Measurements and Monitoring the Behaviour of Crossbeam in Orthotropic Bridge Deck with 3D Digital Image Correlation

Abstract

Digital Image Correlation is a cheap and easy to use measurement method that has been proved to characterise any solid material parameters up to plastic deformation range. The digital image correlation measurement system is based on the tracking of a collective number of pixels from the surface of the reference image to the deformed image. This summation of pixels is called subset. To achieve a reliable measurement in DIC, each of the subsets must contain sufficient speckle pixels. Therefore, there is a firm distinctive intensity pattern contained in a particular subset relative to other subsets. However, with the current DIC method, the user must rely on the intuition and experience for determining these number of speckle pixels. To investigate this problem, a study of measurement and monitoring the behaviour of crossbeam in Orthotropic Bridge due to in-plane quasi-static load is conducted. The reliability of DIC method on measuring strain and deformation in an elastic zone of a steel material is also examined and compared to the other measurement technique such as strain gauge and LVDT. Also, a finite element model is developed to estimate the strain and out-of-plane deformation at the certain location of the crossbeam.
The study focuses on the specific location at the crossbeam with ‘haibach’ shape cope hole, where the strain concentration is expected to be significantly high. The applied quasi-static load is set until 250 kN, with strain gauges and LVDT are installed at the same side of the crossbeam (free edge side), and DIC is installed at the exact opposite side (inner side). Parallel with the test on the Orthotropic Deck Bridge specimen, several benchmarking test is also conducted. These tests are used to examine the problems encountered during the initial experiment.
The study shows that owing to the quasi-static in-plane load, the crossbeam encounter an out-of-plane behaviour. Therefore, to achieve an accurate measurement, a 3-D DIC measurement system is obliged to use. Strain observed into two different directions: x and y-direction, gives different results in term of accuracy. Due to the low strain limit of DIC system, the result in x-direction can be considered as unreliable. For strain measurement in y-direction, some of the results exceeding the low limit strain of the equipment, give a close match result with the strain gauges, within the acceptable amount of deviation. For out-of-plane deformation, all the measurement show a good agreement between the DIC and LVDT technique. The average speckle pattern size that is used in this research is 0.2 mm, and an average area of 38 pixels2 for a particular speckle pattern. One of the benchmark study shows that under the same condition with same expected level of displacement, an average speckle pattern area between 36 and 45 pixels2 leads to reliable 3D measurement. The finite element model that has been developed in this study also gives close strain estimation for both x and y-direction, compared to the measurement data (DIC and strain gauges).