GPS deflection monitoring of the West Gate Bridge

Abstract

The achievable precision and relatively high sampling rates of currently available GPS receivers are well suited for monitoring the movements of long-span engineering structures where the amplitude of movements is often more than a few centimetres and the frequency of vibrations is low (below 10 Hz). However, engineering structures often offer non-ideal environments for GPS data collection due to high multipath interference and obstructions causing cycle slips in the GPS observations. Also, for many engineering structures such as bridge decks, vertical movements are more pronounced and more structurally critical than horizontal movements. Accuracy of GPS determined positions in the vertical direction is typically two to three times poorer than in the horizontal component. This paper describes the results of a GPS deflection monitoring trial on the West Gate Bridge in Melbourne, Australia. The results are compared to the estimated frequencies and movements from the design of the bridge and previous accelerometer campaigns. The frequency information derived from the GPS results is also compared to frequency data extracted from an accelerometer installed close to a GPS receiver. GPS results agree closely to the historical results and recent accelerometer trials for key modal frequencies. This indicates the suitability of GPS receivers to monitor engineering structures that exhibit smaller movements due to their stiffness and in environments not ideally suited to using GPS.