Modelling and real-time control of a high performance rotary wood planing machine

Rotary planing is one of the most valuable machining operations in the timber processing industry. It has been established that cutting tool inaccuracy and forced vibration during the machining process are the primary causes of surface quality degradation. The main aim of this thesis is to design a control architecture that is suitable for adaptive operation of a wood planing machining in order to improve the quality of its surface finish. In order to achieve the stated goal, thorough understanding of the effects of machine deficiencies on surface finish quality is required. Therefore, a generic simulation model for synthesising the surface profiles produced by wood planing process is first developed. The model is used to simulate the combined effects of machining parameters, vibration and cutting tool inaccuracy on the resultant surface profiles. It has been postulated that online monitoring of surface finish quality can be used to provide feedback information for a secondary control loop for the machining process, which will lead to the production of consistently high quality surface finishes. There is an existing vision-based wood surface profile measurement technique, but the application of the technique has been limited to static wood samples. This thesis extends the application of the technique to moving wood samples. It is shown experimentally that the method is suitable for in-process surface profile measurements. The current industrial wood planing machines do not have the capability of measuring and adjusting process parameters in real-time. Therefore, knowledge of the causes of surface finish degradation would enable the operators to optimise the mechanical structure of the machines offline. For this reason, two novel approaches for characterising defects on planed timber surfaces have been created in this thesis using synthetic data. The output of this work is a software tool that can assist machine operators in inferring the causes of defects based on the waviness components of the workpiece surface finish. The main achievement in this research is the design of a new active wood planing technique that combines real-time cutter path optimisation (cutting tool inaccuracy compensation) with vibration disturbance rejection. The technique is based on real-time vertical displacements of the machine spindle. Simulation and experimental results obtained from a smart wood planing machine show significant improvements in the dynamic performance of the machine and the produced surface finish quality. Potential areas for future research include application of the defects characterisation techniques to real data and full integration of the dynamic surface profile measurements with the smart wood planing machine.