Based On The Frequency Response Function Curvature Of The Bolt Research Quality Nondestructive Testing Method

Anchorage technology has been widely used in all kinds of tunnel, slope and underground engineering. As anchor construction is hidden project, the damage generally inside,using the traditional method is difficult to achieve the purpose of testing, it is still a difficult problem to evaluate the quality of the construction. The destructive testing represented by drawing experiment has obvious limitations, only to realize the preliminary measurements of anchoring quality.Further research is needed as to the problems appearing in the theoretical and practical detection method high-precision and non-destructive of testing. In this paper, based on the frequency response function of indexes conducting anchor bar quality nondestructive testing is whose carried out contents are mainly as follows:(1) To estimate the frequency response function. Firstly,besed on with the frequency response function curvature and curvature ratio of two indicators are to be estimated, to identify the damage of bolt. And then three different length of intact and defect of the bolt are to simulated with ANSYS finite element software,follwed by solution to its load response, with excitation and response of analog data finally find out the frequency response function curvature and curvature ratio curve of the two,through the defect in the intact and anchor bolt in the frequency response function curvature and curvature ratio curve characteristic parameters to achieve the goal of damage detection.(2) This paper improves the traditional method of singular value decomposition denoising. the measured signal by the Hankel matrix constructed with the global optimum particle swarm optimization, the number of ranks obtained Hankel matrix,and its singular value decomposition, and then use the K-medoids clustering algorithm to determine the number of effective singular values. Finally, one of the effective part of the singular value decomposition transformation, to reconstruct the optimized ranks number reduction of the measured signal denoising. Judging by the signal-to-noise ratio index, this method can better retain the high frequency part ofsignal than wavelet decomposition and wavelet packet denoising.(3) With incentives and finally response signal of the measured data by using improved singular value denoising method based on particle swarm for denoising,then use denoised signal to get the frequency response function curvature and the curvature than two curve. Through the defect in the intact and anchor bolt in the frequency response function curvature and curvature ratio curve parameters characteristic of different damage detection and validation. End with incentive and response signal of the measured data, using the improved singular value denoising method based on particle swarm for denoising, then use after denoising signal the frequency response function curvature and the curvature than two curve, through the defect in the intact and anchor bolt in the frequency response function curvature and curvature ratio curve characteristics of different parameters to achieve damage detection and verification purposes.