Microstructural variability in filament-wound composites: X-ray computed tomography and mechanical effect studies
Topic(s) : Special Sessions
Co-authors :
Shailee UPADHYAY (BELGIUM), Mahoor MEHDIKHANI (BELGIUM), Stephan LOMOV (BELGIUM), Dirk VANDEPITTE , Yentl SWOLFS (BELGIUM)Abstract :
Carbon fibre composite pressure vessels (CPVs) are commonly used for hydrogen storage in automotive applications. These vessels are manufactured by filament winding technique, which involves winding of long unidirectional fibres in different orientations, providing sufficient strength to withstand high internal pressures. But filament-wound composites suffer from microstructural variabilities and defects due to the limited applied fibre tension and environmental control. The microstructural defects can degrade the performance of the pressure vessel, thereby enforcing the use of large design safety factors [1].
In this study, we first defined the optimum 3D characterisation strategy to quantify microstructural defects in the pressure vessel in a statistically representative manner. High-resolution X-ray computed tomography was performed on small samples of a CPV. A sensitivity analysis was performed to determine the number and location of scanned samples to capture the size range and complexity of microstructural features. The statistical analysis revealed that at least four well-distributed samples need to be scanned at an optimum voxel size (Fig.1).
Second, microstructural defects are studied for various CPVs as a function of manufacturing parameters such as winding speed, liner pressure and resin take-up. The analysis revealed that the void characteristics are affected by manufacturing conditions due to different levels of layer compaction. The fibre misalignment and fibre volume fraction variation did not show any correlations to the manufacturing parameters at a global scale. However, locally the fibre misalignment around a void was affected by the complexity of the void shape. The fibre volume fraction was locally affected by high void fraction and large voids. This indicated the inter-dependency of the microstructural defects (see Fig.2(a) & (b)).
Based on this study, the relationship between the processing parameters and microstructural defects is established. Initial numerical results indicate a stress redistribution owing to the presence of microstructural defects. Further, the data obtained is used as input for performance simulation of the full vessel and ring tests are performed to determine the effect on the mechanical performance of the CPVs.
In this study, we first defined the optimum 3D characterisation strategy to quantify microstructural defects in the pressure vessel in a statistically representative manner. High-resolution X-ray computed tomography was performed on small samples of a CPV. A sensitivity analysis was performed to determine the number and location of scanned samples to capture the size range and complexity of microstructural features. The statistical analysis revealed that at least four well-distributed samples need to be scanned at an optimum voxel size (Fig.1).
Second, microstructural defects are studied for various CPVs as a function of manufacturing parameters such as winding speed, liner pressure and resin take-up. The analysis revealed that the void characteristics are affected by manufacturing conditions due to different levels of layer compaction. The fibre misalignment and fibre volume fraction variation did not show any correlations to the manufacturing parameters at a global scale. However, locally the fibre misalignment around a void was affected by the complexity of the void shape. The fibre volume fraction was locally affected by high void fraction and large voids. This indicated the inter-dependency of the microstructural defects (see Fig.2(a) & (b)).
Based on this study, the relationship between the processing parameters and microstructural defects is established. Initial numerical results indicate a stress redistribution owing to the presence of microstructural defects. Further, the data obtained is used as input for performance simulation of the full vessel and ring tests are performed to determine the effect on the mechanical performance of the CPVs.