GB2559944A - Method of optimising aerodynamic efficiency - Google Patents

Abstract

A method running on a computer for optimising the aerodynamic efficiency of an athlete comprises inputting a plurality of first data sets containing data representing the aerodynamic drag of a plurality of articles/garments such as data representing the aerodynamic drag of said items over a range of predetermined air speeds, inputting a second data set comprising route-specific data for a planned journey such as data represent the topographical profile of a planned journey, inputting a third data set comprising athlete-specific data such as the power output of the athlete, and running a computer simulation to calculate a performance indicator for the athlete when associated with each of the articles over the planned journey. This is done by splitting the journey into a plurality of route sections, calculating the progress of the athlete through each route section taking into account the initial speed of the athlete, the drag of the article, the route data for said section and the athlete specific data and using these to determine the progress of the athlete through each section. The performance indicators are then compared, so that the article associated with the most favourable performance indicator can be identified.

Description

(54) Title of the Invention: Method of optimising aerodynamic efficiency

Abstract Title: Method for optimising the aerodynamic efficiency of an athlete (57) A method running on a computer for optimising the aerodynamic efficiency of an athlete comprises inputting a plurality of first data sets containing data representing the aerodynamic drag of a plurality of articles/garments such as data representing the aerodynamic drag of said items over a range of predetermined air speeds, inputting a second data set comprising route-specific data for a planned journey such as data represent the topographical profile of a planned journey, inputting a third data set comprising athlete-specific data such as the power output of the athlete, and running a computer simulation to calculate a performance indicator for the athlete when associated with each of the articles over the planned journey. This is done by splitting the journey into a plurality of route sections, calculating the progress of the athlete through each route section taking into account the initial speed of the athlete, the drag of the article, the route data for said section and the athlete specific data and using these to determine the progress of the athlete through each section. The performance indicators are then compared, so that the article associated with the most favourable performance indicator can be identified.

Fig. 30

At least one drawing originally filed was informal and the print reproduced here is taken from a later filed formal copy.

1/24

1306 18

2/24

80

m

n (

:yii

nd

er

•Dr

ag

Co

ett

ice

nt

2d

vs

Sp<

?e<

Πτ

/s<

?c

> 'x

<

- '«

Y\

Ί

I.-..,

‘''-I

3

Γ——.1

‘Ί

ks

1 ~ ---4

L. * 1

Π-

\ J

»

\

►----,

u

\

► -··-

► ..-Η

-.—.

.

>··—«

ι----.

-Ί tZA

F=1 F~71

.J

r'

.4

.J

<

C-<

.4

.4 »-

' 4

4ai

ΙΐΤΗ

teT

Sa

Tip

e /

i-S

am

pte

u

Sa

Tip

e t

,-S

am

pie

B

1306 18

Speed m/sec

Fig. 2a

Ό

U

Speed m/sec

Fig. 2b

3/24 ra >«-· £

o

CL

CD

Ο

CO

400 ij§

1000

600

Roiling Resistance Climbing Acceleration Wheel inertia Aero With Draft Aero No'Draft g 200

-200

-400

2000 4000 6000 8000

Time (s)

10000 12000 14000

Fig. 3

Time (s)

Fig. 4

4/24

---—™-		CDA(mA2) vs Air Speedj m/sec)
	S |	8	11 |	14	17	Average
{Standard Race Kit	0,29890	0.29460	0.28790	0,28720	0,28720	0.29116
2016 One Piece’	0.28440 s	0.28270	0,27990	0.27540	0,27450	0.27938
2015 ’Two Piece’	0,28680(	0.28160	0.27760	0,27810	0,27610	0.28004
{2016 XowSpeed'	0.275001	0,27900	0,28000(	0.28300	6.28400	0.28020

Fig. 5

1306 18

Fig. 6

5/24

1306 18

2016 Low Speed Dev	Energy Saved (kJ)	61.77	60.54	59.12	•χΓ kO cn LO	61.20
Watts Saved	4.19	4.35	4.08	cn co xT	4.21
Average Watts	244.07	262.24	247.56	246.26	248.85
2016 2 Piece	Energy Saved (kJ)	106.00	106.75	104.99	101.03	107.20
Watts Saved	CO	cn co	7,55	2 kO	7,55
Average Watts	240.86	258.70	244.09	243.16	245.51
2016 1 Piece	Energy Saved (kJ)	104,19	105.08	98.48	95.95	106.79
Watts Saved	7.27	7.75	LO CO	CO NO	CO LO
Average Watts	240.99	OO CO LO CM	244.59	243.54	245.53
2015 kit	Average Watts	248,26	266.59	251.64	250.35	253.06
Stage 2	Rider	<	CQ	u	Q	LLJ

Ρχ

6/24

CD

Ο

Time (s)

Rolling Resistance Climbing Acceleration Wheel Inertia Aero With Draft Aero No Draft

Fig. 8

7/24

1306 18

Dev	Energy Saved (kJ)	83.87	83.39	83.10	83.29	83.84
“C3
CU	Ό
<U czx.	CU > Π3	OX	co	en	Xj-	ox
co	CO	xj	co	Xj	Xj-	Xj
Si	<zi —l·—	Xj·	xT	Xj	Xj	xj”
o	<r>
—i	s
xO
>~t	tZi
θ	_J__? Jj—t
rsi	Π3	CT	CO	co	CO	Γ'-.
		ox	ft	co	xj-	Xj-
	<u cn <T3	xj·	QQ	LO	CO	CM
	CO	CO	CXJ	CO	CO
	CU >·	CXi	CXi	cxi	cxi	cxi
	*=£
	§
	TT cu	t-H	Xj”	co	o	ft
	7>	cn	cxi	LO	cxi	xj”
	co		cd	GO	LO
	>>.	θ	co	co	CO	co
	ΟΊ	*··^	»—t	ft	ft	ft
	cu
	<77
	!!!
c?	~C3
cu	CU
£X_ f\i	3> co co	r^. CXi	OX OO	LO CO	CXi ft	r^. cxi
xO	cz> XJ co	LO	xj	LO	LO	LO
θ
f\i
	cz> i -4-J>
	jU	kO	or	CO	LO	ox
	—*	fi	uo	ft	r-'-v	MO
	cu <T*<	xj		Xj	ox	fi
	03	CO	CO	CXi	cxi	co
	5_ Π i	CXi	cxi	cxi	cxi	CXi
	>-
	f-χ
	“C7> CU >>	cn o-	co ox	OO CXi	OO LO	ox r-
	U~i cn	xd co	Xj S	s	co 2	xd co
	cu
	jtt:
	LLJ
C? CU	ΧΞ3 CU
Cl.	>*
	Π3	lo	MO	LO	Cxi	co
	CO	co	CO	ft	OX	co
xO	<z> i	LO	xj	LO	xf	LO
CO	-4-J> «3
Cxi
	cn jj—>
		OO	cxi	co	LO	OO
		CO	OO	co	ox	CO
	CU cn	xj		xj	ox	ft
	Π3	co	co	cxi	cxi	co
	i_ CU	CXi	cxi	Cxi	cxi	Cxi
	>·
	cz>
	-4—* -L->
p 1	Π3
	=ϊ	co	OO	OO	i'-'-v	M3
	Xj“	xj-	VC	OO	ox
io	cu C73 Π3 L_ CU	ox	cxi	ox	XT	MO
co CXi	co	xj-	CXi	co	CO
CXi	cxi	cxi	cxi	cxi
	>
co
cu C7) <X3 -4—*	L_
CU Ό c£	<	CQ	u	Q	LU
co

α

CD u_

8/24

	1000
	800
	600
	400
	σ/
	I 200
	ω § 0 CL
00	-200
CD O	-400
CO	-600

2000 4000 6000

10000 12000 14000 16000 (S)

Rolling Resistance Climbing Acceleration Wheel Inertia Aero With Draft Aero No Draft

Fig. 10

9/24

1306 18

Dev	Energy Saved (kJ)	36.84	38.23	40.56	38.27	41.26
Ό
QJ	Ό
CD o.	CD Ξ» Π3		04	CD	CD	Γχ
CZ)	cz>	oo	04	cn	CD	•=T
s	04	„„j	^™i	cxj	CXJ	cxj
o	ro 3
X©
7~~i	04
CD	-!—#
CXi	fO	ρχ.	, CD	CXi	m	cxi
		cn	r-·		sf“	04
	<d cn ro	^“*ϊ		fx	CXJ	LD
		•=3*	r—H	γ*η	cn
	i_ fl 1	cxi	CXJ	cxi	cxj	cxi
	7=
	“C5
	<d	Γχ.	co	cn	LD	co
	<T3	cxj	cxj	LD	ΓΧ	cn
	<Z4	04	CD	04	x©	cxi
	7=- cn CD cz: LLJ		LD	Xj“	•=T	LD
xj	ΤΣ5
CD	cd
n 04	fO cn	cd	cxj	oo cn	04 04	r-χ x©
xO	U4 CCS	ex]	CXJ	cxj	T“i	cxj
CD
CXJ
	04
	CCS	CD	oo	χΤ	xT	cxj
	—*	T**i	xj-	xO	xT	fx.
	CD CD CO		•c“H	Γχ	cxj	LD
		xj“	τΉ	τ—H	cn
	X__ q j	rxi	cxi	cxj	cxi	CXJ
	-^C
	2
	T4
	CD 7> CO	CD	•c“H 04		•=T m	ld x©
	Energy S		ld XT	LD XT	cxi xT	X© xT
o CD	03 CD
Γ4	>
	ro	CO	co	T“i	cn	cn
	<Z)		LD	rx	*=T	04
xO	04 -i-J	v„i			T~~H
CD	ZP
Cxj
	04 -!—#
	fO	co	•c“H		CD	X©
		rx		cn	CD	*=t
	<d cn	t—HI	cxj	oo	m	xo
	ro	»—I	T		T~~{	cn
	&— CD	Cxj	cxj	CXJ	cxj	CM
	7=-
	04
	t··*
	f<5
'·*<		tH	04	cxj	cn	04
	CXJ	x©	CD	•=T	cn
LD	€L> 04 m	on	cn	CD	xT	co
CD cxj	v«	xT	cxj		cn
S_ €L>	CXJ	cxi	cxj	CXi	cxj
	>
	<c
CD
Cx3	s_
<L·* 04	CD Ό	<	CQ	u	Q	LU
fO	o£
GO

LL,

1000

Rolling Resistance Climbing Acceleration Wheel Inertia Aero With Draft Aero No Draft

Fig. 12

1306 18

11/24

2016 Low Speed Dev	Energy Saved (kJ)	LO	4.55	LO cn	co LO	11.21
Watts Saved		MD co	co CN	o> MD	4.14
Average Watts	258.67	288,52	289.05	260,65	229.99
2016 2 Piece	Energy Saved (kJ)	4.22	CTi OT CO	CO co LO	4.39	CO LO
Watts Saved	LO cn	CN	CN OO	CO co	2.50
Average Watts	258.82	288.67	289,96	260.92	231.63
2016 1 Piece	Energy Saved (kJ)	-4.15	LO 1	CO o> co i	CN !	-1.42
Watts Saved	cn CN l-J i	LO -r—! 1	2 vi s	MD CO t—! 1	-0.58
Average Watts	261.37	291.33	292,88	263.61	234.71
2015 kit	Average Watts	260.08	289.88	291,78	262.25	234.13
Aipe d'Huez	Rider	<	CQ	U	Q	LU

CO

12/24

1306 18

Time (s)

Rolling Resistance Climbing Acceleration Wheel Inertia Aero With Draft Aero No Draft

Fig. 14

13/24 co ο

Time (s)

Fig, 16

14/24

1306 18

Speed (kmh)

Fig. 17

15/24

CD

O

250

Time Ahead for Each 2016 Kit Option

Time (s)

Fig. 18

Time (s)

Fiq. 19

16/24

1306 18

Fig. 20

17/24 co ο

Fig. 22

18/24

1306 18

Speed (kmh)

19/24 co ο

Time (s)

20/24

Histogram of Speed (kmh)

300 -1-1-1-1-1-r

1306 18

15 20 25 30 35 40 45

Speed (kmh)

Fig. 26

21/24

1306 18

22/24

Inputs

1807 18

GPS Device Options
1 1 iOO \ g υ < Ύ. g ... V UJ s„. «J “2QV ω DISTANCE (KM)
ί i
Rider Variables Options
Rider Mass	Enter Rider Mass
Equipment Mass	Enter known bike mass Budget Road Bike (11 kg) High end road Bike (7 kg) TimeTriai Bike (9 kg)
Tire Selection	Winter Tires (0.OOO8) Performance {0.0006) Competition (0.00045) Track Tire (0.002)
Riding Position	CasualRoad Bike( CdA 0.45) Enthusiaste( CdA 0.37} Racing Road(CdA 0.31 Aero Road ( CdA 0.27) Time Trial (CdA 0.2}
:
Baseline Equipment	Component Options
	Rain jacket Club jersey Club Skinsuit
Performance	Course Ave Power Output	i
	Known Average Power Casual rider (1.50 Watts) Club Cyclist (200 watts) Road Racer (240watts) Elite Racer ( 290 watts)

23/24

Mathematical Solver
Aero Drag Power(W) Tire Frictional Power(Wj Potential? nergy Power; Wt	fynctioniair density,Cd A, Air Speed Va,Ground Speed Vg function; mass, Crr,Vb) functionirnass,gravity,gradient)	i/2,rhoXdA.Vaft2.Vg Crr.m.g.Vg m.g.Vg.Gr
nv~mass(kg) rho=denisty (Kg/m\3) Cd ~ Drag Coef A - FrontalArea(mA?)	Gr - road gradient;Risefmi/Oistanceim) A-Rcn-alArasWS; Vg - Ground Speed (rn/sec) Va - Air Speed; m/sec)

1807 18

COA vs Speed (Modifier)
WheelChoice	CiimbingWheei Mid Section Deep Section Oise Wheel
frame Choice	grand X Road Frame Grand V AeroFrame grand? Time Trial Frame
Garment Choice	Aero Road Gib Short & Jersey Aero Road One piece suit Time TnaiSr.it Triatnlon Suit

Predicted Outputs

Fig. 29B σ ® £ <

Q ^C

W & k· / ί

Ο«“^Baseline^ Performance enhancement

Ι/ ss/L.

a $ s a

DISTANCE (KM)

24/24

Input first data sets (drag data)

Input 2^nd data sets route-specific data)

co ο

co

Input 3^rd data sets (athlete-specifc data) </<

***

Run computer simulati each article

Save performance indicators for each article

Compare performance indicators

C C“ .>.·.«<. Zi 2S

Z*/

Output results

FB /**ϊ *0

ICJ. ou

-1Method of optimising aerodynamic efficiency

The present invention relates to a method running on a computer for optimising the aerodynamic efficiency of an athlete.

In particular, but not exclusively, the invention relates to a method for optimising the 5 aerodynamic efficiency of an athlete competing in a sport where the athlete typically travels at a speed in the range 5-25m/s, preferably 10-20m/s, for example cycling, running, skiing and speed skating, where the aerodynamic drag on an athlete can have a significant effect on the athlete’s performance.

When airflow passes over a body there are two fundamental mechanisms that produce a drag force. These forces come from surface drag, caused by friction as the air passes over the surface, and pressure drag caused primarily by the separation of vortices from the boundary layer. The ratio of surface drag to pressure drag is highly dependent on the shape of the object. Where objects are specifically shaped for optimum aerodynamic efficiency, the aspect ratio (length: width) will generally be at least 3:1. With an increased length to width ratio it is possible to have a wing-like shape with a narrow trailing edge. The advantage of this is that the flow can remain attached to the surface of the object so that the streamlines follow the shape of the profile. Although the surface area of the object and the resulting surface friction are increased, the flow is able to “recover” beyond the widest point of the object, resulting in a small net pressure drag. Generally, the reduction in pressure drag far outweighs the increase in surface drag.

-2The human body tends to have a much lower aspect ratio, particularly when upright, which may typically be nearer to 1:1 for the arms and legs, and 1:2 for the torso. As a result, the human body approximates to a “bluff body”, and pressure drag tends to be by far the larger contributory factor to the overall aerodynamic drag experienced by an athlete.

Where it is not practical to modify the shape of the body and the aspect ratio is lower than about 3:1 in the flow direction, a high level of pressure drag can be caused by flow separation soon after the flow has passed the widest point of the body. In such situations in engineering and nature, it is known to adjust the surface texture of the body to help delay the separation point and thereby reduce the net pressure force that retards motion of the object.

A number of techniques are known to reduce the net drag force on bluff bodies, including the use of trip edges and textured surfaces. Although these techniques may give rise to an increase in surface drag, it is generally possible to find a solution whereby the reduction in pressure drag outweighs the increase in surface drag. This allows the total drag to be reduced in various applications.

In the case of the human body, these techniques can be applied by designing a garment that provides the required trip edges and textured surfaces. The use of such garments can significantly improve the aerodynamic efficiency of an athlete, thereby potentially improving the athlete’s performance. Similar techniques can also be applied to other articles associated with an athlete, for example a bicycle, helmet, skis and so on.

In a sport such as road cycling, the speed range 5-25m/s, and particularly 10-20m/s, is critical as the aerodynamic drag is very significant and accounts for a large portion of the athlete’s power output. Competitors also tend to spend the majority of the race travelling at speeds in this range. This speed range is also critical in a number of other sports, for example running (typical speed 5-10m/s), speed skating (up to 15m/s), horse riding (up to 15m/s) and downhill skiing (average speed about 15-25m/s).

Aerodynamic drag is dependent on the airspeed of air passing over the garment, and different garments can have different drag/airspeed characteristics. For example, some garments may have a lowest drag coefficient at the lower end of the critical speed range (e.g. at an airspeed of about 5-10 m/s), whereas other garments may have a lowest drag coefficient at the upper

-3end of the range (e.g. at about 15-25 m/s). Therefore, when selecting a garment to wear for a particular event or competition it is advantageous to take account of the airspeeds that can be expected over the race route (or “planned journey”), and to choose a garment that is optimised for those airspeeds.

An obvious approach is to select a garment that provides the lowest possible aerodynamic drag at the expected average speed of the race. Many might assume that this would provide the optimum aerodynamic efficiency. However, we have found that this is not always true. While it may sometimes be true for races where the speed does not vary much (for example in time trials and cycle races on flat roads where the speed for most of the race is generally quite close to the average speed), it is not necessarily true for races where the speed varies considerably (for example in races on mountainous routes that involve steep ascents and descents). On such routes the athletes may spend very little time travelling at or near the average speed, instead mostly travelling either much slower than the average speed (when climbing) or much faster (when descending).

This leads to a dilemma regarding the choice of garments: for maximum aerodynamic efficiency, should the athlete select a garment that provides very low drag at the lower end of the expected speed range, or at the upper end of the range, or at some intermediate speed, for example at the expected average speed? Unfortunately, the answer to this dilemma is often not obvious and may sometimes be counter-intuitive.

It is therefore an object of the present invention to provide a method for optimising the aerodynamic efficiency of an athlete, while travelling over a planned route.

According to one aspect of the present invention there is provided a method running on a computer for optimising the aerodynamic efficiency of an athlete, comprising:

• inputting a plurality of first data sets comprising article-specific data for a plurality of articles that may be associated with the athlete, wherein the article-specific data includes data representing the aerodynamic drag of the article over a predetermined range of air speeds, • inputting a second data set comprising route-specific data for a planned journey, wherein said route-specific data includes data representing the topographical profile of the planned journey,

-4• inputting a third data set comprising athlete-specific data, wherein said athletespecific data includes data representing the power output of the athlete and, optionally, additional data such as the mass of the athlete, • running a computer simulation to calculate a performance indicator for the athlete when associated with each of the articles, wherein the performance indicator relates to the performance of the athlete over the planned journey and is calculated by:

o dividing the planned journey into a plurality of route sections; o calculating the progress of the athlete through each route section by solving equations of motion taking account of:

an initial speed at the start of each route section the aerodynamic drag of the article at the initial speed, route-specific data for the route section, and athlete-specific data, and o determining from the progress of the athlete through each route section a performance indicator that is related to the performance of the athlete over the planned journey, for example elapsed time, • comparing the performance indicators of the athlete associated with each of the articles, whereby the article associated with the most favourable performance indicator may be identified.

The above method makes it possible to optimise the aerodynamic efficiency of an athlete by identifying an article, for example a garment that provides an optimum performance indicator over a planned journey. This takes the guesswork out of selecting an article, such as a garment or other piece of equipment, for a particular journey and ensures that the athlete is able to perform to the peak of his/her ability.

Inputting a second data set comprising route-specific data for the planned journey may further comprise inputting data representing a forecast wind speed and direction relative to each route section of the planned journey. This adds an extra level of sophistication to the method and ensures that the optimum garment is selected, taking account of weather conditions as well as the topography of the journey. The second data set may further comprise route-specific data relating to gradients, altitude, turns, road surface, precipitation and/or temperature.

-5Inputting a third data set comprising athlete-specific data may additionally include inputting data relating to one or more additional performance characteristics, selected for example from a range comprising the strength, stamina, fitness, weight and body size of the athlete. These factors may also be taken into account when determining a performance indicator for the athlete over the planned journey.

The performance indicator may for example relate to the time required to complete the planned journey, or the energy required to complete the planned journey, or the average speed over the planned journey.

In an embodiment, each route section comprises a section of the planned journey defined by a predetermined distance interval. The performance indicator may for example be determined by calculating the time required to travel the predetermined distance interval.

In another embodiment, each route section comprises a section of the planned journey defined by a predetermined time interval. The performance indicator may be determined by calculating the distance travelled during the time interval.

In an embodiment, the article is a sports garment, which may be intended for use in a sport where the athlete typically moves with a speed in the range 5-25m/s, preferably 10-20 m/s, for example cycling, running, skiing, horse racing or speed skating. The sports garment may for example be a shirt, trousers, leggings, shorts, bibshorts, shoes, overshoes, arm covers, calf guards, gloves, socks or a bodysuit. Other articles of clothing are of course possible. Preferably the sports garment is close-fitting to the body so that it follows the contours of the body and does not flap significantly as the air flows over the surface of the garment.

The sports garment may have a surface texture in at least one region of the garment having a texture height in the range 0.1-0.8mm, preferably 0.2-0.6mm. In an embodiment, the sports garment comprises a fabric that has a texture provided by jacquard knitting of the fabric, or by printing a 3D pattern on the outer surface of the fabric, or by the application of a solid material, for example silicone, to the outer surface of the fabric.

In another embodiment the article is an item of sports equipment, for example a bicycle, a component of a bicycle, or a helmet. The article may be for use in a sport where an athlete

-6typically travels at a speed in the range 5-25m/s, preferably 10-20m/s, for example cycling, running, skiing, horse racing or speed skating.

Embodiments of the present invention will now be described by way of example with reference to the accompanying drawings, wherein:

Figure 1 illustrates schematically the flow of air around a cylindrical object;

Figure 2a illustrates graphically the results of a set of tests investigating the variation of drag 5 coefficient (CD) with air speed for various textured fabrics on an 80mm radius cylindrical body;

Figure 2b illustrates graphically the results of a set of tests investigating the variation of drag coefficient (CD) with air speed for various textured fabrics on a 130mm radius cylindrical body;

Figure 3 is a graph illustrating power output for an athlete in a cycling road race;

Figure 4 is a graph comparing calculated and actual power output in the same road race;

Figure 5 is a data table containing aerodynamic drag data for a number of different garments at a range of airspeeds;

Figure 6 is a graph illustrating the relationship between the drag coefficient CDA (m²) and 15 speed for each of the garments, based on the data set out in figure 5;

Figure 7 is a data table containing power output data for a number of different athletes during stage A of a cycle race, together with projected power output, power saving and energy saving data associated with the choice of different garments;

Figure 8 is a graph illustrating power consumption for one of the athletes during stage A of 20 the cycle race;

Figure 9 is a data table containing output data for a number of different athletes during stage B of the cycle race, together with projected power output, power saving and energy saving data associated with the choice of different garments;

-7Figure 10 is a graph illustrating power consumption for one of the athletes during stage B of the cycle race;

Figure 11 is a data table containing output data for a number of different athletes during stage C of the cycle race, together with projected power output, power saving and energy saving data associated with the choice of different garments;

Figure 12 is a graph illustrating power consumption for one of the athletes during stage C of the cycle race;

Figure 13 is a data table containing output data for a number of different athletes during a final part of stage C of the cycle race, together with projected power output, power saving and energy saving data associated with the choice of different garments;

Figure 14 is a graph illustrating power consumption for one of the athletes during a final part of stage C of the cycle race;

Figure 15 is a graph illustrating comparative predicted time advantages for an athlete wearing different garments during stage A of the cycle race;

Figure 16 is a graph illustrating comparative predicted distance advantages for an athlete wearing different garments during stage A of the cycle race;

Figure 17 is a histogram of illustrating the distribution of speeds during stage A of the cycle race;

Figure 18 is a graph illustrating comparative predicted time advantages for an athlete wearing different garments during stage B of the cycle race;

Figure 19 is a graph illustrating comparative predicted distance advantages for an athlete wearing different garments during stage B of the cycle race;

Figure 20 is a histogram of illustrating the distribution of speeds during stage B of the cycle race;

Figure 21 is a graph illustrating comparative predicted time advantages for an athlete wearing different garments during stage C of the cycle race;

-8Figure 22 is a graph illustrating comparative predicted distance advantages for an athlete wearing different garments during stage C of the cycle race;

Figure 23 is a histogram of illustrating the distribution of speeds during stage C of the cycle race;

Figure 24 is a graph illustrating comparative predicted time advantages for an athlete wearing different garments during a final part of stage C of the cycle race;

Figure 25 is a graph illustrating comparative predicted distance advantages for an athlete wearing different garments during a final part of stage C of the cycle race;

Figure 26 is a histogram of illustrating the distribution of speeds during a final part of stage C of the cycle race;

Figure 27 is a front view of a cyclist wearing a bodysuit for cycling;

Figure 28 is a side view of a cyclist wearing the bodysuit shown in Figure 27;

Figures 29A and 29B illustrate schematically some of the data that may be input into a computer simulation according to an embodiment of the invention, and some of the equations of motion that may be used in the simulation, and

Figure 30 illustrates some of the steps of a computerised method for optimising the aerodynamic efficiency of an athlete.

For the majority of the applications in which use of the invention is envisaged, the Reynolds number will have a value of up to 10⁶, such that the flow of air will be in the laminar/turbulent transition zone. We have used wind tunnel testing to understand and derive optimum textures for use in the invention, and in particular on garments that are worn in applications where they are exposed to an airflow with a speed in the range 5-20m/s.

In order to simplify experimentation, much of our research is based on optimising the drag around cylindrical objects with radii of 80mm and 130mm. This has enabled us to identify the surface requirements for a wide range of applications. Testing is conducted at a range of speeds and consideration is also given to wind direction. Within the sizes of cylinder used it is possible to approximate a range of curvatures that the airflow will encounter on a human

-9body in a range of applications. For example, for an adult, the upper arm typically has an average radius (based on circumference) of about 50mm, the thigh typically has an average radius of about 80mm, and the chest typically has an average radius of about 160mm. It is of course recognised that the human body is not a perfect cylinder and in regions such as the chest it is closer to an elliptical shape. However, a cylinder provides a good first approximation to an irregular curved body in which the radius of curvature is similar to that of the cylinder.

Figure 1 illustrates a typical airflow around a cylindrical body 2, wherein the longitudinal axis X of the cylindrical body is perpendicular to the direction of airflow relative to the cylindrical body. It will be understood that the movement of a body through stationary air may be modelled in a wind tunnel by creating a moving airstream that flows over a stationary body, as depicted in the drawings. In this example the direction of airflow as indicated by arrow S is perpendicular to the surface of the cylindrical body at point P, which is called the “stagnation point”. This is equivalent to forward relative movement of the body 2 through the air in the direction of arrow M.

On either side of the stagnation point P the airflow splits into two streams FI, F2 that pass around opposite sides of the cylindrical body 2. Up to approximately the widest point of the cylindrical body relative to the flow direction, the airflow is substantially laminar, allowing a boundary layer to build up against the surface 3 of the cylindrical body 2.

After passing the widest point of the cylindrical body 2 relative to the direction of flow, the flow streams FI, F2 tend to separate from the surface 3 of the cylindrical body, forming vortices V in the region behind the cylindrical body. This creates a low pressure zone F behind the cylindrical body 2 and the resulting pressure difference between the front and the rear faces 5, 6 of the cylindrical body creates a pressure drag force Fd that opposes movement of the cylindrical body relative to the air. The movement of air over the surface of the cylindrical body also creates a surface friction force F_s, which is usually much smaller than the drag force Fd at relative speeds in the range 5-40m/s.

The points where the boundary layer separates from the surface 3 of the cylindrical body 2 are called the transition points Τι, T2. The pressure drag force Fd experienced by the cylindrical body 2 depends in part on the area of the cylindrical body located within the low

-10pressure zone L between the transition points Τι, T2. If the transition points Τι, T2 can be moved rearwards, this will reduce the size of the area affected by the low pressure zone L, thereby reducing the pressure drag Fd acting on the cylindrical body 2.

It is known that the transition points Ti,T2 can be shifted rearwards by providing a suitable texture 8 on the surface of the cylindrical body 2. It should be understood that the texture pattern 8 shown on the upper part of the cylindrical body 2 may also be repeated on the lower side of the body. In an embodiment of the present invention we have sought to optimise the aerodynamic performance of an athlete by selecting an article, for example a garment, that provides the lowest possible drag , for example by selecting a fabric with a surface texture that maximises the reduction in pressure drag Fd without significantly increasing surface friction drag F_s.

As illustrated in Figure 1 the pressure drag force Fd can be reduced substantially, without significantly increasing the surface friction drag force F_s by covering the cylindrical body 2 with a fabric 3 having a textured pattern 8 on at least the side regions of its outer surface, between the front face 5 and the rear face 6 of the cylindrical body 2.

Much research has been done into the change in the drag on a cylindrical body through a range of speeds. It is well known that the drag coefficient falls and then increases again as the speed of the airflow increases for a given cylinder size. This is due to vortex formation and periodic shedding, which affects the laminar transition points behind the cylindrical body.

Our research has enabled us to modify this flow behaviour through the use of variable surface textures and thus minimise the pressure drag for the critical speed range (5-25m/s). We have conducted a series of wind tunnel tests to determine how different textured fabrics affect the drag coefficient for 80mm and 130mm radius cylinders at air speeds ranging from 4m/s to 25m/s. The results are illustrated in Figs. 2a and 2b.

In Fig. 2a the wind tunnel test results are illustrated for four different fabrics on an 80mm radius cylinder, and for the cylinder without a fabric covering:

• Sample A - a first prior art fabric • Sample B - a second prior art fabric

-11• Sample D - a new fabric with a chevron texture pattern, having a spacing of 10mm • Sample E - a new fabric with a chevron texture pattern, having a spacing of 7mm.

The fabrics were all wrapped around the circumference of the cylinder for the test.

As can be seen in Fig. 2a, the prior art Sample A fabric performed well at high speeds, having a CD of less than 0.70 at air speeds between 17m/s and 25m/s, but less well at lower speeds, the drag coefficient rising rapidly at air speeds of less than 17m/s to a value in excess of 1.00 at air speeds less than 14m/s.

The Sample B fabric performed well at low speeds, having a CD of less than 0.80 at air speeds between 6m/s and 8m/s, but less well at higher speeds, the drag coefficient exceeding 1.00 at air speeds of more than 16m/s.

With the new Sample D fabric the drag coefficient fell rapidly at air speeds greater than 5m/s, achieving a value of less than 0.80 at air speeds between lOm/s and 25m/s.

The new Sample E fabric performed even better at low speeds, achieving a drag coefficient of less than 0.90 across substantially the entire range of air speeds from 6m/s to 25m/s.

Therefore, neither the Sample A fabric nor the Sample B fabric provided a low drag coefficient across the entire lOm/s to 20m/s speed range, the Sample A fabric providing low drag only at high speeds between 17m/s and 25m/s, and the Sample B fabric providing low drag only at relatively slow speeds between 6m/s and 8m/s.

By comparison the two new fabrics provided a much wider range of low drag performance, the Sample D fabric providing a low drag coefficient at air speeds between lOm/s and 25m/s, and the Sample E fabric providing a low drag coefficient at air speeds between 6m/s and 25m/s.

In Fig. 2b the wind tunnel test results are illustrated for three different fabrics on a 130mm radius cylinder, and for the cylinder without a fabric covering. The fabrics were:

• Sample A - a first prior art fabric;

• Sample B - a second prior art fabric, and • Sample C - a new fabric with a chevron texture pattern, having a spacing of 7mm.

-12The fabrics were each wrapped around the circumference of the 130mm radius cylinder for the test.

As can be seen in Fig. 2b, the prior art Sample A fabric again performed well at high speeds, having a CD of less than 0.60 at air speeds between about 14m/s and 21m/s, but less well at lower speeds, the drag coefficient rising rapidly to a value in excess of 1.00 at air speeds less than about 12m/s.

The Sample B fabric performed well at low speeds, having a CD of less than 0.80 at an air speed of about 5m/s, but less well at higher speeds, the drag coefficient exceeding 0.90 at air speeds of more than 15m/s.

With both new Sample C fabric the drag coefficient was low across a wide range of air speeds, maintaining a value of less than 0.70 at air speeds between about 8m/s and 25m/s.

Therefore, the Sample A and Sample B fabrics again did not provide a low drag coefficient across the entire lOm/s to 20m/s speed range, whereas the new fabric provided a low drag coefficient across a wide range of air speeds, between 8m/s and 25m/s.

In summary, it can be seen that the different fabrics have aerodynamic drag characteristics that vary greatly, some fabrics providing lowest drag at the lower end of the critical speed range (around lOm/s) and other fabrics providing lowest drag at the higher end of the critical speed range (about 20m/s), while other new fabrics provide relatively low drag across a range of speeds from 8m/s to 25m/s.

The importance of aerodynamic drag on the performance of an athlete is clearly demonstrated by figure 3, which illustrated by way of mathematical modelling the power consumption for an athlete in a cycling road race, indicating how the power is consumed in overcoming rolling resistance, climbing, accelerating, providing wheel inertia and overcoming aerodynamic drag. Two values are provided for aerodynamic drag, illustrating the difference between drafting (e.g. riding within the pcloton) and non-drafting (e.g. riding solo). It can be clearly seen that aerodynamic drag consumes by far the largest portion of the athlete’s power output, particularly when riding solo, but also when drafting.

-13Figure 4 is a graph comparing calculated and actual power output in the road race illustrated by figure 3, wherein the calculated power output is generated by a computer simulation based on the topography of the race route. This graph indicates that it is possible to simulate with a good degree of accuracy the power output of an athlete during a particular race route (or planned journey).

Figures 5 and 6 illustrate how the aerodynamic drag coefficient CDA(m²) for a number of different sports garments varies according to the airspeed of air passing over the garment. The table and graph provide a comparison of the following garments worn by a rider:

1) 2015 Team Issue race kit

2) Type 1 - 2016 development two-piece road aero clothing

3) Type 2 - 2016 development one-piece road aero clothing

4) Type 3 - 2016 development two-piece with silicone texturing

This data is based on a pedal cycle rider holding a constant road position.

It can be seen that for some garments (e.g. garments 1, 2 and 3) the drag coefficient decreases as the speed increases from 5m/s to 17m/s, whereas for the other garment (4) the drag coefficient increases as the speed increases.

By making appropriate change to the rider’s CDA, we can predict how each of the above garments would have performed in a certain cycle race.

Stage A

Figure 7 illustrates the results of a mathematical modelling process, which shows how wearing different garments could affect the power and energy consumption for a number of different athletes (riders A, B, C, D & E) during stage A of the cycle race. It can be seen that garments 2, 3 and 4 all provide power and energy savings compared to the garments that were actually worn (1).

The graph presented in figure 8 shows the result of a mathematical modelling process to determine the power used to overcome each of the drag sources throughout the stage. This plot is for the rider A, but the trends and magnitudes are very similar for each rider. Two aerodynamic drag plots shown: one accounting for peloton draft effect, and one without. As

-14can be seen, the aerodynamic drag becomes hugely dominant without the benefit of drafting, and is also the largest consumer of power when drafting.

Figures 9 and 10 relate to Stage B, which was a typical stage of the race, with two classified climbs. The Stage B drag source plot shows aerodynamic drag is still dominant but at particular points during the route climbing also becomes a large consumer of power.

Figures 11 and 12 relate to Stage C, which included two major ascents. The Stage C plot shows that climbing power is dominant on the two climbs. However, the 2016 Fow Speed Development kit (4) has significantly lower CdA at low velocity in comparison to the 2015 kit (1), so the decrease in aerodynamic drag would be noticeable.

Stage C final part, Solo Climb

Figures 13 and 14 relate to the final section of Stage C. The final ascent was analysed separately as this was a critical point in the race. Fig. 13 illustrates the potential energy and power savings when comparing the three 2016 clothing choices (2, 3 & 4) to the 2015 option (1). It can be seen that on this ascent the 2016 1-piece suit (2) caused an increase in energy consumption for all 5 riders, whereas the 2016 2-piece suit (3) and the 2016 low-speed suit (4) produced savings in energy consumption. The 2016 low-speed suit (4) produced the greatest savings in energy consumption, as this was a low-speed route with steep climbs.

The Stage C plot (Fig. 14) shows how climbing dominates power usage in steep ascents, but aerodynamic drag still has a noticeable impact. The use of aerodynmically-optimised garments is therefore justified even on low-speed ascents. For the second climb, each watt saved is worth around 7 seconds. So using the 2016 Fow Speed Development kit would potentially have put rider E 28s further ahead by the end of the stage, which shows how much of an impact clothing choice can have even on mountain top finishes.

Figs. 15 to 26 illustrate by mathematical modelling the potential impact of garment choice on different stages of the cycle race.

-15Figs. 15 to 17 relate to Stage A, which was a relatively flat stage with uniform speed profile, as illustrated in Fig. 17, where most of the stage was covered at speeds in the range 40-50kph (ll-14m/s). Figs. 15 and 16 illustrate how each of the 2016 clothing options could have affected the relative position of one of the riders (in this case rider B) compared to the 2015 suit that was actually worn.

We assumed that if rider B was to make a solo break away then his power would increase by 25%, and this gave a similar velocity profile to the Peloton. A simulation was run for the four garments, and we can see in the graph (Fig. 15) what the time difference would be through the stage.

The one and two piece 2016 prototype garments have similar velocity profiles and we can see an improvement of 160 seconds over the stage. Interestingly, the gain in the low speed suit drops off because at the end of the stage the speeds will be higher.

Figs. 18 to 20 relate to Stage B, which included two classified climbs. The speed profile for the stage, illustrated in Fig. 20, shows that speeds were generally lower than in Stage A, most of the stage being covered at speeds in the range 30-40kph (8-llm/s). Figs. 18 and 19 illustrate how each of the 2016 clothing options could have affected the relative position of one of the riders (rider B) compared to the 2015 suit. On this stage the Tow speed’ suit is faster. Prototype garments of Type 1 and Type 2 are both better than the 2015 suit by 200 seconds. The low speed suit is 240 seconds faster.

Figs. 21 to 23 relate to Stage C, which included two major ascents. The speed profile for the stage, illustrated in Fig. 23, shows that much of the stage was covered at a speed of about 20kph (6m/s). Figs. 21 and 22 illustrate how each of the 2016 clothing options could have affected the relative position of one of the riders (rider E) compared to the 2015 suit. As may be expected, the aerodynamic performance is not so significant owing to the lower average speeds. However, we still see an improvement of 100 seconds with Type 1 and Type 2 garments, and the low speed Type 3 garment is around 140seconds better.

Finally, Figs. 24 to 26 relate to the last section of Stage C, including the second major ascent. The speed profile shown in Fig. 26, shows that most of this section was covered at a speed

-16of less than 20kph (6m/s). It can be seen that the Type 1 prototype suit is 11 seconds faster, the Type 2 suit is 18 seconds faster and the low speed suit is 33 seconds faster.

The examples set out above illustrate how selecting the optimum garment for a particular planned journey can have a significant impact on the athlete’s performance over that route.

However, it is not always obvious which garment will provide the greatest aerodynamic benefits over any particular route owing to the different speed profiles of different routes and the impact this can have on the aerodynamic drag experienced by the athlete.

The present invention seeks to provide a method running on a computer for optimising the aerodynamic efficiency of an athlete, for example by selecting the optimum garment or garments for a particular planned journey. In an embodiment, the method comprises the following steps, which are illustrated in figure 30.

The first step SI comprises inputting into the computer a plurality of first data sets comprising article-specific drag data for a plurality of articles having different aerodynamic drag characteristics, where in this example the articles are garments. Each first data set may for example contain data relating to aerodynamic drag coefficient of a garment when worn by an athlete over a predetermined range of air speeds, or of a specific fabric from which the garment may be made. This data may be obtained for example by wind tunnel testing with the garment worn by the athlete or placed on a body-shaped dummy having either a standard form or a form based on the body shape of a specific athlete. Alternatively, the drag coefficient of a fabric may for example be obtained from wind tunnel testing with a sample of the fabric wrapped around a cylindrical tube.

The second step S2 comprises inputting a second data set comprising route-specific data for a planned journey. This route-specific data includes data representing the topographical profile of the planned journey, for example height data representing the changes in altitude along the route of the journey, so that the gradients the athlete will have to ascend and descend during the journey can be taken into account. The planned journey may for example be divided into a plurality of sections, where each section has an associated height value. Each section may be quite short, for example 10m (or more or less), so that the speed of the athlete does not vary greatly during the section, or they may be longer, in which case the average speed during each route section may be used.

-17Other factors may also optionally be included in the route-specific data, to be taken into consideration when determining the predicted speed, including for example climatic conditions (e.g. wind speed and direction, temperature, likely precipitation), the length of the journey, the type of road surface (including the presence of water or ice on the surface) and so on. Some or all of these factors may be taken into account when determining the speed of the athlete at each section of the journey.

The term “speed” as used herein may mean either the road speed (i.e. the speed over the ground) or the effective wind speed, which will generally be equal to the road speed plus or minus a factor that depends on the speed and direction of the wind over the ground. The wind speed may be taken into account, depending on climatic factors (e.g. sea breezes caused by warm air rising over land) or forecast weather conditions.

The third step S3 comprises inputting a third data set containing athlete-specific data, including data relating to the power output of the athlete. The power output may for example be selected from a predetermined range (e.g. 180W for an amateur rider, 230W for a club rider, 260W for an elite athlete), or it may be based on the measured power output of a specific athlete.

The athlctc-spccific data may also include factors such as rider drag (which depends primarily on body shape, and may be either selected from a predetermined range, or based on the body shape of a specific athlete), the weight of the athlete (and/or bike), and other factors such as the possible effects of altitude on the performance of the athlete.

The fourth step S4 comprises running a computer simulation to calculate a performance indicator for the athlete when associated with each of the articles (e.g. when wearing each of the garments). The performance indicator relates to the overall performance of the athlete over the planned journey, so that the effects of the articles on the athlete’s performance can be compared.

The performance indicator is calculated by dividing the planned journey into a plurality of route sections and calculating the progress of the athlete through each route section by solving equations of motion taking account of an initial speed at the start of each route

-18section, the aerodynamic drag of the article at the initial speed, route-specific data for the route section, and athlete-specific data.

For example, the calculation may include looking up the speed at the end of a previous route section and taking this as the initial speed at the start of a new route section. The aerodynamic drag of the article at that initial speed is then looked up in the first data set. This provides an aerodynamic drag value acting on the athlete. Route-specific data for the route section is looked up in the second data set. From this data the gradient of the road in the route section is determined, which provides an accelerating or decelerating force on the athlete, depending on whether the gradient is negative or positive. The power output of the athlete is looked up in the third data set. This data and, optionally, other data such as tyre rolling resistance, wind data, the weight of the athlete and so on, is then used to calculate the progress of the athlete through the route section including, for example, any change in the speed of the athlete owing to aerodynamic drag, changes in gradient and other factors.

A performance indicator that is related to the performance of the athlete over the planned journey is determined from the progress of the athlete through each route section. This performance indicator may for example relate to the time taken to complete the planned journey, and may be determined by calculating the time required to complete each route section, and then summing the time intervals for all of the route sections.

The fifth step S5 optionally comprises saving the performance indicators for the athlete when associated with each of the different articles (for example, when wearing each of garments).

The sixth step S6 comprises comparing the performance indicators of the athlete associated with each of the articles, so that the article associated with the most favourable performance indicator may be identified. Optionally, the article associated with the most favourable performance indicator (e.g. the garment that produces the shortest time to complete the planned journey) may be indicated. The identity of the article with the most favourable performance indicator may be output in step S7 for example by displaying the identity of the article on a screen or outputting a printed report.

It should be understood that steps St, S2 and S3 may occur in any order or simultaneously, the only requirement being that the data is input into the computer before the computer runs

-19the simulation to calculate for each article (e.g. each garment) the progress of the athlete through each respective route section. The data input may be accomplished manually, for example using a keyboard, or the data may be downloaded from another device or memory device comprising part of the computer.

The data that may optionally be entered into the computer simulation, and some of the equations that may be used in the simulation, are illustrated in more detail in Figs. 29A and 29B.

The route-specific data DI may for example include values for elevation gain against distance travelled, where the elevation gain value corresponds to the height at a particular point as compared to the height at the start point of the journey.

The athlete-specific data D2 may include data relating to rider mass, equipment mass, tyre selection, riding position and performance criteria. As can be seen in this example, each of these values may for example be selected from a predefined list: for example the performance value may be selected from typical output values for a casual rider, a club cyclist, a road racer or an elite racer. In most situations, selecting a value from a pre-defined list will provide a sufficiently accurate result to assess the potential impact of an particular article, such as a garment, on the overall performance of an athlete, since the purpose of the simulation is to allow for a comparison between different articles, rather than deriving an absolute prediction for the performance of the athlete. However, if preferred, measured values may be used instead of values selected from a predetermined list.

The drag data D3 is selected according to selected baseline equipment options, for example choice of clothing, for example rain jacket, club jersey, club skinsuit etc.).

The computer simulation step S4A determines a baseline prediction for performance based on the division of the athlete’s power output between the three main sources of power consumption, namely aerodynamic drag power, tyre friction power and potential energy power. The computer simulation step S4B then repeats the simulation to assess the impact of various selected articles, such as choice of garment, wheel choice and bicycle frame choice. These simulations allow the potential impact of the various choices to be compared, for example by way of comparative output plots S7 of speed against distance covered.

-20The garment referred to above is preferably a sports garment, which may be used for any sport where the reduction of drag is important. This applies particularly to sports where the input power is limited (for example being supplied by the athlete or the force of gravity) and where the athlete travels at a speed typically in the range 5-25m/s, preferably 10-20m/s, for example cycling, running, speed skating or skiing. The article of clothing may for example consist of a shirt, trousers, leggings, shorts, bibshorts, shoes, overshoes, arm covers, calf guards, gloves, socks or a one-piece bodysuit. The article of clothing may also be an item of headwear, for example a hat or helmet, or a fabric covering for a helmet.

An example of a garment intended for use while cycling is illustrated in Figures 27 and 28. The garment in this case is a one-piece bodysuit 11 comprising a body portion 12 that covers the athlete’s trunk, with short sleeves 14 and legs 16 that cover the upper portions of the athlete’s arms and legs. The garment has a plurality of zones that are defined in relation to the direction of forward travel M of the athlete, and which take account of the athlete’s posture. The zones include a first zone A located generally in an inner front region of the garment, a second zone B located in a side region of the garment and a third zone C that is located in a rear region of the garment. The outer surface of the garment has a texture that is designed to reduce aerodynamic drag. In an embodiment, the texture height may vary across the three zones, for example the fabric being essentially smooth in the first zone A (e.g. having a height less than 0.2mm), have a height of 0.2-0.6 in the second zone B, and a height in third zone C that is either equal to or greater than that in the side region B.

In this example, the first zone A is located primarily on the chest and shoulder regions of the trunk 12 and on the forward facing portions of the sleeves 14 and the legs 16. The second zone B is located primarily on the side and back regions of the body 12 and on the side regions of the sleeves 14 and the legs 16. The third zone C is located primarily on the lower back portion of the body 12 and the rear portions of the sleeves 14 and the legs 16. This arrangement of texture patterns has been found to be particularly advantageous for cyclists adopting the classic crouched posture illustrated in Figures 27 and 28. It will be appreciated that in other sports where athletes adopt different postures, the arrangement of the texture patterns may be adapted as required to provide a low level of pressure drag.

Claims (15)

Claims:

1. A method running on a computer for optimising the aerodynamic efficiency of an athlete, comprising:

• inputting a plurality of first data sets comprising article-specific data for a plurality of articles that may be associated with the athlete, wherein the article-specific data includes data representing the aerodynamic drag of the article over a predetermined range of air speeds, • inputting a second data set comprising route-specific data for a planned journey, wherein said route-specific data includes data representing the topographical profile of the planned journey, • inputting a third data set comprising athlete-specific data, wherein said athletespecific data includes data representing the power output of the athlete, • running a computer simulation to calculate a performance indicator for the athlete when associated with each of the articles, wherein the performance indicator relates to the performance of the athlete over the planned journey and is calculated by o dividing the planned journey into a plurality of route sections; o calculating the progress of the athlete through each route section by solving equations of motion taking account of:

an initial speed at the start of each route section the aerodynamic drag of the article at the initial speed, route-specific data for the route section, and athlete-specific data, and o determining from the progress of the athlete through each route section a performance indicator that is related to the performance of the athlete over the planned journey, • comparing the performance indicators of the athlete associated with each of the articles, whereby the article associated with the most favourable performance indicator may be identified.

2. A method according to claim 1, wherein the second data set further comprises routespecific data relating to gradients, altitude, turns, road surface, wind, precipitation and/or temperature.

-223. A method according to any one of the preceding claims, wherein the third data set further comprises athlete-specific data relating to the strength, stamina, fitness, weight and/or body shape of the athlete.

4. A method according to any one of the preceding claims, wherein the performance indicator relates to the time required to complete the planned journey, or the energy required to complete the planned journey, or the average speed over the planned journey.

5. A method according to any one of the preceding claims, wherein each route section comprises a section of the planned journey defined by a predetermined distance interval.

6. A method according to claim 5, wherein the performance indicator is determined by calculating the time required to travel the distance interval.

7. A method according to any one of claim 1 to 4, wherein each route section comprises a section of the planned journey defined by a predetermined time interval.

8. A method according to claim 7, wherein the performance indicator is determined by calculating the distance travelled during the time interval.

9. A method according to any one of the preceding claims, wherein the garment has a surface texture in at least one region of the garment having a texture height in the range 0. ΙΟ.8mm, preferably 0.2-0.6mm.

10. A method according to any one of the preceding claims, wherein the article is a sports garment.

11. A method according to claim 10, wherein the garment is a shirt, trousers, leggings, shorts, bibshorts, shoes, overshoes, arm covers, calf guards, gloves, socks or a bodysuit.

12. A method according to any one of the preceding claim, wherein the article is an item of sports equipment.

13. A method according to claim 12, wherein the item of sports equipment is a bicycle, a component of a bicycle, or a helmet.

-2314. A method according to any one of the preceding claims, wherein the article is for use in cycling, running, skiing, horse racing or speed skating.

15. A method according to any one of the preceding claims, wherein the article is for use in a sport where an athlete typically travels at a speed in the range 5-25m/s, preferably 10-20m/s.