Defect Detection And Evaluation Of Polyethylene Pipes Based On Active Infrared Thermography Technique

Polyethylene(PE)pipe is widely used in the piping systems,such as the transmission systems of coal gas,natural gas,city water,etc,for its reliable quality,safe operation,and easy maintenance.Butt fusion jointing is one of the common connection ways to join PE pipes.Some flaws,such as cracks,inclusions and porosity,often exist in joints of PE pipes which will ultimately lead to the failure of the PE pipes.Therefore,reliable non-destructive testing methods for PE pipe and its joint are necessary to improve the pipe safety.Currently,few researches have been focused on internal defect detection of PE pipe joint based on active Infrared thermography technique.In this thesis,finite element models were built for simulating transient heat transfer in defected PE Pipes.Infrared thermography experiments on different PE pipe defects were carried out by using the developed active infrared thermography system with different thermal excitations.We deeply investigated the application of the technique for the detection of defected PE pipes and also the recognition algorithms for the defects of PE pipes.Two finite element models were built to simulate transient heat transfer in PE Pipes with internal hole-defects,and also in the PE pipe joint with air-hole defects.In the computational model,constant wall temperature or constant heat flux boundary condition was applied to the internal surface of the PE pipes.The relationships between the temperature distributions and the sizes and positions of defects were established.After comparative analysis of temperature,temperature difference,and the rate of change of temperature difference between the center point at external surface of the defects and the reference point at non-defect external surface,the optimal parameters such as thermal excitation time and boundary conditions were obtained.An active infrared thermography system was developed for the verification of the FE model for defect detection of PE pipes.A series of PE Pipes with internal hole-defects were tested by the developed system.The FE simulation results well agreed with the one obtained from the infrared imaging experiments.The results indicate that finite element numerical method can be an effective method to analyze infrared imaging,and it can provide the numerical model for thermal imaging based defect detection and precise theoretical foundations of using infrared thermography in non-destructive testing.Meanwhile,two different PE pipe samples were comparatively investigated by using the developed active infrared thermography system:PE pipe joints with internal defects mimicked by different embedded materials,and PE pipe joint with internal cylindrical air-hole defects.For the thermal excitation of the infrared thermography system,we proposed to use two new methods by using an electrical heating bar or liquid nitrogen.The experimental results demonstrated that defects of PE pipes could be effectively detected by using the developed active infrared thermography with the proposed thermal excitations.However,the defect recognition from the experimental thermal images by using liquid nitrogen is more changeable since it is more difficult to control the cooling process by liquid nitrogen.In order to achieve automatic recognition of PE pipe defects from thermal images obtained,three image processing algorithms were investigated:Sobel edge detection,Otsu algorithm and fuzzy c-means algorithm.The comparison of the PE pipe defect recognition by these three algorithms was conducted and the algorithms suitable for automatic recognition of PE pipe defects were suggested.Effective defect shape of thermal images could be obtain by Sobel edge detection method.However,it has a poor anti-noise ability and the edge is discontinuous.Otsu algorithm is more suitable for the recognition of PE pipe Butt fusion joint defects,but it could not be employed in the defect detection of internal hole defects.Fuzzy c-means algorithm has the best applicability for obtaining excellent recognition of PE pipe joints' shape and size.The researches of defect detection of PE pipes based on active infrared thermography technique and images recognition algorithms provide theoretical foundation and experimental data for PE pipe defect detection and evaluation.