چكيده به لاتين
Composite patches are used to repair the damaged metal and composite structures in different industries, especially in the aerospace industry. In order to ensure the health of the restoration, evaluation of the restored structure is a challenging topic. Thermography technique is one of the most powerful non-destructive testing methods that used to evaluate the restored structures. In the present study, the defects of the edge debonding type between the based aluminum structure and the carbon/epoxy patch made by layup configuration [04] and [08] have been investigated by step heating thermography method. Defects locate close to the patch edges because it is more likely that debonding onset in a repaired structure at edges in practice. Furthermore, detection of the edge defects is more difficult than the middle defects because of edge effects. In order to experimental simulation the debonding damage and achieve a defective sample close to reality, the edge debondings have been made from the three materials of teflon, polyurethane and air at the patch bonding place with the mother's structure. To achieve the correct inspection, as well as to investigate the testing limitations and to better understand the process of thermography testing, the finite element modeling of thermography testing was performed for all samples. The thermal loading values used in the finite element modeling have been extracted from a new method of empirical results in order to more exact numerical modeling. The step thermography results have been proccesed using one-dimensional processing methods including of; pulse phase thermography (PPT), thermography signal reconstruction (TSR), principle component analysis (PCA), one-dimensional wavelet transform, as well as two-dimensional processing methods including of; the depiction of the norm of the spatial derivatives of temperature, two-dimensional wavelet transform, edge detection algorithm of prewitt and low-pass ideal filter. Additionally, an one-dimensional processing algorithm has been developed with respect to the temperature images sequence and a morphological algorithm as a two-dimensional image processing. Also, for the first time, the 3D wavelet transform was used to identify the boundary of damaged edges in the field of thermography. In order to accurately interpret the results of a proper image or applying processing algorithms on it, for the first time, an instruction has been developed to find the best frame of raw or processed video results. For improve the results of defect detection, an effective sequential combination of two one-dimensional processing algorithms along with the morphological algorithm is presented. Finally, the dimensions, positions and materials of the debonding defect were compared in two layups of the patches, and the acceptable accordance was obtained between the experimental and numerical results. Also, a wide quantitative and qualitative comparison was done between the results of conventional algorithms of thermography and image processing, developed algorithms, and the combination of algorithms for all patches.
