明渠湍流中的主要相干结构模式

Coherent structure models for open channel flows

利用自主开发的高帧频明渠湍流粒子图像测速(PIV)系统测量了3种Reynolds数下的恒定均匀流时间序列流场,运用本征正交分解能够给出数据集的最优模态的特性,分析了明渠湍流的主要相干结构模式及其能量关系。湍动能主要集中在少数几个低阶模态上,第1阶模态含有约30%以上的湍动能。Q2(Q4)事件是明渠湍流中除平均流动外含能最多、最主要的大尺度结构,湍动能主要通过Q2(Q4)事件向其他结构传递。它并不完全是发夹涡群诱导的结构,相反从能量的角度来看,发夹涡群的产生与维持与它有关键关系。Q2(Q4)事件可能存在大范围独立维持的大尺度机制。同时本征正交分解(POD)技术成功提取了发夹涡群结构,流向范围达到约3.4h,垂向尺度从交叠层一直到水面附近,整个结构与壁面呈约10°夹角。发夹涡群载有约23%的平均湍动能,并主要集中在处于对数区的中间部分。其他含能较少的大尺度结构有水面涡旋、单双涡结构、独立涡包等。

Instantaneous velocities in open channel flows were measured using high-speed PIV in steady uniform flows at three different Reynolds numbers. Proper orthogonal decomposition （POD） was used to find the optimal models for the turbulence and turbulent coherent structures. The results show that the kinetic energy is concentrated in a few modes with the first mode containing 30% of the total energy. The Qe（Q4） events constitute the most prominent coherent motions in the flow. The Q2 （Q4） events convey kinetic energy to other large-scale structures such as the hairpin vortex packets, but the mechanism of generating and maintaining the Q2（Q4） events remains elusive. Typical hairpin vortex packet structures incline at 10°to the wall, with the stream wise length scale averaging 3.4h while the wall-normal scale ranged from the buffer layer to the free surface. Hairpin vortex packets contain 23% of the turbulent kinetic energy concentrated in the middle part of the structure in the logarithmic layer. Other less energetic large scale structures include vortices near the surface, single and coupled vortex systems, and independent vortex packets.