Identification of damage in steel beams using frequency response changes.

Abstract

The issue of structural condition monitoring is becoming increasingly important as the Ireland’s and the world’s civil engineering infrastructure ages. An efficient monitoring process is essential as part of a modern infrastructure management system. Automated monitoring processes offer advantages over traditional, subjective techniques including repeatability, reliability and flexibility. They also offer potential for incorporation into remote monitoring systems. This paper presents a technique developed by the authors for detecting damage in steel laboratory scale beams by identification of changes in the natural frequencies of the first four vertical modes of vibration of the beams. This method was successfully employed to both locate and quantify a range of actual damage scenarios. Using twelve damage scenarios, this method successfully detected the damage location 92% of the time, while detecting the magnitude of the damage with a 75% success rate.