[an error occurred while processing this directive] Aero Wheels Under Scrutiny, by François Grignon


Aero Wheels Under Scrutiny

We round up seven of the finest for a battery of lab tests

by François Grignon

Copied from the online version,
with kind permission from the author and
Club Cycliste Beaconsfield Cycling Club
Montréal, Québec, Canada.
Proofread and with added comment sby John Allen.

When Francesco Moser took the track on the nineteenth day of January 1984, people laughed at his funny-looking wheels. One hour later, the laughter had stopped. For the Italian star had shattered the hour record, a record that had held for ten years. Many thought the record would never be equaled because it was held by a man with no equals, the great Eddy Merckx. Moser broke the 50 km/hr barrier and opened the way to scientific cycling.

The most notable improvement brought to equipment by the Moser team was disc wheels. Wind tunnel testing had demonstrated a clear advantage in streamlining the blunt shape of the wheels. And Moser was determined to use every trick of the arsenal to his advantage. The disk wheel became a necessity for every successive hour record attempt. And its use trickled down from the covered velodrome world of hour records to time trialing on the road. There, it showed one serious shortcoming: by presenting such a large uninterrupted surface to a sidewind, it made the bike very hard to ride in windy conditions.

Enter the "baton" wheel. By replacing the full disk filling the center of the wheel by a few large spokes of a suitably streamlined cross section, almost the same drag reduction could be obtained, with a much reduced sensitivity to side winds. In the early nineties, a baton wheel in the front with a disk in the back became the hot setup for time trialing.

For all their success, baton wheels suffered some shortcomings. They were heavy, expensive and unrepairable. And their very design did not lend itself to mass production. This encouraged some small manufacturers to experiment with hybrid designs. Zipp was one of the pioneers of the current trend towards deep-section rims held by a reduced number of spokes to a conventional hub.

The deep-section rim aero wheel is not fundamentally different in design from the conventional spoked wheel. Contrary to disk or baton wheels, its rim is held in compression by the tension of the spokes. It can be trued. It can be repaired. It achieves drag reduction in two ways. First, the deep rim is streamlined to act as a fairing behind the tire, allowing the airflow to reattach smoothly without creating as much turbulence as the blunt shape of a conventional square rim. Secondly, the deep rim being much stiffer than a conventional rim in bending, it can be held by a reduced number of spokes. And each spoke of a rotating wheel is subjected to a drag force that opposes its motion.

Deep-section rim aero wheels achieved large-scale public awareness when Campagnolo introduced its Shamal series in 1993. At first, they were used only as time-trial equipment. But in 1994, the Italian Gewiss team raced on Shamals in every road event. And the team was incredibly successful, scoring victories in Milan-San Remo (Furlan), Flèche Wallone (Argentin), Liège-Bastogne-Liège (Berzin) and the Giro d'Italia (Berzin). A high-performance everyday road wheel was born.

[However, the Shamal wheel was prone to rim failure due to the small number of spokes and "pincushioning" of the rim with too little material around the spoke holes -- see photo -- John Allen]

Imitation is a measure of success. Other manufacturers have joined the fray and the deep-section rim aero wheel is now a common sight at every race, even in the amateur ranks.

But at roughly a thousand dollars for the pair, an amateur racer is entitled to ask how much performance gain do aero wheels procure? And, besides price, are there other disadvantages?

In an attempt to answer these fundamental questions, we rounded up seven front wheels that were evaluated in laboratory tests.

Our contestants included two conventional wheels, one of which had 32 double-butted spokes and the other, 36 straight 14-gauge spokes; one baton wheel, the Specialized Tri-Spoke; and four deep-section rim wheels: Campagnolo Shamal 12-spoke, Mavic Cosmic, Cane Creek Crono and Zipp 540, the last three having 18 spokes. Detailed specifications are provided in the table below.

We did not have access to a wind tunnel to make translational drag measurements. We relied on other published tests for this figure. But we did evaluate the drag torque of the stationary wheels rotating about their axles by means of timed deceleration runs. These measurements were taken at two velocities, enabling us to calculate the bearing friction component and the aerodynamic component of the resisting torque on the spinning wheel.

When it comes to aerodynamics, it is often said that looks can be deceiving. But not this time. The most aero looking wheel, the Specialized Tri-Spoke, posted the lowest figure for aerodynamic torque at 40 kph, only 0.044 N-m. The Shamal and Cosmic followed not too far behind, at 0.047 and 0.049 N-m, respectively. The Zipp 540 (0.068 N-m), despite having the deepest rim and bladed spokes, could not even beat the round-spoked Cane Creek newcomer (0.060 N-m) and placed last of the aero wheels. Both conventional wheels trailed far behind. The 36-spoke wheel finished last, posting a drag figure 225 % that of the winner.







HPW Shamal 12

12 straight-pull, bladed 3.0 x 1. mm, brass nipples

Campagnolo sealed cup-and-cone

Alu, 40 mm deep, tubular, powder-coated braking surface

Vittoria Corsa CX tubular

Zipp 540

18 straight-pull, bladed 3.3 x 1.2 mm, brass nipples

Shimano 600 sealed cup-and-cone

Carbon- fiber, 58 mm deep, clincher, alu braking surface

Michelin Hi-Lite Bi-Synergic clincher

Mavic Cosmic Expert

18 straight-pull,
bladed 2.75 x 1.5 mm, brass nipples

Sealed cartridge

Alu, 30 mm deep, clincher, machined braking surface

Michelin Hi-Lite Bi-Synergic clincher

Cane Creek Crono

18 straight pull, round, 2.0 mm dia.,
brass nipples

Sealed cartridge

Alu, 32 mm deep, clincher, machined braking surface

Michelin Hi-Lite Bi-Synergic clincher

Specialized Tri-Spoke

3 carbon-fiber, teardrop, 70.6 x 19.4 mm


Carbon fiber, 54 mm deep, tubular, alu braking surface

Continental L.A. 255 tubular

Campagnolo 32 spokes

32 elbowed, round DB 2.0/1.8 mm dia., alu nipples

Campagnolo Chorus sealed cup-and-cone

Campagnolo L.A. 84, 12 mm box section with ferrules

Continental Sprinter 250 tubular

Mavic 36 spokes

36 elbowed, round 2.0 mm dia.,
alu nipples

Campagnolo Record cup-and-cone

Mavic GL 330,
12 mm box section with ferrules

Vittoria Corsa CX tubular

Bearing-friction measurements shattered the myth about Campagnolo's superior hubs. The worst bearing performance, by a wide margin, was posted by the Shamal (0.021 N-m). The Shamal's bearings are smooth. It is just the seals that are so goddam tight. And this was a fully broken-in wheel. Next worst was the Zipp (0.017 N-m). Then came the other sealed Campy hub, a Chorus unit (0.014 N-m). The brand new sealed cartridge bearings of the Cosmic and Cane Creek were star performers (0.007 and 0.006 N-m). So were the well-worn unsealed cup and cone bearings on the Specialized (0.005 N-m) and those of the 20-year-old Campagnolo Record hub around which is built our 36-spoke wheel (0.007 N-m).

In Campagnolo's defense, it must be mentioned that both sealed Campy hubs feature a grease injection hole, and that both units had received a generous serving of the sticky stuff prior to winter hibernation. Neither wheel had been ridden since.

[However, bearing friction when under the load of the bicycle, rider and chain tension is higher, and was not measured for this article -- John Allen]

If we sum the effect of aerodynamic torque and friction torque to look at total drag torque at 40 kph, the Specialized wheel comes out a clear winner (0.049 N-m). The Shamal (0.068 N-m), penalized by sticky bearings, drops to fourth place behind the Cosmic (0.057 N-m) and the Cane Creek (0.066 N-m). The Zipp wheel, at 0.085 N-m, only managed to beat the conventional wheels. Our 32-spoke wheel scored 0.100 N-m, while the 36-spoke wheel finished last, at 0.105 N-m.

These drag figures, however, are pretty low. To illustrate this point, let us convert them into power consumption. We then find that the winning Specialized uses up only 1.5 Watts to rotate at 40 kph, while the last-place finisher requires 3.4 Watts. This difference is really nothing to kill for, given that the average club fast rider produces in the neighbourhood of 250 to 300 Watts.

"Bearing friction measurements shattered
the myth about Campagnolo's superior hubs"

Our test protocol produces results that underestimate the real-life rotational aerodynamic torque on the wheel by a factor of about two. The reason is simple. When spinning the stationary wheel, all spokes have an air velocity of, say, "V". On the road, when travelling forward, the spoke attached to the wheel at the point of contact with the ground has an air velocity of zero. It is, for an instant, stopped. The opposing spoke, attached to the top of the rim, travels through air at a velocity 2V. Since drag increases with the square of velocity, the total drag on the two spokes of the stationary wheel is V2 + V2= 2V2. For the wheel traveling on the road, the total drag for the two spokes is 02 + (2V)2 = 4V2, hence, greater than that of the stationary wheel by a factor of two.

It may be argued that this phenomenon holds only for the 12 o'clock and 6 o'clock spokes, but does not hold for the 3 o'clock and 9 o'clock spokes. Granted. But another phenomenon must also be accounted for. On the advancing wheel, the air that meets the upward-traveling 3 o'clock spoke has, just an instant ago, been disturbed by the downward-travelling 9 o'clock spoke. It has acquired a downward motion that increases the drag on the upward-traveling spoke. The stationary spinning wheel, in our test, spins in mostly undisturbed air. Worse yet, it probably creates a swirling motion in the air in its vicinity that diminishes the drag on the wheel. When air enters the plane of the spokes, it has already acquired a favorable velocity.

It must also be remembered that we are considering only rotational drag, completely neglecting translational drag arising from the forward velocity of the object through the air. Tests published on the latter indicate that aero wheels tend to be very close to each other in translational drag, requiring about 11 Watts at 40 kph to overcome it. Conventional spoked wheels fare much worse, requiring in the neighbourhood of 24 Watts to travel at the same forward speed.

Such quantum jumps are not unusual in aerodynamics. Streamlined bodies that let the airflow reattach smoothly behind them experience much less drag than bluff bodies to which the airflow cannot remained attached, creating a low pressure zone behind them that literally sucks them back. There is no doubt that the deep-section rim creates a streamlined shape behind the tire, whereas the conventional square rim is definitely a bluff body.

To conclude on this topic, even though our testing produced drag results that are, in absolute terms, much below those that can be expected in real life, we believe that it did reveal a useful relative ranking of the test specimens.

Other factors affecting performance are not so tricky to evaluate. Weight, for instance. Weight is a key performance factor in climbing. In this regard, conventional wheels still remain the top choice. A lightweight hub and a lightweight rim made the 36-spoke wheel the lightest in the group, at 1030 grams. It beat the 32-spoke wheel by 30 grams. The Shamal, thanks to its jewel like miniature hub and lowest spoke count of the group, is remarkably light at 1090 grams. Other aero wheels all weighed around 1200 grams, except the Specialized. It weighed a substantial 1360 grams.

We also evaluated the moment of inertia of the wheels. Moment of inertia is technical jargon for the "rotating mass" of an object. It is common knowledge among cyclists that weight reductions in rotating components are more beneficial, because, when accelerating, a rotating component not only has to gain a higher translational speed, but it must also be made to spin faster. This tendency to resist changes in rotational velocity is related to mass and to the square of its distance from the axis of rotation. Hence, it is expressed in kilograms times meters squared.

The values of moment of inertia produced a ranking similar to that of weight, with the notable exception of the Specialized. The Tri-Spoke carbon-fiber heavyweight carries most of its heft in a clumsy hub design. It fared better in moment of inertia, placing fourth, just behind the two conventional wheels and the Shamal.

A convenient way to appreciate moment of inertia is to convert its value into an equivalent mass. This equivalent mass is the mass of a non-rotating object that would pose the same resistance to acceleration. We have made the calculation and tabulated the values. This piece of information is of particular interest to those who intend to use the wheels for criterium racing. In criterium racing, reaccelerating corner after corner is what wears the riders down.

Finally, we measured both lateral and radial stiffness of the wheels. Lateral stiffness is a good thing. A laterally stiff wheel feels better in a sprint. Radial stiffness, however, is less desirable. Soft wheels provide a more comfortable ride.

[Comment from John Allen: softness of the tire is much greater than that of the rim and spoking.]





Moment of inertia


Equiv. total mass

Aero torque @ 40 km/h

Friction torque

Total torque @ 40 km/h

Lateral stiffness


Radial stiffness


Shamal 12 HPW









Zipp 540









Mavic Cosmic Expert









Cane Creek Crono

c (?)








Specialized tri-spoke









Campagnolo 32 spokes









Mavic 36-spoke










Four wheels that sprinters might want to avoid are the Shamal, Cosmic, Specialized and Zipp. Clustered around a value of 215 lb/in, they occupy the soft side of the lateral stiffness spectrum. At the other end of the spectrum, the 36-spoke wheel and the Cane Creek, the latter in virtue of a large diameter hub with widely spaced flanges and straight spokes, were clearly stiffer, at about 275 lb/in. The 32-spoke wheel fell in between, at 260 lb/in.

Aero wheels are not supposed to be comfortable, right? The deep rim makes them too stiff radially, right? Wrong. The softest wheels in the test were the Specialized and Cane Creek, at 8,400 lb/in and 8,800 lb/in, respectively. Then came the Shamal, at 12,600 lb/in. The Cosmic and the 32 spokes followed at 13,500 lb/in.

Thirty-six 14 gauge spokes sure make a wheel stiff, and so does a very deep carbon fiber rim as on the Zipp. These two tied for the rough-ride championship, at 20,200 lb/in.

[Again: the radial stiffness of a bicycle tire inflated to 100 PSI is on the order of 1000 lb/in -- John Allen]

There is no doubt that the difference in ride quality between a 20,000 lb/in wheel and a 13000 lb/in one is immediately apparent to a 70 kg rider. Lighter guys should avoid stiff wheels

[There may be a difference in feel of the wheel, but it is not due to differing radial stiffness -- John Allen]

.In the end, are the magic wheels worth a thousand dollars? Look at it this way: there are many ways to spend a thousand dollars on improving your bike that will not result in a comparable performance gain. And the price paid is quickly forgotten after the check has cleared. But the joy of effortless high speed cruising remains every time you put the wheels on. Furthermore, new-generation aero wheels can also climb, hold their own in a criterium and provide a comfortable ride. Aero wheels work in the real world. They are no longer reserved for record attempts.

Copyright © 1998, François Grignon

The Home Lab, by François Grignon

Home back to Damon Rinard's Bicycle Tech Page

[an error occurred while processing this directive]

Harris Cyclery Navigational Hub Page

Articles by Sheldon Brown and others
Beginners Brakes Commuting
Fixed Gear
Frames Gears &
Tandems Touring What's
Wheels Sheldon