Thermography and computational analysis of water ingress in honeycomb composite panels

Abstract

This research presents the results of experimental and numerical investigations on water ingress trapped in honeycomb panels. Ingress of atmospheric water in aircraft honeycombs may cause damage to aircraft. The percentage of water/ice filling honeycomb cells is an important factor in possible cell wall damage. This study is focused on the analysis of the following inspection parameters: 1) influence of panel orientation (horizontal, vertical and Inclined at 30° and 60°) on the efficiency of water detection, 2) efficiency and optimisation of a heating technique in evaluating water ingress, 3) influence of water/ice phase transformation on the detectability of water ingress, 4) quantifying of the water ingress. The numerical analysis was conducted by using the finite difference algorithms (ThermoCalc 3D), image processing algorithms (ThermoFit Pro), finite difference algorithms with only radiation heat transfer boundary condition ( ThermoCalc-3D-radiation), and Normalization image processing algorithm (ThermoDouble software), and the experiments were conducted by using active and passive infrared thermography to evaluate the detectability of water ingress and image processing in the cases where a test panel is placed in different spatial orientations. The samples with water and ice were tested and analysed using several data processing algorithms in the ThermoFit software to enhance water detection performance. The Maximum surface differential temperature signals (ΔT), running contrast (Cm) and their observation times (tΔT and tCm) were recorded and analysed for both samples with water and ice in the honeycomb cells. The signal-to-noise ratio (SNR) concept was used to compare the efficiency of image processing algorithms in inspecting water ingress in honeycomb panels with varying water content, spatial orientation and water/ice phase transformation. The computational results indicate that cells filled with water (100%) in a horizontal panel (180°) exhibit the highest differential temperature signal (∆Tm) of 30.1°C and a running contrast (Cm) of 0.88. In contrast, cells filled with 50% water in the same orientation show a ∆Tm of 27°C and a Cm of 0.14. For the panel in a vertical position, the recorded ∆Tm and Cm values are 30.0°C and 0.82, respectively. The inclined panel (60°) with 50% water-filled cells shows the optimal ∆Tm and Cm values of 30.1°C and 0.83. These computational trends are supported by the experimental data, which demonstrate that the horizontal panel (180°) filled with 100% water yields the highest ∆Tm of 6.5°C and Cm of 0.7. The same panel orientation with 50% water-filled cells results in a ∆Tm of 3.96°C and Cm of 0.59. The vertical panel shows a ∆Tm of 4.8°C and Cm of 0.55. The inclined panel (60°) with 50% water-filled cells yields optimal values of 3.71°C for ∆Tm and 0.58 for Cm. To assess the efficiency of data processing and improve water ingress detection (visibility), computing the Signal-to-Noise Ratio (SNR) during image processing is essential. Using a single image processing algorithm (Fourier phase 3rd harmonic) across all scenarios, the horizontal panel with 100% water-filled cells recorded an SNR of 88.1, while the vertical and inclined (60°) panels recorded SNRs of 21.9 and 20.2, respectively. The qualitative data indicate that the variation in ∆Tm and Cm across panel orientations is minimal, suggesting that panel orientation has little impact on the detection of water ingress. This implies that water visibility is primarily dependent on water content rather than the panel's orientation. Quantifying the amount of water trapped in honeycomb cells is crucial. From the calibration curve, it was observed that passive heating results in a maximum ∆Tm of 11°C, which remains constant as water content increases, due to the rapid thermal equilibrium between the panel and the environment. Active heating, on the other hand, generates a ∆Tm of 27°C, with a positive correlation to increasing water content. These findings suggest that active heating provides a more reliable calibration curve, while passive heating is better suited for detecting the presence of water ingress. The overall findings from this study make a valuable contribution to aircraft fuselage maintenance and provide useful data for aviation engineers to quantify water content in the cells of honeycomb panels