Ansys CFD 2021 R1

燃烧和化学反应新功能介绍

井文明 Ansys 中国
[email protected]
个人简介

• ANSYS 流体高级工程师
• ANSYS 燃烧及电池行业专家
‐ 《ANSYS电池仿真实例与详解-流体篇》第一作者
‐ 《ANSYS电池仿真实例与详解-结构篇》第二作者
• 北航硕士燃烧专业/西工大本科

Ansys 能源在线 个人企业微信

2
创新应用布局——从提供“工具”到提供“核心业务解决方案”

流体仿真
2003
光学与声学仿真
2018 后处理 LS-Dyna 显式动力学
化学反应/燃烧 流体仿真 2017 2019
2014 2006 飞行器结冰
2015

结构仿真
1970

复合材料分析与优化
2013

概念建模与前处理 Granta材料数据库
2014 高安全性嵌入式系 2018
半导体芯片 半导体大数据分析
统仿真 2011 2015
2013

电磁场仿真 Motion多体动力学
2008 2018
增材制造仿真
2017 功能安全与可靠性 半导体时序分析
2016 2017 数据截至 2019年9月
电大尺寸电磁计算
2015
Outline

• Ansys燃烧解决方案
• Ansys CFD 2021R1 燃烧及化学反应新功能

4
化学反应典型场景

Biochemical and Biomedical CVD Electro-Chemical Reactions

Battery

Micro-reactors

Anode
off-gas

Fire and Fire Protection

Fuel Cells Pollution Control

5
 燃烧典型场景

IC Engines Rocket Propulsion Gas Turbine Combustors

Cement Kilns Glass Melting Furnaces Gas Flares Steel Making Boilers

6
ANSYS Fluent applied to combustion applications
Refinery furnace Glass furnace Steel furnace

Courtesy of Danieli Centro

Combustion S.P.A
Courtesy of John Zink Company
Gasification Burner
Gas Turbine (coal,biomass)

Courtesy of MTU
Aero Engines GmbH
 CHEMKIN
Focuses on chemistry using engineering
approximations of the flow: 0-D, 1D, 2D
Plug-flow Reactor (PFR)

Reduced order models to allow detailed kinetics Opposed-flow

Diffusion Flame
to investigate combustion behavior

Perfectly Stirred Reactor (PSR)

Allows combustion designers to investigate Partially Stirred Reactor

• Fuel effects, i.e. varying fuel composition

T
• Ignition delay times Shear-layer Channel-flow Reactor
v
• Laminar flame speeds and table generation
• Flame extinction H2 CO
CH4 H2O
• Mechanism optimization T
Flame-speed Calculation
• Particulate formation, oxidation, growth v
Equilibrium Calculator
• Utilize reactor networks for modeling larger systems
Pre-mixed Flame
 ANSYS Model Fuel Library (MFL) provides the chemistry

‐ Enables fuel-surrogate modeling

‐ Allows representation of complex “real” fuels
‐ Consists of validated combustion mechanisms
‐ ~65 fuel components
‐ Validated against > 500 sets of experimental data
‐ Mechanisms are self-consistent
‐ Common core chemistry
‐ Consistent rules for reaction rates
‐ Validated fuel-blending behavior
‐ Updated and extended annually
Prepare a fuel model with Chemkin Reaction Workbench

• Accurately represent:
Model Fuels
– Combustion & emissions behavior Library
– Chemical and physical properties

Fuel properties Surrogate Blend

to match Optimizer
• Chemical properties
• Octane, Cetane numbers Model-fuel blend
• Distillation Curve

Mechanism Fuel Model ready

Reduction for ANSYS CFD
 ANSYS Forte is designed for IC-engine CFD simulations
• Set up tailored to IC engines
• Automatic mesh generation
‐ On-the-fly, dynamic, Cartesian
‐ Solution-adaptive mesh refinement
• Well validated kinetics models
• Fast chemistry solver
‐ Predict ignition, emissions, fuel effects
‐ Predict flame-front location and knocking in SI engines
‐ Particle size tracking (soot)
• Well validated spray and spray-wall models
‐ Capture physics accurately and with mesh insensitivity
Combining ANSYS Products Fluent 3-D
CFD
+ global chemistry + reduced
chemistry

+ progress variables
Automatic Mapping Automated
to Zones Chemistry
Reduction
Table look-up vs.
Progress variable
Equivalent Oxidizer

Reactor
Networks
(ERNs) 0- or 1-D
x r models
Fuel

Energco
Model Fuel Chemkin
Library (MFL) Reaction Workbench
 ANSYS offers comprehensive solutions for Reacting Flow
Chemistry-centric

CFD-centric
CFD Enterprise CHEMKIN Enterprise
CFD Premium
Fluent Chemkin-Pro
• General flow simulation • 0-D, 1-D, 2-D reactor models
• PDF combustion models Forte • Flame simulators
• Solid, liquid, gaseous fuels • CFD for IC Engine • Table generators
• Radiation modeling • Automatic, • Reaction Path Analysis
• Real Gas modeling adaptive mesh • Uncertainty analysis
• Particulate and pollutant • Spark ignition, • Real-gas Not in academic license
modeling Flame tracking
Energico
CFX • Spray physics
• Detailed kinetics • Automated reactor-network
• Turbomachinery analysis for emissions
Model Fuel
Library Reaction Workbench
• Fuel model preparation
and reduction
Image courtesy of GE Aviation

CFD-centric Chemistry-centric
 Example: Gas turbine Combustion Modeling Aspects

Pressure Drop/Flow Split Model Fuel Library

Combustion
Fuel Sprays Auto-Ignition
Chemkin PRO

Emissions
Reaction Workbench

Chemkin CFD within

Fluent
Alternate/Dual Fuels
Durability: CHT/Thermal Stresses
Chemistry Energico

Forte
Basic concepts
• 燃烧类型 • 固体燃烧相关
‐ 预混、非预混、部分预混 ‐ 干燥、脱挥、热解、气化
‐ 缓燃、爆震 • Other
• 火焰稳定相关 ‐ 表面反应、CVD、高速离解、SCR、碳捕捉等
‐ 稳火、传焰、回火 ‐ 污染排放
‐ 熄火、热声震荡 • 无量纲数
• 组分相关 ‐ 𝑹𝒆 =
𝑰𝒏𝒆𝒓𝒕𝒊𝒂𝒍 𝑭𝒐𝒓𝒄𝒆 𝝆𝑼 𝑳
= 𝝁
‐ 当量比、化学当量 𝑽𝒊𝒔𝒄𝒐𝒖𝒔 𝑭𝒐𝒓𝒄𝒆
𝑳ൗ
油气比、化学当量油气比 𝑴𝒊𝒙𝒊𝒏𝒈 𝑻𝒊𝒎𝒆 𝑺𝒄𝒂𝒍𝒆 𝒌Τ
‐ ‐ 𝑫𝒂 = ~ 𝝆𝒂𝒅
𝑼
~ 𝝆𝒂𝒅
𝜺
𝑪𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝑻𝒊𝒎𝒆 𝑺𝒄𝒂𝒍𝒆
‐ 余气系数 ൗ𝑹𝒔𝒍𝒐𝒘 ൗ𝑹𝒔𝒍𝒐𝒘
𝑪𝒐𝒏𝒗𝒆𝒄𝒕𝒊𝒐𝒏 𝑺𝒑𝒆𝒆𝒅 𝑳
‐ 摩尔分数、质量分数、体积分数 ‐ 𝑴𝒂 =
𝑨𝒄𝒐𝒖𝒔𝒕𝒊𝒄 𝑺𝒑𝒆𝒆𝒅
=
𝒄
• 火焰速度 ‐ 𝑲𝒂 =
𝜺/𝝂
~
𝒄𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝒕𝒊𝒎𝒆 scale
𝒔 𝟎𝑳/𝜹
层流/湍流火焰速度
𝑲𝒐𝒍𝒎𝒐𝒈𝒐𝒓𝒐𝒗 𝒕𝒊𝒎𝒆 𝒔𝒄𝒂𝒍𝒆
‐
𝑪𝒐𝒏𝒗𝒆𝒄𝒕𝒊𝒐𝒏 𝑯𝒆𝒂𝒕 𝑻𝒓𝒂𝒏𝒔𝒇𝒆𝒓 (𝝆𝑼𝑪𝒑𝑻)𝒊𝒏𝒍𝒆𝒕
• 极限相关 ‐ 𝑩𝒐 =
𝑹𝒂𝒅𝒊𝒂𝒕𝒊𝒐𝒏 𝑯𝒆𝒂𝒕 𝑻𝒓𝒂𝒏𝒔𝒇𝒆𝒓
~
𝝈 𝑻𝒂𝒅 𝟒
‐ 贫油熄火极限
‐ 富油熄火极限 ‐ Weber :惯性力与表面张力之比
• 多相流相关 ‐ Lewis Number:热扩散系数和质量扩散系数比
‐ 初始破碎、二次破碎、雾化 ‐ Stokes number:颗粒松弛时间和流体特征时间的比
‐ 蒸发、沸腾 ‐ Schmidt number:动量扩散率与质量扩散率的比值

15
Turbulence: Reynolds (Time) Averaged Species Equation


 ( Y ) +
 
 ( u Y ) +

(u"i Y"k ) =
 
 D k
Yk 
 + R k
 t  x i x i x i x i 
k i k


{unsteady term} convection convection molecular mean

(zero for by mean by turbuent diffusion chemical
steady flows) velocity velocity fluctuations source term

Yk , Dk , Rk are the kth species mass fraction, diffusion coefficient and chemical source term respectively

Turbulent flux term modeled by mean gradient diffusion as

u"i Y"k =  t /Sct   Yk / xi , which is consistent in the k-e context

Gas phase combustion modeling focuses on Rk

− Arguably more difficult to model than the Reynolds stresses for turbulence
Why Reacting Flow Modeling is Complex?

• Turbulence
• Most industrial flows are turbulent
• Full resolution is not possible: wide range of
time and length scales involved
• Chemistry
• Realistic chemical mechanisms cannot be
described by a single or two reaction equations
▪ Tens of species, hundreds of reactions
▪ Known in detail for only a limited number of fuels
• Stiff kinetics: Wide range of reacting time scales
• Turbulence-chemistry interaction
• Reaction rates sensitive to local changes due to
enhanced mixing
Why Reacting Flow Modeling is Complex?

• Turbulence • DNS
• Most industrial flows are turbulent • LES
• RANS
• Full resolution is not possible: wide range of • Hybrid
time and length scales involved
• Chemistry
• Realistic chemical mechanisms cannot be
described by a single or two reaction equations
▪ Tens of species, hundreds of reactions
▪ Known in detail for only a limited number of fuels
• Stiff kinetics: Wide range of reacting time scales
• Turbulence-chemistry interaction
• Reaction rates sensitive to local changes due to
enhanced mixing
Why Reacting Flow Modeling is Complex?

• Turbulence
• Most industrial flows are turbulent
• Full resolution is not possible: wide range of
time and length scales involved
• Chemistry
• Realistic chemical mechanisms cannot be • Simplify the chemistry
• Decouple chemistry from flow
described by a single or two reaction equations • Model detailed chemistry(stiff)
▪ Tens of species, hundreds of reactions
▪ Known in detail for only a limited number of fuels
• Stiff kinetics: Wide range of reacting time scales
• Turbulence-chemistry interaction
• Reaction rates sensitive to local changes due to
enhanced mixing
Why Reacting Flow Modeling is Complex?

• Turbulence
• Most industrial flows are turbulent
• Full resolution is not possible: wide range of
time and length scales involved
• Chemistry
• Realistic chemical mechanisms cannot be
described by a single or two reaction equations
▪ Tens of species, hundreds of reactions
▪ Known in detail for only a limited number of fuels
• Stiff kinetics: Wide range of reacting time scales
• Turbulence-chemistry interaction • Da >>1 fast chemistry
• Reaction rates sensitive to local changes due to • Da ~ 1 slow chemistry
enhanced mixing
Fast vs. Slow Chemistry

Fast Chemistry Slow Chemistry

• Da >> 1 • Da ~ 1
• Reactions limited by turbulent mixing • Reactions limited by chemistry and turbulence
interactions
• Selection of turbulence closure model is • Turbulence/chemistry interactions are important
important
• Combustion in • Selection of reaction mechanism is important
• Furnaces • Reactions associated with
• Boilers • Pollutants formation
• Gas Turbines • Ignition and Extinction
• Gasifiers, Incinerators • Chemical Vapor Deposition (CVD)
• Flares, etc. • Non-Equilibrium Phenomenon
• Air dissociation at hypersonic speed
 Overview of Combustion Modeling
Models for Additional Physics

Transport Equations Solid or Liquid Fuels

• Mass (DPM)
Fast Chemistry Models • Droplet/particle dynamics
• Momentum
• Eddy Dissipation model • DEM collisions
• Energy
• Premixed model • Evaporation
• Turbulence
• Equilibrium model • Devolatilization
• Steady Laminar Flamelet model • Heterogeneous reaction
Turbulence Models • Flamelet Generated Manifold model
• RANS: k-e, k-w, RSM….. • Partially premixed model Radiation Models
• LES, DES, SAS…. • P1, DO (Gray/Non-gray)

Chemistry Solution Finite Rate (Slow) Chemistry Models

• Laminar Finite rate model Real Gas Effects
• Species OR
• EDC • SRK, ARK, RK, PR, UDRGM
• Mixture fraction OR
• Composition PDF
• Progress variable OR
Pollutant Models
• ……
Combustion Models • NOx, Soot, SOx
Combusting Flow Solution
 Reacting Flow Models in Fluent
Flow Configuration
Non-Premixed Partially Premixed
Premixed Combustion
Combustion Combustion
Finite-Rate/Eddy-Dissipation Model (Species Transport)
Partially Premixed Model
Premixed Combustion Non-Premixed Model Reaction Progress Variable
Fast Chemistry
Model Mixture Fraction +
Closures Mixture Fraction
Chemistry

Reaction Progress
Variable Chemical Equilibrium
Steady Laminar Flamelet Model
Finite Chemistry Flamelet Generated Manifold Model (Premixed/Diffusion)
Closures Unsteady Laminar Flamelet Model
Finite Rate Laminar Finite-Rate Model (Species Transport)
Chemistry Eddy-Dissipation Concept (EDC) Model (Species Transport)
Models Composition PDF Transport Model
ANSYS Fast Combustion
Modeling
Eddy Dissipation Model (EDM)

‐ Minimise the influence of chemistry

• A good assumption for fast chemistry (Da >> 1)
• Most industrial flames are limited by turbulent mixing
‐ D Brian Spalding (1971) proposed the Eddy Break-Up
(EBU) model D Brian Spalding

• Introduced eddy lifetime, k/e

‐ Bjorn F Magnussen and B. H. Hjertager (1976) extended
the EBU model to non-premixed and partially premixed
combustion
Bjorn F Magnussen
• Eddy Dissipation Model (EDM)

D. B. Spalding, Chemical Eng. Sci. 26-1 (1971), 95-107

th
B. F. Magnussen and B. H. Hjertager, 16 Symposium (Int.) on Combustion (1976) p. 719
Eddy dissipation model (cont…)
‐ Reaction is mixing limited
‐ Chemistry is described by a global reaction mechanism
• 1 or 2 steps
‐ Reaction rate is governed by large-eddy mixing time scale
• Eddy break-up (EBU) or turbulence time scale, k/e
• Rate of production of a species, i due to reaction, r
▪ 𝑹𝒊𝒓 = 𝒎𝒊𝒏 𝑹𝒊𝒓(𝑹𝒆𝒂𝒄𝒕), 𝑹𝒊𝒓(𝑷𝒓𝒐𝒅)
𝝐 𝒀𝒊
▪ 𝑹𝒊𝒓 (𝑹𝒆𝒂𝒄𝒕) = 𝒗𝒊𝒓 𝑴𝑾𝒊 𝑨 𝝆 𝒌 𝒗𝒊𝒓 𝑴𝑾
𝒊
𝝐 σ𝒑 𝒀𝒑
▪ 𝑹𝒊𝒓 (𝑷𝒓𝒐𝒅) = 𝒗𝒊𝒓 𝑴𝑾𝒊 𝑨 𝑩 𝝆
𝒌 σ𝑵
𝒋 𝒗𝒋𝒓 𝑴𝑾𝒋
▪ A and B are constants → A = 4.0 and B = 0.5 suggested by
Magnussen
o Works fine for most of the problems
o Sometimes needs tuning to get required temperature distribution

▪ The rates are not function of temperature

Non-premixed model
‐ Burke & Schumann (1928), Bilger (1976) suggested a closure based on
mixture fraction (f or Z) for turbulent diffusion flames

Diffusion flame
‐ Assumptions involved
• Separate streams for fuel and oxidizer (diffusion flames)
• Equal species diffusion coefficients
▪ Suitable for turbulent flows as turbulent diffusion overwhelms
• Unity Lewis number
• Low Mach number flow Robert W. Bilger

‐ With these assumptions:

• Species transport equations can be reduced to a single equation for conserved scalar,
mixture fraction

Fuel →
Oxidizer →
• For mixture fraction equation, the reaction source terms cancel out since atoms are
conserved in chemical reactions
S. P. Burke and T. E. W. Schumann, Indust. Eng. Chem. 20 (1928) p. 998
R. W. Bilger, Combust. Sci. Technol. 13 (1976) p. 155
Non-premixed model (cont…)
‐ The Favre mean mixture fraction equation
𝜕 𝝁
𝝆𝒇 ത = 𝛁. 𝒕 𝛁𝒇
ത + 𝛁. 𝝆𝒗𝒇 ത
𝜕𝒕 𝝈𝒕
‐ For chemical equilibrium, scalars like Yi,  and T are uniquely related to the mixture
fraction (f)
• For adiabatic systems: ∅𝒊 = ∅𝒊 𝒇
• For non-adiabatic systems: ∅𝒊 = ∅𝒊 𝒇, 𝒉
‐ In turbulent flows
• Mean values of scalars are of interest
• The probability density function, P(f) can be employed
▪ Describes the temporal fluctuations of f in the turbulent flow
▪ Mathematical functions like  or double  functions can be used
▪ Transport equation for variance of mean mixture fraction would be required
𝜕 𝝁𝒕 𝟐 𝜺
𝝆𝒇′𝟐 + 𝛁. 𝝆𝒗𝒇′𝟐 = 𝛁. ത
𝛁𝒇′𝟐 + 𝑪𝒈 𝝁𝒕 𝛁𝒇 − 𝑪𝒅 𝝆 𝒌 𝒇′𝟐
𝜕𝒕 𝝈𝒕
▪ t = 0.85; Cg = 2.86 and Cd = 2.0
 Non-premixed model (cont…)

• Density weighted mean species mass fraction or temperature

𝟏
▪ ∅ഥ𝒊 = ‫𝒇𝒅 𝒇 𝒊∅ 𝒇 𝑷 𝟎׬‬
• Overall calculation procedure

Chemical Equilibrium PDF Shape

∅𝒊 = ∅𝒊 𝒇 P 𝒇 = P 𝒇, 𝒇′𝟐

Mean scalar value

𝟏
∅ഥ𝒊 = න 𝑷 𝒇 ∅𝒊 𝒇 𝒅𝒇
𝟎

• ∅ഥ𝒊 values can be pre-tabulated for given chemistry model (like chemical
equilibrium) and assumed PDF shape and looked up whenever required
Flamelet model
‐ Effect of turbulence on flame structure
• Flame stretching and straining
▪ Increase species and temperature gradients across the flame
▪ Cause quick diffusion of heat and radicals
▪ Allow less time for species to reach chemical equilibrium
o Degree of local non-equilibrium

• Not considered in equilibrium based non premixed model

Norbert Peters
‐ Prof. Peters provided a formulation
• Accounts for these flow dependent effects
• Turbulent flame modelled as ensemble of stretched laminar-opposed flow diffusion flames
• Non-equilibrium chemistry and turbulence are coupled using the statistical description between
▪ Mixture fraction and the instantaneous scalar dissipation rate

N. Peters, Prog. Energy. Combust. Sci. 10 (1984) p. 319

N. Peters , 21st Symposium (Int.) on Combustion (1986) p. 1231
Flamelet structure of a diffusion flame Z = Zst

‐ Mixture fraction as a function of space and time, Z = Z (xi,t)

‐ Surface of stoichiometric mixture fraction can be determined as Zst = Z (xi,t)
• Combustion takes place in the vicinity of this surface
▪ If the local mixture fraction gradients are high
‐ Assuming very thin flame (called a ‘flamelet’), can transform equations for
species and temperature from 3-D to 1-D:
Z=0
Z=1

Equilibrium
•  (Kai) is the scalar dissipation rate (1/s) = strain rate of flame front
▪  →0; The chemistry tends to equilibrium
▪ Chemistry departs from equilibrium with increase in 
due to aerodynamic straining
Flamelet modeling approach

‐ Turbulent flame brush is modeled as an ensemble of

discrete diffusion laminar flamelets
‐ Density-weighted average species mass fractions and
temperature
∅ഥ𝒊 = ‫𝒇 𝑷 ׭‬, 𝒙𝒔𝒕 ∅𝒊 𝒇, 𝒙𝒔𝒕 𝒅𝒇𝒅𝒙𝒔𝒕
‐ f and  are assumed to be statistically independent
• i.e. 𝑷 𝒇, 𝒙𝒔𝒕 = 𝑷 𝒇 𝑷 𝒙𝒔𝒕
• Beta function assumed for 𝑷 𝒇
• Delta function or log normal distribution assumed for 𝑷 𝒙𝒔𝒕
‐ Mean scalar dissipation rate is calculated as
𝑪𝒙 𝜺𝒇′𝟐
𝒙𝒔𝒕 =
𝒌
• Cx = 2.0 Instantaneous Turbulent
flamelets flame brush
Flamelet modeling approach

‐ Turbulent flame brush is modeled as an

ensemble of discrete diffusion laminar
flamelets
‐ Density-weighted average species mass
fractions and temperature
∅ഥ𝒊 = ‫𝒇 𝑷 ׭‬, 𝒙𝒔𝒕 ∅𝒊 𝒇, 𝒙𝒔𝒕 𝒅𝒇𝒅𝒙𝒔𝒕
‐ f and  are assumed to be statistically
independent
• i.e. 𝑷 𝒇, 𝒙𝒔𝒕 = 𝑷 𝒇 𝑷 𝒙𝒔𝒕
• Beta function assumed for 𝑷 𝒇
• Delta function or log normal distribution
assumed for 𝑷 𝒙𝒔𝒕
‐ Mean scalar dissipation rate is calculated as
𝑪𝒙𝜺𝒇′𝟐
𝒙𝒔𝒕 =
𝒌
• Cx = 2.0
Turbulent premixed combustion model

‐ Concept of laminar premixed

Premixed flame
combustion was introduced earlier
• Two distinct states of premixed
mixture
▪ Burned and unburned
• Separated by a flame front
‐ Effect of turbulence
• To wrinkle the flame sheet, increase
the flame area
and effective flame speed Burned state

• Three flame regimes

+ →
Oxidizer
▪ Wrinkled laminar flame regime (𝜹 ≤ 𝒍𝒌 )

Fuel
▪ Flamelets in eddies regime (𝒍𝟎 > 𝜹 > 𝒍𝒌 )
▪ Distributed regime (𝜹 > 𝒍𝟎 ) Unburned state

𝜹 → Flame thickness
𝒍𝒌 → Kolmogorov length scale (smallest eddies)
𝒍𝟎 → Integral length scale (largest eddies)
Turbulent premixed combustion model (cont...)

‐ Two options to model

• Track the reaction progress (c equation)

• Track the flame front (G equation) [transient only]

• Calculate mean progress variable using Gaussian distribution

• Variance of mean G → Transport equation to algebraic

OR
 Turbulent Flame Speed Closure Model

Example: Zimont turbulent velocity correlation:

1/2 −1/4 1/4

𝑈𝑇 = 𝐴𝐺𝑢′3/4𝑈𝐿 𝜆𝑢 𝑙𝑡
where
− A = Turbulent flame speed factor (0.5 for CH4/air, 0.6 for H2/air)
− u´ = Integral turbulent velocity fluctuation = 2/3k
− UL = Laminar flame speed
− l u = thermal conductivity of unburnt mixture 𝑈𝑇
− 2/3
lt = integral turbulent length scale = k /e
− G = Flame front stretch factor

𝑈𝐿
u
´
Partially premixed combustion model

‐ Premixed flames systems with non-uniform fuel-oxidizer mixtures

• Premixed jets discharging into a quiescent atmosphere
• Lean premixed combustors with diffusion pilot flames and/or cooling air jets, and imperfectly
premixed inlet
‐ Modeling approach
• Premixed approach (c or G equation) to track the flame
• Equilibrium or flamelet approach to model the composition of burned products
• Flame brush indicated by a value of 𝒄ത between 0 and 1

Premixed combustion Partially premixed combustion Non premixed combustion

Progress variable concept (c) Progress variable (c) Mixture fraction (f)
+
Mixture fraction(f)
Partially premixed combustion model (cont…)

‐ Assumptions in both f and c models apply

• Chemistry at or near equilibrium
• Equal species diffusion coefficients
• Two or three distinct stream compositions
• Unity Lewis number
‐ Assumed PDF shapes for evaluation of mean scalars values
~ 1 1
 =    ( f , c) P( f , c) df dc
0 0
• f and c are assumed to be statistically independent
P ( f , c) = P ( f ) Pc (c)
• Assumed PDF functions: P(f) →Beta; P(c) → Double delta
• Thin flame assumption: Only unburnt reactants and burnt products exist
~ 1 1
~
0
~

 = c b ( f ) P ( f )df + (1 − c ) u ( f ) P ( f )df
0
Flamelet Generated Manifold (FGM) model
‐ Laminar diffusion flamelets
• Turbulent flames an ensemble of laminar flames whose
internal structure is not significantly altered by the
turbulence
• Chemical non-equilibrium effects due to local turbulent
strain only
• Cannot model ignition, slow chemistry (e.g. NOx) and
quenching
‐ Flamelet Generated Manifold (FGM)
• The scalar evolution in a turbulent flame can be
approximated by the scalar evolution in a laminar flame
• No assumption on flame structure
• Flamelets solved in reaction progress space
▪ Predominantly for premixed systems
• Can model ignition, slow chemistry, quenching
▪ Far from equilibrium (e.g. c = 0.5)
• Realistic results for air (c = 0) mixing with unburnt mixture (ahead of
flame) or burnt mixture (behind flame)
ANSYS Finite Rate
Combustion Modeling
Introduction

• Detailed chemical kinetics

‐ Tens of species and hundreds of reactions
‐ Each species participate in a series of
reaction steps
• Produced in some steps and destroyed in some
other steps
‐ Example: H2-O2 reaction mechanism
‐ Suitable for modeling
• Flame ignition and extinction
• Pollutants (NOx, CO, UHCs)
• Slow (non-equilibrium) chemistry
• Intermediate species of interest (OH…)
Mechanisms for practical fuels
Large Sizes C11H22O2 -6
Chemical Stiffness
Methyl
Decanoate
10
4 (LLNL) Typical flow time

Shortest Species Time Scale, Sec

ofofReactions 10 C16 (LLNL)
C12 (LLNL)
C10 (LLNL) C14 (LLNL)
reactions

iso-ocatane (LLNL) PRF -9 Ethylene,

n-heptane (LLNL)
10 p = 1 atm
iso-ocatane (ENSIC-CNRS)
T0 = 1000K
CH4 (Konnov)
nC7H16
3 skeletal iso-octane (Lu & Law)
10
Number

USC C1-C4 skeletal n-heptane (Lu & Law)

n-butane (LLNL) n-Heptane,
Number

USC C2H4 -12

1,3-Butadiene
C1-C3 (Qin et al) DME (Curran) 10 p = 50 atm
GRI3.0 neo-pentane (LLNL) T0 = 800K
GRI1.2 C2H4 (San Diego)
pre-2000
before 2000
CH4 (Leeds)
2 2000
2000 to– 2005
2005
10 after 2005 -15
post-2005
10
1000 2000 3000
1 2 3 4
10 10 10 10
Temperature , K
Number
Numberofof Species
species

• Characteristics of detailed reaction mechanisms

‐ Large number of species
‐ Wide range of chemical timescales (Stiffness)
• Small times scales, ~1e-8 (Reactions involving radical species)
• Large times scales, ~ 1e-2 (CO oxidation, NOx formation)
Types of detailed reaction mechanisms

‐ Three types
• Detailed, Skeletal and Reduced mechanism
‐ Detailed mechanism
• Large number of species
• Large number of elementary reactions
‐ Skeletal mechanisms
• Shortened detailed mechanisms
• Unimportant species and reactions eliminated
‐ Reduced mechanisms
• Further simplified detailed or skeletal mechanisms
• Quasi-Steady State Assumptions (QSSA)
Detailed chemistry models

‐ Laminar Finite Rate (LFR) model

• Laminar flames
• Turbulent flames with weak turbulence-chemistry interaction
‐ Eddy Dissipation Concept (EDC) model
• Assumes reaction occurs in small turbulence scales
• Turbulent flames with turbulence-chemistry interaction
‐ Composition PDF transport
• Turbulent flames with rigorous turbulence-chemistry interaction
• Computationally most expensive
Laminar Finite Rate model

‐ Laminar flows
𝜕𝒀𝒊 𝜕𝒀𝒊 𝜕 𝜕𝒀𝒊
• 𝝆 + 𝝆𝑼𝒊 = 𝝆𝑫 +𝑺
𝜕𝒕 𝜕𝑿𝒊 𝜕𝑿𝒊 𝜕𝑿𝒊
• Diffusivity modeling affects the accuracy
‐ Turbulent flows
෪𝒋
𝜕𝒀 ෪
෪𝒋 𝜕 𝒀𝒋 = 𝜕 ෪𝒋
𝜕𝒀 𝜕 ෫
• 𝝆
ഥ +𝝆
ഥ𝑼 𝝆𝑫 − ഥ 𝑼" 𝒋 𝒀"
𝝆 +෩
𝑺
𝜕𝒕 𝜕𝑿𝒋 𝜕𝑿𝒋 𝜕𝑿𝒋 𝜕𝑿𝒋 𝒋
• ෩
𝑺 evaluated from mean temperature
• Can predict ignition characteristics correctly
• No corrections for temperature or species concentration fluctuations
▪ Works fine for weak turbulence-chemistry interaction
• For larger turbulent fluctuations
▪ Need to use smaller mesh size and/or time step size
Eddy Dissipation Concept model

‐ Extension of LFR model to account for

turbulence-chemistry interaction
‐ Proposed by Magnussen in 1981
Bjorn F Magnussen
‐ Assumptions
• Reactants are mixed at molecular level in the fine
turbulent structures
▪ Of the order of Kolmogorov length scale
• Reactions take place within these structures
• Entire volume of fine scale structures is a fraction
of total fluid volume

Fine structures within a computational cell

B. F. Magnussen, 19th AIAA Sc. meeting, St. Louis, USA, 1981

EDC reaction rate evaluation

‐ Length fraction of the fine scales

 e 
1/ 4

 * = C  2   Ret− 0.25
k 
‐ Volume fraction of the fine scales =  * 3
‐ Life-time of small scales
 
1/ 2

 = C  
*
 t Kolmogorov
e 
‐ Reactions proceed within the fine scales over the time, 𝝉∗
• Assuming constant pressure reactor
‐ Mean reaction rate in mean species equation
 *
2

Si = ( Yi* − Yi ) Fine structures within a computational cell

*(1− *3
) * Quantities are fine scale quantities: C = 2.1377 and C = 0.4082
Composition PDF transport model

‐ RANS approach solves Favre averaged species and energy equations

• Prone to errors for flows with strong turbulence-chemistry interaction
‐ A transport equation for joint probability density function (PDF) is obtained from
species and energy equations
• Proposed by S.B. Pope (1976)
𝜕 𝜕 𝜕 𝜕 𝜕 𝟏 𝝏
• 𝝆𝑷 + 𝝆𝒖𝒊 𝑷 + 𝝆𝑺𝒌 𝑷 = 𝝆 𝒖" 𝒊 𝝋 𝑷 + 𝝆 𝑱𝒊,𝒌 𝝋 𝑷
𝜕𝒕 𝜕𝑿𝒊 𝜕𝝋𝒌 𝜕𝑿𝒊 𝜕𝝋𝒌 𝝆 𝜕𝑿𝒊

▪ P is Favre joint composition PDF

▪ 𝝋 is the composition space = 𝝋 (Y1, Y2, Y3 …… YN , T)
▪ 𝑨 𝑩 denotes the probability of event A, given the event B occurs
▪ Ji,k is molecular diffusion flux vector

S B Pope
S.B. Pope, Combustion and Flame, 27, 299-312 (1976)
Composition PDF solution approaches
Two solution approaches
‐ Lagrangian Monte Carlo method
• Appropriate for high-dimensional equations
• Involves notional particles which move randomly
▪ Through physical space due to particle convection
▪ Through composition space due to molecular mixing and reaction
▪ Sum of the mass of particles in a cell equals the cell mass
▪ Computationally expensive due to large number of particles to represent the
PDF
• Fractional steps solution for convection, diffusion and
reaction
‐ Eulerian solution method
• Assumes a shape for the PDF
• Multi-dimensional PDF shape is assumed as a product
of delta functions
• Transport equations are solved implicitly
• Computationally economical
 Choosing a Radiation Model
• The radiation model selected must be appropriate for the optical thickness of the
system being simulated
Model Optical Thickness Computational Expense
Surface to surface model (S2S) 0 When optical thickness = 0, S2S has comparable accuracy with DO and
MC at less computational expense
Solar load model 0 (except window panes) Very low computational expense for solar radiation problems
compared to the DO model
Rosseland >5 Very inexpensive but limited in applicability
P-1 >1 Reasonable accuracy for moderate cost
Discrete ordinates model (DO) All Very comprehensive and accurate model but more computationally
expensive than S2S for optically thin systems and more expensive than
P1 and Rosseland for optically thick systems
Discrete transfer method (DTRM) All Cheaper then DO but not available in parallel so rarely used
Monte Carlo model (MC) All The most accurate model but also the most computationally expensive

• In terms of accuracy, DO and MC are most accurate (S2S is accurate for optical thickness = 0)
Introduction Overview Wall BCs Applications Summary
Pollutant Modeling Options in Fluent /fluent中模拟污染物的方法

Post Processing 后处理的方式

• Pollutants governing equations with frozen flow Field

• ANSYS CFD has this option for NOx, Sox and Soot predictions

Fully Coupled 完全耦合的方式

• Pollutants governing equations solved coupled with flow equations

• Computationally more expensive

Decoupled Detailed Chemistry 将详细化学解耠的方式

• A finite rate chemistry solution for pollutants in post processing step
• Suitable for slow forming pollutants (NOx and SOx)
Some tips and tricks

• Keep the species with abundant mass fraction as the last species
• Use temperature dependent Cp for included species
• WSGGM-domain-based for absorption coefficient if radiation employed
• Start with non-reacting flow
• Patch small values for product species/high temperature in flame region (non-PDF
model)
• Adjust URF during calc.
‐ ~0.9 for lower species and energy
‐ ~0.7 for density for PDF model
• Final solution with URFs 1 and radiation included

52
 Ansys Fluent 2021R1
燃烧模型更新

53
Perforated Wall Model 发散冷却孔壁面模型
• 对发散冷却孔的精确模拟对于预测压降，回流裕度，流量
分配和火焰筒壁面温度非常重要
• 在网格离散过程中，大量的小孔使得捕捉所有细节不现实
• 发散冷却孔壁面模型可在不对小孔物理建模的前提下在网
格中去模拟冷却孔的行为

54
 Perforated Wall Model发散冷却孔壁面模型
概念
火焰筒发散冷却孔
冷侧
• 不为火焰筒壁厚划分网格
• 在热侧施加入射边界条件（源）
无网格
• 在冷侧施加出抽取边界条件（汇）
热侧
模拟方法
分布式：
均匀的入射/ • 指定入射/抽取表面
抽取 • 指定入射方向及流量系数
• 使用平均后的源/汇项
• 无须指定每个冷却孔

• 指定冷侧/热侧表面
• 指定每个孔的轴向，直径，流量系数
离散式： • 源/汇项作用在相应的网格上
非均匀的入
射/抽取

55
 Perforated Wall Example发散冷却孔壁面模型案例
对冷却孔物理建模及使用发散冷却孔壁面模型温度及混合物分数的对比

Temperature, F Mixture Fraction

对冷却孔物理建模

模型 出口温度[F] 压力损失
%
对冷却孔物理 2535 4.8%
建模
发散冷却孔壁 2536 5.2%
面模型
发散冷却孔壁面模型
 Perforated Wall Example发散冷却孔壁面模型案例
对冷却孔物理建模及使用发散冷却孔壁面模型火焰筒壁面温度的对比

火焰筒内壁面

对冷却孔物理建模 发散冷却孔壁面模型

火焰筒外壁面

对冷却孔物理建模 发散冷却孔壁面模型
使用FGM燃烧模型时编辑 flamelet及PDF grids

• 编辑 flamelet及PDF grids

Flamelet 面板 Pdf 面板
F, 和 C grids 可编辑 所有变量grid 可编辑

58
使用FGM燃烧模型时编辑 flamelet及PDF grids

修改网格点分布

查看默认的网格 输入文档
点分布 格式

59
燃烧模型的其它改进项

• 使用元素组成法来进行化学计量混合物计算
‐ 改善了flamelet生成的鲁棒性
‐ 修复了极端柔和燃烧现在中flamelet生成失败的bug
‐ 如果检测到用现有方法失败，则在后台执行
• 使用CEQ作为平衡PDF table生成的默认求解器
‐ 使用ANSYS自己的产品取代了第三方的CPROPEP

60
 Diesel和Jet-A材料属性拓展
• 在fluent 数据库droplet-particle材料中添加了两种新液体材料
‐ 在更大的温度和压力跨度内有效（如，燃气轮机燃烧）
‐
‐ Jet-A1 (也即: Jet-A, Kerosene, RP-1): jet-a1-fuel
𝑝𝑣𝑎𝑙𝑖𝑑 = 0.01 𝑏𝑎𝑟 − 23.44 𝑏𝑎𝑟, 𝑇𝑣𝑎𝑙𝑖𝑑 = 220𝐾 − 684.26𝐾
‐ Diesel: diesel-1-fuel
𝑝𝑣𝑎𝑙𝑖𝑑 = 0.01𝑏𝑎𝑟 − 18.7 𝑏𝑎𝑟, 𝑇𝑣𝑎𝑙𝑖𝑑 = 260𝐾 − 736 𝐾
Diesel和Jet-A材料属性拓展
• 提供了温度相关的数据：
‐ Liquid Density 液体密度
‐ Thermal Conductivity 热导率
‐ Specific Heat Capacity 比热容
‐ Dynamic Viscosity 动力学粘性
‐ Latent Heat of Evaporation 蒸发潜热
‐ Binary Diffusivity of fuel vapor into Air @ 1 bar
‐ Saturation Pressure 饱和压
‐ Surface Tension Coefficient 表面张力系数

• 新材料的默认设置
• 默认激活:
• “Convection/Diffusion” 控制的质量输运
• “Variable Lewis Number Formulation” 和 “Use the Specific Heat
of the Evaporating Species in Boiling Law”模型
• “Binary Diffusion Reference Pressure” 设置为1bar
Diesel和Jet-A材料属性拓展

• Jet-A1
‐ 数据源于: “M. Rachner, Die Stoffeigenschaften von Kerosin Jet A-1, DLR Mitteilung 98-01, 1998”
and cast into polynomial form that can be used by Fluent.

• Diesel
‐ 添加到材料数据库中的性能数据对十六烷值为45的一类柴油有效，假定平均化学式为
C13H24
‐ 材料数据来源于福特目前使用的模型燃料库(MFL)的数据拟合
‐ 例外情况是偏心因子和二元扩散系数的数据，这些数据来自于Poling, Prausnitz和O’connell
发表的数据, “The Properties of Gases and Liquids”, Fifth Edition.
 带预混层流小火焰的非绝热FGM燃烧模型 2020R1

• 在不能忽略冷壁、参与介质等造成的热损失/热增
益的情况下提高精度
• 并行 FGM Flamelet and PDF 生成

CO Mass Fraction OH Mass Fraction

部分预混合的 Ch4-空气火焰，壁面温度较低会
导致火焰稳定附近的高热损失

FGM结果得到了有限速率方法的验证
非绝热FGM(NFGM)可预测组分与热损失间的敏感性
基于绝热火焰 （AGM） 的 FGM 无法捕获热损失对组 AFGM :基于绝热火焰的FGM
分生成的影响 NFGM :基于非绝热火焰的FGM
FR : 有限速率化学求解
 FGM有限速率鲁棒性和收敛性提升 2020R1

Improved
• 对瞬态FGM 带DPM的有限速率方 FGM
法计算中，默认使用新的鲁棒性
更好的算法

2d 高压自由传播预混火焰

湍流预测V型火焰

Schematic of the case: Flame stabilized on a rod

 FGM有限速率鲁棒性和收敛性提升测试 2020R1

• 湍流预测V型火焰
Default FGM Improved FGM
• 火焰驻在出口处
• 使用RANS 和 LES 与FGM 结合

Schematic of the case: Flame Instantaneous flame surface,

stabilized on a rod Bell et. al. #

改进后的FGM收敛更快，并且没有精度损失
#“Numerical Simulation of a laboratory-scale turbulent V-flame”
J.B. Bell, M.S. Day, I.G. Shepherd, M. R. Johnson, R. K. Cheng, J. F. Grcar, V. E. Beckner, and M. J. Lijew ski
碳烟颗粒尺寸(Soot size)分布预测 2019R3 ANSYS Fluent

• 使用任一碳烟表面反应机理(使用Chemkin格式)
‐ 充分兼容Model Fuel Library详细碳烟动力学(soot-kinetics)机理
• 捕捉碳烟生成过程中的详细PAH过程
• 不在局限于设定 (HACA) 表面生长氧化(surface-growth-oxidation)机理
• 不在局限于设定气相反应机理
‐ 碳烟表面化学反应中气相反应机理必须求解组分
• 更准确更灵活的碳烟生成/分布输入

• 在2019 R3中，该功能仅限于同有限速率化学反应一起计算气相燃烧时使用

67
 例子: 火焰中碳烟的形成验证案例 2019R3 ANSYS Fluent

• Burner-stabilized stagnation flame

• Using MFL 2018 kinetics mechanism vs. in-built HACA with Method of Moments
‐ Reasonable agreement with data, without any tuning of model parameters

30.0

Experimental conditions (ethylene flame, phi=2): 25.0

Soot Volume Fraction × 109
20.0

H 15.0

10.0

Temperature contour: 5.0

-
H = 6mm H = 12mm

Experiment
General Soot Mechanism
In-built HACA (Default) multiplied by 1.E-4
In-built HACA (Tuned model)

68
 对化学反应设置温度阙值 2019R3 ANSYS Fluent

• 对于T<温度阙值的位置不求解化学反应已减少计算量
• 对于燃烧推荐阙值: ~600-700 K

T
Chemistry
Speedup
Solver Time
Reference 636.17 -- H2

𝑇𝑡ℎ = 400𝐾 523.76 17.7%

CO
𝑇𝑡ℎ = 600𝐾 500.52 21.3%

𝑇𝑡ℎ = 800𝐾 478.11 24.8%

OH

69
 2019R1
Chemkin Physical-Space Flamelets in Fluent

• Premixed flamelet-generation in Fluent

assumes an SDR profile.
• Using Physical Space avoids this assumption
‐ Impact on temperature and flame properties will
depend on combustion conditions
• In 2019 R1 Chemkin-based flamelets can be
generated from the Fluent UI
‐ Seamless integration with the Fluent FGM model
‐ Results are the same as premixed flamelets generated
in ANSYS Chemkin-Pro UI using the Flame Speed
Calculator T
v 1-D Freely propagating Flame
as modeled in ANSYS Chemkin-Pro
 2019R1
Example showing potential impact of physical space flamelet

• Comparing Fluent premixed flamelet vs. Chemkin premixed flamelet

‐ Laminar flame to remove impact of turbulence (zero variances in FGM model)
‐ Compare with Finite-rate chemistry (FRC) solution

OH mass fraction in
a premixed laminar
FRC Fluent FRC Chemkin methane flame
Flamelet Flamelet

Comparing Finite-rate Chemistry (FRC)

vs. Premixed FGM with zero variances
 2019R1
Thickened Flame Model (TFM) has been improved

• Improved theoretical basis allows prediction of

light-around
• Results for 5-burner Ignition Case (LES) with
TFM predict the correct sequence

Experimental data from:

Barré et al., C & F 2014

Animation for 2019 R1 TFM:

Temperature at midplane
Burner number 5 4 3 2 1
反应流：自适应格子提高计算精度
• 格子点在需要的地方自动加密
• 不会浪费格子点，却可有效增加计算精度
• AGR可用于小火焰生成和PDF表格生成

例： H 质量分数 vs. 混合分数

AGR 用最少的点
精确描述混合分
数对燃烧的影响
反应流：利用Chemkin-CFD的动态自适应化学算法（DAC）
• 甲烷燃烧模拟
‐ 机理： MFL2016
‐ 组份数： 231
‐ 计算核数：16
‐ 计算网格： 2400

Solver Time
100

90
使用和不使用DAC 的比较
80

70

60 在未损失精度的前提下， Temperature
50 求解时间减少54%
40

30

20

10

0
w/o DAC w/DAC CO
反应流：利用Chemkin-CFD求解器加速
• 在Fluent中，Chemkin-CFD求解器能够对大规模反应提供更快的求解速度
• 在R19中，求解器改善也提高了小规模反应（<50组份种类）和ISAT算法的速度
‐ ISAT：相较18.2提高1.5倍速度*
‐ 表面反应： (气相反应使用ISAT): 2-3倍提速**
‐ 精度无变化
• Dynamic Adaptive Chemistry (DAC) 选项
‐ 可与Dynamic Cell Clustering联合使用加速反应求解

* 基于Sandia Flame D 和 a 3-D oxyfuel flame

** 基于典型的2-D CVD 案例
 19.0
VOF-DPM模型

VOF VOF-DPM 转化工具 DPM Reaction

•捕捉不稳定性和 •转化标准： • 二次雾化 •组分输运模型

大的液体结构 • 等效直径 •传热
• 球形度 •传质 (蒸发)

界面不稳定性 转化 Lagrangian 追踪
• DECEMBER 2019

2021 R1 Chemkin-Pro / Energico /

Reaction Workbench / MFL
Release Update
11 November 2020
2021 R1新功能概述

• Chemkin-Pro集成到Ansys WB中的参数优化功能
‐ 新教程
‐ 支持WB Units的参数功能
‐ 在WB中测试和验证了Remote Solver Manager (RSM)
• 能够通过Chemkin-Pro用户界面设置对置活塞式发动机
‐ 适用于所有IC发动机类型：单区和多区HCCI模型，直喷柴油机模型，火花点火2-区模型
• Energico中，能够使用Sectional Method跟踪碳烟颗粒
• Reaction Workbench中，增加新的液体属性估算选项
• 在MFL中，添加了H2机理的预简化机理选项
• 2021 R1中的Beta功能：
‐ 关键液体工艺的初始反应器模型：塞流式和闭式均相间歇式反应器

78
Ansys Workbench中的Chemkin-Pro有助于设计探索

• 使用基于WB的分析工作流程：
‐ 通过WB使用RSM进行参数研究
‐ 创建响应面
‐ 使用实验设计
‐ 优化目标结果的流程
• 允许将任何Chemkin模型输出作为优化
或分析目标
• 使用任何反应器参数作为输入：
‐ 工艺配或入口组成
‐ 反应器尺寸和停留时间
‐ 流动条件
‐ 模型参数

79
Ansys Workbench集成Chemkin-Pro : 过程优化

• 允许和Design Explorer, Optimization, Response Surface 工作流程共同使用

• 实例应用：优化工艺配方，产量，安全性

Plasma processing example

目标: 最大化 SiO2 沉积速率
输入: Plasma 功率, 进口流率 最大化沉积的最佳值

Power Flow rate Dep. rate

Inlet Flow rate

权衡分析

样品分析
Plasma power

80
Ansys WB中Chemkin-Pro : 新教程
• 案例：优化发动机的功率 vs. 排放
‐ 使用Chemkin-Pro中的HCCI IC engine例子
‐ 使用Ansys Design Explorer优化进口温度
• Response surface optimization, 两个目标: 最大平均指示缸压(IMEP)和最小一氧化碳(CO) 排放

初始温度 最大功率(IMEP) 和最小排放(CO)

HCCI engine

81
扩展了Energico中的颗粒追踪方法
• 设置ERN时，追踪颗粒尺寸分布的模型可以使用 Moment Method 或 Sectional
Method
‐ 需要在 chemistry set 中包含dispersed-phase surface chemistry

82
 立即注册2021 Ansys Simulation World!
2021 Simulation World是一场免费的在线活动，旨在激励和帮助企业管
理人员、工程师、研发和制造专业人士了解工程仿真变革的力量。 www.simulationworld.com
 让我们听到您对本场会议提出的宝贵意见！
提交问卷即有机会获取限时限量派发的精美礼品，千万别错过哦！

ANSYS-China ANSYS中国

欢迎扫码关注我们！
新的活塞运动选项：支持对置活塞发动机

• 新选项:

85
案例: 对置活塞柴油机设置

• 在the Engine Parameters 面板中增

加了对置活塞设置参数

曲柄连杆尺寸

86
在Reaction Workbench中新增液体属性评估

• 例子:乙醇属性

新属性选项

87
在MFL中新的内置简化机理

• 用于氢气燃烧和热解
‐ 适用于整个P，T和当量比的范围，已验证了
完整的气态燃料MFL机理
‐ 包括氮氧化物和一氧化碳排放

88
液体反应器模型(Beta)

• 常压 closed homogeneous liquid reactor

model
◦ 固定温度或求解能量方法
• Liquid plug-flow reactor model Liquid Closed
Homogeneous
◦ 固定温度或求解能量方法
勾选打开
Beta 功能
(Preferences ) Liquid
Plug
Flow

89