管壁厚度变化炮筒的无损检测

Nondestructive testing of gun barrel with wall thickness changes

由于炮筒管壁属于板型构件,且厚度较小,使用传统的横波与纵波无损检测都较困难,而且漏检率和误判率较高。通过对炮筒构件壁的厚度变化以及缺陷特征的分析,提出用单模式兰姆波水浸探伤方法对薄、不等厚壁炮筒进行无损检测,并用5层小波分解与重构的方法对采集到的信号进行去噪。通过Hilbert变换提取出信号包络,然后对信号进行10点滑动平均等一系列信号处理,从而达到对缺陷信号的有效识别。兰姆波的硬件检测与小波分析的信号处理有效融合,使得无损检测的漏检率和误判率有很大的提高。并采用20通道超声检测总控系统,达到3 s/件的检测速率,大大提高检测效率。

Gun barrel walls belong to plate-type components and are small,so it is difficult to identify the defects and both the miss rate and the misjudgment rate of conventional transverse wave and longitudinal wave nondestructive testing are high. By analyzing the wall thickness changes and defect characteristics of cartridge components,have put forward a single-mode lamb wave water immersion detection method to check the defects of thin-and thick-wall gun barrels and then used five-layer wavelet decomposition and reconstruction to remove the noises of acquired signals at the same time. Signal envelopes have been extracted by the Hilbert transform and then the signal was processed by 10-point moving average and other ways to identify flaw signals. The fusion of the hardware detection of lamb waves with the signal processing of wavelet analysis have greatly reduced the miss rate and the misjudgment rate of nondestructive testing.Moreover, a 20-channel master control system for ultrasonic testing was adopted and the detection rate was up to one gun barrel per three seconds.