检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Feng, Xin Jiang, Yi Qin, Jia-Xian Zhang, Lai-Ping Deng, Xiao-Gang College of Computer Science Sichuan University Chengdu610065 China Tianfu Engineering-oriented Numerical Simulation & Software Innovation Center Sichuan University Chengdu610000 China PLA Beijing100082 China Institute of Defense Science and Technology Innovation Academy of Military Sciences Beijing100071 China

The wave equation is an important physical partial differential equation, and in recent years, deep learning has shown promise in accelerating or replacing traditional numerical methods for solving it. However, existing deep learning methods suffer from high data acquisition costs, low training efficiency, and insufficient generalization capability for boundary conditions. To address these issues, this paper proposes an unsupervised learning method for the wave equation based on finite difference residual constraints. We construct a novel finite difference residual constraint based on structured grids and finite difference methods, as well as an unsupervised training strategy, enabling convolutional neural networks to train without data and predict the forward propagation process of waves. Experimental results show that finite difference residual constraints have advantages over physics-informed neural networks (PINNs) type physical information constraints, such as easier fitting, lower computational costs, and stronger source term generalization capability, making our method more efficient in training and potent in application. Copyright © 2024, The Authors. All rights reserved.

关键词： Finite difference method

来源：评论

学校读者我要写书评

暂无评论

Revealing the Local Microstates of Fe–Mn–Al Medium Entropy Alloy:A Comprehensive First-principles Study

引用

Acta Metallurgica Sinica(English Letters) 2021年第11期34卷 1492-1502页

作者： Ying Zhang William Yi Wang Chengxiong Zou Rui Bai Yidong Wu Deye lin Jun Wang Xidong Hui Xiubing Liang Jinshan Li State Key Laboratory of Solidification Processing Northwestern Polytechnical UniversityXi'an 710072China State Key Laboratory for Advanced Metals and Materials University of Science and Technology BeijingBeijing 100083China CAEP Software Center for High Performance Numerical Simulation Beijing 100088China Defense Innovation Institute Academy of Military Sciences of the PLA of ChinaBeijing 100071China

Entropy-stabilized multi-component alloys have been considered to be prospective structural materials attributing to their impressive mechanical and functional *** local chemical complexions,microstates and configurational transformations are essential to reveal the structure–property relationship,thus,to promote the development of advanced multicomponent *** the present work,effects of local lattice distortion(LLD)and microstates of various configurations on the equilibrium volume(V0),total energy,Fermi energy,magnetic moment(μMag)and electron work function(Φ)and bonding structures of the Fe–Mn–Al medium entropy alloy(MEA)have been investigated comprehensively by first-principles *** is found that theΦandμMag of those MEA are proportional to the V 0,which is dominated by lattice *** terms of bonding charge density,both the strengthened clusters or the so-called short-range order structures and the weakly bonded spots or weak spots are *** the presence of weakly bonded Al atoms implies a large LLD/mismatch,the Fe–Mn bonding pairs result in the formation of strengthened clusters,which dominate the local microstates and the configurational *** variations ofμMag are associated with the enhancement of the nearest neighbor magnetic Fe and Mn atoms,attributing to the LLD caused by Al atoms,the local changes in the electronic *** work provides an atomic and electronic insight into the microstate-dominated solid-solution strengthening mechanism of Fe–Mn–Al MEA.

关键词： Medium entropy alloy Bonding charge density Magnetic moment Microstates

来源：评论

学校读者我要写书评

暂无评论

Field-circuit simulation based on parallel FDTD-SPICE method 4

Field-circuit simulation based on parallel FDTD-SPICE method

引用

4th International Applied Computational Electromagnetics Society Symposium in China, ACES-China 2021

作者： Yue, Chen Shun Wu, Yu Mao Wang, Wei Jie Zhou, Hai Jin Fudan University Key Laboratory for Information Sciences of Electromagnetic Waves Shanghai China CAEP Software Center for High Performance Numerical Simulation Beijing China Institute of Applied Physics and Computation Mathematics Beijing China

ISBN: (纸本)9781733509619

In this work, we study the simulation effect of FDTD Method in microwave circuit. The microwave circuit is divided multiple Yee cells, then we can get the relevant parameters through parallel computation and sent they into SPICE for simulation. The numerical example shows that the proposed method has a higher speed than the common SPICE simulation. © 2021 Applied Computational Electromagnetics Society (ACES).

关键词： SPICE

来源：评论

学校读者我要写书评

暂无评论

Study on Extraction of Capacitance Parameters of Three Dimensional Interconnect Structures Based on Finite Element Method 4

Study on Extraction of Capacitance Parameters of Three Dimen...

引用

4th International Applied Computational Electromagnetics Society Symposium in China, ACES-China 2021

作者： Jin, Han Qi Wu, Yu Mao Wang, Wei Jie Zhou, Hai Jing Fudan University Key Laboratory for Information Sciences of Electromagnetic Waves Shanghai China CAEP Software Center for High Performance Numerical Simulation Beijing China Institute of Applied Physics and Computation Mathematics Beijing China

ISBN: (纸本)9781733509619

In recent years, the rapid development of industry makes the proportion of interconnect in chip increase rapidly. In order to ensure the stability and reliability of the circuit, accurate 3D modeling and high efficiency parasitic parameter extraction are needed for the chip interconnect. This paper has carried out the research on the extraction of capacitance parameters of the 3D finite element interconnection line. Starting from the principle of multi-conductor capacitance calculation, the problem is reduced to solving the Laplace equation for point boundary conditions. By using the finite element method (FEM), the solving process is analyzed and deduced. In addition, based on the high-efficiency parallel support software framework JAUMIN, this paper realizes the parallel calculation of the three-dimensional finite element capacitance parameter extraction, which greatly improves the calculation efficiency. Finally, the algorithm is applied to a representative interconnect structure. The calculation results are compared with the multi-physics commercial simulation software COMSOL. The results show that the potential distribution diagram is consistent. © 2021 Applied Computational Electromagnetics Society (ACES).

关键词： Finite element method

来源：评论

学校读者我要写书评

暂无评论

FEM-PIKFNNs for underwater acoustic propagation induced by structural vibrations in different ocean environments

arXiv

引用

arXiv 2023年

作者： Xi, Qiang Fu, Zhuojia Xu, Wenzhi Xue, Mi-An Zheng, Jinhai Key Laboratory of Ministry of Education for Coastal Disaster and Protection Hohai University Nanjing210098 China College of Harbour Coastal and Offshore Engineering Hohai University Nanjing210098 China Center for Numerical Simulation Software in Engineering and Sciences College of Mechanics and Materials Hohai University Nanjing211100 China

In this paper, a novel hybrid method based on the finite element method (FEM) and physics-informed kernel function neural networks (PIKFNNs) is proposed and applied to the prediction of underwater acoustic propagation induced by structural vibrations in the unbounded ocean, deep ocean and shallow ocean. In the hybrid method, PIKFNNs are a class of improved shallow physics-informed neural networks (PINNs) that replace the activation functions in PINNs with the physics-informed kernel functions (PIKFs), thereby integrating prior physical information into the neural network model. Moreover, this neural network circumvents the step of embedding the governing equations into the loss function in PINNs, and requires only training on boundary data. By using the Green's functions as the PIKFs and the structural-acoustic coupling response information obtained from the FEM as boundary training data, the PIKFNNs can inherently capture the Sommerfeld radiation condition at infinity, which is naturally suitable for predicting ocean acoustic propagation. numerical experiments demonstrate the accuracy and feasibility of the FEM-PIKFNNs in comparison with the true solutions and finite element results. © 2023, CC BY-NC-ND.

关键词： Finite element method

来源：评论

学校读者我要写书评

暂无评论

Reaction–diffusion equation driven topology optimization of high-resolution and feature-rich structures using unstructured meshes

引用

Advances in engineering software 2023年第1期180卷

作者： Li, Hao Yu, Minghao Jolivet, Pierre Alexandersen, Joe Kondoh, Tsuguo Hu, Tiannan Furuta, Kozo Izui, Kazuhiro Nishiwaki, Shinji Department of Mechanical Engineering and Science Kyoto University Kyoto615-8540 Japan China Academy of Engineering Physics Software Center for High Performance Numerical Simulation Beijing100088 China LIP6 Sorbonne Université Paris France Department of Mechanical and Electrical Engineering University of Southern Denmark Odense Denmark

Motivated by the need for porous structures in the design of biodegradable implants, as well as for the diverse and competitive designs in architecture, this paper builds upon the recent advances in the single-scale (macroscopic) topology optimization (TO) approach and proposes a level-set method (LSM) for the design of lattice structures. The key idea is to introduce the maximum length-scale constraint, realized by a PDE filter, into the reaction–diffusion equation (RDE)-based LSM which can end up with feature-rich shape, starting from scratch. Then, the proposed TO technique is coupled with distributed computing algorithms based on PETSc, interfaced in the open-source software FreeFEM using unstructured meshes, to deliver both two- and three-dimensional designs with complex-shape geometry. The parallel efficiency is validated by solving large-scale 3D benchmarks with 20 million tetrahedral elements on a cloud-based cluster. Furthermore, we integrate local mesh refinement (h-adaptation scheme) into this workflow, allowing one to reach high-resolution and feature-rich 3D designs on a desktop, paving the way to future investigations, aimed at including the maximum length-scale constraint in large-scale multiphysics topology optimization at a reasonable cost. © 2023 Elsevier Ltd

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

Grider-Ho: Rapid and Efficient Parallel software for High-Order Curvilinear Mesh Generation

SSRN

引用

SSRN 2024年

作者： Liu, Xiangyu Wang, He Zhao, Zhong Wang, Huadong Guan, Zhidong Wang, Nianhua School of Aeronautic Science and Engineering Beihang University Beijing100191 China Tianfu Engineering-oriented Numerical Simulation & Software Innovation Center Sichuan University Chengdu610207 China State Key Laboratory of Aerodynamics China Aerodynamics Research and Development Center Mianyang621000 China Computational Aerodynamics Institute China Aerodynamics Research and Development Center Mianyang621000 China

The advancement in high-order computational methods is reshaping the landscape of mesh generation in Computational Fluid Dynamics (CFD), steering the focus towards curvilinear mesh techniques to meet the escalating accuracy demands. Grider-HO, a software designed for generating high-order curvilinear mesh efficiently and rapidly, has been developed. It features a layered architecture and utilizes the concurrent hash table and the Alternating Digital Tree (ADT) data structure, supporting thread-level parallelism to convert the straight-edge mesh into the high-order curvilinear mesh seamlessly. Grider-HO initially inserts high-order nodes to a straight-edge mesh using a concurrent hash table, then projects them onto the geometry model with a parallel thread pool. Finally, it employs a localized RBF method with ADT for volume node interpolation, avoiding mesh entanglement. Validated through CFD simulations using the GPU-accelerated PyFR high-order solver, the practicality of Grider-HO is demonstrated across various Reynolds numbers in typical cases like the sphere, the cylinder, and the SD7003 airfoil cases. These results confirm that the high-order curvilinear mesh generated by Grider-HO can meet the high-order requirements of emerging computational methods. Moreover, Grider-HO exemplifies its effectiveness in generating large-scale, high-order curvilinear meshes for DLR-F6 cases. It elevates a 5 million mesh to P2 in 3min39sec at 68% parallel efficiency on 16 threads, and a 14 million mesh to P3 in 52min39sec at 60% parallel efficiency on 16 threads, which illustrates its efficiency and potential in satisfying the demands of complex geometries in engineering applications. © 2024, The Authors. All rights reserved.

关键词： Computational fluid dynamics

来源：评论

学校读者我要写书评

暂无评论

Large-Eddy simulation of Dual-Fuel Spray Ignition at Varying Levels of Methane Diluted Ambient Oxidizer Using FGM

SSRN

引用

SSRN 2023年

作者： Bao, Hesheng Han, Jinlin Zhang, Yan Matteo, Andrea Di Roekaerts, Dirk van Oijen, Jeroen Somers, Bart Department of Mechanical Engineering Eindhoven University of Technology Netherlands CAEP Software Center for High Performance Numerical Simulation China Institute of Applied Physics and Computational Mathematics China Department of Process & Energy Delft University of Technology Netherlands

In the current work, the Flamelet Generated Manifold (FGM) method is applied with large-eddy simulation (LES) to investigate the effect of methane on dualfuel (DF) spray ignition. The diesel surrogate n-dodecane is injected as the socalled pilot fuel into selected lean methane-air mixtures, ranging from CH4 = 0 to CH4 = 0.75, at engine relevant conditions. The operating conditions are those of the completely characterized Engine Combustion Network (ECN) Spray A configuration, for which the modeling approach adopted in the present study was extensively validated. The specific purpose of this study is to extend and validate the FGM approach for dual-fuel combustion. In order to understand the interplay of chemistry and mixing, the ignition behavior of selected cases is investigated. It is found that both low and high temperature combustion (LTC and HTC, respectively) are increasingly retarded by higher values of CH4, while the induction time between LTC and HTC is relatively insensitive compared to the ignition delay time (IDT). Analysis reveals a more prominent role of mixing for increased CH4. The development of LTC and HTC are quantitatively analyzed for different cases. The transition from LTC to HTC is found to be highly correlated with the evolution of lift-off length (LOL), which on its turn is seriously affected by CH4. The local flame behavior is analyzed via chemical explosive mode analysis (CEMA), suggesting a clear flame propagation due to diffusion towards lean mixtures after the ignition of the pilot fuel. Besides, it is found that diffusion helps to stabilize the flame in leaner mixtures, which is more important in DF combustion. The results show FGM to be a promising tool in modeling the DF sprays. © 2023, The Authors. All rights reserved.

关键词： Large eddy simulation

来源：评论

学校读者我要写书评

暂无评论

Localized collocation schemes and their applications

引用

Acta Mechanica Sinica 2022年第7期38卷 1-28,I0002页

作者： Zhuojia Fu Zhuochao Tang Qiang Xi Qingguo Liu Yan Gu Fajie Wang Key Laboratory of Ministry of Education for Coastal Disaster and Protection Hohai UniversityNanjing210098China Center for Numerical Simulation Software in Engineering and Sciences College of Mechanics and MaterialsHohai UniversityNanjing211100China Department of Fluid Dynamics and Thermodynamics Faculty of Mechanical EngineeringUniversity of LjubljanaLjubljana1000Slovenia Laboratory for Simulation of Materials and Processes Institute of Metals and TechnologyLjubljana1000Slovenia School of Mathematics and Statistics Qingdao UniversityQingdaoShandong266071China College of Mechanical and Electrical Engineering National Engineering Research Center for Intelligent Electrical Vehicle Power SystemQingdao UniversityQingdao266071China

This paper presents a summary of various localized collocation schemes and their engineering *** basic concepts of localized collocation methods(LCMs)are first introduced,such as approximation theory,semianalytical collocation methods and localization *** on these basic concepts,five different formulations of localized collocation methods are introduced,including the localized radial basis function collocation method(LRBFCM)and the generalized finite difference method(GFDM),the localized method of fundamental solutions(LMFS),the localized radial Trefftz collocation method(LRTCM),and the localized collocation Trefftz method(LCTM).Then,several additional schemes,such as the generalized reciprocity method,Laplace and Fourier transformations,and Krylov deferred correction,are introduced to enable the application of the LCM to large-scale engineering and scientific computing for solving inhomogeneous,nonisotropic and time-dependent partial differential *** typical benchmark examples are presented to show the recent developments and applications on the LCM solution of some selected boundary value problems,such as numerical wave flume,potential-based inverse electrocardiography,wave propagation analysis and 2D phononic crystals,elasticity and in-plane crack problems,heat conduction problems in heterogeneous material and nonlinear time-dependent Burgers’***,some conclusions and outlooks of the LCMs are summarized.

关键词： Localization Collocation method Semianalytical Computational mechanics

来源：评论

学校读者我要写书评

暂无评论

2d 3t Pinns for Solving 2-D 3-T Heat Conduction Equations Based on Physics-Informed Neural Networks

SSRN

引用

SSRN 2024年

作者： Lei, Xiaojun Gu, Tongxiang Xu, Xiaowen An, Hengbin Yao, Yanzhong Graduate School of Chinese Academy of Engineering Physics Beijing100193 China Laboratory of Computational Physics Institute of Applied Physics and Computational Mathematics Beijing100088 China CAEP Software Center for Numerical Simulation Beijing100088 China Institute of Applied Physics and Computational Mathematics China

The 2-D 3-T heat conduction equations can approximate the propagation of energy in the material, as well as the energy exchange process of electrons, ions and photons. In many cases, the computational time to solve these equations accounts for a large proportion (more than $80\%$) of that of the entire simulation of radiation hydrodynamics. PINNs is a promising way to solve partial differential equations (PDEs). Although numerical methods have been successful in solving 2-D 3-T heat conduction equations, the PINNs method also has some advantages, such as mesh-free, suitable to high dimension and complex domain problems. But the original PINNs cannot solve the 2-D 3-T heat conduction equations to a reasonable precision. This work wants to explore what techniques needs to be added to PINNs and to what extent it can solve the problems of strong nonlinearity and multi-scale existing in 2-D 3-T heat conduction equations. Hence, we adopt guaranteed positive constraint to the outputs so that the network can be trained, give a relatively large weight to the initial loss and Diliclet boundary loss, take the logarithm of the initial loss, use transfer learning and Fourier feature embedding to improve accuracy. We call our improved approach 2D 3T PINNs. numerical experiments show that the relative $L_numerical$ and the absolute $L_{\infty}$ error between the 2D 3T PINNs prediction and the reference solution is of the order of $10^{-2}-10^{-3}$. © 2024, The Authors. All rights reserved.

关键词： numerical methods

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：