The longitudinal compressive behavior of unidirectional composite laminates with fiber waviness is highly complex and plays a crucial role in determining final failure of composites. Evaluating this behavior, especially considering nonlinear failure mechanisms like kink-band formation, typically requires computationally expensive finite element analysis, which is impractical for large-scale quality inspection. To address computational challenges, this paper develops a new computational framework using Convolutional Neural Network (CNN) models, providing an ultra-efficient prediction of the entire stress–strain curve of composites with fiber waviness. The CNN models were trained on simulation data generated from an experimentally validated mesoscale finite element model. The microstructures of the composites with fiber waviness were taken from realistic micrographs, resulting in diverse stress–strain curves. The proposed CNN models showed high accuracy and efficiency for predicting the highly nonlinear stress–strain curves of the composites, which can be employed as a real-time evaluation method of the criticality of fiber waviness.

This paper presents a user-friendly plugin for Abaqus to facilitate damage simulation that may require a state-of-the-art damage model for composite materials. The plugin can perform interactive model calibration, set up single-elements required for model verification, and display their respective stress–strain curves. This tool breaks down the use of composite damage models, reducing the chances of user errors and making the model more accessible to end-users. 

Experimental and numerical studies were carried out to characterize hybrid thin- and thick-ply composite laminates and assess modelling capabilities. Five different composite laminates were manufactured using a single material system with varying proportions of thin plies (0%, 50%, and 100% thin-ply). Bearing tests were performed and the results from the tests were investigated. The results showed that performance, in terms of bearing strength at onset of damage and ultimate bearing stress, increased proportionally with the increasing amount of thin plies within the laminate. Microscopic examination of the failure modes for all laminates was performed at the center of the hole to determine the dominant failure mode. The numerical investigation uses a highly detailed mesoscale model previously validated for crash simulations but never used successfully to bearing damage areas. The results showed a good correlation regarding both the load response and the morphology of damage. 

Two 3D homogenized models for damage growth in a unidirectional (UD) composite ply are simplified and merged into a unified model. The fibre kinking behaviour is based on fibre kinking theory handled in a finite deformation framework. The nonlinear shear behaviour is pressure dependent and is modelled by combining damage and friction on the fracture plane. Fibre kinking growth and transverse behaviour are modelled with a single damage variable. This allows both modes to occur simultaneously and mutually influence each other in an efficient and physically-based way. For validation the model is tested against micro-mechanical Finite Element (FE) simulations under pure longitudinal compression and influenced by shear. The results show nearly perfect agreement for stiffness, strength and crushing stress. The model validation is performed against two different components under three-point bending and a quasi-static crash scenario. Both simulation show good correlation with experiments, validating thus the present unified model. © 2022 The Author(s)

The fiber waviness is inevitable in non-crimp fabric (NCF) reinforced composites. It is very challenging to accurately and efficiently predict the material behavior with fiber waviness. This work presents a machine learning approach to the prediction of material behavior of NCF composites under a compressive load. The out-of-plane fiber orientations are first extracted from micrographs of NCF laminates. A digital twinning process is followed to create finite element (FE) models with elementwise fiber orientations. Based on the FE models, a physics-based damage model is employed to generate high-fidelity simulation datasets, capturing the kink-band due to the fiber waviness. With the simulation datasets, convolutional neural network (CNN) models are developed to take the images of the fiber orientations and predict the corresponding stiffness, strength, and stress-strain curves of the NCF composites. The results show that the CNN models can capture spatial information of the fiber orientation and efficiently predict the corresponding material behavior with a high accuracy. In addition, the correlations of the fiber orientations and the final material behaviors are investigated based on the developed CNN models. 

Simulation of damage in composite laminates using currently available three-dimensional finite element tools is computationally demanding often to the point that analysis is not practical. This paper presents an enriched shell element that can provide a computationally efficient means to simulate low-velocity impact damage in a composite. The enriched element uses the Floating Node Method and a damage algorithm based on the Virtual Crack Closure Technique that is capable of simulating progressive damage growth consisting of delamination and delamination-migrations from ply to ply during a dynamic impact load. This paper presents results from the shell model in a test-analysis correlation for impact testing of 7-ply and 56-ply laminates. Analysis results from a separate high-fidelity three-dimensional finite element analysis are included also for comparison in the case of the 7-ply laminate, but not in the case the 56-ply laminate due to excessive computational demand. This paper serves as the first application of both models in low-velocity impact simulation. The shell model is considerably more computationally efficient than the high-fidelity model by at least an order of magnitude and is shown to produce results, while not as accurate as the high-fidelity model, potentially sufficiently accurate for a wide range of engineering applications including structural design and rapid prototype assessments.

A mesoscale model for fibre kinking onset and growth in a three-dimensional framework is developed and validated against experimental results obtained in-house as well as from the literature. The model formulation is based on fibre kinking theory i.e. the initially misaligned fibres rotate due to compressive loading and nonlinear shear behaviour. Furthermore, the physically-based response is computed in a novel and efficient way using finite deformation theory. The model validation starts by correlating the numerical results against compression tests of specimens with a known misalignment. The results show good agreement of stiffness and strength for two specimens with low and high misalignment. Fibre kinking growth is validated by simulating the crushing of a flat coupon with the fibres oriented to the load direction. The numerical results show very good agreement with experiments in terms of crash morphology and load response.

This paper details a complete crush model for composite materials with focus on shear dominated crushing under a three-dimensional stress state. The damage evolution laws and final failure strain conditions are based on data extracted from shear experiments. The main advantages of the current model include the following: no need to measure the fracture toughness in shear and transverse compression, mesh objectivity without the need for a regular mesh and finite element characteristic length, a pressure dependency of the nonlinear shear response, accounting for load reversal and some orthotropic effects (making the model suitable for noncrimp fabric composites). The model is validated against a range of relevant experiments, namely a through-the-thickness compression specimen and a flat crush coupon with the fibres oriented at 45° and 90° to the load. Damage growth mechanisms, orientation of the fracture plane, nonlinear evolution of Poisson's ratio and energy absorption are accurately predicted.

A newly developed physically based model for the longitudinal response of laminated fibre-reinforced composites during compressive damage growth is implemented in a Finite Element (FE) software. It is a mesoscale model able to capture the physics of kink-band formation by shear instability, the influence of the matrix in supporting the fibres and the rotation of the fibres during compression, resulting in more abrupt failure for smaller misalignments. The fibre kinking response is obtained by solving simultaneously for stress equilibrium and strain compatibility in an FE framework. Strain softening creates pathological sensitivity when the mesh is refined. To make the model mesh objective, a methodology based on scaling the strain with the kink-band width is developed. The FE implementation of the current model is detailed with focus on mesh objectivity, and generalized to irregular meshes. The results show that the current model can be used to predict the whole kinking response in a 3D framework and thus account for the correct energy dissipation.

A material model to predict kink-band formation and growth under a 3D stress state is proposed. 3D kinking theory is used in combination with a physically based constitutive law of the material in the kink-band, accounting for friction on the microcracks of the damaged material. In contrast to existing models, the same constitutive formulation is used for fibre kinking and for the longitudinal shear and transverse responses, thereby simplifying the material identification process. The full collapse response as well as a crush stress can be predicted. The model is compared with an analytical model, a micromechanical finite element analysis and crushing tests. In all cases the present model predicts well the different stages of kink-band formation and crushing.