基于递归分析的动车组牵引电机转子导条微分层缺陷超声无损检测技术

A Method Based on The Recurrence Analysis of Ultrasonic Backscattered Signals for Micro Defects Detection in Copper Alloy Extruded Bars

在实际生产中,铜及铜合金挤压件存在一种微小的分层缺陷,其超声检测的识别成功率较低。利用递归分析技术对含微分层缺陷的挤压铜导条进行检测,截取得到含有缺陷信号的背散射信号,选定递归参数后对其进行分析,并生成不同阈值下的递归图。同样绘制不含缺陷信号的递归图,通过对比发现,含缺陷回波信号的递归图中出现了周期特性。对递归图分别进行递归定量分析,发现两者的平均对角线长度(L)、递归熵(ENTR)、层状度(LAM)与圈闭时间(TT)有明显区别。

In actual production, copper and copper alloy extrusions have a slight delamination defect.And thisdefect is hard to be identified by ultrasonic test.Recurrence plot was used to study the signal of the extruded copper bar containing slight delamination defect.The backscattered signal from the signal of test block was intercepted, and by selecting the correct parameters to analyze it, the recurrence plot was obtained.When compared with the recurrence plot of the signal of no defect test block, the recurrence plot of defect signal shows a clear periodicity.To study the influence of the defect signals on recursive nature, we used the recurrence quantification analysis to get the characteristic quantities of the whole backscattered signals.And it was found that trapping time(TT),averaged diagonal length(L),laminarity(LAM) and entropy(ENTR) of the two are significantly different.