搅拌槽微观混合的数值模拟研究

发布时间：2018-03-19 12:41

本文选题：微观混合　切入点：搅拌槽反应器　出处：《中国科学院大学(中国科学院过程工程研究所)》2017年博士论文　论文类型：学位论文

【摘要】：搅拌槽作为工业上广泛使用的混合设备,其内部的流动和传递特性将直接关系到产品的特性以及生产过程的经济性。对于混合敏感的快速复杂反应过程,如果加入搅拌槽的物料无法快速地实现分子尺度的混合,那么进料位置处反应物的局部混合状态将会决定主产物收率、产品质量以及操作稳定性。因此,深入了解搅拌槽内各尺度的混合特性,尤其是直接影响化学反应进程的微观混合,将有助于工业规模反应器的设计、工程放大、操作优化及过程强化。基于此,本文采用数值模拟方法对单相以及多相搅拌槽内湍流反应流进行了系统的研究,考察了微观混合对混合敏感的化学反应体系选择性的影响,并将模型方法应用到实际反应体系及其现象的分析。具体工作以及主要成果如下:(1)使用混合分数及其方差描述了物料宏观以及微观尺度的离集状况,提出了一种CFD耦合卷吸模型(E-model)的新方法,数值研究了单进料、半连续搅拌槽内微观混合对酸碱中和/氯乙酸乙酯水解平行竞争反应体系以及碘化物/碘酸盐平行竞争反应体系选择性的影响。结果表明,本文提出的CFD耦合卷吸模型的新方法可以很好地预测出离集指数随搅拌转速、进料位置、进料浓度等条件的变化,且该方法的模型方程简单、计算时间短,不需要任何实验数据作为模型参数,适用于工业规模反应器的诊断和优化。(2)基于"Eulerian-Eulerian"的多流体模型观点,将本文提出的适用于均相体系的CFD耦合卷吸模型的新方法进行了扩展,数值研究了气液以及固液搅拌槽内微观混合对碘化物/碘酸盐平行竞争反应体系选择性的影响。采用欧拉多相流模型以及k-ε多相湍流模型数值计算流场,在模拟气液宏观流场时还添加了可变气泡尺寸模型。结果表明,针对多相体系扩展得到的CFD耦合卷吸模型的新方法,可以较好地预测出离集指数随惰性相相含率、搅拌转速、进料位置等条件的变化。对于气液搅拌槽,增加气速可以显著增加液面附近流体的湍动程度,使得每一份进料的消耗时间缩短,但对于靠近桨叶附近的进料位置,气速对离集指数的影响较小;对于固液搅拌槽,相含率很高时形成固体云,在清液层进料时,离集指数显著增大。(3)基于CFD耦合卷吸模型,通过求解平均混合分数及其方差的输运方程,数值研究了搅拌槽进料管的返混,定性地描述了返混发生时进料管出口附近物料流动和混合特征。结果表明,当进料管发生返混时,从速度矢量图上可以清晰看到管口处产生旋涡,且旋涡尺寸随着返混程度的增加而增大,返混最严重时旋涡占据了整个进料管出口;从平均混合分数及其方差分布图可以看到,返混发生时,进料管出口内侧混合分数值明显小于1,方差最大值所在的区域也位于进料管出口内,随着进料速度增加或者进料管直径的减小,返混逐渐减弱,方差最大值所在的区域逐渐由管内侧转移到管口外;不发生返混时,方差最大值所在的区域即离集大的区域则位于管口外侧。(4)将 CFD 耦合 DQMOM-IEM(Direct quadrature method of moments combining with the interaction by exchange with the mean micro-mixing model)微观混合模型用于研究搅拌槽内微观混合对酸碱中和/氯乙酸乙酯水解平行竞争反应体系产物分布的影响。计算结果表明:DQMOM-IEM微观混合模型可以成功地预测离集指数随搅拌转速以及进料时间的变化。搅拌转速越高,流体湍动程度越大,充分的微观混合抑制了副产物的生成;进料时间越长,加入搅拌槽内的新鲜物料越容易与搅拌槽内反应物充分混合,反应区体积越小且越集中于进料管出口,离集指数也越小。(5)采用CFD耦合DQMOM-IEM微观混合模型,数值计算了搅拌槽内的反应PLIF(Planar laser-induced fluorescence)过程。针对该真实反应体系的特征,详细推导了以混合分数和反应进度变量所表示的化学反应过程,考察了搅拌桨转速以及安装高度对同时进行的混合和快速化学反应过程的影响。该方法不仅适用于反应PLIF过程的计算,还可以数值模拟其他混合敏感的化学反应体系。模拟结果表明:在获得宏观流场的基础上,通过求解两环境DQMOM-IEM微观混合模型,成功地预测了反应PLIF过程,获得了荧光示踪剂罗丹明B浓度的时空分布。随着搅拌转速的增加,物料混合速率加快,物理混合时间以及反应混合时间均减小;改变搅拌桨的安装高度,搅拌槽内不同位置的湍动程度以及流动主体方向均会发生相应变化,从而影响了反应混合进程。此外,如果在数值计算中不耦合微观混合模型,则预测的反应混合时间远小于实验值以及使用DQMOM-IEM模型获得的模拟值,因此在对混合敏感的反应体系进行模拟时,微观混合作用显著,添加微观模型可以获得更加准确的预测结果。
[Abstract]:Mixing equipment in stirred tank is widely used as the industry, its internal flow and transfer characteristics will be directly related to the characteristics of products and production process of the economy. The rapid reaction process for mixed complex and sensitive, if adding stirring tank material can not achieve rapid mixing into scale, local mixing state will then feed the position of the reactants decided the main product yield, product quality and operation stability. Therefore, in-depth understanding of the mixing characteristics of each scale in a stirred tank, especially micro mixing directly affects the chemical reaction process, will be helpful to design, industrial scale reactor scale-up, operation optimization and process intensification. Based on this, this paper by using the numerical simulation method of single-phase and multiphase turbulent reacting flow in stirred tank were studied. Effects of micromixing on mixed sensitive chemical reaction system. Effect of selective, and the analysis model is applied to the actual reaction and phenomena. The specific work and main results are as follows: (1) using the mixture fraction and its variance is described from the set of materials macro and micro scale, proposes a CFD coupled entrainment model (E-model) of the new method, numerical study the single feed semi continuous stirred tank micro mixing of neutralization / ethyl chloroacetate hydrolysis parallel reaction system and iodide / iodate parallel competition selective reaction system. The results show that the CFD coupling the volume of new method of absorbing model can well predict the segregation index with stirring speed change, feed position, feed concentration and other conditions, the model equation and the method is simple, the computation time is short, do not need any experimental data as the model parameters for industrial scale reaction diagnosis and device Optimization. (2) based on the "Eulerian-Eulerian" model, a new method of CFD coupled to a homogeneous system entrainment model is proposed in this paper will be extended, numerical study of gas-liquid and Solid-Liquid Stirred Tank micro mixing of iodide / iodate parallel competitive selective reaction system. Using the European pull multiphase flow model and k- e turbulent multiphase numerical model in the simulation of gas-liquid flow field, macroscopic flow field also adds a variable bubble size model. The results show that the new method for CFD coupled system extended entrainment model, can better predict the segregation index with inert phase holdup, stirring change speed, feed location and other conditions. For the gas-liquid stirred tank, increase in gas velocity can significantly increase the level of fluid near the turbulence degree, makes every feed consumption time is shortened, but close to the blade of Feed location near, gas velocity has little effect on the segregation index; for Solid-Liquid Stirred Tank, phase holdup is very high in the clouds to form a solid, liquid feed layer, the segregation index increased significantly. (3) CFD coupled entrainment model based on the average and variance of the mixture fraction by solving the transport equation numerical study on the mixing tank, feeding back mixing tube, qualitatively describes the backmixing occurs when the feed pipe near the exit of material flow and mixing characteristics. The results show that when the feed back mixing tube, from the velocity vector diagram, you can clearly see the vortex nozzle, and the vortex size with backmixing the extent of the increase, mixing the most serious vortex occupies the entire feed pipe outlet; from the mean mixture fraction and its variance distribution map can be seen, backmixing occurs when the feeding pipe of the inner side of the outlet of mixed value is less than 1, the maximum variance region is located in the feed The outlet, with the decrease of feed rate increase or feed pipe diameter, mixing gradually weakened, the variance maximum value area gradually from inside the tube to tube mouth; no backmixing, variance maximum region from the region is located in the mouth outside the master. (4) CFD DQMOM-IEM (Direct quadrature method of coupled moments combining with the interaction by exchange with the mean micro-mixing model) model is used to study the micro mixing in a stirred tank of micromixing on neutralization / ethyl chloroacetate hydrolysis reaction system affect the parallel competing product distribution. The calculation results show that the DQMOM-IEM micro hybrid model can successfully predict the segregation index change the stirring speed and the feeding time. The stirring speed is higher, the greater the degree of fluid turbulence, micro mixing fully inhibited the formation of by-products; feed The longer, stirring fresh materials groove more easily fully mixed and stirred tank reaction, the reaction zone and the smaller and more concentrated in the feed outlet, the segregation index is smaller. (5) using CFD coupled DQMOM-IEM micro hybrid model, numerical calculation of the mixing reaction tank PLIF (Planar laser-induced fluorescence). According to the characteristics of the real reaction system, deduced the chemical reaction with the mixture fraction and the reaction progress variable is represented, the effects of stirring speed and installation height of mixed and rapid simultaneous reaction process. The calculation method is not only applicable to the reaction process of PLIF also, the numerical simulation of other mixed sensitive chemical reaction system. The simulation results show that: Based on the macroscopic flow field obtained by solving two DQMOM-IEM micro environment, hybrid model successfully predicts the reverse PLIF process, temporal and spatial distribution of fluorescent tracer rhodamine B concentration was obtained. With the increase of stirring speed, mixing speed, mixing time and mixing time of physical reaction decreased; change the installation height of the agitator tank in different position of the turbulence degree and the flow direction of the main body will change, thus affecting the the reaction mixture process. In addition, if not in the calculation of micro mixing in the numerical model of coupling reaction, mixing time prediction is much smaller than the experimental values and simulation values obtained using the DQMOM-IEM model, so the simulation of reaction system of mixed sensitive, micro mixing effect, adding micro model can obtain more accurate forecasting results.

【学位授予单位】：中国科学院大学(中国科学院过程工程研究所)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TQ051.72

【参考文献】