RAD

RESEARCH AND DESIGN

How We Test

WHAT WE MEASURE

Above roughly 10 mph, air resistance is the majority of resistive force, and at racing speed it takes ~90% of power produced. Aerodynamic drag is represented by CdA, the drag area of rider and bike system. We measure CdA for each position and piece of equipment, then convert the difference between setups into watts / timed saved at your target race speed.

THE METHOD IN BRIEF

We test on 600-1000 Meter straight flat roads, or indoor/outdoor velodromes. Testing requires a target speed and steady state power while sensors on the bike record air speed, wind angle, ground speed, power, elevation, air temperature, and pressure. CdA is solved for each run from a power balance: the power you put in, minus what goes to rolling resistance, drivetrain loss, gravity, and acceleration, is what goes to air drag. Wind, grade, and air density are accounted for in the calculation.

Typical margin of error is 0.002 m^2, or 2 watts at 45 kph. Yaw angle is binned in the analysis report, but it is not controlled. Speed is also binned in the analysis report, but special considerations must be made for speed reactive fabrics etc. Test values will typically be ridden at "tempo" power.

A suitable power meter is required, such as Quarq D-series or SRM PM9.

THE SYSTEM WE USE: STREAMLINES CIRRUS

Our field tests run on the Streamlines Cirrus system, a drag measurement system developed and built in the UK for real-world testing. Its probe sits at the bike and reads air speed and wind angle where the rider actually meets the air, along with air pressure and temperature. The probe design comes from motorsport, where it is used in Formula One. The rest of the system is aimed at the usual sources of error in outdoor testing:

Probe. Measures air speed and wind angle at the bike, so we correct for the wind you really rode through rather than an estimate.
FORMA. An infrared sensor that checks your body position is the same from run to run, so a faster result comes from the equipment and not a different tuck (streamlines normalizes CdA for changes in head and body position).
TEMPO. A precise wheel speed sensor. Timing error in ordinary speed sensors can distort the power balance, so we use one built for the job.
NIMBUS. A data logger that keeps every sensor on the same clock.
Cirrus Portal. Software that processes each run, including elevation, air speed calibration, and corrections for sail effect and Reynolds number.

A SESSION, STEP BY STEP

Brief. You tell us your target events, goals, current setup (preferably in sheets or excel).
Plan. We model your race to find the speed you will hold and the wind angles you will see, then build a suggestion testing protocol.
Setup. We calibrate sensors, zero the power meter, record the starting conditions, and confirm the test road suits the day's wind.
Runs. We work in blocks: baseline, test block, baseline, test block. You see the numbers between blocks and we decide the next change together.
Analysis. Runs are corrected for conditions. Runs that fall outside our quality limits, such as gusts or a broken position, are excluded from the report but logged.
Report. You receive the results with the data behind them.

WHY THE NUMBERS CAN BE TRUSTED

Baseline bracketing. Returning to the starting setup between blocks shows whether conditions or the rider drifted, and anchors every result to a known reference.
Repeatability first. If the same setup does not give the same number twice, the difference between two setups means nothing. We check this before comparing anything.
Consistent position. A faster result should come from the change we made, not a lucky tuck. We hold position and power steady and note any deviation.
Order of tests. We vary the order of setups so slow drift in wind or fatigue does not favour one of them.
Stated uncertainty. Every result carries its error range. [Insert your measured repeatability, for example: within X W at Y km/h.] If two setups differ by less than that, we report them as indistinguishable.

LIMITS

Field testing depends on weather. We schedule around wind and rain and rebook if conditions fail.
Wind angle is measured, not controlled. Questions that depend on behaviour across a full range of angles, such as wheel choice for a crosswind course, may need more runs or a different venue.
Small differences are hard to see in any test. We say so instead of rounding them into a winner. Small differences can be found by repeated runs.

WHAT YOU RECEIVE

CdA for every configuration tested, with uncertainty.
Watts and time saved at your race speed.
The raw run data.
Ranked recommendations for position and equipment.

back to home

rdand.net is a website of RAD Performance.

+1 [205] 902-5236
[email protected]