More Articles by Jobst Brandt
Next: Slope Wind, the Invisible Enemy
Previous: Cycling Myths
From: Jobst Brandt
Date: May 11, 2001
Descending on mountain roads, bicycles can reach speeds that are more common on motorcycles. Speeds that are otherwise not attainable, or at least not continuously. Criterium racing also presents this challenge, but not as intensely. Unlike a motorcycle, the bicycle is lighter than the rider and power cannot be applied when banked over when cornering hard. Because narrow bicycle tires inflated hard have little traction margin, a slip on pavement is usually unrecoverable.
Riders have claimed they can slide a bicycle on dry pavement in curves to achieve greater cornering speed, as in drifting through a turn. A drift, in contrast to a slide, means that both wheels slip, which is even more difficult. This notion may come from observing motorcycles, that can cause a rear wheel slide by applying power when banked over. Besides, when questioned about how this is done, the proponent says that the ability was observed, done by others.
A bicycle can be pedaled only at lean angles far less than the maximum without grounding a pedal, so hard cornering is always done coasting, hence, there is no power in hard cornering. Although bicycles with high ground clearance have been built, they showed only that pedaling imbalance has such a disturbing influence on traction, that pedaling at a greater lean angle than that of a standard road racing bicycles has no benefit. That is why road bicycles are built the way they are, no higher than is useful.
That bicycle tires have no margin for recovering a slip at maximum lean angle, has been tested in lean-slip tests on roads and testing machines. For smooth tires on pavement, slipout occurs at slightly less than 45 degrees from the road surface and is both precipitous and unrecoverable. Although knobby tires have a less sudden slipout and can be drifted around curves, they begin to side-slip at a more upright angle as their tread fingers walk rather than slip. For this reason, knobby tires cannot achieve lean angles of smooth tires and offer no cornering advantage on pavement.
Cornering requires estimating the required lean angle before reaching the apex of the turn where the angle with the road surface is the critical parameter rather the angle with the vertical, as is evident from banked curves. Lean angle is limited by the available traction that must be assessed from velocity and appearance of the surface. For good pavement, this angle is about 45 degrees, in the absence of oil, water, or smooth and slick spots. Therefore, a curve banked inward 10 degrees, allows a lean of up to at least 55 degrees from the vertical, while a crowned road with no banking, where the surface falls off about 10 degrees, would allow only up to 35 degrees.
Banked curves have a greater effect than just adding to the maximum lean angle, because with a steeper banking, more of the centripetal cornering force goes into increasing traction directly into the banking up to the point of a vertical wall where only the maximum G-forces limit what speed a bicyclists can attain. In contrast, an off banked curve makes cornering progressively more difficult until the bicycle will slip even at zero speed. This effect is more naturally apparent to riders who exceeded these limits early in life and have added the experience to expected natural phenomena.
The skill of visualizing effects of speed, traction, braking, and curvature are complex, but is something humans and other creatures do regularly in self propulsion. The difficulty arises in adapting this to higher speeds. When running, we anticipate how fast and sharply to turn on a sidewalk, dirt track, or lawn, to avoid sliding. The method is the same on a bicycle although the consequences of error are more severe.
Cornering requires reflexes to dynamics that are easily developed in youth, while people who have not exercised this in a long time find they can no longer summon these skills. A single fall strongly reinforces doubt, so cautious practice is advisable if returning to bicycling after a long time.
Countersteer is a popular subject for people who belatedly discover or rediscover how to balance. What is not apparent, is that two wheeled vehicles can be controlled ONLY by countersteer, there is no other way. Unlike a car, a bicycle cannot be diverted from a straight path by steering the wheel to one side. The bicycle must first be leaned in that direction by steering it ever so slightly the other way. This is the means by which a broomstick is balanced on the palm of the hand or a bicycle on the road. The point of support is moved beneath the mass, in line with the combined forces of gravity and cornering, and it requires steering, counter and otherwise. It is so obvious that runners never mention it, although football, basketball, and ice hockey players conspicuously do it.
Once the basics of getting around a corner are developed, doing it fast involves careful use of the brakes. Besides knowing how steeply to lean in curves, understanding braking makes the difference between the average and the fast rider. When approaching a curve with good traction, the front brake can be used almost exclusively, because it is capable of slowing the bicycle so rapidly that nearly all weight transfers to the front wheel, at which point the rear brake is nearly useless. Once in the curve, more and more traction is used to resist lateral slip as the lean angle increases, but that does not mean the brakes cannot be used. When banked over, braking should be done with both brakes, because now neither wheel has much traction to spare and with lighter braking, weight transfers diminishes. A feel for how hard the front brake must be applied for rear wheel lift-off, can be developed at low speed.
Why brake in the turn? If all braking is done before the turn, speed will be slower than necessary before the apex. Anticipating maximum speed for the apex is difficult, and because the path is not a circular arc, speed must be trimmed all the way to that point. Fear of braking in curves usually comes from an incident of injudicious braking at a point where braking should have been done with a gentle touch to match the conditions.
Substantial weight transfer from the rear to the front wheel will occur with strong use of the front brake on good traction just before entering the curve. When traction is poor or the lean angle is great, deceleration cannot be large and therefore, weight transfer will be small, so light braking with both wheels is appropriate. If traction is miserable, only the rear brake should be used, because although a rear skid is recoverable, a front skid is generally not. An exception to this is in deep snow, where the front wheel can slide and function as a sled runner while being steered.
For braking in a curve, take the example of a rider cornering with good traction, leaning at 45 degrees, the equivalent of 1G centrifugal acceleration. Braking with 1/10g increases the traction demand by one half percent. The sum of cornering and braking vectors is the square root of the sum of their squares, SQRT(1^2+0.1^2)=1.005 or an increase of 0.005. In other words, there is room to brake substantially during maximum cornering. Because the lean angle changes as the square of the speed, braking can rapidly reduce the angle and allow even more braking. For this reason skilled racers nearly always apply both brakes into the apex of turns.
Beyond leaning and braking, suspension helps substantially in descending. For bicycles without built-in suspension, this is furnished by the legs. Standing up is not necessary on roads with fine ripples, just taking the weight off the pelvic bones is adequate. For rougher roads, enough clearance must be used so the saddle carries no weight. The reason for this is twofold. Vision will become blurred if the saddle is not unloaded, and traction will be compromised if the tires are not bearing with uniform force on the road while rolling over bumps. Ideally the tires should bear on the road at constant load. Besides, if the road has whoop-de-doos, the seated rider will get launched from the saddle and possibly crash.
Some riders believe that sticking the knee out or leaning the body away from the bicycle, improves cornering. Sticking out a knee is the same thing that riders without cleats do when they stick out a foot in dirt track motorcycle fashion. On paved roads this is a useless but reassuring gesture that, on uneven roads, even degrades control. Any body weight that is not centered over the bicycle (leaning the bike or sticking out a knee) puts a side load on the bicycle, and side loads cause steering motions over uneven road. Getting weight off the saddle is also made more difficult by such maneuvers.
To verify this, coast down a straight but rough road, weight on one pedal with the bike slanted, and note how the bike follows an erratic line. In contrast, if you ride centered on the bike you can ride no-hands perfectly straight over the same road. While leaning off the bike, trail of the front wheel causes steering on rough roads.
It is often said that putting the outside pedal down in a curve improves cornering. Although most experienced riders do this, it is not because it has anything to do with traction. The reason is that it enables the rider to unload the saddle while standing with little effort on a locked knee, cushioning his weight on his ankle. This can only be done on the outside pedal because the inside pedal would hit the road. However, standing on one extended leg does not work on rougher roads, because the ankle cannot absorb large road bumps nor raise the rider high enough from the saddle to avoid getting bounced. Rough roads require rising high enough from the saddle to avoid hard contact while the legs supply shock absorbing knee action, with pedals and cranks horizontal.
Most of the "body English" riders display is gratuitous gesturing, much like the motorcyclists who stick their butt out in curves while their bikes never get down to 45 degrees (the angle below which hiking out becomes necessary to keep hardware from dragging on the road). In fact, in a series of tight ess bends, there's no time to do any of this. It's done by supporting weight on the (horizontally positioned) pedals, and unless the road is rough, with a light load on the saddle. On rough roads, the cheeks of the saddle, (the ones that went away with the Flite like saddles) are used to hold the bicycle stably between the legs while not sitting.
The path through a curve is not symmetrical for a bicycle, because it can slow down much faster than it can regain speed. Thus the trajectory is naturally asymmetric. Brakes are generally used to the apex (that is usually not the middle) of the curve, where pedaling at that lean angle is not possible, nor does pedaling accelerate as fast as braking decelerates.
Although the railroad term switchback arises from early mountain railroading where at the end of a traverse, a switch is turned to back up the next traverse, after which another switch is turned to head up the next, on roads these are hairpin turns. In such turns trajectory asymmetry is most conspicuous, because braking can be hard enough to raise the rear wheel when entering but one cannot exit with such acceleration. For this reason, riders often find themselves with extra road on the exit of such turns, having slowed down too much.
Where to direct vision is critical for fast cornering. Central vision should be focused on the pavement where the tire will track, while allowing peripheral vision, with its low resolution and good sensitivity to motion, to detect obstacles and possible oncoming traffic. Peripheral vision monitors surroundings anyway, so the presence of a car in that "backdrop" does not require additional consideration other than its path.
If central vision is directed at the place where an oncoming vehicle might appear, its appearance presents a new problem of confrontation, stopping image processing of the road surface for substantial time. Because the color or model of car is irrelevant, this job can be left to peripheral vision in high speed primitive processing, while concentrating on pavement surface and composition.
When following another bicycle or a car downhill, the same technique is even more important, because by focusing on the leading vehicle, pavement and road alignment information is being obscured giving a tendency to mentally become a passenger of that vehicle. Always look ahead of the vehicle, observing it only peripherally.
Riders often prefer to keep their head upright in curves, although leaning the head with the bicycle and body is more natural to the motion. Pilots who roll their aircraft do not attempt to keep their head level during the maneuver, or in curves, for that matter.
Picking the broadest curve through a corner may be obvious by the time the preceding skills are mastered, but that may not be the best line, either for safety or because the road surface is poor. Sometimes hitting a bump or a "Bott's dot" is better than altering the line, especially at high speed. Tires should be large enough to absorb the entire height of a lane marker without pinching the tube. This means that a minimum of a 25 mm actual cross section tire is advisable. At times, the crown of the road is sufficient to make broadening the curve, by taking the curve wide, counterproductive because the crown on the far side gives a restricted lean angle.
Mental speed is demanded by all of these. However, being quick does not guarantee success, because judgment is even more important. To not be daring but rather to ride with a margin that leaves a feeling of comfort rather than high risk, is more important. Just the same, do not be blinded by the age old presumption that everyone who rides faster than I is crazy. "He descends like a madman!" is one of the most common descriptions of fast descenders. The comment generally means that the speaker is slower.
Although tandems with their higher weight to wind drag ratio have this problem more often, steep mountain roads, especially ones with poor or no pavement require so much braking that single bicycles blow off tires from overheating. For tubulars the problem is not so much over pressure than rim glue melting as all pressure sensitive glues do with heating. As glue softens, tires slip on the hot rim and pile up on the valve stem. This is the usual indicator that tubular tire wheels are too hot. The next is that the tire arches off the rim in the area just before the stem.
This is a serious problem both for tubulars and clinchers because most clincher tires, given enough time on a hot rim will blow off if inflated to recommended pressure. Pressure that gives good rolling performance (hard) while tubulars roll off from lack of adhesion to the rim. The faster the travel, the more descending power goes into wind drag and the better the rims are cooled. Going slowly does not help, unless speed is reduced below walking pace.
On steep descents, where rims stay too hot to touch for more than a minute, reducing tire inflation pressure is a sure remedy. However, tires should be re-inflated once the rims cool to normal. The blow-off pressure is the same for small and large tires on the same rim, it being dependent only on the opening of the rim width. Also, tires with a smaller air volume become hot faster than larger ones.
There is no way of descending continuously and steeply without reducing inflation pressure, unless there is an insulator between the tube and rim of a clincher. Insulating rim strips are no longer offered because they were an artifact of dirt roads that often required riders to descend so slowly that all potential energy went into the brakes and almost none into wind drag. These rim strips were cloth tubes filled with kapok, their insulating purpose being unknown to most people when they were last offered.
See also Sheldon Brown on Braking and Turning
More Articles by Jobst Brandt
Next: Slope Wind, the Invisible Enemy
Previous: Cycling Myths
Articles by Sheldon Brown and others Harris
Beginners Brakes Commuting