Geometric error modeling and sensitivity analysis of crankshaft comprehensive measuring machine
Summary:The crankshaft comprehensive measuring machine can realize the automatic detection of key parameters of the crankshaft.In order to improve the measuring accuracy of the crankshaft comprehensive measuring machine, in view of the many geometric error elements of the crankshaft comprehensive measuring machine, the compensation work is large and the process is complicated.The crankshaft comprehensive measuring machine made by our research group is studied by using the multi-body dynamics theory.The dynamics theory of multi-body system is usually expressed by the combination of homogeneous coordinate transformation and low order body sequence.The low order body sequence is used to describe the relationship between the typical bodies in the topological structure of the crankshaft integrated measuring machine, and the relationship of degrees of freedom between the components of the measuring machine is determined.Based on the principle of homogeneous coordinate transformation, the pose relationship between the adjacent body of the crankshaft comprehensive measuring machine in the ideal state and the actual state is described, and the ideal static, actual static, ideal motion and actual motion transformation matrix between each adjacent body is determined.Finally, according to the deviation between the actual measuring point coordinates and the ideal measuring point coordinates, the corresponding spatial geometric error is obtained.The geometric error model of vertical crankshaft measuring machine is established.In order to determine the degree of influence of geometric error on the measuring accuracy of the vertical crankshaft measuring machine and identify the key error items that have great influence on the measuring accuracy of the crankshaft measuring machine, the geometric error model of the crankshaft comprehensive measuring machine is analyzed by using Sobol global sensitivity analysis method.Monte Carlo method is employed to estimate the variance of the geometric error model of the crankshaft comprehensive measuring machine, and the sensitivity coefficient of the corresponding input is calculated by the ratio of the variance of the corresponding function value to the total variance of the error model.In the measuring space of the crankshaft comprehensive measuring machine, five coordinate points are uniformly selected along the X direction to calculate the sensitivity coefficient of each geometric error term along the X direction at the five coordinate points, and the mean value of the five sensitivity coefficients is calculated.Similarly, the corresponding five coordinate points are selected at the other three different heights to calculate the sensitivity coefficient of each error term.The mean value of sensitivity coefficient is also calculated, and the mean value of each error is arranged in the order of increasing height, and the median 0.1 is taken as the threshold to judge whether each geometric error item of the crankshaft comprehensive measuring machine is the key error item.Finally, 10 key errors which have great influence on the measuring accuracy of crankshaft comprehensive measuring machine are identified from 21 geometrical errors.Determining the key error terms can not only provide a basis for the subsequent error compensation and improve the measuring accuracy of the crankshaft comprehensive measuring machine, but also provide a theoretical basis for designing and optimizing the structure of the crankshaft comprehensive measuring machine.
crankshaft comprehensive measuring machinemultibody system theorySobol ' sensitivity analysisMonte Carlo estimationkey geometric errors
国家科技期刊平台
NSTL
万方数据
