Coarse-grained Method Research Articles

Mechanical Properties Simulation of Crosslinked Elastomers by Coarse-grained Molecular Dynamics Method

Mechanical Properties Simulation of Crosslinked Elastomers by Coarse-grained Molecular Dynamics Method

Multi-scale molecular simulation of random peptide phase separation and its extended-to-compact structure transition driven by hydrophobic interactions.

Intrinsically disordered proteins (IDPs) often undergo liquid-liquid phase separation (LLPS) and form membraneless organelles or protein condensates. One of the core problems is how do electrostatic repulsion and hydrophobic interactions in peptides regulate the phase separation process? To answer this question, this study uses random peptides composed of positively charged arginine (Arg, R) and hydrophobic isoleucine (Ile, I) as the model systems, and conduct large-scale simulations using all atom and coarse-grained model multi-scale simulation methods. In this article, we investigate the phase separation of different sequences using a coarse-grained model. It is found that the stronger the electrostatic repulsion in the system, the more extended the single-chain structure, and the more likely the system forms a low-density homogeneous phase. In contrast, the stronger the hydrophobic effect of the system, the more compact the single-chain structure, the easier phase separation, and the higher the critical temperature of phase separation. Overall, by taking the random polypeptides composed of two types of amino acid residues as model systems, this study discusses the relationship between the protein sequence and phase behaviour, and provides theoretical insights into the interactions within or between proteins. It is expected to provide essential physical information for the sequence design of functional IDPs, as well as data to support the diagnosis and treatment of the LLPS-associated diseases.

Intelligent dissipative particle dynamics: Bridging mesoscopic models from microscopic simulations via deep neural networks

We develop a bottom-up coarse graining framework to construct mesoscopic force fields directly from microscopic dynamics by a machine learning technique. This scheme, called the intelligent dissipative particle dynamics (IDPD), holds various consistencies, including configuration, phase, static and dynamic consistencies that are incorporated into a deep neural network. The network is trained with full atomic data in a special way for preserving the natural symmetries of the system, and generates both the conservative and non-conservative force fields in the mesoscopic system. Subsequently, the force fields are used in the dissipative particle dynamics (DPD) that is considered as the effective mesoscopic dynamics resulting from the Mori–Zwanzig projection of the underlying atomic dynamics. To show the validity and applicability of IDPD, we coarse grain star polymer and methane fluids respectively, from full molecular dynamics (MD) simulations. Quantitative comparisons between the MD and IDPD indicate that the IDPD is able to reproduce both the static and dynamic properties of underlying MD system, with the aids of a pressure correction and a diffusion rescaling, even when coarse graining multiple molecules into a particle. It follows that the IDPD is competent for preserving the dynamic properties and coarse graining non-bonded molecules, which are common challenges for most of coarse graining methods.

Dissipative particle dynamics simulations for biological systems: From protein structures to cell mechanics

<p indent="0mm">Dissipative particle dynamics (DPD) is a mesoscopic coarse-grained simulation method developed in recent years, which is an important method for studying the dynamic behaviors of soft matter and complex fluids. In this method, each DPD particle represents a coarse-grained virtual cluster of a set of atoms or groups of matter. The position and momentum of the DPD particle are updated in a continuous phase but spaced at discrete time steps. The coarse-grained DPD particles are subject to simplified pairwise interacting conservative, dissipative, and random forces. Especially, the dissipative force and random force in the DPD method act as a heat sink and a source, respectively, and the combined effect of these two forces act as a thermostat, which conserves momentum and thus provides the correct description of hydrodynamic interactions for the model system. In addition, a common choice of the soft repulsion for the conservative force allows using larger integration time steps in DPD simulations than that usually allowed in classical molecular dynamics (MD) simulations. Hence, compared with the MD method, the computational cost of the DPD simulation is significantly reduced due to the smaller number of modeled particles and the larger computational time step, enabling the simulations of the static and dynamic behaviors of complex fluids and soft matter systems at physically attractive length scale and time scale. Moreover, the particle-based framework of the DPD method enables people to easily incorporate additional physical features into the model systems and extend its application to complex systems. For these reasons, the DPD method and its extension have been successfully applied to numerous soft matter and complex fluid systems such as oil/water/surfactant systems, chemical morphology, microscopic morphology, phase separation, as well as dynamics and rheological properties of polymer solutions and colloidal suspensions. In this paper, we first introduce the theoretical formulation and parameterization of the DPD method. Then, we review recent advances in DPD modeling of biological systems, focusing on its applications at the molecular and cellular scales. At the molecular scale, we highlight examples of successful simulations of the protein structures and their interactions, the structure and dynamics of amphiphilic lipid molecule membranes (e.g., the self-assembly of lipid molecules, the structure and properties of lipid membranes, the fusion of lipid membranes, and the budding and fission of lipid membranes), the interaction of lipid membranes with protein molecules, and the interaction of nanoparticles with lipid membranes; at the cellular scale, we focus on the DPD modeling of blood cell flow and blood rheological behaviors in the blood microcirculatory systems, including the shape deformation and flow dynamics of red blood cells, the margination and adhesion dynamics of white blood cells, the margination and aggregation behaviors of platelets, the hemorheological behavior of blood flow, as well as the separation of circulating tumor cells from blood flow using microfluidic devices. Additionally, we compare the advantages and disadvantages of the microscale blood flow simulations between the continuum-based methods and particle-based methods, including the DPD method. Finally, we briefly present the development trends and application prospects of DPD modeling in biological systems.

An improved CFD-DEM modelling of raceway dynamics and coke combustion in an industrial-scale blast furnace

The raceway dynamics and combustion by lateral injection of high-temperature gas into a packed bed of solid fuel are widely practised in various industries such as ironmaking blast furnace (BF), yet the particle-scale simulation and understanding were hindered by low resolution and huge computational cost. In this work, an advanced reactive computational fluid dynamics-discrete element method (rCFD-DEM) featuring two state-of-the-art techniques - coarse-grained method and smoothing method is developed to describe raceway dynamics and coke combustion in a three-dimensional (3D) industrial-scale BF. The model is validated against the measurements in the open literature. The simulation results show that the smoothing method can more reliably capture raceway morphology and the coarse-grained treatment reduces 78.14 % computational costs under the given conditions. The typical phenomena are illustrated in terms of raceway shape and size, temperature field and gas species distributions. Then, the effect of the key operating parameters is quantified in terms of raceway size, gas temperature and species distributions. The comparisons indicate increasing the blast rate from 2.78 Nm3/s to 3.05 Nm3/s and oxygen mass fraction from 0.17 to 0.27, the raceway volume expands by 61.11 % and 47.62 %, respectively; moreover, lower blast rate, higher bed temperature and higher oxygen mass fraction lead to higher temperature distribution and higher CO concentration. The present work provides a cost-effective tool for understanding and optimizing raceway operation in practical BFs.

Learning Dual-view User Representations for Enhanced Sequential Recommendation

Sequential recommendation (SR) aims to predict a user’s next interacted item given his/her historical interactions. Most existing sequential recommendation systems model user preferences only with item-level representations, where a user’s interaction sequence are often modeled with sequential or graph-based method to infer the user’s sequential interaction pattern. However, since a user’s preference factors may vary over time, the user modeling on item-level could hardly represent the user’s preference precisely and sufficiently, resulting in suboptimal recommendation performance. In addition, the recommendation results based on the item-level user representations lack the interpretability of preference factors. To address these problems, we propose a novel SR model with dual-view user representations in this paper, namely DUVRec, where a user’s preference is learned based on the representations of two distinct views, i.e., item view and factor view . Specifically, the item-view user representation is learned as the previous SR models to encode the user preference of item level, while the factor-view user representation is learned by an coarse-grained graph embedding method to explicitly represent the user in terms of preference factors. As a result, such dual-view user representations are more comprehensive than that in the previous SR models, leading to enhanced SR performance. Furthermore, we design a contrastive learning strategy to achieve mutual complementation between these two views. Our extensive experiments upon three benchmark datasets justify DUVRec’s superior performance over the state-of-the-art SR models, including the advantage of the dual-view contrastive learning. In addition, DUVRec’s capability of providing explanations on recommendation results is also demonstrated through some specific case studies.

Coarse-grained CFD-DEM simulation of coal and biomass co-gasification process in a fluidized bed reactor: Effects of particle size distribution and operating pressure

The co-gasification of coal and biomass through fluidized bed gasifier is a promising technology for the improvement of the utilization efficiency of fossil fuels and the development of renewable energy. In this work, a coarse-grained computational fluid dynamics-discrete element method (CFD-DEM) simulation is performed to investigate the co-gasification process in a bubbling fluidized bed gasifier. The effects of particle size distribution and operating pressure on the hydrodynamics, heat transfer, and co-gasification performance are analyzed. The results show that the increase of sand particle size distribution (PSD) width slightly affects the particle concentration distribution, particle residence time distribution, and co-gasification performance. The heat transfer rate of convection is an order of magnitude higher than those of conduction and radiation. At higher PSD width, the formation of the vertical solid temperature gradient at the bed bottom is attributed to the deteriorated solid mixing. As the operating pressure increases, the bed temperature decreases apparently. In addition, the variation of the convective heat transfer rate dominates the heat transfer efficiency as the operating pressure changes. The application of elevated operating pressure benefits to the thermal-chemical conversion of blend fuel since the char conversion reaction is accelerated.

Coupling between mechanical stresses and lithium penetration in a lithium ion battery

Silicon-based anodes are an attractive choice for the future of energy storage. However, mechanical stresses generated during lithiation induce a rapid degradation of the anodes. The coupling between mechanical and (electro)chemical phenomena during lithiation is investigated using Molecular Dynamics simulations. First, the diffusion regime of lithium atoms in the silicon is found to depend drastically on the crystalline orientations and on the applied pressure. Above a threshold pressure, a ballistic motion of lithium along certain crystallographic planes is observed. Second, the local complex stress field generated by the lithiation has been computed using a coarse-graining method including thermal effects. It appears to be strongly dependent on the density of lithium in the lithiated layer, on the temperature, and on the crystalline orientation of the free surface. The resulting very high values of the locally induced pressures can be related to the progressive amorphization of the silicon support.

Bearing Fault Diagnosis Method Based on RCMFDE-SPLR and Ocean Predator Algorithm Optimizing Support Vector Machine.

For the problem that rolling bearing fault characteristics are difficult to extract accurately and the fault diagnosis accuracy is not high, an unsupervised characteristic selection method of refined composite multiscale fluctuation-based dispersion entropy (RCMFDE) combined with self-paced learning and low-redundant regularization (SPLR) is proposed, for which the fault diagnosis is carried out by support vector machine (SVM) optimized by the marine predator algorithm (MPA). First, we extract the entropy characteristics of the bearings under different fault states by RCMFDE and the introduction of the fine composite multiscale coarse-grained method and fluctuation strategy improves the stability and estimation accuracy of the bearing characteristics; then, a novel dimensionality-reduction method, SPLR, is used to select better entropy characteristics, and the local flow structure of the fault characteristics is preserved and the redundancy is constrained by two regularization terms; finally, using the MPA-optimized SVM classifier by combining Levy motion and Eddy motion strategies, the preferred RCMFDE is fed into the MPA-SVM model for fault diagnosis, for which the obtained bearing fault diagnosis accuracy is 97.67%. The results show that the RCMFDE can effectively improve the stability and accuracy of the bearing characteristics, the SPLR-based low-dimensional characteristics can suppress the redundancy characteristics and improve the effectiveness of the characteristics, and the MPA-based adaptive SVM model solves the parameter randomness problem and, therefore, the proposed method has outstanding superiority.

Glass transition of random heteropolymers: A molecular dynamics simulation study in melt, in water, and in vacuum

Random heteropolymers (RHPs) have a long history of being studied as toy models for protein folding. Previous computational studies applied statistical mechanics or coarse-grained methods to investigate the RHP phase transition; here, we use atomistic molecular dynamics (MD) simulations to explore the factors affecting the glass transition temperature (Tg) and the mobility of a complex class of RHPs. Our RHPs consist of four methyl methacrylate-based monomers with side chains mimicking different classes of amino acids. We explore the dynamical features in the RHP melts with random sequence mixtures and show that instead of a sharp pseudo-second-order glass transition, as for PMMA, RHPs with low ionic contents show a soft and gradual transition and those with increased content of ionic monomers show a sharp transition. Meanwhile, the configurational entropy of the backbone dihedral angle decreases suddenly in both PMMA and RHPs as a common signature of the glass transition. The dynamical spatial heterogeneity during the glass transition is related to both the chemical structure of the monomers and their positions along the chain. Moreover, chain mobility depends on the content of ionic monomers and types of counterions. The comparison with the single-chain nanoparticles (SCNPs) in water and in vacuum shows “breathing dynamics” with higher mobility. Our work characterizes the physicochemical properties of the RHP melts as a group of emerging bio-inspired polymer materials.

Understanding and Modeling Polymers: The Challenge of Multiple Scales

Polymer materials are multiscale systems by definition. Already the description of a single macromolecule involves a multitude of scales, and cooperative processes in polymer assemblies are governed by their interplay. Polymers have been among the first materials for which systematic multiscale techniques were developed, yet they continue to present extraordinary challenges for modellers. In this Perspective, we review popular models that are used to describe polymers on different scales and discuss scale-bridging strategies such as static and dynamic coarse-graining methods and multiresolution approaches. We close with a list of hard problems which still need to be solved in order to gain a comprehensive quantitative understanding of polymer systems.

Smoothed particle hydrodynamics physically reconsidered: The relation to explicit large eddy simulation and the issue of particle duality

In this work, we will identify a novel relation between Smoothed Particle Hydrodynamics (SPH) and explicit large eddy simulation using a coarse-graining method from non-equilibrium molecular dynamics. While the current literature points at the conclusion that characteristic SPH issues become restrictive for subsonic turbulent flows, we see the potential to mitigate these SPH issues by explicit subfilter stress modeling. We verify our theory by various simulations of homogeneous, isotropic turbulence at Re=104 and compare the results to a direct numerical simulation [T. Dairay et al., “Numerical dissipation vs subgrid-scale modelling for large eddy simulation,” J. Comput. Phys. 337, 252–274 (2017)]. Although the simulations substantiate our theory, we see another issue arising, which is conceptually rooted in the particle itself, termed as particle duality. Finally, we conclude our work by acknowledging SPH as a coarse-graining method for turbulent flows, highlighting its capabilities and limitations.

Road Friction Coefficient Estimation Via Weakly Supervised Semantic Segmentation and Uncertainty Estimation

Vision-based road friction coefficient estimation received extensive attention in the field of road maintenance and autonomous driving. However, the current mainstream coarse-grained friction estimation methods are basically based on image classification tasks. This makes it difficult to deal with complex road conditions in changing weathers. Many models can correctly predict the friction coefficients of the road as a whole in consistent and simple road conditions, but perform poorly otherwise. The existing image benchmarks in this field rarely consider the above problems as well, which limits the comparable evaluations of different models. Therefore, in this paper, we first construct a challenging pixel-level friction coefficient estimation dataset WRF-P to evaluate model performances under mixed road conditions. Then, we propose a friction coefficient estimation method based on weakly supervised learning and uncertainty estimation to realize pixel-level road friction prediction with low annotation cost. The model outperforms existing weakly supervised methods and reaches 39.63% mIOU on the WRF-P dataset. The WRF-P dataset will be made publicly available at https://github.com/blackholeLFL/The-WRF-dataset soon.

Analysis of Gas Nanoclusters in Water Using All-Atom Molecular Dynamics.

The Young-Laplace equation suggests that nanosized gas clusters would dissolve under the effects of perturbation. The fact that nanobubbles are observed raises questions as to the mechanism underlying their stability. In the current study, we used all-atom molecular dynamics simulations to investigate the gas-water interfacial properties of gas clusters. We employed the instantaneous coarse-graining method to define the fluctuating boundaries and analyze the deformation of gas clusters. Fourier transform analysis of the cluster morphology revealed that the radius and morphology deformation variations exhibit power law relationships with the vibrational frequency, indicating that the surface energy dissipated through morphology variations. Increasing pressure in the liquid region was found to alter the network of water molecules at the interface, whereas increasing pressure in the gas region did not exhibit this effect. The overall gas concentration was oversaturated and proportional to the gas density inside the clusters. However, the result of comparison with Henry's law reveals that the gas pressure at the interface reduced by the interfacial effects is much lower than that inside the gas region, thus reducing the demanding degree of oversaturation. Originating from the interfacial charge allocation, the magnitude of the electrostatic stress is greater than that of the gas pressure inside the cluster. However, the magnitude of the reversed tension induced by electrostatic stress is far below the value of interfacial tension. The potential of mean force (PMF) profiles revealed that a barrier potential at the interface hindered gas particles from escaping the cluster. Several effects contribute to stabilizing the gas clusters in water, including high-frequency morphological deformation, electrostatic stress, reduced interfacial tension, and gas oversaturation conditions. Our results suggest that gas clusters can exist in water under gas oversaturation conditions in the absence of hydrophobic contaminants or pinning charges at interfaces.

Fault Diagnosis Method Based on AUPLMD and RTSMWPE for a Reciprocating Compressor Valve.

In order to effectively extract the key feature information hidden in the original vibration signal, this paper proposes a fault feature extraction method combining adaptive uniform phase local mean decomposition (AUPLMD) and refined time-shift multiscale weighted permutation entropy (RTSMWPE). The proposed method focuses on two aspects: solving the serious modal aliasing problem of local mean decomposition (LMD) and the dependence of permutation entropy on the length of the original time series. First, by adding a sine wave with a uniform phase as a masking signal, adaptively selecting the amplitude of the added sine wave, the optimal decomposition result is screened by the orthogonality and the signal is reconstructed based on the kurtosis value to remove the signal noise. Secondly, in the RTSMWPE method, the fault feature extraction is realized by considering the signal amplitude information and replacing the traditional coarse-grained multi-scale method with a time-shifted multi-scale method. Finally, the proposed method is applied to the analysis of the experimental data of the reciprocating compressor valve; the analysis results demonstrate the effectiveness of the proposed method.

Fault Diagnosis of Hydroelectric Units Based on a Novel Multiscale Fractional-Order Weighted Permutation Entropy

To improve the noise immunity, stability and sensitivity to different signal types in the hydroelectric unit fault diagnosis model, a hydroelectric unit fault diagnosis model based on improved multiscale fractional-order weighted permutation entropy (IMFWPE) is proposed. Firstly, the fractional order and weighting theory is introduced into the permutation entropy (PE) to improve the sensitivity to different fault signals while improving the defect of ignoring the signal amplitude information. Additionally, considering the problem that a single scale cannot fully reflect the timing characteristics and that the traditional coarse-grained method will shorten the timing length, a new tool for measuring the complexity of timing signals, IMFWPE, is proposed by introducing an improved multiscale method. Finally, the IMFWPE values of signals are extracted as features and input to the classifier for fault identification of hydroelectric units. The experimental results show that the proposed method has the best diagnostic effect when compared with other methods, has good noise immunity and stability, and has good diagnostic capability in the actual unit environment.

Fast, Faithful Simulations of Donor-Acceptor Interface Morphology.

The local structure of conjugated polymers governs key optoelectronic properties, such as charge conduction and photogeneration at donor-acceptor interfaces. Because conjugated polymers are large, stiff, and relax slowly, all-atom molecular dynamics simulations are computationally expensive. Here, we describe a coarse-graining method that exploits the stiffness of constituent aromatic moieties by representing each moiety as rigidly bonded clusters of atoms wherein virtual sites replace several atoms. This approach significantly reduces the degrees of freedom while faithfully representing the shape and interactions of the moieties, resulting in 10 times faster simulations than all-atom simulations. Simulation of a donor polymer (P3HT) and a non-fullerene acceptor (O-IDTBR) validates the coarse-graining method by comparing structural properties from experiments, such as the density and persistence length. The fast simulation produces equilibrated systems with realistic morphologies. The simulation results of an equimolar mixture of P3HT, with a molecular weight of 1332 g mol-1, and an O-IDTBR mixture suggest that the interface width must be larger than 7 nm. Also, we investigate the effect of slow cooling on morphologies, particularly the number of close contacts that facilitates carrier transport. Slow cooling increases close contacts, and the effect is more pronounced in crystal-forming P3HT than in O-IDTBR, where bulky side-groups hinder crystal formation.

Application of coarse-graining for large scale simulation of fluid and particle motion in spiral jet mill by CFD-DEM

Spiral jet milling is commonly used for size reduction of high value particulate solids, such as pharmaceutical ingredients, for which low contamination is critical. The mill utilises high-pressure gas nozzles to form an internal vortex and induce particle breakage through inter-particle collisions. Due to the centrifugal flow field, the particles can move radially inwards towards the mill centre and escape the mill only when their size is sufficiently reduced for the fluid drag to exceed the centrifugal force. Large particles move radially towards the outer wall and form a dense particle bed. The bed itself circulates in the milling chamber due to the induced effect of the inclined gas jets. In this study, we analyse the implementation of a coarse-graining (CG) approach on a coupled Computational Fluid Dynamic-Discrete Element Method simulation. Along with the actual particle size, four CG scale cases are compared (CG-2, CG-4, CG-8 and CG-10). To analyse the success of the approach in predicting the dynamics of fluid and particle motion, the characteristic features of the particle bed and fluid field, i.e. the fluid and particle velocity distributions and dissipated energy through particle collisions are analysed. There is good agreement between the original particle size and the two smallest scaled cases (CG-2, CG-4) for the above characteristics. However, modelling the lean phase is less successful, as there are fewer particles that reside there at any given time. There is also good agreement between these three cases in terms of dissipated energy through particle collisions. An average value for dissipated energy of 0.6 mJ is recorded for each of the three lowest cases. This proves beneficial for simulation time required, as the particle number is reduced 2-fold and 4-fold, respectively, and interparticle collision rate is reduced 60% each time. Application of the Coarse Grain method in a spiral jet mill using coupled Computational Fluid Dynamics and Discrete Element Method Modelling. Top: Original particle size system and four scaled systems. Bottom: Resulting temporal average dissipated energy through interparticle collisions and particle wall collisions. • Coarse graining method is applied in the spiral jet mill using a CFD-DEM framework. • Particle and fluid motion is similar in the bed at smaller scales. • Insufficient particles were present in the lean phase to model the original system. • Temporal dissipated energy remained constant for low coarse grain numbers.

Nonlinear coarse-graining models for 3D printed multi-material biomimetic composites

Bio-inspired composites are a great promise for mimicking the extraordinary and highly efficient properties of natural materials. Recent developments in voxel-by-voxel 3D printing have enabled extreme levels of control over the material deposition, yielding complex micro-architected materials. However, spatial complexity makes it a formidable challenge to find the optimal distribution of both hard and soft phases. To address this, a nonlinear coarse-graining approach is developed, where foam-based constitutive equations are used to predict the mechanics of biomimetic composites. The proposed approach is validated by comparing coarse-grained finite element predictions against full-field strain distributions measured using digital image correlation. To evaluate the degree of coarse-graining on model accuracy, pre-notched specimens decorated with a binarized version of a renowned painting were modeled. Subsequently, coarse-graining is used to predict the fracture behavior of bio-inspired composites incorporating complex designs, such as functional gradients and hierarchical organizations. Finally, as a showcase of the proposed approach, the inverse coarse-graining is combined with a theoretical model of bone tissue adaptation to optimize the microarchitecture of a 3D-printed femur. The predicted properties were in exceptionally good agreement with the corresponding experimental results. Therefore, the coarse-graining method allows the design of advanced architected materials with tunable and predictable properties.

Molecular Dynamics Simulations of Polymer Grouting Material: Its Mechanical Behavior under Uniaxial Tension, Cyclic Tensile Loading, and Stress Relaxation

At present, polyurethane (PU) has been extensively used as a grouting material in civil engineering. The mechanical properties of PU are the key to achieving the desirable grouting effect. This study presents the research results of the mechanical behavior of PU matrix under tensile, successive cyclic tensile, and stress relaxation at the nanoscale, using the coarse-grained molecular dynamics simulation method. The influences of the number of molecule chains and strain rate on the tensile mechanical properties are discussed, and the tensile deformation mechanism of PU matrix is revealed. The tensile strength of PU matrix is independent of loading path, and after yielding, the strain of PU matrix contains the elastic strain, plastic strain, and viscous strain. In the stress relaxation process, the evolution of the axial stress is mainly caused by the varied van der Waals interactions. The stress relaxation behavior of PU matrix can be described by the viscoelastic model consisting of one elastic element in parallel with one Maxwell element.