Modeling and estimation of thermally induced bearing loads in machine tool spindles

Machine tool spindle bearings are subjected to a large range of axial and radial loads due to the machining process. Further the rotating spindle must be extremely stiff to minimize the cutting tool's deflection. The high spindle stiffness is achieved by applying a mechanical load to the bearings, the preload. In fixed preload spindles the bearing loads tend to increase with increasing spindle speed due to thermal expansion and it is well established that these thermally induced loads can lead to premature bearing failure. The traditional approach to avoiding excessive thermally induced bearing loads is to limit the spindle speed. This is incompatible with the need to increase machining productivity by using higher spindle speeds. This trade-off between maintaining low bearing loads, and high spindle stiffness requires machine tool spindle design tools that are capable of analyzing the interaction between the thermally induced bearing loads and the parameters that impact these loads.; A three state model of thermally induced bearing loads is presented which is representative of the models used in recent thermally induced bearing load studies. This model is used in a thermally induced roller bearing load estimation system design study in which different parameter variation compensation methods are used. These bearing load estimation methods and the model that they are based upon do not consider the axial thermal gradients and expansions that are important in the calculation of angular contact bearing loads.; Three models of thermally induced bearing load in angular contact bearing spindles are developed that include the calculation of the radial and axial thermal expansions and the bearing and shaft dynamics. A two dimensional lumped mass, a finite element and a reduced order finite element model of the heat transfer and thermal expansion within the spindle's housing and shaft are developed. In the reduced order model, nodal reduction is used to minimize the number of temperature states which dramatically reduces the computational load.; Simulations are used to demonstrate the relative magnitude of the radial and axial thermal expansions and to demonstrate the increased accuracy of the finite element model. The reduced order model is validated using measured temperature and bearing load data. The thesis ends with a discussion of the future development of an angular contact bearing load estimation system. The contributions of this thesis include the development of a finite element thermally induced angular contact bearing load model that is both accurate and computationally efficient.