检索结果-南通市图书馆

An Eulerian SPH method with WENo reconstruction for compressible and incompressible flows

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

journal of Hydrodynamics 2023年第2期35卷 210-221页

作者： Zhentong Wang Chi Zhang oskar j.haidn Xiangyu Hu TUM School of Engineering and Design Technical University of MunichMunichGermany Huawei Technologies Munich Research Center MunichGermany

While Eulerian smoothed particle hydrodynamics(SPH)method has received increasing attention in scientific and industrial communities owing to its high spatial accuracy,it exhibits excessive numerical dissipation due to the fact that the flux is derived in particle pair *** this paper,we adopt a one-dimensional weighted essentially non-oscillatory(WENo)reconstruction to reduce the numerical dissipation and improve the overall accuracy particularly in capturing the contact *** underlying principle is to construct a 4-point stencil along the interacting line of each particle pair and then the WENo scheme is applied to reconstruct the initial states of the Riemann problem which determines the flow flux.A set of benchmark tests for both compressible and incompressible flows are studied to investigate the accuracy,robustness and versatility of the proposed Eulerian SPH method with the WENo reconstruction(ESPH-WENo).

关键词： Eulerian smoothed particle hydrodynamics(SPH) compressible flows incompressible flows Riemann problem

Physics-Driven Learning of the Steady Navier-Stokes Equations using Deep Convolutional Neural Networks

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Communications in Computational Physics 2022年第8期32卷 715-736页

作者： Hao Ma Yuxuan Zhang Nils Thuerey Xiangyu Hu oskar j.haidn Department of Aerospace and Geodesy Technical University of Munich85748 GarchingGermany Beijing Aerospace Propulsion Institute 100076BeijingChina Technical University of Munich 85748GarchingGermany Department of Mechanical Engineering Technical University of Munich85748 GarchingGermany

Recently,physics-driven deep learning methods have shown particular promise for the prediction of physical fields,especially to reduce the dependency on large amounts of pre-computed training *** this work,we target the physicsdriven learning of complex flow fields with high *** propose the use of Convolutional neural networks(CNN)based U-net architectures to efficiently represent and reconstruct the input and output fields,*** introducingNavier-Stokes equations and boundary conditions into loss functions,the physics-driven CNN is designed to predict corresponding steady flow fields *** particular,this prevents many of the difficulties associated with approaches employing fully connected neural *** numerical experiments are conducted to investigate the behavior of the CNN approach,and the results indicate that a first-order accuracy has been *** for the case of a flow around a cylinder,different flow regimes can be learned and the adhered“twin-vortices”are predicted *** numerical results also show that the training for multiple cases is accelerated significantly,especially for the difficult cases at low Reynolds numbers,and when limited reference solutions are used as supplementary learning targets.

关键词： Deep learning physics-drivenmethod convolutional neural networks Navier-Stokes equations

Numerical study of operational processes in a Gox-kerosene rocket engine with liquid film cooling

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Propulsion and Power Research 2020年第2期9卷 132-141页

作者： Evgenij A.Strokach Igor N.Borovik Vladimir G.Bazarov oscar j.haidn Moscow Aviation Institute(National Research University) Moscow125080Russia Technical University of Munich Garching85748Germany

Combustion process inside kerosene-Gox rocket combustor with kerosene Alm cooling is studied,and a modeling approach is *** paper suggests to use the Lagrangian particle tracking technique to model fuel film behavior while the continuous fluid is simulated via the Navier-Stokes system of Favre-averaged *** approach is validated over the 12 experimental regimes by the criterions of characteristic velocity and pressure,ence on the adiabatic wall temperatures and relatively low impact on the *** general,phenomena,the calculation of operational processes becomes fast and robust yet precise en-the design process.

关键词： Liquid rocket engine Kerosene oxygen Favre-averaged Navier-Stokes Film cooling Numerical simulation

气氧甲烷单喷嘴燃烧室壁面热流的测量和数值模拟

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

推进技术 2018年第5期39卷 1033-1040页

作者：刘占一刘计武汪广旭石晓波 haidn o j 西安航天动力研究所液体火箭发动机技术重点实验室陕西西安710100 Institute for Flight Propulsion Technique University of Munichmunich85748

为了研究气氧甲烷燃烧室的壁面热载荷,对一个包含同轴剪切喷嘴的燃烧室开展了热试研究,混合比2.647,室压2MPa。根据沿燃烧室轴向测得的壁面温度数据,利用逆向传热计算的方法获得了壁面热流分布。为了更好地分析实验结果,应用商业CFD软件... 详细信息

为了研究气氧甲烷燃烧室的壁面热载荷,对一个包含同轴剪切喷嘴的燃烧室开展了热试研究,混合比2.647,室压2MPa。根据沿燃烧室轴向测得的壁面温度数据,利用逆向传热计算的方法获得了壁面热流分布。为了更好地分析实验结果,应用商业CFD软件ANSYS Fluent开展了相应的数值模拟研究,采用涡耗散概念模型模拟湍流燃烧过程。在进行分析之前对网格无关性进行了验证,比较了从数值模拟和实验获得的热流分布,结果表明所采取的模拟方法能够有效预测燃烧室壁面热流分布,最大热流值偏差17%,另外还从燃气温度场和流场结构方面分析了造成这种热流分布的原因。

关键词：火箭燃烧室壁面热流数值模拟