Lab-based in-situ micro-CT observation of gaps in prepreg laminates during consolidation and cure

Kratz, James; Galvez-Hernandez, Pedro; Pickard, Laura Rhian; Belnoue, Jonathan P.-H.; Potter, Kevin D

doi:10.1016/j.compositesa.2020.106180

Cited by 16 publications

(6 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…XCT has been applied to observe crack growth during mechanical loading, 23 defect characterisation, 24,25 and even in situ observation of manufacturing processes. 26,27 XCT generates a greyscale image where the intensity of each voxel (3D pixel) is related to the density of the material through the Linear Attenuation Coefficient (LAC). 28 In the case of uncured composite samples, three phases are present, having each a characteristic greyscale intensity linked to the density of their constituents: 1) Interlaminar voids, formed by the entrapped air between two consecutive plies, are represented by dark pixels due to the low density of the air, 2) Fibres and resin-saturated areas have high density and therefore are displayed with brighter grey values, 3) Dry areas represent the unsaturated fibre bed and feature a mixture of air and fibres, as the spaces between fibres have yet to be filled with resin.…”

Section: Sample Preparationmentioning

confidence: 99%

The effect of convolutional neural network architectures on phase segmentation of composite material X-ray micrographs

Galvez-Hernandez

Kratz

2023

Journal of Composite Materials

Self Cite

View full text Add to dashboard Cite

Porosity severely reduces the mechanical performance of composite laminates and methods for automatic segmentation of void phases are growing. This study investigates porosity in composite materials that take the form of interlaminar voids and dry tow areas. Deep Learning was used for the segmentation of X-ray micrographs via the implementation of eight state-of-the-art Convolutional Neural Network (CNN) architectures trained with data sets containing twenty-five, fifty, and one-hundred images. The combination of hyperparameters providing the highest accuracy for each architecture and training set size was achieved through the optimisation of six relevant hyperparameters, including the cut-off probability applied to output probability maps. Additionally, the properties of the CNN architectures ( e.g., layer typology, connections, density…) were found to play a determining role, not only in the segmentation results but also in the associated computing effort. U-Net and FCDenseNet outperformed the FCN-8s, FCN-16, SegNet, LinkNet, ResNet18 and Xception CNN architectures. However, the CNNs generally outperformed the standard thresholding approaches, especially in sub-volumes containing low porosity (1.07%) where the influence on strength is very sensitive in high-performance composites. In low porosity samples, U-Net and FCDenseNet consistently segmented voids to 85% + accuracy, whereas thresholding was only half as accurate, at around 40%. The results provide a strong motivation to replace thresholding as a segmentation method for composite X-ray micrographs. In terms of efficiency, the reduced complexity of the U-Net network allowed for an average reduction of the training time (−36%) and prediction time (−17%) when compared to FCDenseNet.

show abstract

Section: Sample Preparationmentioning

confidence: 99%

The effect of convolutional neural network architectures on phase segmentation of composite material X-ray micrographs

Galvez-Hernandez

Kratz

2023

Journal of Composite Materials

Self Cite

View full text Add to dashboard Cite

show abstract

“…The ability of laboratory-based XCT to provide information from the 3D internal structure of composite samples has allowed the investigation of different aspects of the microstructure, [5][6][7] manufacturing process [8][9][10] and damage characterisation. 11,12 Each aspect requires careful selection of the XCT parameters for the particular CT system, such as magnification, exposure time, number of projections, tube accelerating voltage and power, in order to optimise the data visualisation and subsequent analysis.…”

Section: Introductionmentioning

confidence: 99%

“…A recent example of time-resolved in-situ XCT of composites manufacturing was used to observe the consolidation of lay-up gaps and porosity development during the cure cycle using a resolution of 13.5 µm/voxel and a scan time of 7 min. 8 If the scan time constraint was removed, a different selection of XCT parameters would allow for longer and higher-quality scans through improved spatial resolution and/or improved contrast to noise ratio, providing a more detailed characterisation of the microstructure. For example, a scan time of 45 min and a resolution of 6.56 µm/voxel was selected by Mehdikhani et al to analyse the void distribution in cured composite laminates with different stacking configurations.…”

Section: Introductionmentioning

confidence: 99%

The effect of X-ray computed tomography scan parameters on porosity assessment of carbon fibre reinfored plastics laminates

Galvez-Hernandez,

Smith,

Gaska

et al. 2023

Journal of Composite Materials

Self Cite

View full text Add to dashboard Cite

Combinations of X-ray Computed Tomography (XCT) scan times, from 30 s to 60 min, and voxel sizes, from 6 to 50 µm, were investigated for their effect on the porosity measurements of a unidirectional carbon fibre epoxy composite volume. The sample had a total void volume of around 2%, which is typical of the tolerance expected in the aerospace industry. The volume contained localised voids that create sub-volumes with representative high (5%) and low (1%) porosity regions. The ability to detect small-size voids in the lower porosity regions decreased as the voxel size increased. Scan resolutions above 25 µm resulted in a coarser segmentation and overestimation of the porosity due to the presence of partial volume effects. Scan times shorter than 2 min resulted in noisy images, requiring aggressive filtering that affected the segmentation of voids. Porosity segmentation was performed by thresholding and Deep Learning methods. Deep Learning segmentation was found to recognise noise better, providing more consistent and cleaner segmented data than thresholding. To capture micro-voids that contribute to porosity levels at the typical aerospace tolerance of 2%, scan parameters with a voxel size equal to or smaller than 25 µm, scan times of 2 to 8 min, and deep learning segmentation were found to be the most promising. These shorter scan times can be used to increase the productivity of CT scanning for porosity or observing time-resolved events. The data provided here contributes to the body of knowledge studying X-ray hardware settings and optimising image segmentation.

show abstract

“…However, most experimental evaluations to date have measured the shape of composite samples after curing and have been unable to accurately evaluate shape changes during consolidation. Recently, in situ measurement using X-ray computed tomography has been conducted [17], but the sample size is limited, and measuring consolidation deformation in a high-pressure environment similar to that of an autoclave is difficult. For evaluating the cure-induced deformation of composite materials, embedded optical fiber sensors are effective [18][19][20].…”

Section: Introductionmentioning

confidence: 99%

Strip-Type Embeddable Shape Sensor Based on Fiber Optics for In Situ Composite Consolidation Monitoring

Minakuchi

Niwa

Takeda

2022

Sensors

View full text Add to dashboard Cite

Carbon fibers and resin used in manufacturing carbon fiber-reinforced plastic composite structures flow before the resin solidifies, resulting in disrupted fiber orientation and non-uniform thickness. This process, known as consolidation, is critical for the quality of the composite structure, but no technology exists to measure the deformation in situ. This study proposes a strip-type embeddable shape sensor based on fiber optics for in situ monitoring of consolidation deformation. The sensor consists of a thin, flexible sheet with optical fibers embedded in the upper and lower surfaces of the sheet, and it can monitor out-of-plane bending deformation in composite materials during consolidation. Finite element analysis and experiments are used to evaluate the basic performance of the shape sensor before it is applied to composite gap/lap monitoring. For the first time, the relaxation of consolidation deformation due to the flow of fiber-resin suspension is measured. The proposed sensor will be a powerful tool for elucidating consolidation mechanisms and for validating composite manufacturing simulations.

show abstract

Lab-based in-situ micro-CT observation of gaps in prepreg laminates during consolidation and cure

Cited by 16 publications

References 38 publications

The effect of convolutional neural network architectures on phase segmentation of composite material X-ray micrographs

The effect of convolutional neural network architectures on phase segmentation of composite material X-ray micrographs

The effect of X-ray computed tomography scan parameters on porosity assessment of carbon fibre reinfored plastics laminates

Strip-Type Embeddable Shape Sensor Based on Fiber Optics for In Situ Composite Consolidation Monitoring

Contact Info

Product

Resources

About