金刚石刀具研磨声发射信号的非高斯性特征分析  

Analysis on Non-Gaussian Characteristic of Acoustic Emission Signal in Diamond Tool Grinding

作　　者：杜文浩[1,2] 孙涛[1] 王宝瑞[2] 

机构地区：[1]哈尔滨工业大学精密工程研究所,哈尔滨150001 [2]中国工程物理研究院机械制造工艺研究所,绵阳621900

出　　处：《纳米技术与精密工程》2015年第3期211-218,共8页Nanotechnology and Precision Engineering

摘　　要：为确保金刚石刀具的制备效率和质量,采用声发射信号作为监测手段对金刚石刀具的精密研磨辅助进行过程调控.然而,该声发射信号的有效特征不能通过传统的信号处理方法辨识,实际研磨中仅能通过人工监听和操作经验定性判断.因此,本文从砂轮磨粒与刀具研磨面的微观接触特点出发,采用泊松过程模型对金刚石刀具研磨中声发射信号的激励机制进行了分析,阐明该信号具有非高斯性特征,采用线性系统理论阐明这一特征可被不失真地传输和接收,同时也给出了该信号高斯化近似的约束条件.最后通过工艺实验验证了该声发射信号的非高斯性特征在金刚石刀具研磨过程中的存在性,也证实了这一特征不受研磨速度和研磨压力等工艺参数的影响,而是与刀具研磨面的耐磨方向存在固有关联关系,从而为声发射信号监测在金刚石刀具研磨自动化系统中的合理应用奠定了方法基础.To guarantee high tool quality and grinding efficiency, acoustic emission （AE） signal monito- ring is necessary in precision grinding process of diamond tool. However, the signal has some special characteristics which are hard to be identified effectively using traditional signal processing methods. Hence, operators had to rely on their own ears to perceive empirically any change of the signal. Based on Poisson process model, it is firstly analyzed how the signal is stimulated according to micro-scale contact properties between abrasive grain and ground surface, and then it is pointed out that the signal enjoys non-Gaussian characteristic. Based on linear system theory, it is explained why signal transmitting and monitoring would not change such non-Gaussian characteristic, and conditions that the AE signal can be approximated to Gaussian signal are given . Finally, statistical analysis of monitored signal samples in ex- periments illustrates the existence of non-Gaussian characteristic of AE signal, and that it does not change with experiment parameters of grinding speed or grinding pressure, but has some relations with wear re- sistance direction in the ground surface of diamond tool. All discussions and methods provide great sup-porl to apply AE signal monitoring technology into automation system of diamond tool grinding.

关 键 词：金刚石刀具 声发射信号 非高斯性特征 

分 类 号：TG71[金属学及工艺—刀具与模具] TG580.68