Applied Composite Materials最新文献

This study investigates the heating (i.e. discharging) of Lithium-ion (Li-ion) polymer batteries (e.g. pouch and 18650 cells) embedded in sandwich composites made of carbon fibre laminate facesheets and polymer foam cores (Polyvinyl Chloride or PVC, Polyethylene Terephthalate or PET). The effects of facesheet thickness, foam core thickness and density, and battery type and orientation on the heating of sandwich composites are systematically investigated. Heat can be rapidly dissipated from sandwich composites when the Li-ion polymer battery has a large area of contact with the carbon fibre facesheets. However, rapid internal heating, potentially leading to thermal runaway and fire, may occur when the battery is fully embedded within the foam core and physically separated from the facesheets. The optimal foam core thickness to prevent overheating can be predicted using the numerical thermal design maps. This study builds on our previous work which investigated the thermal performance of monolithic carbon fibre laminates.

The progressive fatigue damage (PFD) model effectively simulates the fatigue behavior of laminated composites under multiaxial cyclic stress. This model employs the generalized material property degradation (GMD) technique to calculate the residual properties of unidirectional (UD) plies subjected to cyclic stress. The present study enhances the PFD model to simulate the fatigue behavior of polymer matrix composite (PMC) materials under cyclic stress at various temperatures by executing it at a single temperature. The equivalent cycle time (ECT) method evaluates property changes in PMCs across different temperature settings, utilizing data from isothermal loading. In the present study, a novel approach based on the ECT concept, termed equivalent cycle number (ECN), is developed and integrated into the GMD technique. Additionally, a combined fatigue life model is employed to improve the predictive capability of the PFD model. This model is constructed by evaluating the results of three commonly used fatigue life models in predicting the fatigue life of UD plies under uniaxial cyclic stress at both room and elevated temperatures. The proposed PFD model effectively predicts the residual properties and fatigue life of a PMC subjected to multiaxial cyclic stress at two distinct temperatures. The findings demonstrate that the ECN method significantly reduces the model's computing load while maintaining a high level of predictive capability compared to available experimental data. Furthermore, the results indicate that using the combined fatigue life model substantially enhances the predictive capability of the PFD model.

An analytical model is developed to predict the ballistic performance of fiber-reinforced plastic (FRP) laminates under normal impact of rigid penetrators with various nose shapes. The model formulation is based on the localized interaction model incorporated with the spherical cavity-expansion model. Experimental validation of the analytical model is performed on experimental data obtained by own ballistic tests on Twaron/epoxy laminates and previous studies on ballistic performance of other FRP laminates. The model predictions for the ballistic limits and residual velocities are in good agreement with the experimental data, with discrepancies remaining within 10%, demonstrating the robustness and reliability of the present model.

This study employs numerical methods to model through-the-thickness reinforcements in CFRP tubular structures under axial impact, investigating the influence of reinforcement configurations on crashworthiness performance. Experimental validation involves testing unpinned tubular structures to establish a baseline model. LS-DYNA finite element models simulate low-velocity axial impacts, incorporating energy-based tiebreak contacts or solid cohesive elements to describe interlaminar bridging. Through-the-thickness are introduced through a homogenous mesh system or locally refined mesh at pin locations. Various reinforced tube designs with different pin diameters and areal densities are examined to identify the optimal pinned design for crashworthiness. The research demonstrates numerically that pinning enhances crashworthiness performances in axial crushing of composite tubes.

Carbon fibre reinforced polymers (CFRP) are increasingly being used in different industries, including the automotive and aerospace sectors. One important reason for this is because they have interesting structural and mechanical properties compared to metallic materials. Their high strength-to-weight ratio makes them a preferred choice for high-stress applications. However, CFRPs are often subjected to various defects during their manufacturing that can significantly alter their structural integrity and durability. Amongst these defects, the occurrence of void formation (known as porosity) is the most common. Many methods have been developed for the characterisation of porosity including the ones based on the use of ultrasound data. The present work aims at providing a comprehensive review of the application of machine learning (ML) techniques to the mapping and characterisation of porosity across CFRP composites. The types of ML used, and their potentials for improving the accuracy of porosity detection are presented and discussed. It is particularly noted that ML techniques can extract unique features from CFRP complex ultrasound data with a relatively good level of accuracy. This result suggests that these techniques, particularly the convolutional neural network (CNN), would overcome the limitations of traditional signal processing techniques.

This study presents a novel experimental methodology designed to assess damage in woven glass fibers reinforced polyamide 6,6/6 composites, specifically subjected to low-velocity impact and cyclic tensile loading. Conventional ultrasonic testing techniques often fail to detect subtle material degradation, particularly when dealing with barely visible impact damage (BVID), which can go unnoticed but still significantly compromise structural integrity. In contrast, the proposed approach utilizes multi-directional ultrasonic Lamb wave analysis, a more advanced technique that offers greater sensitivity and precision in identifying damage at various stages of the composite’s lifespan. In this work, a damage indicator is defined based on the velocity profile of Lamb waves, which are sensitive to changes in material properties such as stiffness degradation. The Lamb wave-based methodology is rigorously validated through detailed comparisons with X-ray tomography. These comparisons reveal strong correlations between the two techniques, highlighting the effectiveness of the proposed ultrasonic approach in detecting BVID. Moreover, the study demonstrates that this methodology is not only highly sensitive but also scalable, making it suitable for industrial applications where automated inspection of composite components is essential. The proposed method offers a significant advancement in non-destructive testing (NDT) techniques based on Lamb wave diagnostic tools in composite material testing.

Increasing power density and power consumption in electronic devices necessitate heat-dissipating components with high in-plane and cross-plane thermal conductivity to prevent overheating and enhance performance and reliability. Traditionally, polymer composites are made by incorporating rigid, high thermally conductive fillers within the polymer matrix. However, the filler loadings required to achieve significant thermal conductivity enhancement can impact the mechanical properties of the material system, often making them significantly more rigid or brittle than the base polymer. In this study, we developed a method to incorporate an ultra-high molecular weight polyethylene (UHMWPE) three-dimensional fiber mat and eutectic gallium indium alloy (EGaIn) liquid metal into an epoxy matrix. We integrated in-plane and cross-plane thermal conductivity measurements with flexural modulus assessments to understand the impact of the high thermal conductivity fillers on the thermal and mechanical response of the material. This approach enhances both the in-plane and cross-plane thermal conductivity of the composite, achieving thermal conductivities three times higher in the cross-plane direction and six times higher in the in-plane direction compared to the base polymer. Moreover, mechanical characterization reveals that the mechanical performance of the composite is comparable to that of a fiber-reinforced polymer composite, and the incorporation of liquid metal does not significantly impact stiffness, even at high filler loadings. This work demonstrates the potential of strategic composite design to achieve polymeric materials with optimized thermal-mechanical coupling. These new materials offer a solution to the challenges posed by higher power consumption in electronics, providing improved heat dissipation capabilities for more reliable devices.

Aiming to improve the impact resistance of structures, a technique to prepare carbon fiber composite laminates based on bonding methods was proposed. CFRP laminates with different numbers of bonding surfaces were fabricated, and then the impact toughness and impact bearing capacity of the laminates were evaluated based on pendulum and drop hammer impact tests, respectively. Results showed that the introduction of bonding surfaces noticeably reduced the generation of manufacturing defects within CFRP. Moreover, the bonding surface at 1/2 thickness as a neutral layer was conducive to the energy absorption of composite layers and the bonding surfaces at 1/4 and 3/4 thickness prevented the extension of shear cracks. As a result, compared to the specimen B0 without bonding surface internally, the specimen B1 with only one bonding surface at 1/2 thickness showed the best impact resistance in the drop hammer impact test. The peak force increased by 8.6% and the energy absorption before failure was increased by 30.6%. When three bonding surfaces were introduced and distributed at 1/4, 1/2 and 3/4 thickness, the best impact toughness was obtained in the pendulum impact test. The impact absorption work and impact toughness of the specimen B3 increased by 19.61% and 17.02%, respectively. However, the specimen B2 with two bonding surfaces distributed at 1/4 and 3/4 thickness showed poor impact bearing capacity. In conclusion, both B1 and B3 showed advantages in terms of impact resistance, proving the beneficial effects of the introduction of bonding surfaces with a reasonable distribution on CFRP.

This paper presents an equivalent damage model for efficiently predicting compression-after-impact (CAI) behaviors of laminated composites, based on the intelligent numerical reconstruction of impact-induced damage. Using the k-means +  + clustering algorithm, the 3D spatial distribution of delamination is quantitatively identified from C-scanning time-of-flight (TOF) images and then discretized into a numerical mesh along with the soft inclusion. With the incorporation of interlaminar and intralaminar damage models, the CAI residual strength, failure modes, and damage scenarios of the composite laminate after low-velocity impact are predicted, showing good agreement with the experimental results at various impact energies. The proposed model enables fast evaluation of CAI strength within 1.5 h, without requiring impact energy information and maintaining accuracy, which is beneficial for application in damage tolerance design and optimization of engineering laminated structures.

Despite significant progress in additive manufacturing, processing defects remain a persistent challenge. Artificial intelligence (AI) enabled early defect detection and process optimization is promising solution for this problem. In this study, real image data from a composite 3D printing setup was used to evaluate the anomaly detection performance of three models: Autoencoder, Support Vector Machine (SVM), and the Zero Bias Deep Neural Network (DNN). The results demonstrate that the Zero bias model achieved an accuracy of 97.96%, significantly outperforming the Autoencoder (93.38%) and SVM (89.80%). Multiple thresholds in the zero bias model enable explain ability.