Detection of Abnormal Regions by Full-Field Non-Contact TemperatureMeasurement and Inverse Analysis with the Aid of the MathematicalAnalysis

在结构完整性评估中，材料劣化或异常性能的检测和评估是非常重要的。如果可以对与材料性质有关的某些物理量进行现场测量，就有可能建立一种可靠而准确的方法来估计材料性质的分布。在这项研究中，我们提出了一种在一定的热条件下由二维全场温度分布估计导热系数分布的方法。假设热通量在感兴趣的区域几乎是恒定的，热导率由温度梯度估计。对方形区域内的导热系数进行了数值模拟。热导率的估计分布与实际分布基本一致，但在热导率突变点附近，估计分布变差。由于该区域的通量不是恒定的，估计结果变差了。由温度梯度得到流线。用流线之间的宽度来修正导热系数的估计值。研究发现，利用流线宽度可以较好地估计导热系数的分布。

For the assessment of structural integrity, the detection and evaluation of deteriorated or anomalous material properties are very important. If full field measurement of some physical quantity related to the material property is available, it is possible to construct a robust and accurate method for estimating the distribution of material properties.In this investigation, we propose a method for estimating thermal conductivity distribution from two-dimensional full field temperature distribution under certain thermal conditions. Assuming that heat flux is nearly constant in the region of interest, thermal conductivity is estimated from the temperature gradient. Numerical simulations were made of the estimation of thermal conductivity in a square region. The estimated distribution of thermal conductivity agreed reasonably with the actual one, while the estimation deteriorated around a point where thermal conductivity showed abrupt change. The estimation deteriorated due to the fact that the flux was not constant in the region. Streamlines were obtained from the temperature gradient. The width between streamlines was used for modifying the estimated value of thermal conductivity. It was found that a better estimation of the thermal conductivity distribution was made by using the width of streamlines.