NUMERICAL STUDY ON TURBULENCE EFFECTS IN POROUS BURNERS

・44・2021年第50卷第10期Vol.50No.102021INDUSTRIAL HEATINGDOI：10.3969/j.issn.1002-1639.2021.10.011蜂窝陶瓷蓄热室内气体传热过程数值模拟研究汪建新1,王思浩1,吴左明2,江华$(1.内蒙古科技大学机械工程学院，内蒙古包头014000；2.北京凤凰工业炉有限公司，北京100083)摘要:为了探索气体在蓄热室内的传热特性,运用多孔介质模型和CFD理论，利用Fluent软件仿真求解，分析了蓄热室内的温度场、压力场及形变。

结果表明:蓄热式内的压力呈梯度分布;高温烟气入口处温度最高约为1400K,常温空气入口处温度最低约为500 K;形变最大的地方出现在高温烟气入口顶部的两个直角处约为0.28mm o所得结果对蓄热室稳态工作时蓄热体内温度、压力和形变等有一定参考价值。

关键词:蓄热室;数值模拟;温度分布;计算流体力学中图分类号:TQ021.3文献标志码:A文章编号:1002-1639(2021)10-0044-04Numerical Simulation Research on Gas Heat Transfer Process in Honeycomb Ceramic RegeneratorWANG Jianxin1,WANG Enhao1,WU Qiming2,JIANG Hua2(1.School of Mechanical Engineering,Inner Mongolia University of Science and Technology,Baotou014000,China；2.Beijing Phoenix Industrial Furnace Co.Ltd.,Beijing100083,China)Abstract:In order to explore the heat transfer characteristics of the gas in the regenerator,use porous media model and CFD theory,use fluent software to simulate and solve,analyze the flow field,temperature field,concentration field,pressure field and deformation in the regenera­tor.The results show that the pressure in the regenerative type is gradient distribution；The maximum temperature at the inlet of high tempera­ture flue gas is about1400K,the minimum temperature at the inlet of room temperature air is about500K；The largest deformation occurs at two right angles at the top of the high temperature flue gas inlet,about0.28mm.The results obtained have certain reference value for the tem­perature,pressure and deformation of the regenerator body when the regenerator works in a steady state.Key Words:regenerator；numerical simulation；temperature distribution;CFD20世纪80年代被开发并广泛推广的蓄热式燃烧技术⑷又被称为高温空气燃烧技术，因其显著的节能效果与特殊的燃烧过程,得到了国际工业炉领域的普遍应用炉勺。

[1] G. K. Batchelor.An Introduction to Fluid Dynamics.Cambridge Univ. Press, Cambridge, England, 1967.[2] D. Cokljat, V. A. Ivanov, and S. A. Vasquez.A Non-Equilibrium Two-Phase Model for Cavitating Flows.In Third International Conference on Multiphase Flow, Lyon, France, 1998. Available on ICMF98 CD-ROM, paper 224.[3] J. L. Ferzieger and M. Peric.Computational Methods for Fluid Dynamics.Springer-Verlag, Heidelberg, 1996.[4] J. Janicka and W. Kollmann.A Numerical Study of Oscillating Flow Around a Circular Cylinder.Combustion and Flame, 44:319--336, 1982.[5] W. P. Jones and J. H. Whitelaw.Calculation Methods for Reacting Turbulent Flows: A Review.Combust. Flame, 48:1--26, 1982.[6] W. M. Kays.Turbulent Prandtl Number - Where Are We?J. Heat Transfer, 116:284--295, 1994.[7] B. E. Launder.Second-Moment Closure and Its Use in Modeling Turbulent Industrial Flows. International Journal for Numerical Methods in Fluids, 9:963--985, 1989.[8] B. E. Launder.Second-Moment Closure: Present... and Future?Inter. J. Heat Fluid Flow, 10(4):282--300, 1989.[9] B. E. Launder, G. J. Reece, and W. Rodi.Progress in the Development of a Reynolds-Stress Turbulence Closure.J. Fluid Mech., 68(3):537--566, April 1975.[10] B. E. Launder and N. Shima.Second-Moment Closure for the Near-Wall Sublayer: Development and Application. AIAA Journal, 27(10):1319--1325, 1989.[11] B. E. Launder and D. B. Spalding.Lectures in Mathematical Models of Turbulence.Academic Press, London, England, 1972.[12] B. E. Launder and D. B. Spalding.The Numerical Computation of Turbulent Flows.Computer Methods in Applied Mechanics and Engineering, 3:269--289, 1974.[13] J. P. Vandoormaal and G. D. Raithby.Enhancements of the SIMPLE Method for Predicting Incompressible Fluid Flows. Numer. Heat Transfer, 7:147--163, 1984.[14] L. D. Smoot and P. J. Smith.NOx Pollutant Formation in a Turbulent Coal System.In Coal Combustion and Gasification, page 373, Plenum, Plenum, NY, 1985.[15] F. C. Lockwood and C. A. Romo-Millanes.Mathematical Modelling of Fuel - NO Emissions From PF Burners.J. Int. Energy, 65:144--152, September 1992.[16] R. K. Boyd and J. H. Kent.Three-dimensional furnace computer modeling.In 21stSymp. (Int'l.) on Combustion, pages 265--274. The Combustion Institute, 1986.[17] M. Manninen, V. Taivassalo, and S. Kallio.On the Mixture Model for Multiphase Flow.VIT Publications, Technical Research Centre of Finland, 1996.。

ＣｏＮＴＥＮＴＳＣｈｉｎｅｓｅａｂｓｔｒａｃｔ……………………………………………………………………………………………．．ＩＥｎｇｌｉｓｈａｂｓｔｒａｃｔ……………………………………………………………………………………………ＩｌｌＮｏｍｅｎｃｌａｔｕｒｅ………………………………………………………………………………………………．．Ｖ１Ｉｎｔｒｏｄｕｃｔｉｏｎ…………………………………………………………………………………………………１１．１Ｂａｃｋｇｒｏｕｎｄ…………………………………………………………………………………………．１１．２Ｒｅｓｅａｒｃｈｓｉｔｕａｔｉｏｎ．．．．．．．．．．．．．．．．．．．．．…．．．…．．．．…．…．…．．．．．．．．．．．．．．．．．．．．．．．．．．．．…．．．．．．．．．…．．．．．】【１．２．１Ｆｌｏｗｃｈａｒａｃｔｅｒｉｓｔｉｃｓｉｎｃｕｒｖｅｄｐｉｐｅｌｉｎｅ．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．…．．．．．．１１．２．２Ａｂｒａｓｉｏｎａｎｄｄｅｐｏｓｉｔｉｏｎｏｆｆｌｕｅａｎｄｈｅａｔｅｘｃｈａｎｇｅｒ．．…．……．…．………．．．．４１．２．３Ｍｅｔｈｏｄｓｏｆｆｌｕｉｄｕｎｉｆｏｒｍｉｔｙ……．………………．…．…．………………．．…………．．６１．３Ｔｅｘｔｕａｌｔａｓｋ…………………………………．．…………………………．………ｉ…．．．…………．．．８２Ｎｕｍｅｒｉｃａｌｃｏｍｐｕｔａｔｉｏｎｔｈｅｏｒｉｅｓｏｆｇａｓ—ｓｏｌｉｄｔｗｏｐｈａｓｅｆｌｏｗｉｎｃｕｒｖｅｄｆｌｕｅ……９２．１Ｔｕｌ？ｂｕｌｅｎｃｅｍｏｄｅｌｓｆｏｒｆｌｕｉｄｆｌｏｗｉｎｆｌｕｅ…………．．…．……．…．…．…………．………．．．９２．１．１Ｍａｔｈｅｍａｔｉｃａｌｄｅｓｃｒｉｐｔｉｏｎｏｆｔｈｅｔｕｒｂｕｌｅｎｃｅ．…．．……．…．……………．……．．．．．９２．１．２Ｅｄｄｙｖｉｓｃｏｓｉｔｙｍｏｄｅｌｓａｎｄａｐｐｌｉｃａｔｉｏｎ……．…………．…．……．…．．…．……．．．１０２．１．３Ｗａｌｌ．ｆｕｎｃｔｉｏｎｍｅｔｈｏｄｎｅａｒｗａｌｌａｒｅａ……………………………………………．１４２．２Ｆｌｕｅｇａｓ·ａｓｈｔｗｏｐｈａｓｅｍｏｄｅｌｓ…………．……………．．………．…．…………．．…………．１４２．２．１Ｃｌａｓｓｉｆｉｃａｔｉｏｎｓａｎｄｃｈａｒａｃｔｅｒｉｓｔｉｃｓｏｆｇａｓ—ｓｏｌｉｄｔｗｏｐｈａｓｅｆｌｏｗ．………．１５２．２．２Ｇａｓ－ｓｏｌｉｄｔｗｏｐｈａｓｅｍｏｄｅｌｓａｎｄａｐｐｏｌｉｃａｔｉｏｎ……．．．…．………．．…．………．１５２．３Ｐｏｒｏｕｓｍｅｄｉｕｍｍｏｄｅｌ…………．…………．……………．……．…．．．．………………．……．．．１６２．４Ｓｕｍｍａｒｙ……………．……………．……．．……．……………．．…．．．．…．…．．…．……．…………．．１７３Ｆｌｏｗｃｈａｒａｃｔｅｒｉｓｔｉｃｓｉｎｅｑｕａｌｃｒｏｓｓ－ｓｅｃｔｉｏｎｃｕｒｖｅｄｔａｉｌｆｌｕｅ．．…．．………．．…．………．１９：；．１Ｎｕｍｅｒｉｃａｌｍｌｄｅｌｆｏｒｃｕｒｖｅｄｔａｉｌｆｌｕｅ…．……………．……．．．．…．．…．…．……．…．………１９３．１．１Ｅｓｔａｂｌｉｓｈｍｅｎｔｏｆｃｏｎｔｒｏｌｅｑｕａｔｉｏｎ………………………………………………．．１９３．１．２Ｄｉｓｃｒｅｔｉｚａｔｉｏｎａｎｄｓｏｌｕｔｉｏｎｏｆｃｏｎｔｒｏｌｅｑｕａｔｉｏｎｓ……………………………．２１３．１．３Ｖａｌｉｄａｔｉｏｎｆｏｒｔｈｅｎｕｍｅｒｉｃａｌｃｏｍｐｕｔａｔｉｏｎｍｏｄｅｌ……………………………２１３．２Ｆｌｏｗｃｈａｒａｃｔｅｒｉｓｔｉｃｓｉｎｈｏｒｉｚｏｎｔａｌ．ｔｏ．ｖｅｒｔｉｃａｌｃｕｒｖｅｄｔａｉｌｆｌｕｅ……………………２２３．２．１Ｃｏｍｐｕｔａｔｉｏｎａｌｇｒｉｄｓａｎｄｂｏｕｎｄａｒｙｃｏｎｄｉｔｉｏｎｓ………………………………．２３３．２．２Ａｎａｌｙｓｉｓｏｆｇａｓ—ａｓｈｔｗｏｐｈａｓｅｆｌｏｗｆｌｉｅｄｉｎｔａｉｌｆｌｕｅ………………………．２４３．２．３Ｅｆｆｅｃｔｏｆｉｎｌｅｔｖｅｌｏｃｉｔｙｏｎｆｌｏｗｆｉｌｅｄｉｎｔａｉｌｆｌｕｅ……………………………．．３１３．２．４Ｅｆｆｅｃｔｏｆｐａｒｔｉｃｌｅｍａｓｓｌｏａｄｉｎｇｏｎｆｌｏｗｆｉｌｅｄｉｎｔａｉｌｆｌｕｅ…………………．３４３．２．５Ｅｆ瓷ｃｔｏｆｃｕｒｖａｔｕｒｅｒａｔｉｏｏｎｆｌｏｗｆｉｌｅｄｉｎｔａｉｌｆｌｕｅ……………………………３６３．２．６Ｅｆｒｅｃｔｏｆｃｒｏｓｓ－ｓｅｃｔｉｏｎａｓｐｅｃｔｒａｔｉｏｏｎｆｌｏｗｆｉｌｅｄｉｎｔａｉｌｆｌｕｅ……………．３８ｉｉｊ山东大学硕十学位论文摘要电厂锅炉尾部烟道由于转弯以及变截面致使烟道内含灰烟气流的速度场和飞灰颗粒浓度场分布不均，进而造成烟道和换热器的积灰和磨损以及换热器的换热不均等问题，对电厂的安全经济运行造成威胁。

第32卷第5期高校化学工程学报No.5 V ol.32 2018 年10月Journal of Chemical Engineering of Chinese Universities Oct. 2018文章编号：1003-9015(2018)05-1090-07克劳斯燃烧炉花墙结构改进杨海洲, 武锦涛, 胡大鹏(大连理工大学化工机械与安全学院, 辽宁大连116024)摘要：燃烧炉的花墙结构有助于提高燃烧炉中硫的产率。

为深入探究燃烧炉中花墙结构对反应物的流动、传热和反应速率的影响，针对扼流圈结构和花墙结构分别建立了燃烧炉的数学模型，应用Fluent对燃烧炉的混合燃烧情况进行数值模拟。

采用realizable k-ε湍流模型和组分运输模型探究花墙结构的优势所在，并在此基础上，对花墙结构提出改进。

结果表明，花墙结构与扼流圈结构相比，加强了气体的混合效果，延长了气体的停留时间，将硫的产率提升了1%。

进一步研究表明花墙结构的位置以及其导流方式能够使燃烧炉炉膛温度稳定的区域在流动方向增大3 m，并且使稳定后的炉膛温度提升23 K，进一步延长气体的停留时间，使硫的产率提升7.6%。

研究试图通过改进花墙结构来定制燃烧炉的内部流场以提高硫的产率，为克劳斯燃烧炉的优化设计提供理论依据。

关键词：硫回收；克劳斯燃烧炉；花墙；数值模拟中图分类号：TQ465.92 文献标志码：A DOI：10.3969/j.issn.1003-9015.2018.05.013Structural Improvement of Checker Wall for Claus Reaction FurnaceYANG Hai-zhou, WU Jin-tao, HU Da-peng(School of Chemical Machinery and Safety, Dalian University of Technology, Dalian 116024, China)Abstract:Checker wall of Claus furnace is useful in improving sulfur productivity. In order to explore the effects of checker wall on flow, heat transfer and reaction rate, mathematical models were established and numerical simulation of combustion in the furnace was studied by Fluent considering Claus furnaces with choke and checker wall, respectively. Advantages of checker wall were explored using realizable k-εmodel and component transport equation, and improved checker wall was designed. Simulation results show that the checker wall can enhance gases mixing and hence increase the residence time of gases and improve sulfur productivity by 1%. Furthermore, the location and diversion procedure of checker wall can effectively extend the area of stable temperature by 3 m along the flow stream direction, and increase the stable temperature by 23 K, which enhances gas mixing and increase gas residence time, and sulfur productivity can increase by 7.6%. The flow field of the furnace is designed to increase sulfur productivity by improving the structure of checker wall, and theoretical basis is provided for optimizing checker wall structure design.Key words: sulfur recovery; Claus furnace; checker wall; numerical simulation1前言燃烧炉是整个克劳斯工艺的关键设备，其硫的转化量占整个工艺的70% 左右[1]。

Flow Turbulence Combust(2012)89:215–230DOI10.1007/s10494-011-9343-2Numerical Investigation on the Hydrogen-AssistedStart-Up of Methane-Fueled,Catalytic MicroreactorsSymeon Karagiannidis·John MantzarasReceived:25August2010/Accepted:28February2011/Published online:24March2011©Springer Science+Business Media B.V.2011Abstract The hydrogen-assisted start-up of methane-fueled,catalytic microreactorshas been investigated numerically in a plane-channel configuration.Transient2-D simulations have been performed in a platinum-coated microchannel made ofeither ceramic or metallic walls.Axial heat conduction in the solid wall and surfaceradiation heat transfer were accounted for.Simulations were performed by varyingthe inlet pressure,the solid wall thermal conductivity and heat capacity,and com-parisons were made between fuel mixtures comprising100%CH4and90%CH4–10%H2by volume.A significant reduction in the ignition(t ig)and steady-state (t st)times was evident for microreactors fed with hydrogen-containing mixtures in comparison to pure methane-fueled ones,for all pressures and reactor materialsinvestigated,with hydrogen having a direct thermal rather than chemical impact oncatalytic microreactor ignition.The positive impact of H2addition was attenuated asthe pressure(and the associated CH4catalytic reactivity)increased.Reactors withlow wall thermal conductivity(cordierite material)benefited more from hydrogenaddition in the fuel stream and exhibited shorter ignition times compared to higherthermal conductivity ones(FeCr alloy)due to the creation of spatially localizedhot spots that promoted catalytic ignition.At the same time,the cordierite materialrequired shorter times to reach steady state.Microreactor emissions were impactedpositively by the addition of hydrogen in the fuel stream,with a significant reductionin the cumulative methane emissions and no hydrogen breakthrough.Finally,gas-phase chemistry was found to elongate the steady-state times for both ceramic andmetallic materials.Keywords Catalytic microreactors·Transient simulation·Hetero-/homogeneouscombustion·Hydrogen-assisted combustionS.Karagiannidis·J.Mantzaras(B)Combustion Research,Paul Scherrer Institute(PSI),CH-5232Villigen,Switzerlande-mail:ioannis.mantzaras@psi.ch1IntroductionIn recent years,hydrocarbon-fueled catalytic microreactors have been the focus of intense research efforts[1],as they can reliably supply the necessary thermal power for a variety of portable power generation systems with demonstrated energy densi-ties considerably higher than those obtained with state-of-the-art Li-ion batteries[2]. Catalytic microburners have thus been investigated for use in scaled-down devices employing conventional thermal cycles,with applications ranging from micro-Stirling engines[2]to direct chemical-to-thermal energy conversion devices[3].With most of the experimental work on catalytic microreactors focusing on operational aspects such as fuel conversion and thermal management[4],a wide range of computational models of varying complexity are increasingly being em-ployed to provide insight on the physics of microscale catalytic combustion.Such approaches include simplified1-D models with lumped heat and mass transport coefficients investigating the performance of propane-fueled microreactors[5], 2-D models with detailed surface chemistry without heat conduction in the solid wall probing the robustness of methane-fueled microreactors against external heat losses [6]and finally full2-D CFD models whereby all relevant heat transfer mechanisms in the solid(heat conduction and surface radiation heat transfer)along with detailed catalytic and gas-phase chemistries are accounted for[7].The start-up of microreactors,and catalytic microburners in particular,is of prime importance in small-scale power generation devices;for example,long heat-up times can reduce the availability of the microdevice over extended periods and also result in substantial pollutant emissions.While the aforementioned numerical studies delineated the steady-state performance of catalytic microreactors,only a few investigations focused on their transient behavior,with particular emphasis on the ignition process(light-off).Such works include the mapping of hydrogen/air flame dynamics in catalytic and non-catalytic microchannels having prescribed wall temperatures using a fully transient2-D CFD code[8],as well as investigations of optimum ignition strategies in a propane-fueled catalytic microreactor using a1-D transient code[9].Since fully transient simulations for reacting flows can be computationally de-manding,a number of simplifications have been introduced for catalytic combustion applications allowing extensive parametric numerical investigations.One such sim-plified computational model used widely for conventional-sized honeycomb reactors involves a“continuum”description of the entire reactor structure and invokes the quasisteady treatment of the gaseous phase as a result of the disparity between the gas-phase and solid-phase characteristic time scales[10].In a recent combined experimental and numerical study[11],the light-off and extinction in the catalytic partial oxidation(CPO)of methane over rhodium were investigated by invoking the quasisteady assumption for the gas to model the2-D reacting flow in a single catalytic channel.The validated transient computational model was subsequently used to study ignition characteristics in a methane-fueled,catalytic microreactor channel [12].Therein,detailed hetero-/homogeneous chemistry was employed along with all relevant heat transfer mechanisms in the reactor,with emphasis on the impact of solid heat conduction and thermal radiation heat transfer from the catalytic walls on microreactor ignition and attainment of steady state.Parametric studies were carried out to identify the effect of various operational parameters such as pressure,fuel/air equivalence ratio,solid material properties and radiation properties on the transientprocesses leading to ignition and finally to steady-state.Operation at higher-than-atmospheric pressures was identified as particularly favorable due to a subsequentsignificant reduction in the times required for microreactor ignition and attainmentof steady state.In the present work,a numerical study is undertaken to investigate the start-upof methane-fueled,catalytic microreactors using methane/hydrogen fuel blends,asan alternative strategy for rapid light-off.A full elliptic,transient in the solid andquasisteady in the gas numerical code is used to simulate the reacting flow in acatalytic plane channel configuration with a gap of1mm and a length of10mm,with this setup effectively representing a single channel of a catalytic honeycombmicroburner structure.Detailed catalytic and gas-phase reaction mechanisms forthe total oxidation of methane and hydrogen on platinum have been used.Theinvestigation is focused on operating conditions pertinent to microturbine-basedmicroreactor systems[13–15],which include preheated fuel/air mixtures and inletpressures up to5bar.The main objective is to quantify the impact of hydrogen-assisted hetero-/homogeneous combustion on the elapsed time required for microre-actor ignition and subsequent attainment of steady state.Operating pressures in therange of1bar≤p≤5bar are investigated for two representative microreactormaterials,namely cordierite ceramic and FeCr alloy parisons are madewith the start-up times of catalytic microreactors operating with pure methane/airstreams(no hydrogen addition[12]);moreover,the effect of hydrogen-assistedcatalytic combustion on microreactor emissions is assessed.It is noted that thiswork presents for the first time transient results on the hydrogen-assisted hetero-/homogeneous combustion of methane in catalytic microreactors at higher-than-atmospheric pressures and for different microreactor wall materials.The current article is organized as follows.The numerical model is presented first.Hydrogen-assisted catalytic combustion is qualitatively assessed next,followed byresults on the impact of hydrogen addition in the fuel/air stream on microreactorignition and steady-state times,at various pressures and for both solid materialsinvestigated.Catalytic microreactor start-up characteristics during operation withmethane/hydrogen fuel blends are subsequently analyzed,followed by an assessmentof pollutant emissions during the microreactor start-up process.Finally,the impactof gas-phase chemistry is discussed.2Numerical ModelA full-elliptic,two-dimensional CFD code[11,16]has been used to simulatethe flow domain in a plane channel having length L=10mm,height2b=1mm and wall thicknessδ=50μm(see Fig.1).The initial1mm channel lengthwas catalytically inert,while the remaining L a=9mm was coated with platinum. Due to symmetry,only half of the channel domain was modeled.The global fuel-to-air equivalence ratio of the examined CH4/H2/air mixtures wasϕ=0.37,whichwas obtained when substituting10%CH4by volume with H2in an initial leanCH4/air stream of equivalence ratioϕ=0.40.The inlet temperature was set to T IN= 850K,a value practically achievable in recuperated microreactor thermal cycles[13]. The initial temperature for the channel solid wall was uniform and equal to the incoming mixture temperature,such that T W(x,t=0)=850K.Calculations wereFig.1Schematic of the Array catalytic microreactorconfigurationperformed for pressures p=1,2,3,4and5bar,while the nominal inlet velocitywas U IN=1.5m/s at p=1bar,a value typical for microreactors.When increasing the inlet pressure,the inlet velocity was decreased accordingly,so as to maintain the same mass throughput(ρIN U IN).Two types of solid materials were examined for the microreactor wall,cordierite ceramic with a thermal conductivity of k s=2W/mK and a thermal capacity ofρs c s=3806kJ/m3K and FeCr alloy metal with k s=16W/mK andρs c s=4428kJ/m3K.The quasisteady assumption requires that the convective and diffusive time scales of the gas are appreciably shorter than the characteristic time scales for diffusion of heat in the solid,thus allowing the gaseous flow to equilibrate to the channel solid wall temperature at any given time during the ignition event.The2-D steady calculations of the flow field were then coupled to a1-D transient energy balance equation for the solid.Justification for the choice of a1-D model for the solid energy equation has been provided elsewhere[12].The net radiation method for diffuse-gray areas[12,17]accounted for radiation exchange between the discretized channel wall elements themselves,and between each wall element and the inlet and outlet channel enclosures.The inlet and outlet planes of the enclosure had emissivities equal to those of the channel wall surfaces,εIN=εOUT=ε=0.6,while the inlet and outlet exchange temperatures were set equal to the inlet and outlet mixing cup temperatures,respectively.In the numerical model,the geometrical surface area has been considered equal to the catalytically active area.Radiative boundary conditions were finally applied to the vertical front and rear solid wall faces.It should be noted that,while the outer horizontal channel wall was treated as adiabatic,the reactor itself was non-adiabatic due to radiation heat losses towards the colder inlet enclosure.The combustion of lean methane/hydrogen blends on platinum was modeled using the elementary heterogeneous scheme of Deutschmann et al.[18](24reactions, 11surface and9gaseous species),coupled to the C1/H/O elementary gas-phase mechanism of Warnatz et al.[19](26species,108reactions).In the catalytic mechanism,a surface site density =2.7×10−9mol/cm2was used.The catalytic mechanism has been validated against spatially-resolved measurements of major species concentrations across the boundary layer formed in a Pt-coated channel at pressures of1to16bar,while the gas-phase mechanism has been tested against OH laser induced fluorescence(LIF)homogeneous ignition measurements in the same channel reactor,again at pressures of1to16bar[20,21].To reproduce homogeneous ignition at p≤6bar(a range encompassing the present investigation),the gaseous mechanism has been modified in the single reaction CHO+M⇔CO+H+M;this modification was further supported by recent kinetic measurements,as described in[21].Mixture-average diffusion was used,with transport properties calculated from the CHEMKIN database[22].Surface and gas-phase reaction rates were evaluated using Surface-CHEMKIN[23]and CHEMKIN,respectively[24].An orthogonal staggered grid with100×24points in the x-and y-direction,respectively,over the channel half-height produced a grid-independent solution forthe flow domain.Finer spacing towards the channel wall and entry section was used.A100grid node resolution in the x-direction was also used to discretize the solid wall.At the inlet(x=0),uniform profiles of species,temperature and axial velocity wereapplied,while zero-Neumann conditions were set at the outlet(x=L)and planeof symmetry(y=0).No-slip was applied for both velocity components at the gas–wall interface(y=b).The coupled gas and solid phases were solved iteratively andconvergence was achieved at each time step when the solid temperature did not varyat any position along the wall by more than10−5K.The quasisteady approximation also entails the assumption of catalytic chemicalreaction times being shorter than the heat conduction times in the solid,so as toensure chemical equilibration at the local wall temperature during an integrationtime step of the ing the same methodology as in[12],a time step t=50ms was subsequently used in this work,having a value sufficiently longer thanthe chemical time scales present during catalytic microreactor ignition.3Results and DiscussionTransient simulations were performed in order to assess the impact of hydrogenaddition in methane-fueled catalytic microreactors during the start-up phase.First,hydrogen-assisted heterogeneous combustion is assessed in an ideal catalytic reactor,which indicates the potential benefit of substituting part of methane in the fuel/airstream with hydrogen.Next,full2-D simulations are performed to quantify the effectof hydrogen addition in methane-fueled microreactor start-up times.Operating para-meters of interest in this case are the inlet pressure p,the microreactor wall material,and finally the potential impact of gas-phase reactions.The computed characteristictimes of interest(describing the microreactor start-up process)are the ignition time(t ig),defined as the elapsed time required to reach50%of fuel conversion at the channel outlet,and the steady-state time(t st),defined as the elapsed time whereby the outlet gas temperature varied by less than10−3K.By running a steady-stateversion of the code,it was confirmed that the adopted definition of steady statein the transient simulations reproduced the true steady-state outlet temperaturewithin1K.3.1Hydrogen-assisted heterogeneous combustionIn order to decouple the underlying chemical processes from microreactor-specificeffects(e.g.,thermal inertia of solid wall,flow conditions)and to acquire an initialestimate on the effect of hydrogen addition on catalytic methane combustion,computations have been carried out with an ideal reactor model.Catalytic light-off times were computed in a constant pressure batch reactor,under conditionspertinent to the subsequent full2-D channel simulations.To this direction,thehomogeneous-reaction package SENKIN of CHEMKIN[25]has been augmentedwith the inclusion of catalytic reactions(model details have been provided elsewhere[11]).Calculations were performed for two fuel/air mixtures,one containing puremethane fuel(100%CH4)at a fuel/air equivalence ratio ofϕ=0.40,and anothercontaining90%CH4and10%H2,at an overall fuel/air equivalence ratio ofϕ=0.37.Reactor pressure for both cases was p=1bar,while the initial fuel/air mixtureand reactor temperatures were T=850K.To mimic the confinement of a catalyticmicroreactor channel,the surface-to-volume ratio of the ideal batch reactor wasset to S/V=20cm−1,a value equal to the S/V value of the2-D plane channelmicroreactor in the ensuing calculations.Figure2presents the computed temporalevolution of temperature during light-off for the aforementioned fuel/air mixtures.Keeping in line with the definition of t ig in Section3,it is evident from the topgraph in Fig.2that,despite the overall lower equivalence ratio(and also lowerchemical energy input),the CH4/H2blend achieves catalytic light-off appreciablyfaster than the pure methane case.In the case of the CH4/H2blend,the initial reactortemperature of850K is high enough to rapidly consume all of H2(see Fig.2,bottomgraph);this in turn leads to a temperature rise(evident during the first3ms in Fig.2),which further promotes methane catalytic reactions and results in faster light-off.Itis thus expected that even a10%substitution of CH4with H2in the subsequentmicroreactor simulations can significantly impact the computed ignition and steady-state times;moreover,pollutant emissions(namely unburned CH4)are also expected to be affected,partly due to the lower methane concentration in the CH4/H2blendsand partly due to the faster light-off(since most of the cumulative reactor emissionsare attributable to the pre-ignition start-up phase[12]).3.2Ignition and steady-state times for CH4/H2-fueled catalytic microreactors Computed ignition(t ig)and steady-state(t st)times for all cases considered in this study are provided in Table1.The fuel/air equivalence ratio is kept constant atϕ= 0.37for all cases,with a constant ratio of90%CH4–10%H2by volume in the fuelFig.2Temporal evolutionof temperature(top graph)in an ideal batch reactor andrespective temporal evolutionof fuel mole fractions(bottomgraph)for two fuel/airmixtures:100%CH4,ϕ=0.40(solid line)and90%CH4(dashedline)–10%H2(dash-dottedline),ϕ=0.37.Symbolsdenote catalytic light-off times(t ig)Reactorexittemperature(K)80010001200140016001800Elapsed time (ms)03691215 Fuelmolefraction10-410-310-210-1Table 1Case number,microreactor material,inlet pressure p (bar),inlet velocity U IN (m/s),ignition time t ig (s)and steady-state time t st (s)Case Material p (bar)U IN (m/s)t ig (s)t st (s)1Cordierite 1 1.5012.927.52Cordierite 20.759.923.23Cordierite 30.509.221.54Cordierite 40.388.920.85Cordierite 50.308.720.56FeCr alloy 1 1.5016.531.47FeCr alloy 20.7513.225.78FeCr alloy 30.5012.224.49FeCr alloy 40.3811.723.810FeCr alloy 50.3011.423.411Cordierite 1 1.5012.127.812Cordierite 50.308.520.913FeCr alloy 1 1.5016.232.214FeCr alloy50.3011.225.7Cases 1to 10pertain to simulations with surface reactions only,while Cases 11to 14include gas-phase chemistry.In all cases ϕ=0.37with 90%CH 4–10%H 2by volumestream.As the pressure is increased from the nominal case of p =1bar up to 5bar,the inlet velocity is adjusted so that the mass throughput ρIN U IN remains constant.Previous studies on the steady-state stability of methane-fueled catalytic mi-croreactors have exemplified the impact of high pressure operation in maintaining vigorous hetero-/homogeneous combustion in microchannels,owing to the positive p +0.47pressure dependence of the catalytic reactivity of methane on platinum [7].Moreover,transient simulations on the start-up of methane-fueled catalytic mi-croreactors delineated a similar trend,in which the ignition and steady-state times during the heat-up phase of such microreactors were significantly reduced as the operating pressure was increased,thanks to the positive pressure dependence of the catalytic reactivity,even at relatively low microreactor temperatures [12].The same trend is observed from the characteristic times presented in Table 1,and is illustrated in Figs.3and 4for the hydrogen-assisted start-up of methane-fueled catalytic microreactors,where computed t ig and t st times are plotted for Cases 1–5Fig.3Ignition (t ig )andsteady-state (t st )times versus inlet pressure for Cases 1–5in Table 1.Triangles ignition times;squares steady-state times.Solid lines 90%CH 4–10%H 2fuel blend (ϕ=0.37);dashed lines 100%CH 4fuel (ϕ=0.40).The mass inflow (ρIN U IN )is constant for all cases.Cordierite microreactorInlet pressure p (bar)I g n i t i o n / S t e a d y -s t a t e t i m e (s )Fig.4Ignition (t ig )andsteady-state (t st )times versus inlet pressure for Cases 6–10in Table 1.Triangles ignition times;squares steady-state times.Solid lines 90%CH 4–10%H 2fuel blend (ϕ=0.37);dashed lines 100%CH 4fuel (ϕ=0.40).The mass inflow (ρIN U IN )is constant for all cases.FeCr alloy microreactorInlet pressure p (bar)I g n i t i o n / S t e a d y -s t a t e t i m e (s )1015202545 1.02.03.04.05.05304035(Fig.3)and Cases 6–10(Fig.4).For comparison purposes,ignition and steady-state times for the same microreactor configurations are plotted in Figs.3and 4for pure methane/air mixtures,for pressures p =1to 5bar and two microreactor wall materials.Evident from Figs.3and 4is the positive effect of high operating pressure on the ignition and steady-state times of catalytic microreactors fueled with CH 4/H 2blends,albeit with a less pronounced impact as the pressure increases beyond p >3bar,for both materials studied.More pronounced,however,is the difference between characteristic times for microreactors fueled with CH 4/H 2blends (solid curves,Figs.3and 4)and the corresponding times for pure CH 4-fueled ones (dashed curves,Figs.3and 4).By replacing 10%vol.of methane with hydrogen in the fuel stream,a sig-nificant reduction in both t ig and t st can be achieved,despite the subsequent reduction in the global fuel/air equivalence ratio.More pronounced benefits are observed at lower operating pressures,especially at atmospheric pressure.Characteristically,for a cordierite catalytic microreactor operating at p =1bar,replacing 10%vol.of methane with hydrogen in the fuel stream reduces the ignition and steady-state times by ∼47%and ∼33%respectively,while at p =5bar the corresponding reductions are ∼25%and ∼20%.In order to place the positive impact of hydrogen addition in perspective,it should be pointed out that to achieve the same reduction in t ig and t st in the pure methane-fueled cases,the operating pressure would have to be increased to p =4bar.As the operating pressure increases,the reduction in t ig and t st for the CH 4/H 2blends compared to pure CH 4cases is less pronounced;the catalytic reactivity of methane on platinum at elevated pressures is high enough [20]leading to fast light-off and attainment of steady state,such that the effect of hydrogen is attenuated.The elapsed time between light-off and steady state (t st –t ig )remains relatively constant among cases with identical fuel blends,which points out to the fact that,once ignited,the chemical energy input per unit time is more important in determin-ing how fast the microreactor will attain steady state.For CH 4/H 2fuel blends the time required for steady state is,on average,∼15%shorter compared to pure CH 4cases;this can be attributed to the fact that cases with CH 4/H 2fuel blends suffer less thermal radiation heat losses to the channel inlet and outlet enclosures owing to their lower overall equivalence ratio and to the resulting lower surface temperatures [26].3.3Effect of microreactor wall materialCordierite microreactors benefit more from hydrogen-assisted catalytic combustion than FeCr alloy ones.Figure 5presents microreactor wall temperature profiles for Cases 5and 10in Table 1(both at p =5bar)at various time instances,including the characteristic ignition and steady-state times.In both cases,rear-end ignition is observed,which is common for all the conditions in Table 1.It will be shown in the next section that,although the wall temperature profiles indicate an overall back-end ignition process,hydrogen fuel is essentially igniting at the beginning of the catalyst-coated section,leading to a pseudo -ignition seen as a slight bulging of the wall temperature profile around x =1mm (e.g.wall temperature profile at t =8.7s,Fig.5,Case 5).In contrast to catalytic combustion applications with high hydrogen contents in the fuel stream [27],the amount of H 2in the cases considered herein is not large enough to lead to front-end microreactor ignition.As evidenced from Fig.5,FeCr alloy microreactors dissipate heat much faster via heat conduction through their walls than cordierite ones (due to their higher thermal conductivity),leading to broadly distributed reaction zones.This in turn hinders fast methane catalytic light-off,leading to the higher t ig and t st times presented in Table 1.On the other hand,the steady-state temperatures are lower for the FeCr alloy material,suggesting that the material choice is a compromise between reactor thermal management issues and demand for fast light-off.3.4Effect of hydrogen addition on the catalytic microreactor start-up process In the previous sections,the positive effect of hydrogen addition in the methane/air stream has been established regarding the subsequent reduction in characteristic ignition and steady-state times compared to pure methane/air cases,for all operatingFig.5Channel wall streamwise temperature profiles during the start-up phase for Cases 5(top graph )and 10(bottom graph )at various time instances,including ignition (ign )and steady state (st )W a l l t e m p e r a t u r e (K )Channel length (mm)0246810pressures and microreactor materials considered.In this section,the underlying physics behind the heterogeneous combustion of CH 4/H 2blends in catalytic microre-actors will be investigated.Moreover,it will be clarified whether the promotion of catalytic combustion in microreactors with the addition of hydrogen is a thermal or a chemical effect.Catalytic reaction rates for CH 4and H 2fuels are presented in Fig.6for Case 5in Table 1,at two time instances before ignition and at the ignition and steady-state times.Substantial differences are evident in the reaction rate progress of the two fuels throughout the heat-up process.Since the microreactor wall temperature is initially set at T W =850K (well-above the ignition temperature of H 2on platinum),the reaction rate of hydrogen peaks already at t =0.0s and x =1mm (the beginning of the catalytic section)and is fully consumed within the first half of the reactor.Methane on the other hand retains a low reaction rate in the early pre-ignition phase,only surpassing that of hydrogen close to the ignition time.After light-off and until steady state is reached,hydrogen exothermicity has a minor contribution to the heat generated from catalytic reactions in the microreactor,in contrast to the pre-ignition phase where heat is primarily generated from hydrogen conversionChannel length (mm)246810F u e l c o n v e r s i o n r a t e (g r /m 2s )0.000.040.080.120.16Fig.6Hydrogen (solid lines )and methane (dashed lines )catalytic conversion rates along the microreactor during the start-up phase for Case 5at four time instances:ignition (ign ),steady state (st )and two pre-ignition time instances。

喷孔锥角对柴油机燃烧及排放影响的数值模拟李研芳文明李云广徐春龙强永平尹艳君（中国北方发动机研究所，山西大同037036）摘要：应用三维CFD模拟软件FIRE研究了不同喷孔锥角对燃烧过程的影响以及对NOx、碳烟、CO及UHC生成的影响。

结果表明：喷孔锥角增加，预混放热更迅速且峰值增加，扩散燃烧的峰值先降低后增加，持续时间依次降低，累积放热量逐渐降低，燃烧效率及热效率均降低。

soot及UHC排放增加，NOx排放先增加后降低，CO先减少后增加。

关键词：喷孔锥角，燃烧效率，热效率，排放，CFD主要软件：A VL FIRE1.前言直喷式柴油机的燃烧过程的好坏主要取决于油、气、室三者之间的相互配合。

三者的合理匹配决定燃油与空气的混合均匀程度，并最终决定了柴油机的性能。

本文主要就不同的喷孔锥角对柴油机燃烧过程及排放进行了数值模拟研究。

探讨了喷孔锥角对柴油机燃烧效率、热效率及排放的影响。

2.数学模型A VL·FIRE是基于有限体积法，计算域被网格分成许多控制体，对每个控制体分别求多维N-S方程标量输运方程，方程的数值解反映气流及喷雾的运动，燃油的蒸发、混合及燃烧等一系列物理化学过程。

采用simplec算法对流场和其他标量进行解耦。

计算中所采用的子模型如表1中所示。

燃烧模型ECFM-3Z是基于火焰面密度输运方程以及能描述非均质湍流预混及扩散燃烧的混合模型，其对燃烧的具体描述见图1。

表1 计算中所采用的子模型图1 ECFM-3Z模型3.发动机基本参数柴油机部分参数如表2所示。

表2柴油机部分参数喷孔数/个8喷孔锥角/°145/155/165考虑到燃烧室的轴对称性结构，为减少计算时间，计算区域与喷孔数相对应，即选取燃烧室的八分之一，图2为位于上止点时刻的燃烧室网格模型，计算时间区域从进气门关闭时刻到排气门开启时刻。

图2 燃烧室网格图4.结果分析（1）滞燃期内流场分析由图3可知，喷孔锥角增加，气相燃油距活塞壁的距离缩短，燃油撞击壁面后更容易形成小油滴，一定程度上加速了滞燃期内液相燃油与空气的混合，所以相对而言，喷孔锥角越大，燃烧初始阶段，燃油消耗较快，放热较多，但上止点之前，由于活塞上行，这部分热量的贡献是负功。