【作者】 刘国标；

【导师】 余国琮；

【作者基本信息】 天津大学 ， 化学工程， 2007， 博士

【摘要】 本文第一部分对计算流体力学(CFD)、计算传热学(CHT)、计算传质学(CMT)相关知识及其相关的湍流理论、湍流模型进行了简单地介绍与讨论。并对CFD、CHT与CMT在化工单元操作过程中尤其是在预测填料塔、固定床中的流体流动、温度分布以及浓度分布的应用进行了综述、评论。本文第二部分提出建立统一动量、热量及质量传递的计算传递学基础。从广义的传递变量Φ出发,详细推导出了Φ的瞬时守恒微分方程、雷诺时均微分方程、以及用以封闭雷诺时均方程的两个辅助微分方程及其模型化,这样就构成了计算传递学体系的基本微分方程组,其中包括连续性、传递量守恒及其雷诺时均的微分方程,以及传递脉动方差及其耗散率的微分方程。根据上述计算传递学的基本方程,分别导出动量传递、热量传递及质量传递的相应微分方程及相应的方差及其耗散率的微分方程组,以及不同研究者给出的各自模型化的方程。对基于Boussinesq假设引入的湍流粘度系数μ、湍流热量传递扩散系数α_t、湍流质量传递扩散系数Dt的定义及物理意义进行了分析比较。对辅助的k,ε, t~2,ε_t, c~2,ε_c微分方程的入口、出口边界条件设置进行了讨论分析。本文第三部分是计算传递学模型在化工过程的推广应用,并对过程中的速度、温度及浓度的分布进行预测,包括(1)填料塔内流动分布:对直径分别为0.5m、1.0m散堆填料塔进行了模拟,并预测塔内的液体流动分布、发展;(2)化学吸收过程:对直径为0.1m的实验散堆填料塔、直径为1.9m的工业散堆填料塔内用碱溶液吸收CO2过程的流体流动、传热、传质过程进行了模拟,并预测了气相中CO2浓度、液相中碱浓度(碱溶液的碳化率)、液相温度沿塔高分布;(3)精馏过程:对直径为1.22m的工业散堆填料塔内精馏过程进行了模拟,并预测了浓度分布、HETP;(4)催化反应过程:对在直径为0.041m的固定床内,用乙炔、醋酸气相制备醋酸乙烯的壁冷固定床催化反应器进行了模拟,并预测温度分布、转化率分布等。上述各项预测均与文献上发表的实验测量符合较好。此外,用两方程导出的k -ε、t 2 -εt、ci 2 -εci模型可计算出的湍流动力粘度、湍流传热扩散系数、传质扩散系数沿固定床轴向、径向的分布,结果显示出彼此有部分的相似,但并不是完全相似,说明湍流动量、热量、质量传递既有类似性,又有各自的特殊性。预测的湍流传质扩散系数沿固定床轴向、径向的分布与文献报道一致。本文第四部分是进行萃取精馏实验并用计算传递学模型进行模拟及与实验结果比较。实验是研究用N-甲基吡咯烷酮(NMP)萃取精馏苯和噻吩混合物的过程,并考察了不同回流比对苯和噻吩分离情况的影响。模拟得出的结果与实验测量符合较好。本文最后,对计算传质学的展望及进一步探索提出了一些建议。更多还原

【Abstract】 The first part of this dissertation is the summary of some basic knowledge about computational fluid dynamics (CFD), computational heat transfer (CHT) and computational mass transfer (CMT), and their related turbulent theory and turbulent models are briefly reviewed. The application of CFD, CHT and CMT to the chemical engineering unit operations, such as predicting the fluid flow, temperature profile and concentration profile in packed columns or fixed-bed, are also summarized.The second part is the detailed derivation of the basic differential equations of the proposed computational transport, including the conservation equation of the general transport variableΦ, its Reynolds averaged equation together with its closure two equations and their modeled equations. From the generalized transport equation, the precise governing conservation equations of continuity, momentum transfer, heat transfer, mass transfer, as well as their auxiliary equations of turbulent kinetic energy k and its dissipation rateε, temperature variance t 2and its dissipation rateεt, concentration variance c 2 and its dissipation rateεc can be obtained, which formulates the system of equation for the transport computation. The definition and physical meaning of the turbulent kinetic viscosityμ, turbulent heat transfer diffusivityαtand turbulent mass transfer diffusivity Dt which are introduced by the Boussinesq hypothesis are analyzed and compared. Furthermore, the inlet and outlet boundary conditions of k,ε, t 2,εt, c 2,εc differential equations are also discussed and suggested.The third part is the application of the computational transport model to various chemical processes, including (1) the modeling of liquid flow in the randomly packed column of 0.5m and 1.0m I.D. in predicting the velocity profile and its evolution; (2) the modeling of chemical absorption processes in the randomly packed columns of 0.1m I.D. and 1.9m I.D. respectively for removing CO2 from gas mixtures by the alkaline solutions in predicting the profiles of CO2 concentrations in gas phase, the alkali concentrations (carbonation ratio of the alkali aqueous solutions) and the temperature in liquid phase; (3) the modeling of distillation process in randomly packed column of 1.22m I.D. in predicting the profiles of component concentration and the height equivalent of theoretical plate (HETP) as function of gas phase F-factor; (4) the modeling of the wall-cooled fixed bed catalytic reactor of 0.041m I.D for the gas phase synthesis process of vinyl acetate from acetic acid and acetylene in predicting the profiles of conversion and the temperature along the reactor axial and radial directions. All foregoing predictions were compared with the experimental measurements taken from literatures, and good agreements were found. The predicted profiles of turbulent kinetic viscosity, turbulent heat transfer diffusivity and turbulent mass transfer diffusivity calculated by using the k-ε, t 2-εt, c 2-εc two equations models respectively along the radial and axial directions in randomly packed columns displayed some similarity, which is in agreement with the reported literatures. However, they were not totally identical, which demonstrates that the turbulent momentum, heat and mass transports are not completely analogous, and they have their own characteristics.The fourth part is the description of the experimental work of the extractive distillation processes for separating the mixture of benzene and thiophene using N-methyl-2-pyrrolidone (NMP) as the extractant under the different operating conditions in a randomly packed column and the process simulation in predicting the benzene concentrations in gas phase and liquid phase by the proposed computational transport model. The prediction is confirmed by the experimental measurements.Finally, several aspects concerning the prospect and further investigation of computational transport are suggested.更多还原