Study On The Distribution Law Of Ultrasonic Sound Field For Cleaning Disc Parts Under The Excitation Of Vibrating Plate Sound Source

Ultrasonic cleaning equipment is widely used in various industrial fields due to its efficiency,good quality,and easy automation.However,when cleaning parts in batches,the mutual obstruction between parts makes the sound intensity in the obstructed area unable to meet the cleaning requirements,resulting in poor cleaning effect.Therefore,when designing a cleaning plan,it is necessary to arrange the sound source and part position reasonably according to the structural characteristics of the parts to ensure that all cleaning surfaces meet the sound field requirements.This article takes commonly used disc parts as the research object,and based on a large vibrating plate sound source composed of multiple transducers and stainless steel vibrating surfaces treated with waterproofing,the distribution of sound field when parts are placed in the large vibrating plate sound source is studied through experiments,in order to provide guidance for the reasonable design of vibrating plates and the reasonable arrangement of parts during ultrasonic cleaning of disc parts.(1)The ultrasonic cleaning platform was reconstructed and built on the existing YKTD-L1200 W ultrasonic cleaning machine.From the perspective of the vibration plate sound source,the impact of the transducer array(number and layout of transducers)and vibration surface parameters(length,width,and thickness of the vibration surface)on the sound field distribution was studied.Based on the placement of two disk-shaped workpieces in a forward-facing parallel manner,the sound field distribution law of obstructed areas under different parameters was obtained.The results show that the direct coupling between the sound fields of each element in the transducer array affects the size and uniformity of the sound field distribution on the surface of the workpiece.Therefore,the array of transducers and vibration surface parameters need to be reasonably designed according to the sound field requirement during cleaning.(2)Based on the large vibrating plate sound source,the mutual occlusion area of two disk parts when placed in a forward or sideways position was analyzed under different parameters of part radius R and part spacing L,and the sound field distribution under different conditions was quantitatively obtained.The results show that under the same part parameters,the surface sound field distribution of the sideways placed parts is better than that of the forward placed parts.At the same time,the optimal range of parameters for ultrasonic cleaning of disk parts under different placement conditions was obtained.(3)Considering the complex nonlinearity of predicting sound pressure changes in the mutual occlusion area of parts,a data-driven modeling approach was adopted to establish a regression prediction model for sound pressure changes by building a BP neural network.The number of input and output layers was determined,the number of hidden layers was designed,and the learning rate and transfer function of the neural network were determined.The data obtained from experiments were used to train the neural network and complete the analysis of the regression prediction model for the surface sound field of the parts.The established sound pressure regression prediction model was verified by setting random experiments,and it was found that the average prediction error was within 10%,proving the effectiveness of the model.