Vibration assisted drilling (VAD) is a type of machining process in which high-frequency vibrations with a small amplitude are induced in the cutting tool to improve the cutting process of hard and brittle materials. These vibrations create an unsteady repetitive processing effect which eventually reduce the cutting forces. It is also crucial to measure these forces in some way because their knowledge directly aids in determining the best machining parameters. Direct and indirect methods can be used to measure these forces, but due to serious limitations of direct measurement methods, an indirect measurement method is required which is capable of online monitoring of high-frequency cutting forces.

In this thesis, an indirect method is proposed to estimate thrust force and torque from the voltage signal generated by piezoelectric sensor and torsional deflection signal measured through piezoelectric accelerometer. The estimation of two input signals requires a multi-input multi-output (MIMO) model of VAD system which is developed using Receptance Coupling and Substructure Analysis (RCSA) method. Experimental and numerical methods are used to validate the constituent single-input single-output (SISO) transfer functions of the MIMO model. As the estimated forces are distorted by the dynamics of VAD structure, a Kalman Filter is employed to compensate the dynamics. The accuracy and similarity of results is determined by comparing the estimated cutting force values with the force measured from a load cell in time and frequency domain. The reported experimental results confirm the possibility of using Kalman Filter in estimating high-frequency forces generated in VAD process.