氢气氧化详细动力学机理建模与宽工况验证研究（h2_paper）

3月前浏览1372

摘要：

本文构建了适用于宽工况的氢气氧化详细动力学模型，涵盖298–2700 K、0.05–87 atm、当量比 0.2–6，包含 19 步基元反应，优化了H+O₂+M、H+OH+M等关键反应速率与热力学参数。通过激波管点火延迟、常压 / 高压层流火焰速度、稳焰组分分布及流动反应器浓度剖面系统验证，模型与实验高度吻合。灵敏度分析揭示H+O₂支链、HO₂/H₂O₂反应、三体复合是控制不同工况的核心路径。该机理优于 GRI-Mech、Konnov 等模型，为氢燃烧数值模拟提供高精度基础子机理。

A Comprehensive Modeling Study of Hydrogen Oxidation

MARCUS CONAIRE1 , HENRY J. CURRAN2 , JOHN M. SIMMIE1 , WILLIAM J.

PITZ3 , CHARLES K. WESTBROOK3

1 National University of Ireland, Galway, Ireland

2 Galway-Mayo Institute of Technology, Galway, Ireland

3 Lawrence Livermore National Laboratory, Livermore, CA 94551

Correspondence to: Henry Curran; e-mail:henry.curran@nuigalway.ie

ABSTRACT: A detailed kinetic mechanis m has been developed to simulate the combustion of H2/O2 mixtures, over a wide range of temperatures, pressures and equivalence ratios. Over the series of experiments numerically investigated, the temperature ranged from 298 to 2,700 K, the pressure from 0.05 to 87 atm, and the equivalence ratios from 0.2 to 6.

Ignition delay times, ﬂame speeds and species composition data provide for a stringent test of the chemical kinetic mechanis m, all of which are simulated in the current study with varying success. A sensitivity an alysis was carried out to determine which reactions were dominating the H2/O2 system at particular conditions of pressure, temperature and fuel/oxygen/diluent ratios. Overall, good agreement was observed between the model and the wide range of experiments simulated.

INTRODUCTION

The prospect of a hydrogen-based economy has prompted increased interest in the use of hydrogen as a fuel given its high chemical energy per unit mass and cleanliness. It appears that most of the technological problems in using hydrogen in spark-ignited internal combustion engines, including NOx emissions [1], have now been solved; vehicular on-board storage is probably the one remaining diﬃculty [2, 3].

There is also continued interest in developing a better understanding of the oxidation of hydrocarbon fuels [4] over a wide range of operating conditions in order to increase ef- ﬁciency and to reduce the emission of pollutant species. All, or almost all petrochemical, fuels are hydrocarbons which burn to form carbon dioxide and water. Thus, the develop- ment of a detailed kinetic mechanis m for hydrocarbon oxidation necessarily begins with a hydrogen/oxygen sub-mechanis m, followed by the addition of CO chemistry.

In recent years, many kinetic studies of hydrogen oxidation have concentrated on a single set of experimental results obtained either in shock tubes, or in ﬂow reactors or in ﬂames; these have been simulated using a detailed kinetic mechanis m. This procedure has been criticised recently by s mith [5] who asserts that uncertainty limits on individual reaction rate constants produce a parameter space of possible mechanis ms still too im- precise for accurate prediction of combustion properties such as ﬂame speed or ignition delay, thus requiring additional system data. s mith adds that low pressure and coun- terﬂow ﬂames, mixtures in shock tubes, and ﬂow or well-stirred reactors are examples of such experimental environments. It is the aim of this study to apply a hydrogen kinetic mechanis m to as broad a range of combustion environments as possible.

There have been a very large number of measurements made on the reaction between hydrogen and oxygen. These include ﬂame speed measurements, burner stabilised ﬂames in which species proﬁles are recorded, shock tube ignition delay times, and concentration proﬁles in ﬂow reactor studies. This study aims to simulate these experiments using a detailed chemical kinetic mechanis m which takes its origin from Mueller et al. [6] in their study of hydrogen oxidation in a ﬂow reactor. Mueller and co-workers validated their mechanis m using only their ﬂow reactor data over the temperature range 850–1,040 K, at equivalence ratios of 0.3 ≤ φ ≤ 1.0, pressures of 0.3 to 15.7 atm and residence times of 0.004 to 1.5 s. We have exercised their mechanis m against shock-tube data, burner- stabilised ﬂame experiments and ﬂame speed data and have made modiﬁcations to some of the kinetic parameters in order to achieve better overall agreement between mechanis m simulations and this broader range of experimental results. Previously, Marinov et al. [7] had also developed a detailed H2/O2 kinetic mechanis m to simulate shock tube, ﬂame speed and burner stabilised ﬂame experiments with good agreement between model and experiment but a large body of datasets have become available since then. Therefore, this study presents a new detailed chemical kinetic mechanis m for hydrogen oxidation but with increased attention paid to experiments conducted at high pressures since internal combustion engines operate at elevated pressures.

Davis et al. [8] have recently presented a re-examination of a H2/CO combustion mechanis m in which they simulated some of the experimental data included in this study. Their work was motivated by new kinetic parameters for the important reaction H + O2 + M = HO2 + M and by new thermodynamic data for OH, and, had the objective of optimising their H2/CO model against experiment.

IGNITION DELAYS IN SHOCK WAVES

Schott et al. [9] measured ignition delay times of two H2/O2/Ar fuel mixtures behind incident shock waves over a wide range of reactant densities in the temperature range 1,085–2,700 K and at 1 atm. Skinner and Ringrose [10] measured the ignition delays of an H2/O2/Ar mixture in the temperature range 965–1,076 K and at a reﬂected shock pressure of 5 atm. Asaba et al. [11] performed experiments in the temperature range 1,500–2,700 K, at reﬂected shock pressures of 178–288 torr, at an equivalence ratio, φ, of 0.5 and with 98% argon dilution. Fujimoto et al. [12] measured ignition delay times of stoichiometric H2/O2/Ar fuel mixtures in the reﬂected shock pressure range 1.3–5 atm and in the temperature range 700–1,300 K. Hasegawa et al. [13] measured ignition delays in the temperature range 920–1,650 K, at a reﬂected shock pressure of 5.5 atm, with φ = 0.25 at 94% argon dilution. Bhaskaran et al. [14] reported ignition delay times for a 29.59% H2 , 14.79% O2, 55.62% N2 mixture in the temperature range 1,030–1,330 K and at a constant reﬂected shock pressure of 2.5 atm.

More recently, Slack [15] studied stoichiometric hydrogen-air mixtures in a shock tube and measured induction times near the second explosion limit. The experiments were performed at a reﬂected shock pressure of 2 atm in the temperature range 980–1,176 K. Cheng et al. [16] reported ignition delay times for a 6.67% H2, 3.33% O2, 90% Ar mixture in the temperature range 1,012–1,427 K and at a reﬂected shock pressure, P5 ≈ 1.9 atm. Koike [17] measured ignition delay times for two hydrogen/oxygen/argon fuel mixtures of incident shock pressure 20 torr in the temperature range 1,000–1,040 K.

In a methane shock-tube study, Hidaka et al. [18] carried out some measurements of a H2/O2/Ar mixture at 1,250–1,650 K and at reﬂected shock pressures of 1.6–2.8 bar. Pe- tersen et al. [19] measured high pressure (33–87 atm) H2/O2/Ar reﬂected shock ignition delays at 1,189–1,876 K and at an equivalence ratio of 1.0 in every case for six mixtures. Petersen et al. [20] measured reﬂected ignition delay times in three highly dilute H2/O2/Ar mixtures at temperatures of 1,010–1,750 K, equivalence ratio range 1.0 ≤ φ ≤ 1.47 and around atmospheric pressure. Finally, Wang et al. [21] carried out reﬂected shock mea- surements in various H2/air/steam mixtures at 954–1,332 K and pressures of 3.36–16.63 atm. Hydrogen concentration was 15% of air throughout.

FLAME MEASUREMENTS

Atmospheric Flame Speed Measurements

Very many hydrogen/air ﬂame speed studies have been performed at atmospheric pressure, over various ranges of equivalence ratio. Koroll et al. [22] reported data in the equivalence ratio range 0.15 ≤ φ ≤ 5.5, Iijima et al. [23] in the range 0.5 ≤ φ ≤ 3.9 and Takahashi et al. [24] in the range 1 ≤ φ ≤ 4. However, these data did not account for the eﬀects of ﬂame stretch.

The earliest stretch-corrected atmospheric hydrogen/air ﬂame speed experiments were performed by Wu et al. [25] in the range 0.6 ≤ φ ≤ 6. Since then, stretch corrected ﬂame speeds, all of which were performed at 1 atm, have been reported at various equivalence ratio ranges: Egolfopoulos et al. [26] (0.25 ≤ φ ≤ 1.5), Law et al. [27] (0.4 ≤ φ ≤ 1.5), Vagelopoulos et al. [28] (0.3 ≤ φ ≤ 0.55), Dowdy et al. [29] (0.3 ≤ φ ≤ 5), Aung et al. [30] (0.3 ≤ φ ≤ 5) and Tse et al. [31] (0.4 ≤ φ ≤ 4), Fig. 1.

The measurements of Takahashi et al. [24] are considerably faster than the rest of the data and 10 % faster than the intermediate values of Tse et al. and Dowdy et al. at an equivalence ratio of 1.75. The slowest ﬂame speeds are those of Aung et al. [30] which have a maximum ﬂame speed of 2.6 m s—1 at φ = 1.65. The authors point to possible

Figure 1: Atmospheric H2/O2/air ﬂame speeds versus equivalence ratio, Ti = 298 K. ◇ Koroll et al. [22], 田 Iijima et al. [23], ▲ Takahashi et al. [24]; stretch–corrected: ☒ Wu et al. [25] × Egolfopoulos et al. [26], • Law et al. [27], 于 Vagelopoulos et al. [28], ■ Dowdy et al. [29], + Aung et al. [30] and △ Tse et al. [31].

greater stretch eﬀects than accounted for to explain the relative slowness of their data. The Koroll et al. values [22], on the other hand are much faster than any other between 1.0 ≤ φ ≤ 2.5. The recent ﬂame speed measurements of Dowdy et al. [29] and Tse et al. [31] probably are the most representative of the entire dataset; they have a maximum ﬂame speed of 2.85 m s—1 at φ = 1.75.

Lamoureux et al. [32] very recently measured the speeds of freely propagating ﬂames in a spherical bomb for ﬁve H2/air mixtures using a diluent consisting of CO2 + He to mimic the eﬀect of water vapour on ﬂame speed. The mixtures were composed as follows: 父(40%He + 60%CO2) + (1 一父)(H2 + air), where 父 ranged from 0.0 to 0.4, and with synthetic air of composition O2 : N2 = 20 : 80.

High-Pressure Flame Speeds

In addition to their atmospheric ﬂame speed measurements, Tse et al. [31] also measured mass burning velocities for H2/O2/He mixtures in the equivalence ratio range 0.5 ≤ φ ≤ 3.5 and between 1 and 20 atm at an initial temperature of 298 K. It was reported that ﬂames became increasingly unstable at elevated pressures. For this reason, true stretch-free ﬂame speeds become more diﬃcult to measure. Experimentally, in the case of the 10–20 atm data, the oxygen to fuel ratio was reduced to suppress diﬀusional-thermal instability and delay hydrodynamic instability. Using helium as the diluent also helped minimize instability up to 20 atm by reducing the Lewis number of the ﬂame and retarding the formation of ﬂame cells. Stretch-free ﬂame speeds have only been available up to a few atmospheres. The oxygen to helium ratio at 1 to 5 atm was 1:7 (12% dilution) and at elevated pressures, this ratio was 1:11.5 (8% dilution).

Burner Stabilised Flame

In their investigation of a rich 18.83 % hydrogen, 4.6 % oxygen, 76.57 % nitrogen ﬂame at atmospheric pressure, Dixon-Lewis et al. [33] measured the temperature proﬁle and the concentration proﬁles of the stable species in the ﬂame, above and below the burner. Flame structure measurements had been carried out by Kohse-Ho…inghaus [34] who measured H. and O.H radical concentrations versus distance in a H2/O2/Ar ﬂame, at a pressure of 95 mbar, in the equivalence ratio range 0.6 ≤ φ ≤1.4 and in the temperature range 1,100– 1,350 K. Vandooren and Bian [35] investigated the structure of a rich H2/O2/Ar ﬂame over a ﬂat burner at a pressure of 35.5 torr and at an equivalence ratio of 1.91. They reported H2, O2, H2O, H., O. and O.H species mole fractions versus distance above the burner.

Flow Reactors

Mueller et al. [6] measured H2, O2 and H2O proﬁles over the temperature range 850 to 1,040 K, at equivalence ratios of 0.3 ≤ φ ≤1.0 in the pressure range from 0.3 to 15.7 atmospheres and over a range of residence times of 0.004 to 1.5 s. Previously, Yetter et al. [36] reported atmospheric H2 , O2 and H2O proﬁles at 910 K, and at an equivalence ratio of 0.3.

Experiments Simulated

A representative selection of recent experimental work has been chosen to validate the H2-O2 combustion mechanis m. The chosen experiments were:

1. the ignition delay times measured by Schott et al. [9], Skinner et al. [10], Fujimoto [12], Bhaskaran et al. [14], Slack [15], Cheng et al. [16], Petersen et al [19], Hidaka et al [18], Petersen et al. [20] and Wang et al. [21]. Simulations of the data of Asaba et al. [11], Hasegawa et al. [13], and Koike [17] were not attempted in this study because of a lack of suﬃcient information.

2. the ﬂame speed measurements of Dowdy et al. [29]. These ﬂame speeds not only span a wide range of equivalence ratio but are in agreement with the more recent values of Tse et al. [31]. Dowdy and co-workers also measured the temperature proﬁles, thus making their data more amenable to simulation.

3. the high-pressure ﬂame speed measurements of Tse et al. [31]. This data is the only set where hydrogen ﬂame speeds have been measured at pressures greater than 5 atm.

4. the very lean H2/air and H2/air/CO2/He ﬂame speed measurements of Lamoureux et al. [32].

5. the burner stabilised ﬂame proﬁles of Vandooren et al. [35] in which reactant and intermediate species concentrations were measured as a function of height above the burner surface. Also included are the species proﬁles of Dixon-Lewis et al. [33].

6. the comprehensive ﬂow reactor data of Mueller et al. [6] along with a single data set from Yetter et al. [36].

CHEMICAL KINETIC MODELING

The chemical kinetic mechanis m was developed and simulations performed using the HCT program [37]. Initially, ignition delay times measured by Slack [15], Fig. 8, and Hidaka et al. [18], Fig. 7, and the ﬂow reactor experiments of Mueller et al. [6], Fig. 25, were simulated with very good agreement observed between experiment and model. The mechanis m was then converted into Chemkin 3.6 [38] format and the simulations repeated in order to compare results from both codes, which were in very good agreement as expected. Thereafter, all other experiments including the ﬂame speeds and the burner stabilised ﬂame proﬁles were simulated using only the Chemkin applications.

Thermodynamic and Transport Properties

The H2/O2 reaction mechanis m consists of nineteen reversible elementary reactions, Table 1, together with the thermochemical data, Table 2. Reverse rate constants were computed by microscopic reversibility. The thermochemical data for each species considered in the mechanis m are from the Chemkin thermodynamic database [51] with the exception of two:

1. ∆Hf(HO.2 , 298K) of 3.0 kcal mol—1 , from Hills and Howard [52] which is in good agreement with the recent reappraisal by Ramond et al. [53] of 3.2±0.5 kcal mol—1 .

2. ∆Hf(O.H, 298K) of 8.91 kcal mol—1 which is based on recommendations by Ruscic et al. [54] and Herbon et al. [55].

The Chemkin database of transport parameters was used without modiﬁcation. As in the study of Tse et al. [31], the kinetic parameters of helium were assumed equal to those of argon in order to simulate ﬂame propagation where helium is the diluent. As Tse et al. noted, using the third body eﬃciency of argon for monatomic helium is a useful starting estimate; termolecular reactions such as H. + O2 + M = HO.2 + M become signiﬁcant at el- evated pressures and so the uncertainties in these values can create considerable diﬀerences in the ﬂame speeds.

Mechanis m Formulation

The kinetic mechanis m referred to in this study as ζthis study’ or the ζ revised mechanis m ’ has its origins in the CO/H2/O2 reaction mechanis m of Yetter et al. [56], which was updated later by Kim [57] and is, for the most part, taken from the more recent work of Mueller et al. [6].

We found it necessary to modify some of the kinetic parameters of Mueller et al. in order to achieve an overall improvement with all the experimental data simulated here. This altered version of the mechanis m, Table 1, the revised mechanis m, reproduces the selected experimental datasets more accurately than that published by Mueller and co- workers.

The entire data set has also been simulated using relevant portions from Leeds 1.5 [58], Konnov [59, 60] and GRI-Mech 3.0 [61] which are all primarily methane oxidation mechanis ms. The reason for using both Konnov mechanis ms is that the shock tube data presented in Figs. 4-6 was used to validate version 0.3 while the more recent version 0.5 was used to simulate the remaining data. A select set of experiments is reproduced here using GRI-Mech, Leeds and Konnov as an indication of their performance but they have not been comprehensively tested.

A comparison shows that these mechanis ms are quite diﬀerent, Table 3; not only do the total number of reactions diﬀer but so do the rate constant expressions.

详细内容请见附件

免责声明：

本页面/内容部分素材来源于互联网公开信息，旨在传递更多信息，不代表本平台立场。

版权归原作者或机构所有，如涉及侵权，请通过平台联系我们，我们将在核实后第一时间处理。

本平台对转载内容的真实性、准确性不作任何保证，用户需自行判断并承担使用风险。

FMEA在钢板弹簧设计中的应用

摘要：故障模式与影响分析(FMEA) 是提高产品可靠性的重要方法之一。文中介绍了汽车钢板弹簧设计过程中使用FMEA 的方法，并对开展故障模式与影响分析工作中需要注意的问题进行了讨论。关键词：汽车；故障模式与影响分析( FMEA); 钢板弹簧设计；可靠性中图分类号：U463.33 文献标识码：A 文章编号：1671-2668(2006)04-0005—03故障模式与影响分析 FMEA(Failure Mode and Effects Analysis)是产品可靠性、维修性设计中的重要分析方法之一。它通过对系统中每一产品可能产生的所有故障模式及其严重程度、检测难易程度和发生频度进行分类、归纳分析，鉴别设计上的薄弱环节，以采取适当的纠正措施消除或减轻其影响。总体来说，通过实行 FMEA, 可在产品设计或生产工艺真正实现之前发现产品的弱点，在原形样机阶段或在大批量生产之前确定产品缺陷。FMEA 作为一种可靠性分析方法起源于20世纪50年代(由美国格鲁曼公司开发，用于飞机发动机故障预防);70年代，美国海军制定FMEA 标准； 1976年，美国国防部采纳 FMEA 标准；80年代，汽车工业和微电子工业领域应用 FMEA;90 年代， ISO9000推荐采用FMEA; 1994 年，FMEA 成为汽车行业质量认证标准 QS9000 的认证要求。在汽车行业，FMEA 是产品设计与开发阶段、过程设计与开发阶段必须使用的缺陷预防工具。本文以汽车钢板弹簧为例，对 FMEA 在汽车设计与开发中的应用进行探讨。1 钢板弹簧分析钢板弹簧是汽车悬架系统中重要的弹性元件，其主要功能：①连接车架与车桥，在车辆行驶状态下，承受由于地面不平而使车轮、车桥产生的冲击力，避免冲击力直接向车身传递；②支撑车身重量，决定整车高度；③在车轮行驶时起导向作用。1.1 常见故障模式在开展 FMEA 分析之前，应对批量使用相似零部件的用户、长距离道路试验等情况进行调查，收集相关故障情况并整理成表格(见表1)。1.2 工作环境及设计要求本文分析对象为某汽车钢板弹簧，设计时已知整车轴荷分配、钢板弹簧限位块刚度、悬架偏频、静挠度和动挠度等参数。2 FMEA 表格填写方法及分析结果在 QS9000 关于 FMEA 相关规定的基础上制定表格。2.1 表格的填写方法第一栏(零部件名称、图号):填写待分析的零部件名称及其图号。第二栏(功能):用简洁的文字列出被分析零部件需具有的功能，如该零部件有多种功能，且有不同的失效模式，要把所有功能都单独列出。第三栏(潜在失效模式):填入零部件可能未达到或未完成的功能项目的种类(如预期的功能丧失),前提是这种失效可能发生，但不一定发生。第四栏(失效影响):填人失效模式对功能的影响，要根据顾客可能发现或经历的情况来描述失效的后果，如果失效模式影响到安全性或与法规不符，则要清楚地予以说明。第五栏(严重度 S): 填人失效模式严重性的评价。免责声明：本页面/内容部分素材来源于互联网公开信息，旨在传递更多信息，不代表本平台立场。版权归原作者或机构所有，如涉及侵权，请通过平台联系我们，我们将在核实后第一时间处理。本平台对转载内容的真实性、准确性不作任何保证，用户需自行判断并承担使用风险。