The four manufacturers in this test all make claims about the comfort of their bikes, so we set out to measure just how well they handled vibration. In collaboration with Microbac Laboratories, we created a new test to measure the bikes’ responses to bumps.
We wanted to see how the bicycles reacted to a rough chip-seal road as well as a larger rumblestrip-like surface. To accomplish this, we purchased two sets of Kreitler rollers. We used 4.5-inch rollers on the rear and a 3-inch roller on the front of both sets (to keep the bumps on the rollers out of sync). We then welded a 1/8-inch bump along the width of the front two drums on one set of rollers for our small bump test. The second set received 1/4-inch bumps to simulate the rougher road.
Accelerometers were attached to the bicycle at the front and rear dropouts, the top of the steerer tube and the seatpost just under the saddle. These measured acceleration in Gs (G is acceleration measured in units of gravity. For example, 2 G = 2 times the acceleration of gravity). The lower the g number, the better the bicycle was isolating the rider from the vibration of the rollers.
We normalized accelerometer position, bicycle position, rider, rider position, tire pressure, speed and gear used. We intentionally kept the stem and handlebar out of our test. Bars and stems are often changed, and size and material have a big impact (see our May issue Tech Report). Instead, we focused on how the frame, fork, wheels and seatpost performed as a system.
Our ride testing was done blind, meaning that no lab data was divulged to testers until the bikes were returned. This is vital in maintaining the validity of the not only the lab data and its interpretation (by Microbac), but also in order to avoid coloring our impressions of the bikes in the real world.
By testing each bike eight times, we collected and processed over 2.8 million data points over the course of the vibration test. We distilled the data down to the graphs here. —NICK LEGAN