Damage detection in high-speed rotated blades by blade tip-timing method based on compressed sensing

High-speed rotated blades are key mechanical components in turbomachinery. Understanding dynamic characteristics of rotated blades is important for damage detections in turbomachinery. Conventionally contact methods are used for this purpose. However, strain gauges and accelerometers can affect the precision of measurements. So blade tip-timing(BTT) is used to realize non-contact vibration measure and then using accurate measure results to detect damage in rotating blades. However, BTT signals are typically under-sampled. How to extract characteristic features of blade vibrations by under-sampled signals becomes a big challenge. In this paper, a novel of compressed sensing (CS) is proposed to solve the problem. Firstly, a CS mathematical model is built based on BTT method. Then, minimum one normal(L1) algorithm is applied to solve the CS problem. Finally, damage detections are realized based on reconstructed frequency information. Thus the proposed method can provide a promising way for online damage detection in rotated blades for practical engineering.