Material Model Research Articles

Equivariant graph convolutional neural networks for the representation of homogenized anisotropic microstructural mechanical response

Composite materials with different microstructural material symmetries are common in engineering applications where grain structure, alloying and particle/fiber packing are optimized via controlled manufacturing. In fact these microstructural tunings can be done throughout a part to achieve functional gradation and optimization at a structural level. To predict the performance of particular microstructural configuration and thereby overall performance, constitutive models of materials with microstructure are needed.In this work we provide neural network architectures that provide effective homogenization models of materials with anisotropic components. These models satisfy equivariance and material symmetry principles inherently through a combination of equivariant and tensor basis operations. We demonstrate them on datasets of stochastic volume elements with different textures and phases where the material undergoes elastic and plastic deformation, and show that the these network architectures provide significant performance improvements.

An overview of patents of noise reduction in buildings

Noise is an unwanted sound in urban areas. Natural and or recycled acoustic absorbing or insulation materials are invented to reduce air and or impact sounds in buildings. Yet, they are not widely and commercially available. This review presented an overview of relevant studies on natural materials particularly on cotton and rubber using Scopus search engine and of the patents registered and granted using Espacenet and Google Patents search engines on noise reduction in buildings from the years 1996 to 2023. The 18% and 68% (out of 32 articles) relevant articles, respectively, were written on cotton and rubber and other recycled natural material e.g., consumed teabags, rubber crumb, indicating a clear vision for experimentation and modeling of natural materials e.g., to investigate their sound absorption coefficients at lower frequencies (below 250 Hz) for (recycled) cotton sound absorbing materials and for producing ultralight density with high sound absorption materials (0.56) using recycled rubber material. There were found 31 patents registered and granted, in total. Natural materials, namely, cotton, rubber, glass wool and rock wool were applied in 29% of patents registered and granted from 1996 to 2022. Yet, 93% of patents registered and granted lacked data on sound absorption index of materials they invented. LDA models estimated 29% of Espacenet and Google Patents patents registered and granted having an increased trend of patent innovations for example, in method, structure and building materials in the last decade.

Flow-induced erosion modelling of cohesive material with coupled CFD-DEM approach

Erosion is one of the continuous wear mechanisms in cohesive materials caused by the shear stress applied by the fluid flow at the liquid–solid interface. The erosion resistance of cohesive materials is significantly influenced by the strength of the cohesive bonds that hold the particles together, increasing the complexity of the mechanism. The erosion rate of cohesive materials depends on the critical shear stress (CSS) and the erodibility coefficient at the surface. The current understanding of the relationship among the inter-particle bond strength, CSS, erodibility coefficient, and their respective contributions to the flow-induced erosion process is still lacking; therefore, it cannot be adequately described mathematically. The present research offers a coupled computational fluid dynamics (CFD)–discrete element method (DEM) approach to visualize and quantitatively analyze the flow-induced erosion in cohesive materials. The capability and accuracy of the developed model were validated using experimental data available in the literature. A cohesion model was then employed to describe the strength of the cohesive bond, and the effects of the cohesion energy density (CED) on the CSS and erodibility coefficient were investigated. The analysis indicated that for cohesive materials, the non-dimensional CSS not only is affected by the particle Reynolds number but also strongly depends on CED. The simulation results indicated that by increasing the CED value from 500 to 900kJ/m3, the CSS increased by 25 %, and the erodibility coefficient decreased by 62 %. The proposed CFD–DEM approach can effectively estimate the erosion initiation and erosion rate of cohesive materials in different applications and geometries.

Practice-oriented numerical model for seismic response of cold-formed steel-framed mid-rise buildings

Cold-formed steel (CFS) framed buildings are becoming increasingly common, notably in the high seismic regions of the United States (US). However, research to understand the seismic performance of CFS-framed buildings specifically using large-scale (building-level) specimens has begun only recently. Therefore, efforts to validate and improve numerical modeling of whole building systems have been limited due to this lack of experimental data. A recently completed experimental program, conducted at the Large High-Performance Outdoor Shake Table at the University of California, San Diego, involved earthquake testing of a full-scale 6-story CFS-framed building. The test building adopted an assembly of continuous tie-rod and built-up compression stud pack at the ends of individual shear walls to resist seismic overturning moment and uplift forces within the building. The objective of this paper is to present a practice-oriented phenomenological finite element modeling approach intended for design engineers and compare the predictions from this model against the measured seismic response of the 6-story building. The model aims to provide a robust, yet computationally efficient approach, with minimal degrees-of-freedom, while also capturing the salient features of seismic response of mid-rise CFS-framed buildings suitable for nonlinear time history analyses-based design. The building model, developed within the OpenSeesPy finite element platform, includes the shear walls and gravity walls as well as the lateral strength and stiffness contribution from exterior and interior gypsum sheathing. The continuous tie-rod system is also modeled to capture the cumulative floor displacements caused by the axial elongation in the steel rods. The effect of built-up compression stud packs on the lateral behavior of a shear wall was estimated using available experimental data and a detailed numerical model of an isolated shear wall, then subsequently incorporated in the nonlinear hysteretic material model. The numerical model captured the natural frequencies of the translation modes of the building in its undamaged state with <3% error. It also predicted the peak roof drift ratio within an error of 0.03% drift ratio under the design earthquake event. These results provide confidence in the proposed numerical modeling approach as a tool for seismic response prediction of mid-rise CFS-framed buildings with similar structural detailing features.

Validation of Experimental Data for the Application of the Magnesium Alloy “Elektron 43”

The behaviour of a structural component, such as the spreader installed on an aeroplane passenger seat made of the magnesium alloy Elektron® 43, is evaluated under a variety of load conditions. The purpose of this research project is to considerably reduce weight by employing the new alloy while keeping the strength and ductility necessary to meet the dynamic standards for both the 16 g forward and 14 g downward tests. A comprehensive campaign of static and dynamic testing on coupons was conducted to characterise the mechanical behaviour of the E43 magnesium alloy, from quasi-static to dynamic loading, and across a wide range of deformation rates. The elastic–plastic and strain rate sensitive material model of E43 is then calibrated using an FEA approach and LS-DYNA software, utilising stress–strain curves and properties determined from standardised experimental tensile and compression trials at varied strain rates. Finally, this material model was used to perform a finite element structural study of a major component of an aeroplane seat built using Elektron® 43 under typical in-flight stresses.

Viscoelastic-adhesive modeling of hybrid adhesively bonded double-lap joints: analytical and numerical studies

ABSTRACT There are a limited number of studies examining the viscoelastic and viscoplastic behavior of adhesive joints. If there is more than one adhesive in the overlap region, the problem becomes more complex and viscoelastic modeling of the overlap region becomes difficult. There is no study in the literature that covers such a bonding (overlap) scheme and examines the viscoelastic behavior of the double lap joint (DLJ). In this study, the effect of viscoelasticity of polymeric adhesives on shear stress and strain distributions in the hybrid-adhesive layer of a DLJ is studied. Maxwell-Wiechert (MW) viscoelastic material model is used to define the viscoelastic behavior of the adhesives. Analytical and finite element (FE) results are compared for both mono- and hybrid-adhesive joints under the step load. Good matches are observed between the analytical and numerical results. Four different hybrid-adhesive joints with different combinations of the adherend thicknesses are investigated by considering only the analytical method. In order to investigate the effect of the loading type on the viscoelastic behavior, step and ramp loads were applied to the joint separately. It is observed that applying the ramp load to the hybrid-adhesive joints affects the shear stresses occurred at the ends of the overlap more than the mono-adhesive joints.

On the application of RHT model and SPG algorithm for the analysis of rock cutting process

The linear cutting process in rock poses challenges for verification in field experiments, laboratory investigations, or numerical simulations. This study aims to analyze the rock cutting process and disc cutter force estimation when using linear cutting mode. Three-dimensional numerical simulations using the explicit dynamic finite element method (LS-DYNA software) are conducted to characterize the cutting process. In this regard, two computational algorithms (Lagrangian and Smoothed Particle Hydrodynamics (SPH)) and two material models (Johnson-Holmquist Concrete (JHC) and Riedel-Hiermaier-Thoma (RHT)) are compared, with SPH and RHT identified as more suitable for rock cutting simulation. The results of comparative analyses show that the Lagrangian computational algorithm is highly dependent on the erosion value, hence this method is not suitable for the simulation of the rock-cutting process. Comparing to the RHT material constitutive model, the Johnson-Holmquist model does not well model the post-failure softening strain behavior, which leads to a reduction in the width of the failure area. The comparative analyses also show that the normal and rolling forces predicted by the JHC model are well over 30% higher than the actual experimental results, while the RHT model shows a good agreement between the predictions and the actual results. Overall, the RHT material model with the use of the SPH computational algorithm shows a very good combination in rock cutting process simulation.

A proposal for a simple characterization of stainless steel connections through an equivalent yield strength

Notwithstanding some recent contributions, there still appears to be a gap in the literature regarding the express evaluation of the moment–rotation performance of some connections fabricated from austenitic, duplex, or ferritic stainless steel, such as flush end-plate connections (FEP). In the present work, moving from a previous proposal, the authors investigate a simple approach for the practical characterization of these connections. The results from industry standard software, which is dedicated to the analysis of connections based on the component method and widely used in engineering practice, are compared with the results from a more refined numerical model. It is found that, by using conventional carbon steel procedures with an ad hoc ideal material model, the results can be made in line with some experimentally validated numerical simulations for stainless steel. The findings are discussed in detail and may turn useful from both a design and an assessment standpoint.

Effects of spatial microstructure characteristics on mechanical properties of dual phase steel by inverse analysis and machine learning approach

This work aims to investigate complex relationship between microstructure characteristics and mechanical properties of dual phase (DP) steel through an inverse analysis based on Markov chain Monte Carlo (MCMC) method combined with meso-scale material modelling. In this framework, a machine learning approach as surrogate model was developed, in which support vector regression (SVR) and artificial neural network (ANN) were trained using results from representative volume element (RVE) simulations coupled with damage model, which were previously calibrated with experimental data of commercial DP steel grades. Moreover, specific microstructure descriptors including Moran’s index, martensite band index and martensite orientation were proposed for representing effects of spatial distributions of martensitic phase. As a result, inverse predictions of microstructure characteristics of DP steels for achieving defined yield strength, tensile strength, uniform elongation and toughness were presented. The inverse analysis could solve the non-uniqueness of structure–property relationships of steel, whereby significances of dispersed structures and aligned martensite bands were highlighted in details. The approach fairly dealt with multi-target optimization and high dimensional problem, which can be further applied as a guideline for designing DP microstructures with enhanced mechanical properties.

Evidence for separate backward recall and n-back working memory factors: a large-scale latent variable analysis

ABSTRACT Multiple studies have explored the factor structure of working memory (WM) tasks, yet few have done so controlling for both the domain and category of the memory items in a single study. In the current pre-registered study, we conducted a large-scale latent variable analysis using variant forms of n-back and backward recall tasks to test whether they measured a single underlying construct, or were distinguished by stimuli-, domain-, or paradigm-specific factors. Exploratory analyses investigated how the resulting WM factor(s) were linked to fluid intelligence. Participants (N = 703) completed a fluid reasoning test and multiple n-back and backward recall tasks containing memoranda that varied across (spatial or verbal material) and within (verbal digits or letters) domain, allowing the variance specific to task content and paradigm to be assessed. Two distinct but related backward recall and n-back constructs best captured the data, in comparison to other plausible model constructions (single WM factor, two-factor domain, and three-factor materials models). Common variance associated with WM was a stronger predictor of fluid reasoning than a residual n-back factor, but the backward recall factor predicted fluid reasoning as strongly as the common WM factor. These data emphasise the distinctiveness between backward recall and n-back tasks.

New OpenSees material model for simulating reinforced concrete shear walls subjected to coupled axial tension and cyclic lateral loads

In high-rise buildings, reinforced concrete (RC) shear walls, particularly those forming part of coupled wall systems or core wall systems, may be subjected to coupled axial tension and cyclic lateral loads during strong seismic events. A new two-dimensional material model named the “Fixed Angle Model Considering Crack Sliding” (FAM-CS) was proposed for simulating the nonlinear behavior of RC walls subjected to axial tension and cyclic lateral loads. The proposed model was implemented in the OpenSees platform. Nonlinear constitutive models for shear aggregate interlock effects of concrete and dowel action of rebars were incorporated into the proposed model to represent the shear transfer mechanisms along concrete cracks. The proposed model was validated against the test data of six wall specimens subjected to coupled axial tension and cyclic lateral loads. The results indicated that the proposed model reasonably predicted the lateral strength capacity (an average error of 6.2 %) and the residual strength at the maximum drift loading (an average error of 12.3 %) of the specimens. For comparison, three other commonly used models, the Shear-Flexure Interaction Multiple-Vertical-Line-Element-Model, the multi-layer shell element model and the Cyclic Softened Membrane Model were adopted for the simulation of the specimens. Compared with the three existing OpenSees models, the newly implemented material model provided improved predictions of the wall hysteretic behavior with respect to lateral strength capacity, lateral strength degradation, pinching characteristics and cyclic stiffness degradation. In addition, the proposed model reasonably captured the localized failure pattern as well as the energy dissipation characteristics of the test specimens. Finally, a mesh sensitivity study was conducted to assess the robustness of the proposed model. The results revealed that the proposed model was robust regarding the lateral load-drift responses of RC walls, with a mesh size ranging from approximately one to three times the average crack spacing.

Drilling performance analysis of a polycrystalline diamond compact bit via finite element and experimental investigations

The significance of improving the drilling productivity and reducing the cost and non-productive time of drilling process, substantially relies on the efficiency of drilling performance. This paper provides a comprehensive understanding of drilling process, aiming to predict drilling performance and investigate drilling parameters using a validated finite element (FE) model. Experimental validation of the FE model was achieved through testing on a laboratory drilling rig, ensuring the accuracy and reliability of the numerical simulations. To accurately capture the nonlinear characteristics of bit-rock interaction, the Riedel–Hiermaier–Thoma model was adopted as a material model, and its parameters were identified through a series of carefully conducted experimental tests. The numerical results obtained from the FE rock failure model during the compressive and tensile tests demonstrated a robust correlation with the experimental data. The verified material model was then employed into another FE drilling model to simulate rock breaking in an actual drilling scenario. This analysis sheds light on the impact of drill-bit interaction with the rock formation, providing valuable insights into its behaviour during drilling operations. The FE drilling model was further utilised in a parametric study to predict the effects of critical drilling parameters, like loading rate and rotary speed, on the weight on the bit, torque on the bit, and rate of penetration. Both the FE drilling and experimental results provided a significant consistency when the drilling parameters were compared, and nonlinear dynamic phenomena, such as stick–slip and bit-bouncing, were observed. By investigating these effects, this study contributes to optimising drilling operations, enabling better control of premature vibrations and enhancing drilling efficiency.

Data-driven methods for computational mechanics: A fair comparison between neural networks based and model-free approaches

We present a comparison between two approaches to modelling hyperelastic material behaviour using data. The first approach is a novel approach based on Data-driven Computational Mechanics (DDCM) that completely bypasses the definition of a material model by using only data from simulations or real-life experiments to perform computations. The second is a neural network (NN) based approach, where a neural network is used as a constitutive model. It is trained on data to learn the underlying material behaviour and is implemented in the same way as conventional models. The DDCM approach has been extended to include strategies for recovering isotropic behaviour and local smoothing of data. These have proven to be critical in certain cases and increase accuracy in most cases. The NN approach contains certain elements to enforce principles such as material symmetry, thermodynamic consistency, and convexity. In order to provide a fair comparison between the approaches, they use the same data and solve the same numerical problems with a selection of problems highlighting the advantages and disadvantages of each approach. Both the DDCM and the NNs have shown acceptable performance. The DDCM performed better when applied to cases similar to those from which the data is gathered from, albeit at the expense of generality, whereas NN models were more advantageous when applied to wider range of applications.

Data-driven computational mechanics: comparison of model-free and model-based methods in constitutive modeling

In computational homogenization approaches, data-driven methods entail advantages due to their ability to capture complex behavior without assuming a specific material model. Within this domain, constitutive model-based and model-free data-driven methods are distinguished. The former employ artificial neural networks as models to approximate a constitutive relation, whereas the latter directly incorporate stress–strain data in the analysis. Neural network-based constitutive descriptions are one of the most widely used data-driven approaches in computational mechanics. In contrast, distance-minimizing data-driven computational mechanics enables substituting the material modeling step entirely by iteratively obtaining a physically consistent solution close to the material behavior represented by the data. The maximum entropy data-driven solver is a generalization of this method, providing increased robustness concerning outliers in the underlying data set. Additionally, a tensor voting enhancement based on incorporating locally linear tangent spaces enables interpolating in regions of sparse sampling. In this contribution, a comparison of neural network-based constitutive models and data-driven computational mechanics is made. General differences between machine learning, distance minimizing, and entropy maximizing data-driven methods are explored. These include the pre-processing of data and the required computational effort for optimization as well as evaluation. Numerical examples with synthetically generated datasets obtained by numerical material tests are employed to demonstrate the capabilities of the investigated methods. An anisotropic nonlinear elastic constitutive law is chosen for the investigation. The resulting constitutive representations are then applied in structural simulations. Thereby, differences in the solution procedure as well as use-case accuracy of the methods are investigated.

On the grain level deformation of BCC metals with crystal plasticity modeling: Application to an RPV steel and the effect of irradiation

Capturing realistic deformation behavior in BCC metals at the polycrystal scale is an important aspect of predicting the material’s strength and failure. Furthermore, local deformation/strength heterogeneity also influences the lifetime and its assessments in reactor pressure vessel (RPV) steels, especially with accumulating irradiation doses during long-term operational conditions. This work utilizes a micromorphic crystal plasticity (CP) model for BCC materials with the capability to address temperature-dependent stress–strain response and irradiation effects relevant to RPV materials. The microstructure of the investigated material was characterized prior to and during testing using electron microscopy experiments, which are used for finite element model generation and simulation result validation. To analyze the validity of the model to predict strain localization under monotonic tensile loading, micro digital image correlation (DIC) was employed jointly with the CP simulations. Different modeling choices, such as the use Schmid/non-Schmid laws and gradient based CP modeling, greatly affected the capability of the model to represent similar magnitudes in strain localization and grain reorientation. The requirement for experimental measurements, including image based analyses, on a microstructural level as a validation tool for CP modeling is clearly demonstrated.

Thermal postbuckling and thermally induced postbuckled flutter of tri-directional functionally graded plates in yawed supersonic flow

The current work examines the thermal postbuckling, aero-elastic flutter and thermally induced postbuckled flutter about a static equilibrium state of tri-directional functionally graded (TFGM) rectangular and tapered plates in yawed supersonic flow. Based on the von Kármán nonlinear strains, the general higher-order shear deformation theory (GHSDT) and the first-order piston theory, the governing equations of motion for TFGM plates in yawed supersonic flow are established via the Hamilton's principle. The isogemetric analysis (IGA), the load continue strategy associated with the Newton Raphson iterative technique are exploited synthetically to capture the postbuckling paths, and then the postbuckled flutter behaviors are acquired with the static equilibrium state being obtained. Comparative and convergence studies are provided to improve reliabilities of the present material model, formulae and code implementations. Subsequently, detailed parametric investigations are carried out to evaluate the influences of material volume fractions, boundary conditions and airflow angles on the thermal postbuckling, aero-elastic flutter and postbuckled flutter behaviors of TFGM plates. The results show that it is the in-plane volume fractions, rather than the thickness volume fraction, that exert a greater influence on the thermal postbuckling and flutter behaviors of TFGM plates. Furthermore, TFGM plates exhibit more diverse postbuckling paths and more complex deformations due to the inhomogeneity of the in-plane material properties compared with z-directional FGM plates.

Programmable bionanocomposite coated fertilizers for prolonged controlled release of nitrogen

Controlled-release fertilizers (CRFs) present a promising solution for alleviating food and nutrient scarcity. However, their development has been mainly hindered by both rapid and unsynchronized nutrient release and unsustainable coating materials. In this study, we address this issue by developing a new coating layer for CRFs using a programmable biopolyurethane nanocomposite. This nanocomposite is prepared from diphenylmethane diisocyanate (MDI)-functionalized bentonite nanoclay (BNT-MDI) and a biopolyol from biomass waste. The results show that the BNT-MDI-doped CRFs (BCRFs) exhibit an impressive nitrogen (N)-release longevity of approximately 120 days at a 4 wt% coating ratio, surpassing previous CRF formulations. In a 30-day snap bean cultivation study, BCRF significantly improved root length (1000 %), leaf length (257 %), leaf width (400 %), and plant height (1400 %) compared to the control. The superior performance of BCRF is attributed to the PU-nanoclay biocomposite film, with full nano-exfoliation, controllable porosity, and high crosslinking density. Furthermore, we introduce a new dynamic release mechanism and establish a quantitative relationship between the nanostructure, property, and release performance of BCRF by combining the multiplicative and diffusion models for porous materials. This study provides a theoretical framework and a straightforward methodology for designing programmable nanocomposite structures for future biobased CRFs.

Multi-scale approach to hydrogen susceptibility based on pipe-forming deformation history

Hydrogen-induced-cracking initiates without external loading due to residual stresses. Pipe manufacturing process composed of crimping, U-ing, O-ing, and expansion has a major impact on local hydrogen concentration, as strain pattern evolves from one forming step to another, causing residual stresses that serve as driving force for hydrogen diffusion. The novelty of the presented work lies in the development of a multi-scale approach that links the residual stresses from the macroscopic pipe-forming process with locally dissolved hydrogen atoms in microstructure under the consideration of microstructural heterogeneities to identify areas susceptible to hydrogen-induced-cracking. First, a 3d-pipe-forming-model was built. Second, representative volume elements with lattice defects were generated to analyze hydrogen trapping in microstructure. Third, representative volume elements were placed in the pipe via sub-modeling, so that local loading history of the pipe was assigned to microstructure models. At the end of the pipe-forming process, representative volume elements were loaded with hydrogen on the surface and final hydrogen concentration was simulated based on residual stresses, considering microstructural effects such as grain size/shape, crystallographic texture and hydrogen traps, e.g. dislocations, voids and inclusions. On meso-/macroscale, a combined isotropic–kinematic hardening material model was implemented, while on microscale, a phenomenological crystal-plasticity-hydrogen-diffusion model was coded. According to the multi-scale simulations under the consideration of microstructural effects the bottom center position in the pipe was detected to be critical to hydrogen-induced-cracking as the maximum local hydrogen concentration was predicted at that location. Based on the loading history hydrogen-induced-cracking susceptibility increases from voids to hard and soft non-metallic inclusions.

Mechanical behaviour of bridge expansion joint interfaces filled with novel polyurethane-modified asphalt

The use of Asphalt Plug Joints (APJs) in bridge construction has rapidly expanded due to their distinct advantages in driving comfort, shock absorption, noise reduction, and maintenance ease. However, prolonged exposure to vehicular loads and environmental factors renders this expansion joint susceptible to interface effects and stress concentration, leading to damage, particularly at the interface layers between the expansion joint pavement and the expansion joint-cross joint plate contact areas. To address issues such as stress concentration, this study proposes a polyurethane-modified asphalt concrete mixture as the joint filling material. Pull-out and inclined shear tests are used to identify a bonding agent with superior adhesive properties. Additionally, a finite element model of pull-out specimens containing bilinear cohesive elements is established to simulate the cohesive damage process of the interface layers. Through parameter design analysis, the study examines the influence of individual parameter variations on interface damage behavior, obtaining comprehensive simulated analysis data on interface crack resistance performance. Based on interface test data, a bilinear cohesive force model and a viscoelastic material model of the asphalt concrete mixture are employed to establish a scaled-down model of the improved geometry design of the new APJ. This model simulates the interface mechanical characteristics of APJs under temperature effects and vehicle loads. The results demonstrate that the bilinear cohesive force model effectively simulates the nonlinear behavior of interface bonding slip. Furthermore, the use of SZ interface bonding agents and improvements in cross joint plate form can effectively mitigate issues such as uneven settlement and low-temperature cracking.

Multiscale homogenization of thermo-mechanical viscoelastic response of 3D orthogonal composites with time-dependent CTEs

Viscoelastic behavior of the matrix affects the thermomechanical response of polymer composites. This study investigated the structural effects of the time-dependent thermal expansion response of 3D orthogonal woven composite (3DOWC) by combining experiments and finite element modeling (FEM), incorporating the viscoelasticity of its constituents. The deformation characteristics at three orthogonal cross-sections, in the presence of structure-related strain fields at various temperatures, were investigated using digital image correlation method. Findings reveal heterogeneity and localization of the thermal strain, characterized by periodic stripes with alternating high strain regions on the resin and low strain regions on the yarn. This localization intensifies at higher temperatures, but diminishes with increased distance from the interface. Furthermore, the composites exhibit time-dependent thermal expansion behavior derived from the time-independent thermal expansion of its constituents, which cannot be captured in classic elastic material model. FEM reveals that thermal stresses are concentrated at the interfaces between constituents with significant differences in their coefficients of thermal expansion (CTE), particularly at the interface between the resins and axial yarns. The maximum thermal stress occurs in the binder yarns over the considered temperature range due to their low volume content, which should be considered in practical applications.