throttle angle [deg] n ear-TD. C elevation. [d eg] tangential port throttling helical port throttling closed. 20. 40. 60. 80 WOT. 0. 90. 180. 270. 360 throttle angle [deg].
Study of In-cylinder Swirl Flow Structures using Principal Component Analysis
6 4 2 0
-90
-45
0
45
crank angle [degrees ATDC]
90
8 6 4 2
x 10
0
45
crank angle [degrees ATDC]
90
0.5
2 0
0
-90
-45
0
45
crank angle [degrees ATDC]
90
12 1.0 10 0.8 8 0.6 6 0.4 4 0.2 2
00
-90 -90
-45
0
45
crank angle [degrees ATDC]
• Full 360 degrees 4 unstructured engine mesh2 including intake and 0exhaust runners and45 -90 -45 0 x 10 of crank angle [degrees ATDC] plenums the Sandia 1.2 1.9L1.0light duty single cylinder engine facility
Grid type Turbulence Chemistry solver
Unstructured, staggered 2-equation GRNG k-epsilon model, Wang et al. [38] SpeedCHEM, Perini et al. [39,40]
Atomization
KH-RT, Beale and Reitz [41]
SGS near-nozzle flow field
Gas-jet, Abani and Reitz [42]
In-cylinder flow modelingLaw of the wall
Launder and Spalding [43]
• Comparison with PIV in-cylinder experiments [2]
360 4 0
IVO IVO EVC EVC
global squish bowl
-360 3 -720
-2160
global squish bowl
High Rs’s have similar structure w/ different bore-scale vortex strength
-90 x 104
-45
0
45
crank angle [degrees ATDC]
90
00
100 100
200 200
300 300
1.0
full engine geometry
0.8 0.6 0.4
global squish bowl
3
IVO EVC
IVC IVC
global global squish squish bowl bowl
80 3 60
2
2
40
1
1 20
0 -400 -300 -200
-100
0
100
crank angle [deg aTDC]
elevation,
200
300
0 -400 -300 -200 -100
0
100
crank angle [deg aTDC]
-90
-45
0
45
crank angle [degrees ATDC]
Intake region Composition: arbitrary fresh air + measured EGR comp p, T: from intake transducers
Injection Actual timing, duration, injected mass and fuel composition. Modeled injector body protrusion
90
90
Momentum from T dominates flow structure Recirculating regions at Rs=1.5 due to competition with H
75 60 45 30 15 20
40
60
throttle angle [deg]
80 WOT
Affects: Tangential port Helical port
3.5 3 2.5 2 helical port throttling tangential port throttling
1.5 1 closed
20
40
60
80 WOT
throttle angle [deg]
tangential port throttling helical port throttling
270 EXHAUST
180 INTAKE
90
0 closed
20
Principal Components
Global Rs
near-TDC turbulence
40
-45
0
45
crank angle [degrees ATDC]
90
80 WOT
• Both ports do not change bowl vs. squish discrepancies significantly role of piston bowl geometry and squish flow
References [1] Perini, Miles, Reitz, Computers&Fluids 105, 113-124, 2014. [2] Perini, Zha, Busch, Miles, Reitz, SAE technical paper 2015-01-1969, 2015.
-90
60
throttle angle [deg]
0.2 0
300
360
4 tangential port throttling helical port throttling
90
0.4
200
azimuth,
eccentricity, e
sector simulation
2 3
100 4
IVC
Port throttling effects
0.2
0 closed
0.6
IVO EVC
1)Compression accumulates flow energy into 'noble' storage rotational kinetic energy (high Rs) + potential energy (deviation from axial mom. cons. position is a way to store energy by applying some force and enabling an action-reaction effect through gradients, such as what happens in a capacitor, or in a spring). 2)This storage is quicky released back into turbulence as the forcing element (compression) cannot compete anymore against dissipation.
sector simulation
Cylinder Composition: measured, exhaust p, T: from pressure trace
Rs
- Swirl vortex starts forming > -300 aTDC (squish); >-250 aTDC (bowl) - Tilt is negligible near BDC, accumulates close to TDC - Fast dissipation of non-symmetric structure after TDC
4
turbulent dissipation [m /s ]
4
IVC IVC
-1440
phi [-]
Exhaust region Composition: measured, exhaust p, T: measured
0.8
T
0
2
full engine geometry
Swirl structure evolution swirl structure, Rs=2.2
crank angle [deg aTDC]
0
e
(seen from below)
0 -2520 -400 -400 -300 -300 -200 -200 -100 -100
4
H
elevation angle 0, 2 azimuthal angle , e center of mass eccentrici ty[cm] Rs swirl ratio [-]
2 -1080
0.5
6
H
ALE (KIVA, Torres et al. [35])
1 -1800
8 6
Solver
Reynolds-Averaged Navier-Stokes
8
90
Full Engine Geometry
Tangential
Equations
1
sector simulation full engine geometry
10
1.5
sector simulation full engine geometry
10
1.2
sectorsimulation simulation sector full engine full enginegeometry geometry
12
turbulent length scale [mm]
-45
T
6
0
turbulent dissipation [m2/s3]
dissipation turbulent [m/s/s] ] energy[m kinetic turbulent
2 23 2
1.2 14
-90 4
Helical
1
4
0.7
Model details
H
Turbulence development is not properly captured 0
Average cell size near TDC [mm]
1.5
8
12
sector simulation full engine geometry
10
Rs = 4.5
682,091
near-TDC azimuth [deg]
8
Number of cells
sector simulation full engine geometry
12 Rs = 3.5 10
body-fitted hexahedral
near-TDC eccentricity[mm]
10
12
experiment turbulent length scale [mm]
Rs = 2.2
• Adjustable swirl generation plates in the intake runner modeled using cell deactivation
T
o Dimension reduction of the swirl vortex structure to better understand its implications on local mixture preparation o 1st principal component = main alignment direction of the swirl center envelope The axis the vortex would have if solid-body Described thru:
near-TDC elevation [deg]
Rs = 1.5
turbulent kinetic energy [m2/s2]
12
sector simulation full engine geometry
turbulent length scale [mm]
simulation turbulent kinetic energy [m2/s2]
• Sector mesh simulation can’t capture this phenomenon well, even though ignition timing and duration are well captured, thanks to accuate spray modelng at the jet centerline 14 The sector geometry intrinsically simplifies 14 the flow structure
Grid details Mesh type
Principal Component Analysis of swirl vortex
elevation [deg] eccentricity [cm]
• CO and UHC emissions in light-load partially premixed combustion strategies for light duty engines mainly arise from overly lean mixture forming at the center of the combustion chamber and in the squish region
• Detailed compositional and
Table 2. Simulation parameters used for the present study. thermophysical initial/time-varying boundary conditions
eccentricity [cm]
Motivation
Funding Sponsor – Sandia National Laboratories
azimuth [deg] eccentricity [cm]
Federico Perini, Rolf D. Reitz, Paul C. Miles
UNIVERSITY OF WISCONSIN - ENGINE RESEARCH CENTER
As of March 25, 2015
