Improvement of Intake Restrictor Performance for a Formula

Improvement of Intake Restrictor Performance for a Formula SAE Race Car through 1 D & Coupled 1 D/3 D Analysis Methods Mark Claywell & Donald Horkheimer University of Minnesota

Agenda • • • Background 1 D Simulation Setup & Results Coupled 1 D/3 D Simulation Setup & Results • • • Volumetric Efficiency Flow measurements- Mach Number, Turbulent Kinetic Energy, & Total Pressure Acoustic Filtering for Volumetric Efficiency Flow Control Conclusions 2 2006 -01 -3654

Background • Overall Goal: Reduce the impact of the restrictor on VE • • • Ricardo WAVE 1 D model used with acoustic quality mesh in intake manifold • • Achieve a better understanding of flow in different restrictors Look for areas to reduce flow losses 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 3 2006 -01 -3654

Description of Intake Manifold Geometry Diffuser Half-Angle Diffuser Height Total Volume = Height Plenum + Diffuser Volume Diffuser Exit Volume Diameter Plenum Height Plenum Volume 4 2006 -01 -3654

Case 1: Same Plenum Used n. 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 5 2006 -01 -3654

Standing Waves - Case 1 4° 14, 000 11, 000 RPM 2 nd Order 4 th Order 6 th Order 7° 6 2006 -01 -3654

Case 2: Equal Diffuser Lengths n. Diffuser length had large impact on tuning n. Total volume had little impact on tuning ■ VE did not increase with increasing total volume 7 2006 -01 -3654

WAVE-VECTIS Setup Assumptions • WAVE-VECTIS junctions placed near 1 D 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 Case 1 – Same Plenum Used 8 Inlet Box 2006 -01 -3654

VECTIS – Mesh Details ~ 600, 000+ cells 1. 75 mm 3. 5 mm 9 2006 -01 -3654

Theoretical Maximum Volumetric Efficiency Through a Restrictor Orifice Implicit Assumptions: Engine Air Flow Demand • Steady state flow • Restrictor Maximum Isentropicthroat Flow Through an Orifice Area is choked 100% of the cycle • Flow through restrictor equals flow through the intake valves. Maximum Theoretical • No pulse tuning Volumetric Efficiency (Limited by Restrictor) effects 10 2006 -01 -3654

Volumetric Efficiency Predictions (CASE 1) Typically defined as the “choke point” of the restrictor 11 2006 -01 -3654

Volumetric Efficiency Comparison – 1 D vs. 1 D/3 D vs. Theoretical Maximum 12 2006 -01 -3654

CFD Results - Mach Number, Time Averaged 13 2006 -01 -3654

CFD Results – Mach, Time Avg. - 14, 000 RPM 3° 5. 5° 4° 7° 14 2006 -01 -3654

Velocity over one cycle – 14, 000 RPM 7° 4° • Movies frames are at 5° CAD resolution • Scale peak at 347 m/s 15 2006 -01 -3654

Mach Number – Over One Cycle (1 D/3 D), 14, 7° 16 4° 2006 -01 -3654

Mach Number – Over One Cycle (1 D/3 D), 10, 000 RPM 7° 4° • 4° achieves supersonic velocities yet outperforms 7° 17 2006 -01 -3654

Flow Uniformity of Mach Number 18 2006 -01 -3654

Turbulent Kinetic Energy (TKE) – Time Averaged 19 2006 -01 -3654

T. K. E. – One Cycle at 14, 000 RPM 7° 4° 5. 5° 3° 20 2006 -01 -3654

TKE vs. Mach 4° at 14, 000 RPM Peak TKE Minimum TKE Rising Mach ## Falling 21 2006 -01 -3654

Turbulent Kinetic Energy Over One Cycle 7° 4° 22 2006 -01 -3654

Total Pressure – Time Averaged • Total Pressure at diffuser did not agree well with VE • Total Pressure loss vs. TKE shows good agreement 23 2006 -01 -3654

Total Pressure – Time Averaged 7° 4° 5. 5° 3° 24 2006 -01 -3654

Total Pressure Over One Cycle, 4° at 14, 000 RPM 25 2006 -01 -3654

Total Pressure Over One Cycle 4° at 14, 000 RPM 26 2006 -01 -3654

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 15 mm cell size 27 2006 -01 -3654

Acoustic Filtering Helmholtz Resonator A tuned for 442 Hz or 13, 260 RPM (1 st E. O. ) • Vol. = 0. 761 liters • Helmholtz Resonator B tuned for 416 Hz or 12, 480 RPM (1 st E. O. ) • Vol. = 0. 856 liters • Mesh held constant between different cases • 28 2006 -01 -3654

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 29 2006 -01 -3654

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% • 30 2006 -01 -3654

Swirl Vanes vs. No Swirl Vanes – 14, 000 RPM With Swirl Vanes No Swirl Vanes 31 2006 -01 -3654

Conclusions n n Length of the intake plays an important role in VE curve; not just volume Lower diffuser angles provided: ■ ■ ■ n n n 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? 32 2006 -01 -3654

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 33 2006 -01 -3654

QUESTIONS ?

Static Pressure – Time Averaged 35 2006 -01 -3654