论文部分内容阅读
内部缺陷的反问题定量识别是红外无损检测技术的重要内容,而红外稳态识别方法存在受误差影响大等缺点。文中对具有矩形内部缺陷的试件建立了二维物理和数学传热模型,通过有限体积法求解瞬态温度分布,分析了检测表面的温度分布规律,运用Levenberg-Marquardt法研究了瞬态情况下对缺陷的尺寸和方位进行定量识别的方法。数值算例证明了算法的有效性;初始假设等对识别结果的影响不大;最大检测温差越大,识别结果越准确;瞬态测温法便于得到较大的检测温差,有利于识别结果的准确性。
The quantitative identification of the inverse problem of internal defects is an important part of the infrared non-destructive testing technology. However, the infrared steady state identification method has the shortcoming of being affected greatly by the error. In this paper, a two-dimensional physical and mathematical heat transfer model was established for the specimens with rectangular internal defects. The transient temperature distribution was solved by the finite volume method. The temperature distribution of the test surface was analyzed. Levenberg-Marquardt method was used to study the transient state Methods for quantitatively identifying the size and orientation of defects. The numerical examples show the effectiveness of the algorithm. The initial assumption has little effect on the recognition results. The larger the maximum detection temperature difference, the more accurate the recognition result. The transient temperature measurement method is convenient for obtaining large detection temperature difference, accuracy.