Modelling and evaluation of pulsed and pulse phase thermography through application of composite and metallic case studies

摘要

A transient thermal finite element model has been created of the pulsed thermography (PT) and pulse phase thermography (PPT) experimental procedure. The model has been experimentally validated through the application of four case studies of varying geometries and materials. Materials used include aluminium, carbon fibre reinforced plastic (CFRP) and adhesively bonded joints. The same four case studies have also formed a basis for comparison between three experimental techniques: PT, PPT and the more established ultrasonic (UT) c-scan. Results show PPT to be advantageous over PT due to its deeper probing as it is less influenced by surface features. Whilst UT is able to reveal all the defects in these case studies, the time consuming nature of the process is a significant disadvantage compared to the full field thermography methods. Overall, the model has achieved good correlation for the case studies considered and it was found that the main limiting factor of the PT model accuracy was knowledge of thermal material properties such as conductivity and specific heat. Where these properties were accurately known the model performed very well in comparison with experimental results. PPT modelling performed less well due to the method of processing the PT data which aims to emphasise small differences. Hence inaccuracies in inputted values such as material properties have a much greater influence on the modelled PPT data. The model enables a better understanding of PT and PPT and provides a means of establishing the experimental set-up parameters required for different components, allowing the experimental technique to be appropriately tailored to more complex situations including bonded joints or structures where several materials are present. The paper ends with a section on defect detectability based on thermal diffusivity contrast between the defect and the bulk material. It shows that in aluminium, because of its higher conductivity, greater thermal contrast is achieved for small differences in diffusivity. Regions where the diffusivity ratio between defect and bulk materials was insufficient to provide thermal contrast for defect identification were found. PPT phase data is shown to reduce the extent of such regions increasing the detectability of defects. Effusivity is introduced as a means of determining the thermal contrast between the defect and non-defective areas and hence establishing the defect detectability.