【作者】 何慧灵；

【导师】 杨克成；

【作者基本信息】 华中科技大学 ， 光学工程， 2012， 博士

【摘要】 水下成像目前已成为水下探测的一种常用手段。湍流作为一种重要的流动现象成为影响激光水下成像的重要因素。因此需要了解湍流场,掌握其流动特性才能进行相应的后续工作。为实现该目的,本论文通过粒子成像技术完成了对水下湍流流速场的测量。粒子成像测量技术的工作原理是用激光薄片照亮流场中一个与流场流速平行的平面,在与激光面垂直方向上拍摄两个时刻流场中流动粒子的图像,对粒子图像进行处理就可以得到流场中的速度场分布。根据其工作原理,本论文从示踪粒子的选择、待测流场的建立、实验系统的选型和搭建及粒子图像处理四个部分进行了论述。流场中加入的示踪粒子是正确反映流场流速的重要因素。本论文讨论了示踪粒子的跟随性和光散射特性。微小气泡作为无污染的微粒在本实验中做为示踪粒子添加到水体中,通过探测气泡的运动从而反映水体的流动。通过对两相流的模拟,计算了不同流速下不同直径的气泡的跟随性。本论文提出了符合湍流条件的新的气泡模型-椭球模型并采用该模型对气泡的光散射进行了研究。在平面波入射的情况下,采用几何光学近似的方法计算了大尺寸气泡的光散射特性。其计算结果符合Mie散射理论的结果。同时分析了激光入射时,高斯光束照射下球形气泡的光散射分布。采用类似的方法计算得到的散射分布与扩展米散射理论进行了比较,通过各阶散射光强的分布与Debye序列的比较分析了产生差异的原因。本论文建立了实验用待测湍流场。采用可实现κ-ε模型对两种结构简单的湍流场圆柱绕流模型和管道流动模型进行了二维模拟分析。选用了管道流动做为实验室流场模型。根据产生湍流场的条件,设计待测湍流场,并根据设计要求对各组件选件并组成了实验用待测流场。本论文搭建整个粒子成像测量系统。根据流场的流速设计及相关要求,对激光器、扩束光路、成像光路及图像记录元件进行选件。搭建光路,选取双帧单曝光的方式记录流场粒子图像。本论文对粒子图像对采用相关运算并得到了视场范围内的流场速度矢量分布图。考虑CCD的背景噪声,对粒子图像做了降噪及增强处理。选用合适的问询区,对两幅图像对应的区域做互相关运算,得到该区域的流速矢量。依次探询并得到整个图像的流速矢量分布。更多还原

【Abstract】 Image underwater is a useful technique in the detection underwater. The turbulence as an important flow phenomena becomes an important factor for the laser imaging underwater. It is necessary to find out the properties of the turbulent flow. Aimed at this purpose, the thesis implemented measurement of the turbulent velocity field underwater using particle image velocimetry(PIV).The principle of PIV is described as followed:A laser sheet illuminates a plane of the flow field, and the images of the fields are recorded in the perpendicular direction at two different times. Then velocity distribution of the flow field has been achieved using image processing. Based on the working principle, the thesis includes four parts:the selection of the tracer particles, the construction of the measured flow field, the establishment of the whole PIV system and the image processing of particle images.The tracer particles seeding in the flow fields is an important factor to accurately measure the fluid velocity. The optical scattering and following properties of seeding particles are discussed. The small bubbles as non-polluting particles are seeded into the water in these experiments. The velocity of the flow field can be achieved with the measurement of the bubbles motion. Through the simulation of two phase flow, the following properties of the bubbles with different diameters are calculated in different flow velocities.The thesis proposes a new spheroidal model to simulate the bubble’s shape in the turbulent flow to research the optical properties of a bubble. With the plane wave incidence, the scattering distribution of the bubble is calculated by the geometrical optics approximation (GOA). The algorithm is verified using the Mie results for a spherical bubble, and the scattering patterns of the two methods agree well. The scattering properties of the spherical bubble with the Gaussian incidence are also calculated. The results of GOA and GLMT(Generalized Lorenz-Mie Theory) are compared, the scattering intensity distribution of different p orders rays are analyzed with Debye series in order to explain the differences between the two results.The measured turbulence flow is constructed in this thesis. Using the RKE(realized κ-ε) model, two flow fields, including the flow around cylinder and the pipe flow, are simulated. The latter one is selected in the experiment. Based on the conditions of produceing turbulence, the equipments of the measured flow are designed and chosen, and the measured flow field in experiment is achieved.The whole PIV system is established in this thesis. Based on the demands of velocity of flow and other corresponding conditions, the instruments of the laser, laser sheet optics, imaging optical path and image recordings are selected and assembled. The experiments are fulfilled and the images are achieved using the single frame-double exposed recording.The pairs of particle images are calculated using the correlation algorithm to obtain the velocity vectorgraph of flow in the field of view. Considering the background noise of CCD, the particles images are reduced noise and enhanced. With the suitable interrogation windows, the images are processing using the cross-correlation in the corresponding grids to calculating the velocity of flow in this region. Then we yield the whole by analyze the image pairs point-by-point.更多还原