论文总字数：40892字

摘 要

随着工业化的进步，能源需求量的骤然增长与减少污染物排放的问题越加突出。在这种形势下，开发研究一种高效且环境友好的燃烧技术迫不及待。

本文围绕氧燃料无焰燃烧状况的仿真与实验测试结果，通过对比分析，得到对氧燃料无焰燃烧状况的燃烧特征与污染物排放特征的进一步了解。本课题所研究炉膛为200千瓦的半工业燃烧炉，燃烧状况为氧燃料无焰燃烧，燃料为液态丙烷（LPG），氧化剂为工业纯氧。燃烧器为市场有售的REBOX-W燃烧器。尽管无焰燃烧技术已经在不同工业领域有了较为普及的应用，关于氧燃料无焰燃烧的燃烧特性的研究仍是比较有限的。本文中，所用湍流模拟模型为标准 模型， 燃烧模型为有限速率/涡耗散模型。本课题的主要目标是使用计算流体力学分析研究辐射模型对无焰燃烧仿真结果的分析。本文共研究一个灰体气体模型与五个非灰体气体模型。仿真结果与实验结果对比并分析。同时讨论了不同模型及其仿真结果的优缺点和与实验测量结果的拟合程度，最后对氧燃料无焰燃烧的特征进行研究，并提出最佳模型。

关键词：

无焰燃烧；辐射模型；模拟；

List of Figures

Figure 1 Location of measurement points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Figure 2 Oxygen and fuel inlets configuration at the newly developed burner REBOX®-W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Figure 3 (a), (b), (c), (d) Temperature profile in the furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Figure 4 (a), (b), (c), (d) Oxygen Volume Fraction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Figure 5 (a), (b), (c), (d) CO Volume Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Figure 6 CO volume fraction at horizontal plane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Figure 7 Contour of temperature at horizontal plane, Grey model (left), Non grey model (right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

Figure 8 Streamline at horizontal plane (left) and vertical plane (right). . . . . . . . . . . .19

Contents

Chapter 1. Introduction 1

Objective 2

Chapter 2. Literature Review 3

2.1 Oxy-fuel combustion 3

2.2 Flameless combustion 3

2.3 Non grey gas radiation models 4

2.4 Suction pyrometer 4

2.5 Local gas composition 5

Chapter 3. Experiments and Mathematical Formulation 6

3.1 Turbulence Modeling 6

3.2 Combustion Modeling 6

Finite Rate/Eddy Dissipation model 6

3.3 Radiation 7

3.3.1 Simple grey gas model (SGG) 8

3.3.2 Leckner model 8

3.3.3 Weighted-sum-of-gray-gases model (WSGG) 10

3.4 Boundary Conditions 11

3.5 Experimental Set-Up and Procedures 11

Chapter 4 Results and Discussions 13

4.1 Models validation 13

4.1.1 Temperature in furnace 13

4.1.2 Volume fraction 14

4.1.3 Volume fraction 16

4.2 Effect of models for gaseous radiative properties 17

4.3 Combustion performance 17

4.3.1 Flame configuration 17

4.3.2 Temperatures in flame 18

4.3.3 Flow and recirculation 19

Chapter 5 Conclusion 21

Chapter 6 Acknowledgement 22

Chapter 7 References 23

Chapter 1. Introduction

With the progress of industrial process, energy saving and emission reduction is more and more becoming a topical issue. On the other hand, price of fossil fuels staying at high level has become another urge motivation to develop a new combustion technology that can improve the efficiency. Oxy-fuel combustion is a new combustion technology compared to the traditional air-fuel combustion. Oxy-fuel combustion can improve the furnace efficiency and reduce the pollution emission by using industrial oxygen as oxidizer instead of air.

But for traditional oxy-fuel combustion technology, there are several non-negligible issues. Although the oxidizer has been replaced by oxygen, which separate the nitrogen from the combustion process, due to the chamber of the furnace is maintained at a micro-negative pressure, the nitrogen carried in the leak-in air is still a potential source for NOx production. Apart from this, in oxy-fuel combustion, the high oxygen concentration is can create a higher flame temperature, which lead to a higher emission of thermal NOx. The large temperature gradient in will also causes uneven heat transfer and higher emission.

Then another concept of combustion, flameless oxy-fuel combustion is proposed. For flameless oxy-fuel combustion, with internal product recirculation, the oxygen concentration as well as fuel concentration is diluted by high-temperature inert flue gas. Then a combustion condition of low oxidizer concentration and uniform temperature distribution is created. Since the furnace temperature becomes higher than the auto-ignition temperature of the mixture of oxidizer and fuel, the combustion is started autonomously. The flame of flameless combustion is nearly invisible and the peak flame temperature is much lower than the traditional oxy-fuel combustion. For traditional oxy-fuel combustion, completed and advanced investigations have been done and several well-designed models have been proposed. The international flame research foundation has been investigated the combustion performance of high temperature air combustion with experiment since 1990s^[1]. There are also some succeeding experiments concerned with the performance of flameless combustion. But for the combustion mechanism of oxy-fuel flameless combustion, especially when concerned the radiation model, there is few studies of oxy-fuel flameless combustion with CFD simulation.

In the present work, the transfer of radiant energy was obtained from the solution of the radiative transfer equation (RTE) as described in the coming sections^[1]. The radiative transfer equation is of the integral differential type. This equation is complex and, therefore, difficult to be exactly solved in 3D geometries. The following simplifications to this equation were therefore, incorporated. Since the scattering in the combustion of gas fuel has an insignificant effect, it can be neglected in the RTE.

Objective

The objective of this project is to investigate the influence of radiation models on simulation results by CFD. In this study, a grey gas model and five non grey gas model are used and their combustion behaviors are compared with experimental data. For oxy-fuel flameless combustion, there is no special non grey gas radiation model designed for this combustion condition, so in this study, five non grey gas models suitable for oxy-fuel combustion are adopted. They are a simple grey gas model, two Leckner models and two weighted sum of grey gas models adopt different radiative property database. These non grey gas models have been validated under the oxy-fuel combustion condition and performed well. The results of different models are analyzed and an optimum model is proposed.

Chapter 2. Literature Review

2.1 Oxy-fuel combustion

In oxy-fuel combustion, the industrial oxygen is used as oxidizer instead of air. In traditional air-fuel combustion, the nitrogen in air is only works as inert gas but can lead to the NOx emission. Currently, many research has been done in pure oxygen combustion or oxygen-enrichment combustion.

The industrial oxygen is normally separated by cryogenic or membrane technology^[1]. Economically the process obtaining the industrial oxygen require extra energy, which costs much more than traditional air-fuel combustion. But there are still several advantages by adopting oxy-fuel combustion technology. Because the nitrogen is separate from the input oxidizer, the mass and volume of the flue gas are reduced by approximately 75%. Due to the reduction of flue gas volume, the heat loss in flue gas is reduced simultaneously. Another distinct characteristic of oxy-fuel combustion is the production of nitrogen oxide is reduced dramatically.

The further study of oxy-fuel combustion including mixing rate, flame configuration, gaseous species composition, blow off velocity, noise, emission and etc. And with a trumpet construction, the internal flue gas recirculation within the chamber will lead to a significant temperature rise. ^[1] And the NOx emission is influenced by the mixing rate of oxygen and fuel as well as injection angle and velocity.

2.2 Flameless combustion

In flameless combustion, the reactants are distributed in a large volume in the furnace and the furnace temperature is above the self-ignition temperature. Then the combustion take place spontaneously and there is no flame front as well as visible flame configuration^[1]. Due to the reactants are diluted by the internal recirculation of flue gas, works as inert gas, the peak combustion temperature is lower than oxy-fuel flame combustion. Although the nitrogen is separate from the inlet oxidizer in oxy-fuel combustion, due to the furnace is maintained at a micro-negative pressure, -50Pa, the nitrogen carried in the leak-in air is non-negligible. So the lower peak temperature reduced the production of thermal nitrogen oxide greatly compared with oxy-fuel combustion.

Visually, the flame of the oxy-fuel flameless combustion represents an almost invisible light blue flame. And due to the high velocity injection of oxidizer and fuel with low temperature, the local gaseous species concentration of internal flue gas are diluted and the peak temperature is lowered. ^[1] Several other expressions for flameless combustion techniques are: HiTAC (High Temperature Air Combustion), mild combustion, dilute combustion, FDI (Fuel Direct Injection) and etc^[1]. The flameless combustion using pure oxygen as oxidizer is studied in this project.

To achieve the oxy-fuel flameless combustion, the injection velocity of both oxidizer and fuel are both set at a high level. Nowadays, different flameless combustion burners are commercial available. In this study, a REBOX-W burner is used. According to the previous studies, this burner can provide a more uniform diluted reactants distribution and temperature profile, better internal flue gas recirculation and lower noise.

2.3 Non grey gas radiation models

In this study, a grey gas model and five non grey gas model are used and their combustion behaviors are compared with experimental data. For oxy-fuel flameless combustion, there is no special non grey gas radiation model designed for this combustion condition. So five non grey gas models suitable for oxy-fuel combustion are adopted. These models are all calculate the emissivity of gas mixture based on the radiative transfer equation (RTE). Simple grey gas model is the simplest model and mainly based of empiricism data, and in this study this model works as an reference model. Two Leckner models developed by B. Leckner are adopted. In the Leckner model, the calculation of total emissivity of carbon dioxide and water molecules consist of a summation of emissivity of carbon dioxide and water molecules respectively and then minus a correction term. And two weighted sum of grey gas models adopt different radiative property database are also used. The weighted sum of grey gas model is the most popular model today and used suitable for many cases. One of the weighted sum of grey gas model is provided by CFD software and the other model is a refined weighted sum of grey gas model by Yin et al. These non grey gas models have been validated under the oxy-fuel combustion condition and performed well. The results of different models are analyzed and an optimum model is proposed.

2.4 Suction pyrometer

Normally, the temperature inside the furnace can be measured by the suction pyrometer. The suction pyrometer has several specific advantages over other devices. One is the furnace temperature measured by the suction pyrometer is at a high precision level and not affect by the inside furnace radiation^[1]. Regularly, the probe of suction pyrometer is made of stainless steel and ceramic pipe. The ceramic cover acts as a radiation shield for the thermocouple. In order to increase the heat transfer to the temperature sensor, the suction pyrometer is inserted in the combustion chamber. A under pressure is created with a fan connected to the suction pyrometer, then the gases in furnace are sucked through the ceramic tube at a very high velocity. The flow rate of cooling water is 2.5 . And the accuracy of the thermal couple is 8 to 1300 . And the higher the gas suction velocity is, the more accurate the measured temperature will be.

2.5 Local gas composition

The inside furnace gas sample is carried with a cooling water probe in order to measure the inflame gaseous species composition. During the measurement, the gas sample is cooling down rapidly to avoid further chemical reaction and then be sent to the micro-GC^[1]. The micro-GC consists of four independently controlled modules, which can provide a combination of speed and analytical accuracy. One GC module consists of an injector, capillary sample and reference columns, a column heater and a thermal conductivity detector. The micro GC is calibrated using known composition of calibration gas.

Chapter 3. Experiments and Mathematical Formulation

3.1 Turbulence Modeling

In this study, a standard model proposed by Launder and Spalding is used to modeling the turbulent flow. This model is a semi-empiricism model and it is widely used today due to its proper compromise between accuracy and complexity. The standard model based on several consumptions that a fully turbulent flow is considered, and the molecular viscosity is neglected. During the combustion process in this study, the injection velocity of fuel and oxidizer are both supersonic and the flow in chamber is almost fully turbulent, so this model is well-fitted with the combustion condition in this case.

3.2 Combustion Modeling

Finite Rate/Eddy Dissipation model

In this project, the finite-rate/eddy-dissipation model is used to simulate the combustion process. In this model, the reaction rate, depends on the smaller one of the Arrhenius rate and the mixing rate.

(3.1)

Where refers to the Arrhenius rate and the mixing rate respectively.

In turbulence flows, the mixing rate is governed by the eddy dissipation rate. ^[1] The mixing rate is defined by the expression Eq. (3.2):

(3.2)

Where k is the turbulent kinetic energy and is the dissipation rate.

The value of mixing rate is calculated in the following approach.

(3.3)

Where ,, is the mass fraction of fuel, oxidizer and product respectively. A and B are empirical constants relate to mixing rate.

The finite rate reaction model in finite-rate/eddy-dissipation model considers the reaction rate as difference between forward reaction and backward reaction:

(3.4)

Where are the forward and backward reaction rate respectively.

The forward and backward reaction rate are computed as Eq. (3.5):

(3.5)

Where A is the pre-exponential factor, is the activation energy, n is the temperature exponent. In this study, previous finite-rate parameters are defined based on the result of Charles K. Westbrook^[1].

In this study, a two-step reaction mechanism is used. ^[1]

(3.6)

(3.7)

This project is under the conditions of oxy-fuel flameless combustion. In this case, the turbulence mixes fuel and oxidizer quite slowly when compared to the burning rate. So the reaction process can be chemical kinetics negligible and controlled by mixing rate of fuel and oxidizer. But before the computation reaches to the steady state, the Arrhenius rate might be smaller than the eddy-dissipation rate and governs the reaction rate.

3.3 Radiation

In this work, the discrete ordinates (DO) radiation model is used, which requires the absorption coefficient as in put. So the radiative transfer equation (RTE) is introduced to compute emitting and absorbing at position in the direction. Eq. (3.8) is the RTE presented by R. Viskanta^[1].

(3.8)

In which is the spectral intensity, are absorption coefficient and scattering coefficient respectively, is the phase function that represents the probability of frequency propagating, and is the emission coefficient based on the assumption of Kirchhoff’s law that can expressed as Eq. (3.9):

(3.9)

3.3.1 Simple grey gas model (SGG)

The simple grey gas model, as indicated from its name, is the simplest approach to simulate the radiative properties of gases. This model assumes the effective absorption coefficient to be independent of the frequency of radiation and as the single parameter to represent the radiative properties of gas mixture^[1].

Under the grey gas assumption, the scattering of radiation is negligible, then the RTE for the radiation intensity in three-dimensional Cartesian coordinates is simplified as Eq. (3.10):

(3.10)

A good estimation of the effective absorption coefficient from the known properties (a mean beam length and a characteristic gas temperature^[1]) of emission by the gas can be obtained from interpretation of the total emissivity of the gas upon the bounding surface.

3.3.2 Leckner model

This model is developed by B. Leckner^[1] in 1972. In this model, the total emissivity of the gas mixture is equal to the sum of emissivity of carbon dioxide and water vapor minus a correction term due to the overlap in some special regions.

(3.11)

In Leckner model, the general form of the pressure correction correlation is given by Eq. (3.12):

(3.12)

Where is the emissivity of either or and is the pressure path length (bar.cm) and the effective pressure respectively for and is given in REF^[1].

According to an empirical relationship Eq. (3.13):

(3.13)

Then the pressure correction relationship may be normalized according to Eq. (3.14).

(3.14)

Neglecting the temperature and pressure dependence left and assuming that the function of path length, is expressed as follows:

剩余内容已隐藏，请支付后下载全文，论文总字数：40892字