Understanding thermal strains in composites is crucial in numerous industries, from aerospace to automotive and beyond. Composites, known for their exceptional strength-to-weight ratio, are engineered by combining materials like carbon fibers and polymers. However, their distinct constituents have varying coefficients of thermal expansion (CTE), causing them to expand or contract at different rates when exposed to temperature changes. This creates internal stresses, potentially leading to structural damage, compromising performance, and safety.
Modeling thermal strains allows engineers to predict how these materials will respond to temperature variations. By employing analytical and computational models, they can simulate the behavior of composites under thermal stress, predicting potential deformations or failures. This predictive ability is invaluable during the design phase, enabling optimization of material selection, layup configurations, and manufacturing processes to mitigate detrimental thermal effects.
Despite the importance of thermal strains on composite structures, modeling thermal strains is an inherently difficult task. That is because the thermal strains of the phases are not simply a volume average property and depend on the shape, the orientation, and the elastic and expansion properties of the matrix and the inhomogeneities (fibers, particles etc.) [1]. Due to these dependencies, the thermal strains are, in general, non-isotropic which results in an obvious complex modeling process.
The aim of this work is to present a novel methodology on modeling the thermal strains of composite parts. To this end, analytical and numerical methods are both implemented for fibrous and particulate composites. The analytical formulation is presented for the general case of anisotropic inhomogeneities in an anisotropic matrix and is based on effective field methods (EFMs) using property contribution tensors. The numerical models are developed using the finite element method and serve as a validation of the presented formulas. The results of the analytical and numerical models are compared in terms of components of thermal strains for the matrix and the inhomogeneities and exhibit great agreement. Therefore, the novel methodology allows for an accurate solution of the thermal strains and can serve as a tool for gaining insight into the behavior of composite parts and provide valuable information for their design. Finally, it should be underlined that the formulation based on EFM is characterized by minimal computational cost due to its analytical nature and can be preferred over the numerical models.