ignition and combustion of diesel sprays · Influence of multicomponent fuel composition on ignition and combustion of diesel sprays ... Liquid and vapor fuel, non-condensable gas
Post on 19-Aug-2020
1 Views
Preview:
Transcript
Influence of multicomponent fuel composition on
ignition and combustion of diesel sprays4th Two-day Meeting on Internal Combustion Engine Simulations Using OpenFOAM® Technology
Martin BlumePhilip SchwarzRomuald Skoda
Sandro Gierth Magnus KircherFederica Ferraro Martin PollackChristian Hasse
Lukas Weiß Michael Wensing
Alessandro StagniTiziano Faravelli
Techn. University Darmstadt
Ruhr University Bochum
FAU Erlangen-Nürnberg
Politecnico di Milano
2
Motivation
3D mixture formationand combustion
3D spray simulation3D gas
exchange
Fuel sideGas side
1D gas exchange
3D nozzle flowsimulation
1D hydraulicssimulation
System simulation(MBS, electr., magnet.)Image source: Bosch
The entire diesel process Zoom into diesel spray cause effect chain
Multicomponent fuel is the standard in diesel engine operation
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
3
Motivation and Outline
Diesel fuel:
Mixture of hundreds of
hydrocarbons
Exact composition neither known
nor standardized
Single species’ interactions
rather complex to investigate in
such multicomponent mixtures
Aim:Investigation of multicomponent
mixture influence along diesel
engine cause and effect chain
Based on simplified surrogate fuel:
10 mass-% n-dodecane/
90 mass-% n-heptane
Based on application relevant
configuration (spray chamber,
near to application heavy-duty
injector)
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
4
Motivation and Outline
Motivation
Experimental setup
Modeling Approach
Results and Discussion
Summary andOutlook
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
5
Motivation and Outline
Motivation
Experimental setup
Modeling Approach
Results and Discussion
Summary andOutlook
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
6
Experimental setup and techniques
Experimental setup:
High pressure combustion chamber
Heavy-duty 9-hole injector, 4 holes closed
Fuels
n-dodecane
n-dodecane / n-heptane mixture (10 / 90 mass-%)
Chamber temperature: 600 °C
Chamber pressure: 50 bar
Rail pressure: 1000 bar
Fuel temperature: 90°C
Experimental techniques:
Inert environment (N2):
!LIF (primary break-up)
Mie (liquid penetration)
Schlieren (vapor penetration)
Reactive environment (air):
OH* luminosity
(Visual flame signal)
Spray
closed
closed
closed
closed
7
Motivation and Outline
Motivation
Experimental setup
Modeling Approach
Results and Discussion
Summary andOutlook
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
8
Numerical approach for spray modeling
Particle in cell method (PIC) / Euler – Lagrange Approach
Statistical descriptionFollows evolution of parcels
Each parcel represents collection of identical droplets
CFD-Cellà gas phase
Parcelà liquid phase
model
~up~ug
Source of Schlieren image: [1]
[1] L. Weiss, A. Peter, M. Wensing, „Diesel Spray characterization with Schlieren-Mie Technique“, Lisbon Symposia, Lisbon 2016
Suitable approach for parcel (spray) initialization needed.
stoichiometric mixture fraction
f (d, u, r, ✓)
Exemplary simulation result
9
Inner nozzle flow and primary break-up
3-Phase solver has been developed [1] and implemented in FoamExtend branchLiquid and vapor fuel, non-condensable gas360° model of injector with directly coupled spray domainfor one selected nozzle hole
Agreement between simulation and experimental data in range of the
cyclic fluctuations
12 % 15 % 19 % 100 %.
Exp. SI
!LIF data from highpressure injection chamber
" = 600 °C,# = 50 bar$-Dodecane
Exp. EA
Simulation
Q-Stat.
[1]: Blume et al., submitted to Atomization&Sprays[2]: L. Weiß, M. Wensing, FAU Erlangen-Nürnberg, 2018
Opening phaseExperimental !LIF data [2]Single shot (SI)Ensemble average (EA)Sim.: Iso-Surf. of liquid volume fraction
10
In-nozzle flow / Spray interface
f(d, u, r, ✓) ⇡ f(d) · f(u|d) · f(r|d) · f(✓|d, r)<latexit sha1_base64="TSNiega9o0pTI1rQIsMIzbvqmIw=">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</latexit><latexit sha1_base64="TSNiega9o0pTI1rQIsMIzbvqmIw=">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</latexit><latexit sha1_base64="TSNiega9o0pTI1rQIsMIzbvqmIw=">AAADHnichVFNSxxBEH1OolETdWOOuTQuAReWZUaEeBQSJZeAQVcFV6RntncddnZm6OmRGJP/4t/wD3gTr4kXr/E/ePB1OwY/CPbQU1Wvql7XR5gncWF8/8+I9+Ll6Nir8YnJ12+mpmdqb2c3i6zUkWpHWZLp7VAWKolT1TaxSdR2rpUchonaCgefrH/rQOkiztINc5ir3aHsp3EvjqQhtFdb7813m6JsCt0UHbOvjGx0ZJ7r7LugpyE6UTczVEvxU9wz9UPzNtNiJGrs1ep+y3dHPFWCSqmjOmtZ7QIddJEhQokhFFIY6gkkCn47COAjJ7aLI2KaWuz8Cr8wydySUYoRkuiA/z6tnQpNaVvOwmVHfCXh1cwU+MC76hhDRttXFfWC8pr3h8P6/33hyDHbCg8pQzJOOMavxA32GfFc5rCKvKvl+UzblUEPS66bmPXlDrF9Rv94PtOjiQ2cR2DFRfbJETr7gBNIKduswE75jkG4jruU0knlWNKKUZJPU9rpsx6uOXi81KfK5kIr8FvBt8X68lK18HG8xxzmudWPWMYXrLGOCCe4xF9cecfeqXfmnd+GeiNVzjs8ON7vGy6zq+M=</latexit><latexit sha1_base64="TSNiega9o0pTI1rQIsMIzbvqmIw=">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</latexit>
6 mm
Joint PDF
Lagrange spray initialization by PDF of ligament location, velocity and diameter
Euler VoFSimulation
Coupled Euler-Lagrange Simulation
Spray asymmetry due to injector characteristics captured by simulation interface (joint PDF)
11
Evaporation model
) dTd
dt=Qs � mFL
mdcp,l
) Ql = Qs � mFL
Evaporation rate with
convective fluxes:
Droplet temperature
d
dr
✓r2⇢guY v
i � ⇢gDgi r
2 dYvi
dr
◆= 0
Nu = 2 + 0,6Re1/2Pr1/3
Particle in cell method (PIC) / Euler – Lagrange Approach
Liquid phase: Governing equation for each parcel
in Lagrangian manner
Gas film transport equation (!... fuel and environment species):
mf,i = 2⇡rd�g
cp,gNuln (1 +Bm,i)
Bm,i =Y sf,i � Y 1
f,i
1� Y sf,i
Raoult's law for vapor-liquid
equilibrium at interface:
Xsf,i = X l
f,ipvap,ip
n-heptane evaporating
faster than n-dodecane
12
Combustion model – Flamelet concept
[1] Popp, STFS, 2015[2] N. Peters. Symp. (Int.) Combust. 21 (1988)
[1]
Theoretical basis of the flamelet concept:§ combustion chemistry is fast§ thin flame sheet assumption§ gradient alignment at flame sheet§ important physics along flame-normal direction
fuel
oxidizer
non-premixed flamelet: Definition:
fuel
oxidizer
non-premixed flamelet:
flameletfuel
oxidizer
non-premixed flamelet:<latexit sha1_base64="6KiGWX1lNMmnvN9D5nhwZFj7Cw8=">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</latexit>
Turbulent flames ≈ Ensembles of 1D flamelets
<latexit sha1_base64="cKWMFAAeXB1XleR7UaZd+9Rsw+s=">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</latexit>
[2]
13
!(#) obtained from canonical setups, e.g. counterflow diffusion flame
As diffusion coefficient, ! acts as inverse residence time for fluid packages in mixture fraction space à determines ignition delay time
Combustion model – Flamelet concept spray
<latexit sha1_base64="cKWMFAAeXB1XleR7UaZd+9Rsw+s=">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</latexit>
[1]
�(Z) = �stexp(2[erfc�1
(2Zst)]2 � 2[erfc
�1(2Z)]
2)
Zst
constant for each flamelet simulation
�st
Zst
!T
@T
@t
⇠ �decreasing residence tim
e
increasing ignition delay time
⌧hom
no ignition
⇢@Yi
@t� 1
2⇢�
@2Yi
@Z2= !i
⇢@T
@t� 1
2⇢�
@2T
@Z2= !T
�st > �st,q
fuel
stagnationplane
x
r
flame
oxidizer
[1 N. Peters. Symp. (Int.) Combust. 21 (1988)
14
In diesel sprays, !"# is varying after start of injection and not known a-priori (function of nozzle and injection parameters)
Tabulation of flame structures based on constant !"# simulations chosen here to capture main effects of the combined ignition and mixing processes with manageable effort
Combustion model – Flamelet concept spray
Source: [1]
Progress of ignition à Progress variable $%
t = 2ms
t
Impact of scalar dissipation rate !"#
�st,1
�st,2
�st,1 > �st,2
�st = 20 1/s
[1]: Pickett, L. M., Genzale, C. L., Manin, J., Malbec, L. M. & Hermant, L. 2011 Measurement uncertainty of liquid penetration in evaporating Diesel sprays. In Proceedings of the23rd Annual Conference on Liquid Atomization and Spray Systems
<latexit sha1_base64="cKWMFAAeXB1XleR7UaZd+9Rsw+s=">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</latexit>
[1]
15
Combustion model – Flamelet concept bi-component spray
Parametrization of fuel composition for 2 components:
In more appropriate formulation:
fuel
stagnationplane
x
r
flame
oxidizer
ZC12H26 =mg,ox,Cl2H26
mg +mg,Cl2H26 +mg,C7H16= Z1
ZC7H16 =mg,C7H16
mg +mg,Cl2H26 +mg,C7H16= Z2
y =Z2
Z1 + Z2
Z = Z1 + Z2 Location in Z-space
Composition of fuel
Ignition delay time increasing with n-heptane content
y =Z2
Z1 + Z2
0.0 0.2 0.4 0.6 0.8 1.0
t ignin
ms
0.4
0.5
0.6
0.7
Ignition delay obtained by flamelet simulation
16
Combustion model – Resulting tabulation approach
1D Flamelet Database CFDtabulation
+ 2D !-pdf integration coupling
<latexit sha1_base64="igTbEPMzZEAid/aPpUwoar6DQtQ=">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</latexit>
Full Thermochemical State Reduced-order Model Solve control variables instead of
<latexit sha1_base64="cKWMFAAeXB1XleR7UaZd+9Rsw+s=">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</latexit>
[1] N. Peters. Symp. (Int.) Combust. 21 (1988)
[1]
fuel
stagnationplane
x
r
flame
oxidizer
e
Z = Z1 + Z2
y =Z2
Z1 + Z2Fuel composition
Thermo-chemical state
eZ1, eZ2, eYC
�st = 01/sgZ 002i = 0
eZ
e!YCey
e = e ⇣ey, eZ, gZ 002
sum,n, e�st, eY norm
C
⌘
17
Combustion model – Chemical mechanism
Tabulated chemistry enables incorporation of complex chemical mechanism in 3D-CFDBut: Amount of flamelet simulations makes utilization of reduced mechanism advantageousReduction approach:
Optimization target: Homogeneous ignition delay timeA-posteriori evaluation based on laminar flame speed and diffusion flames relevant for current spray flame tabulation
POLIMI C0-C16 mechanism (Low/High Temperature)
~ 500 species ~ 17000 reactions
POLIMI TPRF mechanism (Low/High Temperature)
~ 335 species ~ 9315 reactions
C12-C7 reduced mechanism (Low/High Temperature)
~ 127 species~ 2205 reactions
Source: [4]
Preselection based on heaviest fuel components
Reduction procedure utilizingDRGEP + SA [1,2,3]
[1]: Pepiot-Desjardins, P., & Pitsch, H. (2008). An efficient error-propagation-based reduction method for large chemical kinetic mechanisms. Combustion and Flame, 154(1-2), 67-81., [2]: Niemeyer, K. E., Sung, C. J., & Raju, M. P. (2010). Skeletal mechanism generation for surrogate fuels using directed relation graph with error propagation and sensitivity analysis. Combustion and flame, 157(9), 1760-1770.[3]: Stagni, A., Frassoldati, A., Cuoci, A., Faravelli, T., & Ranzi, E. (2016). Skeletal mechanism reduction through species-targeted sensitivity analysis. Combustion and Flame, 163, 382-393.[4]: http://creckmodeling.chem.polimi.it/menu-kinetics
Reduction(DRGEP + SA)
18
Motivation and Outline
Motivation
Experimental setup
Modeling Approach
Results and Discussion
Summary andOutlook
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
19
Investigation mixing field in non-reactive environment
Comparison of evaporation behavior and mixture formation based on 1% fuel mass fraction:
Differences in start of fuel vapor formation with similar vapor penetration length
n-Dodecane Mixture
Initial fuel vapor earlier for mixture than for pure n-dodecaneVapor penetrations similar for both fuels
20
Validation of vapor and liquid penetration lengths
Comparison vapor and liquid penetration in non-reacting environment
Exp. Mie probability, Mixture, tASOI = 1.0 ms
Vapor penetration slightly overpredicted,no fuel influence in simulation
Steady liquid penetration in agreement with exp. data, fuel influence captured
Sim. Z=0.001, Mixture, tASOI = 0.7 ms Exp. Schlieren probability,
Mixture, tASOI = 0.7 ms
Sim. liquid mass < 99 %, Mixture, tASOI = 1.0 ms
21
Investigation gaseous fuel composition in non-reactive environment
Comparison of liquid and gas phase composition:
High fraction of n-heptane in liquid and vaporized fuel, decreasing with distance to injector
Downstream reduction of n-heptane fraction in
Liquid phaseGas phase
22
Comparison ignition and combustion process
Simulation results for ignition and combustion behavior
Qualitative experimental findings concerning ignition process reproduced,
n-heptane content in gas phase increases ignition delay time
n-Dodecane Mixture
Combustion starts at spray flank
and proceeds towards spray tip
Ignition delay time smaller for
pure n-dodecane than for mixture
Exp. line of side integrated OH*
n-Dodecane, tASOI = 0.70 ms
Exp. line of side integrated OH*
Mixture, tASOI = 1.05 ms
23
Comparison of ignition delay times based on cumulated OH* signal (exp.) / OH mass fraction (sim.)
Comparison integrated OH* and OH signal
Differences in ignition delay time captured by tabulation strategy
Ignition delay time slightly underpredicted by simulation for both, pure n-dodecane and mixtureBut: Difference in ignition delay time of n-dodecane and mixture well reproduced
Δ"#$%&'(
Δ"#$%)#*
24
Motivation and Outline
Motivation
Experimental setup
Modeling Approach
Results and Discussion
Summary andOutlook
Source: [1]
[1] https://images.app.goo.gl/NSaXBLGtgLtTLs828
25
Summary and Outlook
Aim:Investigation of multicomponent mixture influence along diesel engine cause and effect chainBased on simplified surrogate mixture: 10 mass-% n-dodecane/ 90 mass-% n-heptaneBased on application relevant configuration (spray chamber, near to application heavy-duty injector)
Findings:Liquid penetration for mixture shorter than for pure n-dodecaneInitial fuel vapor first formed by mixture Simulated vapor penetration unaffected by fuel composition due to same injection pressure / similar momentum fluxIgnition delay time for surrogate larger than for n-dodecane due to n-heptane content in gas phase
Outlook:LES of setup to investigate influence of gas phase mixing model
y =Z2
Z1 + Z2
0.0 0.2 0.4 0.6 0.8 1.0
t ignin
ms
0.4
0.5
0.6
0.7
26
The joint research project LowEmissionDesign is
funded by the German Ministry of Energy and
Economics (BMWi) in the framework of the
Verkehrsforschungsprogramm (“Gefördert vom
Bundesministerium für Wirtschaft und Energie
aufgrund eines Beschlusses des Deutschen
Bundestages.”). We would like to express our
gratitude to the project partners AVL Deutschland
GmbH, Robert Bosch GmbH and MAN Truck & Bus
SE
Calculations for this research were conducted on the
Lichtenberg high performance computer of the TU
Darmstadt within the computing project 973
and
with computing resources granted by RWTH Aachen
University within the computing project bund0002.
Acknowledge
top related