Ivn Sidorovich Aerodynamicist Agenda Bicycle aerodynamics - - PowerPoint PPT Presentation

Iván Sidorovich Aerodynamicist

P5 SIMPLY FASTER - 2

Agenda

• Bicycle aerodynamics background

information

• Cervélo’s design history
• A new tool: STAR-CCM+
• P5 aerodynamic development
• Conclusions

P5 SIMPLY FASTER - 3

Bicycle aerodynamics background information

P5 SIMPLY FASTER - 4

The importance of aerodynamics

~ 365 W (100%) ~ 320 W (88%) ~ 25 W (7%) ~ 20 W (5%)

P5 SIMPLY FASTER - 5

How do we measure aerodynamic drag?

• We do not measure in Cd, CdA or Force (N)
• We measure in grams (g)

– Good scale for bicycles

• Aero drag rule of thumb

– 100 grams of drag

• = ~ 10 Watts
• = ~ 1 second / km

Ref: (Martin et. al., “Validation of a Mathematical Model of Road Cycling Power”, 1998)

P5 SIMPLY FASTER - 6

P5 P4

Example: P5 to P4 aero improvements

P5 SIMPLY FASTER - 7

Example: P5 to P4 aero improvements

Results P5 P4 Diff % diff Average Drag ±20° (g) 2056 2122

• 66

3.11 EXAMPLE Time savings (s) Distance savings (m) 50 km TT 33 @ 50 kph 458 @ 50 kph

P5 SIMPLY FASTER - 8

Cervélo’s design history

P5 SIMPLY FASTER - 9

• 1st curve seat tube
• Aero seat stays
• Pen and paper process
• Experience
• Tunnel validation

P2 - P3 P3 - P4 Dev time (months) 6 Drag savings (g) 32

P2 P3

P2 to P3 aero development

P5 SIMPLY FASTER - 10

• Integrated rear brake
• Drop down tube
• Integrated water bottle
• Wind tunnel experimentations (5)

P2 - P3 P3 - P4 Dev time (months) 6 24 Drag savings (g) 32 32

P3 P4

P3 to P4 aero development

P5 SIMPLY FASTER - 11

P5 SIMPLY FASTER - 12

• System is bike plus rider
• Rider was scanned to make a life-size rider

model

• Dummy on bike is 100% fixed and

repeatable

• To properly measure the effects of individual

design features

• Need to make only one change
• Test mule

Wind tunnel experimentation

Mule SLA

P5 SIMPLY FASTER - 13

Wind tunnel experimentation

• P3 to P4 aero development process involved

– Several wind tunnel trips (5) – Numerous SLA parts – VERY expensive!

• Limited amount of runs

– Constraint

• Time (25 runs per day)
• Money
• Schedule availability

– Makes you be more conservative!

P5 SIMPLY FASTER - 14

A new tool: STAR-CCM+

P5 SIMPLY FASTER - 15

CFD – Validation case

• Need to make sure CFD gives

correct results – In absolute terms – In relative terms

• A simple case of frame only was

chosen from wind tunnel experimentations

P5 SIMPLY FASTER - 16

CFD – Validation results

• For the mesh we use a

combination of

• Trim for outer domain
• Poly for MRF
• 35 million elements
• Turbulence
• ko-SST for steady state
• DES for transient

2.2 % off wind tunnel data

P5 SIMPLY FASTER - 17

CFD – Validation results with rider

3.3 % off wind tunnel data

P5 SIMPLY FASTER - 18

New aerodynamic design cycle

• Phase 1

– CFD initial concepts with a parametric model – Wind tunnel experimentation with best CFD concepts

• Phase 2

– CFD detailed analysis – Production model wind tunnel experimentation

• Competitor benchmarking

P5 SIMPLY FASTER - 19

• 1st bike to be fully design with CFD

P4 P5

P4 to P5 aero development

P2 - P3 P3 - P4 P4 - P5 Development time (months) 6 24 7 Drag savings (g) 32 32 66

P5 SIMPLY FASTER - 20

P5 SIMPLY FASTER - 21

P5 aerodynamic development

P5 SIMPLY FASTER - 22

3 key design areas

• Aerobar
• Front end
• Seat tube

Aerobar

P5 SIMPLY FASTER - 24

Phase 1 – initial parametric model

P5 SIMPLY FASTER - 25

Phase 1 – best CFD models

Model A Model C

P5 SIMPLY FASTER - 26

Phase 1 – best CFD models in the tunnel

Model A Model B Model C

P5 SIMPLY FASTER - 27

Phase 2 – detailed CFD models

TT2 Low Version TT2 Vcap TT2 Pylon Long Base Ext TT2 Vcap Long Base Ext

P5 SIMPLY FASTER - 28

• Cables can add up to 15 grams
• f drag when exposed
• Ultra low friction cable path

Phase 2 – final model selection

Front end

P5 SIMPLY FASTER - 30

Phase 1 – initial head tube parametric model

• We started the head tube with a parametric

conceptual model to test against P4

P4 Baseline Head tube conceptual x-sections

P5 SIMPLY FASTER - 31

Phase 1 – initial head tube parametric model @ 6° yaw

P4 Baseline Head tube conceptual design

P5 SIMPLY FASTER - 32

Phase 1 – best CFD models in the tunnel

Smooth Ridge

P5 SIMPLY FASTER - 33

Phase 2 – initial brake model

• We started the analysis by investigating the aerodynamics of a

naked brake

Standard Magura Production

P5 SIMPLY FASTER - 34

Phase 2 – best CFD models

• It was discovered that Magura model was faster than standard

Standard Magura

P5 SIMPLY FASTER - 35

Phase 2 – brake head tube integration

P5 SIMPLY FASTER - 36

Phase 2 – brake head tube final model

P4 P5

P5 SIMPLY FASTER - 37

P4 P5

P5 SIMPLY FASTER - 38

P4 P5

Seat tube

P5 SIMPLY FASTER - 40

Phase 1 – initial parametric model

P5 SIMPLY FASTER - 41

Phase 1 – best CFD models

Baseline (P4) Straight Full gusset Small Gusset

P5 SIMPLY FASTER - 42

Phase 1 – best CFD models in the tunnel

Baseline (P4) Straight Full gusset

P5 SIMPLY FASTER - 43

Phase 2 – final model selection

P5 SIMPLY FASTER - 44

P5 P4

Phase 2 – final wind tunnel experimentation

P5 SIMPLY FASTER - 45

RESULTS P4 - P5 Drag savings (g) 66 % savings 3 Time savings (s/km) 0.66 50 km TT 33 180 km Iron Man 119

Final results

P5 SIMPLY FASTER - 47

Competitor benchmarking: 7 total

P5 SIMPLY FASTER - 48

P5: Lowest drag of any triathlon bike ever

P5 SIMPLY FASTER - 49

P5: Lowest drag of any triathlon bike ever

Savings Best Worst Drag (g) 60 110 Power (W) 6 11 Time (s/km) 0.6 1.1

P5 SIMPLY FASTER - 50

Conclusions

P5 SIMPLY FASTER - 51

Conclusions

• The P5 is the fastest triathlon and time trial bike ever
• CCM+ has dramatically speed up the development time to market
• The new aerodynamics design cycle has allowed us to be more

innovative, yielding better and more exiting bicycles

Questions?