利用超声衍射时差法(Time of Flight Diffraction,TOFD)对合金钢焊缝中的小尺寸裂纹实施检测时,受脉冲宽度等因素制约,裂纹上、下端点的衍射波会发生混叠,而导致裂纹高度定量困难。采用自回归谱外推(Autoregressive Spectral Extrapolation,ARSE)技术对混叠信号进行处理,扩展有效频带范围,达到压缩时域信号脉宽和提高时间分辨率的目的。结果表明,结合ARSE技术可有效分离混叠信号,实现厚度100.0mm合金钢中深度50.0mm处高度1.0mm裂纹的TOFD定量检测,且相对定量误差不超过5.3%。 Detection of small-size cracks in the weld of alloy steel by Time of Flight Diffraction (TOFD) is controlled by pulse width and other factors, and the diffracted waves at the upper and lower ends of the crack will be aliased and lead to cracks Highly quantitative difficulties. Autoregressive Spectral Extrapolation (ARSE) is used to process the aliasing signal to extend the effective frequency range and achieve the purpose of compressing the time-domain signal pulse width and improving the temporal resolution. The results show that the combination of ARSE technology can effectively separate the aliasing signal and achieve the quantitative detection of TOFD with 1.0mm depth at a depth of 50.0mm in the 100.0mm alloy steel with a relative quantitative error of less than 5.3%.