Thermomechanical modeling of motorized spindle systems for high-speed milling
Abstract
Motorized high speed spindles with angular contact ball bearings and minimum oil lubrication may suddenly fail due to thermal problems, for which the exact reason is unknown. Currently, internal temperatures, internal heat flow, thermal preload and stiffness changes in this type of spindle cannot be predicted with sufficient accuracy, and there are observations that cannot be fully explained by previous models. It is believed that the understanding of thermal and mechanical interactions between different spindle components in a practical spindle system is the key to improving spindle performance and reliability. This research proposes to develop an integrated thermo-mechanical model to account for all heat sources, heat transfer paths, heat sinks, and relative thermal expansions of the spindle system. The temperature field of the entire spindle is predicted by an axisymmetric finite difference model. The model includes linear heat conduction and nonlinear convection and can efficiently solve for the temperature growth of each element. The model accuracy is then validated by comprehensive experiments through accurate temperature and heat flux measurements. Finally, the predicted temperature field is used to determine the mechanical behavior changes in component fit conditions, stiffness, bearing preload, natural frequency, etc. Quantitative relations for temperatures, internal heat flow, thermal preload and spindle stiffness as functions of spindle speed, set cooling conditions and the rigidity of the preloading mechanism have been established. Sensitivity to changes of speed and cooling conditions was also investigated. These results reveal some new observations of spindle behavior that were not reported before. Temperatures of the spindle shaft at high speed are strongly dependent on convective heat flow to surrounding fluids. Also, it was possible to identify temperature oscillations of the bearings under certain operating conditions. The model can be used for diagnosis or design and aims at helping to reduce premature spindle failure and to avoid changing bearing stiffness during machining.
Degree
Ph.D.
Advisors
Tu, Purdue University.
Subject Area
Industrial engineering|Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.