湍流稀液雾两相流燃烧的大涡模拟和火焰面模型相关问题研究

发布时间：2018-11-13 11:24

【摘要】：稀液雾燃烧工程应用广泛,深受研究者关注。随着计算能力的增强和数值模型的不断优化,数值模拟在研究稀液雾燃烧机理、指导燃烧器设计等方面发挥着越来越重要的作用。大涡模拟(LES)方法能够求解湍流中的大尺度含能结构,基于火焰面的模型能够耦合详细反应机理并有很大潜力预测非稳态过程,这两种方法在稀液雾燃烧中的应用势在必然。然而把传统气相燃烧模型推广到两相流中需要对模型改进以考虑更多的物理过程。改进相关模型并研究其应用范围、探究模型的理论基础和分析流场中气液相互作用等是相关工作的热点。基于此,本文对火焰面模型在稀液雾燃烧中的应用进行了相关研究,主要工作如下：首先,基于非预混的火焰面模型和LES方法对CH4/H2/N2湍流射流扩散火焰进行模拟,对比稳态火焰面(SFM)模型和火焰面/进度变量(FPV)模型在气相燃料燃烧中的性能。发现两者都较好预测流场中一阶矩和二阶矩,而后者对二阶矩的模拟更准确。同时两种模型对小浓度组分(小组分)的预测存在误差。如果单独求解小组分输运方程可较好解决这个问题。流场的本征正交分解(POD)分析发现前5阶模态占有总湍流能量的60%左右,而中心射流区域集中着各尺度的湍流相干结构。随后通过对有回流区稀液雾受限射流、自由射流湍流流场的大涡模拟来研究流场中气液相互作用。有回流区的稀液雾受限射流中发现Stokes数大的液滴弥散受其初始惯性影响较大,而Stokes数在1附近的小液滴分布在大涡结构的外缘,弥散受气相流场结构的影响更大。冷态条件下,液雾单位体积蒸发速率和液雾弥散分布密切相关,而单液滴蒸发速率受液滴滑移速度影响明显。这两方面的工作验证了FPV模型的有效性并分析了冷态条件下的液滴弥散和蒸发作用。其次,对均匀各向同性湍流下正庚烷液滴自点火过程进行直接数值模拟(DNS)研究,评估不同局部放热指标的性能,分析燃烧机制。DNS中采用详细反应机理并直接求解组分输运方程。通过对时间演化过程、瞬时图、散点图和拟合直线的分析发现[OH]×[CH2O]在自点火火焰中作为火焰放热指标是合适的。不同火焰机制时[OH]×[CH2O]和局部放热速率基本满足正比关系,但比例系数稍有变化。再则,在欧拉-拉格朗日架构下采用FPV模型和LES方法模拟了乙醇-空气稀液雾驻定火焰,并提出在火焰面架构中考虑蒸发吸热效应的修正模型。修正模型一种是基于修改层流火焰面,一种则侧重在流体计算中考虑能量平衡。对修正模型的对比发现,提出的降低火焰面温度的N-FPV模型和在流场计算中考虑能量平衡的T-FPV模型都可以有效地模拟蒸发吸热,同时看到模型对气相和液相的统计参数具有很好的预测性能。瞬时图分析发现在x/D=20上游FPV模型也能给出部分点火、熄火现象,而较强的蒸发作用发生在剪切层和x/D=20附近高温区。数值结果表明,蒸发速率分布受液滴密度和气相温度的共同影响。然后,采用LES方法耦合FPV模型对甲醇和正庚烷燃料高温伴流中的自点火过程进行了模拟,验证火焰面模型对自点火预测特性。为使FPV模型适用存在三个热化学进口状态(液相燃料、载气和高温伴流)的燃烧系统,本文引入新的守恒标量对FPV模型进行扩展。数值结果表明,新提出的三进口状态模型的火焰瞬时图像能合理地预测温度和组分质量分数分布,尤其在喷嘴出口附近,较传统双进口系统具有更好的预测特性。然而要有效地预测稀液雾燃烧中的点火过程需解决抬举高度被高估的问题。随后,本文采用自点火模型耦合新提出的三进口状态模型对稀液雾抬举火焰进行了模拟,得到的一系列火焰抬举高度和统计量,与实验值符合良好,说明自点火模型适合这类火焰。对模拟结果分析可发现,随着液相燃料质量流率的增加火焰抬举高度下降,点火首先发生在贫燃、标量耗散率小的位置。自由基OH往往出现在火焰生成点的上游。在火焰生成点上游,中心射流和高温伴流的剪切层中存在少数小点火核心,是点火的预兆。最后,针对反应进度参数的定义没有考虑重油燃料的热解等情况,提出反应进度参数的新定义。采用主成分分析(PCA)这一数学工具提出在基于火焰面的燃烧降维处理中优化反应进度参数定义。对预混火焰面数据库应用PCA和核密度加权PCA (KEDPCA)并依据前两个主元得到了优化的反应进度参数定义。通过模拟/重构一个CH4/N2/空气抬举火焰验证新定义的有效性。数值结果表明,新定义可以保证热化学变量和反应进度参数之间的单调性。在层流抬举火焰的模拟过程中发现优化的反应进度参数比传统的定义有更好的表现。温度和OH质量分数的云图说明基于PCA的优化定义在高温区表现得更好。R2统计系数说明除部分小组分外,KEDPCA结果比PCA稍好但差距不明显。KEDPCA方法的最主要优势体现在它受原始数据库中插值等操作的影响很小,PCA则对原始数据库中的状态变化更敏感。通过研究,本文给出反应进度参数或降维参数定义的普遍准则。
[Abstract]:The application of the thin-liquid fog combustion engineering is widely used, and it is deeply concerned by the researchers. With the enhancement of the computing power and the continuous optimization of the numerical model, the numerical simulation plays a more and more important role in the research of the mechanism of the lean liquid fog and the design of the burner. The large eddy simulation (LES) method can solve the large-scale energy-bearing structure in the turbulent flow, and the model of the flame surface can be coupled with the detailed reaction mechanism and has a great potential to predict the unsteady process. however, that extension of the conventional gas-phase combustion model to the two-phase flow requires the improvement of the model to take into account more physical processes. To improve the relevant model and to study the application range of the model, the theoretical basis of the model and the interaction of the gas and liquid in the flow field are the hot points of the related work. Based on this, the application of the flame surface model in the lean liquid fog combustion is studied. The main work is as follows: First, the diffusion flame of CH4/ H2/ N2 turbulent jet is carried out based on the non-premixed flame surface model and the LES method Simulation, comparison of steady-state flame surface (SFM) model and flame surface/ progress variable (FPV) model in gas-phase fuel combustion It is found that both the first and second moments in the flow field are better predicted, and the latter is more suitable for the simulation of second-order moments. Accurate. The prediction of the small-concentration component (group) by the two models Error. If a separate solution is used to solve the sub-transport equation, this problem can be solved well The orthogonal decomposition (POD) analysis of the flow field shows that the first 5-step mode occupies about 60% of the total turbulent energy, while the central jet region is concentrated on the turbulence coherence of each scale. in that flow field, the gas-liquid in the flow field is study by the large-eddy simulation of the turbulent flow field of the free-jet turbulent flow through the dilute-liquid-fog-restricted jet with the return region, The small droplets in the vicinity of the Stokes number are distributed in the outer edge of the large-vortex structure and the dispersion is influenced by the structure of the gas-phase flow field. Under the condition of cold state, the volume evaporation rate of the liquid fog unit is closely related to the dispersion distribution of the liquid fog, and the evaporation rate of the single liquid drop is affected by the speed of the liquid drop sliding speed. It is clear that the effectiveness of the FPV model is verified and the dispersion and evaporation of the droplets in the cold state are analyzed. in that follow, the direct numerical simulation (DNS) study on the self-ignition of the n-heptane liquid drop under the homogeneous isotropic turbulence is carried out, and the performance of different local heat release indexes is evaluated. Analysis of combustion mechanism. The detailed reaction mechanism is used in DNS and the group is solved directly. By analyzing the time evolution process, the instantaneous diagram, the scatter diagram and the fitted straight line, the[OH]-[CH2O] is used as the flame heat-releasing finger in the self-ignition flame. The scale is suitable.[OH][CH2O] and the local heat release rate at different flame mechanisms basically satisfy the proportional relationship, but the scale system In the Euler-Lagrange architecture, the FPV model and LES method are used to simulate the flame of the ethanol-air thin liquid fog, and the evaporation and suction are considered in the flame surface structure. The modified model of the thermal effect. The modified model is based on the modified laminar flame, one on the fluid meter. The energy balance is considered in the calculation. The comparison of the modified model shows that the N-FPV model for reducing the temperature of the flame surface and the T-FPV model considering the energy balance in the flow field calculation can effectively simulate the evaporation and heat absorption, and the statistical parameters of the model to the gas phase and the liquid phase are also observed. The instantaneous graph analysis shows that the FPV model in the x/ D = 20 can also give partial ignition and flame-out, and the stronger evaporation takes place in the shear layer and the x/ D. = 20 high temperature area. Numerical results show that the evaporation rate distribution is affected by droplet density and gas The self-ignition process of methanol and n-heptane fuel high-temperature co-flow is simulated and the flame is verified by using LES method to couple the FPV model. In order to make the FPV model applicable to the combustion system of three thermochemical import states (liquid fuel, carrier gas and high temperature with flow), a new conservation standard is introduced in this paper. The numerical results show that the flame transient image of the newly proposed three-inlet state model can reasonably predict the distribution of temperature and component mass fraction, especially near the outlet of the nozzle. The system has better prediction characteristics. However, it is necessary to effectively predict the need of the ignition process in the lean-liquid-fog combustion. In this paper, a new three-inlet state model is used to simulate the high-lift flame of the thin-liquid fog, and a series of flame lifting heights and statistics are obtained, which are in good agreement with the experimental values. The self-ignition model is suitable for this kind of flame. The analysis of the simulation results shows that, with the increase of the mass flow rate of the liquid fuel, the lifting height of the flame is decreased, and the ignition first occurs in the lean state. the position where the rate of the scalar dissipation is small. the free radical OH tends to occurs upstream of the flame generation point. A small number of small shear layers upstream of the flame generation point, the central jet, and the high temperature tracing flow The ignition core is an omen of ignition. Finally, the definition of the reaction progress parameter does not take into account the conditions of the pyrolysis of the heavy oil fuel In this paper, a new definition of the reaction progress parameter is put forward. The main component analysis (PCA) is used to put forward the combustion drop based on the flame surface. The definition of the optimization reaction progress parameter in the dimensional process is defined. The PCA and the kernel density weighted PCA (KEDPCA) are applied to the pre-mixed flame surface database and the two main elements are based on the first two main elements. The optimized reaction progress parameter is defined. A CH4/ N2/ air is simulated/ reconstructed The effectiveness of the new definition of the gas lifting flame verification. The numerical results show that the new definition can ensure the thermochemical change Monotonicity between the amount and the reaction progress parameter. The optimized reaction progress was found during the simulation of the laminar flow-up flame. The parameter is better than the traditional definition. The image of the temperature and OH mass fraction is based on the PCA The optimization definition is better in the high temperature area. The R2 statistics show that, in addition to the partial group, the KEDPCA The result is better than PCA but the difference is not obvious. The most important advantage of the KEDPCA method is that it is affected by the interpolation and other operations in the original database. The state changes in the original database are more sensitive. By studying, this paper gives the progress of the reaction
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TK16

【参考文献】