Methods And Key Technologies Of On-Machine Contact Scan-Tracking Measurement For Large-Scale Parts

There are a vast of key large-scale components and parts in the modern weaponry of national defense, aerospace, energy and delivery, which require high precision but are difficult to process. The manufacturing quality of these components and parts becomes the constraint to the overall mechanical equipment capability. The deformation during the processing cannot be controlled efficiently owing to characters of large spatial scale, low stiffness and irregular shape, etc. These characters results in individualized machined surfaces. Therefore, conventional CNC machining process based on the original profile design cannot meet the requirements of mechanical accuracy and property standards of large-scale parts. Using on-machine contact scan-tracking neasurement technology, and building measuring-processing integrated manufacturing systems have become an efficient way of keeping a high level of manufacturing precision and efficiency for large-scale parts. This paper is supported by the State Key Program of National Natural Science of China "Processing theory and technologies for multi-source constraint, re-design class complex surface parts", Navy Equipment Pre-Research Project "Basic research of precision processing technology for***" and the project commissioned by China Academy of Launch Vehicle Technology. It focuses on the precise and efficient manufacturing problems of large-scale parts and studies the on-machine contact scan-tracking measurement methods and key technologies. The main contents of this thesis are as follows:Aiming at minimizing the negative effect on detection precision which caused by nonlinear and quasi-orthogonal features of3D scanning probe, the adverse effects of nonlinearity error and the deformation error of guiding mechanism are analyzed. The negative influences on measurement accuracy are qualified, which are led by parasitic angular and parasitic displacement of parallel reed mechanical structure. The mechanism of quasi-orthogonal mutual coupling effects is explored. Taking a DIGIT-03analog probe as an example, a separated non-linear calibration experiment is carried out. Furthermore, a reduction algorithm for quasi-orthogonal mutual coupling effects of3D scanning probe is provided.Owing to the hard real time requirement, multiple hysteretic links and large encircle negative feedback tracking control characteristics of on-machine contact scan-tracking measurement system, a global control system model including drives, mechanical structures, detection signals, sample delay, quantization error, interfering noise and non-linear friction is built. The parameters of the control system model are researched. The dynamic features of the system are analyzed in detail illustrating the influence of multiple interference factors.To improve the dynamic performance of scan-tracking axis, a dedicated control algorithm for tracking axis of the measurement system is designed. Based on conventional servo axis control algorithm of CNC system and the scan-tracking control model, Kalman filter is introduced to predict the tracking axis position, speed and interference noise. The natural frequency and damping ratio of the closed-loop poles are adjusted to the state feedback gain by the pole assignment controller. Zero phase error tracking control technology is adopted to optimize system bandwidth. Based on the identification of friction, friction feed forward compensation is performed, reducing the delay caused by friction transients resulting from frequent start-brake and reversing of the tracking axis.In order to protect the speed and stability of scan-tracking movement from being restricted by surface geometric and friction characteristics, a scan-tracking speed intelligent planning and dynamic friction compensation method is presented based on cerebellum model articulation controller (CMAC); meanwhile, an algorithm of identifying and quantifying the surface geometric and friction characteristics are designed. Moreover, a CMAC spatial mapping structure for scan-tracking measurement is built and the relevant learning algorithm is figured out. Since the spatial characteristics (including the geometric and friction characteristics) can automatically predict the feature of measured area, it is proved to achieve intelligent planning of scan-tracking speed and the dynamic friction compensation successfully.Due to the deficiencies of the existing on-machine data reduction methods, an on-line3D point cloud data reduction algorithm is proposed, which is derived from bi-Akima spline interpolation. This algorithm connects recorded points by using bi-Akima spline; meanwhile, it calculates the maximum distance between the original sampling points and the fitting curves and compares the maximum distance with required accuracy, then decide which point should be selected and recorded. The reduced data can be directly used in the following data processing and direct NC machining. This method can guarantee the required accuracy, reduce point cloud data, improve processing quality and compress spatial point cloud data.At last, an on-machine scan-tracking measuring system is developed based on commercial open CNC system. This system is successfully applied in high efficient, high reliable digital matching process of large-scale co-bottom components of fuel tanks in launch vehicle CA-3A. The application of this system makes a great contribution to the development of aerospace industry in China.