Determination of vibration characteristic in automatic tapping operations based on artificial neural networks

One of the critical processes in machining is the tapping operation, which is increasingly being performed automatically due to advancements in CNC and drilling technologies. Unpredictable vibrations significantly affect threading accuracy, reducing precision and shortening tool life. This study investigates the prediction of vibration characteristics during automatic tapping operations using artificial neural networks (ANN). Experimental studies were conducted using different feed rates, spindle speeds, and material types to analyze their impact on vibrations. Three ANN models were designed and evaluated based on their effectiveness in predicting vibration characteristics. The results indicate that the proposed Radial Basis Function Neural Network (RBFNN) performs exceptionally well in the real-time prediction of vibrations during tapping. This study uniquely applies ANN models to automatic tapping vibration analysis, demonstrating high accuracy under varying conditions.