Refined detection technique for bridge frequencies using rocking motion of single-axle moving vehicle.

• Single-axle test vehicle's rocking motion is firstly used to scan bridge frequencies. • New formulas are derived for calculating wheels' contact point (CP) responses. • Wheels' CP responses are experimentally verified to be free of the masking by vehicle's frequencies. • The present model enables us to look into the mechanism of the wheel-bridge contact for uneven surface roughness. • Temporary parking of the vehicle on bridge can help sharpen the frequencies identified. A self-designed single-axle vehicle is used to detect the bridge frequencies from its rocking motion, caused by wheels moving over an uneven surface. To include both the rocking and vertical motions, the single-axle vehicle is modeled as a two degree-of-freedom (DOF) system, unlike the previous single-DOF model. The rocking frequency, an inborn property of the vehicle, may pollute the frequency extraction of bridges in the test. The sensors mounted near the wheels (called the wheel sensors) for catching the rocking are of higher fidelity than the central sensor for the vertical motion. Both the vertical and rocking frequencies of the vehicle will be eliminated in the newly derived responses of the wheel-bridge contact points (CPs). In the flat road test, the vertical frequency of the test vehicle is identified from the average of the two wheel sensors responses, and rocking frequency from the angular response calculated. From the field test conducted on a two-span girder bridge, it is concluded that: (1) the new formulas for the CPs are reliable for scanning the bridge frequencies, while removing vehicle's frequencies; (2) the present model enables us to look into the mechanism of the wheel-bridge contact for uneven surface roughness; and (3) temporary stop of the test vehicle on the bridge can help sharpen the bridge frequencies identified. [ABSTRACT FROM AUTHOR]