【作者】 王贵华；

【导师】 周怀春； 程强；

【作者基本信息】 华中科技大学 ， 热能工程， 2012， 博士

【摘要】 同其他传热方式相比,辐射传热包含了发射、吸收、散射、反射等复杂过程,其控制方程为微分积分方程。在辐射换热问题分析以及实际应用中,辐射传递方程的求解和辐射强度的计算至为关键。到目前为止,已经发展了多种方法用来求解辐射传递方程,如离散坐标法、蒙特卡洛法、热流法、区域法、球形谐波法等,这些方法各有优缺点,根据其特点被用于不同的辐射问题求解中。在一些辐射问题的研究中,方向辐射强度信息显得格外重要。基于蒙特卡洛法发展起来的DRESOR (Distributions of Ratio of Energy Scattered by the medium Or Reflected by the boundary surface,被介质散射或者被壁面反射的能量份额分布)法,因能够提供高方向分辨率的辐射强度,而被应用到一系列的辐射传递问题的研究中,如工业炉内燃烧温度场的可视化监控、平行脉冲入射辐射、各向异性散射、梯度折射率介质中等。然而因该方法基于蒙特卡洛法,计算耗时不可避免,同时存在着统计误差,限制了此方法的进一步发展。本文致力于提高DRESOR法计算效率和计算精度的研究,在基本DRESOR法的基础上发展了ES-DRESOR (Equation-solving DRESOR)法。通过求解线性方程组直接得到DRESOR数(DRESOR法求解的核心),无需像先前那样由蒙特卡洛法的能束跟踪得到,以实现计算效率和计算精度上都能有大幅度的提升。本研究对于提高DRESOR法的分析求解能力,拓展其应用范围具有重要意义。具体来说,主要工作如下：在基本DRESOR法的基础上提出了ES-DRESOR。给出了该方法的计算原理,分为透明壁面和漫壁面参与两种情况,对于前者,设定系统内某微元体具有单位黑体发射,其它部分没有发射,这样对DRESOR法给出的辐射强度简化后,可由入射辐射(对辐射强度的空间积分)的两种计算方法(即DRESOR法给出的计算公式和标准计算公式)所得结果相等,建立关于DRESOR数的方程。在漫壁面参与辐射过程后,对DRESOR法计算入射辐射表达式进行了修正,加入壁面的贡献部分。对壁面单元建立方程的原则是被壁面反射的总能量等于其它单元散射或反射到此壁面的总能量；对空间单元建立方程的原则依然是由两种方法计算的入射辐射相等。在一维吸收发射各向同性散射的辐射系统内,对ES-DRESOR求解辐射传递方程进行了研究。数值部分,首先对ES-DRESOR法的可靠性进了验证,包括与文献中的精确解对比,与广泛使用的离散坐标法对比,以及将ES-DRESOR法计算的结果代入到辐射传递方程进行检验。而后将ES-DRESOR法计算出来的DRESOR数与蒙特卡洛法计算的结果进行对比,显示了很好的吻合性,且消除了蒙特卡洛法中因统计误差存在而导致的“抖动”现象。对比ES-DRESOR法同蒙特卡洛法计算出来的辐射强度相对误差,前者的精度高出一个数量级。更为重要的是ES-DRESOR法的计算效率在测试网格数下,与基于蒙特卡洛法的基本DRESOR法相比,有两个数量级的提升。更多还原

【Abstract】 Compared with other modes of heat transfer, radiative heat transfer contains some more complex process as the emission, absorption, scattering and reflection of the radiative energy. The control equation of the radiative heat transfer is an integral differential equation. How to solve the radiative heat transfer equation (RTE) and calculate the radiative intensity is the key in the analysis of radiative heat transfer problems and practical applications. So far, many methods have been developed to solve the RTE, for example, the Discrete Ordinate method, the Monte Carlo method, the Heat Flux method, the Zone method, the Spherical Harmonics method and others. These methods do have advantages and drawbacks, and are used in different radiative heat transfer problems according to their own characteristics. In some radiative problems, the information of intensity in direction is particularly important. Based on the Monte Carlo method, the DRESOR method (Distributions of Ratio of Energy Scattered by the medium Or Reflected by the boundary surface) can get the radiative intensity with high directional resolution and has been applied to a series of radiative heat transfer problems, as the visual monitoring of the combustion temperature field in the industrial furnaces, the collimated pulse incidence, the an-isotropic scattering media and the gradient index medium and so on. The calculating of the DRESOR values is the kernel and demands the most calculating time of the program. As it is based on the MCM, the time-consuming is unavoidable and the statistical error exits, which limit the development of this method.This paper aims to improve the computational efficiency and accuracy of the DRESOR method. On the basis of the basic DRESOR method, Equation solving DRESOR method, ES-DRESOR method, is devoloped in this paper. In this method, the DRESOR values which is the kernel in the DRESOR method can be got by solving a set of linear equations instead of by the traditional ray-tracing using MCM, so as to have a great enhance on the calculation efficiency and accuracy. This study has an important significance for improving ability of DRESOR method solving radiation problems and expand its application scope. Specifically, the main work is as follows:The principles of ES-DRESOR method are given. The computing principles contain the transparent boundary surfaces and the diffuse boundary surfaces two conditions. For the former, assuming an arbitrary element emission of the system as a blackbody emission while other elements with no emission, which simplifies the intensity description given by the DRESOR. Then equation can be set up by two different calculating description of incident radiation (the standard calculating equation and the equation under the DRESOR method) equals to each other. For the latter, the calculation description of the incident intensity under the DRESOR method needs to be modified, with the contribution of the boundary surface added. The principle of establishing equations for the boundary surface elements is that the total energy reflected by boundary surface equals to the total energy scattered and/or reflected to this boundary surface. As for the spatial elements, the difference with the former one is that the contribution of boundary surface needs to be added.Numerical verification is done on this method in one-dimensional absorbing, emitting, and isotropicall scattering system. First, the reliability of this method is verified including compared with the exact solution in the literature, compared with the widely used DOM, and substituted the ES-DRESOR results into the RTE to test. Then compared the DRESOR values with the MCM, the results show a good agreement of the two method. The shaking phenomena of DRESOR values caused by the statistical error in the MCM is eliminated in the ES-DRESOR method. Compared the relative error of the radiative intensity calculated by the ES and MCM, the accuracy of the former is one order magnitude higher. What is more important is that the efficiency of the ES is two order magnitudes higher than MCM under the same discrete grids.更多还原