Aero Analysis
Testing your ride configurations allows you to establish the following metrics:
- CdA - this defines how much drag your configuration has. Technically, CdA is your drag coefficient multiplied by your frontal area. Frontal area is how much of you and the bike is pushing through the wind, drag coefficient is how 'slippery' you are. So a skin suit has a lower drag coefficient than a flappy rain jacket for example. You don't need to know the details of how it is calculated or used, but know that lower CdA values mean a more aerodynamic setup. CdA can vary depending on the angle of the apparent wind (the perceived wind direction taking into account the direction and speed of the actual wind, and the direction and speed of the bike). The angle of apparent wind is known as the 'Yaw' angle. You can model different CdA values up to Yaw angles of 20° with BikeMetrix, which is in line with most optimised aero wheelset designs.
- Crr - this models the rolling resistance of the tyres on the bike. This is the energy that is lost as the tyres move along the road. Every small imperfection in the road surface causes a loss of forward momentum. Tyres with more supple casings tend to deform better over imperfections and have lower losses. Wider tyres at lower pressures tend to have a lower Crr than narrow tyres at higher pressures, assuming that the tyre casing is supple. Non supple tyres, such as touring tyres with high puncture protection, tend to have much higher rolling resistance.
- Efficiency - the drive train efficiency of your bike. If you measure power at the drive wheel hub, you don't need to worry about this as all drive train losses are upstream of your power meter.
BikeMetrix supports 6 different aero analysis modes.
Virtual Time using GPS and power sensor data
BikeMetrix plots the actual ride from your loaded file, and then overlays a predicted plot that uses the GPS distance / elevation and power meter data from your ride and predicts the time to complete each data point in your ride. In other words, elevation and power are fixed, and time is predicted.
Virtual Elevation using GPS and power sensor data
BikeMetrix plots the actual ride from your loaded file, and then overlays a predicted plot that uses the GPS distance / speed and power meter data from your ride and predicts the elevation change for each data point. In other words, speed and power are fixed, and elevation is predicted.
Coast Down using GPS data
BikeMetrix plots the actual ride from your loaded file, and then overlays a predicted plot assuming no input power, using GPS data.
Virtual Elevation using speed and power sensor data
This is the most accurate mode. The procedure is the same as for Virtual Elevation using GPS above, but uses a speed sensor for distance rather than GPS.
Coast Down using speed sensor data
Another flavour of Virtual Elevation testing assuming no input power, using a speed sensor rather than GPS for greater accuracy.
Tabular lap analysis
A generic report on all laps in the ride showing different metrics and averages. Useful for quick comparisons, for example if you are doing repeated runs on a segment and want to quickly test something out without the more stringent test conditions needed for other modes.
In addition, BikeMetrix can model the effects of wind during analaysis, and enables you to slice and dice your ride files to remove parts where the test conditions were affected by things such as braking, uneven wind, poor road surface and so forth. You can very quickly chop out the parts you don't want.
BikeMetrix uses laps to mark the boundaries of different test segments. So for example, you can test on a half pipe course, going out and back multiple times, and at the start and end of each test run you hit the lap button. At each end of the half pipe, you can use your brakes or even stop your bike completely because when you hit the lap button at the start of the next segment, you know that the 'turnaround' lap can be removed from analysis.
This means you don't need to find a perfect loop or half pipe that you can ride without braking to test on. As long as you know which parts of your ride were the test portions where you didn't use your brakes, you can chop out the noise during analysis.
You can flip between analysis modes with a click of the mouse.
Removing Bad Data
In GPS modes, BikeMetrix can be set to automatically detect and compensate for anomalies that exceed a configurable threshold of error, allowing you to cope with braking, gusts of wind and poor road surfaces. This makes it very simple to get good metrics even when conditions are not ideal. Anomaly Detection can be turned on or off via the preferences screen.
Which method should I use?
If you don't have a power meter, using one of the coast down methods can get you accurate figures for CdA and Crr, but not for drivetrain efficiency. Some recumbents lose some power due to boom flex and longer chainlines that results in a lower overall efficiency, so you should bear this in mind if using coast down analysis.
Virtual Time testing is great for longer routes where you can minimise the effects of GPS innacuracies by using base map data and corrected elevation from your cycling app such as Strava. It enables you to get a very good 'real world' set of metrics for a given configuration, where all the road condition variations and GPS innacuracies are smoothed out by the length of the test segment. This gives you good metrics for race planning where you have a similar level of accuracy in your input data file for the plan.
Virtual Elevation testing using power and speed data is best for doing very detailed testing where you want to test small differences between different setups. It will give you a more accurate overall set of metrics, but some of the detail may be lost in the noise if you try to apply these figures to longer routes when planning.
In summary - coast down testing is good if you don't have a power meter. Virtual Time and Virtual Elevation testing complement each other - use Virtual Time for real world numbers over longer distances and Virtual Elevation for more accurate numbers over shorter distances in a more controlled testing environment.
