Hashem Jahangir - Assistant Professor, Department of Civil Engineering, University of Birjand, Birjand, Iran
Hamed Hasani - .Sc in Structural Engineering, Department of Civil Engineering, University of Birjand, Birjand, Iran
Seyed Mojtaba Mosavi Nezhad - Assistant Professor, Department of Civil Engineering, University of Birjand, Birjand, Iran
Fatemeh Shabazi Aval - B.Sc student of Civil Engineering, Department of Civil Engineering, University of Birjand, Birjand, Iran

Identifying the initial cracks and damages in reinforced concrete (RC) beams is an essential concern among researchers. Mode shapes are one of the most regular inputs for various known damage detection indices. The mode shapes are noise contaminated in most of practical situation. These noises cause inaccurate damage localization and should be eliminated from the obtained data. In this paper, a two-step algorithm is presented for detecting multiple damage scenarios in RC beams. Obtaining mode shapes and denoising them as input signals is the first step. To do so, a wavelet transform is applied to noisy data. In the second step, the locations of the damages were determined using an damage index based on the wavelet transform of smoothed mode shapes. The results shown that the proposed method had a proper capability to identify multiple-damages locations even when the mode shapes are noise contaminated. The proposed damage index needs just some limited number of mode shapes which is in agreement with practical situations.

Damage identification, Noise elimination, Wavelet transform