大跨桥梁实测湍流风场的大涡模拟

Large Eddy Simulation of Measured Turbulent Wind Fields on Long-span Bridge

为实现在大涡模拟(LES)中准确评估强风湍流对大跨桥梁的作用,关键难点在于生成符合桥梁真实强风特性的入口湍流。为此应用了一种新的规则化波矢量随机流生成方法PRFG~3(Prescribed-wavevector Random Flow Generator),该方法遵守连续性方程和泰勒假设,可准确模拟目标湍流的脉动风谱、湍流度和湍流积分尺度等风特性参数。首先利用西堠门大桥结构健康监测系统(SHMS)2016年内采集的风速数据,选取了该桥址区10 min时距平均风速较大但风特性不同的2个强风样本,分析得到相应的强风特性参数;然后采用PRFG~3方法合成了符合上述2个实测强风特性的均质各向异性湍流,同时为验证该方法用于主梁节段模型LES入口湍流的适用性,还模拟了缩尺比为1∶50的强风湍流场,并基于OPENFOAM平台,将3类风场赋予LES入口进行了数值计算;最后将LES流场中多个监测点的湍流特性与实测结果进行了对比。研究结果表明:2个实测风场在顺风向、横风向、竖风向的脉动风谱均与Von Kármán谱接近,顺风向湍流积分尺度最大约为192 m,各脉动风分量近似符合正态分布;PRFG~3方法产生的入口湍流在LES计算域内能正确传输,模拟的3类风场均有较好的均匀性,脉动风分量在各方向上的功率谱、湍流度和所有的9个湍流积分尺度均与相应的实测值吻合较好。相关湍流合成方法及LES设置可为大跨桥梁在实测强风湍流下的数值模拟研究提供参考。

Accurately assessing the effect of strong wind turbulence on a long-span bridge by large eddy simulation(LES) primarily involves synthesizing turbulence inflow conditions that meet the actual strong wind characteristics of the bridge. In this study, a new synthetic turbulence method named PRFG~3(prescribed-wavevector random flow generator) is proposed, which fulfils the continuity equations as well as the Taylor assumption, and can thus be used to accurately simulate the characteristics of the target turbulence, such as wind spectral, turbulence intensity, and integral length scale. First, based on the wind speed data collected by the Xi-houmen bridge’s structural health monitoring system(SHMS) in 2016, two strong samples with a high 10 min interval average wind speed but different turbulence properties were selected for analysis, and the corresponding strong wind characteristics were obtained. Then, considering the wind characteristics of these two wind fields as target values, PRFG~3 was used to generate not only homogeneous anisotropic turbulence, but also the corresponding strong wind fields with a scale ratio of 50, thus verifying the applicability of the LES inlet of the section model. Three synthetic turbulence fields were used as inflow conditions for the LES, and the corresponding numerical simulations were carried out using the OPENFOAM software. Finally, the results of the measured points in the LES flow fields were compared with the turbulence properties of the field measurements. The results show that the wind spectra of the two selected wind fields show good agreement with the von Kármán spectral in the along-wind, across-wind, and vertical wind directions. The maximum along-wind turbulent integral length scale is approximately 192 m, and its fluctuating wind components approximately follow a normal distribution. The three synthetic inflow turbulences obtained by PRFG~3 can be correctly transmitted in the LES domain and have good uniformity. Furthermore, the wind spectrum and turbulence intensities of simulated fluctuating wind components, as well as all nine integral length scales, are in good agreement with the corresponding field-measured values. Related synthetic turbulence generation methods and numerical simulation settings can provide a reference for the numerical simulation of long-span bridges under strong wind turbulence.