Experimental Behavior Monitoring of Cold-formed Steel Members by Feature Tracking

Abstract

Behavioral properties of cold-formed steel (CFS) members, the use of which has increased in recent years, are determined by standard tests, including axial compression and bending tests. During experimental testing, various sensors record physical changes of the test specimen, such as local displacements and strains. The local displacements are generally measured using linear variable differential transformers, commonly known as LVDTs. The LVDTs are positioned carefully on the test specimen and connected to a data acquisition device. Even though LVDTs are quite common in experimental testing, it has various disadvantages. The LVDT setup process requires particular expertise and is time-consuming. Besides, the entire data acquisition system is usually pricey. In this research, a novel image processing-based displacement measurement procedure, which is more economical and sustainable, is developed to create an alternative hands-off and easy-to-implement displacement measurement method. First, the test videos recorded from a fixed position during the axial compression tests of CFS columns have been used to extract the test member's displacements automatically. The obtained displacement tracking results are then compared with the LVDT measurements recorded during testing to verify the accuracy of the developed methodology. The general results showed that video-based image processing is a promising method for displacement recording. However, the proposed method fails to capture the small displacement values recorded at the beginning of the axial compression tests as the image resolution is not sufficient. Nevertheless, the determined image-based displacement values match with the recorded displacement values as the displacement values increase.