Optimisation of Machining Operations by means of Finite Element Method and Tailored Experiments

Doctoral thesis, 2015

Traditionally, costly experimental procedures have been followed in industry to optimise the machining operations to secure the maximum efficiency in production line, while the functional requirements of the machined surfaces are fulfilled. In recent years, development of robust numerical techniques such as Finite Element Method (FEM) and advances in computing capacity of computers have made it possible to simulate the machining operations under operational conditions. This has provided a platform to evaluate the responses such as the tool-chip contact stress and temperature for a certain process and to determine the optimum cutting condition by using a considerably smaller number of tool life tests. Nonetheless, the reliability of FE simulation results is shown to largely depend upon implementation of a well-defined material model, which can properly describe the severe material deformation in the vicinity of cutting edge. In the current study a methodology is developed to determine the flow stress properties of the work material together with the optimum frictional condition at the tool-chip interface by means of inverse modelling of orthogonal cutting test. The calibrated constitutive model and the friction coefficients are then used to build three dimensional (3D) FE models to simulate the cutting process under operational conditions. The FE simulation results are integrated with a limited number of tool life tests to determine the optimum conditions. The viability of the presented methodology is evaluated for finish face milling of the bi-metal aluminium-grey cast iron component-like samples using uncoated cemented carbide tools. This methodology is also adopted for tool flank wear prediction in transverse machining Inconel 718 in aged condition and longitudinal machining 20MnCrS5 case hardening steel using uncoated cemented carbide tools. In order to optimise the finish face milling of the bi-metal workpiece, a limited number of tool life tests are initially carried out under controlled conditions to determine the involved wear mechanisms. Investigations by scanning electron microscopy (SEM) revealed that the main wear mechanism constitutes thermal fatigue cracking. Hence, the milling process is simulated under operational conditions and the thermally induced stresses on the cutting edge are calculated for certain combinations of cutting parameters including cutting speed, feed per tooth and depth of cut. Response Surface Methodology (RSM) is then adopted to establish the mathematical relation between the cutting parameters and the range of thermally induced stresses on the cutting edge. Subsequently, the optimum cutting condition is determined once the amount of thermally induced stresses on the tool edge and the number of tool engagements per unit length of cut are both minimised. This approach led to 32% reduction in cycle time compared to a reference tool life test. An FE-based wear modelling approach is developed to predict the flank wear evolution on the uncoated cemented carbide tools in transverse machining Inconel 718 under operational condition. In this approach, initially, the relation between the tool volume loss per unit time and the flank wear evolution rate is established based on the well-known Usui wear model. Once the rate equation was calibrated by the results of the tool life test at a certain cutting condition, it was then possible to integrate it with the FE models to estimate the flank wear evolution for an arbitrary cutting condition. This wear modelling approach is then extended based on the concept of experimental design and RSM to attain a higher accuracy and better computational performance, particularly for the case of 3D FE modelling of cutting process. The extended methodology is adopted for flank wear estimation in longitudinal turning 20MnCrS5 case hardening steel using uncoated cemented carbide tools. The prediction results showed a good agreement with the tool life measurements.