A5059508758- Lattice Boltzmann Simulation Studies
- Fluid Dynamics Simulations and Interactions
- Computational Fluid Dynamics and Aerodynamics
- Fluid Dynamics and Turbulent Flows
- Model Reduction and Neural Networks
- Fluid Dynamics and Heat Transfer
- Wireless Sensor Networks and IoT
- Online Learning and Analytics
- Physical Activity and Education Research
- nanoparticles nucleation surface interactions
- Combustion and flame dynamics
- Nuclear Engineering Thermal-Hydraulics
- Parallel Computing and Optimization Techniques
- Teaching and Learning Programming
- Electronic and Structural Properties of Oxides
- Radio Wave Propagation Studies
- Cyclone Separators and Fluid Dynamics
- Fluid Dynamics and Mixing
- Heat Transfer and Boiling Studies
- Aerosol Filtration and Electrostatic Precipitation
- Plasmonic and Surface Plasmon Research
- Advanced Numerical Methods in Computational Mathematics
- Coal Properties and Utilization
- Heat transfer and supercritical fluids
- Environmental Engineering and Cultural Studies
National University of Defense Technology
2018-2025
Northwestern Polytechnical University
2019
Henan University of Science and Technology
2018
China University of Mining and Technology
2011-2018
SUNY Oneonta
2013-2015
Air Force Engineering University
2011
Physics-informed neural networks (PINNs) have emerged as a popular approach in scientific machine learning for solving both forward and inverse problems of partial differential equations (PDEs). However, complex physical systems are often characterized by parameters, such viscosity Reynolds number fluid dynamics, which pose significant challenges parameterized PDE solutions. The inherent limitations PINNs include the need repeated time-consuming training under varying parameter conditions,...
We investigated a mid-infrared (mid-IR) dual-band absorber consisting of continuous gold film coated with an asymmetric silicon grating. In each unit cell the grating, there are three unequally spaced strips. Numerical results reveal that (+1, -1) planar surface plasmon polariton (SPP) waves excited by transverse-magnetic (TM) incidence can be coupled different Fabry-Pérot (FP) resonances and resonant energy is dissipated to ohmic loss. Under normal condition, provides two high-absorbance...
To address the limited generalisation ability issue of physics-informed neural networks, we propose a multi viscosity networks (μ-PINNs), along with two tailored training strategies. By using μ-PINNs, train model once for specific scenario and obtain flow field data varying fluid (μ). validate conduct experiments on three 2D scenarios, comparing results to those computed by OpenFOAM providing their relative L2 error. The demonstrate that μ-PINNs possess capability capturing influence output...
Abstract Due to the intrinsic nature of multi‐physics, it is prohibitively complex design and implement a simulation software platform for study structural responses detonation shock. In this article, partitioned fluid‐structure interaction computing designed parallel simulating The wave propagation are modeled in an open‐source multi‐component solver based on OpenFOAM blastFoam, simulated through finite element library deal.II. To capture dynamics between fluid structure, both solvers...
The overset grid method is widely employed to solve moving boundary problems in numerical simulations. However, the heavy and inevitable communication resulting from movements severely impedes improvement of parallel efficiency. This paper proposes a Motion Trace Decomposition (MTD) alleviate this issue. MTD minimizes overhead between processors by decomposing sub-grids distributing them according object motion trajectory, negating need reproduce areas when boundaries move. Various tests...
Solving fluid–structure interaction (FSI) problems using traditional methods poses significant challenges in the field of numerical simulation. The multiphysics coupling library precise code environment (preCICE), renowned for its robust capabilities, offers a promising solution FSI problems. It supports various open/closed source software and commercial computational fluid dynamics solvers black box manner. However, preCICE currently mainly schemes mesh-based as well few meshless methods....
Contrastive learning is a powerful technique in the field of machine learning, specifically representation learning. The central idea contrastive to learn model by distinguishing between similar and dissimilar data points. This involves pulling points closer learned space while pushing farther apart.Imagine you have collection images, some which are different views same object completely unrelated. would aim generate embeddings (i.e., numerical representations) for these images such that...
Fluid-structure interaction dynamics under high-explosie detonation is widely used in engineering application, and its numerical simulation also challenging due to the violent reaction of blast. Based on open source software coupling library, we developed a solver support fluid-structure detonation, carried out case analysis. In test case, motion process perpendicular elastic flap high-explosive selected time 5ms. The transient results whole are given. We tested parallel, using two fluid...
In order to improve the local convergence of differential evolution algorithm, we puts forward greedy (GE) algorithm based on search strategy. According fitness value and selection probability, population a generation is classed best vectors, better vectors poor vectors. The retained in child population, replaced if newly generated vector its neighborhood than objective vector, regenerated until new not worse vector. Improving locally ability ensuring diversity GE increases obviously....
The Overset Grid method is a promising computational approach for tackling the challenging moving boundary problems in Computational Fluid Dynamics (CFD) simulations. efficiency and accuracy of are critically dependent on effectiveness Assembly (OGA) process. However, OGA process plagued by unavoidable issues load imbalance communication overheads, which adversely impact parallel method, particularly when dealing with sub-grids motion. This paper proposes an improved assembly as effective...
Due to the complex geometry and turbulent flow characteristics, it is hard simulate process of steam dumping pressurizer relief tank (PRT). In this study, we develop a compressible fluid solver PRTFOAM numerically study dynamics from PRT. The implemented based on OpenFOAM designed be capable integrating various turbulence models. Two representative Reynolds-averaged Navier–Stokes (RANS) models Smagorinsky–Lilly SGS model Large Eddy Simulation (LES) are coupled tested with PRTFOAM. case past...
Due to the complex geometry and physical models of real-world engineering applications, parallel performance mainstream computational fluid dynamics(CFD) codes is unsatisfactory. For fluids, an extra stress tensor governed by constitutive equations including nine components brings much more amount computations. This paper focused on optimizing most compute-intensive part a simulation for fluids: iterative linear solver solving multicomponent equations. Based widely used opensource CFD code...
Smoothed Particle Hydrodynamics (SPH) is a classical mesh-free particle method which has been successfully applied in the field of Computational Fluid Dynamics (CFD). Its advantages over traditional mesh-based methods have made it very popular simulating problems involving large deformation and free-surface flow. The high computational cost SPH obstructed its vast application. A lot research effort devoted to accelerating using GPU multi threading. However, developing efficient parallel...