A Noise Evaluation Based Wiener Deconvolution Method For Contact Parameters Measurement Using Ultrasound

Fei Du,Jun Hong,Baotong Li,Zeng Luo,Shaofeng Wang,Ying Xu
DOI: https://doi.org/10.1115/IMECE2012-86445
2013-01-01
Abstract:Ultrasound method is an extremely useful tool in contact parameters identification for joint surface widely used in engineering product. Wiener filter is commonly used to ultrasonic reflection map to reduce the influences of the blurring effect introduced by ultrasonic beam size on the measurement accuracy, but the results are not good enough because it is difficult to estimate the signal-to-noise ratio of reflection map. In this paper, a noise evaluation based Wiener deconvolution approach is developed to handle this problem. The noise power spectrum of reflection map is estimated through the steel-air contact area of the joint surface, and an evaluation of the ratio of noise power spectrum and signal power spectrum is obtained based on the noise variance and the experimental reflection map. To verify the efficiency of the developed approach, a sphere-plate contact interface is selected as a concrete case in which different Wiener deconvolution approaches are applied for evaluation of contact parameters. The results are compared with simulation results produced by a finite element model (FEM). The comparisons show that the contact parameters obtained by the proposed approach are more accurate than that obtained by commonly used Wiener filter approaches.
What problem does this paper attempt to address?