Vector characterization and evolution of the running-in attractor

摘要

Running-in is an important stage in friction and wear, the research on its process and evolution law is significant for running-in state identification and prediction. Running-in process tests were conducted on a tribometer under different conditions and its signals were collected. The phase space reconstruction and principal component analysis methods are introduced to reconstruct phase trajectory. Considering the limitations existing in traditional parameters, the vector characterization parameters of the minimum cosine similarity and the maximum distance between the adjacent phase points are proposed and their validity is studied. They can identify the state changes and characteristics of the running-in attractor and describe the evolution law of the running-in process. Thus, the vector characterization parameters have important application value and theoretical significance.