Particle Model Research Articles

A Numerical Study of the Mechanical Behavior of Jointed Soft Rocks under Triaxial Loading Using a Bonded Particle Model.

In order to master the strength and deformation characteristics, including the macro-micro failure mechanism of soft rock samples with penetrating joints under triaxial loading, a series of numerical triaxial tests have been carried out. The strength and deformation characteristics, failure modes, crack propagation, distribution of force chains, and the influences of joint dip angles and confining pressures have been analyzed and compared with the laboratory test results. The results show that (1) the residual strength ratio of jointed rock samples generally increases first and then decreases with the increase in joint dip angles under the same confining pressure and reaches the maximum value around 23-24°. Poisson's ratio increases with the increase in the confining pressure or the joint dip angle. The elastic modulus increases with the increase in the confining pressure and decreases with the increase in the joint dip angle. (2) The jointed rock samples with different joint dip angles compact with relatively small volumetric strains and then dilate up to failure with relatively large volume expansions. Lower confining pressure and smaller dip angles will lead to a more pronounced dilation phenomenon and less obvious volume shrinkage rules. (3) The low-angle jointed rock samples all exhibit the X-type shear failure. The jointed rock samples with a joint dip angle of 45° exhibit hybrid failure with both slippage and shearing, which are controlled by both the matrix and the joint. (4) The change in the number of cracks includes three stages: the slow crack initiation stage, rapid growth stage, and crack coalescence stage. The total number of shear or tensile cracks all decrease with an increase in the joint dip angles, with the number of tensile cracks being approximately twice that of shear cracks. The tension cracks are mostly horizontal, and the shear cracks are mostly vertical. (5) The number of force chains shows a decreasing trend after the cracks begin to grow. The jointed rock samples for the intact, 15° and 30° cases all form a main force chain during the failure process, while there is no main force chain for the 45° case.

Boosting predictive accuracy of single particle models for lithium-ion batteries using machine learning

Boosting predictive accuracy of single particle models for lithium-ion batteries using machine learning

Modeling and motion analysis of flexible legged robots using the finite particle method

Robotics with flexible legs have attracted significant attention. Engineers often design and analyze the motion of legged robots from kinematic and biomimetic perspective. However, the influence of flexibility of the feet on robot locomotion is often not given sufficient considerations, which is also very crucial to the motion posture of the flexible legged robots, especially as the soft robots design becomes increasingly popular. The mainly difficulties lies in the traditional numerical methods in handling the dynamic motion analysis with both large rigid motion and large deformation. In this paper, the finite particle method (FPM) is used to simulate the motion and deformation coupled problems of the flexible six-legged robot. A shell-based particle model of a six-leg robot and the contact model between legs and ground are built. Without iterative and modification of the FPM analytical framework, structural nonlinearity is efficiently handled after eliminating rigid body motions by a fictitious reverse motion. The motion and deformation of a single leg with varying leg thickness, locomotion speed, and leg-to-ground friction coefficients were simulated. By analyzing the stress distribution in the leg and the number of contact points with the ground, the mechanical leg was optimized in design. Furthermore, the motion and deformation of the entire six-legged robot were simulated using the FPM. The numerical results' feasibility was validated through comparison with experimental data obtained from robot walking tests. The proposed method effectively simulates the motion and deformation of flexible robots, providing significant insights for the design of soft robots.

Calibration of Discrete Element Method Parameters for a High-Fidelity Lunar Regolith Simulant Considering the Effects of Realistic Particle Shape

The Discrete Element Method (DEM) is an important tool for investigating the geotechnical properties of lunar regolith. The accuracy of DEM simulations largely depends on precise particle modeling and the appropriate selection of mesoscopic parameters. To enhance the reliability and accuracy of the DEM in lunar regolith studies, this paper utilized the high-fidelity IRSM-1 lunar regolith simulant to construct a DEM model with realistic particle shapes and conducted an angle of repose (AoR) simulation test. The optimal DEM parameters were calibrated using a combination of the Plackett–Burman test, steepest ascent test, and Box–Behnken design. The results indicate that the sliding friction coefficient, rolling friction coefficient, and surface energy significantly influence the simulation AoR. By optimizing against the measured AoR using a second-order regression model, the optimal parameter values were determined to be 0.633, 0.401, and 0.2, respectively. Under these optimal parameters, the error between the simulation and experimental AoR was 2.1%. Finally, the calibrated mesoscopic parameters were validated through a lifting cylinder test, showing an error of 6.3% between the simulation and experimental results. The high similarity in the shape of the AoR further confirms the accuracy and reliability of the parameter calibration method. This study provides a valuable reference for future DEM-based research on the mechanical and engineering properties of lunar regolith.

CFD-DEM simulation and experimental validation of air classification for tobacco particles

In cigarette processing, the challenge is particularly significant during the air classification of tobacco and stems due to their similar characteristics. This paper employs computational fluid dynamics (CFD) combined with the discrete element method (DEM) to analyze factors affecting separation efficiency and improve performance. The results were validated through laboratory and production line experiments at a 1:1 scale. The tobacco (length: 0 mm–4.75 mm, diameter: 0.32 mm) and stem (length: 0–25.145 mm, diameter: 1.51 mm) were modeled based on production samples. Findings suggest that particle feeding speed primarily impacts tobacco loss rate, while inlet air velocity mainly influences stem removal rate. Optimizing the chamber structure in the simulation resulted in a 63.66 % improvement in separation efficiency. Airflow streamlines, particle distribution, trajectory, and collision behaviors were discussed to illuminate motion characteristics. The flexible particle model moderately influenced separation efficiency and collision behaviors. These insights enhance the understanding of particle separation and the design of separation devices.

Cdc42 mobility and membrane flows regulate fission yeast cell shape and survival

Polarized exocytosis induced by local Cdc42 GTPase activity results in membrane flows that deplete low-mobility membrane-associated proteins. A reaction-diffusion particle model comprising Cdc42 positive feedback activation, hydrolysis by GTPase-activating proteins (GAPs), and flow-induced displacement by exo/endocytosis shows that flow-induced depletion of low mobility GAPs promotes polarization. We modified Cdc42 mobility in Schizosaccharomyces pombe by replacing its prenylation site with 1, 2 or 3 repeats of the Rit C-terminal membrane-binding domain (ritC), yielding alleles with progressively lower mobility and increased flow-coupling. While Cdc42-1ritC cells are viable and polarized, Cdc42-2ritC polarize poorly and Cdc42-3ritC are inviable, in agreement with model’s predictions. Deletion of Cdc42 GAPs restores viability to Cdc42-3ritC cells, verifying the model’s prediction that GAP deletion increases Cdc42 activity at the expense of polarization. Our work demonstrates how membrane flows are an integral part of Cdc42-driven pattern formation and require Cdc42-GTP to turn over faster than the surface on which it forms.

Simulation and time series analysis of responsive active Brownian particles (rABPs) with memory

To realise the goals of active matter at the micro- and nano-scale, the next generation of microrobots must be capable of autonomously sensing and responding to their environment to carry out pre-programmed tasks. Memory effects are proposed to have a significant effect on the dynamics of responsive robotic systems, drawing parallels to strategies used in nature across all length-scales. Inspired by the integral feedback control mechanism by which Escherichia coli (E. coli) are proposed to sense their environment, we develop a numerical model for responsive active Brownian particles (rABP) in which the rABPs continuously react to changes in the physical parameters dictated by their local environment. The resulting time series, extracted from their dynamic diffusion coefficients, velocity or from their fluctuating position with time, are then used to classify and characterise their response, leading to the identification of conditional heteroscedasticity in their physics. We then train recurrent neural networks (RNNs) capable of quantitatively describing the responsiveness of rABPs using their 2D trajectories. We believe that our proposed strategy to determine the parameters governing the dynamics of rABPs can be applied to guide the design of microrobots with physical intelligence encoded during their fabrication.

Investigation of Al-Li particle ignition dynamics with different Li content

Promoting the ignition of aluminum powder is considered an effective method to inhibit the agglomeration of aluminum powder in propellants and enhance combustion efficiency. This study utilizes laser ignition technology and high-speed photography to investigate the ignition and combustion processes of single aluminum particles and aluminum-lithium alloy particles. The focus is on comparing the effects of the diameter of micron-sized metal particles and the lithium content in aluminum-lithium alloy particles on the ignition and combustion processes of metal particles. The results show that the ignition delay time is directly proportional to the diameter of the metal particles and inversely proportional to the lithium content. For the aluminum-lithium alloy particle with a lithium content of 3.5 %, even if the diameter is close to 300 μm, the ignition delay time is only 125.5 ms, which is much smaller than that of the pure aluminum particle with a diameter of 208 μm. Compared to aluminum particles and aluminum-lithium alloy particles, there is basically no difference between the two during the combustion stage. However, in the ignition stage, aluminum-lithium alloy particles sequentially exhibit a red gas-phase flame corresponding to lithium and a yellow gas-phase flame corresponding to aluminum. This indicates that during the ignition process of aluminum-lithium alloy particles, lithium first reacts with the oxidative atmosphere and releases heat, providing a heat source for the subsequent ignition of aluminum particles. This also explains why the ignition delay time of metal particles is inversely proportional to the lithium content. An ignition model for aluminum particles in a multi-component atmosphere is established, which further considers the chemical reactions between lithium and oxidative gases, making the model applicable to aluminum-lithium alloy particles. This ignition model effectively describes the impact of particle diameter and lithium content on the ignition process of metal particles. The model is further verified, and the results show that the calculated ignition delay is in good agreement with the experimental data. Overall, this study provides deeper experimental and theoretical insights into the ignition and combustion processes of aluminum-lithium alloys, and the findings can guide the application of aluminum-lithium alloys in propellants.

An interprocess communication-based two-way coupling approach for implicit–explicit multiphysics lattice discrete particle model simulations

In this study, the researchers have developed a Multiphysics-Lattice Discrete Particle Model (M-LDPM) framework that deals with coupled-fracture-poroflow problems. The M-LDPM framework uses two lattice systems, the LDPM tessellation and the Flow Lattice Element (FLE) network, to represent the heterogeneous internal structure of typical quasi-brittle materials like concrete and rocks, and to simulate the material’s mechanical and transport behavior at the aggregate scale. The researchers revisited the LDPM governing equations and added the influence of fluid pore pressure. They also derived the Flow Lattice Model (FLM) governing equations for pore pressure flow through mass conservation balances for uncracked and cracked volumes. The M-LDPM framework was implemented using Abaqus user element subroutine VUEL for the explicit dynamic procedure of LDPM and user subroutine UEL for the implicit transient procedure of FLM. The coupling of the two models was achieved using Interprocess Communication (IPC) between Abaqus solvers. The M-LDPM framework can simulate the variation of permeability induced by fracturing processes by relating the transport properties of flow elements with local cracking behaviors. The researchers validated the M-LDPM framework by comparing the numerical simulation outcomes with analytical solutions of classical benchmarks in poromechanics.

Directional switching behavior of swarming systems with social and nonlinear interactions

Coordinated directional switching can occur among members of many mobile biological communities. Some studies show that self-propelled particle models can describe the directional switching behavior well. The key to understanding group movement is to determine the influential factors relevant to directional switching behavior. This paper focuses on the impact of social and nonlinear interactions on the directional switching behavior observed in swarming systems. In which, the nonlinear interaction is represented as a function of a trade-off between the velocity and velocity direction of its neighbors. Based on the framework of dimension reduction theory, the high-dimensional complex model is simplified into a one-dimensional simple model, and the stationary probability density and mean switching time are obtained by theoretical analysis of the one-dimensional model. It can be seen that social and nonlinear interactions play an important role in regulating the directional switching behaviors of swarming systems. Specifically, the increase of group density and nonlinear parameter can inhibit the directional switches. For Erdös-Rényi networks, the large mean degree can suppress the directional switching behavior. For scale-free networks, increasing the degree heterogeneity can reduce the mean switching time. The results reveal the underlying mechanisms by which social and nonlinear interactions influence the directional switching behaviors of swarming systems, and provide a theoretical foundation for the design of bio-inspired devices with specific functions.

How motility affects Ising transitions

Abstract We study a lattice gas (LG) model of hard-core particles on a square lattice experiencing nearest neighbour attraction J. Each particle has an internal orientation, independent of the others, that point towards one of the four nearest neighbour and it can move to the neighbouring site along that direction with the usual metropolis rate if the target site is vacant. The internal orientation of the particle can also change to any of the other three with a constant rate ω . The dynamics of the model in ω → ∞ reduces to that of the LG which exhibits a phase separation transition at particle density ρ = 1 2 and temperature T = 1 , when the strength of attraction J crosses a threshold value ln ⁡ ( 1 + 2 ) . This transition belongs to Ising universality class (IUC). For any finite ω > 0 , the particles can be considered as attractive run-and-tumble particles (RTPs) in two dimensions with motility ω − 1 . We find that RTPs also exhibit a phase separation transition, but the critical interaction required is J c ( ω ) which increases monotonically with increased motility ω − 1 . It appears that the transition belongs to IUC. Surprisingly, in these models, motility impedes cluster formation process necessitating higher interaction to stabilize microscopic clusters. Moreover, MIPS like phases are not found when J = 0.

A comprehensive study on the validation and application of multi-lognormal distribution models for atmospheric particles

The particle number size distribution (PNSD) is crucial for evaluating air quality and mitigating environmental pollution, as particles of different sizes have diverse effects on human health and climate. However, obtaining a comprehensive understanding of PNSD is challenging due to its inherent complexities and variability. Multi-lognormal distribution models are employed to fit PNSD, as seen in climate models, but discrepancies between model fits and observed PNSD persist. This study adopts hourly data from 2017 to 2020 across eight monitoring sites in diverse environments—rural, urban, mountainous, and polar, and compares the observed PNDS with those simulated by multi-lognormal distribution models. The results demonstrated that the model generally achieved a high correlation with observed PNSD data (r2 > 0.75), effectively capturing key characteristics of nucleation and Aitken mode particles. However, the model had a tendency to overestimate the total number concentration by approximately 1.09 times, particularly noticeable under conditions of high concentrations of smaller particles. The model successfully represented prevalent bimodal size distribution patterns in urban areas with high ultrafine particle concentrations, though its performance was slightly less accurate in scenarios involving trimodal distributions. Despite these strong correlations and the model's ability to reflect diurnal and seasonal variations, which suggests its broad applicability and utility, there were notable limitations on smaller time scales and in specific particle size ranges. These limitations were particularly evident in capturing detailed phenomena relevant to new particle formation events, indicating areas where model refinement is necessary. The results highlighted the importance of investigating discrepancies between model predictions and actual observations, which is crucial for refining climate models that utilize PNDS. The uniform comparison facilitated a detailed exploration of particle properties from model results, offering deeper insights into aerosol behavior and its environmental impacts.

A modified pore evolution particle model applied to CaCO3 decomposition and CaO sintering during calcium looping CO2 cycles

The calcium looping of calcium-based carbonation/calcination cyclic reaction is an efficient decarbonization technology that can be applied to post-combustion carbon capture, solar thermal storage, and enhanced methane reforming hydrogen. The pore structure generated during the calcination stage greatly affects adsorption performance. Hence, the change of pore structure is crucial during the calcination. Existing models related to pore structure still need to be further improved regarding pore formation mechanisms. In this paper, a modified pore evolution model is proposed. The modified pore evolution model contains the kinetic equation, the CO2 diffusion equation, and the vacancy aggregates equation. The modified model considers the possibility of spontaneous pore formation in the mechanism, which can better fit the experimental results and provide higher simulation accuracy. The model reveals the diffusion of CO2 within the particles during decomposition, and spatiotemporal evolutionary properties of pore parameters. The pore change trend is the most obvious when the initial pores are mainly 3 nm mesopores. The pore evolution is relatively stable when the initial pores are mainly 25 nm mesopores. Besides, the particles experience significant pore migration during the sintering stage when the temperature reaches 1273 K. At 1073 K, the maximum pore size occurring during pore evolution did not exceed 40 nm. The model estimates a maximum pore size of more than 60 nm at 1273 K. The modified model can aid in a deeper comprehension of how the pore structure changes throughout the calcination process.

Research on alarm of coal content in gangue based on binocular vision and YOLACT segmentation network

ABSTRACT The coal content in the gangue of coal preparation plants cannot be detected online, especially in the jigging sorting process. There is a large time lag in manual testing. An approach based on the YOLACT segmentation network and binocular vision was proposed. First, the YOLACT segmentation network was used to identify coal and gangue. At the same time, the coal and gangue particle regions are segmented from the background of the bucket elevator. A binocular matching method based on deep learning was proposed to obtain better disparity maps and establish better volume models of coal and gangue particles. A stereo-matching global fusion network model was also presented by improving the Pyramid Stereo Matching Network. Finally, the coal and gangue volume parameters were combined with their respective empirical densities to predict the coal content in gangue of the bucket. When the load in the bucket elevator is small, the average prediction error of coal content in gangue is 9.19%. When the load is large i.e. the coal and gangue are stacked in multiple layers, it is not appropriate to calculate the coal content in gangue due to serious particle coverage. Therefore, a fuzzy alarm strategy was proposed based on actual needs. The situation of coal content in gangue was divided into three levels: normal, warning, and alarm. It meets the alarm requirements for coal content in gangue in most industrial plants.

Universality classes in out-of-equilibrium systems: An encompassing theorem for a one-dimensional fusing particles model

Universality classes in out-of-equilibrium systems: An encompassing theorem for a one-dimensional fusing particles model

Investigation on cutting forces in ultrasonic vibration assisted milling of particle-reinforced titanium matrix composites (PTMCs)

Particle-reinforced titanium matrix composites are widely used in aerospace and marine fields because of their excellent performance. However, the difficulty of machining is increased by the existence of highly brittle and hard reinforcing particles dispersed in the matrix. To address the challenges of the difficulty of subsequent processing of particle-reinforced titanium matrix composites, an ultrasonic vibration-assisted process is used for milling. In this paper, an orthogonal cutting simulation model of particle-reinforced titanium matrix composites’ material is established by ABAQUS software, and the constitutive material models of particles and metallic matrix are established respectively. The excellent performance of ultrasonic vibration assisted cutting is dynamically analyzed from the simulation micro angle. Meanwhile, combined with the conventional milling and ultrasonic vibration assisted cutting experiments, the influence of the two processing methods on the cutting performance of grain-reinforced titanium matrix composite material is compared. The results show that, compared with conventional milling, the intermittent cutting mode of ultrasonic vibration-assisted cutting can significantly reduce the cutting force by about 10%; in both cutting modes, with the same cutting conditions, the change in cutting force is positively correlated with the cutting speed, feed per tooth, and cutting width.

Wear particles image enhancement using long short-term memory 3D recurrent reconstruction neural network (LSTM 3D-R2N2)

3D modeling of wear particles has proven to be a useful tool for monitoring mechanical failure conditions. In this work, a new method for 3D reconstruction of wear particles in uncontaminated oil (healthy oil) and contaminated oil (used oil) was proposed. The image acquisition device can capture multi-view images of moving wear particles in both healthy and used oil by using the reflected light. The images were pretreated first, and the image color inversion was conducted using the Pillow library. The pretreated wear particle images were used for 3D reconstruction using long short-term memory 3D recurrent reconstruction neural network. The current results were verified against existing results, and good agreement can be found. It can be concluded that we can reconstruct the similar 3D wear particle results with fewer images by comparison with other methods. Specifically, only 4–6 image samples were used for the 3D reconstruction of wear particles, and at least 8 image samples were needed for other existing reports.

Quasi particle model vs lattice QCD thermodynamics: extension to Nf=2+1+1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$N_f=2+1+1$$\\end{document} flavors and momentum dependent quark masses

In the last decade a quasi-particle model (QPM) has supplied the basis for the study of heavy quark (HQ) production in ultra-relativistic collisions, allowing for a phenomenological estimate of the HQ diffusion coefficient Ds(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$D_s(T)$$\\end{document}. Using the new lattice QCD results for the equation of state (EoS) with 2+1+1 dynamical flavors, we extend the QPM from Nf=2+1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$N_f=2+1$$\\end{document} to Nf=2+1+1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$N_f=2+1+1$$\\end{document}, where the charm quark is included. Fixing the coupling g(T) by a fit to the lQCD energy density ϵ(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\epsilon (T)$$\\end{document}, we evaluate the impact of different temperature parametrizations of charm quark mass on EoS and susceptibilities χq(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\chi _q(T)$$\\end{document} of light, χs(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\chi _s(T)$$\\end{document} of strange and χc(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\chi _c(T)$$\\end{document} of charm quarks, the last favouring a charm quark mass increasing toward Tc\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$T_c$$\\end{document}. We also explore the extension of the QPM to a more realistic approach called QPMp\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$_p$$\\end{document}, where quark and gluon masses explicitly depend on their momentum converging to the current quark mass at high momenta, as expected from asymptotic free dynamics. The QPMp\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$_p$$\\end{document} allows for a simultaneous quantitative description not only of the EoS but also of the quark susceptibilities (χq(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\chi _q(T)$$\\end{document}, χs(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\chi _s(T)$$\\end{document}), which instead are underestimated in the simple QPM. Furthermore, evaluating the spatial diffusion coefficient 2πTDs(T)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$2\\pi T D_s(T)$$\\end{document} in the QPMp\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$_p$$\\end{document}, we find it is also closer than QPM to the recent lQCD data performed including dynamical fermions. Finally, in a 1+1D expanding system, we evaluate the RAA(pT)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$R_{AA}(p_T)$$\\end{document} in the QPM and QPMp\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$_p$$\\end{document}, finding a significant reduction at low momenta for QPMp\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$_p$$\\end{document} which could lead in a realistic scenario to a better agreement to experimental data.

Masked particle modeling on sets: towards self-supervised high energy physics foundation models

Abstract We propose masked particle modeling (MPM) as a self-supervised method for learning generic, transferable, and reusable representations on unordered sets of inputs for use in high energy physics (HEP) scientific data. This work provides a novel scheme to perform masked modeling based pre-training to learn permutation invariant functions on sets. More generally, this work provides a step towards building large foundation models for HEP that can be generically pre-trained with self-supervised learning and later fine-tuned for a variety of down-stream tasks. In MPM, particles in a set are masked and the training objective is to recover their identity, as defined by a discretized token representation of a pre-trained vector quantized variational autoencoder. We study the efficacy of the method in samples of high energy jets at collider physics experiments, including studies on the impact of discretization, permutation invariance, and ordering. We also study the fine-tuning capability of the model, showing that it can be adapted to tasks such as supervised and weakly supervised jet classification, and that the model can transfer efficiently with small fine-tuning data sets to new classes and new data domains.

A Review of Discrete Element Method Applications in Soil–Plant Interactions: Challenges and Opportunities

The discrete-element method (DEM) has become a pivotal tool for investigating soil–plant interactions in agricultural and environmental engineering. This review examines recent advancements in DEM applications, focusing on both the challenges and opportunities that shape future research in this field. This paper first explores the effectiveness of DEM in simulating soil and plant materials, including seeds, roots, and residues, highlighting its role in understanding interactions that affect agricultural practices. Challenges such as long computation times and the complexity of determining accurate contact parameters are discussed, alongside emerging methods like machine learning that offer potential solutions. Notable advancements include the integration of machine learning algorithms for contact parameter estimation, the use of expanded particle models for dynamic processes, and the development of new techniques for detailed post-processing of DEM simulations. The review also identifies key future research directions, including the incorporation of environmental factors such as air and water, and the exploration of residue management for carbon storage and erosion prevention. By addressing these challenges and seizing these opportunities, future research can enhance the accuracy and applicability of DEM models, advancing our understanding of soil–plant interactions and contributing to more sustainable agricultural and environmental practices.