Representative Volume Element Research Articles

Investigation of Macroscopic Mechanical Behavior of Magnetorheological Elastomers under Shear Deformation Using Microscale Representative Volume Element Approach

Magnetorheological elastomers (MREs) are a class of smart materials with rubber-like qualities, demonstrating revertible magnetic field-dependent viscoelastic properties, which makes them an ideal candidate for development of the next generation of adaptive vibration absorbers. This research study aims at the development of a finite element model using microscale representative volume element (RVE) approach to predict the field-dependent shear behavior of MREs. MREs with different elastomeric matrices, including silicone rubber Ecoflex 30 and Ecoflex 50, and carbonyl iron particles (CIPs) have been considered as magnetic particles. The stress–strain characteristic of the pure silicon rubbers was evaluated experimentally to formulate the nonlinear Ogden strain energy function to describe hyper-elastic behavior of the rubbery matrix. The obtained mechanical and magnetic properties of the matrix and inclusions were integrated into COMSOL Multiphysics to develop the RVE for the MREs, in 2D and 3D configurations, with CIP volume fraction varying from 5% to 40%. Periodic boundary condition (PBC) was imposed on the RVE boundaries, while undergoing shear deformation subjected to magnetic flux densities of 0–0.4 T. Comparing the results from 2D and 3D modeling of isotropic MRE-RVE with the experimental results from the literature suggests that the 3D MRE-RVE can be effectively used to accurately predict the influence of varying factors including matrix type, volume fraction of magnetic particles, and applied magnetic field on the mechanical behavior of MREs.

Development and Calibration of a Phenomenological Material Model for Steel-Fiber-Reinforced High-Performance Concrete Based on Unit Cell Calculations

A phenomenological material model has been developed to facilitate the efficient numerical analysis of fiber-reinforced high-performance concrete (HPC). The formulation integrates an elasto-plastic phase-field model for simulating fractures within the HPC matrix, along with a superimposed one-dimensional elasto-plasticity model that represents the behavior of the embedded fibers. The Drucker–Prager plasticity and one-dimensional von-Mises plasticity formulations are incorporated to describe the nonlinear material behavior of both the HPC matrix and the fibers, respectively. Specific steps are undertaken during the development and calibration of the phenomenological material model. In the initial step, an experimental and numerical analysis of the pullout test of steel fibers embedded in an HPC matrix is conducted. This process is used to calibrate the micro-mechanical model based on the elasto-plastic phase-field formulation for fracture. In the subsequent step, virtual experiments based on an ellipsoidal unit cell, also with the resolution of fibers (used as a representative volume element, RVE), are simulated to analyze the impact of fiber–matrix interactions and their physical properties on the effective material behavior of fiber-reinforced HPC. In the final step, macroscopic boundary value problems (BVPs) based on a cuboid are simulated on a single scale using the developed phenomenological material model. The resulting macroscopic stress–strain characteristics obtained from both types of simulations, namely simulations of virtual experiments and macroscopic BVPs, are compared. This comparison is utilized for the calibration of material parameters to obtain a regularized solution and to assess the effectiveness of the presented phenomenological material model.

An insight into the influence of precipitation phase on the surface quality in diamond turning of an Aluminium alloy

Diamond turning is an effective technology for processing metal mirrors used in photoelectric communications, radar, and other fields. In diamond turning, the precipitated phase is an essential factor that influences the surface quality of the metal mirrors. However, in previous studies, the precipitation phase has typically been handled as a random variable in a surface morphology model to evaluate its influence on the surface roughness, instead of determining the formation mechanism and proposing suppression solutions. In this study, a new phenomenon is observed in the diamond turning of metal mirrors, that is, the micro diamond tool can reduce the protrusion of the precipitated phase under a small feed rate and improve the surface quality. Investigating the turning process using diamond tools with varying tool nose radii at small feed rates (<1 μm/r), the underlying transformation mechanism of the precipitation phase is determined with the advanced material characterization technologies. The growth of the precipitated phase with an increase in the tool nose radius is explained using the energy gradient theory. The results showed that the increased material strain on the machined surface decreased the activation energy of solute diffusion in the material, causing solute accumulation and precipitate phase growth. With a further increase of tool nose radius to around 1000 μm, the β'' phase breaks and rotates. The representative volume element method shows that when undergoing severe plastic deformation, dislocations and grain boundaries quickly aggregate and slide on the precipitated phase, which will lead to the fracture and rotation of β'' phase. These findings provide a theoretical basis for the development of highly smooth mirrors.

Carbon fiber pull-out analysis for composite materials by considering interface properties

ABSTRACT An improved shear-lag model is adopted in this paper to study the fiber pull-out by considering interface properties in a representative volume element (RVE). The surface roughness, represented by grooves distributed around the fiber, is introduced to analyze the influence of roughness on fiber pull-out comparing it with a round fiber. The influence of groove shapes is also discussed. The interfacial stiffness is introduced in the improved shear-lag model by the cohesive zone modeling to simulate a practical interface connection. To represent the residual stress in composites, a pressure load is added on the surface of the RVE to analyze its influence on the fiber pull-out. The results show that the proposed theory can accurately predict the stress distribution in the RVE. Furthermore, the maximum reaction force increases with an improvement in roughness and interfacial stiffness, while the pressure can improve the reaction force in the third stage of fiber pull-out process. The proposed theory can predict the stress field and fiber pull-out process accurately. It can improve the efficiency and effectivity of micro-experiment with the reference of theoretical results. The results obtained using the proposed theory in this research can also be useful for the design and analysis of composite materials.

FEM Modeling Strategies: Application to Mechanical and Dielectric Sensitivities of Love Wave Devices in Liquid Medium

This paper presents an extended work on the Finite Element Method (FEM) simulation of Love Wave (LW) sensors in a liquid medium. Two models are proposed to simulate the multiphysical response of the sensor. Both are extensively described in terms of principle, composition and behavior, making their applications easily reproducible by the sensor community. The first model is a Representative Volume Element (RVE) simulating the transducer and the second focuses on the sensor’s longitudinal (OXZ) cut which simulates the multiphysical responses of the device. Sensitivity of the LW device to variations in the rheological and dielectric properties of liquids is estimated and then compared to a large set of measurements issued from LW sensors presenting different technological characteristics. This integral approach allows for a deeper insight into the multiphysical behavior of the LW sensor. This article also explores the advantages and drawbacks of each model. Both are in good accordance with the measurements and could be used for various applications, for which a non-exhaustive list is proposed in the conclusion.

Structural analysis of small-scale 3D printed composite tidal turbine blades

The existing research literature lacks comprehensive investigations into assessing the structural performance of marine renewable energy conversion devices, particularly 3D printed turbine blades, which often rely solely on computational modelling without experimental validation methods and/or established mechanical characterization techniques. This leads to significant uncertainty regarding the performance of 3D printed turbine blades manufactured by additive manufacturing technology. This study aims to fill this gap by proposing a procedure for evaluating the structural integrity of commercial small-scale tidal turbine blade (5 KW) manufactured using fused filament fabrication 3D printing with a linear infill pattern. This is achieved by developing a combined experimental, hydrodynamic, and finite element approach with the view to inspect the micro-mechanical properties of representative volume elements of 3D printed microstructures using homogenization technique. The results of mechanical testing and hydrodynamic modelling are used to create a finite element model of the 3D printed blade, allowing for stress and failure analysis. Findings indicate that while integrating 3D printed materials into blade design via 3D printing technology is feasible, the choice of materials is limited to high stiffness composite filaments. Finally, experimental validation of numerical results, particularly full field strain distribution maps obtained by digital image correlation technique for flexural testing and laboratory-scale 3D printed blade, confirms the accuracy of the finite element results. Finite element-based homogenization techniques provide valuable insights into potential failure modes in 3D printed tidal turbine blades. However, the expedited calculation of orthotropic properties through finite element analysis proves to be a faster mechanical characterization method compared to experimental approaches. The proposed methodology in this study facilitates quicker iterative design of 3D printed blades, thereby reducing the need for repeated experiments and ultimately lowering manufacturing costs.

Revealing the percolation–agglomeration transition in polymer nanocomposites via MD-informed continuum RVEs with elastoplastic interphases

This contribution builds the concluding step of a multiscale approach to effectively capture the mechanical behavior of polymer nanocomposites (PNCs), in this case, silica-modified polystyrene. By introducing continuum-based representative volume elements (RVEs) that employ previously identified elastoplastic property gradients for the interphases surrounding the fillers, the effects of particle size, particle volume fraction, and agglomeration on the mechanical performance are investigated. Uniaxial tension tests are simulated with the respective finite-element RVEs, and stress–strain curves are derived. The elastic and plastic material properties of the RVE can then be extracted and analyzed quantitatively by fitting the stress–strain curves with a Voce-type elastoplasticity formulation.At small degrees of agglomeration, i.e., good particle dispersion, in combination with sufficiently large particle volume fraction, percolation bands form, leading to improved elastic and plastic properties. Higher degrees of agglomeration or particle clusters behave like large single particles, which has an adverse effect, i.e., the nanoscale size effect is thereby neutralized. Therefore, the precise MD-informed elastoplastic interphase representation of our RVEs enables the investigation of the transition from beneficial percolation to unfavorable agglomeration. Ultimately, this contribution establishes a link between the effects of particle size, particle volume fraction, agglomeration, and percolation, which have so far only been discussed separately in the literature.Our methodology offers new insights into the structure–property relations of PNCs and their resulting mechanical behavior. The underlying multiscale approach with a systematic transition from molecular to microscopic scales is required to complement experimental observations and exploit the full potential of PNCs.

Size effects of voids on the mechanical properties of 3D printed parts

AbstractAlthough the effects of additive manufacturing process parameters on the mechanical properties of 3D printed parts have been numerically investigated in the literature, less attention has been paid on the size effects of voids between deposited filaments. This study fills this gap by developing a methodology based on a combined finite element (FE) and design of experiment (DoE) technique. The development of FE methodology is based on micro-mechanical analysis of representative volume element (RVE) of 3D printed parts to predict the effective orthotropic properties. To account for the size effects of inter-bead voids, the RVE includes contribution of the multiple parameters of layer heights, layer widths and overlapping regions. To study the main and interaction effects of the above input parameters on the stiffness properties of 3D printed parts, a structured approach based on full factorial design is used. Although the size effects of voids on the constituents of elastic moduli of RVE were investigated, the main focus in the present work is to develop a regression model to predict the stiffness properties. The FE stress analysis of the RVE conducted in this study provides an insight about the potential failure modes such as delamination and filament debonding that may occur in load bearing 3D printed parts. For a case study, the results of FE-based homogenization technique in terms of stiffness properties are validated against the experimental data via three-point bending and Iosipescu shear tests which were conducted in conjunction with digital image correlation technique. The combined numerical and statistical approach proposed in this study provides a swift iterative design of 3D printed parts prior to the time-consuming computation modelling, contributing to reduce the number of tests and manufacturing costs.

Evaluation of strain-sensitive behavior in elastomeric conductive nanocomposites containing carbon nanotubes-a finite element-percolation model approach

A finite element (FE)-percolation model approach is developed to predict the strain-sensitive response of the three-dimensional (3D) representative volume element (RVE) of carbon nanotube (CNT)-elastomeric nanocomposite. In the simulation model, CNTs are modeled as solid, impenetrable cylinders inside a cubic insulating matrix. FE simulation is performed to evaluate the structural response of the RVE under applied strain. The FE model updates the locations of the CNTs in the deformed RVE. The paths are found using a suitable 3D resistance network associated with different percolation paths involved in the critical distance criterion. The percolation model utilizes the paths found to identify the electrical circuit for predicting tunneling conductivity. The simulating algorithm is used to study the influence of tunneling barrier height, nanotube volume fraction, and geometrical aspects. The lowest critical distance criterion is achieved for higher volume fractions and the most heightened sensitivity is obtained for lower CNT aspect ratios.

Physics-based modeling of the effective gas transport properties of single jersey knitted fabrics based on images

In many processes and applications, the performance of textiles relies heavily on fluid transport; for example, the in-plane distribution of water and the through-plane permeation of water vapor and air. Prediction of knitted fabrics’ effective transport characteristics can enhance development workflows and bring them to new applications. Effective transport parameters that are particularly important are the permeability and diffusive mass transfer. Usually, experiments are used to determine these parameters. It is desirable to conduct a thorough investigation into how yarn structure and knitting gauge influence these properties to tailor knitted fabrics for a particular application. Our contribution in this context describes a consistent workflow to forecast the effective mass transfer characteristics of single jersey fabrics. Single jersey fabrics have been chosen for they are the simplest patterning, and are widely used in body-near worn garments. The proposed approach involves visualizing fabric samples with a light microscope, and subsequently determining relevant geometric parameters through automated image processing algorithms. With these parameters in hand, a representative elementary volume of the fabric is constructed. The yarn is modeled as an effective medium to reduce calculation time. The representative elementary volume is then used for numerical predictions of air permeability and the diffusive water vapor transport. The predicted through-plane gas transport properties are compared with experimental data to validate the approach. Six different single jersey polyester fabrics were analyzed, with different yarn structures and machine gauges. The comparison shows a good agreement between simulated and measured transport properties.

Multiscale modeling for the impact behavior of 3D angle-interlock woven composites

In this paper, a multiscale modeling framework predicting the impact response of three-dimensional angle-interlock woven composites (3DAWCs) is presented. Multiscale modeling is performed sequentially at the mirco, meso, and macroscales. A microscopic representative volume element (RVE) model is established to characterize the yarn's nonlinear mechanical properties and damage behavior at different strain rates. At the mesoscale, a novel subcell-based discretization scheme is proposed to discretize 3DAWCs, and developing six representative subcells (RSCs) that bridge the microscale and macroscale effective properties. Based on it, a ballistic impact model involving six kinds of homogeneous material bricks is established. Furthermore, elastic-viscoplastic damage constitutive models are developed for the epoxy matrix and the yarn, respectively, to enhance the predicted accuracy of the multiscale model. Ballistic impact tests of 3DAWCs are conducted, and X-ray micro-computed tomography is employed to evaluate the internal damage of the 3DAWC target after impact. Validation work is conducted by comparing the experimental and numerical results of the ballistic impact of the 3DAWC target, which indicates the effectiveness and efficiency of the developed multiscale framework in the impact damage prediction of 3DAWCs.

An efficient parameterized neural network enhanced multiscale finite element modeling for triply periodic minimal surface meta-structures and its applications for femur

Triply periodic minimal surface (TPMS) multiscale structures are meta-structures known for their low-weight-high-strength property. Because of their delicate geometry and sheet-networks feature, a fine mesh is required for finite element analysis, which is computationally expensive for conventional direct numerical simulation. To resolve this issue, this study develops a neural network enhanced finite element method (NN-FEM) for parameterized TPMS multiscale meta-structures. The multiscale homogenization approach replaces the complicated representative volume elements (RVEs) with a surrogate homogenized material model. The surrogate process is accomplished by deep learning of artificial neural networks (NN) with an effective boundary displacement gradient and the Helmholtz free energy of RVEs. The NN-trained surrogate material model is deployed on the macroscale Galerkin finite element model for various field tests. The parameterized structural information of TPMS can also be embedded into the NN model to resolve the hidden constitutive relation of RVEs with unique geometries. Numerical examples are provided to benchmark the efficacy and efficiency of NN-FEM on the parameterized TPMS structure. The proposed model can be applied to femurs for rapid stress analysis and osteoporosis detection.

A novel multiscale prediction strategy for simulating the progressive damage behavior of plain-woven bamboo fabrics reinforced epoxy resin composites

This study proposed a novel multiscale prediction strategy, including mesoscale and macroscale damage evolution modeling, to investigate the effective properties and progressive damage failure behavior of plain-woven reinforced composites (PWRCs) under various external loads (tension, compression). A high-precision mesoscale representative volume element (RVE) that could accurately describe the local mechanical behavior of PWRCs was developed by combining experimental characterization (scanning electron microscope and X-ray computed tomography) and numerical simulation. To solve the problem of inaccurate prediction results caused by multiscale characteristics and complex stress changes in the damaged area of PWRCs, the strain-based 3D Hashin failure criterion and the multiscale damage models were used to predict the damage initiation and evolution of mesoscale reinforcement (bamboo fiber yarn bundle) and macroscale composites, respectively. Considering the damage evolution law of isotropic materials, a damage model based on the Von Mises criterion was used to characterize the damage initiation and evolution of mesoscale matrix epoxy resin (EP) under external loading. The effective properties and mechanical behavior of the PWRCs were transferred from mesoscale to macroscale through the progressive homogenization method. The bilinear constitutive relationship of the mixed-mode cohesive element was used to characterize the interlaminar failure of the PWRCs. Finally, the corresponding mechanical characterization (tension, compression) of the PWRCs was carried out. Moreover, the experimental results were highly consistent with the simulation results, verifying the reliability of the novel multiscale prediction strategy in investigating the mechanical response of the PWRCs at multiple scales and revealing the damage mechanism of the PWRCs.

Microstructure generation and full-field multi-scale analyses for short fiber reinforced thermoplastics: Application to PA66GF composites

Short Fiber Reinfored Thermoplastics (SFRTs) like PA66GF composites are highly heterogeneous materials with a rather complex microstructure such that the characterization and the prediction of their mechanical behavior remain quite challenging. So far, most of the research efforts have employed phenomenologigal and mean-field multi-scale models to deal with SFRTs. The present contribution rather focuses on a full-field multi-scale approach that incorporates advanced techniques in terms of microstructural representation and material modeling, allowing a deep insight of the dominating deformation mechanisms occurring in PA66GF composites. The proposed approach is based on an automatic periodic mesh generation algorithm for matrix-inclusion Representative Volume Elements (RVE) with randomly positioned fibers that follow a given Orientation Distribution Function (ODF). At the microscopic scale, while the fibers are assumed to be elastic, the behavior of the thermoplastic matrix is described by a phenomenological multi-mechanism constitutive model accounting for viscoelasticity, viscoplasticity and ductile damage. It results an advanced multi-scale model that enables to visualize the local deformation and degradation mechanisms occurring at the microscopic scale while simultaneously analyzing their influence on the macroscopic response of the composite upon monotonic, persistent and cyclic loading. The potential of the proposed approach is further evaluated by comparing the predicted responses against experimental data.

A multiscale hydro-elastoplastic model for large floating structures in two-dimension

A novel fluid–structural model was presented considering hydro-elastoplastic behavior, which coupled multiple hydro-structural-material scales. A large floating sandwich structure (LFSS) comprising upper and lower high-strength panels and a low-weight core was considered as an illustration. The mesoscale characteristics of materials and elastoplastic parameters of the low-weight perforated components were coupled by utilizing the representative volume element (RVE) method. Through the parameterized relations from RVE analysis, the flexure dynamics model for the floating sandwich structure with an equivalent homogenized core was deduced. With the flexure dynamic equation, yield criterion, and potential flow model, a multiscale hydro-elastoplasticity theoretical model was established, which combined the wave action, hierarchical component, material configuration, and structural behavior. The dynamic responses of the large floating structure under fluid–structure interaction were calculated, and the internal deformation (i.e., core strain) was set as the determining variable for the plastic region. The initially intact floating structure became hinged multi-modules after plastic cracking, and the high-strength layers at the cracking positions behaved as flexible hinges, which was defined as a hydro-elastoplastic process. The elastoplastic state evolutions of the LFSSs with different structural parameters and material configurations were solved for practical optimization. The results indicated that the multiscale coupled calculation model can provide great scientific guidance for designing large floating structures.

Micro-CT image-based computation of effective thermal and mechanical properties of fibrous porous materials

Fibrous porous materials are extensively employed as heat shielding material for space vehicles and capsules. To predict the performance of these materials using computational modeling at the vehicle/capsule scale, the effective material properties are needed. In this work, the effective thermal conductivity and elastic constants of a carbon fibrous porous material called FiberForm (precursor of PICA) were calculated using a direct image-based approach. In this image-based approach, the microstructures obtained using X-ray computed tomography technique are represented as binary voxel data. Nonlocal interactions are introduced between neighboring voxels. Integro-differential equations, with respect to spatial and temporal dimensions respectively, are developed to govern material thermal and mechanical behaviors. Using this approach, energy-based computational procedures were developed to calculate the effective material properties of fibrous and porous materials irrespective of the periodicity of underlying microstructures. The size of representative volume element was determined by convergence of effective properties with respect to microstructure size.

Investigation of the effects of volume fraction, aspect ratio and type of fibres on the mechanical properties of short fibre reinforced 3D printed composite materials

AbstractMechanical behaviour of 3D-printed composite parts is affected by the volume fraction, aspect ratio and type of fibre reinforcement. Although in the literature experimental approaches have been used to characterise the effects of the above factors on the mechanical properties of 3D printed parts, time and cost of the manufacturing process as well as the uncertainty associated with a large number of experimental techniques are the key issues. This study aims to address these challenges by developing a methodology based on a multi-scale Finite Element (FE) analysis of representative volume element (RVE) of 3D printed composite parts to predict the effective orthotropic properties. To account for the effects of fibre features, RVEs were modelled considering variables of volume fraction, aspect ratios and type of short fibres. To study the main and interaction effects of the above variables on the mechanical properties of 3D printed composite parts, a structured approach based on the Design of Experiments is used. The FE stress analysis of the RVE provides an understanding about the potential failure modes such as interfacial debonding between fibres and matrix, interlayer and intralayer delamination that may occur in load-bearing 3D printed composite parts. The FE computed mechanical properties are validated against experimental data through a series of mechanical testing of flexure, Iosipescu, and short beam shear which were conducted in conjunction with the Digital Image Correlation technique. As a result, certainty is obtained in using the proposed approach for a fast iterative design of 3D printed composite parts prior to industrial applications.

Integrating Experimental and Computational Analyses for Mechanical Characterization of Titanium Carbide/Aluminum Metal Matrix Composites.

This study investigates the mechanical properties of titanium carbide/aluminum metal matrix composites (AMMCs) using both experimental and computational methods. Through accumulative roll bonding (ARB) and cryorolling (CR) processes, AA1050 alloy surfaces were reinforced with TiCp particles to create the Al-TiCp composite. The experimental analysis shows significant improvements in tensile strength, yield strength, elastic modulus, and hardness. The finite element analysis (FEA) simulations, particularly the microstructural modeling of RVE-1 (the experimental case model), align closely with the experimental results observed through scanning electron microscopy (SEM). This validation underscores the accuracy of the computational models in predicting the mechanical behavior under identical experimental conditions. The simulated elastic modulus deviates by 5.49% from the experimental value, while the tensile strength shows a 6.81% difference. Additionally, the simulated yield strength indicates a 2.85% deviation. The simulation data provide insights into the microstructural behavior, stress distribution, and particle-matrix interactions, facilitating the design optimization for enhanced performance. The study also explores the influence of particle shapes and sizes through Representative Volume Element (RVE) models, highlighting nuanced effects on stress-strain behavior. The microstructural evolution is examined via transmission electron microscopy (TEM), revealing insights regarding grain refinement. These findings demonstrate the potential of Al-TiCp composites for lightweight applications.

Numerical generation and verification of a two-dimensional virtual asphalt mixture

This paper aims at finding a robust method to generate a two-dimensional virtual asphalt mixture and to verify the suitability of the generated virtual mixture. First, three methods namely stereological method, Walraven formula and phase separation method are adopted to convert the actual three-dimensional gradation curves into two-dimensional ones. Then, an approach called semi whole scale cross section (SWSCS) method is developed to verify the results obtained and to determine the most robust procedure to follow. Third, the numerical two-dimensional geometry shape data library (GSDL) of different aggregates and voids and a robust method to assemble representative volume elements (RVEs) are developed using Matlab. Finally, a gradation curve from the literature is considered to generate a virtual asphalt mixture specimen, which is then imported in the finite element (FE) software Abaqus to simulate the dynamic modulus test. The virtual test results are further verified using available literature and another homogenization numerical prediction method based on the Mori-Tanaka's approach. The results show that the phase separation method is effective to convert the three-dimensional gradation curve to a two-dimensional one. The simulation results of the virtual asphalt mixture developed in this study turn out to be in good agreement with both previously presented literature data and Mori-Tanaka's theory.

Multiscale modeling of coupled thermo-hydro-mechanical behavior in ice-bonded granular media subject to freeze-thaw cycles

We present a novel multiscale framework that integrates the single-point multiphase material point method (MPM) and the discrete element method (DEM) to model the complex freeze-thaw behavior of ice-bonded granular media. The proposed numerical framework is featured by (a) employing the continuum-based MPM to solve the macroscopic governing equations for granular systems involving thermo-hydro-mechanical (THM) coupling and phase transitions, and (b) using the grain-scale discontinuum-based DEM to capture the thermodynamically sensitive mechanical behaviors of ice-bonded granular media. The multiscale framework is constructed by attaching a DEM-based representative volume element (RVE) at each material point in MPM. This RVE serves as a live sample of each material point to track the state-dependent effective stress with respect to the local deformation and thermodynamic conditions like ice saturation, bridging the macroscopic phenomena and the underlying microstructural evolution. In particular, we implement a semi-implicit staggered integration scheme for the macroscale THM-coupled MPM to boost computational efficiency and enhance numerical stability. We also propose an innovative ice saturation-dependent bond contact to effectively reproduce the thermodynamically sensitive mechanical behaviors. The new multiscale framework is first benchmarked against analytical solutions for 1D non-isothermal consolidation problems. We then demonstrate its exceptional capability in simulating intricate freeze-thaw behavior of granular media through a boundary value problem involving cyclic freeze-thaw actions. Further cross-scale analyses reveal its potential in capturing key loading- and state-dependent THM responses with explainable microstructural mechanisms during complex freezing and thawing loading conditions.