Research On Vibration Modal Testing Method For Thin-walled Parts Based On Machine Vision

Thin-walled parts have been widely used in various fields of industrial production, since they have a lot of advantages such as light weight, high reliability and high structure efficiency and so on. Excessive vibration will occur easily in the process of mechanical operation because the thickness of common thin-walled parts is very small compare to the overall size, the area is large and the stiffness is small, so the vibration and noise, even mechanical failure will generate, therefore the modal testing must be done for thin-walled parts to predict their dynamic performance. Additional quality will be introduced and the original dynamic characteristics will be changed since the sensors must be attached on the testing object with the traditional contact modal testing method, and the testing points are extremely limited in one testing, it is very difficult to realize full-field measurement. The modal testing method by point-by-point vibration measurement with Laser vibrometer has the disadvantages such as time-consuming, larger labor intensity and long measurement cycle and so on. Although the modal testing method by laser continuous scanning vibration measurement can make up for the deficiency of the single point laser measurement, but this method require s high surface quality for measured objects, complex optical system and motion control devices, especially the measurement results will be effected easily by external vibration. It is hard to satisfy the demand of experimental modal testing and operating m odal testing. Aiming at this condition, a modal testing method for thin-walled parts based on machine vision vibration measurement is proposed in this paper.Firstly, the basic theory of vibration measurement based on machine vision is investigated, the mathematical model of the imaged objects is established with the basic principle of industrial camera imaging, and the industrial camera model parameters calibration method is researched. The image preprocessing methods related with vision vibration measurement technology are discussed including ROI identified, image filtering and feature enhancement algorithm, and the wavelet image denoising method are studied according to the disadvantages of the common image filtering method. The feature point vibration information extraction method based on image segmentation and Blob analysis is proposed.Then, the experimental modal testing method for thin-walled beam based on machine vision is proposed, the hardware composition for the thin-walled beam modal testing system is designed, including industrial camera, optical lens, light source and modal vibration system, and the calibration of industrial came ra model parameters is realized. The vibration parameters of the thin-walled beam is obtained with the thin-walled beams vision vibration measuring experiment, the results are compared with the results of traditional acceleration sensor measurements. The accuracy of the vision measurement system proposed in this article is verified through the compare of different methods, which shows that the proposed system can realize the visual vibration measuring accurately. The thin-walled beam experiment modal test is done by frequency sweeping with modal vibration meter, the experimental modal test of thin-walled beam based on machine vision is realized, the first and second order natural frequency and the modal vibration shape of the thin-walled beam are obtained, and the results are compared with the results of Finite Element Modal Analysis. The modal testing results is verified by the comparing with different methods.Finally, the operating modal testing method of thin-walled beam based on machine vision is investigated, by using the theory of stochastic subspace modal parameter identification. The identification method of the covariance driven stochastic subspace modal parameters is discussed. Taking the t hin-walled beam as test object, the operating modal test experiment is done, and the operating modal parameters of the thin-walled beam is obtained under the condition of unknown stimulations, the result of the test is compared and verified.A machine vision vibration measurement modal testing method for thin-walled parts is proposed in this paper, the multipoint non-contact vibration modal measurement for thin-walled is realized with the proposed system. It is verified that the proposed system can effectively solve the problem of operating modal testing, also improve the efficiency of the measurement resolution and experiment greatly. And the proposed method and system is accurate and feasible by real experiment.