Ride with an anemometer

A friend is currently testing intensively the CdaCrr app with the Weather Meter anemometer mounted on the bike. The air speed is now recorded every second, so the rider average CdA is more accurately computed. 

On the smartphone screen, two figures now appear on the top left, with first the real CdA (wind is used in the model), and the apparent CdA (model with no wind). The two are of interest, with the first one you can see the influence of your position changes during the ride, even with unsteady wind. By the second value, we can learn how drafting is important during a group ride, or how tail or headwind affects quantitatively the air resistance measured by the apparent CdA figure.

2 hours ride. Segments where CdA is averaged are indicated by the red marks.

During his last ride (between 11am and 1pm), the nearest airport station has reported a wind speed of 14.8 km/h and an east direction (so between ~80 and 100 degrees from the north). Meteorological station are at a standard height of 10 meters, so wind speed should be translated to a value near the ground (there is a velocity profile created by shear) with the Hellman estimation. A rule of thumb is to divide the value by two (or even three, according to this post). So, we guess a wind speed near the ground of 5-8 km/h.

CdaCrr app has recorded the following data: axial wind speed (subtracted bike sensor speed from anemometer air speed) and direction of travel (given by the phone GPS):

We can notice the sinusoidal tendency of axial wind speed with direction. Indeed, the axial wind speed seen by the rider is given by:

AxialWindSpeed = WindSpeed * cos(BikeDirection-WindDirection)

By fitting the data with a sinusoidal function (red curve below), we find out the prevailing wind direction 84°, east as indicated by the weather station. 

The fit also gives us the sinusoidal amplitude, WindSpeed in the previous formula, as 3.9 km/h with a wind in the previous graph between -10 and 10 km/h. Compared to the 5-8 km/h of the meteorological station, there is no surprise here.

The idea of using GPS data to check wind direction was read in this great article.

Monitoring apparent CdA during a race

I raced sunday on a flat course during 2 hours with the CdACrr app on my bike (a criterium with 30 turns). It was the first time I recorded data during a race, and the post analysis is interesting. The first 5 laps were not recorded (I forgot to push the "start run" button), but we can seen clearly on each lap the influence of drafting. I spent 8 laps (lap 4 to 11) alone in front of the race, with the peloton just a dozen of seconds behind me. The CdA was stable and between 0.30 and 0.31. Before and after my escape the apparent CdA was oscillating between 0.24 and 0.27 (I was generally in the first five positions). I think I spent the whole 14th lap (CdA~0.215 in the middle of the peloton) and the lap 13th in the first position, hands on the hoods as nobody wanted to ride at this moment of the race. The second picture is the average power per lap (300-310W during laps 4 to 10) and 320W during the last lap where I chased behind 5 riders in the final. I remenber also that during the 23th lap, I was always in the second/third positions of the peloton (CdA~0.225).

Velodrome test

Field testing isn't easy. 

Lesson one: one turn (250 meters) per lap is enough to compute a CdA value. Adding more turns for a lap do not reduce variability.

Lesson two: the first turns should be systematically discarded.

Lesson three: keeping a constant position on the bike should be the goal number one.

1.0 Release

The CdaCrr app has been released this week on the Google Store.

For the moment, it has only been tested on Samsung S4/S5 and Sony Xperia Active phones and it records correctly from a PowerTap powermeter.

It implements Virtual Elevation method (aka: Chung method) to measure CdA (aerodynamic coefficient of drag) and Crr (rolling resistance coefficient) when riding a bike.