Fatigue failure of carbon fiber-reinforced plastics (CFRPs) under cyclic loadings has attracted the attention of researchers recently. In this study, the entropy-based failure criterion is proposed to investigate the fatigue lifetime of unidirectional CFRPs subjected to multiple-amplitude cyclic loadings. Due to the heterogeneity of CFRPs, a micro-finite element model considering matrix resin and fibers independently is developed, and the entropy-based damage criterion is implemented into a user-subroutine of Abaqus to model the progressive damage of matrix resin. The fatigue lifetime of CFRPs under typical loading sequences consisting of two stages, such as varying from low to high (L-H) or from high to low (H-L) loading sequence, is estimated with the proposed failure criterion. Numerical results show that the initial damage occurs near the area between two fibers, and a transverse crack propagates progressively under the cyclic loading. The difference in predicted lifetime to final failure in L-H and H-L stress levels is 6.3%. Thus, the effect of loading sequence on the fatigue lifetime can be revealed via the proposed entropy-based damage criterion. Comparisons with the conventional linear cumulative damage (LCD) and kinetic crack growth (KCG) theories are also conducted to demonstrate the validity of the proposed method. The entropy-based failure criterion is a promising method to predict the residual strength and fatigue lifetime of CFRP components.

The transverse cracking behavior of a carbon-fiber-reinforced plastic (CFRP) cross-ply laminate is investigated using a fatigue test and an entropy-based failure criterion in this study. The results of fatigue experiments show that the crack accumulation behavior depends on the cyclic number level and frequency, in which two obvious transverse cracks are observed after 10⁴ cyclic loads and 37 transverse cracks occur after 10⁵ cycles. The final numbers of transverse cracks decrease from 29 to 11 when the load frequency increases from 5 Hz to 10 Hz. An entropy-based failure criterion is proposed to predict the long-term lifetime of laminates under cyclic loadings. The transverse strength of 90° ply is approximated by the Weibull distribution for a realistic simulation. Progressive damage and transverse cracking behavior in CFRP ply can be reproduced due to entropy generation and strength degradation. The effects of stress level and load frequency on the transverse cracking behavior are investigated. It is discovered that, at the edge, the stress σ₂₂ + σ₃₃ that is a dominant factor for matrix tensile failure mode is greater than the interior at the first cycle load, and as stress levels rise, a transverse initial crack forms sooner. However, the initial transverse crack initiation is delayed as load frequencies increase. In addition, transverse crack density increases quickly after initial crack formation and then increases slowly with the number of load cycles. The proposed method’s results agree well with those of the existing experimental method qualitatively. In addition, the proposed entropy-based failure criterion can account for the effect of load frequency on transverse crack growth rate, which cannot be addressed by the well-known Paris law.

Experimental methodologies for fatigue lifetime prediction are time-intensive and susceptible to environmental variables. Although the cohesive zone model is popular for predicting adhesive fatigue lifetime, entropy-based methods have also displayed potential. This study aims to (1) provide an understanding of the durability characteristics of carbon fiber-reinforced plastic (CFRP) adhesive joints by incorporating an entropy damage model within the context of the finite element method and (2) examine the effects of different adhesive layer thicknesses on single-lap shear models. As the thickness of the adhesive layer increases, damage variables initially increase and then decrease. These peak at 0.3 mm. This observation provides a crucial understanding of the stress behavior at the resin–CFRP interface and the fatigue mechanisms of the resin.

Experimental methodologies for fatigue lifetime prediction are time-intensive and susceptible to environmental variables. While the cohesive zone model is popular for predicting adhesive fatigue lifetime, entropy-based methods have also shown promise. This study aims to provide an understanding of the durability characteristics of CFRP adhesive joints by incorporating an entropy damage model within the context of the finite element method (FEM), examined the effects of different adhesive layer thicknesses on SLS models. As the adhesive layer increased in thickness, DAMAGE VARIABLES initially rose and then declined, peaking at 0.3mm. This finding provides crucial understanding into the stress behavior at the resin-CFRP interface and the resin's fatigue mechanisms.

Compressive strength is one of the most important properties of carbon fiber reinforced plastics (CFRP). In this study, a new method for predicting the axial compressive strength of CFRP using the response surface method is developed. We focused on a microbuckling model to predict the compressive strength of unidirectional fiber composites. For the microbuckling model, axial shear properties are required. To obtain the compressive strength for various material properties, we perform individual shear tests and numerical simulations, but these require enormous computational costs and extended time. To address the issue of computational cost, in this study, we propose a new method to predict compressive strength using the response surface method. First, we perform shear simulation in a microscale fracture model for unidirectional CFRP with various parameters of the fiber and resin properties. Based on the results of the shear simulation, the response surface method is used to evaluate and develop prediction equations for the shear properties. This method allows for the study of the objective values of the parameters, without significant computational effort. By comparing both the results predicted from the response surface method (RSM) and the simulation results, we verify the reliability of the prediction equation. As a result, the coefficient of determination was higher than 94%, and the validity of the prediction method for the compressive strength of CFRP using the response surface method (RSM) developed in this study was confirmed. Additionally, we discuss the material properties that affect the compressive strength of composites comprised of fibers and resin. As a result, we rank the parameters as follows: fiber content, elastic modulus after resin yield, yield stress, and initial elastic modulus.