单相复杂流动的离散模拟
文献类型：学位论文
| 作者 | 王利民 |
| 学位类别 | 博士 |
| 答辩日期 | 2008-07-11 |
| 授予单位 | 中国科学院过程工程研究所 |
| 授予地点 | 过程工程研究所 |
| 导师 | 李静海 |
| 关键词 | 复杂流动 湍流 微流动 多尺度分析 宏观粒子方法 拟颗粒模型 |
| 其他题名 | Discrete Simulation for Single-phase Complex Flows |
| 学位专业 | 化学工程 |
| 中文摘要 | 多相化工过程中的流动大多属于复杂流动，呈现动态的多尺度流动结构和非线性的速率关系。对它的合理预测与机理性认识对理论和工程研究都是巨大的挑战。近年来多尺度粒子模拟方法的发展为此提供了新的可能性，本论文从不同控制机制协调的角度，通过多尺度分析和离散化模拟对单相湍流中稳定性条件的表达和气体微流动的流型和阻力特性进行了探索研究。 本文首先从三方面改进了离散化模拟方法：建立了一种能同时保证恒温的无滑移边界条件；运用图像识别原理，实现了一种复杂边界的初始条件生成方法；并进行了比较合理的单向数据传递的并行化。 为了探索湍流形成的物理机制，论文运用离散模拟分析了湍流中的多尺度结构和能量耗散过程。研究发现其惯性和粘性作用机制的协调与气固体系中的协调机制类似，从而将能量最小多尺度模型(EMMS)的思想拓展到湍流研究。论文改进了关于管内湍流速度分布的稳定性条件，即粘性作用使粘性平均剪切耗散率趋于最小，而惯性作用使湍流脉动耗散率趋于最大。进而利用改进的宏观拟颗粒方法通过直接数值模拟单圆柱绕流的失稳过程，检验了上述的稳定性条件的合理性和扩展到其它流动现象的可能性。运用稳定性条件，探讨了水力学中最小能耗率假设和最大能耗率假设之间的关系，提出了协调两种假设的可能性。 运用微观离散化模拟方法—改进的颗粒模型模拟了Kn数处于0.010.20范围内的微通道流动，独立地考察了不同Kn数和Ma数下微通道流动的传递特性。研究发现，微通道截面流速的分布主要受Kn数的影响，在等外力场下，壁面滑移速度随着Kn数的增加先增加后减小，在等Kn数下，壁面滑移速度随着外力场强度的增强而增大；通道截面温度分布主要受外力场影响，外力场较弱时截面最高温度出现在中心区域，外力场较强时，截面最高温度出现在近壁区域，而且外力场强度越强，近壁区域与中心区域的温度梯度越大；微通道的达西摩擦因子随着Kn数增大而增大，随着Kn数的增大，类似于微流实验中观察到的层流向“湍流”转捩提前的趋势。此外，还通过大规模并行模拟从分子层次研究了微米级通道内的复杂流动行为。上述模拟结果初步表明，纳微流动中的复杂现象更多地与分子的运动特征及壁面条件相关，这与宏观上的湍流中复杂现象的成因有显著差别。 论文研究表明，多尺度方法是研究具有非线性非平衡特性复杂流动的一种合理的研究思路，而离散方法是很有潜力的模拟工具，两者的结合应用将有力促进对湍流和微流动等复杂流动行为的机理研究。 |
| 英文摘要 | Flows in multi-phase chemical engineering processes are typically complex flows, which present dynamical multi-scale structures and are described by nonlinear rate equations. Understanding and reasonable prediction of such flow are great challenges to both theoretical and engineering researches. Recently, the development of multi-scale particle methods has provided a promising choice for the simulation of complex flows. This study investigated the stability condition in single-phase turbulence and resistance characteristics in gas micro flows by using analytical (variational) multi-scale approach and particle methods. Three aspects of particle methods have been updated and modified in this paper. Firstly, a new wall boundary condition which ensures both no-slip condition and thermostat has been proposed. Secondly, a novel particle initialization method has been proposed which can generate particle information from images directly. Finally, we present a single-direction data-exchange mode in the parallelizing of hard sphere model. The multi-scale structures and energy dissipation process in turbulent flow are analyzed in simulations with particle methods. The similarity between the inertia-viscosity compromising in turbulence and the compromise mechanism in gas-solid systems was found. With reference to the EMMS model for the gas-solid systems, a stability criterion for turbulent pipe flow was proposed. Furthermore, the stability criterion has been verified in the simulations of flow around a cylinder by using macro-scale particle method and extended to other flow phenomena. Finally, the stability criterion has been applied to explore the relationship between the hypothesis of maximum energy dissipation rate and the hypothesis of minimum energy dissipation rate in hydraulics studies, which puts forward the possibility of reconciling the two hypotheses. Gas flow in microchannels for Knudsen numbers ranging from 0.01 to 0.20 has been investigated by using updated and modified pseudo-particle modeling. It has been found that the velocity profiles are mainly affected by Knudsen number and the external force fields applied. Under constant external force, when Knudsen number is increased, the slip velocities on the walls increase at the beginning, and then decrease. The temperature distributions are also significantly affected by the external force. The Darcy friction factor increase with increasing Knudsen number, and its variation with Mach number under increased Knudsen number is similar to the so-called premature laminar-turbulent transition observed in experiments. Finally, the complex behavior of flow in micron channel has been investigated in molecular level by using massive parallel computing. The simulation results indicate that complex phenomena in nano/micro flows mostly are related to molecular motions and boundary condition, which differs from the cause of complex phenomena in macro-scale turbulence. The researches in this dissertation indicate that the variational stability condition can be found in complex flows by analyzing the correlation between different scales and compromise between dominant mechanisms. And the analytical multi-scale method is a reasonable approach for complex system. Furthermore, particle methods are potentially promising computational tools for understanding of the complex behaviors in turbulence and micro flows. |
| 语种 | 中文 |
| 公开日期 | 2013-09-13 |
| 页码 | 139 |
| 源URL | [http://ir.ipe.ac.cn/handle/122111/1171]  |
| 专题 | 过程工程研究所_研究所（批量导入） |
| 推荐引用方式 GB/T 7714 | 王利民. 单相复杂流动的离散模拟[D]. 过程工程研究所. 中国科学院过程工程研究所. 2008. |
入库方式： OAI收割
来源：过程工程研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。