reconnaissance

An earlier post about a chronically loose non-drive side crank got me thinking about why it was that the majority of loose cranks that came through my shop were the left side. Let’s rule out tapered pin cranks, and department store bikes. Good quality drop forged aluminum cranks on the correct taper angle and length spindle properly torqued at assembly and then left alone—why is the left side more likely to loosen than the drive side?

So far, the only thought is that the left crank end has less structural mass compared to the right side, and therefore can ‘stretch’ a little more.

What do you think?

mack_turtle

with square taper cranks, my theory is that when you push down on the left crank arm, the arm is moving counterclockwise. if the arm is loose on the spindle at all, it would take the crank fixing bolt with it and loosen it more. if that theory holds water, the right crank arm will tighten the bolt when this happens, which would explain why right side crank arms rarely fall off. if the left side of spindles and spindle bolts were left-hand threaded, I suspect this would help a lot. however, any decently-made crankset that is properly torqued should not cause this problem at all.

how common is this on splined interfaces like Octalink and Powerspline?

Retro Grouch 

Drive side power gets transferred through the chainring?

Andrew R Stewart 

As RG sort of mentioned ,the LH arm sees both directions of torque. The drive forwards is transferred through the axle to the RH side as the RH side's forward movement is placed back to the LH side through the axle. The RH side arm/axle fit really only sees the transfer to the LH side and some localization effort to keep the arm in place.

This topic is but one of the many that shop wrenches (with inquiring minds) talk about when it's slow. Andy

rydabent

Probably from the old natural habit of leading off with your left leg while riding or walking. It is what you do in a parade too.
When first taking off, a big hard push on the left pedal get you going to your balance. The left pedal gets the hard push and therefore come loose first.

dsbrantjr

I think that "properly torqued" is the key here.

davidad

The tapers should be greased and a torque wrench used. Installing Cranks by Jobst Brandt (sheldonbrown.com)

pdlamb

Meaning, of course, that if the left crank is installed correctly it's not going to loosen.

70sSanO

Righty tighty lefty loosey?

Sorry, I couldn’t resist.

John

veganbikes

Ahhh the oft argued debate, grease or no grease! I cannot argue with Jobst Brandt or Sheldon Brown (one they have sadly shuffled off this mortal coil and two they are more knowledgeable than myself) but it is an oft had debate and you will have people on all different sides arguing.

I will however no doubt agree with a torque wrench being used but beyond that grease or no grease I don't know.

reconnaissance

This makes the most sense to me. Put another way, the left crank, in order to drive the chain must transfer torque through the spindle to the right bolted chainring. The right crank directly drives the chain and only needs the spindle to stabilize the rotation. So the left crank-to-spindle joint suffers much more twisting force than the right side. The right side crank-to-spindle joint only suffers twisting when the rider stands on the pedals horizontal to the ground.

Andrew R Stewart 

Yes but... that's not the topic. It's why does the LH arm come loose more often then the RH one. For pdlamb to be correct then we need to believe that all RH arms are installed correctly and some LH ones are not. This is IMO not likely as the RH install methods to be perfect so many times but the LH side not so doesn't seem likely, both being dealt with by the same factory and shop assembly standards (or lack of...) Andy

sweeks

My study (n=1) of Octalink V1 crankset finds an incidence of "zero" loose cranks, irrespective of side.
Same for ISIS crankset (n=1), and square taper cranks (n=3).
Small sample sizes, but all these have been disassembled and reassembled at least once each. All torqued to spec and, FWIW, greased.

davidad

In the refinery where I worked the machinists greased the taper on bits that fit into drills.

pdlamb

I think you're missing my point. As subsequent posters have noted, a properly installed and torqued square taper crank does not loosen. Properly stated, the original question would be, "Why are improperly installed left side cranks more likely to loosen?" (Leaving for follow-up, more likely than what? Properly installed right side cranks??)

Andrew R Stewart 

"why is the left side more likely to loosen than the drive side?" From the OP. I read nothing about proper or improper installation.

Sometimes I think there are a lot of politicians here as people often answer the question they wish was asked. Andy.

Trakhak

In fact, properly stated, the question would be "Why are left crank arms more likely to loosen than right crank arms?" It wasn't until very near the end of the period of ubiquity of square-taper cranks that torque wrenches came into more or less common use in bike stores, so "proper" versus "improper" installation would have been impossible to quantify in most shops. Regardless, that left crank arms have been observed to loosen more frequently than right crank arms comes down to the fact that the right crank doesn't impose a twisting force on the spindle whereas the left crank does.

Incidentally, a tip for those of you who still have square-taper cranksets to service: to set the torque for the crank bolts with a minimum of effort, apply a torque wrench to one crank bolt and an ordinary socket wrench to the other. When you hear the torque wrench click, you know that both crank bolts are torqued to the proper spec. (I've shown this technique to a number of mechanics over the years; some immediately get how this works; others have to think about it for a while.)

Iride01

Does the left side of square taper cranks loosen more than the right side?

Or is it that observations in general include other than square taper cranks? A two piece crank can only get loose on the left side so if that is included in the stats and observation, then that will skew the data if one tries to subset only square tapers.

capnjonny 

All the above shows once again why the Ashtabula crank is the best design. If they would just make them out of titanium we would all be using them.

Trakhak

I believe that we're all discussing square-taper cranksets exclusively. At least, I hope so, for the reason that you pointed out.

Iride01

I know that everyone is discussing square taper cranks exclusively.

What I don't know is if the OP's question is based in fact. Do square taper cranks loosen on the left side more often?

Just because we can come up with seemingly sound reasons for them to do so doesn't mean they do.

reconnaissance

Keeping a post on track can be “...properly torqued at assembly...” is an attempt to eliminate the noise from the good data. It’s as close to normalized as we can get without actual experimentation. Finally, I will tell you that in a bike shop environment, the left side crank, (square taper or diamond taper) is the one most often found loose. Call that anecdotal if you want, but it is my real world observation. Second is data based: do a search on ebay of quality used cranks and you will find more right hand cranks in sale condition than left side cranks. If there are fewer left side cranks for sale, something has happened to them. I’m guessing damaged taper walls.

In the end, thanks to the truly curious and thoughtful folks here, I’ve digested the valid information and that helps me understand the problem better.