用大涡模拟方法研究湍流边界层流动噪声_英文_
fluent二维大涡模拟命令
fluent二维大涡模拟命令Fluent是流体动力学模拟软件的一种,它提供了二维大涡模拟命令用于模拟二维涡旋动力学过程。
本文将分步骤阐述如何使用Fluent二维大涡模拟命令。
第一步,打开Fluent软件。
进入“File”菜单,选择“New”打开一个新的工作文件。
在Fluent主界面的左侧面板选择“2D”选项卡,然后选择“Viscous”和“Steady”选项后点击“Create/Edit”按钮。
第二步,进入“Grid”界面。
在“Mesh”选项卡中选择“2D Mesh”菜单,选择“Triangle”网格类型。
随后,选择“Mechanical”类型并调整所需参数,包括网格的大小、分辨率、以及其他关键点的划分数量。
最后,点击“Generate Mesh”按钮生成网格。
第三步,设置边界条件。
在Fluent主界面的左侧面板选择“Boundary Conditions”选项卡。
根据需要设置边界条件,包括入口和出口边界、容器壁边界和物体边界。
基本的物理条件包括质量流速、温度和密度。
第四步,设置模拟参数。
在Fluent主界面的左侧面板选择“Solution”选项卡。
首先选择“Viscous”和“Steady”选项,然后在“Methods”菜单中选择“Unsteady”. 调整所需参数并计算时间,包括时间步长和计算时间范围。
第五步,开始求解二维大涡模拟。
在Fluent主界面的左侧面板选择“Compute”选项卡,点击“Start Calculation”按钮开始求解。
第六步,查看二维大涡模拟结果。
在Fluent主界面的左侧面板选择“Graphics”选项卡。
根据需要选择显示不同的结果,包括速度分布、温度变化、实体形态等等。
以上是使用Fluent二维大涡模拟命令的步骤。
通过学习和实践,我们可以使用Fluent来分析和解决各种相关的物理、化学和工程问题。
ACTRAN AERO-Acoustics_Theory_complete-ACTRAN气动声学理论完整版
半经验模型:不依赖于非定常的流体计算。
6
Copyright Free Field Technologies
20
Copyright Free Field Technologies
主控方程
两个定义, 没有假设:
A0是声源区之外静止流体的声速 a = - 0 其中0大气密度 0 是个常数:
得到(L1):
(L1)
如果观察点不在声源区,也没有均匀流动,那么, a = 声场密度
11
Copyright Free Field Technologies
混合方法
积分方法:
Lighthill, Curle, FW-H, Farassat, ... 都基于NS方程的方法 在声场内求解关于声学的显式方程以及计算声源对胜场内任何一点的贡献率 需要Green function
局限
声源项的不准确性(声源的统计) 对声学特性的预测相当困难 数值计算的成本不容忽视(大量的涡=大量的计算)
13
Copyright Free Field Technologies
Analogy concept
声类比理论
Copyright Free Field Technologies
压力 粘性应力 (2)
经过变换,方程(2)可以写成如下的形式:
常数 Lighthill应力张量
Lighthill应力张量T如下:
(T)
19
fluent 介绍
想起CFD,人们总会想起FLUENT,丰富的物理模型使其应用广泛,从机翼空气流动到熔炉燃烧,从鼓泡塔到玻璃制造,从血液流动到半导体生产,从洁净室到污水处理工厂的设计,另外软件强大的模拟能力还扩展了在旋转机械,气动噪声,内燃机和多相流系统等领域的应用。
今天,全球数以千计的公司得益于FLUENT的这一工程设计与分析软件,它在多物理场方面的模拟能力使其应用范围非常广泛,是目前功能最全的CFD软件。
FLUENT因其用户界面友好,算法健壮,新用户容易上手等优点一直在用户中有着良好的口碑。
长期以来,功能强大的模块,易用性和专业的技术支持所有这些因素使得FLUENT成为企业选择CF D软件时的首选。
网格技术,数值技术,并行计算计算网格是任何CFD计算的核心,它通常把计算域划分为几千甚至几百万个单元,在单元上计算并存储求解变量,FLUENT使用非结构化网格技术,这就意味着可以有各种各样的网格单元:二维的四边形和三角形单元,三维的四面体核心单元、六面体核心单元、棱柱和多面体单元。
这些网格可以使用FLUENT的前处理软件GAMBIT自动生成,也可以选择在ICEM CFD工具中生成。
六面体核心网格四边形平铺网格在目前的CFD市场, FLUENT以其在非结构网格的基础上提供丰富物理模型而著称,久经考验的数值算法和鲁棒性极好的求解器保证了计算结果的精度,新的NITA算法大大减少了求解瞬态问题的所需时间,成熟的并行计算能力适用于NT,Linux或Unix平台,而且既适用单机的多处理器又适用网络联接的多台机器。
动态加载平衡功能自动监测并分析并行性能,通过调整各处理器间的网格分配平衡各CPU的计算负载。
并行速度的比较湍流和噪声模型FLUENT的湍流模型一直处于商业CFD软件的前沿,它提供的丰富的湍流模型中有经常使用到的湍流模型、针对强旋流和各相异性流的雷诺应力模型等,随着计算机能力的显著提高,FLUENT已经将大涡模拟(LES)纳入其标准模块,并且开发了更加高效的分离涡模型(DES),FLUENT提供的壁面函数和加强壁面处理的方法可以很好地处理壁面附近的流动问题。
流体力学常用名词术语中英文对照
流体力学常用名词术语中英文对照CFL条件 Courant- Friedrichs- Lewy condition ,CFL condition KDU方程 KDV equationU形管 U-tubeA埃克特数 Eckert number奥尔-索末菲方程 Orr-Sommerfeld equation奥辛流 Oseen flowB巴塞特力 Basset force板块法 panel method半隐格式 semi-implicit scheme薄翼理论 thin-airfoil theory保单调差分格式monotonicity preserving diffe-rence scheme爆轰 detonation爆炸 explosion毕奥-萨伐尔定律 Biot-Savart law壁剪切速度 friction velocity壁剪应力 skin friction, frictional drag壁效应 wall effect边界层 boundary layer边界层方程 boundary layer equation边界层分离 boundary layer separation边界层厚度 boundary layer thickness边界层理论 boundary later theory边界层转捩 boundary layer transition边界元法 boundary element method变分法 variational method标记网格法 marker and cell method, MAC method 表观粘度 apparent viscosity表面波 surface wave表面力 surface force表面张力 surface tension表面张力波 capillary wave波高 wave height波列 wave train波能 wave energy波群 wave group波速 wave speed, wave velocit波阻 wave drag伯格斯方程 Burgers equation伯努利定理 Bernonlli theorem伯努利方程 Bernoulli equation泊 poise泊肃叶流 Poiseuille flow不定常法 time-dependent method不规则波 irregular wave不可压缩流[动] incompressible flow不可压缩流体 incompressible fluid不可压缩性 incompressibility不稳定性 instability布拉休斯解 Blasius solution部分相似 partial similarityC测速法 anemometry测压孔 pressure tap层流 laminar flow层流边界层 laminar boundary layer层流分离 laminar separation掺气流 aerated flow超空化流 supercavitating flow超空泡 supercavity超空泡流 supercavity flow超声速流[动] supersonic flow超压[强] over pressure潮波 tidal wave沉[降堆]积 sedimentation, deposition沉积物 sediment, deposit沉降速度 settling velocity成核 nucleation尺度效应 scale effect冲击波 shock wave出口 exit, outlet出口压力 exit pressure传播 propagation传导 conduction传感器 transducer, sensor传热系数 heat transfer coefficient传质系数 mass transfer coefficient船波 ship wave次层 sublayer猝发过程 bursting processD达朗贝尔佯廖 d'Alembert paradox大涡模拟 large eddy simulation单调差分格式 monotone difference scheme 单相流 single phase flow单组份流 single-component flow当地马赫数 local Mach number等熵流 isentropic flow低速空气动力学 low-speed aerodynamics狄里克雷边界条件 Dirichlet boundary condition 底压 base pressure地面效应 ground effect定理 pi theorem, Buckingham theorem定倾中心 metacenter动力相似[性] dynamic similarity动力粘性 dynamic viscosity动量方程 momentum equation动量厚度 momentum thickness动量交换 momentum transfer动量守恒 conservation of momentum动态响应 dynamic response动态校准 dynamic calibration堵寒效应 blockage effect堵塞 blockage杜福特-弗兰克尔格式 Dufort-Frankel scheme 对流 convection对流传热 convective heat transfer对流扩散方程 convection diffusion equation对流涡胞 convective cell钝体 blunt body钝头体 bluff body多管压强计 multiple manometer多重尺度问题 multiple scale problem多重网格法 multi- grid methodE二次流 secondary flow二维流 two-dimensional flowF反扩散差分格式 anti-diffusion difference scheme 反流 reverse flow反射 reflection反压 back pressure放大矩阵 amplification matrix放大因子 amplification factor非定常流 unsteady flow, non-steady flow非绝热流 diabatic flow非均匀流 nonuniform flow非平衡流[动] non-equilibrium flow非线性波 nonlinear wave非线性不稳定性 nonlinear instability菲克定律 Fick law沸腾 boiling分布 distribution分步法 fractional step method分层流 stratified flow分离点 separation point分离流 separated flow分子扩散 molecular diffusion风洞 wind tunnel风速管 Pitot- static tube冯.诺伊曼条件 von Neumann condition弗劳德数 Froude number浮体 floating body辐射传热 radiative heat transfer附加质量 added mass ,associated mass附面层 boundary layer附体激波 attached shock wave附着点 attachment point附着涡 bound vortex复势 complex potential复速度 complex velocityG伽辽金法 Galerkin method盖斯特纳波 Gerstner wave高度水头 elevating head高分辨率格式 high resolution scheme高速空气动力学 high-speed aerodynamics 戈本诺夫格式 Godunov scheme格拉斯霍夫数 Grashof number构型 configuration孤立子 soliton拐角流 corner flow管流 pipe flow, tube flow规则波 regular wave过水断面 flow cross-sectionH海洋水动力学 marine hydrodynamics亥姆霍兹定理 Helmholtz theorem含沙流 sediment-laden stream含水层 aquifer焓厚度 enthalpy thickness耗散 dissipation洪水波 flood wave后缘 trailing edge滑移速度 slip velocity环量 circulation环流 circulation缓流 subcritical flow缓燃 deflagration回流 back flow汇 sinkJ积分方法 integral method激波 shock wave激波捕捉法 shock-capturing method激波层 shock layer激波管 shock tube激波管风洞 shock tube wind tunnel激波拟合法 shock-fitting method激波阵面 shock front激光多普勒测速计laser Doppler anemometer,laser Doppler velocimeter急变流 rapidly varied flow急流 supercritical flow几何相似 geometric similarity计算流体力学 computational fluid mechanics计算区域 computational domain迹线 path, path line减阻 drag reduction剪切层 shear layer剪切流 shear flow渐变流 gradually varied flow浆体 slurry降水曲线 dropdown curve交错网格 staggered mesh交替方向隐格式alternating direction implicit scheme, ADIscheme介质 medium紧差分格式 compact difference scheme进口 entrance, inlet近场流 near field flow近似因子分解法 approximate factorization method 静[态]校准 static calibration静压管 static [pressure]tube静压头 static head镜象法 image method局域相似 local similarity矩量法 moment method卷筒涡胞 roll cell绝热流 adiabatic flow均匀流 uniform flowK卡门涡街 Karman vortex street开尔文定理 Kelvin theorem可压缩流[动] compressible flow可压缩流体 compressible fluid克兰克-尼科尔森格式 Crank-Nicolson scheme空化 cavitation空化数 cavitation number空泡流 cavity flow空气动力学 aerodynamics空蚀 cavitation damage孔板流量计 orifice meter孔流 orifice flow控制体积 control volume库埃特流 Couette flow库塔-茹可夫斯基条件 Kutta-Zhoukowski condition 跨声速流[动] transonic flow扩散 diffusion扩散段 diffuser扩散率 diffusivity扩散速度 diffusion velocity扩散性 diffusivityL拉克斯等价定理 Lax equivalence theorem拉克斯-温德罗夫格式 Lax-Wendroff scheme拉瓦尔喷管 Laval nozzle来流 incoming flow兰金-于戈尼奥条件 Rankine-Hugoniot condition 雷诺比拟 Reynolds analogy雷诺数 Reynolds number厘泊 centipoise厘沱 centistoke离解 dissociation离散涡 discrete vortex黎曼解算子 Riemann solver连续介质假设 continuous medium hypothesis连续介质力学 mechanics of continuous media涟漪 ripple量热状态方程 caloric equation of state临界雷诺数 critical Reynolds number临界流 critical flow临界热通量 critical heat flux流[动] flow流场 flow field流出边界条件 outflow boundary condition流动参量 flow parameter流动分离 flow separation流动稳定性 flow stability流动显示 flow visualization流管 stream tube流函数 stream function流量 flow rate, flow discharge流量计 flow meter流面 stream surface流入边界条件 inflow boundary condition流速计 anemometer流态 flow regime流体动力学 fluid dynamics流体网格法 fluid in cell method,FLIC method 流体运动学 fluid kinematics流体质点 fluid particle流线 stream lineM马赫波 Mach wave马赫角 Mach angle马赫数 Mach number马赫线 Mach line马赫锥 Mach cone马蹄涡 horseshoe vortex脉线 streak line毛细[管]作用 capillarity弥散 dispersion明槽流 open channel flow磨擦损失 friction loss磨擦因子 friction factor穆曼-科尔格式 Murman-Cole schemeN纳维-斯托克斯方程 Navier-Stokes equation 内流 internal flow能量传递 energy transfer能量法 energy method能量方程 energy equation能量厚度 energy thickness能量守恒 conservation of energy能量输运 energy transport拟序结构 coherent structure粘度测定法 visco[si] metry粘度计 visco[si] meter粘性流[动] viscous flow凝结 condensation牛顿流体 Newtonian fluid浓度 concentration努塞特数 Nusselt numberO欧拉方程 Euler equation欧拉数 Euler number偶极子 doublet, dipoleP排放量 discharge排水 drainage配置方法 collocation method喷管 Nozzle皮托管 pitot tube频率响应 frequency response平面流 plane flow破碎波 breaking wave普朗特-迈耶流 Prandtl-Meyer flow普朗特数 prandtl number普雷斯顿管 preston tube谱方法 spectral methodQ起动涡 starting vortex气动加热 aerodynamic heating气动力 aerodynamic force气动热力学 aerothermodynamics气动噪声 aerodynamic noise气动中心 aerodynamic center气化 gasification气体动力学 gas dynamics气体润滑 gas lubrication前缘涡 leading edge vortex浅水波 shallow water wave强迫对流 forced convection强守恒型 strong conservation form氢泡法 nydrogen bubble method区域分解 domain decomposition全变差下降格式total variation decreasing scheme TVD schemeR扰动 disturbance, perturbation绕射 diffraction热传导 conductive heat transfer热对流 heat convection热交换 heat exchange热量传递 heat transfer热敏电阻 thermistor热膜流速计 hot- film anemometer热线流速计 hot-wire anemometer热状态方程 thermal equation of state 人工压缩 artificial compression人工粘性 artificial viscosity瑞利流 Rayleigh flow瑞利数 Rayleigh number弱解 weak solution弱守恒型 weak conservation formS三维流 three-dimensional flow散度型 divergence form散射 scattering色散 dispersion熵函数 entropy function熵条件 entropy condition熵通量 entropy flux射流 jet深水波 deep water wave失速 stall施密特数 schmidt number施特鲁哈尔数 Strouhal number时间分步法 time splitting method时间线 time line示踪物 tracer势 potential势流 potential flow适应网格生成 adaptive grid generation 收缩 contraction守恒差分格式 conservation differences cheme 守恒型 conservation form数值边界条件 numerical boundary condition 数值耗散 numerical dissipation数值模拟 numerical simulation数值色散 numerical dispersion数值通量 numerical flux数值网格生成 numerical grid generation数值粘性 numerical viscosity数植扩散 numerical diffusion水动[力]噪声 hydrodynamic noise水动力学 hydrodynamics水洞 water tunnel水击 water hammer水静力学 hydrostatics水力半径 hydraulic radius水力坡度 hvdraulic slope水力学 hydraulics水头损失 head loss水位 water level水翼 hydrofoil水跃 hydraulic jump丝线法 tuft method斯坦顿管 Stanton tube斯坦顿数 Stanton number斯托克斯波 Stokes wave速度[水]头 velocity head速度环量 velocity circulation速度亏损律 velocity defect law速度剖面 velocity profile速度势 velocity potential随机选取法 random choice method随体导数 material derivativeT泰勒不稳定性 Taylor instability泰勒数 Taylor number泰勒涡 Taylor vortex特征线法 method of characteristics贴体曲线坐标 body- fitted curvilinear coordi-nates 通量矢量分解法 flux vector splitting method通量校正传输法 flux-corrected transport method 头波 bow wave投影法 projection method湍流边界层 turbulent boundary layer湍流分离 turbulent separation拖曳水池 towing tank脱体激波 detached shock wave椭圆余弦波 cnoidal waveW蛙跳格式 leap-frog scheme外流 external flow完全气体 perfect gas网格雷诺数 cell Reynolds number微压计 micromanometer维数分解 dimensional split尾流 wake [flow]未扰动流 undisturbed flow位移厚度 displacement thickness温度边界层 thermal boundary layer文丘里管 Venturi tube纹影法 schlieren method涡 eddy涡层 vortex layer涡对 vortex pair涡方法 vortex method涡管 vortex tube涡环 vortex ring涡街 vortex street涡量 vorticity涡量方程 vorticity equation涡量计 vorticity meter涡量拟能 enstrophy涡面 vortex surface涡片 vortex sheet涡丝 vortex filament涡线 vortex line涡旋 vortex涡旋破碎 vortex breakdown涡旋脱落 vortex shedding涡粘性 eddy viscosity无反射边界条件 nonreflecting boundary condition 无滑移条件 non-slip condition无量纲参数 dimensionless parameter无旋流 irrotational flow无压流 free surface flow无粘性流体 nonviscous fluid, inviscid fluid物理解 physical solution物理区域 physical domainX吸出 suction希尔特稳定性分析 Hirt stability analysis稀疏波 rarefaction wave稀疏矩阵分解法 split coefficient matrix method 细长度 slenderness细长体 slender body显格式 explicit scheme相容条件 consistency condition相似理论 similarity theory相似律 similarity law相似性解 similar solution响应频率 response frequency消能 energy dissipation楔流 wedge flow斜激波 oblique shock wave斜迎风格式 skew-upstream scheme谢齐公式 Chezy formula形状因子 shape factor型阻 profile drag修正微分方程 modified differential equation 旋臂水池 rotating arm basinY压[力]降 pressure drop压[强水]头 pressure head压差 differential pressure压差阻力 pressure drag压力能 pressure energy压强传感器 pressure transducer压强计 manometer压缩波 compression wave亚声速流[动] subsonic flow烟丝法 smoke wire method衍射 diffraction堰流 weir flow叶栅流 cascade flow液体动力润滑 hydrodynamic lubrication液体动力学 hydrodynamics液体静力学 hydrostatics液体静压 hydrostatic pressure依赖域 domain of dependence异重流 density current, gravity flow翼弦 chord翼型 airfoil阴影法 shadow method隐格式 implicit scheme迎风格式 upstream scheme , upwind scheme 迎角 angle of attack影响域 domain of influence壅水曲线 back water curve涌波 surge wave油膜显示 oil film visualization油烟显示 oil smoke visualization有限体积法 finite volume method有旋流 rotational flow有压流 pressure flow诱导速度 induced velocity诱导阻力 induced drag预估校正法 predictor-corrector method源 source远场边界条件 far field boundary condition 远场流 far field flow约束涡 confined vortex匀熵流 homoentropic flow运动相似 kinematic similarity运动粘性 kinematic viscosityZ再层流化 relaminarization再附 reattachment暂态流 transient flow折射 refraction振荡流 oscillatory flow蒸发 evaporation正激波 normal shock wave支配方程 governing equation直线法 method of lines指数格式 exponential scheme质点法 particle method质点网格法 particle in cell method,PIC method 质量传递 mass transfer质量守恒 conservation of mass滞后 lag滞止流 stagnation flow重力波 gravity wave周期流 periodic flow轴对称流 axisymmetric flow注入 injection驻点 stagnation point驻涡 standing vortex状态方程 equation of state锥形流 conical flow准定常流 quasi-steady flow准谱法 pseudo-spectral method自动网格生成 automatic grid generation [自]适应网格 [self-] adaptive mesh自由对流 natural convection, free convec-tion 自由流 free stream自由流线 free stream line自由面 free surface自由射流 free jet总焓 total enthalpy总压[力] total pressure总压头 total head阻力 drag, resistance阻尼误差 damping error。
2023 年湍流与噪声和 CFD 方法暑期高级讲习班 会议手册说明书
2023年湍流与噪声和CFD方法暑期高级讲习班2023 Advanced Summer Program on Turbulence,Noise and CFD Methods会议手册时间:2023年7月28至8月5日地点:香港科技大学主办单位:中国空气动力学会承办单位:香港科技大学(HKUST)上海大学南方科技大学复旦大学中国空气动力学会CFD专委会中国空气动力学会低跨超专委会上海市应用数学和力学研究所上海市力学信息学前沿科学基地上海市能源工程力学重点实验室粤港澳数据驱动下的流体力学与工程应用联合实验室中国航空学会航空声学分会协办单位:《空气动力学学报》《实验流体力学》《Advances in Aerodynamics》二零二三年七月二十六日2023年湍流与噪声和CFD方法暑期高级讲习班为了促进流体力学与空气动力学的发展、推动学术交流与合作、培育培养优秀人才,助力解决流体力学与空气动力学等相关领域“卡脖子”技术,经中国空气动力学会批准,2023年湍流与噪声和CFD 方法暑期高级讲习班将于2023年7月28日至8月5日在香港科技大学(HKUST)举行。
会议邀请内地与香港地区在湍流、噪声和CFD方法等方面的专家学者、青年学者为讲习班授课。
现诚邀内地与港澳台地区研究生、工程师、相关领域专家学者以及高年级本科生参会。
本次讲习班由中国空气动力学会主办,香港科技大学(HKUST)、上海大学、南方科技大学、复旦大学、中国空气动力学会CFD专委会、中国空气动力学会低跨超专委会、上海市应用数学和力学研究所、上海市力学信息学前沿科学基地、上海市能源工程力学重点实验室、粤港澳数据驱动下的流体力学与工程应用联合实验室等单位承办。
本次讲习班采用线上线下同时进行的方式,其中线上使用腾讯会议App进行直播,会议号码:964-8147-9182,也可直接扫描下面的二维码参会:2023年湍流与噪声和CFD方法暑期高级讲习班专家报告日程安排报告安排以专家自选日程排列,不分先后次序,后续如有变动以最终表格为准。
fluent 二维大涡模拟命令
fluent 二维大涡模拟命令Fluent(通常称为ANSYS Fluent)是一种基于计算流体动力学(CFD)的软件,它使用数值方法解决流体力学和热力学方程。
Fluent支持多个求解器,包括稳态、非稳态、可压缩和不可压缩流体求解器。
其中,二维大涡模拟(Large Eddy Simulation,LES)是一种用于模拟湍流流动的CFD方法,通过分解流体的速度场为大尺度和小尺度来模拟湍流流动。
本文将介绍Fluent中二维大涡模拟的相关命令。
1. 设定模拟参数在开始二维大涡模拟前,需要设定一些模拟参数,包括流体属性和边界条件。
在Fluent中,通过以下命令可以设定流体属性和边界条件:(1)设定流体属性DEFINE > MODELS > VISCOSITY2. 定义二维网格在进行CFD模拟前,需要先定义计算网格,以便数值求解器能够在其上执行算法。
在Fluent中,通过以下命令定义二维网格:(1)导入二维网格FILE > IMPORT > MESH3. 指定求解器有关Fluent的求解器已经在第一段中提到。
在进行二维大涡模拟时,可以选择可压缩或不可压缩流体求解器作为替代。
(2)可压缩流体求解器SOLVE > COMPRESSIBLE FLOW/HEAT TRANSFER > STEADY模拟模型是模拟过程中使用的具体模型。
在Fluent中,用户可以选择不同的模拟模型。
(1)可分离流边界层(Detached Eddy Simulation,DES)MODEL > VISCOSITY > DES(2)壁面函数(Wall Function)MODEL > VISCOSITY > WALL FUNCTION在进行CFD模拟时,需要设定一些计算参数以控制模拟进程,以及获得所需的结果。
在Fluent中,用户可以使用以下命令设定计算参数:6. 运行模拟在完成所有设定后,可以通过以下命令在Fluent中运行二维大涡模拟:SOLVE > EXECUTE COMMAND FILE > RUN此时,Fluent将自动执行过程,直至收敛或达到设定的计算时间。
fluent-介绍
想起CFD,人们总会想起FLUENT,丰富的物理模型使其应用广泛,从机翼空气流动到熔炉燃烧,从鼓泡塔到玻璃制造,从血液流动到半导体生产,从洁净室到污水处理工厂的设计,另外软件强大的模拟能力还扩展了在旋转机械,气动噪声,内燃机和多相流系统等领域的应用。
今天,全球数以千计的公司得益于FLUENT的这一工程设计与分析软件,它在多物理场方面的模拟能力使其应用范围非常广泛,是目前功能最全的CFD软件。
FLUENT因其用户界面友好,算法健壮,新用户容易上手等优点一直在用户中有着良好的口碑。
长期以来,功能强大的模块,易用性和专业的技术支持所有这些因素使得FLUENT成为企业选择CFD 软件时的首选。
网格技术,数值技术,并行计算计算网格是任何CFD计算的核心,它通常把计算域划分为几千甚至几百万个单元,在单元上计算并存储求解变量,FLUENT使用非结构化网格技术,这就意味着可以有各种各样的网格单元:二维的四边形和三角形单元,三维的四面体核心单元、六面体核心单元、棱柱和多面体单元。
这些网格可以使用FLUENT的前处理软件GAMBIT自动生成,也可以选择在ICEM CFD工具中生成。
六面体核心网格四边形平铺网格在目前的CFD市场, FLUENT以其在非结构网格的基础上提供丰富物理模型而著称,久经考验的数值算法和鲁棒性极好的求解器保证了计算结果的精度,新的NITA算法大大减少了求解瞬态问题的所需时间,成熟的并行计算能力适用于NT,Linux或Unix平台,而且既适用单机的多处理器又适用网络联接的多台机器。
动态加载平衡功能自动监测并分析并行性能,通过调整各处理器间的网格分配平衡各CPU的计算负载。
并行速度的比较湍流和噪声模型FLUENT的湍流模型一直处于商业CFD软件的前沿,它提供的丰富的湍流模型中有经常使用到的湍流模型、针对强旋流和各相异性流的雷诺应力模型等,随着计算机能力的显著提高,FLUENT已经将大涡模拟(LES)纳入其标准模块,并且开发了更加高效的分离涡模型(DES),FLUENT提供的壁面函数和加强壁面处理的方法可以很好地处理壁面附近的流动问题。
Large-Eddy Simulation and Trailing-Edge Noise
Cp
压强系数, [-] 翼型弦长, [m] 声速, [m/s] Fresnel混合积分符号, [-] 无量纲噪声能量, [-]
c
c0
E ,F
Eqq
erf
错误方程, [-] 频率,简化频率, [Hz] Amiet模型辐射部分, [-]
f , fs I, I LE , ITE
k
c0
声学波数, [m−1]
III .
大漩涡仿真
为了研究边界层扰动的模型问题和风洞设施本身对航空声学预测的影响,我们分别用了三种不同的 LES仿真法。其中两种仿真法网格上包含了一个边界层扰动模型,从而考虑了风洞设施的影响,然而,第三 种仿真计算了自由飞行条件下无扰动翼型的流动。从现在起,我们把包含有精确的边界层扰动装置几何外 形的LES实验称为锯齿片扰乱LES;把含有简化扰乱装置几何外形的LES实验称为阶梯片扰动LES;把没有边 界层扰动装置且不考虑风洞设施影响的LES实验称为自由飞LES。在接下来的三节中将给出流动解算装置、 数值装置、边界层条件和拓扑网格方面的细节。
5
型表面存在较长的层流区,从而引起Tollmien-Schlichting不稳定波,这种不稳定波含有一个相关的附加 音调和宽频噪声辐射。为了避免这种附加的噪声源的出现,实验中要求翼型上下表面都存在边界层分离。 这种分离以一种充分复杂的结构和一种简化的几何形被加入到LES的数字网格上。最后,在不同的边界条件 和边界层分离模型情况下,对三种不同LES算法进行比较。根据典型的LES方法的空气动力学和航空声学预 测能力来确定模型的影响。通过对比,我们再一次强调了精确的边界条件在LES方法中的必要性,精确的边 界条件主要是为了得到与风洞实验具有可比性的结果。 名称符号表 字母:
大涡模拟的原理
大涡模拟的原理
大涡模拟(LES)是一种计算流体力学(CFD)方法,用于模拟流动中的大尺度涡旋行为。
相比于传统的雷诺平均纳维-斯托克斯(RANS)方法,LES可以更准确地捕捉流动中的湍流结构。
LES将流动场分解
为大尺度涡旋和小尺度涡旋,大尺度涡旋被直接模拟,而小尺度涡旋则被认为是一种随机噪声,并通过子网格模型(SGS)计算。
LES方法的基本原理是通过在时间和空间上对流场进行分解,将大尺度的湍流结构通过直接数值模拟(DNS)进行计算,而小尺度的
结构则通过SGS模型计算。
LES方法在时间上的分解通常采用滤波器方法,通过对流场进行滤波来分离大尺度结构和小尺度结构。
在空间上的分解通常采用泰勒级数展开,将流场分解为平均流量和流量扰动。
LES方法的优点是可以提供更准确的流场预测,适用于需要对湍流结构进行精细分析的复杂流动问题。
同时,LES方法也存在一些挑战,如计算成本高和需要更高的计算资源等问题。
因此,LES方法通常适用于高性能计算领域和需要进行高精度模拟的工程和科学研究
领域。
- 1 -。
不同舵角的舵翼结构涡量及流噪声特性分析
不同舵角的舵翼结构涡量及流噪声特性分析屈铎;张振海;楼京俊【摘要】The flow field and sound field of trapezoidal rudder-wing under different rudder angles are numerically pre-dicted by CFD LES theory and Lighthill acoustic analogy theory, and characteristics of vorticity and flow noise are analyzed. Results show that: at the same speed, the vortex is more and more complex and the vorticity and flow noise increases with the increasing of rudder angle; vortex mainly concentrates in stabilizing wing leading edge, trailing edge of rudder-wing and between rudder and stabilizing wing; sound pressure level spectrum band of flow noise is wide and there is no obvious dom-inant frequency; at the low frequency, sound pressure level is higher, and continues to decline with the increasing of fre-quency; sound intensity at the front of leading edge and after trailing edge is higher than that at both sides of rudder-wing. This is also consistent with the results of flow field vorticity analysis, which shows that vortex is the root cause of flow noise.%以梯形舵翼结构为研究对象,采用CFD大涡模拟及Lighthill声类比理论对不同舵角下舵翼结构的流场和声场进行数值预报,分析其涡量特性及流噪声特性.结果表明:来流速度相同时,随着舵角的增大,涡系越来越复杂,涡量及流噪声也随之增大;涡系主要集中在稳定翼的导边、舵翼的尾缘及舵与稳定翼之间;舵翼结构流噪声的声压级频谱频带较宽,无明显的主频率出现;低频时声压级幅值较大,并且随着频率升高而持续下降;舵翼尾缘及稳定翼导边前缘的声场强度比翼型两侧的声场强度大,这也和流场涡量分析结果一致,进而说明了涡流是产生流噪声的根本原因.【期刊名称】《舰船科学技术》【年(卷),期】2018(040)006【总页数】5页(P40-44)【关键词】舵翼;涡量;流噪声;舵角;大涡模拟;Lighthill声类比理论【作者】屈铎;张振海;楼京俊【作者单位】海军工程大学动力工程学院,湖北武汉 430033;海军工程大学船舶振动噪声重点实验室,湖北武汉 430033;海军工程大学科研部,湖北武汉 430033;海军工程大学动力工程学院,湖北武汉 430033;海军工程大学船舶振动噪声重点实验室,湖北武汉 430033【正文语种】中文【中图分类】TU1310 引言舰艇的噪声源很多且很复杂,主要分为机械噪声、螺旋桨噪声及水动力噪声。
玻尔兹曼方法的鱼类运动的大涡模拟
玻尔兹曼方法的鱼类运动的大涡模拟大涡模拟(Large Eddy Simulation,LES)是一种计算流体力学方法,用于模拟湍流流动。
它通过将流体运动分解为大尺度的宏观运动和小尺度的微观运动,通过直接模拟大尺度涡旋,而使用模型来描述小尺度涡旋的效应。
大涡模拟在流体力学领域具有广泛的应用,包括风力发电机、汽车气动和空气动力学研究等。
要进行鱼类运动的大涡模拟,需要进行以下步骤:1.网格划分:将计算区域划分为网格,通过细分网格可以更准确地模拟流场的细节。
在划分网格时,需要考虑到鱼类的大小和运动范围,以确保模拟结果的准确性。
2.描述鱼类运动:通过给定鱼类的姿态、速度和角速度等参数,可以描述鱼类在水中的运动行为。
这些参数可以通过观察实际鱼类的行为或者根据生物学模型估计得到。
3.边界条件:在模拟中,需要设置合适的边界条件来描述鱼类和水流之间的相互作用。
例如,可以通过施加一定的速度或力来模拟鱼类对水流的作用。
4.数值求解:利用玻尔兹曼方法对流体的动力学行为进行模拟。
玻尔兹曼方法是一种基于统计力学的方法,它通过分子碰撞的概率来描述流体粒子的运动。
在模拟中,需要使用适当的数值方法求解玻尔兹曼方程。
5.分析结果:通过模拟结果,可以分析鱼类运动时水流的速度、压力和湍流特性等参数。
这些参数可以帮助我们更好地理解鱼类的游动行为,并对鱼类在水中的运动和生物力学特性进行研究。
鱼类运动的大涡模拟可以帮助我们更好地理解鱼类的游动行为和其对周围水流的影响。
对于生物力学研究和水生生态学研究而言,鱼类运动的大涡模拟可以为我们提供一个全新的视角,并为我们提供更全面和准确的数据,从而更好地保护和管理水生生物资源。
总之,玻尔兹曼方法的鱼类运动的大涡模拟具有重要的科学研究意义和应用价值。
通过对鱼类游动时水流的模拟和分析,我们可以揭示鱼类游动行为的机理和原理,并对水生生物生态系统进行保护和管理提供科学依据。
基于涡流发生器的流动控制与降噪技术研究
基于涡流发生器的流动控制与降噪技术研究潜艇在水下高速航行时,水动力噪声成为主要噪声源,极大的破坏了潜艇的
声隐身性能,涡流发生器是空气动力学中较为常见的流动控制装置,本文提出了
基于涡流发生器控制指挥台围壳的不稳定流动,降低其水动力噪声的方法,从流
动控制角度为降低潜艇水动力噪声提供了新的思路。
本文以SUBOFF标准潜艇的指挥台围壳-艇身模型为研究对象,通过大涡模拟(LES)求解流场信息,利用声类
比及有限元与无限元结合的方法求解流激噪声,分析围壳与艇身结合处产生的马蹄涡激励围壳产生的流激噪声特性,采用在围壳前缘与艇身结合处施加机械式涡流发生器的方法,减弱马蹄涡的强度,降低因马蹄涡产生的流激噪声。
分析围壳表面边界层分离与尾涡脱落产生的流激噪声特性,采用在围壳转捩区施加微型涡流发生器的方法,控制边界层分离,降低因边界层分离与尾涡脱落产生的流激噪声,分析了机械式涡流发生器与微型涡流发生器的流动与噪声控制机理。
通过改变机械式涡流发生器的形状,与来流方向夹角,距围壳前缘距离;改变微型涡流发生器的攻角,入射角、高度,确定了降噪效果最佳的两类涡流发生器几何参数。
在此基础上,通过开展试验验证,在重力式水洞中利用混响法与湍流脉动压
力测量法测量了添加机械式涡流发生器模型的水动力噪声,进一步评价了机械式涡流发生器的降噪能力,验证了数值计算方法的准确性。
本文的研究结果为潜艇水动力噪声的治理提供了相关参考,为高航速条件下的潜艇减振降噪奠定了基础。
大气边界层中湍流运动的模拟与分析
大气边界层中湍流运动的模拟与分析大气边界层中的湍流运动对天气预报、空气质量评估以及风电场的建设等领域具有重要的影响。
因此,对大气边界层中的湍流运动进行模拟与分析,能够为解决相关问题提供有效的支持和参考。
本文将介绍湍流运动的模拟方法以及相关分析技术。
一、湍流模拟方法湍流模拟是通过数值方法对大气边界层中的湍流运动进行数值模拟,从而获取湍流场的详细信息。
目前常用的湍流模拟方法包括直接数值模拟(DNS)、大涡模拟(LES)和雷诺平均湍流模拟(RANS)等。
1. 直接数值模拟(DNS)直接数值模拟是一种以最基本的方程组为基础,对大气边界层中湍流运动进行精确模拟的方法。
它通过离散化时间和空间,使用计算机求解Navier-Stokes方程组,得到湍流场的精确解。
但直接数值模拟的计算量非常大,通常仅适用于小尺度或小时间尺度的模拟。
2. 大涡模拟(LES)大涡模拟是一种介于直接数值模拟和雷诺平均湍流模拟之间的方法。
它通过将流场分解为一个大尺度的结构和一个小尺度的湍动结构,只对小尺度湍动进行模拟,通过模拟大尺度结构来减小计算量。
大涡模拟在模拟大气边界层湍流运动方面具有一定的优势。
3. 雷诺平均湍流模拟(RANS)雷诺平均湍流模拟是一种通过对时间和空间进行平均,将湍流场表示为平均量和脉动量的和的方法。
它通过求解雷诺平均Navier-Stokes方程和湍流能量方程,得到湍流场的平均解。
雷诺平均湍流模拟在计算上相对简单,适用于大尺度湍流的模拟。
二、湍流分析技术湍流模拟得到的湍流场数据需要进行进一步的分析才能得到有用的信息。
下面介绍几种常用的湍流分析技术。
1. 自相关函数自相关函数是一种分析湍流场中各点相关性的方法。
它可以通过计算不同点之间的相关性来获取湍流运动的相关长度。
自相关函数可以用于描述湍流场的时空结构。
2. 能谱分析能谱分析是一种通过计算湍流场不同频率分量的能量来了解湍流场特性的方法。
它可以用于表征湍流场的能量分布情况和主导长度尺度。
大涡模拟fluent动量格式
大涡模拟fluent动量格式【原创版】目录1.大涡模拟的概述2.Fluent 软件的介绍3.大涡模拟中的动量格式4.动量格式在大涡模拟中的应用5.结论正文一、大涡模拟的概述大涡模拟是一种用于研究流体运动的数值模拟方法。
在计算机科学发展的过程中,人们为了更好地理解流体的运动规律,提出了大涡模拟的思想。
该方法通过将流体运动中的大尺度涡旋与小尺度涡旋分离,然后对大尺度涡旋进行数值模拟,从而获得流体运动的整体特征。
大涡模拟在气象学、海洋学、航空航天等领域具有广泛的应用。
二、Fluent 软件的介绍Fluent 是一款专业的流体动力学模拟软件,可以用于模拟各种流体运动问题,如湍流、热传导、化学反应等。
Fluent 软件采用计算流体动力学(CFD)方法,可以模拟流体在各种几何形状和物理条件下的运动状态。
此外,Fluent 还具有强大的图形功能,可以直观地显示流场的压力、速度、温度等物理量。
三、大涡模拟中的动量格式在大涡模拟中,动量格式是用于描述流体运动中动量传递的数学方程。
动量格式主要包括以下几个方面:1.质量守恒:描述流体在运动过程中质量的守恒原理,即流入和流出一个体积元的质量之和保持不变。
2.动量守恒:描述流体在运动过程中动量的守恒原理,即流入和流出一个体积元的动量之和保持不变。
3.能量守恒:描述流体在运动过程中能量的守恒原理,即流入和流出一个体积元的能量之和保持不变。
四、动量格式在大涡模拟中的应用在大涡模拟中,动量格式主要用于计算流体运动的速度、压力等物理量。
通过动量守恒方程,可以求解出流体运动的速度场;通过质量守恒方程,可以求解出流体运动的压力场。
此外,动量格式还可以用于研究流体运动中的湍流现象、热传导等问题。
五、结论大涡模拟是一种重要的流体动力学研究方法,Fluent 软件为大涡模拟提供了强大的计算支持。
动量格式是大涡模拟中描述流体运动规律的核心方程,通过求解动量格式,可以获得流体的速度、压力等物理量。
ANSYSFLUENT介绍
想起CFD,人们总会想起FLUENT,丰富的物理模型使其应用广泛,从机翼空气流动到熔炉燃烧,从鼓泡塔到玻璃制造,从血液流动到半导体生产,从洁净室到污水处理工厂的设计,另外软件强大的模拟能力还扩展了在旋转机械,气动噪声,内燃机和多相流系统等领域的应用。
今天,全球数以千计的公司得益于FLUENT的这一工程设计与分析软件,它在多物理场方面的模拟能力使其应用范围非常广泛,是目前功能最全的CFD软件。
FLUENT因其用户界面友好,算法健壮,新用户容易上手等优点一直在用户中有着良好的口碑。
长期以来,功能强大的模块,易用性和专业的技术支持所有这些因素使得FLUENT受到企业的青睐。
网格技术,数值技术,并行计算计算网格是任何CFD计算的核心,它通常把计算域划分为几千甚至几百万个单元,在单元上计算并存储求解变量,FLUENT使用非结构化网格技术,这就意味着可以有各种各样的网格单元:二维的四边形和三角形单元,三维的四面体核心单元、六面体核心单元、棱柱和多面体单元。
这些网格可以使用FLUENT的前处理软件GAMBIT自动生成,也可以选择在ICEM CFD工具中生成。
b5E2RGbCAP在目前的CFD市场, FLUENT以其在非结构网格的基础上提供丰富物理模型而著称,久经考验的数值算法和鲁棒性极好的求解器保证了计算结果的精度,新的NITA算法大大减少了求解瞬态问题的所需时间,成熟的并行计算能力适用于NT,Linux或Unix平台,而且既适用单机的多处理器又适用网络联接的多台机器。
动态加载平衡功能自动监测并分析并行性能,通过调整各处理器间的网格分配平衡各CPU的计算负载。
p1EanqFDPw湍流和噪声模型FLUENT的湍流模型一直处于商业CFD软件的前沿,它提供的丰富的湍流模型中有经常使用到的湍流模型、针对强旋流和各相异性流的雷诺应力模型等,随着计算机能力的显著提高,FLUENT已经将大涡模拟<LES)纳入其标准模块,并且开发了更加高效的分离涡模型<DES),FLUENT提供的壁面函数和加强壁面处理的方法可以很好地处理壁面附近的流动问题。
大涡模拟的fluent算例
Introduction:This tutorial demonstrates how to model the2D turbu-lentflow across a circular cylinder using LES(Large Eddy Simula-tion),and computeflow-induced noise(aero-noise)using FLUENT’s acoustics model.In this tutorial you will learn how to:•Perform2D Large Eddy Simulation(LES)•Set parameters for an aero-noise calculation•Save surface pressure data for an aero-noise calculation•Calculate aero-noise quantities•Postprocess an aero-noise solutionPrerequisites:This tutorial assumes that you are familiar with the menu structure in FLUENT,and that you have solved or read Tu-torial1.Some steps in the setup and solution procedure will not be shown explicitly.Problem Description:The problem considers turbulent airflow over a2D circular cylinder at a free stream velocity U of69.19m/s.The cylinder diameter D is1.9cm.The Reynolds number based on theflow parameters is about90000.The computational do-main(Figure3.0.1)extends5D upstream and20D downstream of the cylinder,and5D on both sides of it.If the computational domain is not taken wide enough on the downstream side,so that no reversedflow occurs,the accuracy of the aero-noise prediction may be affected.The rule of thumb is to take at least20D on the downstream side of the obstacle.c Fluent Inc.June20,20023-1Aero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylindernoise.msh.File−→Read−→Case...As FLUENT reads the gridfile,it will report its progress in the console window.2.Check the grid.Grid−→CheckFLUENT will perform various checks on the mesh and will report the progress in the console window.Pay particular attention to the reported minimum volume.Make sure this is a positive number.3.Scale the grid.Grid−→Scale...(a)Under Units Conversion,select cm in the Grid Was Created indrop-down list.(b)Click on Scale.4.Display the grid.Display−→Grid...(a)Display the grid with the default settings(Figure3.0.2).(b)Use the middle mouse button to zoom in on the image so youcan see the mesh near the cylinder(Figure3.0.3).Quadrilateral cells are used for this LES simulation becausethey generate less numerical diffusion than triangular cells.Cell size should also be small enough to make numerical dif-fusion much smaller than subgrid scale turbulence viscosity.Extra:You can use the right mouse button to check which zone number corresponds to each boundary.If you clickthe right mouse button on one of the boundaries in thegraphics window,its zone number,name,and type will beprinted in the FLUENT console window.This feature is c Fluent Inc.June20,20023-3Aero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylindernoise1.cas/dat).File−→Write−→Case&data...You can skip items9-12to avoid the time-consuming calculationsnecessary to get the“dynamically steady state”flowfield.Instead,you can read the corresponding case and datafiles(cylnoise1.cas/dat).See Chapter28of the User’s Guide for more information on using3-14c Fluent Inc.June20,2002Aero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylindernoise2.cas/dat).File−→Write−→Case&Data...Step7:Aero-Noise Calculation1.Save surface pressure variation data.(a)Set up the schemefile and user-defined functions(UDFs)foraero-noise calculation.i.Read the schemefile,normally located in the lib directory,to create the Acoustic-Parameters panel.File−→Read−→Scheme...ii.Select acousticAero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylindernoisenoise noisenoisenoise whole for the File Name to Read Surface Pressure.FLUENT’s aero-noise calculation module operates on asinglefile of surface pressure data at a time.If the surfacepressure data is saved in separatefiles,you may want toconcatenate them into one singlefile.3-18c Fluent Inc.June20,2002Aero-Noise Prediction of Flow Across a Circular Cylinderacousticpowerpower db.xy for the File Name to Power Spectrum in dB Unit.(c)Changefile name for the surface monitor.Solve−→Monitor−→Surface...i.Click on Define next to monitor-1ii.In the Define Surface Monitor panel,change the name of the monitor from monitor-point-behind-pres1-1.outto monitor-point-behind-pres4-1.out.(d)Save case and datafiles(cylnoise4.cas/dat).File−→Write−→Case&Data...(g)Exit FLUENTFile−→ExitIt is necessary to exit parallel FLUENT because the followingaero-noise calculation is performed with an Execute On De-mand UDF,which can only be used in the serial version ofthe solver.2.Calculate aero-noise(a)Start the serial version of FLUENT.c Fluent Inc.June20,20023-19Aero-Noise Prediction of Flow Across a Circular Cylinderpar.scm).File−→Read−→Scheme...(c)Read case and datafiles(cylnoise noise noise noise noise noisenoise whole.If you did not perform the calculation to write thefiles thatwill be used in this step,you can continue by using the corre-spondingfiles provided in the documentation CD.(e)Use the Execute On Demand UDF to perform the aero-noisecalculation.Define−→User-Defined−→Execute On Demand...(f)Select the cal-sound UDF and click Execute.Note:There is a limit to the minimum number of time steps ac-cording to the sound calculation scheme.The minimum num-ber of time steps needs to be larger than n=T/dt,where Tis the propagation time through a distance L,roughly equalto the length scale of the sound generating wall,and dt is thetime step size applied in the unsteady calculation.If the givennumber of time steps for cal-sound is smaller than the requiredminimum number,a warning will be printed on FLUENT’sconsole window,along with the indication of the minimumnumber<n>of time steps requiredWarning:Number of Time Steps of The Input Surface Data Must be Larger Than:<n>.3-20c Fluent Inc.June20,2002Aero-Noise Prediction of Flow Across a Circular Cylinder1.69e+021.52e+021.35e+021.19e+021.02e+028.49e+016.80e+015.12e+013.43e+011.75e+016.49e-01Figure3.0.7:Velocity Vectors2.Display contours of static pressure at the current time step(Fig-ure3.0.8).Display−→Contours...3.Inspect the Sound Pressure Level(SPL)value.The the value ofsound intensity in units of W/m2and its alternative expression in dB are printed in the FLUENT console window after the execution of the cal-sound UDF,and areIntensity=4.060634e+00(W/m2)SPL=1.261719e+02(dB)c Fluent Inc.June20,20023-21Aero-Noise Prediction of Flow Across a Circular Cylinder3.91e+031.78e+03-3.56e+02-2.49e+03-4.62e+03-6.75e+03-8.89e+03-1.10e+04-1.32e+04-1.53e+04-1.74e+04Figure3.0.8:Static Pressure Contours4.Plot Acoustic Pressure variation(Figure3.0.9).Plot−→File...(a)Click on Add.(b)Select thefile cyl pres.xy and click OK.Remember to delete thefiles you do not want to display from theFiles list.5.Plot Power Spectrum of sound pressure(Figure3.0.10).(a)Power Spectrum in units of P a2.Plot−→File...i.Click on Add.ii.Select thefile cyl spectrum.xy and click OK.Figure3.0.10shows a frequency range of0−2000Hz,withmajor and minor rules turned on.From thisfigure it can be 3-22c Fluent Inc.June20,2002Aero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylinderpower db.xy and click OK .Frequency (Hz)5.00e+016.00e+017.00e+018.00e+019.00e+011.00e+021.10e+021.20e+0201e+032e+033e+034e+035e+036e+037e+038e+039e+031e+04Power Spectrum (dB)Figure 3.0.11:Plot of Power Spectrum of Sound Pressure.Figure 3.0.11shows a frequency range of 0−10kHz .6.Inspect Surface Dipole Strength.(a)Display contours of Surface Dipole Strength on surface cylin-der (Figure 3.0.12).Display −→Contours...i.In the Contours Of drop-down lists,select User-DefinedMemory and udm-0.ii.Turn offNode Values .3-24cFluent Inc.June 20,2002Aero-Noise Prediction of Flow Across a Circular Cylinder4.13e+053.72e+053.31e+052.89e+052.48e+052.07e+051.65e+051.24e+058.25e+044.12e+04-1.94e+02Figure3.0.12:Contour of Surface Dipole Strengthiii.Click on Display.The value of Surface Dipole Strength for each cell face is storedfor the center of the face on the cylinder wall.Surface DipoleStrength is the distribution of unit area contribution on thesound generating surface to the intensity of sound measuredat the observer’s location.(b)Plot Surface Dipole Strength(udm-0)on surface cylinder(Fig-ure3.0.13).Plot−→XY Plot...Figure3.0.13shows Surface Dipole Strength distribution onboth the upper and lower half cylinder faces.Extra:Once theflow simulation reaches a“dynamically steady state”, the accuracy for predicting Sound Pressure Level(SPL)and Power Spectrum is usually dependent on the number of time steps used.LES requires a mesh size as small as the length scale of eddies in the inertial sub-range.The corresponding time step size is calcu-c Fluent Inc.June20,20023-25Aero-Noise Prediction of Flow Across a Circular CylindercylinderFigure3.0.13:Plot of Surface Dipole Strengthlated by dt=Cdx/U,where C is the Courant number,and thus isvery small compared with the period T of the dominating acousticwave component(i.e.that corresponding to the frequency of thehighest peak in the power spectrum).For an accurate aero-noiseprediction,at least10periods of the dominating wave componentare required for sampling.The number of time steps for this re-quirement can be roughly estimated for theflow over the cylinder.In a certain Reynolds number range(roughly Re<50000),theStrouhal number(St=fD/U)for the dominating frequency f isabout0.2.Therefore,the period is T=D/0.2/U.From the aboveequations,the number of time steps for each period can be calcu-lated as N=T/dt=5/CD/dx.In LES,the ratio between thedomain scale D and the typical cell size dx can easily be50-100.As an example,if C is taken as order of1,N can be as high as250-500for each period.For40periods,10000-20000time stepsmay be required.Summary:This tutorial demonstrated how to set up and calculate an aero-noise problem for theflow around a cylinder,using the2D LES 3-26c Fluent Inc.June20,2002Aero-Noise Prediction of Flow Across a Circular CylinderAero-Noise Prediction of Flow Across a Circular Cylinder。
openfoam 的大涡模拟算例
openfoam 的大涡模拟算例OpenFOAM是一个开源的计算流体动力学(CFD)软件,它包含了许多不同类型的模拟算例,其中包括大涡模拟(Large Eddy Simulation,简称LES)。
大涡模拟是一种在CFD中用于模拟流体流动的高级方法,特别适用于湍流流动的模拟。
在OpenFOAM中进行大涡模拟的算例可以涉及各种不同的流体流动情况,比如湍流流动在不同的几何形状中的行为、湍流与燃烧的相互作用等等。
下面我将从不同的角度来说明OpenFOAM中大涡模拟的算例。
首先,从几何形状的角度来看,OpenFOAM中的大涡模拟算例可以涉及不同类型的几何形状,比如圆柱、方柱、翼型等等。
这些几何形状对于大涡模拟的算例来说都有不同的影响,因此OpenFOAM提供了针对不同几何形状的大涡模拟算例,以便工程师和研究人员能够针对特定的应用进行模拟研究。
其次,从物理现象的角度来看,OpenFOAM中的大涡模拟算例可以涉及不同的流体流动现象,比如湍流边界层、湍流绕流、湍流与燃烧等等。
这些不同的物理现象需要不同的数值方法和模型来进行模拟,因此OpenFOAM提供了针对不同物理现象的大涡模拟算例,以便用户能够根据自己的研究需要选择合适的算例进行模拟。
此外,从数值方法和模型的角度来看,OpenFOAM中的大涡模拟算例可以涉及不同的数值方法和模型,比如不同的离散化格式、不同的湍流模型等等。
这些数值方法和模型的选择对于大涡模拟的准确性和计算效率都有重要影响,因此OpenFOAM提供了针对不同数值方法和模型的大涡模拟算例,以便用户能够根据自己的需求选择合适的方法和模型进行模拟研究。
综上所述,OpenFOAM中的大涡模拟算例涉及了多个方面,包括几何形状、物理现象、数值方法和模型等等。
用户可以根据自己的研究需求选择合适的算例进行模拟研究,以便更好地理解和分析流体流动中的湍流现象。
湍流的模拟和建模
湍流的模拟和建模湍流是自然界中普遍存在的现象,其涵盖的规模从大气层中的云团到船舶和管道中的流体,十分广泛。
湍流现象表现为流体的不规则而混乱的流动,其中的旋涡和涡旋不断形成和消失。
湍流的复杂性和不可预测性使其对于物理学家和工程师来说是非常具有挑战性的问题。
然而,通过数字模拟和建模,我们可以更好地理解和控制湍流现象,进而提高生产和人类生活的质量。
湍流的模拟和建模一直是流体力学领域的研究热点,旨在通过计算机模拟来预测复杂流动中的物理性质。
对于湍流的模拟,目前主要有两类方法:直接数值模拟(Direct Numerical Simulation, DNS)和大涡模拟(Large-Eddy Simulation, LES)。
其中DNS方法对于湍流的描述最为详细,可以剖析流场中的每一处涡旋,但计算成本极高,通常只适用于小规模的问题。
LES方法通过简化较小尺度的湍流结构来减少计算量,虽然无法完全描述每个涡旋,但是在较大的尺度下仍能准确预测湍流的行为。
湍流现象的建模通常可以基于Navier-Stokes方程进行,这是一组描述流体本质的偏微分方程。
针对这些方程的求解方法和算法不断更新和优化,使得模拟计算变得更加高效和准确。
其中著名的流体力学软件包,包括ANSYS Fluent、OpenFOAM等已经成为工业和研究界广泛应用的工具。
当然,与模拟和建模相伴的,是精度和计算成本之间的取舍。
对于湍流现象的模拟通常需要对涡旋的尺度、湍流能量转化等参数进行详细定量的计算,因此准确度成为了模拟中一个重要的考量因素。
在确定准确度之余,如何减少计算成本也是一个必须解决的问题。
因此,研究人员通常采用增加计算资源的方式,如改进集群计算机和高性能计算机的配置来提升计算速度,并利用一些优化算法和计算技巧来控制误差和减少计算成本。
在湍流模拟和建模方面,模型验证也是一个很重要的步骤,这也是模拟不能完全取代实验的原因之一。
验证过程通常会与实验数据进行比对,用实验数据的帮助来验证模型的准确性。
大涡模拟简单介绍
大涡模拟简单介绍大涡模拟(Large Eddy Simulation,简称LES)是一种流体动力学数值模拟方法,用于模拟湍流流动。
相比于传统的雷诺平均Navier-Stokes方程(RANS)模拟方法,LES可以更准确地捕捉流动中的湍流结构和湍流涡旋,并且消除了能量储存和耗散的子网格模型假设。
LES的基本原理是在Navier-Stokes方程的基础上,通过滤波器将流动变量划分为长时间和空间尺度下的平均分量和湍流分量。
经过充分滤波的方程组被认为是LES方程组,其中长时间和空间尺度下的平均分量由RANS求解,湍流分量则采用直接数值模拟(DNS)或者更为常见的子网格模型进行近似。
LES方程组通常采用基于物理的平滑学习系数(Smagorinsky模型)或者基于数值的子网格尺度计算方法来估计湍流涡旋的剪切应力。
与传统的RANS模拟相比,LES能够提供更多细节的湍流结构信息,从而更好地预测湍流流动中的流场特性,比如涡旋结构、湍流能量传递、湍流耗散等。
这些信息对于工程问题的分析和设计有着重要的意义,比如风力发电机翼型的气动性能、船舶外形的水动力性能等。
LES的优势主要体现在以下几个方面:1.湍流结构预测能力:LES可以更准确地模拟湍流结构,包括涡旋的生成、演化和消散过程,因此能够提供更详尽的湍流流场信息。
2.湍流能量传递和耗散特性:LES能够有效地预测湍流能量的传递和耗散特性,对于评估流动中的湍流耗散和能量损失有着重要的意义。
3.均匀流动和非均匀流动的统一模拟:与传统的RANS方法相比,LES对均匀流动和非均匀流动有着较好的统一模拟能力。
对于非均匀流动,LES能够更好地预测局部湍流结构的分布和演化。
4.对涡旋缩放和旋转的准确模拟:LES能够模拟涡旋的缩放和旋转过程,能够提供更真实的细节湍流结构信息。
尽管LES在提供细节湍流结构信息方面具有优势,但其计算成本较高,主要体现在网格分辨率和时间步长上。
由于需要考虑到湍流结构的空间和时间变化,LES所需的网格分辨率通常较高,这对计算资源的要求较高。
- 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
第13卷第6期船舶力学Vol.13No.6 2009年12月Journal of Ship Mechanics Dec.2009 Article ID:1007-7294(2009)06-0990-12LES Method for Investigation of Noise Generated byTurbulent Boundary LayerPAN Yu-cun1,2,ZHANG Huai-xin1(1State Key Laboratory of Ocean Engineering,Shanghai Jiao Tong University,Shanghai200030,China;2Department of Naval Architecture and Ocean Engineering,Naval University of Engineering,Wuhan430033,China)Abstract:Large Eddy Simulation(LES)was used to investigate space-time field of the low Mach number,fully developed turbulent boundary layer on a smooth,rigid flat plate.The flowfield simu-lated by LES was taken as near-field sound sources and radiated sound from turbulent boundary layer(TBL)was studied using Lighthill’s acoustics analogy.The radiated sound of dipole type,of quadrupole type,and of the sum of the two above were paring their power spectral densities,it is concluded that in low Mach number,the wall shear stress(dipole type of source)is the predominant factor responsible for sound radiation from TBL.Key words:turbulent boundary layer;hydroacoustics;large eddy simulation;Lighthill’s acoustic analogyCLC number:O427.4Document code:A1IntroductionThe sound generated by turbulence is an important part of flow-generated noise.The pas-sengers in aircrafts and automobiles always suffer from it.In the past,the jet noise was be-lieved to be dominant in aircraft noise,and the turbulent noise was considered secondary. However,accompanying the progress in jet-engine construction,material and design to con-trol jet-noise over recent years,the airframe turbulence-noise problem has received an in-creasing amount of attention.In classical acoustic problems,the sound wave was induced by the mechanical vibration of particular object.The fluid was only the medium of acoustic wave propagation.However,the objective of this study is to investigate the sound induced by the turbulence of fluid medium itself.The sound radiated from the low Mach number,fully developed turbulent boundary layer on a smooth,rigid flat plate is computed in this paper.The mechanism of hydrodynamic noise generation by turbulent flow has been a very interesting subject for fluid researches since Lighthill[1-2]laid the foundations to a general theory.Initially,the solution of the Lighthill equation was attained in suppose of free space.In Received date:2009-03-03Foundation item:Supported by the National Natural Science Foundation of China(Grant No.10772119)and theResearch Fund of State Key Laboratory of Ocean Engineering,Shanghai Jiao Tong University.Biography:PAN Yu-cun(1980-),male,Ph.D.student of Shanghai Jiao Tong University.第6期PAN Yu-cun et al:LES Method for Investigation of (991)the flow without solid boundary such as jet flow or free shear flow,the quadrupole source was the only sound source and the radiation efficiency was very low.Curle[3]extended Lighthill the-ory to the flow with stationary,solid boundary using Kirchhoff method.The result shown that the rigid boundary was equivalent to distribution of dipole sources,which enhanced the radia-tion efficiency and changed the components of flow noise.In this study,the Lighthill’s acoustic analogy method was used to investigate the hydro-dynamic noise radiation.Lighthill pointed out that,in a low Mach number turbulent flow,the acoustic and turbulent fields are only weakly coupled so that the turbulent fluctuations,which are not influenced by the acoustic disturbances,act as sources of sound.In this study,the hy-brid approach is used,in which the hydrodynamic and acoustic fields are decoupled.In the near-field,the hydrodynamic velocity and pressure are computed,which describe the nonlin-ear process of sound generation.And the near-field was also called source field.The hydro-dynamic terms are used to calculate the far-field acoustic field,which describe the linear propagation of sound wave.Ffowcs Williams[4]pointed out that for sound generated by low Mach number flow,the influence of compressible could be neglected.So,in computation of fluid in near field,we can investigate it based on the equation of incompressible flow.The fully developed turbulent boundary layer on a smooth,rigid flat plate is a typical tur-bulent shear flow.In the vicinity of the plate,the characteristic of wall shear turbulence is ob-vious;however,at the region away the wall,the characteristic of free shear turbulence is dom-inant.There are violent vortices in the wake behind blunt body such as sphere and circular cylinder,but no such vortex in the turbulent boundary layer in this study.So,vortex-shedding is not the main source of noise;this study is concerned with sound radiation from the tiny ed-dies in the vicinity of wall within the turbulent boundary layer.In this study,we should obtain the hydrodynamic terms of turbulent fluctuations,so,the Reynolds-averaged Navier-Stokes equations(RANS)method couldn’t be used here,in which the governing equations are time-averaged.The direct numerical simulations(DNS)can pro-vide the most precise,detailed fluctuating information,but the present computational resource can not support it to simulate high Reynolds-numbers flow.The information obtained by Large Eddy Simulation(LES)is no as exact as that obtained by DNS,but it can also reveal the ele-mentary mechanism of turbulent flow.Moreover,the computational cost of a LES is much low-er than that of a DNS.So,LES method was used in this study.The paper was divided into essentially two sections.In Sec1the unsteady incompressible Navier-stokes equations were solved numerically using LES method to give an approximate description of the space-time field of the low Mach number,fully developed turbulent bound-ary layer on a smooth,rigid flat plate;In Sec2the far-field radiation of sound was calculated based on Curle’s expansion to the Lighthill acoustic analogy.2Numerical simulation of turbulent boundary layerFig.1shows the computational domain of turbulent boundary layer.The Reynolds number,defined as Re=u τδ/ν=800,(u τis the wallshear velocity,δis the thickness of theboundary layer).Based on the theory ofLES,the filtering operation was applied tothe Navier -Stokes equations,and theSmagorinsky ’s [5]model was chosen to pa -rameterize the subgrid scale (SGS)stress -es.In order to account for low-Reynolds-number SGS turbulence near the wall,theVan Driest [6]exponential damping function was applied to the eddy viscosity νt ,so that it could be damped to zero at the wall.The computation result shows the space-time field of fluctuat -ing turbulent fluid motion near the wall.The more detailed computation process and result can be seen in Ref.[7].3Flow-induced noise3.1Lighthill ’s acoustic analogyThe density fluctuation due to acoustic wave propagation from the hydrodynamic source region is governed by the convected wave equation:坠坠t +U ∞坠坠x 1∞∞2-c 02坠2坠x j 坠x j ∞∞ρ′=坠2T i j 坠x i 坠x j (1)where,ρ′=ρ-ρ∞is the density fluctuation,ρ∞is the free-stream density,c 0is the sound speed in water at undisturbed conditions.We assume the free-stream velocity is U ∞,and T i j =ρv i v j +δi j p -c 02ρ-ρ∞∞∞∞∞-τi j is the Lighthill stress tensor defined in terms of the fluctuating velocity relative to the free-stream value,v i =u i -U ∞δi 1.Here,the usual summation convention applies for repeated subscripts,andδi j =0i ≠∞∞j 1i=∞∞j ≠is the Kronecker delta;τi j =μ坠v i j +坠v j i -2δi j 坠v k k ∞∞is the viscous part of Stokes stress tensor.Strictly speaking,it is almost impossible to resolve the equation (1)exactly.The hydro -dynamic and acoustic terms are coupled each other,flow can induce sound and sound can be scattered by flow.To get the source item on the right-hand side of equation (1),the nonlinear equation should be solved strictly,which was a hard work.In order to study complex engineer -ing applications,we should make some simplifying assumptions:according to the practical flow condition,some components of stress tensor are so small that can be neglected;in low Mach number flow,the two-way coupling is neglected,that is,the acoustic terms do not affect the Fig.1Computational domain of the turbulent boundary layer 992船舶力学第13卷第6期hydrodynamic terms.Thus,we can consider the source terms on the right-hand side of equa -tion (1)only contain the influence of flow to noise and the source terms can be obtained from the flow field.If solid body exists in the sound source region T i j ≠≠≠0and the sound is generated by the solid surfaces,Curle derived a simple solution for noise produced by the rigid surface mov -ing through a quiescent medium.ρx 軆,軆軆t -ρ0=-14πc 02坠坠x i S 乙n j p i j y 軆,t-r 軆0軆軆r 軆d 2y 軆+14πc 02坠2坠x i x j V 乙T i j y 軆,t-r 軆0軆軆r 軆d 3y 軆(2)where x軆and y 軆represent the position vectors of the observer and the sound sources,respective -ly,r =r 軆=x 軆-y 軆;r 軆0is called the delay time and represents the time interval of sound wave traveling from the sound sources to the observer.In the preceding equation p i j =p δi j -τi j and n j is the directional cosine of the outward normal (into the fluid)to the rigid surface S over which the surface integration take place.The volume integral is taken over the entire unsteady flow region V external to the body.In the acoustic far field defined by r 軆>>l e /M ,where l e is the typical eddy size,M=U ∞/c 0is the freestream Mach number,equation (2)is simplified to a form most suitable for numeri -cal evaluation.ρx 軆,軆軆t -ρ∞≈14πc 03坠坠tS 乙r i r 軆2n j p i j y 軆,t-r 軆c 0軆≠d 2y 軆+14πc 04坠2坠t 2V 乙r i r j r軆3T i j y 軆,t-r 軆0軆≠d 3y 軆(3)In the computation,the spatial coordinates x i are non-dimensionalized by a characteristic length L ,and time t is non-dimensionalized by L /U ∞.Also,v i are the components of the fluidvelocity,non-dimensionalized by U ∞,ρthe fluid density,non-dimensionalized by the refer -ence value ρ∞,p the pressure non-dimensionalized by ρ∞U ∞2.Equation (3)can be rewritten asρx 軆,軆≠t -1≈M 34π坠坠tS 乙r i r 軆2n j p i j y 軆,t-M r 軆軆≠d 2y 軆+M 44π坠2坠t 2V 乙r i r j r軆3T i j y 軆,t-M r 軆軆≠d 3y 軆(4)Furthermore,if both the body and the unsteady flow region are small compared with the typical acoustic wavelength l e /M ,the source region is acoustically compact.The far-field den-第6期PAN Yu-cun et al:LES Method for Investigation of (993)sity can be approximated byρx 軆,軆軆t -1=M 3x i x D 觶i t-M x 軆軆+M 4x i x j x Q 咬i j t-M x 軆軆(5)whereD 觶i 軆軆t =坠坠tS 乙n j p i j y 軆,軆軆t d 2y (6)Q 咬i j 軆軆t =坠2坠t V 乙T i j y 軆,軆軆t d 3y (7)In equation (5),there are two kinds of sound source:dipole source D 觶iand quadrupole source Q 咬i j,which represent the compact surface and volume sound radiation,respectively.The pressure fluctuations and viscous shear tensors on the wall boundary generate sound radiation of dipole type;the Lighthill stress tensors behave as quadrupole.3.2Computation conditionTo predict the far-field sound radiation,we assumethere are symmetrical flow field at the both sides of the flat plate.And the center of theplate was chosen to be the origin of coordinates,the vector of the observation point is chosen to be(2m,5m,2m).In this study,the lengths of three di -rections of computational box are chosen to be:8δ×2δ×8乙乙δ=0.1941264m ×0.0485316m 乙×0.1941264乙m (length of streamwise -L 1×length of spanwise -L 2×length of wall-normal direction -L 3).The time-advanced step is chosen to be 1δτ.To satisfy the sampling precision and the resolu -tion,the computation continued for 50000steps af -ter the time-averaged statistics had converged.Re=u τδ/ν=800.The Mach number M =U ∞/c 0=0.00054947(the sound speed c 0=1500m/s).3.3The dipole and quadrupole sources There are three components of dipole source:D 觶1,D 觶2,D 觶3.According to the assumption of symmetrical flow field on both side of the flat plate,we can deduce that the D 觶2=0.There are nine components of quadrupole source:Q 咬11,Q 咬12,Q 咬21,Q 咬13,Q 咬31,Q 咬22,Q 咬23,Q 咬32,Q 咬33,in which three pairs are symmetrical:Q 咬12and Q 咬21,Q 咬13and Q 咬31,Q 咬23and Q 咬32.So,only six quadrupole terms should be solved respectively.Fig.2The computational domain of the far-field sound radiation994船舶力学第13卷第6期(a)Variation of Q咬11t=2.5~7!"s(b)Variation of Q咬11t=4~4.05!"sFig.3Variation of quadrupole Q咬11(a)Variation of Q咬12t=2.5~7!"s(b)Variation of Q咬12t=4~4.05!"sFig.4Variation of quadrupole Q咬12(a)Variation of Q咬13t=2.5~7!"s(b)Variation of Q咬13t=4~4.05!"sFig.5Variation of quadrupole Q咬13Figs.3~8show the variation of quadrupole sources,(a)show a course of long time(t= 2.5s~7s);(b)show the variation in a short course(t=4s~4.05s).We can find the amplitudes ofthe quadrupole Q咬11,Q咬22,Q咬33are relatively larger,and Q咬11is the largest,which shows the Lighthill第6期PAN Yu-cun et al:LES Method for Investigation of…995stress tensor in streamwise is the largest.In Figs.9and 10,the variation of dipole sources D 觶1,D 觶3was shown.To observe the variation of quadrupole and dipole source in frequency domain,Figs.11~16show the auto-power spectral density of quadrupole sources,Figs.17and 18show the auto-powerspectral density ofdipole sources.(a)Variation of Q 咬22t =2.5~7!"s (b)Variation of Q 咬22t =4~4.05!"s Fig.6Variation of quadrupole Q 咬22(a)Variation of Q 咬23t =2.5~7!"s (b)Variation of Q 咬23t =4~4.05!"s Fig.7Variation of quadrupole Q 咬23(a)Variation of Q 咬33t =2.5~7!"s (b)Variation of Q 咬33t =4~4.05!"s Fig.8Variation of quadrupole Q 咬33996船舶力学第13卷第6期Fig.9Variation ofdipole D 觶1Fig.10Variation of dipole D 觶3Fig.11Power spectral density of quadrupoleFig.12Power spectral density of quadrupole source Q 咬11source Q 咬12Fig.13Power spectral density of quadrupoleFig.14Power spectral density of quadrupole source Q 咬13source Q 咬22Fig.15Power spectral density of quadrupoleFig.16Power spectral density of quadrupole source Q 咬23source Q 咬33第6期PAN Yu-cun et al:LES Method for Investigation of (997)Fig.17Power spectral density of dipoleFig.18Power spectral density of dipole source D 觶1source D 觶3In the above figures,we can find the frequency of quadrupole fluctuations is relativelyhigher.All the quadrupole sources except Q 咬13concentrate their energy in range of 1200~2500Hz.The frequency of dipole fluctuations is relatively lower,and the energy is concentrat -ed in the domain of low frequency (under 500Hz).The reason may be that the dipole fluctua -tions were mainly caused by the fluctuations of viscous shear stress near the wall.As we known,in the vicinity of the walls,the high -speed fluid elements correspond to the sweep event toward wall;the low-speed fluid elements are generally being ejected from the wall re -gions.The viscous shear stress intensively affected by the sweep and ejection events,and these events have low frequency and large amplitude.So,the dipole source fluctuation is mainly in low frequencies.And the quadrupole sources associated with the second-order time derivative of Lighthill stress tensor.The Lighthill stress T i j ′≈ρu i ′u j ′is the product of velocities of differ -ent directions.The frequency velocity of fluctuations is relatively high,thus,the fluctuation of quadrupole source is in the domain of high frequencies.3.4Sound radiationFrom Figs.11to Fig.18,the sum of dipole D 觶i is smaller than the sum of quadrupole Q 咬i j.But according to the equation (5),acoustic densities due to the contribution of quadrupole sources at far-field is proportional to the fourth power of the Mach number,whereas that due to the contribution of dipole sources is proportional to the M 3.So,for the low Mach numbers,the con -tribution of the shear-stress dipole sources to the sound generation is larger than that of the quadrupole sources.In Fig.19,the variation of acoustic density due to the contribution of dipole and quadrupole sources were shown respectively.Fig.20shows the variation of acoustic density due to the contribution by the summation of dipole and quadrupole sources.From these fig -ures,we can conclude that for Mach number M =0.00054947,the turbulent shear-stress dipole sources dominates the whole acoustic field,and the contribution of quadrupole sources is al -most negligible.It is observed that,the characteristic of acoustic field behaves as the charac -teristic of dipole sources.According to the Lighthill analogy,the acoustic wave propagation was governed by a lin -ear,inhomogeneous wave equation of a nearly incompressible fluid.So,we can obtain the re-998船舶力学第13卷第6期lationship between acoustic pressure fluctuation p ′and acoustic density fluctuation ρ′:p ′=p-p ∞=c 02ρ′=c 02ρ-ρ∞p p .Thus,after we obtain the acoustic density fluctuation ρ′by Equation (5),we can research the variation of the acoustic pressure fluctuation p ′.To observe the characteristic of acoustic pressure in frequency domain,we obtain the spectral density of it by Fourier trans -formation:准m p p ω=1π∞-∞乙R pppp τe -i ωτd τwhere,R pp pp τ=〈p ′pp t p ′t+p p τ〉is the auto-correlation function of acoustic pressure fluctua -tions.Fig.21Power spectral density of sound pressureFig.22Power spectral density of sound pressure induced by dipole source induced by quadrupole sourceFig.21shows the power spectral density of sound pressure induced by dipole sources.For frequencies under 500Hz,the power spectral density drops sharply with increasing fre -quency.When the frequency is over 500Hz,the power spectral density tends toward calming and the amplitude of its fluctuation is small.So,the energy is concentrated in the domain of low frequencies.Fig.22shows the power spectral density of sound pressure induced by quadrupole sources.For frequencies under 180Hz,the power spectral density drops with in-Fig.19Variation of acoustic density at theFig.20Variation of acoustic density at the far-field due to the quadrupolefar-field (summation of Quadrupole (———)and dipole (---)and dipole)第6期PAN Yu-cun et al:LES Method for Investigation of (999)creasing frequency.When the frequency isover 180Hz,the power spectral density increases with increasing frequency andreaches its peak at 2300Hz.Over 2300Hz,the power spectral density drops again.So,the energy is concentrated in the domainbetween 1000to 2500Hz.Fig.23showsthe power spectral density of sound pres -sure induced by the summation of dipole and quadrupole sources.This curve is very similar to that of the power spectral densityof sound pressure induced by dipole sources in Fig.21.We can say that the sound pressure in -duced by dipole sources dominates the power spectral density of far-field sound radiation.4ConclusionsLarge Eddy Simulation was used to investigate space-time field of the low Mach number,fully developed turbulent boundary layer on a smooth,rigid flat plate.The radiated sound from turbulent boundary layer was studied using Lighthill ’s acoustics analogy.It is observed that the energy of the computed far-field sound radiation is mainly concentrated in the domain of low frequency (under 500Hz).In addition,the sound level is so small that only precision in -strument could measure it.Thus,It was difficult in laboratory to distinguish between the noise from TBL and the background noise.Numerical method would not introduce background noise and provide us a choice to resolve the problem.There were many different assumptions about mechanism of acoustic sources in the tur -bulent boundary layer.Intensity of quadrupole sources associate with the fluctuation of Lighthill stress tensor,and intensity of dipole source depend on the fluctuation of wall shear stress.There are still debates among scholars about which is the predominant factor responsible for sound radiation from TBL:the dipole sources of wall shear stress fluctuations or the quadrupole source of turbulent Reynolds stress?In this study,the computational results show:for low Mach number,the power spectral of sound radiation induced by dipole source determines the power spectral of whole sound radiation.The dipole source dominates over the sound radiation field.The contribution of quadrupole source to the sound field can be neglected in some degree.References[1]Lighthill M J.On sound generated aerodynamically I.Gerneral theory[J].Proc Royal Soc London,Ser A,1952,211(1107):564-587.[2]Lighthill M J.On sound generated aerodynamically II.Turbulence as a source of sound[J].Proc Royal Soc London,Ser A,1954,222:1-25.Fig.23Power spectral density of sound pressure induced by the summation of dipole andquadrupole source 1000船舶力学第13卷第6期第6期PAN Yu-cun et al:LES Method for Investigation of (1001)[3]Curle N.The influence of solid boundaries upon aerodynamic sound[J].Proc Royal Soc London,Ser A,1955,231(1187):505-514.[4]Ffowcs Williams J E,Hawkings D L.Sound generation by turbulence and surface in arbitrary motion[J].Phil.Trans.Roy.Soc.,1969,264A:321-342.[5]Smagorinsky J.General circulation experiments with the primitive equations[J].Monthly Weather Review,1963,91(3):99-164.[6]Van Driest E R.On turbulent flow near a wall[J].J Aero.Sci.,1956,23:1007-1011.[7]Pan Yucun,Zhang Huaixin.Investigation of wall pressure fluctuations beneath turbulent boundary layer by LES method[J].Journal of ship Mechanics,2010,14(1).(En Chinese)[8]Landahl M T.Wave mechanics of boundary layer turbulent and noise[J].J.A.S.A,1975,57(4):824-831.[9]Lauchle G C.Noise generated by exisymmetric turbulent boundary layer flow[J].J.A.S.A,1977,61(3):694-703.[10]Hardin J C.Acoustic sources in the low Mach number turbulent boundary layer[J].J.A.S.A,1991,90(2):1020-1031.[11]Howe M S.Surface pressures and sound produced by turbulent flow over smooth and rough walls[J].J.A.S.A,1991,95:1041-1047.用大涡模拟方法研究湍流边界层流动噪声潘雨村1,2,张怀新1(1上海交通大学海洋工程国家重点实验室,上海200030;2海军工程大学船舶与海洋工程系,武汉430033)摘要:利用大涡模拟方法,对刚性光滑平板上充分发展的低马赫数湍流边界层流动进行了数值模拟。