Enhancing the efficiency of the metal cutting processes is a vital need in the manufacturing technology. In this context, chatter vibration is considered as a limiting factor in enhancing metal cutting efficiency, because its occurrence causes many harm effects, such as poor surface finish, premature tool wear, and potential damage to the machine or the tool. This thesis presents a novel semi-active intelligent control technique for chatter vibration supersession using tunable magnetorheological (MR) dampers. Chatter is a self excited unstable vibration which is generated mainly due to regenerative effect. during machining processes, Structural dynamic characteristics; in particular magnitude, phase, and real and imaginary parts of the frequency response function are the main parameters which specify the stability lobes diagram. In the present work, chatter reduction is achieved by altering these factors semi-actively, by means of a MR damper inserted between a vibration-free component of the machine and the tool holder. First, simulation software is developed for studying the vibration of a lathe machine tool. A lumped model for MR damper is presented using modified Bouc-Wen model. The developed software is capable to consider the regenerative effect, i.e., in the calculation of the dynamic chip thickness, the current wave and the past wave left on the surface of the workpiece due to the machine vibration are taken into account. Since the overall system is nonlinear due to incorporation of MR damper and a nonlinear controller, the common frequency-domain stability lobes generation techniques cannot be applied here. A new approach for on-line chatter detection based on a new defined chatter detection index (CDI) is presented to recognize the chatter conditions and to generate stability lobes diagram. It is based on the analysis of vibration signals of the system. Next, MR damper model and chatter simulation software are integrated to study the influence of MR damper on the stability lobes diagram. Finally, a fuzzy controller is designed to choose the best voltage for MR damper at each instant, to prevent chatter occurrence. The proposed controller is a semi-active controller which requires much less power in comparison with other control approaches.An optimization by the use of Genetic Algorithmis done to optimize fuzzy membership functions, in order to reduce total energy requirement, and enhance stability conditions. The obtained results show that the proposed method has been successful in reducing the chatter conditions and improving the stability of turning operation, by very low energy consumption. Keywords Magnetorheological damper, machine tool vibrations, chatter, semi-active control, fuzzy systems, optimization