Substructuring Methods for Efficient Prediction of Spindle-Holder-Tool Assembly Dynamics
von Prateek ChavanThis thesis provides a framework of efficient experimental and analytical substructuring methods for the reliable prediction of spindle-holder-tool assembly dynamics. Specifically, the prediction of the tool tip Frequency Response Functions (FRFs) as well as the tool tip-Spindle Integrated Displacement Sensors (SIDS) transfer functions for a main spindle equipped with contactless displacement sensors. Here, the holder-tool assembly is modelled analytically whereas the spindle substructure is modelled experimentally.
For achieving high-quality experimental response models, three important aspects are researched. Firstly, strategies for measurement of displacement-to-force compliances are systematically compared and assessed. Secondly, two new methods for the identification of rotational compliances are proposed based on a modal parameter approach and commercially available rotational accelerometers. Thirdly, an experimental method for obtaining the interface flange-SIDS transfer function matrix and corresponding measurement uncertainty is developed and validated. This is required for prediction the tool tip-SIDS transfer function along with propagated uncertainty bounds.
A significant challenge in the analytical response modelling of holder-tool assemblies is that features like joint parameters cannot practically be modeled a priori and require additional reference measurements for their parametrization. This thesis proposes an extended tool model updating approach, where FRFs of the holder-tool assembly measured in an offline, freely constrained state are used for feature parameterization. Using this approach, a priori unknown parameters like joint stiffness and effective diameter of the fluted segment are reliably identified and validated for several different holder-tool assemblies. Furthermore, methods for the accurate beam modelling of various holder features such as balancing holes, tapered segment and holder-inserted tool segment were developed, analyzed and systematically validated.
Another valuable contribution of this thesis is demonstrating the utility of the predicted tool tip and tool tip-SIDS FRFs for estimating process forces and virtual workpiece quality. In this regard, the practical integration of the developed substructuring methods with an existing virtual quality framework is presented and successfully implemented to estimate process forces and workpiece quality for different milling operations and tool assemblies.
For achieving high-quality experimental response models, three important aspects are researched. Firstly, strategies for measurement of displacement-to-force compliances are systematically compared and assessed. Secondly, two new methods for the identification of rotational compliances are proposed based on a modal parameter approach and commercially available rotational accelerometers. Thirdly, an experimental method for obtaining the interface flange-SIDS transfer function matrix and corresponding measurement uncertainty is developed and validated. This is required for prediction the tool tip-SIDS transfer function along with propagated uncertainty bounds.
A significant challenge in the analytical response modelling of holder-tool assemblies is that features like joint parameters cannot practically be modeled a priori and require additional reference measurements for their parametrization. This thesis proposes an extended tool model updating approach, where FRFs of the holder-tool assembly measured in an offline, freely constrained state are used for feature parameterization. Using this approach, a priori unknown parameters like joint stiffness and effective diameter of the fluted segment are reliably identified and validated for several different holder-tool assemblies. Furthermore, methods for the accurate beam modelling of various holder features such as balancing holes, tapered segment and holder-inserted tool segment were developed, analyzed and systematically validated.
Another valuable contribution of this thesis is demonstrating the utility of the predicted tool tip and tool tip-SIDS FRFs for estimating process forces and virtual workpiece quality. In this regard, the practical integration of the developed substructuring methods with an existing virtual quality framework is presented and successfully implemented to estimate process forces and workpiece quality for different milling operations and tool assemblies.