Nonlinear identification of structures using ambient vibration data

•New technique is proposed for detection of nonlinearity in the structures.•New indices for detection and also the level (severity) of nonlinearity are proposed.•Validated with carefully chosen numerical and experimental investigations.•An extended version is proposed for detection of nonlinearity using single sensor data.

Most of the practical engineering structures exhibit a certain degree of nonlinearity due to nonlinear dynamic characteristics of structural joints, nonlinear boundary conditions and nonlinear material properties. In this paper, we present a new technique based on null subspace analysis to detect the presence of nonlinearity and estimate the degree of nonlinearity in structures using ambient vibration data. Effectiveness of the proposed approach is demonstrated using carefully designed numerical examples and also using the experimental data. Extended version of the technique is presented to detect the nonlinearity using a single sensor data.