Numerical investigation of unsteady flows in hydraulic turbines

该项目的主要目标是对水轮机中固有的非定常流动进行数值研究，并确定雷诺平均纳维尔-斯托克斯（RANS）能够解析相关流动结构的水平。涡轮机中的流动波动在结构上产生疲劳应力，该疲劳应力可导致设备故障。这种波动存在于任何操作条件下：尾流相互作用和静止和旋转叶片之间的压力波，水力不平衡，旋涡脱落，剪切层平流。此外，在部分负荷下，强旋流以大涡流的形式离开转轮。由于尾水管中的逆压力梯度，该涡流可能破裂并产生危及结构的振荡压力。为了更好地了解不稳定运行条件下的流动现象，魁北克水电公司（HQ）最近开始了一项雄心勃勃的项目，使用ANSYS-CFX和OpenFOAM软件模拟非定常三维流动。然而，经验提出了一些问题，在充分预测的流动结构和不稳定性的能力。在这段时间内，非定常流的精确模拟对于确定涡轮机叶片上的动态载荷至关重要，因为在尾水管中存在强烈的非定常流。数值模拟的精度完全取决于网格的质量、数值格式的选择和湍流模型。目前对尾水管内非定常流进行建模的尝试表明了一阶闭合模型的无效性;然而，在所需的网格质量和数值方案中很少提及。拟议的项目，以调查网格，湍流模型和数值方案的作用是一个关键的和急需的知识，以帮助推进HQ在正确的方向。它将为总部的IREQ小组提供有价值的技术信息以及模拟这种流动的数字工具。所提出的研究的影响将使HQ能够更准确地预测水电机组的剩余寿命。

The primary objective of the project is to numerically investigate unsteady flows that are inherently present in hydraulic turbines and determine the level at which the Reynolds-averaged Navier-Stokes (RANS) is able to resolve the pertinent flow structures. Flow fluctuations in a turbine generate fatigue stresses on the structure that can lead to equipment failure. Such fluctuations are present under any operating condition: wake interactions and pressure waves between stationary and rotating blades, hydraulic unbalance, vortex shedding, advection of shear layers. Moreover, at part load, a strong swirling flow exits the runner in the form of a large vortex. Because of the adverse pressure gradient in the draft tube, this vortex may break down and give rise to an oscillating pressure that endangers the structure. To better understand the flow phenomena in unstable operation conditions, Hydro-Quebec (HQ) has recently embarked on an ambitious project to simulate unsteady three-dimensional flows using both the ANSYS-CFX and OpenFOAM softwares. However, the experience has raised a number of questions in regards to the ability to adequately predict the flow structures and instabilities. The accurate simulation of the unsteady flow during this period of time is crucial in determining the dynamic loading on the turbine blades, where strong non-stationary flow are present in the draft tube. The accuracy of the simulation is entirely dependent on the quality of grid, the choice of numerical schemes, and turbulence models. Current attempts at modelling the unsteady flow in the draft tube has shed light on the ineffectiveness of first-order closure models; however, very little is mentioned in the required quality of grid as well as the numerical scheme. The proposed project to investigate the role of the grid, turbulence model, and numerical scheme is a critical and much needed knowledge to help advance HQ in the right direction. It will provide HQ's IREQ team with the valuable technical information as well as a numerical tool to simulate such flows. The impact of the research presented will enable HQ to predict with greater accuracy the residual life of hydroelectric units.