Damage Detection of Composite Beams via Variational Mode Decomposition of Shear-Slip Data

Damage of shear connectors in steel-concrete composite (SCC) beams affects the composite action and appears as abnormalities in the shear slip between the composite components. The shear slip at the composite interface causes nonlinearity in the global composite beam response which is an issue beyond the inherent complexity of the composite system. This paper presents a novel approach for damage detection of SCCs by variational mode decomposition (VMD) of shear slip data. Numerical and experimental studies were conducted on steel-concrete composite beams to generate noise-contaminated shear slip data from undamaged and damaged states of the structure. The VMD algorithm is employed to decompose shear slip signals into intrinsic mode functions (IMFs) and then, the center frequency of each mode is captured. The higher center frequency realized in the second mode is taken as a damage sensitive feature. Because IMFs curves are extracted from the signal splines interpolated between the average of the peaks and troughs, the change of energy in the center frequencies of the undamaged and damaged states can be defined as a damage index. Welch power spectral densities of the IMFs are calculated to further investigate changes in the center frequencies of IMFs obtained using the VMD algorithm. The empirical mode decomposition (EMD) technique is also utilized to decompose shear slip signals into IMFs for comparison purposes. The results show that the EMD is not able to detect abnormalities in the shear slip signals affected by damage because of losing information through several mode decompositions and a phenomenon termed mode mixing. However, when the VMD was set to decompose the signal into two IMFs, it was found that it is more efficient in maintaining the frequency content of the shear slip signal. According to the results, the proposed method has been proved successful in detecting damage of composite beams and can be employed as a reliable and robust technique.