【作者】 孙晓娜；

【导师】 雷毅；

【作者基本信息】 中国石油大学 ， 材料加工工程， 2007， 硕士

【摘要】 超声波技术在厚板奥氏体不锈钢焊缝结构的无损检测中占有重要地位,但在实际应用中存在信噪比低的现象,主要是焊缝组织产生的结构噪声引起的,因此需要应用数学方法对超声回波进行处理。制作60 mm厚的单U形和双U形坡口的奥氏体不锈钢(SUS304)焊缝,通过在焊缝的不同区域加工Ф2mm横通孔制成超声波检测试样。分别采用不同尺寸晶片和K值的横波斜探头和纵波斜探头,对试块中的横通孔进行超声波检测。研究发现,使用窄脉冲纵波斜探头效果较好,检测时信噪比能达12 dB以上,但受横波存在的影响;使用窄脉冲横波斜探头检测时,信号信噪比低于8 dB。对焊缝不同区域取样,结合不锈钢冶金原理对焊缝组织进行金相分析表明,厚板奥氏体不锈钢焊缝的组织为粗大平行柱状晶奥氏体+残余的?铁素体,呈现横向各向同性,具有5个相互独立的弹性系数。根据超声波在粗晶材料中散射场的特点和弹性波在线性弹性体中的传播特性,建立了散射背景下的点源综合模型。利用CTS-3000数字化超声波探伤仪和计算机,对焊缝中人工孔的超声回波信号进行采集,并采用小波理论信号进行处理。将采集到的超声波数字信号导入MATLAB软件,应用一维小波自动降噪方法进行去噪。改变去噪函数的各种参数,对比不同参数下的去噪效果,与去噪前相比,使用不同类型探头时,回波信号的信噪比分别提高了4～6dB。更多还原

【Abstract】 Ultrasonictestingisveryimportantinthenondestructiveexaminationofthick-plate austenitic stainless steel welds. But in pratics, the signal-to-noiseratio(SNR) is low, which is caused by the structure noise of weld histology,therefore mathematical treatment is needed to increase the SNR of theultrasonic back wave. In this research, 60 mm thick austenitic stainlesssteel(SUS304) plates were multilayer butt-welded with single and doubleU-groove, and then horizontal through holes with diameter 2mm weremachined along the middle lines and fusion lines. The horizontal throughholes in weld joints were detected by the shear and longitudinal wave anglebeam probes with differentcrystal plate sizes and K values. The results showthat it gets the better effect when the narrow pulse longitudinal wave anglebeam probes wereusedandtheSNRreaches 12 dB,whileSNR islowerthan8dB when the narrow pulse shear wave angle beam probes were used Thewelds were sampled in different area. Associating the stainless steelmetallurgy theory, the metallograph of welds samples were analyzed. It isfound the thick-plate austenitic weld joint is made up of parallel columncrystal austenite andδferrite distributed along the column crystals, and ischaracterized by transverse isotropy with 5 independent elastic constants.Basedonthefeatureofultrasonicscattered fieldincoarse-grainedmaterial,apoint source synthesis model of scattered fields was built. The CTS-3000 digital ultrasonic flaw detector and computer were used to collect theultrasonic back wave digital signal of through holes. The signal data wasimputed into MATLAB software and was denoised automatically by using1-Dwavelettransform.Comparedwiththesignalbeforedenoising,thesignalSNR underusingvariousprobeincreases4～6dB.更多还原