Machine learning (ML) has emerged as a useful predictive tool based on mathematical and statistical relationships for various engineering problems. The pairing of structural health monitoring (SHM) and nondestructive evaluation (NDE) methods with ML algorithms has yielded beneficial results in addressing the damage state of a material or system. Damage state descriptions addressed with ML include detecting a damage mechanism, locating a mechanism, identifying the type of mechanism, assessing the extent of the damage mechanism, and estimating the useful remaining life of a material or system. Damage evaluation research of composite materials has progressed with the increased usage of composite structural elements in the aerospace industry. NDE methods are a viable candidate for pairing with ML algorithms to improve damage state monitoring of composite materials due to the complexity associated with the structure of composites. Fiber-reinforced polymers (FRP), for example, contain at least two constituent materials a fiber and matrix material whose mechanical behavior and interactions contribute to the performance of an FRP. Unlike conventional composite analytical models that require explicit information about the constituents and microstructure of a laminate, an ML algorithm can construct damage evaluation predictions when employing exclusively past operational performance or data from an SHM or NDE method. A researcher determines the type of data selected when applying an ML model for trend analysis, anomaly detection, or prediction making. However, no one specific input feature is required for utilizing an ML model, and examples of possible data features include material properties, physical dimensions, and collected evaluation data. In the present review, applications of ML combined with the damage state evaluation of composite materials, particularly examining FRPs, are discussed to demonstrate the predictive capabilities of ML and its viability for future applications, especially in industrial environments, to minimize costs and improve damage detection rates.

Bistability is exhibited by an object when it can be resting in two stable equilibrium states. Certain composite laminates exhibit bistability by having two stable curvatures of opposite sign with the two axes of curvature perpendicular to each other. These laminates can be actuated from one state to the other. The actuation from the original post-cure shape to the second shape is called as ‘snap-through’ and the reverse actuation is called as ‘snap-back’. This phenomenon can be used in applications for morphing structures, energy harvesting, and other applications where there is a conflicting requirement of a structure that is load-carrying, light, and shape-adaptable. MW Hyer first reported this phenomenon in his paper in 1981. He found that thin unsymmetric laminates do not behave according to the predictions of the Classical Lamination Theory (CLT). The CLT is a linear theory and predicts the post-cure shape of thin unsymmetric laminates to be a saddle. MW Hyer developed a non-linear method called the “Extended Classical Lamination Theory” which accurately predicted the laminate to have two cylindrical shapes. Since then, a number of researchers have tried to identify the key parameters affecting the behavior of such laminates. Geometric parameters such as stacking sequence, fibre orientation, cure cycle, boundary conditions, and force of actuation, have all been studied. The objective of this research is to define a relation between the length, width and thickness of square and rectangular laminates required to achieve bistability. Using these relations, a 36 in × 36 in bistable laminate is fabricated with a thickness of 30 CFRP layers. Also, it is proved that a laminate does not lose bistability with an increase in aspect ratio, as long as both sides of the rectangular laminate are above a certain ‘critical length’. A bistable laminate with dimensions of 2 in × 50 in is fabricated. Further, for laminates that are bistable, it is necessary to be able to predict the curvature and force required for actuation. Therefore, a method is developed which allows us to predict the curvature of both stable shapes, as well as the force of actuation of laminates for which the thickness and dimensions are known. Finite Element Analysis is used to carry out the numerical calculations, which are validated by fabricating laminates. The curvature of these laminates is measured using a profilometer and the force of actuation is recorded using a universal test set-up.

Bistable composite laminates have large-scale applications in morphing and energy-harvesting structures, but their fatigue performance remains largely unexplored. This study investigates the stiffness and damage progression and evaluates bistable performance to develop protocols for long-term applications. We analyze the effects of displacement-controlled fully reversible high cycle fatigue-loading on stiffness, damage, curvature, and snap-through load in the out-of-plane loading direction at eight different combinations of parameters with frequency from 1 to 10 Hz, two boundary conditions, and temperature from 22°C to 150°C up to 3 to 10 million cycles. Stiffness and damage evolution analysis demonstrate the first two stages in out-of-plane fatigue loading. The study proposes a damage definition in terms of load adapting with two fatigue damage models: (1) Shiri Model and (2) Wu Model, while both models exhibit reasonable accuracy in predicting damage for the first two stages despite deviating at the final cycle due to assuming this cycle as the final failure cycle. Of the two models, the Shiri model provided a smaller range of model parameter values, 0.22 and 0.43, for parameters p and q, respectively, which reflects adjustability to different test conditions by maintaining a moderate range. Specimens encountered no final failure by fiber breakage and did not lose bistability for any combination. Curvature and snap-through load measurements have not substantially changed due to fatigue loading. These findings confirm application protocols with a broad range of parameters for which the laminates can operate without significant fatigue damage and maintain their bistable performance for an infinite lifetime.

The experimental results for damage sensing using embedded magnetostrictive particles as sensors in carbon fiber–reinforced polymer laminates are presented. Laminates constructed of varying counts of pre-impregnated plies (Hexcel AS4/3501-6) are embedded with Terfenol-D particles and a delamination to observe changes in the sensing signal in and around the damage region. The testing apparatus consists of a primary circuit to generate a magnetic flux excitation and a secondary circuit to sense voltage induced into a pickup coil. In addition, three groups of test results were generated to observe the effect of fabrication on sensing. The experimental results from the three groups reveal that the particle distribution has a significant effect on the sensing signal. The sensing signal is dominated by changes in local particle distribution, and the effect of the delamination and ply variation did not have a discernible effect on the sensing signal for this setup.