25

I have heard several times the wisdom that you should not switch derailleur gears while pedalling hard ("under load"). The "right way", according to those people, is to pick up speed, stop pedalling, then switch gears, then pedal lightly (without any force actually pulling the bike) until the chain switches, and only then can you resume pedalling with full power.

Otherwise Unspeakably Bad Things are supposed to happen.

My wife however doubts this, and upon a little thought I have to admit, I don't see anything that could go wrong either. But then neither of us are very experienced.

So... how bad is it to switch gears while putting a lot of force on the chain? Is it total bullshit; will the bicycle explode immediately; or something inbetween?

8
  • 2
    This greatly depends on the type of gearing you use - in particular whether it's a derailleur setup or a hub gear. I assume you speak about derailleurs, since you speak about the "chain switching"? Could you edit to clarify? – sleske May 4 at 7:12
  • 1
    Are you asking about a new well tuned bike with Durace/XTR quality components or an older, higher mileage bike with neglected Sora/Acera (or lower) quality components? Is the rider a gorilla mushing pedals or 'in tune' with the feel and noises coming off the bike? Answer is very different..... – mattnz May 5 at 3:38
  • @mattnz - Ehh... the arithmetic mean between all of them? I can describe my own personal situation, but that will probably narrow the question way below usefulness. – Vilx- May 5 at 6:46
  • 1
    OK, I think I haven't been clear enough. Of course I've tried each way and I normally switch without load and it's not a problem and I don't stop pedalling because of that, etc. I was just wondering how much (if at all) switching under load damages the bike. And, by extension, if I should spend effort in teaching my kids to avoid load (not so easy when they're not that big). – Vilx- May 5 at 15:26
  • @ben anecdotes might be a good basis for an answer. Comments are for clarification and improvement of the Question or Answer. – Criggie May 9 at 9:01
36

The answer heavily depends on the groupset construction/generation and personal perception of what is "bad", among other factors. Below are a few aspects that I am aware of.

  1. Switching gears in the traditional bicycle drivetrain assumes there is a momentarily side load on the chain, twisting it and forcing it to jump/fall from one cog to another. The chain is very flexible in one plane, but has limited sideways flex before the metal becomes too stressed. In the worst case, "Unspeakably Bad Things" is one of links of the chain breaking because of that excessive side load. Usually, bad shifting manifests itself as horrible metal clunking and abrupt pedal jerking, but nothing breaks (yet).

  2. Aside from the chain, similar side load is experienced by the cassette and chainrings. Similarly, a potential for bending individual teeth on them exists.

  3. The probability of a mechanical problem during shifting can be increased by several factors. Among them are: front-shifting while loading the pedals, back-shifting while pedaling hard, weak pin in the chain, fast shifting over many gears at once causing the chain to twist too much; misaligned derailleurs; worn drivetrain components; bad luck.

  4. Older (more than ~40 years old) cassettes were made of plain cogs without any features to help shifting. Recent cassettes and chainring have ramps and pins that assist shifting, making the opportunity window for chain overload smaller. Most modern MTB cassette designs were reported to greatly alleviate the problems with shifting under load in either direction, according to many reports I've seen on the net. Even for cheap Deore 12-speed cassettes, it is said to be safe to shift without taking it easy on the pedals.

  5. Your description of events "to pick up speed, stop pedaling, then switch gears, then continuing pedaling" applies to internal gear hubs, but not to the cassettes/derailleur designs. In order to complete the shift, one must keep the chain moving over the cogs, which means pedaling (even if with less effort) during the whole shifting process.

In terms of practical advice, I'd say that there are several situations.

  1. It is usually safer to shift to smaller front or rear cogs while pedaling hard than the other way around.
  2. It is usually safer to shift at the back than at the front under load. The twisting angle at the cassette is usually smaller than it would be at the chainring.
  3. It is recommended to be conscious about shifting when pedaling uphill. The pedaling torque is the highest in such situations. If possible, one should choose the right gear combination just before starting the uphill movement, at the trajectory's lowest point. Making such predictions requires some experience.
  4. Similarly, never shift while standing still. The torque is high at the starting point, the chain may be cross-shifted over multiple cogs, making it more twisted than usual. If you cannot safely start in current gear, it is safer to lift the bike and pedal "it in the air" with your hand to complete the shift.
7
  • 10
    Or just lift the wheel and pedal in the air with your foot clipped. It is quite a common thing to do when you have to stop at an intersection and you had a hard gear. – Vladimir F May 3 at 13:46
  • On front shifting under load, did you mean the reverse? I have dropped my chain when shifting to the small ring under load a few times. Otherwise, I concur, with modern drivetrains it's better to ease on the pedals when you shift but it won't really affect anything, aside from the small chance of dropping your chain to the inside. – Weiwen Ng May 3 at 14:23
  • @WeiwenNg I did not consider the chances of dropping the chain in my answer; I should have done it, in retrospective. My comment about front shifting under the load only convers the possibility of cracking the chain because of twisting. – Grigory Rechistov May 3 at 14:52
  • I would add that some internal gear hubs actually are designed to use the pedalling input to help shift (e.g. the shimano nexus 8, and I heard the same about the rohloff but couldn't find any reference). But anyway I guess it is safe to say the lighter the mechanical load the longer the system will last. – flawr May 3 at 21:03
  • 2
    @cmaster-reinstatemonica Yes, that is a valid correction. It seems that "stop pedaling" in either case (IGH/derailleur systems) should mean "ease on pedaling for a split second". It's the timing of the event relative to other actions, such as pressing levers, knobs, that is different between the systems. One should never need to "stop and go" with either type of gear systems. – Grigory Rechistov May 4 at 6:50
10

You can quite easily get your chain dropped. If you were standing and pushing hard in a climb you could even fall. If the chain drops to a bad place, it can damage your frame near your bottom bracket or it can damage your spokes or similar.

3
  • 3
    I wouldn’t say "easily" for properly adjusted derailleurs but yes, that’s a valid concern if you are shifting on the front or to the easiest or hardest gears in the rear where dropping the chain is possible. – Michael May 4 at 7:07
  • 2
    @Michael A properly adjusted rear derailleur is a good prevention but even the best front ones are very crude machines. The pros know why they carry chain catcher and after some unpleasant surprises I also carry one on my main bike. No dropped chain on the BB since then. – Vladimir F May 4 at 7:28
  • @Michael I wouldn’t say "easily" for properly adjusted derailleurs... "Properly adjust[ing]" your derailleur isn't sufficient because if you're under load, like this question mentions, the frame will flex. throwing that "properly adjusted derailleur" out of alignment. Especially if you're climbing hard. When I was racing a few years back, at 200 lbs or so I'd climb at 350W or even more depending on how long the climb was. Two or three times I had the chain drop between the largest cog and the rear wheel spokes - the rear wheel spokes were trashed and I had to rebuild the wheels afterwards. – Andrew Henle May 9 at 22:13
7

Expanding on a comment I made on Gregory's answer: I agree fully that it's better not to shift under load. However, modern drivetrains are much less sensitive to shifts under load than older ones. You will have noisier rear shifts. For front downshifts, I believe there's a small chance you can drop your front chainring; as Vladimir stated, on some frames you might get the chain wedged between the inner ring and the frame, which can be hard to recover from and which might chip paint. As such, I would recommend most riders learn to ease up as they shift the front once they've mastered the basics. The handful of chainring drops I have had in the last two years were all under load.

It may be worth reminding riders that if you drop your chain to the inside of the crankset, you can often recover by shifting to your big ring and pedaling. The chain will often shift back onto either of the front chainrings.

6

When you are pedalling hard, you are putting a lot of "force" into the chain, which is transmitted to/from the chain via the mechanical contact between chain and gear teeth.

If you are putting a lot of force and you shift, there are some brief moment when all your force is applied torsionally to one tooth only, by a chain that is not alignated but slightly bend (because of the transition from one gear to the other).

The individual force is not enough to snap anything, but metal components suffer from fatigue, and they are designed to stand so many cycle of so many force applied (10'000s?100'000s?1'000'000s cycles? it depends).

The shifting under load is a load for which neither the chain or the gear is designed to resist: by doing that you are inducing microscopical damages to the metal, accelerating the wear of components.

On the sunny side of life, even if the tooth snaps broken because of this accelerated wear, it is true that it is extremely unlikely to have a catastrophic failure (i.e. sudden lack of equilibrium and then falling form the bicycle, being maimed by an incoming truck while doing an emergency clearing of the intersection) and more likely to have a big stumping of crotch area & attached volumes on the top tube.

Additional info: check pictures here (SFW, no crotches, only chain-teeth engagements are depicted)

2

I have heard several times the wisdom that you should not switch gears while pedalling hard ("under load").

It is true that an optimal gear shift works by easing the load, shifting while rotating the pedals with practically zero load, and resuming the load after the shift has happened. With practice, it is possible to time the shifting in such a manner that the shift happens when pedals are at 6-12 o'clock positions when there would be naturally no load anyway. After the shift has been detected from its sound, or if traffic noise is heavy, from the change in gear ratio as felt by the legs, the load can be resumed. Done properly, there is very little interruption in load.

However, this does not work for mid-drive electric bicycles. An electric bicycle has a certain amount of torque, let's say 50 Nm. In an effort to increase the average power of the e-bike, the mid-drive manufacturers make this torque continuous as opposed to oscillating in unison with pedaling loads. So, when you are naturally in a position where there is no pedaling load, the mid-drive still is outputting its torque. The mid-drive software cannot know you want to execute a shift at this very instant. Then, after the mid-drive software detects the shift from the jerk of the chain, it reduces torque to zero but too late. The shift has already happened. However, even with mid-drive e-bikes, the cyclist can and should reduce the pedaling torque during shifting -- then only the motor torque (usually when heavily pedaling, lower than pedaling torque) is present.

In the past, bicycles had chains very unlike those in bicycles today. The chains had full bushings like any reasonable chain in other use such as motorcycle or industrial equipment. The ends of the pins were not "punched" so you could take apart the chain at any position and re-join it using the existing pin (being careful you won't push the pin fully out when disconnecting the chain). Such chains could last tens of thousands of miles if properly taken care of (never lubricating an already oily and thus dirty chain, but rather cleaning it before lubrication).

Then the bushingless chain was invented. It replaced the full bushing by two bushing-halves integral to the inner side plates. The bushingless chain is stronger, lighter, cheaper (4 distinct part types as opposed to 5) and more laterally flexible. However, the bushing-halves do a really poor job at keeping dirt out so bushingless chains rarely last more than perhaps two thousand miles.

Furthermore, the chain was considered not strong enough for shifting under load so modern chains have "punched" pins. You cannot re-join such a chain with the existing pin. Instead, you must push the existing pin out and use a new reinforced connecting pin that is larger so it stays in the hole that was damaged by pushing the "punched" pin out. (At the location where the reinforced pin is used, the chain may never be broken again, so a 116-link chain can be only re-joined 116 times.)

The combination of lateral flexibility and strength of the bushingless chain and "punched" pins improving pin retention means modern chains actually don't suffer from shifting under load.

A cyclist can (and perhaps even should) still use the good practice of easing the load while shifting a modern chain, but that's not realistic on mid-drive e-bikes because the mid-drive motor load is not eased quickly enough.

Otherwise Unspeakably Bad Things are supposed to happen.

The bad thing is a broken chain. If the chain didn't break, no bad things happened. Chain wear life is unaffected by shifting under load. So you won't get 1000 kilometers of life shifting under load often and 5000 kilometers of life never shifting under load. The wear life in both cases will be the same.

Modern chains shouldn't break even when shifting under load, and mid-drive e-bikes cannot reasonably be shifted without load.

Such is the parts market today that you cannot find a bushing-type chain anymore unless it's a ridiculously expensive new-old-stock item.

1

You are much better off learning to ease the load before switching gears! Assuming you have the widespread derailleur kind (as opposed to in-hub gears)...

1. Front derailleur, switch under heavy load:

  • Large ring to small ring: Nothing bad will happen. The chain will rub against the derailleur which will wear out a bit and the switch will just not happen until you release the load.

  • Small ring to larger ring: If you insist you will likely bend your derailleur or break the cable.

2. Rear derailleur, switch under heavy load:

  • Both ways: undue lateral stress on both the chain and the cogs. Premature wear, slight risk of breaking the chain (a cause of fall under load).

  • Light load should be OK with modern derailleurs.

1

Related - switching gears in an Internal Gear Hub (aka an IGH) under load causes the gearbox to not switch immediately.

I have had situations with a Shimano Alfine 11 where pedalling hard, and changing gear in either direction meant the gearbox stayed in the same gear. When going up a hill and trying to change to a lower/easier gear, it just wouldn't until I eased off the pressure.

Likewise, when powering away from traffic lights, I would normally work through the gears. Not letting off meant you were stuck in an easy gear and pedalling with higher cadence, till you couldn't keep a smooth pedal cycle, a moment of slack allowed the gearbox to catch up with the shifter, and suddenly you've changed 5 gears at once. This happened many times, and several times I came close to being thrown because of the sudden change affecting balance.

However, IGHs are essentially the same as a single speed chain setup, so the chain and cog are at no risk. Damage could come from having a gearset partially engaged, so that there's load on the corners of teeth inside the gearbox. This would manifest as metal shavings in the oil at the next oil change.


Strictly speaking this doesn't answer OP's question which is specific to derailleurs. However I'm including this for completeness, as well as that at least one product exists in 2021 that includes an IGH and a rear derailleur.

2
  • 1
    IGH= Internal Gear Hub – Swifty May 9 at 10:36
  • @Swifty excellent point - feel free to simply edit obvious improvements like this. – Criggie May 9 at 10:59
0

The gears just do not switch reliably under the load. You may end up with the previous gear remaining all hill up or, also not unusual, with the chain dropped. Don't known, likely it also bad for the bicycle but makes no sense to ride this way even if it would not. Depending on the load and if you are switching up or down this may be more or less problematic but always problematic.

One of the first skills with modern bicycle is to plan ahead and switch the gears in time. If you are late, to avoid the chain drop, return immediately the gear switch to match the current gear. You may learn to remove the load by pedaling with much less force while switching the gear, but this obviously slows you down and on the steep enough hill you may stop before your gear switches.

My bicycle only has a single ring in the front so all this is true about the rear ring. Switching both up and down is problematic under the load.

-2

in my experience it is better to shift gears with no load on the chain. So there is no rattle. My bike is over 9 years old and is in good working order.

1
  • 1
    Welcome to Bicycles SE. We're looking for answers with more detail. Please consider editing, your answer to explain what the impacts of shifting under load are. You mention a rattle. Are there other concerns? What are the long-term effects of doing this repeatedly? A short, one-line answer like this is likely to get downvoted, flagged for moderator intervention, and possibly deleted. – jimchristie May 4 at 14:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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