Experimental investigation of three-dimensional turbulent boundary layers on ‘infinite’ swept curved wings
Experimental investigation of three-dimensional turbulent boundary layers on ‘infinite’ swept curved wings
复制标题
“无限”后掠曲翼三维湍流边界层实验研究
DOI：
--
复制
发表时间：
1990
影响因子：
3.7
通讯作者：
P. Bradshaw
中科院分区：
文献类型：
--
作者：
V. Baskaran;Y. G. Pontikis;P. Bradshaw
在弯曲管道中的三维湍流边界层中进行了平均流动和湍流测量，模拟了两个分别具有凹曲率和凸曲率的“无限”后掠弯曲机翼表面上的逆压梯度。在两种情况下，初始边界层厚度与表面曲率半径的比值约为0.01，这是早先分别由霍夫曼、穆克和布拉德肖（1985）和穆克、霍夫曼和布拉德肖（1985）对凹曲率和凸曲率影响进行的二维紊流边界层研究中所用的值。压力驱动横流的流向分布与Bradshaw和Pontikos（1985）的“无限”后掠平面实验中的流向分布几乎相同，后者使用了类似的管道。研究结果表明，平均流的三维性和任一符号的表面曲率的耦合效应对湍流结构的影响很小，而上述附加应变率单独作用时的影响很大。在凹的情况下，横流的影响似乎除了抑制展向波动不均匀性外，还与曲率的失稳效应相反。相反，凸曲率和横流的弱联合影响（两者分别倾向于衰减湍流）意味着这两种效应之间的相互作用是严重非线性的。湍流模拟目前的结果的影响进行了简要讨论。
Mean flow and turbulence measurements have been made in three-dimensional turbulent boundary layers in curved ducts, simulating adverse pressure gradients on two ‘infinite’ swept curved wing surfaces with concave and convex curvature respectively. The ratio of the initial boundary-layer thickness to the surface radius of curvature in both cases is approximately 0.01, the value used in the earlier two-dimensional turbulent boundary-layer studies on the effects of concave and convex curvature by Hoffmann, Muck & Bradshaw (1985) and Muck, Hoffmann & Bradshaw (1985) respectively. The pressure-driven crossflow has nearly the same streamwise distribution as in the ‘infinite’ swept flat-surface experiment of Bradshaw & Pontikos (1985), which used a similar duct. The results of the present study show that the coupled effects of mean flow three-dimensionality and prolonged mild surface curvature of either sign have rather a weak influence on the turbulence structure, unlike the significant influence of the above extra strain rates when applied individually. In the concave case, the effect of the crossflow appears to oppose the destabilizing effect of curvature in addition to suppressing spanwise wavy inhomogeneities In contrast, the weak combined influence of convex curvature and crossflow, both of which, separately, tend to attenuate turbulence, implies that the interaction between the two effects is grossly nonlinear. Implications of the present results for turbulence modelling are briefly discussed.