Improvement of Intake Restrictor Performance for a Formula SAE Race Car through 1D & Coupled
1D/3D Analysis Methods
Mark Claywell & Donald Horkheimer University of Minnesota
2006-01-36542
Agenda• Background• 1D Simulation Setup & Results• Coupled 1D/3D Simulation Setup & Results
• Volumetric Efficiency • Flow measurements- Mach Number, Turbulent
Kinetic Energy, & Total Pressure
• Acoustic Filtering for Volumetric Efficiency• Flow Control• Conclusions
2006-01-36543
Background• Overall Goal: Reduce the impact of the restrictor on
VE• Achieve a better understanding of flow in different restrictors• Look for areas to reduce flow losses
• Ricardo WAVE 1D model used with acoustic quality mesh in intake manifold• Multiple iterations quickly solved• Poor modeling of flow losses in diffuser section• Tuning impact of restrictor
• Ricardo WAVE coupled with Ricardo VECTIS (CFD) to model full intake• More accurate losses in diffuser• No need for end corrections at runner to plenum junctions• Very long run times = fewer design iterations
2006-01-36544
Description of Intake Manifold Geometry
DiffuserHalf-Angle
TotalHeight
Plenum Height
Diffuser Height
DiffuserVolume
PlenumVolume
Total Volume = Plenum + Diffuser Volume
Diffuser Exit Diameter
2006-01-36545
Case 1: Same Plenum Used
Difficult to get a clear trend of how diffuser angle affects tuning
■ Large changes in tuning across the rev range■ Both total length and total volume are changing
Diffuser Half-Angle 3° 4° 5.5° 7°
Diffuser Exit Diameter (mm) 72.90 72.90 72.90 72.90
Diffuser Length (mm) 504.7 378.2 274.7 215.4
Diffuser Volume (liters) 0.95 0.71 0.52 0.40
Plenum Volume (liters) 2.26 2.26 2.26 2.26
Plenum Length (mm) 160.5 160.5 160.5 160.5
Total Length from Throat to Plenum Bottom (mm)
665.2 538.7 435.2 375.9
Total Volume After Throat (liters)
3.21 2.97 2.78 2.66
2006-01-36546
Standing Waves - Case 1
2nd Order 4th Order 6th Order
4°
7°
11,000 RPM
14,000 RPM
2006-01-36547
Case 2: Equal DiffuserLengths
Diffuser length had large impact on tuningTotal volume had little impact on tuning
■ VE did not increase with increasing total volume
Diffuser Half-Angle 3° 4° 5.5° 7°
Diffuser Exit Diameter (mm) 42.578 50.126 61.485 72.898
Diffuser Length (mm) 215.4 215.4 215.4 215.4
Diffuser Volume (liters) 0.17 0.22 0.31 0.40
Plenum Volume (liters) 2.25 2.25 2.26 2.26
Plenum Length (mm) 160.5 160.5 160.5 160.5
Total Length from Throat to Plenum Bottom (mm)
375.9 375.9 375.9 375.9
Total Volume After Throat (liters)
2.423 2.474 2.564 2.663
2006-01-36548
WAVE-VECTIS Setup
Assumptions•WAVE-VECTIS junctions placed near 1D flow areas•No throttle body•No fuel spray particles in CFD domain•k-ε turbulence model•Full intake modeled for each diffuser change•Iterations in WAVE-VECTIS same as Case1 – Same Plenum Used
Inlet Box
2006-01-36549
VECTIS – Mesh Details
3.5mm1.75mm
~ 600,000+ cells
2006-01-365410
Theoretical Maximum Volumetric Efficiency Through a Restrictor Orifice
Engine Air Flow Demand SecondPerCyclesntDisplacemeQnom **21
.
0
00
11
*
1
2
pA
Qthroat
choked
Maximum Isentropic Flow Through an Orifice Area
.
max..nom
choked
Q
QEV
Maximum Theoretical Volumetric Efficiency(Limited by Restrictor)
Implicit Assumptions:
•Steady state flow
•Restrictor throat is choked 100% of the cycle
•Flow through restrictor equals flow through the intake valves.
•No pulse tuning effects
2006-01-365411
Volumetric Efficiency Predictions (CASE 1)
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
9000 10000 11000 12000 13000 14000 15000RPM
VO
LU
ME
TR
IC E
FF
ICIE
NC
Y
Theoretical Maximum
7 Degree - WAVE
5.5 Degree - WAVE
4 Degree - WAVE
3 Degree - WAVE
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
9000 10000 11000 12000 13000 14000 15000RPM
VO
LU
ME
TR
IC E
FF
ICIE
NC
Y
Theoretical Maximum7 Degree - WAVE 7 Degree - WAVE-VECTIS 5.5 Degree - WAVE 5.5 Degree - WAVE-VECTIS 4 Degree - WAVE 4 Degree - WAVE-VECTIS 3 Degree - WAVE 3 Degree - WAVE-VECTIS
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
9000 10000 11000 12000 13000 14000 15000RPM
VO
LU
ME
TR
IC E
FF
ICIE
NC
Y
Theoretical Maximum
Typically defined as the “choke point” of the restrictor
2006-01-365412
Volumetric Efficiency Comparison – 1D vs. 1D/3D vs. Theoretical Maximum
Theoretical Maximum
Wave-Vectis
WaveDifference
(Wave - W-V)
Difference (Wave-Vectis -
Theoretical Max.)
7° 10 126.5% 97.2% 103.7% 6.5% -29.3%
4° 10 126.5% 101.3% 105.3% 4.0% -25.2%
7° 11.5 110.0% 100.5% 107.9% 7.4% -9.5%
7° 14 90.4% 87.8% 93.7% 5.9% -2.6%
5.5° 14 90.4% 90.3% 93.2% 2.9% -0.1%
4° 14 90.4% 91.3% 92.9% 1.6% 0.9%
3° 14 90.4% 91.7% 94.3% 2.6% 1.3%
Half Angle
RPM x1000
VOLUMETRIC EFFICIENCY
2006-01-365413
CFD Results - Mach Number, Time AveragedPlanar Mean Mach Number at 14,000 RPM
(Time Averaged)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Normalized Area
Mac
h [#
]
7° 5.5° 4° 3°
2006-01-365414
CFD Results – Mach, Time Avg.- 14,000 RPM
3°
4°
5.5°
7°
2006-01-365415
Velocity over one cycle – 14,000 RPM
•Movies frames are at 5° CAD resolution•Scale peak at 347m/s
7° 4°
2006-01-365416
Mach Number – Over One Cycle (1D/3D), 14,000 RPM
Restrictor Half Angle
RPM% of Cycle
Choked7° 10,000 0%4° 10,000 15.3%7° 11,500 0%7° 14,000 47.9%
5.5° 14,000 68.0%4° 14,000 80.0%3° 14,000 83.3%
7° 4°
2006-01-365417
Mach Number – Over One Cycle (1D/3D),10,000 RPM
• 4° achieves supersonic velocities yet outperforms 7°
7° 4°
2006-01-365418
Flow Uniformity of Mach NumberFlow Uniformity of Mach Number at 14,000 RPM
(Time Averaged)
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1 2 3 4 5 6 7 8 9 10 11 12 13
Normalized Area
Flo
w U
nifo
rmity
Ind
ex_
7° 5.5° 4° 3°
Flow Uniformity of Mach Number at 10,000 RPM(Time Averaged)
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1 2 3 4 5 6 7 8 9 10 11 12 13
Normalized Area
Flo
w U
nifo
rmity
Ind
ex_
7° 4°
2006-01-365419
Turbulent Kinetic Energy (TKE) – Time Averaged
Planar Mean Turbulent Kinetic Energy at 14,000 RPM(Time Averaged)
0
200
400
600
800
1000
1 2 3 4 5 6 7 8 9 10 11 12 13
Normalized Area
Tur
bul
ent
Kin
etic
Ene
rgy
[m2 /s
2 ] 7° 5.5° 4° 3°
Planar Mean Turbulent Kinetic Energy at 10,000 RPM(Time Averaged)
20
40
60
80
100
120
140
1 2 3 4 5 6 7 8 9 10 11 12 13
Normalized Area
Tur
bule
nt K
inet
ic E
nerg
y [m
2 /s2 ]
7° 4°
2006-01-365420
T.K.E. – One Cycle at 14,000 RPM
4°
5.5°
7°
3°
2006-01-365421
TKE vs. Mach4° at 14,000 RPM
Rising Mach #
Minimum TKEPeak TKE
Falling Mach #
2006-01-365422
Turbulent Kinetic Energy Over One Cycle
7° 4°
2006-01-365423
Total Pressure – Time AveragedPlanar Mean Total Pressure at 14,000 RPM
(Time Averaged)
78
79
80
81
82
83
84
85
86
87
88
89
90
1 2 3 4 5 6 7 8 9 10 11 12 13 14Normalized Area
Tot
al P
ress
ure
[kP
a]
7° 5.5° 4° 3°
• Total Pressure at diffuser did not agree well with VE• Total Pressure loss vs. TKE shows good agreement
Planar Mean Total Pressure at 10,000 RPM(Time Averaged)
95.4
95.8
96.2
96.6
97.0
97.4
97.8
98.2
98.6
99.0
1 1.25 1.5 1.75 2 2.25 2.5 2.75 3Normalized Area
Tot
al P
ress
ure
[kP
a]
7° 4°
2006-01-365424
Total Pressure – Time Averaged
7°
5.5°
4°
3°
2006-01-365425
Total Pressure Over One Cycle, 4° at 14,000 RPM
2006-01-365426
Total Pressure Over One Cycle 4° at 14,000 RPM
2006-01-365427
Acoustic Filtering (WAVE Only)• Reduce unsteady flow
through use of acoustic filtering
• Pressure pulse frequency given by:
• Helmholtz resonator used to attenuate desired frequency
• Plenum volume used mesh with 15mm cell size
30
*#
K
CylindersofOrderEngineRPMf
2006-01-365428
Acoustic Filtering
• Helmholtz Resonator A tuned for 442 Hz or 13,260 RPM (1st E.O.)
• Vol. = 0.761 liters• Helmholtz Resonator B tuned for 416 Hz or 12,480 RPM (1st E.O.)
• Vol. = 0.856 liters• Mesh held constant between different cases
2006-01-365429
Flow Effects of Adding Helmholtz Resonator A at 13,750 RPM
• Pressure & Velocity variation decreased• % cycle choked increased• Can not ascertain possible influence on separation with WAVE
2006-01-365430
Flow Control – Swirl Vanes• Reduce onset of separation after throat by increasing dynamic radial pressure component
• Flow separation reduction found in SAE 2003-01-0840
• Impart radial momentum through use of swirl vanes.
• Used with 7 Degree Restrictor•Run at 14,000 RPM•VE within 0.1%
2006-01-365431
Swirl Vanes vs. No Swirl Vanes – 14,000 RPM
With Swirl Vanes No Swirl Vanes
2006-01-365432
Conclusions Length of the intake plays an important role in VE curve; not just
volume Lower diffuser angles provided:
■ Lower TKE values■ Higher percentage of cycle achieving choked flow through the throat■ Better flow uniformity at 10,000 & 14,000 RPM■ Volumetric Efficiency closer to that predicted by WAVE■ Increased time averaged total pressure at the diffuser exit
3° & 4° restrictors achieved VE numbers greater than steady state based Theoretical Maximum VE
Peak TKE values occurred during falling Mach numbers, due to adverse pressure gradient.
Flow separation occurred for all diffuser angles and occurred at approximately the same area ratio
Acoustic filtering provided encouraging results in increasing VE, albeit in a narrow rpm range
Swirl vane flow control devices■ Intangible benefit to Volumetric Efficiency■ Reduced TKE values, but also Total Pressure■ Possible Benefits at lower RPM or with more fine tuning?
2006-01-365433
ACKNOWLEDGEMENTS• Minnesota Supercomputer Institute• Ricardo – Patrick Niven & Karl John• Univ. of Minnesota – Mechanical
Engineering• Dr. Patrick Starr• Dr. David Kittleson• Dr. William Durfee
• Kim Lyons – Daimler-Chrysler• Minnesota State University Mankato – Dr. Bruce Jones
QUESTIONS ?
2006-01-365435
Static Pressure – Time AveragedPlanar Mean Static Pressure at 14,000 RPM
(Time Averaged)
40
45
50
55
60
65
70
75
80
85
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Normalized Area
Sta
tic P
ress
ure
[kP
a]
7° 5.5° 4° 3°