Abstract: Wear of joints is an important factor influencing the performance of a mechanism, and wear prediction is of great significance to the design and maintenance of a mechanism. For the multiple joints in a complex motion mechanism, it is difficult to predict the wear in the joints accurately due to the complexity of the load and motion. To solve this problem, this paper proposes an integrated wear prediction method based on the monitoring data of motion output. The wear coefficient is treated as a random variable in analysis of the uncertainty in the processing, manufacturing and running environment. The mapping relationship between the motion output and the wear depth of multiple joints is deduced according to the kinematic model of the mechanism, and then the wear state data of the joint can be obtained according to the easily obtained monitoring data of the motion output (such as displacement, angle, etc.). On the basis of the obtained wear state data of the joints, distribution information of the wear coefficient is updated based on the Bayesian theory. With more monitoring data, the posterior distribution of the wear coefficient is more realistic. Based on the multibody dynamics model and the Archard wear model, the wear depth of joints is predicted. Wear experiment of a lock mechanism in an aircraft cabin door shows that the prediction error is about 6%, verifying the effectiveness of the proposed method.

Key words: motion mechanism, joint, wear, condition monitoring data, Bayesian inference