Virtual Wind Tunnel work?
- We deal with completely digital models in the whole process, which we call computational fluid dynamics (CFD).
- CFD is used to numerically design something and test the fluid flow with it (including air flow).
- With CFD, it gives you the advantage of being to look at the fluid flow in any place.
- We start off the process with scanning. This is the critical part. So we want to capture a good physical model of someone. We have a scanner that is iPad based. It only takes 2-3 minutes per scan.
- During this process, the athlete themselves will be on their bike in their race position holding a static position. The person doing the scanning will walk around and will try to capture all the different sight lines and angles so we can reconstruct this model in 3 dimensions.
- Once we have the digital model we use a computer software that we developed and we reconstruct it. The raw data from the scanner is just depth.
- Then we get some output, including the CdA and a representation of what the fluid flow will look like. So we’re able to simulate non-steady flows over a reasonable time period to see how the flow changes, how you would expect things to move around.
- This gives us a really good idea of where drag may be coming from and it allows you to do a lot of optimization for the cyclist.
When a triathlete comes to you to go through the Virtual Wind Tunnel process, what is it that you actually do, how long does it take, and what does the client get?
- There are 2 types of clients that comes to us.
- The first one just wants a number, typically for modeling a race using software like Best Bike Split, in which you need your coefficient of drag (CdA). This is just a matter of doing a single scan and running an analysis on it.
- Most people look for a fit and some optimization. Then what we do depends on how good their current fit is.
- If it’s somebody looking for optimization, what we do is look at their current fit and see where the opportunities are and potentially make it better and make it more aerodynamic.
- Optimization will take about 2-4 scans depending on how many positions we want to test. The modification of the position, as long as we're not making major changes, won’t take very much time, just between 10-15 minutes.
- So the whole process might take an hour.
- The full service including a full fit would take about 90 minutes to 2.5 hours because before we even get into the VWT scanning, we would have to get the athlete into a position that is sustainable where ergonomics and power generation are key.
- At the end of the fit process, we would go through the 2-4 scans to really fine tune the fit.
When do the results come in?
- One of the challenges is the turnaround time. Computational work in general takes a long time. Increasing the output of the simulation is something that we’re working on.
- The advantage of a wind tunnel is you can change something on the fly and see exactly what the benefit is. But with the VWT you have to use a little bit of intuition and experience to really tweak it. You also do have the benefit of seeing directly what happens in terms of the air flow.
- Once we get to the 2-4 positions that we think may work, we would take detailed notes on whether it would be stack and reach or just stem length position, base area position, helmet model, and the hydration options that we use in each of the scans.
- Once we have the CdA numbers and the report back from STAC, then the client would come back in for the final fit.
Can you talk about the 3D-viewer on your website where you can compare yourself to different riders who have gone through the VWT process?
- This is something that is currently under development and we’re really excited about the possibilities with it.
- What we’re trying to do every time we work with a professional is to collect those models and collect their position, how it changes and determine how certain people are so fast. And that allows you to compare yourself to a professional.
- It is an interesting way to crowdsource aerodynamic knowledge.
- There’s a couple of different ways that we can look at this in the future. One of them is collecting a public source of data, like a general library that everyone can access. From a coach’s standpoint, what we’re able to do is put together a private area where you can compare all of your athletes together and you can compare different scans over time and have that repository of all that information.
How much can you do in terms of aerodynamics with different athletes, bearing in mind the power generation and sustainability requirements of their bike fit?
- It’s incredibly different. The ability to anteriorly rotate your pelvis is key. So that’s to rotate it forward which allows for a flatter and typically lower back. And that is something that even some very young and fit athletes cannot do because of structural issues, typically in their hips and pelvis.
- And so these athletes will tend to never be able achieve as aerodynamic a position as others. It can be optimized but it may not be as fast as someone who is able to rotate their hips more forward.
- In terms of sheer numbers, you can get really low CdA values in the 0.22s for the really fast guys. And someone who has lower back issues and can’t rotate their hips forward as much, their CDA value even on a TT bike might be around 0.3.
- In terms of speed differences, it depends on how powerful these athletes are. But the differences are dramatic. Easily 2-3 km/h for 200 watts.
What have been some typical results that you have seen in how much the CdA improved after VWT optimisation?
- The typical spread between worst scan and best scan is about 0.02. This is not insignificant. Over the course of an Ironman or even a sprint, that is definitely a significant amount of time. More time than an expensive set of wheels might get.
- With certain individuals, a lower position is not better for their CdA. There comes a point, especially in female riders or even some males where a lower front end is actually slower than a higher front end.
How much does it cost to get a scan with the VWT and what is its availability?
- Currently we’re pricing them at about $100 per scan. This gets you some minor changes to the bike if that needs to happen but not a full fit, the scan and the computational fluid dynamics work, and a CdA number. This is just one position.
- We’re in the early stages of negotiation with some people of setting up remote scanning locations. There’s someone in Hong Kong, Switzerland, Sweden, Netherlands, and a lot in Europe and all over the world.
- The amount of equipment and training needed is not that significant. Most people have an iPad. The scanner itself, which is an attachment to the iPad, is about $350. It’s not a huge investment. It’s something that's within the reach of a triathlon club.
- So potentially, a club can purchase this and then they can organize and train someone in the scanning. Then they can send us the different scans and work with us that way.
What is the accuracy of the VWT compared to actual wind tunnels?
- As a general comparison, working with Cody Beals, we were within 2-3% of the wind tunnel tests.
- Recently, I worked with one of the Slowtwitch editors and I was able to match up the predictions for the different helmets that he had wind tunnel results for.
What is the cost of getting a test done in an actual Wind Tunnel?
- It varies in different wind tunnels. In Ontario there’s one for 800-1000 Canadian dollars an hour.
- The Faster wind tunnel in Scottsdale, Arizona, from my memory is around USD 450-550 an hour plus you have to factor in your travel expenses.
What are the limitations of the VWT?
- One of the biggest limitations is just the turnaround time. This is something that we are working on improving. Currently it takes 2-3 days to get the results.
- In the ideal scenario, we’re hoping to get this down to 24 hours by taking advantage of cloud computing.
- But this doesn’t have instant feedback or near instant feedback that an actual wind tunnel has.
- But the side benefit is that you get a lot more detail into the flow.
- Another downside is the result is only as accurate as the scanning process. This is something that we’re also working on improving. But there are certain features that won’t get picked up by scanning. For example, spokes and cables.
What is the future of STAC Performance?
- There’s a few aerodynamic devices that we’re developing. A lot of those have been with the help of Cody and his input.
- Another exciting thing that we’re working on is a variable resistance version of our trainer which is completely unrelated to the wind tunnel.
- We also want to get into how we can extract more performance from people whether it’s aerodynamics or other methods just to help people perform at their best.
Rapid fire questions
Favorite book, blog, or resource related to triathlon:
- Michael: How Bad Do You Want It by Matt Fitzgerald
- Andrew: Training and Racing with a Power Meter by Hunter Allen and Andrew Coggan
Favorite piece of gear or equipment:
- Michael: Power meter
- Andrew: STAC Zero Trainer
Personal habits that contributed to your success:
- Michael: Consistency
- Andrew: Competitive spirit
- Episode 11: Master Ironman 70.3 training with pro triathlete Cody Beals
- Episode 25: Aerodynamics mastery and free speed on the bike with Nuno Prazeres
- Episode 31: How to shave 14 minutes off your Ironman bike split with Jon and Chris Thornham