Development Diaries: REAP Code 2026

When developing a new bike, as a designer there are so many emotions flying around. There’s the inspiration that urges you to put, in my case, a well-sharpened pencil to paper; closely followed by the confusion that stems from having to translate a 2D sketch into a 3D object on a CAD screen; and the frustration from not quite being able to visualise how to get that bit between the seattube and toptube to sit quite right.

Anyway, it’s a rollercoaster to say the least. The lows are bearable with the high of the moment you present it to the world and get the reaction you believe it deserves. In this case, the first pictures of the Code have >450,000 views and counting on Instagram; that's a success by our standards.

But, there is no more nervous an occasion than taking a first prototype to the wind tunnel. That first trip to the tunnel is when you have to face the fact that your first design might not be any good aerodynamically despite your best efforts. It could just look visually stunning and be as aero as a brick. Spoiler alert, it’s not, but it’s a possibility you have to face.

In this case, we took the Code prototype to the Silverstone Sports Engineering Hub wind tunnel - a place we’ve become fairly well acquainted with over the years. We didn’t just take our bike, we needed something to test against. I don’t want REAP to be one of those brands that just makes a vague comparison of their new bike versus the outgoing model, so we took a Factor Ostro VAM with us after it topped Cycling News's testing.

In real terms, for the people reading this, the comparison you want isn’t what the new bike is like compared to the old, it’s what the new bike is like compared to the other new bike from brand X or Y that you might consider buying. So that’s what we try to do.

The Factor Ostro VAM is a good bike. It has tested very, very well in the CyclingNews wind tunnel tests. However, the Ostro isn’t the perfect comparison for the Code. It’s what the industry calls an all-rounder. It’s balancing aero and weight and not committing to being a 6.8kg race machine, nor is it trying to be what their new prototype aero bike looks to be chasing.

The Code is slightly different. It is committing to a 6.8kg weight; with pedals and cages, everything you can have on for a UCI weigh-in. What we are trying to achieve with Code is the perfect road bike. A bike balanced in weight, aero and ride quality. A bike that doesn’t necessitate racing, but would dominate a peloton whenever required. The UCI conveniently sets a weight limit for this meaning we only have to optimise for the two other metrics; a slightly easier proposition. I won’t get into too much detail on this process here, you’re here for aero data, and that is what you’ll get.

We tested both with and without rider, performed a speed sweep from 30 to 45 kmph at 5 kmph increments and performed a yaw sweep at 40 kmph at -10, -5, -2.5, 0, 2.5, 5 and 10 degrees. We closed out our test with a repeat of the first run to ensure accuracy across the session. We tested with the same wheels and chainset in both bikes and matched position as accurately as possible.

If this is a lightweight bike, presumably it’ll be used for climbing, so why don’t we test at lower speeds? Two reasons. Firstly, aero has less of an impact at lower speeds so the differences are smaller and therefore more difficult to quantify. And secondly, the wind tunnel accuracy is significantly worse at lower speeds. It involves reasoning above my pay grade so I couldn’t explain the ins and outs even if I wanted to, but just know that the higher speed the test, the more accurate the results, generally speaking.

To try my best to paint a picture, you test one bike and the results sit on a screen in the control room meaning very little until you test the next bike. You then load in the next bike and wait for the comparison to pop up in real time as the tunnel completes each ‘run’ of the test. So you can see in real time how good or bad the numbers are; there is an appropriate amount of drama included in the situation.

So, the first tests were without the rider and the results were promising. At 30 kmph we started at 1.2w behind the Ostro, and at 45 kmph were 4.2w behind. Considering the Ostro is a finished bike that took many CFD simulations and many hours in the wind tunnel to reach this state, and likely more than a year from beginning to end; not a bad start at all for a bike at the beginning of its development programme.

With a rider was a little more confusing. Now is the time that I will make the point that the Code prototype is a 3D-printed nylon model with some carbon reinforcement so is not as solid as a carbon bike. What we observed was a significant amount of swaying and bouncing at a couple of points on the bike; handlebars, saddle, BB & chainstays. All of this movement serves to create more disturbance in the airflow around the bike and thus likely has a negative impact on the bikes measured aerodynamics.

Nevertheless, the results were very promising still. We believe that the movement of the rider and bike impacted the 35 and 40 kmph runs as the trends don’t appear to follow logic, however we got stable and logical results at both extremes of the speed sweep. At 30 kmph we were just 2.1w behind the Ostro and at 45 kmph, 2.7w behind.

We are not advertising these results as final factual data due to the limitations of the printed model with the rider onboard, however they serve as a great start position for the Code’s development programme. We will say that the gains we see during development are significantly greater than the current deficit observed.

From the testing, we know that even when targeting a lighter form factor, the Code is in the ballpark of challenging potentially the fastest of the current world tour aero race bikes.

We had only really had a few days to observe the prototype in its 3D form before the test and since then have been able to delve into the areas that may present the lowest hanging fruit for aero gains. This is why it’s really important to go through this process rather than relying on CFD alone; not only do you get an understanding of what the bike looks like when not on a screen, you gain the knowledge of the areas that can easily be changed or the limitations for change as you move forward through development.

The big area for development is the head tube; we haven't nailed it yet, in real life it’s far too wide. Our initial couple of CFD simulations confirm this and that will be the first target for gains. It presents an interesting problem though...

• You are limited in width by the headset bearings and the steerer tube.
• If you go narrower on those, you reduce handling stiffness and negatively impact the feel of the bike.
• If you stay wider, you negatively affect aero.

So, how do you get around it? The answer does not lie in a ‘speed sniffer’ front end, that's for sure. I’m certain we can come up with something that is faster and less visually-challenging than that.

Now is time for us to spend hours and hours staring at pressure maps and rainbow-coloured air velocity magnitude graphics, rather like the one below. Wish me luck with my eyesight afterwards.