Abstract:In this work, we propose a deep learning (DL)-based constitutive model for investigating the cyclic viscoelastic-viscoplastic-damage behavior of nanoparticle/epoxy nanocomposites with moisture content. For this, a long short-term memory network is trained using a combined framework of a sampling technique and a perturbation method. The training framework, along with the training data generated by an experimentally validated viscoelastic-viscoplastic model, enables the DL model to accurately capture the rate-dependent stress-strain relationship and consistent tangent moduli. In addition, the DL-based constitutive model is implemented into finite element analysis. Finite element simulations are performed to study the effect of load rate and moisture content on the force-displacement response of nanoparticle/ epoxy samples. Numerical examples show that the computational efficiency of the DL model depends on the loading condition and is significantly higher than the conventional constitutive model. Furthermore, comparing numerical results and experimental data demonstrates good agreement with different nanoparticle and moisture contents.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a deep learning (DL) model that can accurately predict the viscoelastic - viscoplastic - damage behavior exhibited by nanoparticle - epoxy nanocomposites containing moisture under cyclic loading conditions. Specifically, the authors aim to address the following issues: 1. **Limitations of existing models**: Traditional models are usually calibrated only for uniaxial loading conditions, which are insufficient to fully capture the viscoelastic behavior of polymer - based materials. Moreover, most existing constitutive models are mainly for pure polymer materials or composites in dry conditions, ignoring the influence of environmental conditions (such as humidity) and particle content. 2. **Time - dependent and irreversible responses**: Many existing models cannot fully capture the time - dependent irreversible responses of materials, or their prediction accuracy is insufficient. Especially for material behaviors with complex history - dependence, traditional models often perform poorly. 3. **Computational efficiency**: Traditional numerical simulations of nonlinear material behavior involve time - consuming iterative solution processes, resulting in low computational efficiency. Therefore, a more efficient computational method is required to handle complex material behaviors. To solve the above problems, the authors propose a constitutive model based on deep learning. This model is trained by long - short - term memory network (LSTM) to improve computational efficiency and accurately capture the viscoelastic - viscoplastic - damage behavior of materials. In addition, this model also considers the influence of humidity and nanoparticle volume fraction on the stress - strain relationship of materials. ### Formula summary The main formulas involved in the paper include: - **Overall deformation gradient decomposition**: \[ F = J^{1/3}F_{\text{iso}} \] where \(J=\det[F]\) is the volumetric deformation and \(F_{\text{iso}}\) is the isochoric deformation gradient. - **Further decomposition of volumetric deformation**: \[ J = J_mJ_w \] where \(J_w = 1+\alpha_w w_w\), \(\alpha_w\) is the moisture expansion coefficient and \(w_w\) is the moisture content. - **Cauchy stress tensor decomposition**: \[ \sigma = J^{-1}(\mu_{\text{eq}}B_{ve}^{\text{iso}}+\mu_{\text{neq}}B_e^{\text{iso}}) \] \[ \sigma_{\text{vol}}=\frac{1}{2}k_v\left(\frac{J_m - 1}{J_m}\right)I \] The overall stress is: \[ \sigma_{\text{tot}}=\sigma+\sigma_{\text{vol}} \] - **Total stress in the damaged state**: \[ \sigma_d^{\text{tot}}=(1 - d)(\sigma+\sigma_{\text{vol}}) \] where \(d\in[0, 1)\) is a scalar damage variable, and its evolution rule is: \[ \dot{d}=A(1 - d)\dot{\Lambda}_{\text{max}}^{\text{chain}} \] - **Relationship between shear modulus and nanoparticle volume fraction and moisture content**: \[ \mu_{\text{eq}}(v_{\text{np}}, w_w)=X(v_{\text{np}}, w_w)\mu_0^{\text{eq}} \] \[ \mu_{\text{neq}}(v_{\text{np}}, w_w)=X(v_{\text{np}}, w_w)\mu_0^{\text{neq}} \] The amplification factor \(X\) is defined as: \[ X=(1 + 5v_{\text{np}}+ 18v

A machine learning-based viscoelastic-viscoplastic model for epoxy nanocomposites with moisture content

Multi-Scale Modelling of Long-Term Mechanical Behaviour in Polymer Composite Laminates with Woven Fibre Architecture

Machine Learning and Materials Informatics Approaches for Predicting Transverse Mechanical Properties of Unidirectional CFRP Composites with Microvoids

Mechanical Behaviors of Copolymer Elastomers: Experimental Characterization and Constitutive Modelling

Prediction Model of Residual Load-Bearing Capacity of Composite Laminates Using Deep Learning

A Lennard-Jones Potential Based Cohesive Zone Model and Its Application in Multiscale Damage Simulation of Graphene Reinforced Nanocomposites

A thermodynamically consistent physics-informed deep learning material model for short fiber/polymer nanocomposites

Viscoelastic Bilinear Cohesive Model and Parameter Identification for Failure Analysis of Adhesive Composite Joints

Cyclic viscoelastic-viscoplastic behavior of epoxy nanocomposites under hygrothermal conditions: A phase-field fracture model

A viscoelastic-viscoplastic constitutive model for nanoparticle-reinforced epoxy composites: Particle, temperature, and strain rate effects

A Nonlinear Viscoelastic-Viscoplastic Constitutive Model for Epoxy Polymers

Deep Learning Approaches for Dynamic Mechanical Analysis of Viscoelastic Fiber Composites

Prediction of shape distortions in thermosetting composite parts using neural network interfaced visco-elastic constitutive model

A mechanics-informed machine learning approach for modeling the elastoplastic behavior of fiber-reinforced composites

Application of a large deformation nonlinear-viscoelastic viscoplastic viscodamage constitutive model to polymers and their composites

A micromechanics-based recurrent neural networks model for path-dependent cyclic deformation of short fiber composites

Concurrent predictions of tension, compression, and shear characteristics of epoxy using three-network viscoplastic model

Analyzing and Predicting the Viscosity of Polymer Nanocomposites in the Conditions of Temperature, Shear Rate, and Nanoparticle Loading with Molecular Dynamics Simulations and Machine Learning

Understanding moisture effect on nonlinear vibrations of epoxy thin film via a multiscale simulation

Constitutive modeling for linear viscoelastic fiber-reinforced composites

Comparative analysis of machine learning models for predicting dielectric properties in MoS 2 nanofiller-reinforced epoxy composites