Research On The Calibration Technology For A 5-DOF Tri-Coordinate Measuring Machine

Five dof coordinate measuring machine is a new kind of non-Cartesian type tri-ordinate measuring machine. The use of open chain structure has the characteristics of wide error sources and strong variable coupling, all of which make it difficult to calibrate.The traditional calibration technology mainly uses tri-ordinate measuring machine or standard workpiece to provide standard values, and compares the standard values with the value resulted by uncalibrated equipment to find the error, then analyzes the spatial location relationship between the errors and equipment, and gets equipment error model. But actually, according to this method, the workload is big, data processing is also complex, it is difficult to separate and determine the error sources and the different error precisions influence on equipment. Moreover, the influence of random errors makes it very difficult to design a feasible calibration method.This paper puts forward a new equipment calibration method. Firstly, the paper researches the parameter feature of equipment models; and the error source and the effects of every error for equipment precision are analyzed. Based on features of equipment structure, the two balls model of equipment pose determined reverse from known points is given, and can use it to undertake optimal measurement pose analysis. Second, the mathematical model of grating eccentric is developed and analyzed here, and summarizes eccentric compensation formula. Then we design specific jigs and measuring heads, present a division calibration method. With the division calibration method, the coupling between the parameters is relieved artificially. Through analyzing several groups of combined joint angle, and using the parameters for geometric figure that corresponds to the trajectory formed by the head in measurement space, the equipment error model is described. Finally, Combined with five dof coordinate measuring machine and computer technology, the proposed method is verified by a numerical experiment, and the result shows that the division correction method is feasibility and practicability. The method provides a simple and feasible solution for determination of device model parameters, reducing and eliminating manufacturing error and assembling error.