In cross-ply laminates under loading, multiple intralaminar (transverse) cracking in the 90-layers causes laminate stiffness reduction which can be predicted by analytical models studying average of the stress perturbation between transverse cracks. The commonly used analytical models such as shear-lag type or variational-type models assume that stress in damaged 90-layers does not depend on the laminate thickness coordinate. But with increase in crack density, the stress perturbations created by transverse cracks start to interact, generating high stress gradients in through-the-thickness direction in the 90-layers. Location of a new crack and features of its propagation in a high stress gradient region between two cracks in a cross-ply laminate are analyzed in this paper. FEM is used to calculate stress distribution between two cracks. The location of the next crack and its initial orientation is found using principal stresses and orientation of principal axes. Most often these cracks are initiated at ply interface in the middle between already existing cracks. Their propagation across the layer thickness is analyzed using energy release rate based criterion. It is shown that at very high crack density the crack propagation across the layer thickness is unstable in the beginning, but it is stopped when approaching the middle of the layer where the crack opening becomes zero. Presence of local delaminations at the tips of existing cracks changes the energy release rate for propagation of new transverse cracks.

In-plane thermo-elastic constants of symmetric damaged laminates containing transverse cracks in plies and local delaminations starting from crack tip are predicted using a crack opening (COD) and crack sliding displacement (CSD) based approach. An exact elastic analysis shows that the displacement gap on the delamination crack surfaces does not enter the stiffness expressions explicitly. The delamination affects the stiffness via larger COD and CSD of the intralaminar crack. This means that the same expressions for cracked laminates with and without delaminations can be used but with different expressions for COD and CSD. Finite element method is used to analyze the CSD dependence on delamination length and crack density. The obtained approximative expressions for CSD are in a good agreement with FEM. It is shown that in cases when it depends on CSD only, the predicted shear modulus of laminates is in an excellent agreement with direct FEM calculations. The used homogenization over couples of off-axis plies (monoclinic materials) in CSD expressions for balanced laminates is validated.

Cellulose filaments with a unique microtape-like shape (width/thickness ratio ≈ 40) were studied for their potential as reinforcement in bio-epoxy resin with the hypothesis that their gross-sectional geometry would allow for efficient stress transfer and a higher transverse modulus. The microstructure of the spun filaments, orientation of the cellulose fibrils within them, and their mechanical properties were analyzed. Unidirectional (UD) composites with bio-epoxy resin were fabricated using vacuum infusion resulting in filament content of ≈ 23 vol% and low density 1.18 gcm−3.

Wide-angle X-ray scattering showed that the cellulose fibrils were relatively well aligned in the filament axis, having an orientation factor of 0.78. The axial filament modulus was measured to 33 GPa, the in-plane transverse modulus to 12 GPa and the axial strength was approx. 380 MPa. The longitudinal E-modulus of the UD composites was measured to 10 GPa and the strength to 120 MPa, both 3 times higher than the used bio-epoxy agreeing well with the estimated values. The transverse elongation at break of the UD composite was higher than typical values reported for glass-fiber epoxy composites, but the effect of filament shape on the transverse modulus was less significant than the predicted 4.5 GPa but still better than estimated for circular fibers. The lower property is likely due to the low filament content and the partially uneven in-plane filament arrangement. Simulations based on shear stress analysis suggest that the transverse properties of the UD composite could be improved by ensuring that the filament planes remain predominantly parallel in-plane during fabrication, and the overall mechanical properties could be improved by increasing the filament content.

Transverse cracks in the 90-layers in cross-ply laminates have singular stress state at the crack tips. This causes formation of fiber/matrix debonds which coalesce into a local delamination along the 0/90-layer interface. Various studies in the literature have predicted onset strains for local delaminations at transverse crack tips using energy-based criteria such as critical strain energy release rate (ERR) and critical generalized stress intensity factors. Although similar ERR-based analyses have been carried out to predict the delamination growth as well, a systematic parametric analysis is lacking. Such systematic analysis of parameters that can affect the growth of local delaminations including geometrical parameters, elastic constants and transverse crack density is necessary to predict delamination growth under complex thermo-mechanical loading conditions. In the present work, FEM is used to carry out ERR-based analysis of the growth of local delaminations with different shapes in carbon-fiber epoxy and glass-fiber epoxy [0m,90n]s cross-ply laminates with the help of virtual crack closure technique and J-integral method. Firstly, the ratios of elastic constants and geometrical parameters that can prominently affect the ERR values are identified by a simple analytical routine. Then, the analytical predictions are verified using FEM for local delaminations growing symmetrically (I-shaped and C-shaped) or non-symmetrically (T-shaped and S-shaped) with respect to the laminate midplane. It is shown that a non-symmetrical I-shaped crack would always transition into a symmetrical I-shaped crack before any further delamination growth, if energetically viable. Finally, a simplified strategy to calculate ERR values for local delaminations growing under combined thermo-mechanical loading is presented. 

Viscoelastic (VE) behavior at elevated temperatures of ultra-high molecular weight polyethylene (UHMWPE) was studied in strain controlled tensile tests, finding that a thermo-rheologically simple linear VE model with temperature affecting only the shift factor gives sufficient accuracy. Indeed, master curve construction after the tests at different temperatures proved that only horizontal shift in log t axis was required. The Prony coefficients of the Master curve found at reference temperature and the dependence of the shift factor on temperature were input parameters in validation simulations of a complex loading ramp where the temperature and the applied strain were changing simultaneously, showing good agreement with the test results. The thermo-rheological simplicity of the UHMWPE implies no aging of the amorphous phase and not changing degree of crystallinity during long term tests, phenomena that were inspected in repeated tests and using Differential Scanning Calorimetry.

The effect of test temperature, maximum stress, and stress-ratio on transverse cracking development in cross-ply laminates subjected to tension–tension cyclic loading was analysed. A two-parameter Weibull distribution model was used to predict transverse cracking, wherein the Weibull scale parameter was assumed to be test temperature and number of cycles dependent. By introducing an equivalent stress in the model, it was possible to account for the effect of the stress ratio in cyclic loading over a range of different loading conditions. To verify the model, tests on temperature resistant cross-ply composites were performed at room temperature and at 150 °C with different stress levels and local 90-layer stress ratios. For both test temperatures, increase in stress level increased the transverse cracking tendency. At 150 °C, despite the lower maximum thermo-mechanical ply-stress level compared to room temperature, transverse cracking tendency was found to be higher.

This study investigates the fracture toughness of adhesive joints between carbon fiber-reinforced polymer composites (CFRP) and boron-alloyed high-strength steel under Mode I and II loading, based on linear elastic fracture mechanics (LEFM). Two adhesive types were examined: the excess resin from the prepreg composite, forming a thin layer, and a toughened structural epoxy (Sika Power-533), designed for the automotive industry, forming a thick layer. Modified double cantilever beam (DCB) and end-notched flexure (ENF) specimens were used for testing. The results show that using Sika Power-533 increases the critical energy release rate by up to 30 times compared to the prepreg resin, highlighting the impact of adhesive layer thickness. Joints with the thick Sika adhesive performed similarly regardless of whether uncoated or Al–Si-coated steel was used, indicating the composite/Sika interface as the failure point. In contrast, the thin resin adhesive layer exhibited poor bonding with uncoated steel, which detached during sample preparation. This suggests that, for thin layers, the resin/steel interface is the weakest link. These findings underline the importance of adhesive selection and layer thickness for optimizing joint performance in composite–metal hybrid structures.

Thermo-mechanical response of [90n/0m]s carbon/epoxy and glass/epoxy cross-ply laminates in 4-point bending is analyzed experimentally and analytically. Intralaminar cracks in surface 90°-plies and local delaminations introduced in one of the 90°-plies at large deflections reduce the laminate bending stiffness and make the laminate asymmetric due to differences in the damage state in the layers. The latter leads to residual thermal curvature that increases with intralaminar crack density and with growing local delaminations. In the present study optical microscopy was used for crack density quantification. It was also found experimentally that small local delaminations develop in the initial stage of damage evolution and under increasing load they grow rapidly from the existing and newly created crack tips. The effect of damage on residual curvature and the bending stiffness was analyzed using an analytical method, where the concept of the effective stiffness of damaged ply is used in the classical laminate theory. Analytical results were validated with a 3-D FEM simulation of the damaged laminate in a 4-point bending test. In the literature a phenomenon that the microdamage in laminate layers causes redistribution of in-plane thermal stresses is often overlooked. The present paper shows that the used analytical approach gives an accurate description of experimental results regarding two independent sets of data: the residual curvature; and the laminate bending stiffness with evolving micro-damage. The present study also renders a better insight in the mechanics of the phenomena and allows estimation of the extent of local delaminations that is difficult to measure in tests.

Accumulation of transverse cracks in carbon fiber heat resistant polymer (with bismaleimide formulation) cross-ply laminates during tensile loading at elevated temperatures and after long heat treatment is analysed. Data shows that both the iso-thermal heat treatment and testing at elevated temperatures reduce the transverse cracking resistance. A two-parameter Weibull failure stress distribution model with scale parameter degrading with heat treatment and elevated temperature is used for crack initiation analysis. The degradation is described by polynomial expansion including interaction terms. Data shows that the scale parameter dependence on the heat treatment time and the test temperature is rather linear. The same expansion parameters have been successfully used for laminates with the same constituents but with a different layup and fiber content.

Presented test results on transverse cracking in cross-ply laminates upon tension–tension cyclic loading show that the increase of crack density depends not only on the maximum transverse stress in the cycle but also on the local cyclic stress ratio RTloc in the analyzed layer. To include the effect of the RTloc in the model with statistical failure stress distribution for crack initiation (based on Weibull distribution) adapted for fatigue, an equivalent stress is introduced in a similar manner as the equivalent strain energy release rate has been used for delamination crack propagation. The equivalent stress in the layer is defined as a power function of the maximum stress and the stress ratio in the layer. It was found, testing laminates with two different fiber contents that higher the local stress ratio in 90-layer, higher the transverse cracking resistance. Transverse crack density simulation using the developed equivalent stress model has been validated against test results.