OpenFoam project at Politecnico di Milano G. D’Errico Dipartimento di Energetica Politecnico di Milano
OpenFoam project at Politecnico di Milano
G. D’Errico
Dipartimento di EnergeticaPolitecnico di Milano
Some information:
•Politecnico di Milano is a public university.
•About 40.000 students in 2003/2004 (25.000 studentsin the Engineering Faculties, 4.000 students in the Design Faculty, 11.000 students in the ArchitectureFaculty).
•300 new PhD students per year.
•In 2004 Politecnico has 1.133 academics (396 full professors, 390 associate professors and 347 lecturers) and 870 administrative and technicalemployes.
•Graduated in engineering at Politecnico representabout 12,7% of all the graduated in engineering of Italy.
The Department of Energetics:(Dipartimento di Energetica)
Research and teaching in the following fields:
• Internal combustion engines• Powerplants and energy systems• Turbomachinery• Aerospace Propulsion• Thermodynamics• Acoustics
Internal combustion engine group (www.engines.polimi.it)Staff: Prof. Giancarlo FerrariProf. Angelo OnoratiDr. Gianluca D’Errico
Post-doc researchers: Gianluca Montenegro, Tarcisio Cerri, Federico Piscaglia
PhD students: Tommaso Lucchini, Marco Miglioli
MSc students (≈15-20 thesis per year)
Calculation of exhaust gas composition by means of multi-zone combustion models.
1D and 3D thermo-fluid dynamic modelingof S.I. engines (gasoline, natural gas, hydrogen)and C.I. engines (Diesel, HCCI) GASDYN, OpenFoam
Simulation of wave motion and chemical species transport, with reactions in the gas and solid phase in the exhaust ducts of a s.i. engine.Prediction of catalyst conversion efficiency and light-
off time during a warm-up.
Thermo-fluid dynamic modeling of i.c. engines
Silencers
Exhaust after-treatment system (3way CC, deNOx SCR, Diesel oxi-cat, DPF, etc.)
Air injection system
Intake system
Combustion
1D Thermo-fluid dynamic modeling of i.c. engines
Lamborghini 6.2 L V12 S.I. engine
Thermo-fluid dynamic simulation of the Lamborghini 6.2 L V12 S.I. engine.
Modeling of the intake system for the prediction of the volumetric efficiency, performances and emissions.
Companies, research centres and universities currently involved in the development and application
of the 1D GASDYN model:
• LAMBORGHINI• SEATEK• ELASIS• BMW• MVAGUSTA-CAGIVA• FAURECIA• LAFRANCONI SILENCERS• MIVV EXHAUST SYSTEMS• ART ENGENEERING• ISTITUTO MOTORI (CNR)• UNIVERSITY OF MODENA• POLITECNICO DI TORINO• UNIVERSITY OF ROME (Tor Vergata)• UNIVERSITY OF PERUGIA• UNIVERSITY OF PARMA• UNIVERSITY OF BOLOGNA
Companies, research centres and universities currently involved in the development and application
of the 1D GASDYN model:
• LAMBORGHINI• SEATEK• ELASIS• BMW• MVAGUSTA-CAGIVA• FAURECIA• LAFRANCONI SILENCERS• MIVV EXHAUST SYSTEMS• ART ENGENEERING• ISTITUTO MOTORI (CNR)• UNIVERSITY OF MODENA• POLITECNICO DI TORINO• UNIVERSITY OF ROME (Tor Vergata)• UNIVERSITY OF PERUGIA• UNIVERSITY OF PARMA• UNIVERSITY OF BOLOGNA
OpenFoam - GASDYN
MultiDMultiD Research topicsResearch topics
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
DEVELOPMENT: ENGINE MESHDEVELOPMENT: ENGINE MESH
FOUR STROKE ENGINES: FOUR STROKE ENGINES: Mesh motion MOVIE. Section of a 3D engine withMesh motion MOVIE. Section of a 3D engine withvertical valvesvertical valves
Layer addition/removalLayer addition/removalon the on the pistonpiston patchpatch
Layer Layer addition/removaladdition/removalon the on the valveTopvalveTop andandvalveBottomvalveBottom patchespatches
slidingInterfaceslidingInterfacecontinuoslycontinuoslyattached andattached anddetacheddetached
Switching Switching beetweenbeetweendeformation deformation and layeringand layering
MultiDMultiD Research topicsResearch topics
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
ChemicalChemical modelsmodels
Development of a reduced kinetic scheme for the evaluation of the main chemical species (particularly NO and CO) in premixed turbulent flame.
1D laminar flame
validationvalidation
QuasiQuasi--D and CFD D and CFD combustion in combustion in CNG enginesCNG engines
CFD CFD analysisanalysis of of methane and methane and
air steady air steady turbulent flameturbulent flame
SAE 2005-01-1123
ReducedReduced kinetickinetic schemescheme
To remove O-O2 equilibrium assumption
To characterize the kinetically controlled species
To assume partial equilibrium for the species whose reactions are much faster
To achieve a computational time efficiency
Paper # 2005-01-1123
Closure of the chemical systemClosure of the chemical system
The proposed chemical scheme considers:4 elements: O, N, C, H 11 chemical species: N2, H2O, CO2, CO, O2, H2, H,
O, OH, NO, HO2. NO, CO and H are computed kinetically, O, OH, H2 are in partial equilibriumHO2 is derived by the steady-state assumption.
The remaining species, N2, H2O, O2 and CO2 can be derived applying the element conservation
Paper # 2005-01-1123
CFD Combustion ModelCFD Combustion Model
The proposed chemical scheme was first embedded into the multidimensional code OpenFOAM
Weller combustion model
Solve the transport equations for mass, momentum, energy chemical species and turbulence properties, an equation representing the average position of the interface is solved
Paper # 2005-01-1123
CFD Combustion ModelCFD Combustion Model
The evolution of the combustion process is defined by the normalized fuel mass fraction:
( ) ( ) |~|~~~~bSbDbU
tb
uuu ∇Ξ−=∇∇−∇+∂∂ ρρρρ
The wrinkle factor can be derived by solving the following transport equation:
( ) ( ) Ξ⋅−+Ξ−−Ξ=Ξ∇+∂Ξ∂ )0,max(1 stRGUt
σσ)
Paper # 2005-01-1123
1D validation for Laminar Flow1D validation for Laminar Flow
Detailed chemical code, FlameMaster
A lean mixture of air and methane (Φ=0.8) is considered at atmospheric pressure with an initial unburned gas temperature equal to 400 K.
The chemical composition is evaluated in the post-flame zone and the proposed scheme is compared with the full kinetic scheme and with the standard approach
Paper # 2005-01-1123
1D validation for Laminar Flow1D validation for Laminar Flow
NO concentration profiles in a 1D laminar
0.00
0.50
1.00
1.50
2.00
2.50
3.00
0 0.002 0.004 0.006 0.008 0.01distance [mm]
mol
ar c
once
ntra
tion/
10-5
NO complex kinetics
NO proposed model
NO Zeld. + equilibrium
Paper # 2005-01-1123
1D validation for Laminar Flow1D validation for Laminar Flow
O concentration profiles in a 1D laminar
0.000
0.001
0.001
0.002
0.002
0.003
0.003
0.004
0 0.002 0.004 0.006 0.008 0.01distance [mm]
mol
ar c
once
ntra
tion
O complex kinetics
O proposed model
O equilibrium
Paper # 2005-01-1123
CFD validation for Turbulent FlowCFD validation for Turbulent Flow
Turbulent confined flame in a high velocity premixed flow of methane and air, ignited by a parallel flow of hot
Air is preheated up to 600 K, the inlet velocity is approximately 55 m/s and the equivalence ratio Φ=0.8. Coarse mesh of 9100 cells.
Paper # 2005-01-1123
CFD validation: velocity profilesCFD validation: velocity profiles
Ux ( x = 0.039 m )
0
0.02
0.04
0.06
0.08
0.1
0 30 60 90 120 150Axial velocity [m/s]
Wal
l dis
tanc
e [m
] EXPERIMENTALCALCULATED
Paper # 2005-01-1123
CFD validation: velocity profilesCFD validation: velocity profiles
Ux ( x = 0.250 m )
0
0.02
0.04
0.06
0.08
0.1
0 30 60 90 120 150 180Axial velocity [m/s]
Wal
l dis
tanc
e [m
] EXPERIMENTALCALCULATED
Paper # 2005-01-1123
CFD validation: velocity profilesCFD validation: velocity profiles
Ux ( x = 0.450 m )
0.00
0.02
0.04
0.06
0.08
0.10
0 30 60 90 120 150 180 210Axial velocity [m/s]
Wal
l dis
tanc
e [m
] EXPERIMENTALCALCULATED
Paper # 2005-01-1123
CFD validation: velocity profilesCFD validation: velocity profiles
Ux ( x = 0.650 m )
0
0.02
0.04
0.06
0.08
0.1
0 30 60 90 120 150 180 210 240Axial velocity [m/s]
Wal
l dis
tanc
e [m
]
EXPERIMENTALCALCULATED
Paper # 2005-01-1123
CFD validation: NO profilesCFD validation: NO profiles
NO ( x = 0.222 m )
0
0.02
0.04
0.06
0.08
0.1
0.E+00 1.E-05 2.E-05 3.E-05 4.E-05 5.E-05 6.E-05Molar fraction
Wal
l dis
tanc
e [m
]
Experimental
Proposed mechanism
Equilibrium + Zeldovich
Paper # 2005-01-1123
CFD validation: NO profilesCFD validation: NO profiles
NO ( x = 0.322 m )
0
0.02
0.04
0.06
0.08
0.1
0.E+00 1.E-05 2.E-05 3.E-05 4.E-05 5.E-05 6.E-05Molar fraction
Wal
l dis
tanc
e [m
]
Experimental
Proposem mechanism
Equilibrium + Zeldovich
Paper # 2005-01-1123
CFD validation: NO profilesCFD validation: NO profiles
NO ( x = 0.522 m )
0
0.02
0.04
0.06
0.08
0.1
0.E+00 1.E-05 2.E-05 3.E-05 4.E-05 5.E-05 6.E-05 7.E-05Molar fraction
Wal
l dis
tanc
e [m
]
Experimental
Proposed mechanism
Equilibrium + Zeldovich
Paper # 2005-01-1123
CFD validation: NO profilesCFD validation: NO profiles
NO ( x = 0.692 m )
0
0.02
0.04
0.06
0.08
0.1
0.E+00 1.E-05 2.E-05 3.E-05 4.E-05 5.E-05 6.E-05 7.E-05Molar fraction
Wal
l dis
tanc
e [m
]
Experimental
Proposed mechanism
Equilibrium + Zeldovich
Paper # 2005-01-1123
Engine simulationEngine simulation
Engine thermoEngine thermo--fluid dynamic modelfluid dynamic model
It is possible to predict: • engine performances• cylinder emissions
exhaust system warm-upcatalyst efficiencytailpipe emissionstailpipe noise
Integrated simulation (thermodynamic, fluid dynamic and chemical) of the whole engine system, considering pre-catalysts, catalysts and silencers (GASDYN)
Paper # 2005-01-1123
Paper # 2005-01-1123
Engine simulation: Fiat Engine simulation: Fiat LanciaLancia 2.0L VIS2.0L VIS
Fuelled with Gasoline and CNG
Engine simulation: Fiat Engine simulation: Fiat LanciaLancia 2.0L VIS2.0L VIS
Engine simulation: Fiat Engine simulation: Fiat LanciaLancia 2.0L VIS2.0L VIS
Engine simulation: Fiat Engine simulation: Fiat LanciaLancia 2.0L VIS2.0L VIS
Engine simulation: Fiat Engine simulation: Fiat LanciaLancia 2.0L VIS2.0L VIS
NonNon--equilibrium emission modelequilibrium emission model
MultiDMultiD Research topicsResearch topics
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
3D fluid dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
GDI SPRAY MODELLINGGDI SPRAY MODELLING
SPRAYSPRAY
LagrangianLagrangian descriptiondescription
Cell to Cell Tracking according to the mass,Cell to Cell Tracking according to the mass,momentum and energy exchange with the GAS momentum and energy exchange with the GAS ((EulerianEulerian) phase) phase
PARCELSPARCELS
ADDITIONAL MODELSADDITIONAL MODELS
ATOMIZATIONATOMIZATION: initial droplets size and spray angle: initial droplets size and spray angle
BREAKUPBREAKUP: droplets deformation due to the aerodynamic forces: droplets deformation due to the aerodynamic forces
EVAPORATIONEVAPORATION: energy exchange with the gas phase: energy exchange with the gas phase
COLLISIONCOLLISION: droplet/droplet interaction: droplet/droplet interaction
TommasoTommaso LucchiniLucchini Ph.D. thesisPh.D. thesis
GDI SPRAY EVOLUTIONGDI SPRAY EVOLUTION
PRESSUREPRESSURE--SWIRL SWIRL INJECTORINJECTOR
FUEL LIQUID SHEETFUEL LIQUID SHEET
PRIMARY BREAKUPPRIMARY BREAKUP
LIQUID SHEETLIQUID SHEETINSTABILITYINSTABILITY
LIGAMENTSLIGAMENTS
DROPLETSDROPLETS
SECONDARY BREAKUPSECONDARY BREAKUP
EVAPORATIONEVAPORATION
PRESSURE SWIRL INJECTORPRESSURE SWIRL INJECTOR
m&
x
z
αβ
αβ
x
z
INJECTOR DATAINJECTOR DATA
Diameter (Diameter (dd00))
Spray angle (Spray angle (αα))
Mass flow rate ( )Mass flow rate ( )
Injection pressure (Injection pressure (pp))
Initial Droplet Initial Droplet Velocity (Velocity (UU))
lv
pkUρ∆
=2
Initial Spray Initial Spray Angle (Angle (θθ))
βαθ ⋅−+= )1,1(2
rnd
Initial Sheet Initial Sheet Thickness (Thickness (hh))
)( hdUhm l −=πρ&
ATOMIZATION MODELATOMIZATION MODEL
LISA MODEL (LISA MODEL (Linear Instability Analysis of a Viscous Liquid Linear Instability Analysis of a Viscous Liquid SheedSheed))
AERODYNAMIC INSTABILITYAERODYNAMIC INSTABILITY WAVE GROWTHWAVE GROWTH
The growth rate (The growth rate (ωω) and the wave number () and the wave number (kk) are related) are related
fuelfuel
airfuelfuels
kkUkkρσ
ρρννω
322422 42 −++−=
ΩΩss, maximum growth rate is responsible for the , maximum growth rate is responsible for the breakupbreakup
⎟⎟⎠
⎞⎜⎜⎝
⎛Ω
=0
ln1ηητ b
S
BreakupBreakupTimeTime
τVL =
BreakupBreakupLengthLength
LigamentsLigamentsDiameterDiameter
sL K
hd 8= ( ) 6/13188.1 Ohdd LD +=
DropletsDropletsDiameterDiameter
hh, liquid sheet half, liquid sheet half--thickness calculated by mass conservationthickness calculated by mass conservation
SECONDARY BREAKUPSECONDARY BREAKUP
KHRT MODELKHRT MODEL
TWO DIFFERENT BREAKUP MECHANISMSTWO DIFFERENT BREAKUP MECHANISMS
KelvinKelvin--HelmoltzHelmoltz: : child parcels child parcels are stripped off are stripped off from the parent from the parent parcelparcel
RayleighRayleigh--Taylor: Taylor: catastrophic catastrophic breakupbreakup, much , much smaller droplets smaller droplets creationcreation
EVAPORATION MODELEVAPORATION MODEL
( ) gasdropletev pTp >
Droplet diameter reduction after the primary Droplet diameter reduction after the primary breakupbreakupInitial spray cone angle is increasedInitial spray cone angle is increased
FUEL HEATED BY THE INJECTORFUEL HEATED BY THE INJECTOR
DROPLET FLASH EVAPORATION (DROPLET FLASH EVAPORATION (TTdroplet,surfacedroplet,surface > > TTgasgas))
( )( )b
dropletboilingdropletsfd TL
ATTGm
−==α
&
( ) ⎥⎦⎤
⎢⎣
⎡ −⎟⎠⎞
⎜⎝⎛ ++
+= ∞
b
bf
fp TLhh
GG
GGNur
ckG 11ln
/12 0πGGm fd +=&
DROPLET SUPERHEAT EVAPORATION (DROPLET SUPERHEAT EVAPORATION (TTdroplet,surfacedroplet,surface < < TTgasgas))
SPRAY EVOLUTION IN THE SPRAY EVOLUTION IN THE OpenFoamOpenFoam CODECODE
MESHMESH((fvMeshfvMesh))
SPRAYSPRAY((sprayspray))
SPRAY FUNCTIONSSPRAY FUNCTIONS((spraySubModelsspraySubModels()()))
inject(), inject(), atomizationLoopatomizationLoop()()
breakupLoopbreakupLoop(), (), evaporate(), evaporate(),
collide()collide()
………………
VALIDATIONVALIDATION
--INFLUENCE OF INJECTION PRESSUREINFLUENCE OF INJECTION PRESSURE
--DROP SIZE PREDICTIONDROP SIZE PREDICTION
--INFLUENCE OF AMBIENT PRESSUREINFLUENCE OF AMBIENT PRESSURE
--INFLUENCE OF FUEL TEMPERATUREINFLUENCE OF FUEL TEMPERATURE
INFLUENCE OF INJECTION PRESSUREINFLUENCE OF INJECTION PRESSURE
DROP SIZE PREDICTIONDROP SIZE PREDICTION
R
Z
DROP SIZE PREDICTIONDROP SIZE PREDICTION
Z=7.5
R=3
DROP SIZE PREDICTIONDROP SIZE PREDICTION
Z=10
R=2
Z=10
R=3
INFLUENCE OF AMBIENT PRESSUREINFLUENCE OF AMBIENT PRESSURE
AMBIENTAMBIENT PRESSURE 1 barPRESSURE 1 bar
t = 0.5 mst = 0.5 ms t = 1.0 mst = 1.0 ms t = 1.5 mst = 1.5 ms
t = 2.0 mst = 2.0 ms t = 2.5 mst = 2.5 ms t = 3.0 mst = 3.0 ms
INFLUENCE OF AMBIENT PRESSUREINFLUENCE OF AMBIENT PRESSURE
INFLUENCE OF FUEL TEMPERATUREINFLUENCE OF FUEL TEMPERATURE
INFLUENCE OF FUEL TEMPERATUREINFLUENCE OF FUEL TEMPERATURE
T = 375 K (T = 375 K (∆∆TTSHSH = 0 K)= 0 K)
T = 405 K (T = 405 K (∆∆TTSHSH = 30 K)= 30 K)
MultiDMultiD Research topicsResearch topics
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
Intake and Exhaust Fluid Dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
Mesh modifiers with topological change
Combustion modeling with reduced kinetic schemes
Spray modeling (GDI, Diesel, HCCI)
Intake and Exhaust Fluid Dynamics:steady flows of complex devices1D-3D couplingSilencer modeling
Noise analysis
Expansion chamber
Noise analysis
0
2
4
6
8
10
12
14
0 100 200 300 400 500 600 700 800 900 1000
frequenza [Hz]
TL [d
B]
Teoria classica FoamHz 6122
00 ==
La
f
Computed and theoretical transmission loss
Noise analysis
Column resonator
Noise analysis
-5
0
5
10
15
20
25
30
35
0 500 1000 1500 2000 2500 3000
frequenza [Hz]
TL [d
B]
Teoria classica Foam
Computed and theoretical transmission loss
Noise analysis
Helmholtz
resonator
-5
0
5
10
15
20
25
30
35
40
0 500 1000 1500 2000 2500 3000
frequenza [Hz]
TL [d
B]
Teoria classica Foam
Computed and theoretical transmission loss
Noise analysis
Hz 6122
0 0 ==
La
f alelongitudin Hz 9272
0 0 ==
ha
f etrasversal
Expansion chamber with non-axial inlet-outlet
0
5
10
15
20
25
30
35
40
45
0 500 1000 1500 2000 2500 3000
frequenza [Hz]
TL [d
B]
Gasdyn FoamHz 612 0 =alelongitudinf
Hz 927 0 =etrasversalf
Politecnico di Milano – Dipartimento di Energetica
Expansion chamber at 860 Hz
1D-3D coupling
Same solver (HLLC Riemann solver) is usedin the two domains
OpenFoam: 1D-3D couplingOpenFoam: 1D-3D coupling
Future projectsFuture projects
4 stroke engine with lateral valves
Design and optimization of a new Diesel engine
HCCI modeling
Any idea ???
gtÇ~ çÉâyÉÜ çÉâÜ tààxÇà|ÉÇ