Damage detection by mode shape squares extracted from a passing vehicle

The success of damage detection heavily depends on the quality of damage features, such as mode shapes and power mode shapes, etc. However, these features are usually difficult and inconvenient to be accurately obtained in practice. To solve this problem, this paper develops a simple method to approximately extract structural mode shape squares from the acceleration of a passing tapping vehicle, which serves as a message carrier of the dynamic properties of the structure. Based on the approximately obtained mode shape square, a new damage index is proposed to improve the sensitivity to damage. Numerical simulations and simple experiments demonstrate the validity of the proposed method. Compared with traditional methods, it is easier to be implemented and more accurate in noisy environment, because it requires neither pre-installing many sensors on the structure, nor solving eigenvector or singular value problems and uses only the information of point impedance. In addition, the density of measurement points can be flexibly adjusted since the tapping vehicle scans the structure continuously. (C) 2011 Elsevier Ltd. All rights reserved.