Estimation of trial parameters for Pulse Phase Thermography with low power heat sources

31st UIT (Italian Union of Thermo-fluid-dynamics) Heat Transfer Conference 2013

Vitali, L.^1 ; Fustinoni, D.^1 ; Gramazio, P.^1 ; Niro, A.^1

Energy Department, Politecnico di Milano, Campus Bovisa, Via Lambruschini 4, 20156 Milan, Italy^1

关键词: Characteristic frequencies;    Commercial codes;    Defect identification;    Non destructive testing;    Pulse phase thermography;    Pulsed phase thermography;    Surface cooling;    Surface proprieties;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/501/1/012008/pdf
DOI  :  10.1088/1742-6596/501/1/012008

来源: IOP

【 摘 要 】

Non-destructive Testing by Infrared Thermography (IR-NDT) is a widely adopted technique to reveal the presence of defects, i.e. discontinuity zones of thermal proprieties, inside materials. Pulsed Phase Thermography (PPT) is one of the most interesting techniques among IR-NDT: the specimen is heated by a thermal pulse and the sequence of thermograms of the surface cooling is transformed with the Discrete Fourier Transform (DFT). The resulting phase images (phasegrams) show little sensitivity to irregular heating and surface proprieties, and allow better defect identification by increasing the contrast. It is also possible to estimate the depth of the defect by correlating a characteristic frequency to the thermal diffusion length of the defect. The outcome of this analysis depends on the fine tuning of the technique and the appropriate choice of the parameters of the thermal pulse, namely length and power, as well as of the acquisition: frequency and observation time. While there are in literature a few guidelines for the choice of these parameters, a good knowledge of the technique and a certain degree of guessing is still required, especially when low heating power, longer pulses and small and deep defects are involved. This paper reports a method to estimate these parameters, partly based on theoretic considerations and partly on numerical simulations performed by means of a FEM commercial code on a 2D axial-symmetric model. Experimental results are also here presented, focusing on the difference between a thick plate and a thin one.