Surface roughness and tool wear monitoring in turning processes through vibration analysis using PSD and GRMS

This study presents a novel approach to monitor and predict surface roughness and tool wear in the turning process, which is crucial for anticipating tool failures, reducing replacement costs, and optimizing production efficiency. The study analyzes vibration signals collected during the turning process of a stainless-steel workpiece with a carbide insert until the tool wear threshold (V_B = 300 µm) is reached. Firstly, the vibration signature associated with the machine and the noise were isolated using the Fourier transform (FFT). Then, the optimal frequency band is selected to extract maximum valuable information using the estimated power spectral density (PSD) through the Welch method. The correlation between the vibration signals and surface roughness is then analyzed by calculating the average root mean square (RMS) acceleration of all the obtained PSD curves. Finally, a mathematical prediction model is extracted using a simple linear regression equation between GRMS and surface roughness. The results show a good agreement between the predicted data and the experimental values. The coefficients MSE, RMSE, and MAE have low values of 0.025, 0.1581, and 0.1174, respectively, confirming the accuracy of the proposed model.