Aerodynamic drag of mountain bike tyres

R. J. Sunter and A. T. Sayers

Department of Mechanical Engineering, University of Cape Town, Rondebosch, South Africa

Abstract
This paper proposes the concept of a front wheel air de¯ector for downhill mountain
bikes. Downhill mountain bike tyres are becoming larger with more aggressive tread
patterns in order to cope with the increasingly severe conditions of modern downhill
courses, which in turn leads to increased aerodynamic drag forces caused by air ¯ow, and
consequently a reduction in speed over a downhill course. A front wheel air de¯ector
was designed and constructed from ®breglass that, when ®tted to the bike, could be
adjusted to give different clearance settings between the tyre and de¯ector inner surface
to cope with varying mud conditions. Various designs of downhill bike tyres, a cross-
country tyre and a slick racing tyre were selected and tested in a wind tunnel in the range
0±72 km h±1 to determine their benchmark aerodynamic drag characteristics with no
de¯ector ®tted. Further wind tunnel tests were then conducted for one of the tyres with
the de¯ector in various clearance positions. The trend of increasing aerodynamic drag
with current downhill tyre development was veri®ed, while the drag tests indicated that
the power input to the wheel can be reduced by up to 50% with the de¯ector ®tted in its
smallest ®rst clearance setting. This represents a drag reduction of up to 10%, and a
reduced drag coef®cient of almost 40%.

Keywords: air de¯ector, downhill, tyre, drag

Nomenclature
A projected area of wheel (ˆ Wd ) (m2)
CD drag coef®cient {ˆ D/(0.5qv2A)}
D measured drag force (N)
d tyre diameter (m)
I motor current (A)
P power absorbed by rotating wheel (ˆ IV ) (W)
Re Reynolds number (ˆ vd/m)
V motor voltage (V)
v velocity of free stream (m s±1)
W tyre width (m)
m kinematic viscosity of free stream (m2 s±1)
q density of free stream (kg m±3)

Introduction
Correspondence address:
A.T. Sayers, Associate Professor, Department of Mechanical Downhill mountain biking is a rapidly growing
Engineering, University of Cape Town, Rondebosch, 7700, sport that is becoming more popular and compet-
South Africa. E-mail: asayers@eng.uct.ac.za itive every year. Due to the specialization of

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 63

Aerodynamic drag of mountain bike tyres · R. J. Sunter and A. T. Sayers

downhill cycling from traditional cycling, moun- compared with a stationary wheel at the same wind
tain bikes have had to adapt to cope with increas- speed.
ingly rigorous racecourses. The evolution of Zdravkovich (1992) carried out experiments on a
downhill bicycles began a decade ago, with the separate bicycle frame and stationary wheel and
introduction of the dual suspension bicycle. Today, showed that the wheel drag was extremely sensitive
however, these dual suspension systems are reach- to the yaw angle. He also concluded that the shape
ing a technological plateau, and improvements in factor signi®cantly in¯uenced the drag coef®cient
performance need to be sought in other areas. 1 of the wheel. Tew and Sayers (1998) and Greenwell
The total drag force on a cyclist is made up from et al. (1995) determined the drag and side force
the drag of the rider himself, including any coef®cients for different wheel designs at a range of
streamlined apparel such as a helmet that he might yaw angles to the incident wind.
be wearing, the drag of the cycle frame and the drag From the above discourse, it is evident that
of the wheels. Only the latter can be independently aerodynamic research has been a priority in all
investigated with any real degree of accuracy, since advanced road bicycle design, yet very little has
the rider himself causes severe ¯ow interference been incorporated into mountain biking. Beneke
between himself and the frame whilst peddling. et al. (1989) describe the type of tyres used by all
Kyle & Burke (1984) performed experiments to terrain bikes (ATBs) or mountain bikes. Types for
determine the overall drag force of a cycle/rider off-road riding have a deep tread pattern that
combination, the drag forces attributable to indi- reaches beyond the tyre edges, so they grip well on
vidual items of clothing and equipment, and the loose surfaces. They come in different widths ± as
effects of slipstreaming and passing traf®c. They wide as 55 mm for off-road riding and narrower for
concluded inter alia: hard surfaces. The `knobblier' the tread pattern, the
better traction it has on sandy or stony surfaces but
(1) The wind resistance constitutes over 90% of
the more tiring it is to ride on hard roads. Some
the force retarding the cyclist at racing speeds
tyres have a dual-purpose design: they are wide and
above 32 km h±1.
`knobbly' enough to grip well on gravel, but they
(2) At speeds above 12.9 km h±1, wind resistance
also have a raised central ridge so that the wheels
exceeds the rolling friction.
roll easily on a hard road.
(3) In a 4000-m race, an 0.09-N reduction in drag
Parallel with the development of dual suspension
will reduce the time taken by about 0.3 s; a
systems, new advanced tyre designs are introduced
4.4-N reduction of drag will reduce the time
into the market every year, the introduction of
taken by 13 s.
oversized tyres being one of the latest downhill
(4) A pure tail or head wind will speed up or slow
mountain biking innovations. Tyre traction is
down the rider slightly more than half the wind
essential in downhill racing, and has become
speed. For example, if the cyclist is capable of
important due to the increasing technical standards
travelling at 32 km h±1, a 26 km h±1 head-wind
of downhill racetracks. With more sharp corners
will slow him down to about 22.5 km h±1.
and large rocks to negotiate, tyre design can make
(5) The rolling resistance of bicycle tyres is nearly
an enormous difference to the rider's performance.
constant at normal cycling speeds.
Consequently tyres are becoming larger with more
(6) Wind tunnel tests conducted at 6.7 m s±1
deeply designed tread patterns to give riders greater
(15 mph) gave the following results: (i) drag
traction around these corners and to absorb larger
on an unclad spoked wheel was 0.52 N;
impacts from the track. This obviously leads to a
(ii) drag on a clad wheel was 0.08 N.
larger projected frontal surface area of tyre facing
Kyle (1985) further reported that a wheel rota- the wind, and therefore an increased air resistance.
ting with a peripheral velocity the same as the wind In addition, when the cyclist is riding at a given
speed displayed about a 30% increase in drag when speed, the tangential velocity of the top of the tyre

64 Sports Engineering (2001) 4, 63±73 · Ó 2001 Blackwell Science Ltd

 R. J. Sunter and A. T. Sayers · Aerodynamic drag of mountain bike tyres

is twice the bicycle's linear velocity and the tread Shock absorbers vary in design depending on the
knobs are therefore colliding with the air at twice manufacturer, which makes it dif®cult to ®nd a
the speed of the bicycle, further increasing the total standard shock to which the de¯ector could be
air resistance. An air de¯ector covering a portion mounted. There are currently two types of front
of the tyre periphery would cause the air to ¯ow shocks on the market, inverted shocks and normal
over the front wheel without coming into contact shocks. It was not possible to use inverted shocks
with the tread knobs. This streamlining effect since there is no fork brace to which the de¯ector
would reduce the overall drag of the bicycle and could be attached while a further restriction was that
allow for higher speeds to be attained when riding, the entire shock body moves separately from the
while maintaining all the advantages of the over- wheel so that the clearance between the de¯ector and
sized tyre's extra traction. tyre could not be kept constant. Normal shocks
This paper proposes the concept of a front wheel could be used, however, since they have fork braces
air de¯ector to reduce the aerodynamic drag of that hold the shock legs together and the shock body
the front wheels of mountain bikes. It describes moves with the wheel. If the de¯ector was attached to
experiments that quantify the aerodynamic drag of the fork brace of the shock body, both would move
different designs of tyre, and the effect on the drag together, that is as one body, thus ensuring that a
of one tyre when a front wheel de¯ector is ®tted. constant clearance is maintained between the inside
of the de¯ector and the tyre periphery. This method
of de¯ector attachment and the individual assembly
Air de¯ector design criteria
components are shown in Fig. 2.
The function of the air de¯ector was to reduce the With the de¯ector attached to the fork brace,
front wheel air resistance as much as possible and normal cantilever or vee brakes could not be used,
Fig. 1 shows the designed de¯ector attached to a since they are usually attached to the fork brace.
mountain bike in the ®rst tyre clearance position. This would not pose a problem, because the
The factors that had to be taken into account when majority of downhill racing bicycles are sold with
designing the de¯ector were as follows. disc brakes as a standard accessory.

Method of attachment Tyre clearance

The method of attachment of the air de¯ector was Since environmental and track conditions vary
compromised by the type of shock absorber used. across different downhill racecourses, it was
important to bear these factors in mind when
considering any new design. Mud is often encoun-
tered on these downhill courses and therefore the
de¯ector/tyre clearance would have to be adjustable
in order to accommodate different amounts of mud
on the course to prevent mud blockage between the
wheel and the de¯ector. Thus four different
clearance settings were decided upon to accommo-
date varying degrees of muddy conditions with the
proviso that if there were excessive quantities of
mud on the course, then it would be better not to
use the de¯ector. The effect of large quantities of
mud collecting on the inside of the de¯ector would
Figure 1 De¯ector attached to bike in the ®rst (10 mm) result in an increased steering moment of inertia
clearance position. about the steering axis, thereby causing the steering

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 65

Aerodynamic drag of mountain bike tyres · R. J. Sunter and A. T. Sayers

Figure 2 De¯ector assembly drawing.

to become sluggish and unresponsive (as well as angle of 127 degrees. When the de¯ector was set at
adding signi®cant weight to the bicycle), which the 40 mm clearance position, the side edges of the
would lose the rider far more time in a race than de¯ector just overlapped the metal rim of the
any gain through improved aerodynamics. The wheel. This overlap became greater as the de¯ec-
four de¯ector clearance settings were 10 mm, tor/tyre clearance decreased. The section pro®le of
20 mm, 30 mm and 40 mm, the clearance being the inner surface of the de¯ector was concentric
maintained by changing the spacer (Item 6) and with that of the tyre (a) in Fig. 4.
adjusting the position of the support struts (Item 2).
Strength
De¯ector pro®le
The de¯ector was moulded in ®breglass, and
The external shape of the de¯ector and ultimately although this is heavier than lightweight carbon
its streamlining performance would be affected by ®bre, kevlar or thermoplastics, which would in
three variables, namely the tyre cover angle, the practice be the materials of preference to give a
de¯ector depth over the wheel and the section high strength to weight ratio, for wind tunnel
pro®le over the tyre. The tyre cover angle was testing purposes the weight of the de¯ector was not
dependent on how far around the tyre the de¯ector an issue.
could extend before any improvement in drag The external forces on the shocks during any
performance became negligible, and also what downhill ride are extreme and rigorous, leading to
penalty would be suffered by increasing the overall severe vibrations, which are absorbed by the
weight of the bike. In the ®nal design, the leading dampers in the shock. The de¯ector would be
edge of the de¯ector was set at 15 degrees below exposed directly to these vibrations and therefore
the wheel's axis of rotation, with a total tyre cover needed to be suf®ciently strong and stiff to with-

66 Sports Engineering (2001) 4, 63±73 · Ó 2001 Blackwell Science Ltd

 R. J. Sunter and A. T. Sayers · Aerodynamic drag of mountain bike tyres

stand them. A rubber vibration absorber was

therefore placed between the end face of the
de¯ector and the fork brace.
To add extra lateral strength and stiffness to the
de¯ector, cross struts extending from the wheel
hub to the front edge of the de¯ector were ®tted.
These cross struts also minimized ¯exure of the
de¯ector while riding.

Weight
In downhill mountain biking, low weight is not a
top priority, although it must still be seriously
considered. The de¯ector would not only add to
the overall weight of the bicycle but would also
increase the inertial mass of the steering system.
Responsive and accurate steering are critical in
downhill cycling, and any extra mass on the front Figure 3 De¯ector attached to forks.
wheel would affect performance dramatically. It
was therefore important to keep this mass to a Experimental apparatus
minimum to avoid sluggish or heavy steering.
A number of designs of downhill tyres shown in
Fig. 4 were tested for aerodynamic drag in a wind
Manufacture tunnel. The effect of the addition of the de¯ector
Ease of manufacture is crucial to the design of the with tyre (a) on the wheel was then investigated.
de¯ector since it dictates the amount of time spent
on the component construction and consequently
the expenses incurred during construction. The
design must be kept as simple as possible to reduce
the complexity of the moulding process. If the
de¯ector were to go into production, the simplest
design would have huge cost bene®ts in an
extremely competitive industry. It is also important
to consider whether the extra features that have
been designed into the component do in fact make
a signi®cant difference to its performance. Unnec-
essary complications should be avoided throughout
the design. Finally, considerations of aesthetic
appeal are important in the design. All components
on bicycles are these days designed to perform well,
and look good. A component's appearance has
signi®cant in¯uence on whether or not a product
Figure 4 The tyres showing their tread patterns: (a) Nokian
will be a marketing success. The de¯ector was
Gazzaloddi; (b) Michelin Wildgripper; (c) Hutchinson On The
therefore designed with a rugged and professional Rocks II; (d) Hutchinson Coyote; (e) IRC Kujo DH; (f)
appearance. The ®nal de¯ector design attached to Hutchinson Squale; (g) IRC Mythos XC; (h) Continental
its forks is illustrated in Fig. 3. Avenue Slick.

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 67

Aerodynamic drag of mountain bike tyres · R. J. Sunter and A. T. Sayers

2 1999a and b) presents a detailed comparison of

Wind tunnel
each tyre in terms of their dimensions, tread
The wind tunnel used was of the open jet return description and the course conditions for which
circuit type powered by a 12-bladed variable pitch they were designed. It is apparent that all the tyres
axial ¯ow fan. It had a test cross-sectional area of have unique tread patterns, designed speci®cally to
870 mm ´ 580 mm, with corner ®llets and a work- give maximum traction in various speci®ed condi-
ing length of 1.6 m. The velocity variation across tions. The tyres were ®tted to a standard mountain
the test section was 1% and the turbulence inten- bike front wheel with a rim diameter of 57.6 mm.
sity of the free stream was 0.4%. The maximum
velocity attainable in the test section was 36 m s±1
Tyre test rig
while the area contraction ratio between the
settling chamber and the area of the test section The wheel was supported horizontally in the wind
was 10:1. The tunnel velocity was calibrated using a tunnel air stream by a vertical aluminium support
pitot-static tube and inclined alcohol manometer, sting connected to a specially designed fork ¯ange
to an accuracy of 0.2% at the maximum test speed (Fig. 5). The fork ¯ange supported both the wheel
used in the experiments. A velocity variation of on its axle, and a variable speed d.c. motor. The
2.6% occurred over the 1.6 m jet length. This was motor rotated an aluminium disc with a 66.7 mm
small enough to be considered negligible, partic- diameter rubber O-ring around its periphery.
ularly since at the position at which the wheel was Clamping of the wheel to the fork ¯ange was
located for testing, the expected error in the achieved with a quick release skewer, and the motor
velocity was only 0.8%. Blockage in the open jet then bolted into place, the O-ring friction drive
test section was also negligible (Pope 1966; Sayers being in contact with the metal rim of the wheel.
& Ball 1983). A three component (lift, drag and A variable power supply was used to set the
pitch) proprietary strain gauge wind tunnel balance rotational speed of the wheel and this speed was
was used to measure the aerodynamic drag force measured by a digital re¯ecting optical tachometer.
with an uncertainty of 0.02 N.
De¯ector test rig
Tyre designs tested
This apparatus is shown in Fig. 6, with the wind
Eight tyre designs that have been placed on the tunnel outlet in the background. The de¯ector was
market at various times since 1995 were tested for mounted on the front shock absorber legs as it
their aerodynamic characteristics. The tyres are would be on the bicycle and the assembly then
shown in Fig. 4, while Table 1 (Cunningham mounted again on the forked ¯ange. The friction

Table 1 Details of tyres tested

Diameter Width Pro®le Tread Course

Tyre Year Mass kg d (mm) W (mm) description pattern conditions
(a) Nokian 1999 1.55 709.8 72.68 Gently rounded Large knobs Downhill: severe
(b) Michelin 1996 0.76 668.5 52.70 Rounded Moderate Downhill: average
(c) Hutch OTR 1997 0.9 665.3 53.18 Flattish Aggressive Downhill: all
(d) Hutch Coy. 1996 0.86 668.5 52.38 Flattish Reinforced Downhill: hardpacked
(e) IRC Kujo 1998 1.04 671.6 55.46 Gently rounded Moderate Downhill: dry/loose
(f) Hutch Sq. 1998 1.075 673.2 55.94 Square Aggressive Downhill: muddy
(g) IRC Mythos 1998 0.638 655.7 49.08 Rounded Small knobs Cross country
(h) Cont. Slick 1995 0.59 646.2 41.10 Very rounded Smooth no tread Road

(a) Nokian Gazzaloddi, (b) Michelin Wildgripper, (c) Hutchinson On The Rocks II, (d) Hutchinson Coyote, (e) IRC Kujo DH,
(f) Hutchinson Squale, (g) IRC Mythos XC, (h) Continental Avenue Slick.

68 Sports Engineering (2001) 4, 63±73 · Ó 2001 Blackwell Science Ltd

 R. J. Sunter and A. T. Sayers · Aerodynamic drag of mountain bike tyres

A tyre was placed onto the rim of the wheel,

in¯ated to its recommended pressure and statically
balanced. The balancing was important to avoid
any unnecessary vibration that could have affected
the results. It was then clamped onto the apparatus
with the quick release skewer. The drive motor
speed was set to rotate the wheel at a speed
corresponding to the linear velocity of the bicycle,
and the wind tunnel air speed was then set at the
same value. The motor voltage and current were
then recorded, along with the aerodynamic drag
force on the wheel assembly. This was repeated in
1 m s±1 increments for bicycle and wind speeds in
the range 1±20 m s±1 (72 km h±1). A separate drag
Figure 5 Tyre test rig. test was also conducted for the apparatus alone,
without the wheel and at the same wind tunnel air
speeds as previous. These drag values were sub-
tracted from the recorded tyre drags to give a more
accurate representation of the tyre drag without the
support apparatus.

De¯ector effectiveness
The air de¯ector effectiveness experiments were
conducted with the tyre (a) ®tted to the wheel and
the air de¯ector set in its four clearance settings of
10 mm, 20 mm, 30 mm and 40 mm above the tyre.
The clearance settings were predetermined in the
design phase by means of spacers and alternate
bolting positions on the fork brace. The previous
Figure 6 De¯ector test rig. experimental procedure was followed for measure-
ment of the wheel power input and drag forces at
similar wheel and wind tunnel speeds.
drive motor was similarly used to rotate the wheel.
A larger diameter support sting was required to
Results and discussion
accommodate the extra weight and a centre of
gravity ¯ange used to balance the apparatus. The processed data are presented in Figs 7 and 8 in
terms of absolute power absorbed by the wheels,
the absolute drag on the wheels and the drag
Experimental method
coef®cient for the wheel/de¯ector assembly, all as
functions of wind speed or Reynolds number.
Tyre drag
The tyres were tested independently with the same
Tyre data
apparatus and wind tunnel set-up, the atmospheric
conditions remaining constant throughout the The graphical representation of the motor power-
experiments. input results are shown in Fig. 7(a). The graph

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 69

Aerodynamic drag of mountain bike tyres · R. J. Sunter and A. T. Sayers

Figure 8 De¯ector performance. (a) Absorbed power reduction;

(b) percentage drag reduction; (c) drag comparison for tyre
Figure 7 Individual tyre performance. (a) Power absorbed by
alone and with de¯ector ®tted.
tyres; (b) drag force of the tyres; (c) tyre drag coef®cients.

pared, but tyre (a) still absorbs 50 percent more

illustrates the signi®cant impact the tyre pro®le has power than tyre (g). The corresponding drag forces
on the power required to maintain the wheel at the of Fig. 7(b) show that at 20 m s±1 tyre (a) experi-
same speed. It is veri®ed that the more modern tyre ences twice the aerodynamic drag of the slick road
designs with wider pro®les and aggressive tread tyre (h) and one and a half times that of tyre (g).
patterns such as (c) and (f) require more power Figure 7(c) shows that for each tyre the drag
input to maintain a given speed of cycling. At the coef®cient is approximately constant after a Rey-
top wind speed of 20 m s±1 the (a) tyre absorbs nolds number of 3 ´ 105 is exceeded. As might be
approximately two and a half times as much power expected, the slick tyre (h) has the lowest CD due
as the slick road tyre (h). This difference decreases primarily to its aerodynamic shape. The highest
signi®cantly if only the knobbed tyres are com- CD was displayed by tyre (c) followed closely by

70 Sports Engineering (2001) 4, 63±73 · Ó 2001 Blackwell Science Ltd

 R. J. Sunter and A. T. Sayers · Aerodynamic drag of mountain bike tyres

tyre (d). Although tyre (a) absorbs the greatest for CD to increase as the de¯ector clearance
power, its CD is 31% lower than tyre (c). This is increases.
due to the much greater width of tyre (a) as shown
in Table 1.
Application of results
Beneke et al. (1989) states that the aerodynamic
De¯ector data
drag of a rider on a racing bike in a fully crouched
Figure 8(a) shows that a reduction in the power position at 56 km h±1 is approximately 32 N. He
absorbed occurs at each de¯ector clearance setting. also gives typical frontal projected areas of a racing
Below a speed of about 9 m s±1, the largest bike with rider and a downhill bike with rider as
clearance gives the lowest percentage reduction, 0.36 m2 and 0.41 m2, respectively. Thus, if similar
but after 9 m s±1 the power reduction increases drag coef®cients are assumed for each bike, at
with the clearance, having a value of 45% at 56 km h±1 the total aerodynamic drag on the
20 m s±1. As the clearance increases, so also does downhill bike and rider will be (32 ´ 0.41/
the quantity of air between the tyre and the 0.36) ˆ 36 N.
de¯ector. There is thus a greater quantity of air From Fig. 8(a, b) at 56 km h±1 (15.5 m s±1), the
for the tread knobs to disturb, giving rise to greater drags on tyre (a) alone and tyre (a) with the
friction and therefore a greater power input from de¯ector ®tted in the 10 mm clearance position are
the motor being required to maintain a given wheel 7.9 N and 7.12 N, respectively. This represents a
speed. The percentage reduction in drag shown in decrease in aerodynamic drag of 0.78 N that when
Fig. 8(b) shows similar trends for each de¯ector divided by the total aerodynamic drag of 36 N
clearance, although the drag reduction only gives a percentage improvement of total drag of
becomes effective after 9 m s±1 for the 20 mm, 2.17% with the air de¯ector in place.
30 mm and 40 mm clearances. For the 10 mm Beneke et al. (1989) also state that the rolling
clearance, the drag reduction is the greatest (i.e. the resistance of two `large knobbed' tyres is of the order
drag is lowest), and is virtually constant at about of 10 N, and for the purposes of this analysis can be
10% when compared with tyre (a) alone. However, assumed to be the rolling resistance of tyre (a). Thus
this drag reduction diminishes (i.e. drag increases) the overall resistance to motion is (36 + 10) ˆ 46 N,
as the de¯ector clearance increases. Thus the and of that, the amount due to aerodynamic drag is
de¯ector reduces the drag of the front tyre, which (36 ´ 100%/46) ˆ 78%. If bearing friction is con-
would also reduce the total drag of the bike by a sidered negligible, the percentage decrease in the
certain percentage. The drag coef®cients for each overall resistance of the downhill rider with the
de¯ector clearance are shown in Fig. 8(c) where de¯ector ®tted is (78 ´ 0.0217) ˆ 1.7%.
they are seen to be approximately the same despite The power input to the bike is proportional to
the variation in the percentage drag reduction of the product of the velocity and overall resistance,
Fig. 8(b). The wheel with the de¯ector ®tted is and if the power input from the rider for each bike
clearly aerodynamically superior to the wheel with is the same, the velocity is proportional to the
the tyre (a) alone at all clearance positions. Con- overall resistance. Taking the average time for a
comitant with the increase in drag as the de¯ector typical downhill race to be about 5 min (300 s),
clearance is increased, the frontal projected area then the de¯ector decreases the overall resistance
is also increasing, and since CD is proportional to (and velocity) by 1.7% resulting in a time reduction
D/A, the drag coef®cient changes only marginally. of 5.1 s. In such a highly competitive sport, this
Thus, while the drag itself is signi®cantly lower time reduction is highly signi®cant.
with a low de¯ector clearance, the overall effect of The above analysis has been performed for
de¯ector clearance itself on CD is small, but close the 10 mm de¯ector clearance and Table 2 gives
examination of Fig. 8(c) does reveals a tendency the results for other clearance settings. Clearly,

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 71

Aerodynamic drag of mountain bike tyres · R. J. Sunter and A. T. Sayers

Table 2 Performance improvements with the de¯ector in (a) at maximum power absorbed, these individual
various settings at 56 km h)1 errors translated into a error in power of 0.01%,
Aerodynamic drag Total drag Time
which is negligible.
Clearance improvement improvement reduction
setting percentage percentage over 300 s
Conclusions
10 mm 2.17 1.70 5.09
20 mm 1.92 1.51 4.53 Experiments on different designs of downhill
30 mm 0.89 0.69 2.07
40 mm 1.11 0.87 2.61
mountain bike tyres revealed signi®cant data per-
taining to the aerodynamic performance of the
individual tyres. The major tyre parameters
for the 40 mm clearance position (very muddy responsible for these differences are tyre diameter,
conditions), the improvement in descent time is width, pro®le, tread knobs and tread pattern. The
approximately halved. trend of development over the last decade has seen
diameters, widths and knobs increase in size while
the tread patterns have become more aggressive,
Error analysis
and the pro®les ¯atter. The overall result has been
An error analysis was performed on the data by the a general reduction in aerodynamic ef®ciency
method of Kline & McClintock (1953) for single resulting in an increase in power absorbed by the
sample experiments. Of the independent variables wheel. Thus downhill mountain bikes are becom-
required for the calculation of CD, the density and ing increasingly more air resistant.
area were determined through the measurement of The ®tting of an air de¯ector over the front
pressure, temperature, tyre diameter and width wheel tyre (a) has been shown to reduce the
using standard high sensitivity instruments. The aerodynamic drag of the front wheel by up to
drag and velocity contributed varying uncertainties 10%, resulting in a projected time gain of 5 s on a
depending on the wind tunnel velocity. As the wind typical 300 s descent course. Since the top 10 riders
tunnel velocity decreases towards zero, then, since in international championships are ®nishing within
the uncertainty in its measurement remains con- 2±3 s of the winner, it is envisaged that air
stant, the uncertainties in CD tend towards in®nity. de¯ectors will ultimately become an important
The magnitudes of the uncertainties of pressure, appendage in optimizing a downhill mountain
temperature, diameter and width were of the order bike's performance.
10±5, while those for drag and velocity were of the
order 10±3. Since the total uncertainty is given by
References
the square root of the sum of the squares of the
individual variable uncertainties, only the uncer- Beneke, M., Beneke, G., Noakes, T. & Reynolds, M. (1989)
tainties in drag and velocity measurements were of The Lore of Cycling. Oxford University Press, South Africa.
signi®cance. Cunningham, R. (1999a) Complete guide to mountain bike
tyres. In: Mountain Bike Action (ed. McIlvain, J.),
In Fig. 7(b), the uncertainty in the power
pp. 112±121. Hi Torque Publications, California.
absorbed by the wheel was 0.01%. The direct drag
Cunningham, R. (1999b) Downhill tyre shootout.
force at best, at 20 m s±1 (Fig. 7b), had an uncer- In: Mountain Bike Action (ed. McIlvain, J.), pp. 108±127.
tainty of 0.16% whilst at worst, at the same air Hi Torque Publications, California.
velocity, the uncertainty was 0.34%, both being Greenwell, D.I., Wood, N.J., Bridge, E.K.L. & Addy, R.J.
negligible. In Fig. 7(c), the uncertainty at best in (1995) Aerodynamic characteristics of low-drag bicycle
the drag coef®cient was 0.45% for tyre (c) at wheels. Aeronautical Journal, March, 109±120.
20 m s±1, and at worst for tyre (c) it was 0.9%. Kline, S.J. & McClintock, F.A. (1953) Describing uncer-
tainties in single sample experiments. Mechanical
The error in the motor voltage and current
Engineering, 75, 1, 3±8.
readings were 0.5% and 1%, respectively. For tyre

72 Sports Engineering (2001) 4, 63±73 · Ó 2001 Blackwell Science Ltd

 R. J. Sunter and A. T. Sayers · Aerodynamic drag of mountain bike tyres

Kyle, C.R. (1985) Aerodynamic wheels. Bicycling, Decem- tunnel. Proceedings of the Institution of Mechanical Engineers,
ber, 121±124. 197C, 259±263.
Kyle, C.R. & Burke, E. (1984) Improving the racing bicycle. Tew, G.S. & Sayers, A.T. (1999) Aerodynamics of yawed
Mechanical Engineering, 106, 34±45. racing cycle wheels. Journal of Wind Engineering and
3 Pope, A. & Harper, J. (1966). Wind tunnel boundary Industrial Aerodynamics, 82, 204±222.
corrections. Low-Speed Wind Tunnel Testing. John Wiley Zdravkovich, M.M. (1992) Aerodynamics of bicycle wheel
& Sons Inc, New York, NY, USA. and frame. Journal of Wind Engineering and Industrial
Sayers, A.T. & Ball, D.R. (1983) Blockage corrections for 5 Aerodynamics, 40, 55±70.
rectangular ¯at plates mounted in an open jet wind

Ó 2001 Blackwell Science Ltd · Sports Engineering (2001) 4, 63±73 73