Design of an Acoustic Measurement System for Non-Contact Displacement Estimation and its Experimental Verification in Turning Operations

This study presents a novel non-contact displacement estimation technique using a microphone in turning operations. Microphones are inexpensive, non-contact measurement devices. Therefore, there is a high possibility that microphones can be used as the non-contact measurement means of structural vibration such as forced/chatter vibrations and/or machining phenomena such as tool wear and chip morphology. First, a sound pressure measurement system that can be installed close to the cutting edge was designed. In the designed system, a metal pipe is installed in the path from the source of the cutting sound to the microphone to provide anisotropic measurement sensitivity, which reduces the influence of environmental noise and provides protection of the microphone from cutting fluid and scattering chips simultaneously. Next, a compensation method for the measurement signal is proposed. In the developed measurement system with a metal pipe, sound resonance occurs and resonant frequency components are amplified. In addition, this sound pressure measurement system has the problem of generating harmonics due to sound pressure distortion. Digital processing of the measured sound compensates the frequency response of the designed sound pressure measurement system by the inverse transfer function, which is identified in advance. In addition, harmonics are removed by band-stop filters, so that the original sound pressure information can be reconstructed. Furthermore, the relationship between sound pressure and vibration displacement is formulated based on the wave equation, and a sound pressure-vibration conversion model is proposed to estimate the vibration of a mechanical structure from sound pressure information. A series of experiments were conducted to clarify the anisotropy and frequency characteristics of the prototype sound pressure measurement system, and it was found that the sound pressure signal can be corrected for steady-state vibration. In addition, turning experiments were conducted on an experimental system with a flexible workpiece structure to estimate the vibration of the workpiece which is triggered by chatter instability. The results of the validation experiments confirmed that the proposed method can predict the vibration displacement from the cutting noise. However, it was found that the estimated vibration displacement had a phase error, and thus accurate estimation of transient vibration phenomena was not sufficient. This fact indicates that further modification of the proposed model is necessary.