为了促进热障涂层红外无损检测的定量检测研究，建立了轴对称圆柱坐标下的热障涂层脉冲相位检测模型，针对研究模型，采用有限体积法求解出脉冲热激励下的温度场，将温度进行FFT变化得到相位分布，分析了不同因素对检测表面相位差分布的影响。在此基础上，采用LM算法研究了轴对称圆柱坐标下对热障涂层厚度的大小和脱粘缺陷的位置进行定量化检测的方法，分析了不同因素对检测结果的影响。研究结果表明：当不存在测温误差时，不同的初始假设、采样窗口时间下都能得到很高的识别精度，其对定量识别的影响不大，当测温仅存在均匀误差时，涂层厚度和脱粘缺陷位置识别精度都很高，均匀误差对识别无影响，识别结果的精度会随测温随机误差的增大而降低，但在较大的随机误差下仍有较高的识别精度。

In order to promote quantitative detection of thermal barrier coating by pulse phase, the pulse phase nondestructive testing model of thermal barrier coating in axisymmetric cylindrical coordinates was established. Temperature field was solved by using a finite volume method in the heat pulse excitation. And the phase of the model was accepted by FFT. The impacts of different factors on the phase were analysised. Based on this, quantitative detection of coating thickness and position of the bonding defect of thermal barrier coating were researched in axisymmetric cylindrical coordinates by using the method of LM. The impacts of different factors on quantitative detection were analysised. The research results show that the recognition accuracy of results are high in different initial assumption and different modulation time when there is no measurement error, the effect of initial assumption and modulation time on quantitative recognition is little, the recognition accuracy of results are high in different uniform temperature measurement error, uniform temperature error would not change the result of recognition, the accuracy of the identification results would decrease with the increase of the random temperature error, but there is still a high recognition accuracy under large random temperature error, the effectiveness of the quantitative identification method was proved.