Computational Particle Mechanics最新文献

Three-dimensional concrete printing (3DCP) has emerged as a promising manufacturing technique in the civil engineering sector, offering significant advantages over traditional construction methods. Despite its potential, challenges persist in optimizing the deposition process, particularly regarding the rheological characteristics of printable concrete, which affect filament formation and stability during printing. In this study, we implement the discrete fresh concrete (DFC) model within an in-house formulation of the discrete element method (DEM) to simulate the rheological behavior of fresh concrete during extrusion. While using our implementation to analyze how printing speed and concrete mixture affect the quality of single-layer printing, we identify limitations in the original DFC model to properly represent particle-particle and particle-surface tangential interactions, particularly in energy dissipation mechanisms. To improve the model’s robustness, we suggest modifications to account for static friction and rolling resistance in the material model. Once these limitations are overcome, our simulation results indicate that the enhanced DFC framework can provide valuable insights into the printing process, including filament formation and layer continuity and stability, and may be a useful tool for process optimization in 3DCP applications.

In continuum-based particle methods, the adoption of polygon meshes for solid boundary modeling provides more efficient and smoother representation of complex-shaped solid boundaries. However, the computation of the particle-mesh distances might be an additional time-consuming task if naively performed. This additional computational time, in turn, may become a bottleneck for practical modeling of fluid–structure interaction (FSI) problems involving complex-shaped moving or deformable bodies. In the present work, we propose a technique named point-mesh search based on cell-linked list (PSCL), aiming to increase the computational efficiency of the particle-mesh search procedure. It is a variant of strategies that takes advantage of the cell-linked list structure to narrow the search domain. We investigated the behaviors of the particle-mesh searching techniques’ performances by considering FSI cases with violent free-surface deformations, such as relatively simple geometry dam breaking hitting a block for validation of the impact loads; water entry of a parabolic cylinder modeled as moving body or as fixed body with bottom inflow to assess the overhead due to moving meshes; and impact loads on freefall lifeboats to evaluate the computational efficiency in practical engineering applications. The results show that PSCL is robust and reliable and achieves significant speedups for real-world applications with complex-shaped moving bodies, demanding very low overhead for moving meshes.

Graphical Abstract

In this study, we present an implicit incompressible smoothed particle hydrodynamics method designed for simulating fluid dynamics problems where rotational behavior is a key feature of the flow. Crucial to this task is the calculation of the viscous term of the Navier–Stoker equation, because most SPH operators for the Laplacian of velocity are either highly inaccurate or produce spurious resistance to rotation. Here, we introduce a novel equation for calculating the Laplacian of velocity, designed as a compromise between mathematical accuracy and the local conservation of angular momentum. The proposed method is validated through a series of simulations of the coiling behavior of highly viscous fluids, demonstrating its ability to naturally generate this phenomenon and accurately predict the cessation of coiling at appropriate heights. In addition, we validate the proposed method in terms of quantitatively reproduce the angular frequency of the coiling for different values of height with satisfactory results.

The lattice Boltzmann method (LBM) has been successfully applied to the simulation of fluid flows for over three decades. In recent years, it has also been extended to solid mechanics, particularly for elastodynamics. This work presents a comprehensive introduction to the moment chain LBM for solids, focusing on the two-relaxation-time (TRT) scheme. The method is based on a chain of balance equations, which allows for the simulation of wave propagation in elastic solids. The TRT scheme improves stability and accuracy, making it suitable for a wide range of material parameters. The method is applied to wave propagation in solids with an analysis of the energy dissipation. The results demonstrate the effectiveness of the moment chain LBM for simulating elastodynamics and highlight its potential for future applications in solid mechanics.

This study investigates the size-dependent breakage force of brittle, near-spherical particles under quasi-static compressive loading. A physically motivated analytical model is derived, predicting a power-law relationship between particle size and breakage force—specifically, a quadratic scaling with diameter and a 2/3-power scaling with mass. To validate this model, a multi-material experimental campaign using synthetically produced spherical specimens was conducted, followed by a numerical simulation campaign using the Discrete Element Method (DEM) with a bonded particle modelling (BPM) approach.

The experimental results confirmed the proposed scaling law, with coefficients of determination (R²) exceeding 0.94 for a combined material consideration. The DEM simulations, calibrated using experimental data, reproduced breakage forces, and scaling patterns with high fidelity and enabled extension of the size range by a factor of more than two in diameter and four in mass, whilst also increasing statistical resolution through a higher number of replicates per size class. This numerical extension not only enabled broader parameter exploration but also mitigated experimental limitations, such as specimen variability, preparation inconsistency, and practical size constraints.Across analytical, experimental, and numerical approaches, consistent agreement was found, supporting the general applicability of the model. The findings provide a robust basis for defining breakage thresholds in DEM-based simulations—e.g. for replacement-based approaches—and offer a scalable alternative to physical testing. The validated framework facilitates improved prediction of breakage behaviour in slow compression systems such as jaw crushers and contributes to the broader understanding of particle-scale mechanics in brittle materials.

This paper provides a methodological overview of the current state of the art in discrete element modeling of rock fracture in the context of comminution, an energy-intensive process of breaking down rocks into smaller sizes. This process is essential for liberating valuable metals and minerals that are in growing demand for the green transition and the electrification of society. The paper covers the most recent developments and addresses fundamental issues in the bonded discrete element method, the lattice element method, the particle replacement method, and the level-set discrete element method. We argue that the most effective modeling approach must emerge from a synergy between solid mechanics, rock mechanics, and the comminution field—an effort made by this collaborating multidisciplinary group, with the goal of making the next generation of comminution models, powered by GPU-accelerated high-performance computing, more reflective of real-life rock behavior, advancing energy-efficient mining.

The simulation of fluid–structure interaction problems involving structural contact has a wide range of applications in engineering as well as in biomechanical contexts, such as the actuation of heart valves. However, simulating these problems presents significant challenges, particularly due to the occurrence of topological changes in the fluid domain, dynamic solid-to-solid contact, and the complexities of fluid–structure coupling. This paper presents a monolithic numerical framework for simulating fluid–structure interaction problems involving structural contact. The method combines a particle-position-based formulation of the particle finite element method (PFEM) for fluid dynamics with a position-based total Lagrangian formulation for nonlinear solid mechanics, and employs a node-to-segment algorithm with Lagrange multipliers to handle structural contact. The PFEM effectively addresses topological changes in the fluid domain by integrating remeshing techniques with the particle concept, while the node-to-segment algorithm proves to be a reliable method for managing structural contact, even in scenarios involving complex geometries. The particle-position-based approach uses nodal positions as primary variables, making the monolithic coupling with the total Lagrangian position-based formulation for the solid straightforward and efficient, resulting in a framework that is particularly effective for strongly coupled problems. The proposed approach is tested through two-dimensional numerical examples, demonstrating its robustness and potential for biomedical and engineering applications.

In many engineering applications, materials are subjected to high loading rates and severe deformation, where the coupled effects of temperature and mechanical response become significant. This study presents an element-free Galerkin meshless method for modeling the thermo-viscoplastic behavior of materials under dynamic loading conditions. The formulation is derived based on first-order conservation laws for linear momentum, deformation gradient tensor, volume map, area map, and entropy, within a total Lagrangian framework. A variational multiscale stabilization approach is employed to ensure numerical robustness. The Johnson-Cook model is incorporated to capture strain rate sensitivity in plastic deformation. The method is validated through plasticity benchmark problems, including the modeling of Taylor impact test, necking bar, and equal channel angular pressing process. Furthermore, the temperature-dependent phase transformation behavior of shape memory alloys is simulated under dynamic loading. Results demonstrate the capability of the proposed method to accurately capture the complex interplay between thermal and mechanical effects in highly nonlinear scenarios.

Cardiovascular diseases are the leading cause of mortality worldwide, with projections indicating a concerning rise in related deaths. Computational models offer promising tools to understand the hemodynamics and biomechanical mechanisms underlying vascular failure. In particular, Fluid-Structure Interaction (FSI) algorithms have found significant applications in cardiovascular engineering. This study aims at demonstrating the relevance of the Particle Finite Element Method (PFEM). to model fluid–structure interactions between artery walls and blood flows, and assess the corresponding biomechanical aspects. For this, the flow–structure interaction problem is addressed using a partitioned approach with a strong coupling of the PFEM (for the fluid) and FEM (for the solid) models. Both Newtonian and Casson fluid models, as well as a Mooney–Rivlin hyperelastic model for the deformation of blood vessels, are incorporated. The numerical simulations successfully describe a wide range of situations, from the ejection of blood from the left ventricle to the dynamics of an abdominal aortic aneurysm. To the best of our knowledge, this work describes the very first applications of the PFEM to the study of blood flows in FSI simulations. It is also original by the explicit description of the rupture of the artery wall. Although the model could still be improved, for instance by introducing a turbulence model to deal with high–speed flow through the valve or considering anisotropic hyperelastic models for vessels, the results demonstrate the high potential of this method for describing the interactions of blood flows with the deforming artery walls.

In recent years, particles have gained popularity as crash absorbers. To improve their mechanical properties, a coating layer can be applied. To predict the effect of this coating, a numerical model must be developed. For this purpose, the present study employs the discrete element method, extended by the bonded particle method, using both high- and low-fidelity approaches. In this framework, a single physical particle is modelled as a cluster or agglomerate of smaller particles bonded together. To identify the parameters involved, a sensitivity analysis is performed, followed by optimisation using the particle swarm algorithm, with calibration based on uniaxial single particle compression tests. Once an optimised parameter set is obtained, the models are validated against multi particle compression test results. The outcomes of this study demonstrate the potential of the proposed methodology for simulating large-scale compression tests of coated granular materials.