A Super-Harmonic Feature Based Updating Method for Crack Identification in Rotors Using a Kriging Surrogate Model

Abstract

Dynamic model updating based on finite element method (FEM) has been widely investigated for structural damage identification, especially for static structures. Despite the substantial advances in this method, the key issue still needs to be addressed to boost its efficiency in practical applications. This paper introduces the updating idea into crack identification for rotating rotors, which has been rarely addressed in the literature. To address the problem, a novel Kriging surrogate model-based FEM updating method is proposed for the breathing crack identification of rotors by using the super-harmonic nonlinear characteristics. In this method, the breathing crack induced nonlinear characteristics from two locations of the rotors are harnessed instead of the traditional linear damage features for more sensitive and accurate breathing crack identification. Moreover, a FEM of a two-disc rotor-bearing system with a response-dependent breathing crack is established, which is partly validated by experiments. In addition, the associated breathing crack induced nonlinear characteristics are investigated and used to construct the objective function of Kriging surrogate model. Finally, the feasibility and the effectiveness of the proposed method are verified by numerical experiments with Gaussian white noise contamination. Results demonstrate that the proposed method is effective, accurate, and robust for breathing crack identification in rotors and is promising for practical engineering applications.

Nonlinear Vibration Analysis of a Beam with a Breathing Crack

“…Equation (13) shows that the stiffness matrix K t is time-variant due to the repetitive crack opening and closing. Thus, Equation (13) can be regarded as a linear time-varying equation.…”

“…Due to the various types of crack in different structures, currently, industries have a number of different damage-detection methods, such as ultrasonic methods [8][9][10], X-ray methods and vibration-based damage identification methods. Owing to the advancement of reliable fault detection and diagnosis techniques, vibration-based damage identification methods have been developed for various engineering applications, such as the rotating blades of engines [11], submarine pipes [12], rotating rotors [13] and atomic force microscope (AFM) probes [14].…”

Crack localization in stepped rotors based on Bayesian fusion of multiscale superharmonic characteristic deflection shapes

“…The finite element model of the rotor is depicted in Figure 2. The first natural bending frequency of the rotor without crack and step is 42.1 Hz, which is experimentally verified by a modal experiment 15 . As the crack breathes with time and its breathing behaviour is dependent on the response, which makes the stiffness of the system time‐varying and state‐dependent non‐linear.…”

“…Among these investigations, vibration‐based methods are widely investigated and recognized as useful tools. PHM for cracked rotors can be categorized into four levels 3 as crack detection, 4–10 crack localization, 11–13 crack identification 14–17 and remaining useful life prediction 18,19 . These four stages can provide information required when making a reasonable maintenance strategy, which is of great significance to guarantee the safe operation and reduce maintenance cost of key rotating machines.…”

A New Concept of Instantaneous Whirling Speed for Cracked Rotor’s Axis Orbit

“…The vibration analysis provides optimal indicators for crack detection [2] and harmonic component extraction obtains more fault information [3,4]. In addition, model updating method [5], modal state observer method [6], and modal force identification method [7] for identifying cracks have also been deeply studied. Meanwhile, the feature extraction method for acoustic signals [8], the classification methods for motor faults [9], and the sensor fusion approach for acoustic, electric, and vibration signals [10] can provide important reference for rotor crack diagnosis.…”