【作者】 郝振华；

【导师】 陆慧林；

【作者基本信息】 哈尔滨工业大学 ， 热能工程， 2011， 博士

【摘要】 气固两相流动现象广泛存在于化工、电力、冶金、食品、制药等领域中,深入认识和掌握气固两相流动的内在机理和规律有着重要的实际意义。随着计算机技术和计算方法的不断发展,数值模拟已成为气固两相流动研究的主要方法之一。但是由于气固系统本身的复杂性,气固两相流动模拟的理论模型仍然有许多值得改进和提高之处。气固两相流动中,颗粒表面具有不同的粗糙程度,粗糙颗粒在碰撞和湍流等作用下会产生旋转运动。实验和理论模拟结果均表明颗粒旋转会影响颗粒的运动轨迹、浓度分布以及其它各种宏观和微观参数。但遗憾的是,目前颗粒动理学理论是基于光滑颗粒假设而来,在理论推导过程仅考虑颗粒的平动运动,未考虑颗粒的转动运动。基于此,有必要发展粗糙颗粒动理学并应用于气固两相流动模拟。本文从颗粒动理学基本原理出发建立了粗糙颗粒动理学模型。传统的颗粒动理学理论仅采用了平动拟颗粒温度来描述颗粒脉动的强弱,在本文中引入了颗粒拟总温的概念,颗粒拟总温综合表征了颗粒平动运动和颗粒旋转运动的脉动强度。结合输运理论建立了考虑颗粒旋转的颗粒相质量、动量和拟总温守恒方程。对颗粒速度分布函数采取Chapman-Enskog近似线性求解方法,求解了在同时具有平动和转动运动时的颗粒相应力、热流通量及能量耗散等参数。并在此基础上提出了颗粒相压力、颗粒相剪切粘度和颗粒相耗散等本构关系式,以及边界条件计算模型。应用粗糙颗粒动理学模型,数值模拟鼓泡床内的气固两相流动特性。以Yuu等和Taghipour等的实验结构尺寸和条件进行数值模拟,模拟结果显示采用粗糙颗粒动理学模型时气泡直径和床层膨胀率均增大。模拟得到了床内颗粒相时均速度和脉动速度分布与Yuu等实测结果相吻合。得到了床内时均颗粒浓度分布、速度分布以及床层膨胀率的大小接近于Taghipour等实测结果。同时通过改变切向弹性恢复系数,分析对比了不同的切向弹性恢复系数下颗粒拟总温、剪切粘度、体积粘度、颗粒压力以及热传导系数随颗粒相浓度的变化规律。应用粗糙颗粒动理学模型,数值模拟提升管内的气固两相流动特性。模拟结果表明提升管内颗粒浓度分布沿床层轴向上稀下浓,沿床层径向中间稀边壁浓的分布。同时在提升管内可以清楚的观测到颗粒团聚物的形成、运动及消失。在高质量流率时,模拟得到了时均颗粒径向浓度、时均颗粒径向质量流率以及床层压降的分布与Knowlton等的实验测量吻合较好。在低质量流率时,模拟得到了颗粒相浓度、速度以及质量流率的分布与Miller等实测结果相吻合。与传统颗粒动理学相比,由于粗糙颗粒动理模型考虑旋转运动造成的能量损失等原因,增加了壁面处的颗粒浓度,而在中心区域浓度降低,沿径向方向颗粒速度变小。同时分析了不同切向弹性恢复系数下提升管内浓度和速度等宏观参数的变化,以及颗粒拟总温、剪切粘度、耗散等微观参数的变化,结果表明不同的切向弹性恢复系数下模拟得到的各参数分布趋势一致。建立了考虑颗粒旋转的粗糙颗粒动理学-煤气化反应的气固流动-反应计算模型,数值模拟流化床煤气化反应和中心射流流化床流动-反应过程。模拟结果与他人实验测量相吻合。模拟结果表明在流化床反应器底部,气固异相燃烧反应生成大量的二氧化碳,一氧化碳和氢气等可燃气体体积浓度非常低。伴随着床层升高氧气被耗尽,还原反应开始占据主导,二氧化碳体积浓度逐渐下降,一氧化碳和氢气等气体体积浓度逐渐增加。结果表明均匀入口流化床反应器床层底部由于燃烧反应使得温度上升比较明显。在床层上部虽然还原反应吸收热量,但是由于床内气泡强烈搅动作用,使得流化床上部温度较为均匀。在中心射流流化床中,中心射流区域形成局部高温区域,边壁区域温度相对较低,床内温度分布的不均匀性有助于床内煤颗粒的氧化-还原过程。更多还原

【Abstract】 Gas-solid two phase flows are widespread in chemical industry, electrical power, metallurgy, food, pharmacy and other fields. In-depth understand and grasp of the flow mechanism of gas and particles is important for both industrial application and fundermental theoretical studies. With the development of computer technology and computational methods, numerical simulation has become one of the most promising tools for researching gas-solid two-phase flow. However, due to the complexity of gas-solid system, theoretical models for numerical simulation of gas-solid flows are still needed in great improvement and enhancement.Various degrees of surface roughness of particles in the gas-solid two phase flows will cause rotation of particles. Experiemnts indicate that the rotation of particles can impact the trajectories and concentration distribution of particles, thus influence other macro and micro parameters. Unfortunately, present kinetic theory of granular flow is based on the assumption of smooth and elastic during the collision of particles, where the rotation of particles does not taken in the collisional processes into account. As a result, it is necessary to improve the kinetic theory of rough particles and its application to gas-solid two phase flows.On the basis of kinetic theory of gases and kinetic theory of granular flow (KTGF), the collisional kinetic theory of rough spheres is proposed considering the rotation of particles. In the traditional kinetic theory of granular flow (KTGF) only the translational granular temperature is considered to describe the translational velocity fluctuations of particles. However, the concept of pseudo-temperature of solid phase is introduced in this work to measure fluctuations of both translational and rotational motions of particles. The balance equations of mass, momentum and pseudo-temperature considering both the translational and rotational motion of articles are derived on the basis of the transport theory. The Chapman-Enskog linear approximation method is adopted to solve the particle velocity distribution. Parameters such as particle stress, heat flux and energy dissipation of both translational and rotational motions are also propoed by solving Bolzmann equation. The constitutive correlations of granular pressure, granular viscosity and particle dissipation as well as boundary conditions are proposed.Flow bahvior of gas and particles in a bubbling fluidized bed is simulated by means of the kinetic theory of rough spheres (KTRS). The motion of bubbles, growing up, merging and breaking up are observed from simulations. The time-averaged velocity and fluctuation velocity distribution are obtained. Simulated results agree well with Yuu’s experimental results measured in a bubbling fluidized bed. The time-averaged concentration distribution in bubbling fluidized bed is also predicted and compared to experiments. Results show that simulations agree with Taghipour’s experimental results. Compared with the simulations by means of the original kinetic theory of granular flow (KTGF), it can be seen that the rotation of particles enhances the non-uniform characteristics in bubbling fluidized bed. Simulations show more bubbles are appeared, and the bobble size is increased. Thus, the bed expansion rate is also increased. Through the change of tangential restitution coefficient, the intensity of rotation of particles is altered. The distribution of particle’s pseudo-temperature, shear viscosity, bulk viscosity, solids pressure and thermal conductive coefficient are obtained as a function of solids concentrations. Simulated results show the pseudo-temperature, shear viscosity, bulk viscosity, solids pressure and thermal conductive coefficient relate with tangential restitution coefficient of particles.Flow behavior of gas-solid two-phase in risers is simulated by means of kinetic theory of granular flow with the consideration of rotation (KTRS). The time-averaged solids concentration and mass flow rate distribution at the high mass flow rate are obtained. The simulated results agree with Knowlton’s experimental results. The predicted axial gas pressure drop coincides with experimental results. The distributions of solids concentration, velocity and mass flow rate are predicted for the low mass flow rate in the riser. Simulations agree with Miller’s experimental results. Compared with simulations by means of the original kinetic theory of granular flow, the energy dissipation computed by means of the kinetic theory of rough spheres is increased, which has a certain effect to improve the results, and also changes the velocity distribution in a riser.The hydrodynamic and reaction of gas and particles phases is simulated by means of the kinetic theory of granular flow (KTRS) with the chemical reaction model in the fluidized bed gasifier. Through simulations, the distributions of temperature and gases components are obtained. Simulations coincide with the experimental measurements. Simulated results show that the chemical reaction in fluidized bed gasifier belongs to a reduction process. The combustion takes place at the bottom of the reactor rapidly. As the distance from the bed bottom increases, oxygen is rapidly exhausted. Then the reduction reaction takes place in a dominant mode. The gases components of hydrogen, carbon monoxide and methene are burnt out after being produced at the bottom of the reactor. After that, the three different gases are produced at the top of the bed where the reduction reaction dominants. The carbon dioxide is generated mainly at the bottom of the reactor through the combustion reaction where more oxygen is introduced from inlet. The volume fraction of gases is decreased because of reduction reaction. From the bed temperature field, it can be seen that the temperature at the bottom of the bed rises rapidly because of the high burning intensity. In the upper part of the bed, the heat absorption rate is low due to the reduction reaction which will absorb a large amount of heat. Hot particles are carried by flue gases through bubbles to the top part of the bed. This makes the temperature is evenly distributed in the whole bed. In the fluidized bed gasifier with a center jet, the local high tempetuere near the inlet and low temperature at the other regime in the bed is found. This non-uniform distribution of temperature promotes the processes of oxidation and reduction in the bed. Such advantage gives a unique property of the bubbling fluidized bed gasifier with a center jet used in the coal gasification.更多还原