Evaluation of in-situ measurements of composite bridge with integral abutments

Sammanfattning: Bridges are an important element of the infrastructure today. The technical competence has reached high levels in most countries and the limits given to a bridge designer are set by economic restrictions rather than technical skill. The ever lasting chase for more economic solutions and procedures includes both a wish for cheaper constructions as well as for a decrease in the following maintenance costs. The roller bearings used in conventional bridges are known to generate high costs both during construction and due to maintenance, and the possibility to remove these bearings and accompanying expansion joints are therefore investigated since several years. In Sweden, the absence of bridges built with integral abutments has made it difficult to test and evaluate this solution. In 2005, an international project called Intab was launched with the explicit aim to look at different solutions possible to use if joints and bearings are excluded. One of the most common discussed problems concerning bridges built without expansion joints is the accommodation of longitudinal elongation due to temperature variations. A bridge built with integral abutments is often supported by piles made of steel or concrete. The longitudinal elongation of the superstructure induces a displacement and a moment in these piles, which in time may cause a fatigue failure. Therefore it is of big interest to look at the amplitude of these strains. Within the Intab project, a pilot object was built in Nordmaling, ~50 km south of Umeå. It is a 40 meter long and 5 meter wide composite bridge with two steel girders and a concrete deck. The steel piles in this bridge is equipped with strain gauges who are used both for longterm and shortterm measurements. The sampled data from longterm measurement is collected on a computer on site, and then downloaded through a traditional modem. The data from the shortterm measurement is sampled and collected immediately during the test. The shortterm test is performed during both summer and winter conditions, andconsists of a lorry passing the bridge while different measurements are done. The result from the measurement is also compared to calculations and a 2-D FE-model. It seems that short-term measurements performed during warmer months, May and October, show better agreement with the FE-simulations than the tests performed during the winter. The used FE-model is based on restrictions from Bro 2004. Regarding the long-term measurements, the obtained strain values are not in a range where fatigue can be considered as a severe problem.