# Why does shifting under power affect shift quality?

I recently was attempting to shift up a hill and observed that

• The shifting levers on the handlebars displayed significant resistance to my pushing them.
• Despite several rotations of the crank, the bicycle had not shifted gears.

Googling around, I've seen people claim that pedaling under power increases chain tension and that this chain tension will affect how easily the chain moves between gears.

Intuitively, this does not make sense to me, as how easily the chain moves orthogonal to the direction of travel (i.e., the ease of shifting) would seem to be independent of the chain tensions parallel to the direction of travel. This seems to imply there's a causal relationship between the relative magnitudes of these forces.

Moreover, given that the chain itself is the same length regardless of the power transmitted, it's unclear to me how pedaling increases chain tension. Is it that one part of the chain becomes more tense, and the other becomes less tense? A free-body diagram type explanation would be very helpful.

How might one understand or model the shifting action and mechanism in a way that explains why chain tension affects shifting?

• What kind of gears do you have? Derailleur, internal or something else? If you have derailleurs, was it front or rear you had trouble with? – ojs Oct 1 '17 at 19:08
• Also, how could chain transmit power without tension? – ojs Oct 1 '17 at 19:09
• Did you back off pressure while changing? Or were you changing gear while pressing really hard on the pedals? – Criggie Oct 1 '17 at 19:21
• One simple way in which your intuition lets you down: any flexible object is harder to deform when under tension. Try hanging a light weight from a piece of string, then deflect the string. Repeat with a much heavier weight – Chris H Oct 1 '17 at 19:27
• Before the invention of "indexed" shifting (which required solving this problem as a prereq for indexing), you had to completely remove all tension on the chain to shift. With a good quality indexed setup (which includes "ramps" on the sides of the cogs and a few other enhancements) you can now shift "under load", but it still needs to be a relatively light load. (And, during an under-load shift, the chain is tense along the top, between front and rear cogs. If you know what "free-body diagram" is you should be able to figure this out.) – Daniel R Hicks Oct 1 '17 at 21:45

There may be other factors that affect "shift quality", if I understand what you meant correctly, but the most significant by far is “direction of force applied.

Basically for a gear to change the mechanism just pushes the chain “off rail” enough in the direction of the new gear. When the chain has a higher tension applied on it the derailleur doesn’t have enough strength to push the chain “off the rail”. (Derailleur is the name of the part the pushes the chain) If you still haven’t understood, imagine a line stretched between two fixed poles, if it is loose and you push it to the side it will move some, if it has a higher tension and you push to the side with the same strength and direction than before it will move less. And the tighter the line the less to the side it would move. A gear system has a minimum distance that it needs to move to the side so it can change gears.

More advanced, modern and obviously expensive have over time placed gears closer to each other and shaped individuals to make the “derailing” need a smaller push in a new direction, requiring less strength to push the chain and as a whole make the process more smooth.

If you didn’t understand something just ask.

I would suggest considering two power components: (1) rotational speed and (2) torque.

It should be naturally obvious that these two principles will impact the ability of a chain to jump across sprockets. As rotational speed tends towards zero, shifting becomes problematic. As torque increases, the chain will require more lateral force to be lifted above a sprocket’s cog.

Going uphill while maintaining a reasonable cadence has limited negative impact on modern cassettes that have sculpted cogs to promote easier shifting. However, these designs are challenged if you are a masher (slow cadence + high torque).

In the old days, riders would put less pressure on the pedals to reduce torque while keeping a high cadence. Today, higher end components work well under high torque as long as cadence stays above... I’d venture 60.

Applying greater pedaling force does put the chain under greater tension - that's how the pedaling force is transmitted to the rear sprocket and hence to the rear wheel.

Perhaps when you think of tension you are thinking of something like a rubber band or spring where tension produces very visible increase in length. The individual metal chain links do in fact stretch when under pedaling tension, we just cannot see it because the amount of deformation is so small.

If you take a long, flexible item such as a chain, rope or wire and put it under tension, there is in fact a force needed to push it sideways - think of a drawing back a bowstring, the greater the tension the bow puts on the string the harder it is to pull the string.

A bicycle chain tensioned between sprocket and chainring is like a bowstring tensioned between the limbs of the bow, although obviously a bow string is deflected far more compared to it's length than a bicycle chain when changing gear.

So, a chain under greater tension does resist a gear change more than under less tension. Modern chain, sprocket and chainring design have enabled successful gear changes under more power, but more pedaling force does still produce more difficult or noisy changes,

The other answers give you a good idea. But not addressing the issue. People don't use the gears correctly.
Tension on the underside never changes that much. It is the tension on the top. Friction holds the chain in the cog seat caused by tension pressure. Bad use of gears result in rounding of the teeth. Mainly because of over exerting of power in the give gear. The rear gear set is not meant for constant gear changes. It is rather a optimising gear. The front is designed for constant and rapid gear changes. You are fighting to move a chain up hill under strain if you wish to use the rear set for climbing quickly. Best is to set the rear at a comfortable gear and do rapid gear changes with the front. Also as others have pointed out, you need to ease off a tad for a second during gear changes. You will greatly increase the life of your running gear, as well as easier, quicker gear changes.

• This is complete nonsense. Any cyclist will make many more gear shifts with the rear derailleur than with the front: that's a simple consequence of their being somewhere between seven and eleven eleven gears at the back, but only two or three at the front. Furthermore, the rear derailleur can usually cope with shifting under heavy load, whereas the front absolutely cannot. – David Richerby Oct 4 '17 at 15:03
• I think you will find most riders disagree with you assertion about the purpose of front and rear gear shifting. – Argenti Apparatus Oct 4 '17 at 15:04
• @ArgentiApparatus And most manufacturers! Why would they put eleven at the back if you were mostly supposed to use the two at the front? – David Richerby Oct 5 '17 at 8:29
• I understand your thinking, as I was the same for a very long time. But remove the blindfold and look at it logically. Most of the time you need to do a quick shit to climb up hill. So do you force the chain to climb up 11 gears or just drop down 1 or 2?? I have tried both styles of riding and found that the convention is flawed. Down voting just because you think differently without evidence is like saying it is heresy. – Tempus Oct 5 '17 at 13:44