Fig 1.
A subject riding the instrumented bicycle on the training rollers.
The training rollers were bolted to a force platform via four brackets attached to the rectangular frame on which the rollers were mounted.
Fig 2.
Force and moment measurements from the force platform were used to calculate the rider-bicycle system COM and COP locations in the lateral or y-direction.
The ground reaction forces (Fx, Fy, Fz) and moments (Mx, My, Mz) were measured by the force platform beneath the rollers. The bicycle rides on an effective ground surface located a distance dz above the force platform.
Fig 3.
Rider lean angle as viewed from behind the rider-bicycle.
The rider lean angle (ϕlean) quantifies how a rider is shifting his/her COM to alter the location of the rider-bicycle COM relative to the bicycle. The arrows define the positive sense of all angles. Rider lean (ϕlean) is defined as the COM roll angle (ϕCOM) minus the bicycle roll angle (ϕ).
Fig 4.
Lateral center of pressure location (yCOP) and center of mass location (yCOM) versus time.
Data from a representative trial (non-cyclist, v = 7.46 m/s) demonstrates the lateral center of mass location closely tracks the lateral center of pressure location during bicycle riding.
Fig 5.
Cross-correlation of the lateral position of the center of mass (yCOM) to the center of pressure (yCOP) versus speed.
The cross-correlation decreased significantly with increasing speed (F = 29.113, p < 0.001) and decreased significantly more with increasing speed for non-cyclists than cyclists (F = 14.843, p < 0.001). All subject data are shown; connected points indicate data from the same subject.
Fig 6.
Bicycle roll rate () and steer rate (
) versus time.
Data from a representative trial (non-cyclist, v = 7.96 m/s) demonstrates that the steer rate () lags and is correlated to the bicycle roll rate (
) during riding.
Fig 7.
Cross-correlation of steer rate () to bicycle roll rate (
) versus speed.
The cross-correlation decreases significantly with increasing speed (F = 34.307, p < 0.001) and decreases significantly more with increasing speed for cyclists than non-cyclists (F = 4.650, p = 0.035). All subject data are shown; connected points indicate data from the same subject.
Fig 8.
Standard deviation of steer rate () versus speed.
Cyclists exhibit less variation of steer rate (F = 15.121, p < 0.001) than non-cyclists. All subject data are shown; connected points indicate data from the same subject.
Fig 9.
Average positive steering power versus speed.
All riders developed less positive power to steer the bicycle as speed increased (F = 10.547, p = 0.002). Cyclists developed less positive power than non-cyclists (F = 19.213, p < 0.001). All subject data are shown; connected points indicate data from the same subject.
Fig 10.
Bicycle roll angle (ϕ) and rider lean angle (ϕlean) versus time.
Data from a representative trial (cyclist, v = 2.526 m/s) demonstrates that rider lean (ϕlean) is highly correlated, but negatively so, with bicycle roll angle (ϕ).
Fig 11.
Cross-correlation of rider lean angle (ϕlean) to bicycle roll angle (ϕ) versus speed.
The cross-correlation decreases significantly with increasing speed (F = 32.948, p < 0.001) and decreases significantly more with increasing speed for non-cyclists than cyclists (F = 17.639, p < 0.001). All subject data are shown; connected points indicate data from the same subject.
Fig 12.
Standard deviation of rider lean angle (ϕlean) versus speed.
Cyclists exhibit significantly less rider lean than non-cyclists (F = 19.643, p < 0.001). All subject data are shown; connected points indicate data from the same subject.
Fig 13.
Standard deviation of the lateral position of the center of pressure (yCOP) versus speed.
The standard deviation of the lateral position of the center of pressure decreases significantly with increasing speed (F = 25.294, p < 0.001). Although it may appear that cyclists exhibit less variation in the center of pressure position than non-cyclists, there was not a significant difference between the two groups (F = 3.695, p = 0.059). All subject data are shown; connected points indicate data from the same subject.