基于有限元模拟的超声应力系数标定及分析*

应力系数的标定作为超声应力检测最为关键的环节，直接决定应力检测结果的准确性。传统的应力系数试验标定对于被测物的表面粗糙度、耦合剂厚度、声匹配块与被测物接触力等因素十分敏感，但缺少基本参照值。基于COMSOL建立多物理场耦合的超声应力检测模型，施加不同的拉伸载荷，计算临界折射纵波到达时间与不同应力值之间的关系，模拟标定45#钢的超声应力系数为13.7MPa/ns。单轴水平拉伸试验标定的45#钢应力系数为16.5MPa/ns。结果表明，通过两种方法标定的应力系数较为接近，试验标定的应力系数偏大，这是由于有限元方法能够消除试验过程中各种不确定因素对声时精确测量所造成的影响，能够更加纯粹的反映材料的声弹性效应，因此具有作为基础数据的参考价值。有限元方法作为传统试验方法的补充，可以减小试验标定数据的离散性，提高超声应力检测结果的可信度。

Calibration and analysis of ultrasonic stress coefficient based on finite element simulation

The calibration of stress coefficient, as the most important part of ultrasonic stress testing, directly determines the accuracy of stress testing results. The traditional stress coefficient test calibration is sensitive to the surface roughness of the measured object, the thickness of the coupling agent, and the contact force between the acoustic matching block and the measured object, but lacks the basic reference value. A multi-physical field coupling ultrasonic stress detection model was established based on COMSOL. Different tensile loads were applied, and the relationship between the arrival time of critical refraction longitudinal wave and different stress values was calculated, and the ultrasonic stress coefficient of 45# steel was calibrated. The stress coefficients of 45# steel are 13.7MPa/ns and 16.5MPa/ns, respectively, compared with those calibrated by uniaxial horizontal tensile test. The results show that the stress coefficients calibrated by the two methods are close to each other, but the stress coefficients calibrated by the test are larger. The finite element method can eliminate the influence of various uncertain factors on the accurate measurement of acoustic time in the test process, and can reflect the acoustic elastic effect of materials more purely, and its data has basic reference value. As a supplement to traditional test methods, finite element method can reduce the discreteness of test calibration data and improve the credibility of ultrasonic stress detection results.