Research On Fast Identification Method For Machine Tool Spindle Temperature Rise Characteristics And A Novel Cooling Structure Design

In the CNC machine tools with the higher speed and higher precision, thermal deformation is becoming one of the greatest contributors to machining errors, thus the problem of how to reduce thermal deformation is widely paid attention to. The spindle system is the core component of machine tools and is also the main heat source. The spindle’s thermal characteristics determine the machining accuracy that machine tools can achieve. Therefore, it becomes a key point to obtain thermal characteristics of the spindle system accurately and quickly, meanwhile, it is significant to put forward the methods of reducing thermal deformation. The dissertation is supported by National Nature Science Foundation of China (No.51175461) and the National Science and Technology Project of China (No. 2012ZX04001-011). Based on the thermal characteristic test, some researches on fast identification of a machine tool selected points temperature rise have been made. Three methods have been presented and validated by experiments. Furthermore, some studies have been done to reduce the thermal deformation from the perspective of temperature control, and a novel fractal tree-like channels network heat sink has been designed. It is different from the traditional helical channels heat sink. The novel structure can greatly improve the cooling efficiency and reduce the thermal deformation of the spindle, which has been validated by simulation and experiment. The main contents of the dissertation are summarized as follows:Thermal characteristic test is a necessary step in designing precision spindle, and also an experimental method to obtain thermal characteristics accurately. It always lasts several hours or even longer in practice. In order to shorten thermal characteristic test duration, a method of fast identification for machine tool spindle temperature rise based on operational thermal modal analysis was presented. Based on the method, spindle temperature rise curve can be predicted, and thermal equilibrium time and steady-state temperature can be obtained by use of the measured temperature data during a short time. The method can greatly improve the efficiency of thermal characteristic test.The kalman filter theory and extended kalman filter were introduced firstly, and then unscented kalman filter was analyzed. Because the unscented kalman filter had an advantage in nonlinear state estimation and parameter identification, it was used in selected point temperature rise identification. But the normal unscented kalman filter suffered from performance degradation and even divergence when mismatches had been found between the noise distribution assumed to be known as a priori and the true distribution in a real system. Thus, an adaptive law was applied to adjust parameters dynamically by the actual measured temperature, which could effectively avoid the failure of prediction. In order to shorten thermal characteristic test duration, the minimal identifying time criterion was presented. Based on adaptive unscented kalman filter, a novel method for fast identification of a machine tool selected point temperature rise was proposed.Support vector machine regression algorithm was introduced. It had the unique advantages in solving the small sample, nonlinear and high dimensional pattern recognition problems. So it was used for time series prediction. The temperature rise of the spindle running at the idling condition was also a time series, so the support vector machine regression was used to predict the temperature change. At the same time, the minimal identifying time criterion was presented to determine the minimal time in which the temperature rise could be estimated accurately. Only by using measured temperature data in a short period of time can the selected points temperature rise, steady-state temperature, and thermal equilibrium time be predicted accurately.The thermal characteristic test in the vertical machining center was done. The measured temperature data of several temperature measuring points were obtained. By using these data, three methods of fast identification for spindle selected point temperature rise were validated. It indicated that the temperature rise of the selected point from the start-up of machine tool to the temperature steady-state machine tool reaching had been identified in a short time through these three methods, which greatly shortened the test time.hi order to control the spindle temperature rise, a novel fractal tree-shaped network was presented. The heat transfer and pressure drop characteristics were theoretically analyzed. Compared with the traditional parallel channels network, the fractal tree-like network has better heat transfer capacity and lower pressure drop. Based on the fractal theory, a novel fractal tree-like channels network heat sink in spindle’s cooling sleeve was designed. A three-dimensional thermal and hydrodynamic model for fractal tree-like channels heat sink was developed. The heat transfer and pressure drop characteristics were numerically investigated, taking into consideration conjugate heat transfer in the channel walls. The traditional helical channels heat sink was designed with the same convective heat transfer area and inlet geometric dimension. The pressure drop, temperature uniformity and coefficient of performance of the fractal tree-shaped heat sink were evaluated and compared with those of the traditional helical channels heat sink. A conclusion was made that fractal tree-like channels heat sink had lower pressure drop, more uniform temperature field distribution and larger coefficient of performance than the traditional helical channels heat sink. Both heat sinks were fabricated on a steel bar, and the conclusion was verified by the experiment.