No document available.
Abstract :
[en] Today’s traffic infrastructure, including its engineering structures such as bridges, is stressed not only by natural ageing and corrosion but also by fatigue. The fatigue of material is accelerated by the steadily growing traffic volume and heavier vehicles. Many bridges were built after World War 2 using the prestressed concrete construction method that emerged at that time. Some bridges are close to the end of the planned service life and show damage such as spalling, cracking and corrosion. In addition, some bridges have not yet reached the end of their planned service life and already exhibit damage. These bridges require special attention and control, knowing that this issue is highly safety and cost relevant at the same time.
Structural Health Monitoring (SHM) of bridges aims to detect and localise damage as early as possible to take countermeasures to reach at least the planned service life or even more. Therefore, control systems are needed to support the engineers in addition to the visual bridge inspection. Permanent control systems can allow real time controlling of the bridge behaviour but generate high effort and cost. One approach in SHM is damage detection based on stiffness changes. Damage can alter both the static and modal properties of a structure. It leads to a loss of stiffness and, consequently, to greater static deflection and, in dynamics, to a decrease of eigenfrequencies. A prerequisite for early damage detection is essential information about the bridge structure, best knowledge and understanding of the individual bridge behaviour already in the undamaged state to track changes. This information can be obtained with experiments on a bridge and, in parallel, by simulation with a Finite Element (FE) model. In the next step, the FE model is updated to the measurements so that the reference state of the structure is well matched. The aim of the simulation is not the ultimate load bearing analysis, but the simulation of changes in the deflection line, eigenfrequencies, mode shapes and in the best case, also in the static or dynamic flexibility matrix and even better stiffness matrix due to damage. For this purpose, recurring measurements and simulations are compared with the initial measurements. If changes in the properties occur, model updating can be used to detect, localise and quantify damage.
For the described approach, the detection and localisation of damage depend on the best possible reference state’s characteristics acquisition. The most commonly used construction method for bridges is prestressed concrete. Therefore, the main focus of this work is the recording of the undamaged reference state of a post tensioning bridge beam. The test object was a 26 m long prestressed concrete T-beam, which was saved before the demolition of the real bridge. It was subsequently installed outdoors on the campus of the University of Luxembourg as a simple supported real size test beam. Since the changes in static and dynamic system properties are not only due to damage but can also occur, for example, as a result of temperature fluctuations, the work focuses as well on the recording and assessment of influences arising from bearing and real environmental conditions in the undamaged reference state. For the temperature acquisition, the tests were carried out over around 2 years.
Moreover, the influence of bearing conditions was tested by three interchangeable movable bearing types. The reference condition was recorded by static, quasi-static and dynamic tests. Throughout the observation period, the temperature and deflection of the bridge were continuously measured at different positions. For the deflection measurements, a commercial system was used that requires contact with the bridge. In addition, two new non-contact measurement approaches were tested. One is a camera-based system and the other is a laser-based system. The laser-based measurement method was improved during the recordings and tested by a second laser based system at the beam. Through the various tests, the deflections, eigenfrequencies and mode shapes of the bridge were determined. With the information of the experimental part, an FE model was created and best fitted to the reference state. The FE model consists mainly of solid elements. For a future model updating, a special FE model was created, offering a slice-by-slice adjustment of the beam stiffness. The model was used to perform static deformation and modal analysis. Then, the static and dynamic flexibility matrix was calculated and compared based on the experimental and numerical results. Finally, and in view of the subsequent artificial damage of the beam, damage scenarios are proposed based on the calculated cracking moment.
