Research On Acoustic Online Monitoring Technology Of Laser Paint Removal

Laser paint removal is an important application of laser cleaning,but the lack of online monitoring of the process severely restricts the scale of laser paint removal.In the laser cleaning process,sound waves are generated due to the photoacoustic effect.This sound wave will change with the change of material properties and laser energy,and the cleaning process can be monitored online.In order to realize the online monitoring of the laser paint removal process,this paper builds an experimental platform for experimental research.Research has found that the time domain and frequency domain information of the cleaning sound signal can reflect process changes.Extracting characteristic parameters from the signal and establishing a discriminant model of characteristic parameters based on Bayesian discriminant method can provide an effective online monitoring method.The main work of this paper is as follows:(1)This article explains the mechanism of the solid-state photoacoustic effect.The photoacoustic signal is affected by the optical and thermal properties of the material and the laser energy under the two mechanisms.Based on the photoacoustic signal changes during the laser paint removal process,an on-line acoustic monitoring experimental platform was built,and an acoustic signal acquisition system based on Lab VIEW was developed.The system can collect the acoustic signals in the laser paint removal process in real time,intuitively and accurately.(2)The acquisition and analysis methods of the paint removal sound signal and the discrimination of the paint removal process are studied through experiments.When the laser frequency is low,the paint removal sound signal is affected by noise,which is difficult to determine quantitatively.After improving the parameters,the cleaning quality standards were formulated based on the surface topography,and the effects of laser cleaning times,energy density,and pulse repetition frequency on the cleaning sound signal were studied through time-frequency analysis.Extract cleaning features from the signal.Establishing a discriminant model based on Bayesian discriminant method can realize quantitative discrimination of the cleaning process.(3)For practical applications,the method of acoustic on-line monitoring of laser continuous paint removal process is studied.Under the conditions of a cleaning frequency of 10 k Hz,a pulse width of 200 ns,and a spot overlap rate of 90%,the cleaning quality standard is established by the damage threshold.Combining surface morphology,the cleaning process is divided into three categories: the paint is being removed,the paint removal is completed and the substrate is not damaged,and the substrate is damaged.Based on this standard,four cleaning characteristic parameters of maximum frequency component,signal energy,spectral centroid frequency and amplitude are extracted from the signal spectrum.The laser power density,cleaning frequency and characteristic parameters are used as categorical variables to establish a Bayesian discriminant function,which can quantitatively discriminate the paint removal process.After verification,it was found that the accuracy of training samples reached 99%,and the accuracy of test samples reached 98.7%.