I mostly ride touristic style, and its not uncommon descents of few hundred meters or even 1000 meters (vertically, usually average gradient is around 5-10%). What's the best technique to brake to avoid glazing of brakes?


I've heard following advices:

  • pulse brake instead of constantly braking
  • use alternately front and rear brake
  • create as much of aerodynamic drag as you can to reduce the pressure on brakes

Apart from that, I have following theories:

  • brake pads have some temperature above which they will start to glaze, so I should brake in such way, to avoid exceeding this temperature
  • if I brake suddenly with great power, brake pads/disc rotors/rims will have to absorb a lot of energy which translates to heat, if I brake gently - less heat is generated
  • if I brake with both brakes energy will be divided between two brakes so as heat (though not necessarily evenly - division depends on the force applied on brake levers)
  • the greater the speed, the greater aero drag, so if I go faster, less energy (in total) will need to be absorbed by brakes
  • the lower the speed, the longer the time that it takes to ride, so longer the time to dissipate the heat from brake system

Braking strategies

So based on all above I have two strategies for braking:

  1. Brake as little as possible, but when needed (ex. before hairpin turn) brake shortly with great power and both brakes
  2. Brake constantly using both brakes and maintain low speed (below 20km/h?)

First strategy is trying to use as much aero drag as possible and then it will generate a lot of heat in brake system. Though during the brakes it will have time to cool down again. Second strategy maintains constant high temperature of brakes, but provided low travelling speed I assume that temperature will not exceed glazing temperature.

These strategies I discarded:

  1. Brake constantly and maintain low speed using alternating brakes. This means that although the brake that's not operating is chilling down the other gets full load and its temperature will be higher improving risk of exceeding glazing temperature.


I've already glazed a few disc rotors:

  1. In my commuting bike, I have one hill (700m long, 4% gradient), and I brake shortly but with great force
  2. In my trekking bike - downhill on straight asphalt, starting with gradient above 10% where I gained ~50km/h later I tried to not gain any more speed using alternating brakes technique

Also I have a feeling, that after glazing, when I experience lower braking strength, if I ride a few downhills when I'm more gentle towards this brake it regains its effectiveness. Is it possible?

Though it might be problem solvable using different means (reduce my weight, use larger rotors). I'm looking that would work best, even on non-disc-brake bikes.


Final Question

To reduce number of variables, my final question is: Given:

  1. The speed Vn - which is the speed the biker would get if he would not brake nor pedal at all, when areo drag is preventing him from gaining any more speed from gravitational force.

  2. Desired speed the biker would like to go Vd = R * Vn

What are the optimum braking strategies that:

  1. Minimize time that it takes to ride 10km of such slope
  2. Prevent brakes glazing

Depending on the R. If R >= 1 then the strategy is just not to brake at all, that's easy, but how to brake if R=0.5 or even R=0.1?

  • 1
    You said you “glazed” your rotors on a 4% hill by only braking shortly? That seems odd. What was the duration of your shirt braking period? What is your rotor size?
    – Rider_X
    Commented Jul 14, 2018 at 21:08
  • Minor clarification: " descents of few hundred meters or even 1000 meters" is that horizontal travel, rolling distance, or vertical elevation change ?
    – Criggie
    Commented Jul 14, 2018 at 23:27
  • total energy over the total time is irrelevant. Energy in to energy dissipation over any time period, with respect to the specific heat capacity of the system and the glazing temperature would be what matters.
    – user20209
    Commented Jul 15, 2018 at 0:10
  • @Rider_X: my rotor size is 160mm, I weight 90kg and 8 carry around 20kg on my rack. Short braking is around 3 seconds.
    – Wiktor N
    Commented Jul 15, 2018 at 9:47
  • 1
    @Crigge: 1000 meters vertical. Clarified in question
    – Wiktor N
    Commented Jul 15, 2018 at 9:48

4 Answers 4


Brakes work by turning kinetic energy into heat through friction. As you heat up the system there generally is less capacity to absorb more heat (in addition to physical changes in the brake system component).

Constantly dragging your brakes is probably the worst strategy. Tandem bikes do this but they typically have one braking system dedicated to being a drag brake, and a separate braking system to be used for stopping. If the drag brake fails from overheating the primary brake system should be unaffected.

The reason constant drag braking is a bad strategy if you do not have a dedicated drag brake is as follows:

  1. The brakes never get a chance to cool off, building up heat over time which can make them ineffective if for example you need to suddenly stop;
  2. If you are drag braking to maintain a slow speed then you are not taking advantage for aerodynamic drag. Drag goes up with the cube of velocity. As such you get the highest braking force from aerodynamic drag the faster you go (yes it does sound crazy).
  3. If you are not effectively using aerodynamic drag you are forcing your brakes to do proportional more work (i.e., heat) than if you let your speed run up then brake.

Depending on your tolerance for speed (I am personally happy running my speed up to 90+ km/hr), the optimal strategy is to let your speed run up, do a period of braking where you drop your speed down (e.g., go from 80 km/hr down to 40 km/hr), let off the brakes and let your speed run up. This lets you cool off your brakes as well as use aerodynamic drag to your advantage. The key is to try and have long periods between a braking bought so you can get as much heat out of your brakes. If the period between a braking bought is too short (e.g., pulsing your brakes) you will start to get excessive heat build up in your brakes eventually leading to fade and even failure in some extreme circumstances.

  • And the faster you go, the faster the braking system moves through the air, the better the temperature gradient that can be maintained, the faster the braking system can cool down, so double win for speed :) For any reasonably achievable speed, on a pedal bike, on a road, on earth ... just being careful in-case pedants are lurking ;)
    – user20209
    Commented Jul 15, 2018 at 0:14
  • To increase time between braking moment you can come to a complete stop and wait until they cool off enough. Commented Jul 15, 2018 at 2:55
  • 2
    @ratchetfreak without airflow the brakes will take a long time to cool off at a stop. It can work, but you have to be very patient.
    – Rider_X
    Commented Jul 19, 2018 at 16:34

Issues with glazing on 5-10% gradient descents indicate that overall bike technique/confidence is what needs improvement, braking less and harder is more the culmination of good technique and the resulting confidence, rather than a skill per-se. Yes you need to un-learn the bad habit of braking too much and dragging the brakes, but in order to do so, you must learn other good habits. Turns etc vary but I haven't been on a 10degree or less road decent where I've considered braking for years (except traffic light related/traffic related etc braking)

But firstly and most simply:

Non techinque related

Change to:

  1. Metal pads. Check your discs are compatible, also check your disc width more regularly; metal pads will wear them faster.
  2. Larger rotors. Check your frame/fork are compatible.
  3. Finned pads.
  4. Ice-tec and similar rotors.
  5. Brakes that take pads with larger surface areas.

Technique related

Learn to ride faster.

Harder, less frequent braking is all well and good, but ultimately this results in riding faster and harder, doing this safely involves an overall improvement in your whole descending skill set, and ideally intimate knowledge of your bike, so that you can be even more confident it's not going to fail you at speed.

edit: much of this was written with the assumption of 10% decent mountain biking, since I kind of skimmed the question and it didn't cross my mind that one could brake so much that one glazed the pads on 10% road decent. However it's largely valid for road riding also.

Ultimately one of the most important skills for more rapid descending is reading the terrain, looking further ahead, planning where you will brake. This becomes second nature after a while, but can be frustrating until you start to notice improvement. Everyone can always get better at this skill and it is IME a huge factor in fear/confidence.

Fundamentally it's information processing and decision making. If you can't process fast enough and decide fast enough then no matter your other skills, you can't make use of them properly and safely.

If you are afraid you can't really be in full control because you are distracted, not focused. It is important to remember to listen to that fear, some days, it's just not your day for whatever reason. If you know the reason, work on it. If you don't, then accept the off-day, pushing the fear out of your mind and gun-ho charging like you did some other day can be disastrous if you don't both recognize and address the cause of the fear.

How to go faster (i.e. result in braking less)

Learn. Self teach and/or pay for lessons. If you self teach make sure to record yourself practicing so you can instantly review your technique; it's easy to learn the theory and delude yourself into thinking you are putting it into practice.

Practice. Practice does not just mean "go ride", it means ride something you find tricky, then analyze it, consider what the theory says about riding it and then repeat and review that section till you're putting the theory into practice automatically.

Don't think about going faster, it is a meaningless thing to think about; going faster is a by-product of better technique; the technique is HOW you go faster and that is what you should be thinking about.

Of all the techniques for going faster, braking less has the most per-requisites. You can only brake less when you can process, decide and react faster, i.e. when all your other skills can take less braking. It's very rare to come across some one who needs to specifically focus on braking less, as opposed to all the skills that will enable them to do so and have the confidence to do so.

Note that generally not braking in corners could be considered "learning not to brake" but the process is generally one of learning to corner, with all the skills that entails and learning not to brake is more a matter of unlearning one of the worst and most dangerous bad habits, rather than learning an actual skill.

The only time you really need to learn to brake less, is if you feel totally confident all the time at the speed you are going and want to go faster, but are still stuck in your old ways, subconsciously feathering. Note that this behavior, when all technique appears to be good -apart from braking-, is more often than not, actually a sign of not having total confidence, which is often actually caused by not looking ahead enough and not processing information fast enough.


Some bike adjustment to increase descending confidence is also an option. This can involve lowering your seat post before the descent (or get a dropper, there are some decent cheap options around), edit: removed the purely MTB related suggestions (and yes I would use a dropper post even on a road bike myself)

  • 1
    I did not go into most of (or even mention) the techniques of faster descending as that is a HUGE topic with much information all over the internet. But, fundamentally: Identify what triggers fear of going faster (i'm guessing cornering and fear of not getting down to speed before the corner?) and read up on the theory, then work on putting it into practice. I still remember spending hours learning body position and muscle memory, making s shapes between cones (video self!) when I decided to take MTBing seriously, and the "brake"-through I had when I realized I had to work on looking ahead. :)
    – user20209
    Commented Jul 14, 2018 at 23:48
  • Last week the source of my fear was laptop in my panniers and really bad riding surface (a lot of stones, surface destroyed by flow of water etc). But I really appreciate your answer. I guess that I need learn more about how to ride through sharp hairpin bends (180 turn with less than 10m radius)
    – Wiktor N
    Commented Jul 15, 2018 at 9:56
  • 2
    At 50k/h speeds strength could become the major issue for many people; getting into and maintaining the correct lean (over 45 degrees) and fore-aft weight distribution for a 10m radius hairpin, will take practice, and probably more core strength. Your technique will have to be really good already, but all the extra core strength you can get is going to help when you're trying to hold your position for 10 seconds with those forces acting on you. Go faster as you get better, however, sharp twisty inconsistent turns one after another are harder than straight to hairpin, so be careful on those.
    – user20209
    Commented Jul 15, 2018 at 14:10
  • 1
    Oh and of course you have to take conditions and the surface into account, I wouldn't go trying any of this on the bad surface you mentioned!
    – user20209
    Commented Jul 15, 2018 at 14:12
  • 2
    Issues with glazing on 5-10% gradient descents indicate that overall bike technique/confidence is what needs improvement often yes, but getting stuck behind cars is a big issue too (I had 2 cases on Saturday alone where I had to descend on the brakes because of cars that had overtaken just before descents -- shorter but steeper hills) +1 mainly for metal pads
    – Chris H
    Commented Jul 18, 2018 at 13:20

Metal pads on disc brakes (a bit much for a comment).

The metal pads on my tourer were better than ever after the second descent on this ride (the first big descent had a lunch stop in it but was otherwise similar). This was the first time I'd come down anything like this so was quite cautious, i.e. on the brakes quite a bit -- probably too much but the bike can take it.

To give an idea of my descending style here's the top of that second descent on video. Note the left hand operates the back brake on my British bike; if you look carefully you can see when I change from decelerating to not accelerating too much (those hairpins are steep) and I relax the brake slightly (clearer with my cable discs than it would be with hydraulics). Front brake behaviour was similar.

  • 1
    I really need to do the experiment I planned with an IR thermometer looking at my rotors and a nice downhill. The problem is I was going to dismantle the thermometer to do it, and I keep finding uses for it.
    – Chris H
    Commented Jul 19, 2018 at 15:52
  • To give an impression about what kind of conditions I'm talking - this is one of the recent descents (though only 600m vertical): google.com/maps/@38.0414032,15.9501289,3a,75y,23.71h,68.85t/…
    – Wiktor N
    Commented Jul 19, 2018 at 19:18
  • So on that road (I'm on mobile so I've only looked at a little of it) you need to ride at a speed suitable for the surface/visibility/traffic/bike. This is likely to mean braking just before each hairpin at least, and not decelerating in the bends themselves (that's not to say completely releasing the brake as you may need to limit acceleration). I would say it's more a question of whether you spread your braking over each little straight, or allow your speed to build a bit then hit the brakes in time for the next bend. The latter maximises air braking
    – Chris H
    Commented Jul 19, 2018 at 20:10
  • Apart from generally better braking, metal has the advantage of actually working in the rain and mud, even if you haven't touched the brakes for a while... with resin you sort of need to feather on purpose in the wet to evaporate/remove the water on the discs, so you actually have instant power when you need to brake. With metal it's a non issue. I'd like to see those measurements too :) Ice tec + finned metal vs normal metal would be interesting too... can I borrow your thermometer gun please :)
    – user20209
    Commented Jul 21, 2018 at 9:15
  • 1
    @Wiktor N I like that turn, slightly banked, the sense of danger, followed by more turns, so awesome... but I also hate that turn because it's (mostly) blind and I don't trust motorists :(
    – user20209
    Commented Jul 21, 2018 at 9:27

I'm going to throw a bit of math here to show exactly how much heat your brakes have to handle - and this does get a bit long, and won't answer the question directly.

First, when you descend you have to dissipate a lot of gravitational potential energy. That energy is proportional to your mass: the more you weigh, the more energy you need to dissipate. So given a specific mass and height of descent, the amount of energy you have to dissipate is constant.

And you have effectively only two ways to dissipate that energy: through aerodynamic drag, or by using your brakes to turn the energy into heat.

How much energy?

Assuming you, your bike, and all your gear weighs in at 100 kg, for a 1,000 meter descent you need to dissipate

9.8 * 100 * 1,000 Joules

That's 980,000 Joules.

How much energy is that?

Enough to boil away almost all the water in a full 500 mL water bottle.

Think about how long it takes to boil away a 1/2 liter of water on a stove at the highest heat possible...

980,000 J is enough energy to light a 100W light bulb for almost three hours (980,000 J / 100 W = 9,800 sec).

And how hot does a 100W light bulb get?

Dragging your brakes to stay slow

So what happens if you drag your brakes all the way down a descent to keep your speed down?

Well, let's assume a steep, straight descent to keep things simple.

On that descent, with no brakes at all, let's assume aerodynamic drag causes you to descend at 90 km/hour. If you do that, you dissipate all of that 980 kJ via aerodynamic drag - because you never used your brakes.

But if you don't want to go that fast? And you drag your brakes on the entire descent and keep your speed at 30 km/hour? That's going to split the energy dissipation between your brakes and aerodynamic drag.

How much will go into drag?

Well, energy loss via aerodynamic drag is proportional to speed cubed, so if all of the energy was dissipated via drag at 90 km/hour, going only 1/3 as fast means only 1/27 of the energy will be dissipated, leaving all the rest of the energy to be dissipated as heat by your brakes. But since you're going only 1/3 as fast, the smaller drag force you do encounter acts on you for three times as long, so 3 * 1/27 = 1/9.

So dragging your brakes to keep your speed at 30 km/hour means you have to dissipate 8/9*980,000 Joules - which is still almost 90% of all of your energy.

Again, that's enough energy to completely boil away almost 1/2 a liter of water.

And how many light bulbs will that much energy keep lit on your way down? Well, assuming a 10% grade, or 10 km descent, you'll take 20 minutes to get down 1,000 vertical meters. 20 minutes is 1,200 sec. 980,000 / 1,200 = 817W.

8/9ths of that is 721W.

(And now you also know why you can't climb a 10% grade at 30 km/hour - because you can't sustain an output of 721W through the pedals...)

Yeah, you're bleeding off enough energy to keep seven 100W light bulbs lit.

Into your brakes.

How well do you think that's going to work?

Even if you go just 10 km/hour (and you can't go much slower than that and remain upright...), you still have to dissipate 260W+ into your brakes - enough to light almost three 100W bulbs - try holding those in your bare hands.

That's why dragging your brakes down a descent does things like cause tires to blow out or melt carbon fiber wheels if you have rim brakes. Or heat up your disk brakes and have the heat conduct into your hub and start destroying your hub lubrication, or conduct into your brake system and start boiling hydraulic fluid.

Or glaze your brake pads because they got too hot.

  • 1
    I'm not sure the numbers here are very helpful. First, you're comparing against a hypothetical nearly straight-line descent where one can lose 1000m of altitude at 90km/h without needing to touch the brakes. Second, your analysis is entirely in terms of the amount of energy that needs to be dissipated by the brakes, but that's a largely meaningless quantity. You need to analyze in terms of the power dissipated by the brakes, i.e., energy per unit time. According to your analysis, holding the brakes to do 1km/h continuously would be just as bad as doing 30km/h. But that's obviously not true: Commented Jul 18, 2018 at 15:02
  • 1
    the total amount of energy dissipated by the brakes would be the same (basically all of the 980kJ) but doing it at 1km/h would take 30 times as long to get down the descent, so the power required would be very small and the brakes would stay nice and cool. Commented Jul 18, 2018 at 15:03
  • 1
    Sure, 1km/h was hyperbole to get across the point about power. Still, my point stands. Doing the descent at 10km/h would dissipate the energy over the course of an hour, rather than 20 minutes; 5km/h would be two hours and fast enough to avoid falling off. Your answer gives no indication of how to judge whether that would be safe, because you consider only the amount of energy to be dissipated, and not the rate. Commented Jul 18, 2018 at 16:07
  • 1
    I wasn't proposing holding anything in my hands: not light bulbs and not brake rotors. I've been saying all along that you need to be talking about power and not the amount of energy. Thank you for finally doing that: please edit that information into your answer! Commented Jul 19, 2018 at 13:24
  • 1
    Is this even trying to answer the question?
    – ojs
    Commented Jul 19, 2018 at 14:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.