直接空冷凝汽器流动和传热的降维分析

Reduced Order Analysis on Flow and Heat Transfer in Direct Air-cooled Condenser

由于传统的实验和数值研究的耗时性和资源有限性,目前无法将其用于火电机组空冷系统的实时运行控制。为了快速精确地预测空冷凝汽器单元(ACC)在环境风、轴流风机以及翅片管束的共同作用下的流动换热特性,将本征正交分解(POD)方法引入对环境风影响下的凝汽器单元内的速度场和温度场建立了低维模型。在风速研究范围(0.25m/s,10m/s)内均匀取了40个数值模拟结果的速度和温度场组成样本空间,得到速度场和温度场的低维模型。而后采用三次样条插值方法,得到非样本工况下的POD解。在保持速度场和温度场的相对误差分别在10-4.5和10-6的情况下,将CFD数值模拟105量级的自由度降至POD模型的101数量级。研究结果为实现火电空冷系统实时的运行控制提供了更有效的方法。

As the experimental and numerical measurements are time-consuming and their calculating resources are limit- ed, real-time control is not available for the air-cooled sys- tem operation using the method in thermal power unit. In order to rapidly and accurately predict the flow and heat transfer characteristics influenced by the ambient wind, axi- al fan and the finned tube bundles in air-cooled condenser （ACC）, proper orthogonal decomposition （POD） procedure was introduced to construct the reduced order models （ROMs） for velocity and temperature fields in ACC affected by the ambient wind. 40 simulating velocity and temperature fields in ACC, which ambient wind magnitude increases with uniform step among the ranges from 0.25m/s to 10m/s, constitutes the sample space to obtain the ROMs.Then cubic spline interpolation is utilized to acquire the POD solutions during the out-sample cases. The mean relative er- ror orders of POD solutions for velocity and temperature fields keep at 10-4 and 10-6. At the same time, the de- gree of freedom （DOF） can be decreased from 10s in CFD models to 101 in POD models. The present investigations provide a more reliable approach for the real-time opera- tional control of ACC system in power plants.