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.

The ability of a material to display two equilibrium states, called bistability, has been previously observed in carbon fiber reinforced polymers (CFRPs). For bistability to occur, the laminate must consist of an unsymmetric layup about its midplane which generates internal residual stress from thermal contraction. Prior studies have observed bistability in CFRPs with small-scale rectangular geometries where all sides were less than 250 mm. The aim of this paper is to demonstrate the existence of bistability in large-scale CFRPs with rectangular and non-rectangular geometries. Experiments and finite element analyses were conducted to determine the viability of bistability in large-scale CFRPs where at least one length aspect of the specimen was greater than or equal to 304.8 mm. Specimens whose shapes included rectangles, deltoids, triangles, and circles, were fabricated and tested to determine the presence of bistability and the associated curvature for each cured equilibrium state. Rectangular specimens had a side length of 914.4 mm and widths that varied from 177.8 to 457.2 mm. For the deltoids, triangles, and circles, one length aspect (i.e. the height, hypotenuse, and diameter, respectively) equaled 304.8 mm. Finite element models were created to compare the equilibrium shapes’ curvatures and displacements with the experimental laminates; the existence of bistability was also examined using a nondimensionalized bifurcation plot. Experimentally, bistability was found to occur for the fabricated laminates up to six plies. As the studied laminates could be considered thin, they displayed cylindrical cured shapes. The non-traditional shaped CFRPs followed bistability trends found for traditional, small-scale, rectangular laminates. An inverse relationship between the ply count and curvature was exhibited for the large-scale, rectangular laminates; curvature decreased as the number of plies in the laminate increased.

Composite laminates constructed in an asymmetric layup orientation of [0i, 90i], i > 0, exhibit two stable equilibrium positions and may be actuated to snap from a primary cure shape to an inversely related secondary stable shape. This study aims to aid in developing a comprehensive description of thick bistable laminates, whose increased thickness risks the loss of bistability, through previously established analytical approaches and verification via experimentation. The principle of minimum potential energy is applied to two materials and analyzed using the Rayleigh-Ritz minimization technique to determine the cure shapes of carbon fiber reinforced polymer laminates composed of AS4/8552 and TR50S-12k carbon fibers. These materials were modeled to act as square thick bistable laminated composites with sidelengths up to 0.914m. Visualizations of the out-of-plane displacements are shown with a description of the Rayleigh-Ritz analysis. Additionally, a finite element model (FEM) created in Abaqus CAE 6.14 and experiments using DA409/G35 and TR50S-12K/NP301 prepreg were used to further describe and develop the fundamental description for thick bistable laminates in terms of loss of bistability, actuation load, and principle shape. The analytical model is an extension of Hyer’s (2002) and Mattioni’s (2009) work applied to thick bistable laminates where the primary assumption was the x-axis curvature equaled the negative y-axis curvature for the primary and secondary stable positions, respectively. This assumption leads to the already cemented conclusion that bistable laminates, once cured, take on one of two inversely related paraboloid shapes. FEA simulations contradicted this by showing an average 11% difference in curvature magnitude for the aforementioned shapes. Furthermore, fourth order polynomials were used to describe the curvature along the axes, differing from the previously used Menger curvatures, (three-point approximation). Bifurcation plots using peak deflections and average curvature generated from FEA simulations clearly showed bistability existed to approximately 50 plies; however, the energy landscape plots indicated a significant degradation of bistability starting at 36 plies. Experimentation was performed on a test stand mimicking the same boundary conditions used in FEA while applying a central out-of-plane load. Experimental observations showed decreased peak displacements of stable cure shapes. Observations also indicated that the x-axis curvature had a significant difference in magnitude compared to the negative y-axis curvature. However, the existence of bistability agreed with FEA energy landscape plots, with clear “snaps” ending at thicknesses of 28–36 plies. Moreover, actuation force was found to correlate well with FEA simulations. Differences in the critical point can be attributed to the combination of material property differences for DA409 and TR50S-12K, failure to capture polymer relaxation, limitations of the experimental setup, and hand layup fabrication errors. Lastly, this paper adds viability of thicker laminates for use in macroscale applications where shape morphing or shape-retention attributes are a necessary constraint, although only where low loads are expected.