Numerical Simulation Of Ultrasonic Detection Signal In Concrete Beams And Time-frequency Analysis

Abstract:The conventional concrete structure is multiphase and non-uniform viscoelastic medium. The complex acoustic interface in structure makes the ultrasonic absorption, scattering attenuation and poor directive property in the process of propagation. Besides, the primary wave and multiple wave are often occur superimposed in signal received by transducer. So, the received reflected signal is distorted when using ultrasonic pulse echo method to defect detection. Therefore, regular qualitative discriminate the sound step, amplitude, frequency different in detection signal to describe the internal defects of the beam presence a greatly error.Numerical simulation method can accurate reflect wave field propagation mechanism and snapshot characteristics. Acoustic wave equation of order staggered grid finite difference method being researched in this paper, then, using this method to simulate three kinds of common concrete beam specimen models, which are concrete beam contains not dense defect, whether grouting fullness in duct of prestressed beam. In this paper, it has analyzed the effect of reinforced and random aggregate concrete. The kinematic and dynamic characteristics of direct wave, defect reflected wave, transmission wave and multiple wave can be clearly observed in forward modeling results, which is help to understand the wave field propagation law in concrete beam. It can be also provided theoretical and technical support for ultrasonic pulse echo detection data processing and interpretation in concrete beam.Because of the fourier transform can only describe the overall characteristics of the signal, the local characteristics of the signal can not be expressed in time-frequency domain. The paper explore the application of time-frequency analysis in ultrasonic pulse echo detection signal. Generalized S transform has been used to equivalent Q value, and then do absorption compensation for test data, so that the high-frequency signal attenuation signal energy has been greatly improved, the result shows it can improve the identification of weak valid signal. The curvelet transform has a multi-scale and multi-directional characteristic, so the paper used this method denoising test data. Hard and soft threshold compromise threshold processing method has been used in this paper, using peak signal-to-noise ratio to evaluate denoising quality. Numerical simulation signal and measured signal processing results show that:the curvelet transform can be more better to remove interference noise and fidelity valid signal.