Kinematic and dynamic modeling of grinding processes

Official URL: http://risc01.sabanciuniv.edu/record=b1669820 (Table of Contents)

The use of abrasive tools in grinding and similar manufacturing processes continues to increase in the production of high surface quality or difficult to process materials used in especially aviation, automotive and biomedical industry. The complicated geometric structures of the abrasives used in these processes show significant differences with the tools in other machining processes. Instability in material removal operations has been one of the critical obstacles in manufacturing, hindering productivity as well as resulting in unfavorable workpiece quality. Abrasive processes are often associated with finishing operations, aimed to give workpiece a final geometry and surface condition which makes chatter even more critical in grinding. For these reasons, it is quite time-consuming, costly, and in some cases impossible to achieve the desired quality and performance with conventional trial and error methods in abrasive processes. Process models based on analytical and experimental methods constitute the aim and goal of this thesis as they can be used effectively in the analysis of these processes and in selecting the most appropriate process conditions to increase the performance of abrasive processes. In this study, a new simulation method named geometric-kinematic model has been developed for grinding. The geometric-kinematic model provides the prediction of grinding forces and surface roughness of the workpiece by simulating the micro-interactions of the abrasive particles and the workpiece surface. Using the milling analogy and the normal distribution of the individual grits on the wheel surface, determination of active grits hence the chip thickness calculation per grit is also possible. A time-domain simulation is constructed employing the regenerative effect by utilizing the dynamic chip thickness calculation. The stability regions are determined through time domain simulations and analytical model predictions. The simulation results are compared and verified by experimental data.