krengo
Norsk

Cornering on a road bike in autumn

Autumn rides can be some of the best of the year. The road is a little less predictable, rubber is stiffer in the cold and the light is lower, so a bit of extra margin goes a long way. This page shows, with model numbers, what lower grip does to a corner, and which habits give the margin back.

About the numbers. Every angle and speed below is a model value: the ideal lean formula from the calculator (g = 9.81 m/s²), the grip scale’s assumed μ, level Settled and a 2° margin. Your road, tyres and bike will differ. Krengo shows lean and an estimate of how much assumed grip a corner uses; it does not measure grip, cold, leaves or light. The field itself is not released yet.

Grip moves, so plan a margin

An autumn road is a patchwork of dry, damp and shaded stretches, and the grip changes between them. FEMA, the federation of European motorcyclists, quotes friction coefficients of about 0.7 to 0.8 on dry asphalt and 0.4 to 0.5 on wet, and a drop from 0.8 to 0.35 when loose gravel lies on the surface.1 Those are motorcycle figures for the same kinds of surface. A bike tyre will not match them exactly, but they show how big the swing can be. The grip scale assumes μ 0.90 for dry worn and 0.60 for wet; both sit a little above FEMA’s ranges, so the 0.45 row below is the more cautious wet case.

What if the grip is lower: a 15 m bend, level Settled, 2° margin
Assumed μRaw limitModel lineSpeed at the lineUse at 30 km/h
0.90, grip scale dry worn≈ 42.0°≈ 30.2°≈ 33.3 km/h≈ 82 %
0.60, grip scale wet≈ 31.0°≈ 20.8°≈ 26.9 km/h≈ 123 %
0.45, middle of FEMA’s wet range≈ 24.2°≈ 15.5°≈ 23.0 km/h≈ 164 %
0.35, FEMA’s loose gravel≈ 19.3°≈ 11.8°≈ 19.9 km/h≈ 210 %

The raw limit is atan(μ). The model line is atan(0.70 · μ) − 2°, and the speed is what that line allows in a 15 m bend. The last column is a 30 km/h pass through the bend, which needs about 25.3° of lean; 100 % is the model limit, not detected slipping. Read the table as “what if the grip is lower”, not as a reading of any surface: no source we read gives a μ for wet leaves or for manhole covers, so none is shown.

The speed column is the useful one. The same bend that has its line at about 33 km/h on dry worn has it at about 23 km/h at μ 0.45. Speed pays back quickly: slowing from 30 to 25 km/h takes the 15 m bend from about 25.3° to about 18.1° of lean, and from 20 km/h it is about 11.8°. With the Krengo defaults the pedal warning sits at about 33.2° and touch-down at about 36.2°, so in the three lower-grip rows the tyre reaches its limit before the pedal does; only on dry worn (42.0°) does the pedal come first. See the pedal touch-down angle page for the formula.

FEMA also reports, for motorcycles, that a friction coefficient below 0.45 raises crash risk about 20 times and below 0.30 about 300 times, quoting another source we have not read.1 We do not carry those numbers over to bicycles. The takeaway is the direction: when the surface is uncertain, spend less of the lean you have.

Cold rubber and cold air

Tyre rubber gets stiffer as it cools. Michelin’s book on tyre grip describes rubber at very low temperature as stiff and brittle, “a bit like glass”, and explains that the internal damping that gives rubber its grip is highest near the glass-transition temperature.2 It is about tyre rubber in general and gives no grip value for a bike tyre at 5 °C, so take it as direction rather than a number: the first kilometres of a cold morning are a good time to keep the margin a little larger.

Cold air also lowers tyre pressure. With the ideal gas law, 10 °C colder takes about 3.5 % off the absolute pressure, which is roughly 0.2 bar at 5 bar and 0.3 bar at 7 bar (our own calculation from 15 °C, absolute pressure = gauge + 1.01 bar). Check the pressure on the day you ride, not the day you pumped. What pressure does in a corner is on tyre pressure and grip in a corner.

Paint, ironwork and loose material

FEMA names road markings and manhole covers as slippery surfaces, and notes that worn material can show through new markings.1 Damp autumn mornings are when they matter most, so cross them upright, before the lean starts, and brake a little earlier and softer. Loose material such as gravel is the other autumn surprise. We found no measurement for wet leaves, so the page gives no number for them: look for them in the shade and near the apex, and choose the wider arc where you see them. The line and braking habits that give you that room are on the descending technique page, and the full wet-road numbers are on cornering in the rain.

Light: be seen, see early

Cycling in the dark has a higher injury rate than cycling in daylight, and early-morning darkness a higher one still, according to a review by Twisk and Reurings.3 They cite Dutch data, from Johansson and colleagues (2009, which we did not read directly), of about twice the risk in the dark on urban roads and about five times on rural roads. The positive side is that light helps. In a randomised study of 5,380 cyclists over one year, continuous use of daytime lights cut multi-party accidents in the dark by 71 %, which was statistically significant. The overall reduction in multi-party accidents, 30 %, was not.4 A light that is on in daylight too is cheap margin, and the shorter days make it worth having.

A short checklist

Try your own speed and radius in the lean angle calculator. Krengo shows lean and an estimate of how much of the assumed grip the corner uses. Read green as information, not as permission to go faster. Learn new speeds on a closed or quiet road with permission to ride there.

Sources

  1. FEMA (Federation of European Motorcyclists’ Associations), Road surface friction and motorcycling. Used for friction coefficients of dry, wet and gravel surfaces, road markings and manhole covers, and the crash-risk figures. It is about motorcycles; the same surfaces matter for bikes. Read on 5 October 2026.
  2. Michelin, The Grip (2001), printed pages 13–14. Used for stiff, brittle rubber at very low temperature and damping near the glass-transition temperature. About tyre rubber in general, not a bike-tyre value. Read on 5 October 2026 from a text extract.
  3. Twisk, D. and Reurings, M. (2013), An epidemiological study of the risk of cycling in the dark, Accident Analysis & Prevention 60, 134–140 (author version). Used for the higher injury rate in the dark. The Dutch figures are theirs, quoting Johansson et al. (2009). Abstract and introduction read on 5 October 2026.
  4. Hels, T. et al. (2025), randomised trial of continuous daytime lights for cyclists, Traffic Safety Research. Used for the 71 % and 30 % reductions in multi-party accidents. Only the abstract was read; the confidence intervals and the type of light were not. Read on 5 October 2026.

The lean formula, the friction limit tan θ = μ and the ideal gas law are standard textbook models and are not taken from the sources. The angles, speeds and percentages are Krengo’s own model values: the grip scale’s assumed μ, level factor and margin, and the pedal defaults on the linked pages. The pressure change is our own calculation.

Related pages