An Aerodynamic Study of Bicycle Wheel Performance - CD
Transcription
An Aerodynamic Study of Bicycle Wheel Performance - CD
An Aerodynamic Study of Bicycle Wheel Performance using CFD Matthew N. Godo, Ph.D. FieldView Product Manager STAR European Conference © 2010 Intelligent Light Background Wind Tunnel testing used extensively in cycling for over 20 years Benefits to cyclists from Wind Tunnels Typical test for Zipp, 85h at $850/h, conducted 3 or 4 times per year Advertiseme nt ca 2007 Improved knowledge of positioning Significant improvement in the performance of equipment (helmets, clothing, frames, wheels, spokes,…) Enhanced awareness of the role of aerodynamics in the community Current status Still considerable variation in design UCI rule changes & enforcement can be rapid & unpredictable Wind Tunnel reaching its limit today Interpretation of results „controversial‟ STAR European Conference © 2010 Intelligent Light How much does it matter? Tour de France 2008 Stage 20 Individual Time Trial 1 2 3.0% 3 4 From Greenwell et.al. Wheel drag is responsible for 10% to 15% of total aerodynamic drag 5 Finish Position 6 Rider makes up the majority of overall aerodynamic drag 7 8 9 10 11 12 13 Improvements in wheel design can reduce drag between wheels by as much as 25% Overall reduction in drag can be on the order of 2% to 3% 14 15 0 1 2 3 4 5 Percentage Time Difference Q Q Q Q Q IronManTM Lake Placid Triathlon 2008 Male 45-49 Age Group 1 2 3.3% 3 4 5 6 In the ‟05 Tour of Germany, Ullrich‟s Xentis front wheel was mistakenly fitted backward for the Stage 8 time trial. Although he won the trial, he finished second overall to Levi Leipheimer, behind by a final margin of 31 seconds. If the wheel had been the right way round, might Ullrich have won stage 8 by a greater margin, perhaps enough to win the race overall? Wheel manufacturer Xentis says „Yes!‟. Finish Position 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 Percentage Time Difference STAR European Conference © 2010 Intelligent Light 6 7 8 Scope Wheels* (700c) Forks 12 Zipp 1080 11 10 9 8 Zipp 404 7 6 5 4 3 2 1 0 Reynolds Full Carbon Aero Blackwell Time Bandit (slotted) Frame (partial) Zipp 404 Zipp 1080 13 Based on 2005 Razor Elite RANS calculations run for Reynolds Carbon Isolated front wheel Rotating Ground contact Zipp 1080 2 speeds (20mph & 30mph) 10 yaw angles (0o thru 20o) 120 total cases Blackwell Bandit Zipp 404 Wheel only Study is limited to Rim Depth [cm] *Continental Podium 19mm tubular tire STAR European Conference © 2010 Intelligent Light Methodology Overview STARCCM+ v4.06.011 Meshing models Polyhedral Mesh, prism layers Physics models Steady, incompressible, segregated solver RANS Turbulence K-Omega model SST Mentor All defaults applied Low Re Damping Modification turned ON Force Report convergence after 600 iter Low y+ wall treatment FieldView 12.2.1 (Intelligent Light) FV-UNS exported from STARCCM+ Parallel export compatible with FV STAR European Conference © 2010 Intelligent Light Boundary Conditions Surround Boundary Set upstream flow speed Set yaw angle for specific case Set forward axial speed 20mph or 30mph Matches constant direction of travel of bicycle Wheel, hub & spokes 0o thru 20o Ground Plane 20mph or 30mph Set rotational speed to match forward axial speed Wheel contact Rotational speed matches ground plane axial velocity STAR European Conference © 2010 Intelligent Light Boundary Conditions (cont’d) Nonconformal interface applied to inner region of wheel Permits accurate ground contact Allows for easy spoke count changes For steady case, For unsteady case, Use Moving Reference Frame Use rotational mesh motion Fork & Frame No-slip surface in relative reference frame STAR European Conference © 2010 Intelligent Light Postprocessing the Results For the wheel… Resolved Forces Drag, Vertical & Side Pressure & Viscous Top View Turning Moment Wind Velocity (effective) Axial Drag Force Wheel components Circumferential variations Turning Moments Based on side forces Side (Lift) Force Side View Vertical Force For the fork… Resolved Forces Drag & Side Bike Velocity (relative) Wind Velocity (effective) Axial Drag Force Showing Flow Field Features Use streamlines STAR European Conference Direction of Wheel Rotation © 2010 Intelligent Light CFD Results vs Wind Tunnel Data Wind Tunnel Protocols vary widely! “Wheel-only” studies mount wheel to floor with upright supports Tests start at 30 degrees yaw, angle gradually reduced Rotational wheel speed independently adjusted at each yaw Ground plane boundary condition differs (wind tunnel floor doesn‟t move) Results are often „normalized‟ STAR European Conference © 2010 Intelligent Light Drag Forces Zipp 404 Zipp 1080 Drag Force vs. Yaw Angle Drag Force vs. Yaw Angle 1.5 1.5 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only 1.2 0.9 1.2 0.9 0.6 0.6 0.6 0.6 0.5 0.3 0.4 0 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only 0.3 0.2 Drag Force [N] Drag Force [N] 0.5 0.3 0.4 0 0.3 0.2 0.1 0.1 0 2 4 6 8 10 12 14 16 18 20 0 2 8 10 12 14 Drag force varies significantly with yaw angle 6 Yaw Angle [degrees] Yaw Angle [degrees] 4 10 to 15 degrees yaw is considered a design target by manufacturers Significant differences seen comparing wheel/fork combinations Blackwell slotted fork w/ Zipp 1080 shows considerable promise STAR European Conference © 2010 Intelligent Light 16 18 20 Zipp 404 Circumferential Variation, Drag Force Zipp 1080 Direction of Flow No Fork Reynolds Carbon STAR European Conference Blackwell Bandit © 2010 Intelligent Light Side Forces Zipp 404 Zipp 1080 Side Force vs. Yaw Angle Side Force vs. Yaw Angle 6 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only 12 16 6 4 4 Side Force [N] Side Force [N] 20 8 12 8 8 2 2 0 0 0 2 4 6 8 10 12 14 16 18 4 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only 0 20 0 0 2 4 Yaw Angle [degrees] 6 8 10 12 Dependence on yaw from wind tunnel studies is generally linear Fork has only small influence on wheel side force 14 Yaw Angle [degrees] Side force varies significantly with yaw angle 4 Note: Side force scales are different for each wheel STAR European Conference © 2010 Intelligent Light 16 18 20 Turning Moment Zipp 404 Zipp 1080 Turning Moment vs. Yaw Angle 0.35 Turning Moment vs. Yaw Angle 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only 0.3 0 -0.2 Moment [N·m] Moment [N·m] 0.25 0.2 0.15 -0.4 -0.6 0.1 0.05 30mph, Blackwell Fork 30mph, Reynolds Fork 30mph, wheel only 20mph, Blackwell Fork 20mph, Reynolds Fork 20mph, wheel only -0.8 0 -1 0 2 4 6 8 10 12 14 16 18 20 0 2 4 Yaw Angle [degrees] 8 10 12 Turning moment varies significantly with yaw angle 6 Direction of moment for Zipp 404 acts opposite to Zipp 1080 Significant differences seen comparing wheel/fork combinations Both forks dampen moment on Zipp 404, Blackwell fork amplifies moment on Zipp 1080 STAR European Conference 14 Yaw Angle [degrees] © 2010 Intelligent Light 16 18 20 Zipp 404 Circumferential Variation, Side Force Zipp 1080 Direction of Flow No Fork Reynolds Carbon STAR European Conference Blackwell Bandit © 2010 Intelligent Light Flow Structures, Effect of Yaw Suction Side Zipp 404 Zipp 1080 At low yaw, Strong recirculation observed at top, outer edge of wheel Weaker recirculation observed at bottom half, inner edge of wheel As yaw angle increases, Top recirculation extends along front of wheel, combines with inner wheel recirculation STAR European Conference © 2010 Intelligent Light Flow Structures, Effect of Yaw Pressure Side Zipp 404 Zipp 1080 At low yaw, Backflow zone is created between fork and top of wheel As yaw angle increases, Inner wheel recirculation being driven from pressure side STAR European Conference © 2010 Intelligent Light Forces on Fork only Zipp 404 Zipp 1080 Drag Force vs Yaw Angle 1 1 0.9 0.9 Drag Force [N] Drag Force [N] Drag Force vs Yaw Angle 0.4 0.3 30mph, Blackwell 20mph, Blackwell 30mph, Reynolds 20mph, Reynolds 0.2 0.4 0.3 30mph, Blackwell 20mph, Blackwell 30mph, Reynolds 20mph, Reynolds 0.2 0.1 0.1 0 2 4 6 8 10 12 14 16 18 20 0 2 Yaw Angle [degrees] 6 8 10 12 Yaw Angle [degrees] Significant differences seen between forks 4 Blackwell Bandit slotted fork has much higher drag force (>2X) Choice of wheel does not significantly affect fork drag Dependence on yaw angle is very low STAR European Conference © 2010 Intelligent Light 14 16 18 20 Drag Force Profiles on Forks Zipp 1080 Blackwell Bandit Reynolds Carbon Zipp 404 STAR European Conference © 2010 Intelligent Light Flow Structures, Slotted Fork Suction Side Zipp 404 Zipp 1080 Flow is drawn into fork slots for all yaw angles, both wheels Flow is pulled away from the wheel rim & tire At higher yaw angles, Flow gets trapped behind fork Strong recirculation pulls flow upward STAR European Conference © 2010 Intelligent Light Flow Structures, Slotted Fork Pressure Side Zipp 404 Zipp 1080 Flow is drawn into fork slots for all yaw angles, both wheels, AGAIN! Even on the pressure side, flow is pulled away from the wheel rim & tire Pressure side flow at high yaw does not predominantly cross the center axis STAR European Conference © 2010 Intelligent Light 20mph Need for Automated Postprocessing Zipp 404 Zipp 1080 wheel only wheel only Reynolds Carbon Reynolds Carbon Blackwell Bandit Blackwell Bandit Production Challenge 18 months from concept to shelf Only a few weeks available to make design changes FieldView Automation Methodology Use FVX high level programming language Run FieldView Parallel 30mph wheel only wheel only Reynolds Carbon Blackwell Bandit Blackwell Bandit 5X speed-up on 8 processor system Operate concurrently using Batch-only licensing Reynolds Carbon Customizable environment, one-time investment Create spreadsheet-ready files, figures of merit & animations all at the same time 10 yaw angles for each folder, 120 sim files + 120 FV-UNS files ~500 GB (approx 2400 files in total) STAR European Conference © 2010 Intelligent Light How much does it matter? 45-49M Age Group Eagleman ‘09 Triathlon (World Championship Qualifier) Pick target speed (23mph) & wattage (275W) (use data from Cobb et.al. to get total drag) Estimate time spent at different yaw angles Use Zipp404/Blackwell combination to obtain baseline drag Wind (usually relatively) calm Course is flat Linear interpolation 20mph & 30mph Exchange front wheel & fork Calculate wheel drag for yaw angles Add to baseline drag Recalculate speed, same wattage Compare seconds saved STAR European Conference © 2010 Intelligent Light Future Work Examine transient effects Wheel/Component interactions Shedding frequency Force fluctuations Front fork can choke flow Calipers can cause significant disruption Effect of downtube position relative to wheel/faired to wheel unknown Automate postprocessing Cheaper compute Faster solvers More & more data Transient will add to this! “Rarely can one‟s bike set-up compensate as profoundly as improving the human on it.” Maffetone, P., Inside Triathlon, 1995, 10(3), p 20. STAR European Conference © 2010 Intelligent Light
Similar documents
rolf sestriere
Background • Wind Tunnel testing used extensively in cycling for over 20 years – Typical for Zipp, 85h at $850/h, run 3 or 4 times per year
More information