arXiv subjects

Bai-Xiang Xu

Publications and source records attributed to Bai-Xiang Xu.

At least 37 records · Page 2Linked to original sources

arXiv2024

Unraveling impacts of polycrystalline microstructures on ionic conductivity of ceramic electrolytes by computational homogenization and machine learning

The ionic conductivity at the grain boundaries (GBs) in oxide ceramics is typically several orders of magnitude lower than that within the grain interior. This detrimental GB effect is the main bottleneck for designing high-performance ceramic electrolytes intended for use in solid-state Lithium-ion batteries, fuel cells, and electrolyzer cells. The macroscopic ionic conductivity in oxide ceramics is essentially governed by the underlying polycrystalline microstructures where GBs and grain morphology go hand in hand. This provides the possibility to enhance the ion conductivity by microstructure engineering. To this end, a thorough understanding of microstructure-property correlation is highly desirable. In this work, we investigate numerous polycrystalline microstructure samples with varying grain and grain boundary features. Their macroscopic ionic conductivities are numerically evaluated by the finite element homogenization method, whereby the GB resistance is explicitly regarded. The influence of different microstructural features on the effective ionic conductivity is systematically studied. The microstructure-property relationships are revealed. Additionally, a graph neural network-based machine learning model is constructed and trained. It can accurately predict the effective ionic conductivity for a given polycrystalline microstructure. This work provides crucial quantitative guidelines for optimizing the ionic conducting performance of oxide ceramics by tailoring microstructures.

cond-mat.mtrl-sci↗

arXiv2024

Elasto-plastic residual stress analysis of selective laser sintered porous materials based on 3D-multilayer thermo-structural phase-field simulations

Residual stress and plastic strain in additive manufactured materials can exhibit significant microscopic variation at the powder scale, profoundly influencing the overall properties of printed components. This variation depends on processing parameters and stems from multiple factors, including differences in powder bed morphology, non-uniform thermo-structural profiles, and inter-layer fusion. In this research, we propose a powder-resolved multilayer multiphysics simulation scheme tailored for porous materials through the process of selective laser sintering. This approach seamlessly integrates finite element method (FEM)-based non-isothermal phase-field simulation with thermo-elasto-plastic simulation, incorporating temperature- and phase-dependent material properties. The outcome of this investigation includes a detailed depiction of the mesoscopic evolution of stress and plastic strain within a transient thermo-microstructure, evaluated across a spectrum of beam power and scan speed parameters. Simulation results further reveal the underlying mechanisms. For instance, stress concentration primarily occurs at the necking region of partially melted particles and the junctions between different layers, resulting in the accumulation of plastic strain and residual stress, ultimately leading to structural distortion in the materials. Based on the simulation data, phenomenological relation regarding porosity/densification control by the beam energy input was examined along with the comparison to experimental results. Regression models were also proposed to describe the dependency of the residual stress and the plastic strain on the beam energy input.

cond-mat.mtrl-sci↗

arXiv2023

A robust finite strain isogeometric solid-beam element

In this work, an efficient and robust isogeometric three-dimensional solid-beam finite element is developed for large deformations and finite rotations with merely displacements as degrees of freedom. The finite strain theory and hyperelastic constitutive models are considered and B-Spline and NURBS are employed for the finite element discretization. Similar to finite elements based on Lagrange polynomials, also NURBS-based formulations are affected by the non-physical phenomena of locking, which constrains the field variables and negatively impacts the solution accuracy and deteriorates convergence behavior. To avoid this problem within the context of a Solid-Beam formulation, the Assumed Natural Strain (ANS) method is applied to alleviate membrane and transversal shear locking and the Enhanced Assumed Strain (EAS) method against Poisson thickness locking. Furthermore, the Mixed Integration Point (MIP) method is employed to make the formulation more efficient and robust. The proposed novel isogeometric solid-beam element is tested on several single-patch and multi-patch benchmark problems, and it is validated against classical solid finite elements and isoparametric solid-beam elements. The results show that the proposed formulation can alleviate the locking effects and significantly improve the performance of the isogeometric solid-beam element. With the developed element, efficient and accurate predictions of mechanical properties of lattice-based structured materials can be achieved. The proposed solid-beam element inherits both the merits of solid elements e.g. flexible boundary conditions and of the beam elements i.e. higher computational efficiency.

math.NA↗

arXiv2023

On existence, uniqueness and stability of solutions to Cahn-Hilliard/Allen-Cahn systems with cross-kinetic coupling

A system of phase-field equations with strong-coupling through state and gradient dependent non-diagonal mobility matrices is studied. Existence of weak solutions is established by the Galerkin approximation and a-priori estimates in strong norms. Relative energy estimates are used to derive a general nonlinear stability estimate. As a consequence, a weak-strong uniqueness principle is obtained and stability with respect to model parameters is investigated.

math.AP↗

arXiv2023

Variational quantitative phase-field modeling of nonisothermal sintering process

In this work, we present a variational and quantitative phase-field model for non-isothermal sintering processes. The model is derived via an extended non-diagonal phase-field model. The model evolution equations have naturally cross-coupling terms between the conserved kinetics (i.e., mass and thermal transfer) and the non-conserved one (grain growth). These terms are shown via asymptotic analysis to be instrumental in ensuring the elimination of interface artefacts, while also examined to not modify the thermodynamic equilibrium condition (characterized by dihedral angle). Moreover, we demonstrate that the trapping effects and existence of surface diffusion in conservation laws are direction-dependent. An anisotropic interpolation scheme of the kinetic mobilities which differentiates the normal and the tangential directions along the interface is discussed. Numerically, we demonstrate the importance of the cross-couplings and the anisotropic interpolation via presenting thermal-microstructural evolutions.

cond-mat.mtrl-sci↗

arXiv2023

Influence of amorphous phase on coercivity in SmCo5-Cu nanocomposites

Severe plastic deformation of powder blends consisting of SmCo5-Cu results in magnetically hardened nanocomposite bulk materials. The microstructure is continuously refined with increasing torsional deformation, yet, coercivity saturates at a certain level of strain. Transmission electron microscopy (TEM) investigation of the microstructure reveals a partial amorphization of the SmCo5 phase due to high-pressure torsion by 20 applied rotations. In this amorphous matrix nanocrystals are embedded. The effect of these experimentally observed microstructural features on the magnetic properties are investigated by micromagnetic simulations, which show that an increasing volume fraction of nanocrystals is beneficial for higher coercivities. For a fixed volume fraction of nanocrystals the simulations reveal an increasing coercivity with decreasing the size of the nanocrystals due to increasing number of interfaces acting as pinning sites. Furthermore, our micromagnetic simulations disclose the mechanisms of the saturation and decline of magnetic hardening due to the strain induced by high-pressure torsion. The calculated coercivity fits very well to the experimentally observed coercivity of Hc=1.34 T. The knowledge can also be used to develop and provide optimization strategies from the microstructure perspective.

cond-mat.mtrl-sci↗

arXiv2023

Micromagnetics and Multiscale Hysteresis Simulations of Permanent Magnets

Confronting the unveiled sophisticated multiscale structural and physical characteristics of hysteresis simulation of permanent magnets, notably samarium-cobalt (Sm-Co) alloy, a novel scheme is proposed linking physics-based micromagnetics on the nanostructure level and magnetostatic homogenization on the mesoscale polycrystal level. Thereby the micromagnetics-informed surrogate hysteron is the key to bridge the scales of nanostructure and polycrystal structure. This hysteron can readily emulate the local magnetization reversal with the nanoscale mechanisms considered, such as nucleation of domains, and domain wall migration and pinning. The overall hysteresis, based on a sintered Sm-Co polycrystal, considering both mesoscale and nanoscale characteristics, is simulated and discussed.

cond-mat.mtrl-sci↗

arXiv2023

Thermodynamically consistent non-isothermal phase-field modelling of elastocaloric effect: indirect vs direct method

Modelling elastocaloric effect (eCE) is crucial for the design of environmentally friendly and energy-efficient eCE based solid-state cooling devices. Here, a thermodynamically consistent non-isothermal phase-field model (PFM) coupling martensitic transformation with mechanics and heat transfer is developed and applied for simulating eCE. The model is derived from a thermodynamic framework which invokes the microforce theory and Coleman--Noll procedure. To avoid the numerical issue related to the non-differentiable energy barrier function across the transition point, the austenite-martensite transition energy barrier in PFM is constructed as a smooth function of temperature. Both the indirect method using isothermal PFM with Maxwell relations and the direct method using non-isothermal PFM are applied to calculate the elastocaloric properties. The former is capable of calculating both isothermal entropy change and adiabatic temperature change ($ΔT_{\text{ad}}$), but induces high computation cost. The latter is computationally efficient, but only yields $ΔT_{\text{ad}}$. In a model Mn-22Cu alloy, the maximum $ΔT_{\text{ad}}$ ($ΔT_{\text{ad}}^{\text{max}}$) under a compressive stress of 100 MPa is calculated as 9.5 and 8.5 K in single crystal (3.5 and 3.8 K in polycrystal) from the indirect and direct method, respectively. It is found that the discrepancy of $ΔT_{\text{ad}}^{\text{max}}$ by indirect and direct method is within 10% at stress less than 150 MPa, confirming the feasibility of both methods in evaluating eCE at low stress. The results demonstrate the developed PFM herein, combined with both indirect and direct method for eCE calculations, as a practicable toolkit for the computational design of elastocaloric devices.

cond-mat.mtrl-sci↗

arXiv2023

Modeling and simulation of sintering process across scales

Sintering, as a thermal process at elevated temperature below the melting point, is widely used to bond contacting particles into engineering products such as ceramics, metals, polymers, and cemented carbides. Modelling and simulation as important complement to experiments are essential for understanding the sintering mechanisms and for the optimization and design of sintering process. We share in this article a state-to-the-art review on the major methods and models for the simulation of sintering process at various length scales. It starts with molecular dynamics simulations deciphering atomistic diffusion process, and then moves to microstructure-level approaches such as discrete element method, Monte--Carlo method, and phase-field models, which can reveal subtle mechanisms like grain coalescence, grain rotation, densification, grain coarsening, etc. Phenomenological/empirical models on the macroscopic scales for estimating densification, porosity and average grain size are also summarized. The features, merits, drawbacks, and applicability of these models and simulation technologies are expounded. In particular, the latest progress on the modelling and simulation of selective and direct-metal laser sintering based additive manufacturing is also reviewed. Finally, a summary and concluding remarks on the challenges and opportunities are given for the modelling and simulations of sintering process.

cond-mat.mtrl-sci↗

arXiv2023

Tailoring magnetic hysteresis of Fe-Ni permalloy by additive manufacturing: Multiphysics-multiscale simulations of process-property relationships

Designing the microstructure of Fe-Ni permalloy by additive manufacturing (AM) opens new avenues to tailor the materials' magnetic properties. Yet, AM-produced parts suffer from spatially inhomogeneous thermal-mechanical and magnetic responses, which are less investigated in terms of process simulation and modeling schemes. Here we present a powder-resolved multiphysics-multiscale simulation scheme for describing magnetic hysteresis in materials produced via AM. The underlying physical processes are explicitly considered, including the coupled thermal-structural evolution, chemical order-disorder transitions, and associated thermo-elasto-plastic behaviors. The residual stress is identified as the key thread in connecting the physical processes and in-process phenomena across scales. By employing this scheme, we investigate the dependence of the fusion zone size, the residual stress and plastic strain, and the magnetic hysteresis of AM-produced Fe21.5Ni78.5 permalloy on beam power and scan speed. Simulation results also suggest a phenomenological relation between magnetic coercivity and average residual stress, which can guide the magnetic hysteresis design of soft magnetic materials by choosing appropriate AM-process parameters.

cond-mat.mtrl-sci↗

arXiv2022

AI enhanced finite element multiscale modelling and structural uncertainty analysis of a functionally graded porous beam

The local geometrical randomness of metal foams brings complexities to the performance prediction of porous structures. Although the relative density is commonly deemed as the key factor, the stochasticity of internal cell sizes and shapes has an apparent effect on the porous structural behaviour but the corresponding measurement is challenging. To address this issue, we are aimed to develop an assessment strategy for efficiently examining the foam properties by combining multiscale modelling and deep learning. The multiscale modelling is based on the finite element (FE) simulation employing representative volume elements (RVEs) with random cellular morphologies, mimicking the typical features of closed-cell Aluminium foams. A deep learning database is constructed for training the designed convolutional neural networks (CNNs) to establish a direct link between the mesoscopic porosity characteristics and the effective Youngs modulus of foams. The error range of CNN models leads to an uncertain mechanical performance, which is further evaluated in a structural uncertainty analysis on the FG porous three-layer beam consisting of two thin high-density layers and a thick low-density one, where the imprecise CNN predicted moduli are represented as triangular fuzzy numbers in double parametric form. The uncertain beam bending deflections under a mid-span point load are calculated with the aid of Timoshenko beam theory and the Ritz method. Our findings suggest the success in training CNN models to estimate RVE modulus using images with an average error of 5.92%. The evaluation of FG porous structures can be significantly simplified with the proposed method and connects to the mesoscopic cellular morphologies without establishing the mechanics model for local foams.

cs.LG↗

arXiv2022

A thermo-mechanical phase-field fracture model: application to hot cracking simulations in additive manufacturing

Thermal fracture is prevalent in many engineering problems and is one of the most devastating defects in metal additive manufacturing. Due to the interactive underlying physics involved, the computational simulation of such a process is challenging. In this work, we propose a thermo-mechanical phase-field fracture model, which is based on a thermodynamically consistent derivation. The influence of different coupling terms such as damage-informed thermomechanics and heat conduction and temperature-dependent fracture properties, as well as different phase-field fracture formulations, are discussed. Finally, the model is implemented in the finite element method and applied to simulate the hot cracking in additive manufacturing. Thereby not only the thermal strain but also the solidification shrinkage are considered. As for thermal history, various predicted thermal profiles, including analytical solution and numerical thermal temperature profile around the melting pool, are regarded, whereas the latter includes the influence of different process parameters. The studies reveal that the solidification shrinkage strain takes a dominant role in the formation of the circumferential crack, while the temperature gradient is mostly responsible for the central crack. Process parameter study demonstrates further that a higher laser power and slower scanning speed are favorable for keyhole mode hot cracking while a lower laser power and quicker scanning speed tend to form the conduction mode cracking. The numerical predictions of the hot cracking patterns are in good agreement with similar experimental observations, showing the capability of the model for further studies.

cond-mat.mtrl-sci↗

arXiv2022

Domain size and charge defects on the polarization switching of antiferroelectric domains

The switching behavior of antiferroelectric domain structures under the applied electric field is not fully understood. In this work, by using the phase field simulation, we have studied the polarization switching property of antiferroelectric domains. Our results indicate that the ferroelectric domains nucleate preferably at the boundaries of the antiferroelectric domains, and antiferroelectrics with larger initial domain sizes possess a higher coercive electric field as demonstrated by hysteresis loops. Moreover, we introduced charge defects into the sample and numerically investigated their influence. It is also shown that charge defects can induce local ferroelectric domains, which could suppress the saturation polarization and narrow the enclosed area of the hysteresis loop. Our results give insights into understanding antiferroelectric phase transformation and optimizing the energy storage property in experiments.

cond-mat.mtrl-sci↗

arXiv2022

A mixed formulation for physics-informed neural networks as a potential solver for engineering problems in heterogeneous domains: comparison with finite element method

Physics-informed neural networks (PINNs) are capable of finding the solution for a given boundary value problem. We employ several ideas from the finite element method (FEM) to enhance the performance of existing PINNs in engineering problems. The main contribution of the current work is to promote using the spatial gradient of the primary variable as an output from separated neural networks. Later on, the strong form which has a higher order of derivatives is applied to the spatial gradients of the primary variable as the physical constraint. In addition, the so-called energy form of the problem is applied to the primary variable as an additional constraint for training. The proposed approach only required up to first-order derivatives to construct the physical loss functions. We discuss why this point is beneficial through various comparisons between different models. The mixed formulation-based PINNs and FE methods share some similarities. While the former minimizes the PDE and its energy form at given collocation points utilizing a complex nonlinear interpolation through a neural network, the latter does the same at element nodes with the help of shape functions. We focus on heterogeneous solids to show the capability of deep learning for predicting the solution in a complex environment under different boundary conditions. The performance of the proposed PINN model is checked against the solution from FEM on two prototype problems: elasticity and the Poisson equation (steady-state diffusion problem). We concluded that by properly designing the network architecture in PINN, the deep learning model has the potential to solve the unknowns in a heterogeneous domain without any available initial data from other sources. Finally, discussions are provided on the combination of PINN and FEM for a fast and accurate design of composite materials in future developments.

cs.CE↗

arXiv2022

Phase-field modeling of paramagnetic austenite-ferromagnetic martensite transformation coupled with mechanics and micromagnetics

A three-dimensional phase-field model is proposed for simulating the magnetic martensitic phase transformation. The model considers a paramagnetic cubic austenite to ferromagnetic tetragonal martensite transition, as it occurs in magnetic Heusler alloys like Ni2 MnGa, and is based on a Landau 2-3-4 polynomial with temperature dependent coefficients. The paramagnetic-ferromagnetic transition is recaptured by interpolating the micromagnetic energy as a function of the order parameter for the ferroelastic domains. The model is numerically implemented in real space by finite element (FE) method. FE simulations in the martensitic state show that the model is capable to correctly recapture the ferroelastic and -magnetic microstructures, as well as the influence of external stimuli. Simulation results indicate that the paramagnetic austenite to ferromagnetic martensite transition shifts towards higher temperatures when a magnetic field or compressive stress is applied. The dependence of the phase transition temperature shift on the strength of the external stimulus is uncovered as well. Simulation of the phase transition in magnetocaloric materials is of high interest for the development of energy-efficient magnetocaloric cooling devices.

cond-mat.mtrl-sci↗

arXiv2021

A deep learned nanowire segmentation model using synthetic data augmentation

Automatized object identification and feature analysis of experimental image data are indispensable for data-driven material science; deep-learning-based segmentation algorithms have been shown to be a promising technique to achieve this goal. However, acquiring high-resolution experimental images and assigning labels in order to train such algorithms is challenging and costly in terms of both time and labor. In the present work, we apply synthetic images, which resemble the experimental image data in terms of geometrical and visual features, to train state-of-art deep learning-based Mask R-CNN algorithms to segment vanadium pentoxide (V2O5) nanowires, a canonical cathode material, within optical intensity-based images from spectromicroscopy. The performance evaluation demonstrates that even though the deep learning model is trained on pure synthetically generated structures, it can segment real optical intensity-based spectromicroscopy images of complex V2O5 nanowire structures in overlapped particle networks, thus providing reliable statistical information. The model can further be used to segment nanowires in scanning electron microscopy (SEM) images, which are fundamentally different from the training dataset known to the model. The proposed methodology of using a purely synthetic dataset to train the deep learning model can be extended to any optical intensity-based images of variable particle morphology, extent of agglomeration, material class, and beyond.

physics.data-an↗

arXiv2021

Lossless Multi-Scale Constitutive Elastic Relations with Artificial Intelligence

The elastic properties of materials derive from their electronic and atomic nature. However, simulating bulk materials fully at these scales is not feasible, so that typically homogenized continuum descriptions are used instead. A seamless and lossless transition of the constitutive description of the elastic response of materials between these two scales has been so far elusive. Here we show how this problem can be overcome by using Artificial Intelligence (AI). A Convolutional Neural Network (CNN) model is trained, by taking the structure image of a nanoporous material as input and the corresponding elasticity tensor, calculated from Molecular Statics (MS), as output. Trained with the atomistic data, the CNN model captures the size- and pore-dependency of the material's elastic properties which, on the physics side, can stem from surfaces and non-local effects. Such effects are often ignored in upscaling from atomistic to classical continuum theory. To demonstrate the accuracy and the efficiency of the trained CNN model, a Finite Element Method (FEM) based result of an elastically deformed nanoporous beam equipped with the CNN as constitutive law is compared with that by a full atomistic simulation. The good agreement between the atomistic simulations and the FEM-AI combination for a system with size and surface effects establishes a new lossless scale bridging approach to such problems. The trained CNN model deviates from the atomistic result by 9.6\% for porosity scenarios of up to 90\% but it is about 230 times faster than the MS calculation and does not require to change simulation methods between different scales. The efficiency of the CNN evaluation together with the preservation of important atomistic effects makes the trained model an effective atomistically-informed constitutive model for macroscopic simulations of nanoporous materials and solving of inverse problems.

cond-mat.mtrl-sci↗

arXiv2021

Humidity Influence on Mechanics and Failure of Paper Materials: Joint Numerical and Experimental Study on Fiber and Fiber Network Scale

Paper materials are natural composite materials and well-known to be hydrophilic unless chemical and mechanical processing treatments are undertaken. The relative humidity impacts the fiber elasticity, the fiber-fiber bonds and the failure mechanism. In this work, we present a comprehensive experimental and computational study on the mechanical and failure behaviour of the fiber and the fiber network under humidity influence. The manually extracted cellulose fiber is exposed to different levels of humidity, and then mechanically characterized using Atomic Force Microscopy, which delivers the humidity dependent longitudinal Young's modulus. The obtained relationship allows calculation of fiber elastic modulus at any humidity level. Moreover, by using Confoncal Laser Scanning Microscopy, the coefficient of hygroscopic expansion of the fibers is determined. On the other hand, we present a finite element model to simulate the deformation and the failure of the fiber network. The model includes the fiber anisotropy and the hygroscopic expansion using the experimentally determined constants. In addition, it regards the fiber-fiber bonding and damage by using a humidity dependent cohesive zone interface model. Finite element simulations on exemplary fiber network samples are performed to demonstrate the influence of different aspects including relative humidity and fiber-fiber bonding parameters on the mechanical features such as force-elongation curves, wet strength, extensiability and the local fiber-fiber debonding. In meantime, fiber network failure in a locally wetted region is revealed by tracking of individually stained fibers using in-situ imaging techniques. Both the experimental data and the cohesive finite element simulations demonstrate the pull-out of fibers and imply the significant role of the fiber-fiber debonding in the failure process of the wet paper.

physics.comp-ph↗