湍流自然对流新模型及其实验验证

Turbulent natural convection, as a common physical phenomenon in the fields of new energy use and traditional industries, has been received considerable attention because it determines the efficiency of energy using. Moreover, the Rayleigh number is very large because in these cases the temperature difference is large and also the length scale. The studies of turbulent natural convection with large Rayleigh number are impossible to be carried out through experimental method because the cost of the test equipments is very high or there is no enough space in the laboratory to build the test equipments. In such circumstances, the numerical simulation becomes the most effective method to analyze the turbulent natural convection having large Rayleigh number. At present, the k-ε model is the most popular model to simulate the turbulent natural convection for engineering purposes. But the difference between the numerical results of the k-ε model and the experimental results are always very significant owing to many reasons. Thus, it is necessary to develop a new k-ε model to accurately capture the main characteristics of turbulent natural convection. After analyzing the shortcomings of the common k-ε model, this project uses the theoretical analysis method to transform the momentum equation and the energy equation of natural convection boundary layer into the same form. Based on the quasi Prandtl number deduced from the two transformed equations, an expression of turbulent Prandtl number is obtained, then based on the quasi Prandtl number a new k-ε model has been constructed. The new k-ε model is validated from two aspects: different shape of the wall on which turbulent natural convection processes and different fluid property (Prandtl number) by the available experimental data in the literatures. The new k-ε model uses denser grid in the near wall region to decrease the error resulting from large gradient of physical parameters. The values of the turbulent kinetic energy and the turbulent kinetic energy dissipation rate of the first inner point are given by some modified wall functions. If the difference between the numerical results and the experimental results is over the required engineering accuracy, the turbulent Prandtl number is revised again until the new k-ε model can obtain expected results. In order to explain the physical meanings of the turbulent Prandtl number used to develop the new k-ε model, experimental studies are performed by the hot wire anemometer system IFA300 made in America. The main result of this project is a new k-ε model which is expected to obtain more accurate results with less sensitivity to fluid property (Pr) and shape of the wall on which turbulent natural convection occurs. This result will not only have scientific significance on deeper insight on turbulent natural convection, but also have broad engineering applications.

湍流自然对流是新能源利用和传统工业等领域常见的物理现象，具有高温差大尺度等特点，从而Ra数很高。由于在实验室很难建立很高Ra数的湍流自然对流模型，所以用数值方法研究高Ra数湍流自然对流成为切实可行的方法，而工程计算中最常用的k-ε模型的数值研究结果与实验结果之间存在较大差异。为了获得精度较高、适应面广的湍流模型，本项目在分析常用模型缺陷基础上，用理论分析法构造一湍流普朗特数σt，并以它为核心建立求解湍流自然对流的新k-ε模型；用已发表的实验数据从流体属性和对流面曲率两方面对新模型进行考核和修正，使新模型在满足工程计算精度情况下达到较理想的普适度；再通过实验方法对模型所用σt的物理意义进行解释。项目研究结果是一适用于求解不同Pr数介质、不同对流表面曲率湍流自然对流问题的新模型，对进一步描述和认识湍流自然对流具有重要的科学意义和学术价值，而且作为一种研究工具具有广泛的应用前景和工程实用价值。

主要研究内容：针对目前用于求解湍流自然对流流动与传热的k-ε模型在应用中存在的不足，结合高Re数k-ε模型和低Re数 k-ε模型的特点，重新定义了湍流普朗特数σt的计算式，提出了一种修正的k-ε新模型。该模型用于封闭方腔内的湍流自然对流流动与传热计算对于流体属性（Pr=0.001~1000）具有很好的适应性。在此基础上，将该模型应用于建筑供暖、自然通风等工程实际问题。实验研究方面，在热线风速仪系统标定方法、提高测试精度、探针防碰撞系统设计、降低扰流效应等方面也开展了一系列探索性工作。..主要成果：.（1）提出的k-ε新模型对封闭方腔内的湍流自然对流数值计算的结果与实验值之间的相对误差在当109≤Ra≤1014时均在6%以内。该模型对于流体属性具有很好的适应性，准确地描述了封闭腔内湍流自然对流换热中边界层发展与壁面传热特性之间的内在联系。.（2）提出了一种简单易操作的探针标定方法。对同一温度气流在不同流速以及同一流速气流在不同温度下，进行了探针标定。.（3）设计制作了探针标定中对测试气流的快速、准确加热装置。克服了热线风速仪标定系统设计上的缺点，提高了标定过程中对气流温度的控制精度。.（4）自行设计制作了探针防碰撞控制系统。该系统成本低、可靠性高，对于后续测试工作的顺利展开以及降低探针坏损率起到了重要作用。.（5）提出了建筑外壁面换热系数与室内自然对流换热情况下室内环境参数间的关联性。对于暖通行业在设计、系统运行管理等环节提供了理论参考。.（6）研究了太阳辐射通过激励室内自然对流换热过程对室内污染物的空间分布及迁移特性的影响，获得了寒冷地区冬季自然通风的合理模式和空气质量控制策略。. 经过4年的研究，获得了一系列与自然对流换热有关的基础性成果，并将其应用于暖通、空调等领域。在国内外学术期刊和国际学术会议上发表（含收录）论文19篇，其中EI收录8篇（含2篇录用），CSCD核心期刊论文8篇。出版专著1部。获国家发明专利1项，培养研究生6名。