用大涡模拟方法研究湍流边界层流动噪声

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.在此基础上将时一空变化的湍流流场信息作为近场声源,运用Lishthill的声学类比理论计算了边界层辐射噪声.文中讨论了偶极子、四极子对噪声的贡献,通过对比它们的功率谱密度,认为壁面剪切应力(偶极子)是湍流边界层辐射噪声的主要来源.     

翼型尾流场湍流特征的水筒PIV测试分析研究

翼型尾流场是一个典型的湍流场.文章在CSSRC空泡水筒中采用PIV技术对翼型尾流场进行了大子样试验测量,获得翼型尾流场的时均速度、雷诺应力、脉动速度均方根、湍流强度和平均涡量等统计特征量,并将试验结果与美国Notre Dame大学风洞试验结果作了比较,吻合较好.文中还利用PIV测量瞬时速度场计算了脉动速度空间相关系数和湍流积分尺度,分析了翼型尾流场湍流积分尺度与湍流强度的分布特征及湍流演化规律.     

湍流激励下简支平板振动特性的解析法研究

解析法基于随机激励理论,采用湍流脉动压力的Corcos模型作为输入,结合考虑流固耦合时平板的频响函数,求解得到平板振动速度的自功率谱密度。计算结果同已有文献吻合较好,证明了解析法的正确性。同时研究结果表明：湍流激励下平板的振动特性体现了平板本身的固有特性;随着来流速度的增大,平板振速的均方谱密度幅值按一定的数值关系增大。     

平板壁面湍流脉动压力及其波数-频率谱的大涡模拟计算分析研究

壁面湍流脉动压力是重要的流噪声声源,对壁面湍流脉动压力及其波数-频率谱进行数值计算是流声耦合领域的重要课题,开展相应的研究十分必要。文章采用大涡模拟方法（LES）结合动态亚格子涡模型（DSL）与千万量级的精细网格,对平板壁面湍流脉动压力及其波数-频率谱进行了数值计算,并与试验结果进行了对比分析,验证了数值计算方法的可靠性。首先,介绍了大涡模拟的物理内涵与基本方程,给出了所采用亚格子涡模型的表达式。其次,描述了Abraham试验中矩形试验段的几何特征,给出了网格的剖分形式,并给出了相应的离散求解数值方法以及边界条件的设置。再次,探讨了湍流脉动压力变化规律及其相似律,基于Fourier变换计算得到了湍流脉动压力波数-频率谱,并详细讨论了壁面湍流脉动压力及其波数-频率谱计算值与试验值之间的差异,进行了定量与定性的验证分析,结果表明,计算结果与试验结果吻合良好,计算方法合理可靠,为今后复杂几何模型壁面湍流脉动压力及其波数-频率谱的计算研究工作奠定了基础。最后,基于试验和计算结果,比较分析了常用波数-频率谱理论模型,为波数-频率谱的工程应用提供了参考。     

自由来流角度影响下壁面湍流脉动压力波数—频率谱的大涡模拟计算分析研究

壁面湍流脉动压力是重要的流噪声声源,对壁面湍流脉动压力及其波数—频率谱进行数值计算是流声耦合领域的重要课题。文章在已有工作的基础上,采用大涡模拟方法(LES)结合动态亚格子涡模型(DSL)与千万量级的精细网格,对不同自由来流角度影响下壁面湍流脉动压力及其波数—频率谱进行了数值计算与分析。首先,介绍了大涡模拟基本方法,包括:大涡模拟的物理内涵、基本方程以及所采用亚格子涡模型的表达式。其次,介绍了湍流脉动压力波数—频率谱及其计算与分析方法。再次,对不同自由来流角度情况下的湍流脉动压力及其波数—频率谱进行了计算,并将计算结果进行了比较分析,深入讨论了自由来流角度对湍流脉动压力及其波数—频率谱的影响。结果表明,在自由来流角度影响下,湍流脉动压力及其波数—频率谱主要参数(包括波数—频率谱的谱级峰值、迁移脊在波数—频率域内的分布范围、迁移速度和无量纲迁移速度等)均发生了明显变化,说明自由来流角度对湍流脉动压力波数—频率谱有显著影响,且边界层内湍流脉动压力的能量主要沿流向分布。因此,为了更加准确可靠地研究边界层内湍流脉动压力的主要统计特性及其波数—频率谱,传感器阵列或监测点阵列布置方向应与当地流向(局部剪应力线或摩擦力线)一致。     

基于单向求解、双向流固耦合的悬臂平板绕流涡激振动特性对比研究

本文采用双向流固耦合研究一端固定的二维平板(展向无限长)涡激振动特性,通过流场和结构之间力和位移的相互传递来实现双向耦合。同时,对不同雷诺数下的平板绕流涡激振动进行了计算,并与单向求解算法进行了对比。计算结果表明,当涡脱落频率接近结构固有频率时,将出现频率锁定现象,同时结构将产生极大的位移响应。在锁定雷诺数区域中,双向流固耦合结构位移响应明显小于单向求解算法,而在其他非锁定区间,两种算法差别较小。     

Simple Lagrangian formulation of bubbly flow in a turbulent boundary layer (bubbly boundary layer flow)

For the theoretical consideration of a system for reducing skin friction, a mathematical model was derived to represent, in a two-phase field, the effect on skin friction of the injection of micro air bubbles into the turbulent boundary layer of a liquid stream. Based on the Lagrangian method, the equation of motion governing a single bubble was derived. The random motion of bubbles in a field initially devoid of bubbles was then traced in three dimensions to estimate void fraction distributions across sections of the flow channel, and to determine local bubble behavior. The liquid phase was modeled on the principle of mixing length. Assuming that the force exerted on the liquid phase was equal to the fluid drag generated by bubble slip, an equation was derived to express the reduction in turbulent shear stress. Corroborating experimental data were obtained from tests using a cavitation tunnel equipped with a slit in the ceiling from which bubbly water was injected. The measurement data provided qualitative substantiation of the trend shown by the calculated results with regard to the skin friction ratio between cases with and without bubble injection as function of the distance downstream from the point of bubble injection.List of symbols B law of wall constant - C _f local coefficient of skin friction - C _f0 local coefficient of skin friction in the absence of bubbles - d _b bubble diameter [m] - g acceleration of gravity [m/s²] - k ₁ k₄ proportional coefficient - k _L turbulent energy of the liquid phase [m²/s²] - L representative length [m] - l _b mean free path of a bubble [m] - m _A added mass of a single bubble [kg] - m _b mass of a single bubble [kg] - N _x ,N _y ,N _z force perpendicular to the wall or ceiling exerted on a bubble adhering to that wall or ceiling [N] - P absolute pressure [Pa] - Q _G rate of air supply [/min] - q _L ⁽ⁱ⁾ turbulent velocity at the ith time increment [m/s] - R> _ex Reynolds number defined by Eq. 32 - T _*L integral time scale of the liquid phase [s] - U velocity of the main stream [m/s] - ,¯v,¯w time-averaged velocity components [m/s] - u,v,w turbulent velocity components [m/s] - û _L ,v_L root mean square values of liquid phase turbulence components in thex- and y-directions [m/s] - V volume of a single bubble [m³] - X,Y,Z components of bubble displacement [m] - x _s ,y _s ,z _s coordinate of a random point on a sphere of unit diameter centered at the coordinate origin - root mean square of bubble displacement in they-direction in reference to the turbulent liquid phase velocity [m] - local void fraction - _m mean void fraction in a turbulent region - regular random number - R _v increment of the horizontal component of the force acting on a single bubble, defined by Eq. 22 [N] - t time increment [s] - ₁ reduction of turbulent stress [N/m²] - _L rate of liquid energy dissipation [m²/s³] - _m coefficient defined by Eq. 30 - law of wall constant in the turbulent region in absence of bubbles - ₁ law of wall constant in the turbulent region in presence of bubbles     

基于AutoSEA2的声呐自噪声水动力分量分析

利用基于统计能量分析法的声仿真软件AutoSEA2分析湍流边界层激励下水下航行器声呐腔自噪声水动力分量。采用一种新的回转体模型模拟声呐罩,重点讨论了空间分布不均匀的湍流边界层对声呐罩的输入功率的计算。利用Fluent软件计算边界层的分离点及一些重要参数。分析结果可作为空间不均匀湍流边界层激励下声呐腔自噪声工程估算的参考。     

Effect of microbubbles on the structure of turbulence in a turbulent boundary layer

The time-averaged velocity and turbulence intensity distributions were measured by a laser Doppler velocimeter in a turbulent boundary layer filled with microbubbles. The void fraction distribution was also measured using a fiber-optic probe. The velocity decreased in the region below 100 wall units with an increase in bubble density. This led to a decrease in the velocity gradient at the wall, which was consistent with a decrease in shearing stress on the wall. The turbulence intensity in the buffer layer increased at a low microbubble density, and then began to decrease with an increasing microbubble density. Based on the present measurements, the mechanism of turbulence reduction by microbubbles is discussed and a model is proposed. Received for publication on Dec. 3, 1999; accepted on April 18, 2000     

Distribution of void fraction in bubbly flow through a horizontal channel: Bubbly boundary layer flow, 2nd report

A method of enveloping the hull with a sheet of microbubbles is discussed. It forms part of a study on means of reducing the skin friction acting on a ship's hull. In this report, a bubble traveling through a horizontal channel is regarded as a diffusive particle. Based on this assumption, an equation based on flow flux balance is derived for determining the void fraction in approximation. The equation thus derived is used for calculation, and the calculation results are compared with reported experimental data. The equation is further manipulated to make it compatible with a mixing length model that takes into account the presence of bubbles in the liquid stream. Among the factors contained in the equation thus derived, those affected by the presence of bubbles are the change of mixing length and the difference in the ratio of skin friction between cases with and without bubbles. These factors can be calculated using the mean void fraction in the boundary layer determined by the rate of air supply into the flow field. It is suggested that the ratio between boundary layer thickness and bubble diameter could constitute a significant parameter to replace the scale effect in estimating values applicable to actual ships from corresponding data obtained in model experiments.List of symbols a ₁ proportionality constant indicating directionality of turbulence - B law-of-the-wall constant - C _f local skin-friction coefficient in the presence of bubbles - C _f0 local skin-friction coefficient in the absence of bubbles - d _b bubble diameter (m) - g acceleration of gravity (m/s²) - j _g flow flux of gas phase accountable to buoyancy (m/s) - j _t flow flux of gas phase accountable to turbulence (m/s) - k ₄ constant relating reduction of liquid shear stress by bubble presence to decrease of force imparted to bubble by its displacement due to turbulence - l _b mixing length of gas phase (m) - l _m mixing length of liquid phase (m) - l _mb diminution of liquid phase mixing length by bubble presence (m) - Q _G rate of air supply to liquid stream (l/min) - q _/g velocity of bubble rise (m/s) - 2R height of horizontal channel (m) - T _* integral time scale (s) - U _m mean stream velocity in channel (m/s) - U friction velocity in channel (m/s) - V volume of a bubble (m³) - u, ¯ v time-averaged stream velocities inx- andy-directions, respectively (m/s) - u, v turbulent velocity components inx- andy-directions, respectively (m/s) - v root mean square of turbulence component in they-direction (m/s) - root mean square of bubble displacement iny-direction with reference to turbulent liquid phase velocity (m) - y displacement from ceiling (m) - local void fraction - _m mean void fraction in boundary layer - _m constant relating local void fraction to law-of-the-wall constant - _t reduction of turbulent stress (N/m²) - law-of-the-wall constant in turbulent liquid region in absence of bubbles - ₁ law-of-the-wall constant in turbulent liquid region in presence of bubbles - ₂ law-of-the-wall constant in gas phase - _m constant indicating representative turbulence scale (m) - viscosity (Pa × s) - v kinematic viscosity (m²/s) - density (kg/m³) Suffixes G gas - L liquid - 0 absence of bubbles     

Investigation of turbulent flows in a waterjet intake duct using stereoscopic PIV measurements

Stereoscopic particle image velocimetry measurements were made in a wind tunnel using a prototype waterjet model. The main wind tunnel provided the vehicle velocity and a secondary wind tunnel was set up as the waterjet propulsion model. Pressure distributions along the ramp and lip sides inside the duct were measured for three jet velocity to vehicle velocity ratios. Three-dimensional velocity fields were obtained at the intake entrance and the nozzle exit of the waterjet system. The flow into the duct was faster in the lip region than on the ramp side. Because of the variation in intake geometry from a rectangular to a circular section and because of the sudden curvature change on the lip side, a pair of counter-rotating vortices was observed in the mean velocity field at the nozzle exit. In addition, the turbulent kinetic energy correlated with the vortex pair was stronger on the lip side than in other areas. Dominant large-scale structures were extracted by using a snapshot proper orthogonal decomposition analysis. It was found that most of the turbulent kinetic energy was attributed to at least three vortices near the nozzle exit. This detailed three-dimensional velocity field will be useful for the verification of CFD simulations applied to the waterjet system.     

Mixing in the surface boundary layer of a tropical freshwater reservoir

A combined observational-modeling study was conducted to investigate turbulence mixing, and the relation to surface forcing, in the surface boundary layer (SBL) of a tropical, high-altitude, freshwater reservoir. A suite of vertical profiles of temperature microstructure, collected at three different stations of one-day duration each, provided estimates of dissipation rates of turbulence kinetic energy, , and temperature variance, χ. Numerical simulations of and χ, using state-of-the-art, public domain, two-equation turbulence closure models, compared favorably with the observations and reproduced the dynamics of daytime wind mixing as well as the vertical and temporal turbulence structure during nighttime convective conditions.Two independent estimates of vertical eddy diffusivities in the stably stratified (daytime) SBL, computed from the microstructure measurements, agreed closely, and the near surface heat and buoyancy fluxes, computed from the diffusivities, were similar to those computed independently from surface meteorology. Model generated eddy diffusivities agreed closely with the observed values, except those generated by K profile parameterization (KPP) model simulations. The good agreement provides confidence that nutrient fluxes in the SBL may be accurately computed from the models when forced with regularly measured surface meteorological parameters. The consequences are important for estimation of daily primary productivity rates in the euphotic zone and the ability to predict algal blooms such as those observed in the present reservoir.     

Non-local simulation of the stable boundary layer with a third order moments closure model

A new higher order closure model for the stable boundary layer is presented and compared with Large Eddy Simulation data. The model includes numerical solutions for the mean values, second and third order moments equations. A satisfactory agreement is found between the calculated vertical profiles of the turbulent quantities with those provided by the LES. Furthermore the new model results are compared with profiles obtained with a lower order closure model in order to verify the effective importance of including third order dynamical equations in the model.     

湍流边界层脉动压力对声强测量的影响研究

实际声强测量时常常存在风流或水流,如在飞机或船舶上进行测量,声强探头将受到湍流边界层脉动压力的影响。如何评估该影响以及如何修正测量结果是人们十分关心的工程实用问题。该文首先简要介绍了建立的由湍流边界层脉动压力诱发声强的理论模型。接着利用现有的湍流边界层脉动压力频率-波数谱模型,对湍流边界层脉动压力及其诱发的声强进行了数值分析。为了验证理论模型及数值结果,设计制作了一套实验装置,对湍流边界层脉动压力及其产生的声强进行了具体测量分析。结果表明,湍流边界层脉动压力的测量结果与数值结果吻合良好,测量得到的边界层脉动压力诱发的声强特性与计算结果也十分一致,但必须注意对测量传感器的空间响应进行修正。     

Initial results from long-term measurements of atmospheric humidity and related parameters in the marine boundary layer at two locations in the Gulf of Mexico

Measurements of boundary layer moisture have been acquired from Rotronic MP-100 sensors deployed on two NDBC buoys in the northern Gulf of Mexico from June through November 1993. For one sensor, which was retrieved approximately 8 months after deployment, the post- and precalibrations agreed closely and fell well within WMO specifications for accuracy. The second sensor operated continuously from June 1993 to February 1997 (3.5 years). Buoy observations of relative humidity and supporting data were used to calculate specific humidity and the surface fluxes of latent and sensible heat. Specific humidities from the buoys were compared with observations of moisture obtained from nearby ship reports, and the correlations were generally high (0.7–0.9). Surface gravity wave spectra were also acquired. The time series of specific humidity and the other buoy parameters revealed three primary scales of variability, small (h), synoptic (days), and seasonal (months). The synoptic variability was clearly dominant and occurred primarily during September, October, and November. Most of the synoptic variability was due to frontal systems that dropped down into the Gulf of Mexico from the continental US followed by air masses which were cold and dry. Cross-correlation analyses of the buoy data indicated that: (1) the moisture field was highly coherent over distances of 800 km or more in the northern Gulf of Mexico; and (2) both specific humidity and air temperature served as tracers of the motion associated with propagating atmospheric disturbances. These correlation analyses also revealed that the prevailing weather systems generally entered the buoy domain from the South prior to September, but primarily from the North thereafter. Spectra of the various buoy parameters indicated strong diurnal and semidiurnal variability for barometric pressure and sea surface temperature (SST) and lesser variability for air temperature, wind speed and significant wave height. The surface fluxes of latent and sensible heat were dominated by the synoptic events which took place from September through November with the transfer of latent heat being primarily from the ocean to the atmosphere. Finally, an analysis of the surface wave observations from each buoy, which included calculations of wave age and estimates of surface roughness, indicate that major heat and moisture flux events coincide with periods of active wave growth, although the data were insufficient to identify any causal relationships.