gyozadude

From your perspective it may make no difference, but in reality, and in basic Physics, it makes all the difference. Counter rotational moment caused by precession can only be exerted along the axis of rotation and only against objects that are fixed or rotating counter to the spin of the axis of main motion. The crank arm fixing bolt/nut spins with the BB spindle. It's not moving counter and it's not fixed like the cups. Hence, no precession. To say it doesn't make a difference is simply incorrect.

Yes. Exactly. However, the BB spindle and the bearings do move relative to each other and fixed cups of the BB. Hence, there is where there is precession and we have moment exerted counter to the rotation. Please see Onespeedbiker's explanation above with diagram and this gives us the reason why the drive side cup is left hand threaded and why the left pedal is left hand threaded.

There is no feeding through of torque except if the bearings are going to spin counter to the BB spindle's rotation because it's sandwiched in a gap between the BB spindle cone and the cup, the counter rotation cannot simply exist on its own. For any action, there is an equal and opposite reaction. This translates, that for any moment, there is an equal and opposite moment. So if the spindle spinning forward while cycling is the action, the bearings rolling with the spindle with spin and precess and exert and opposite reaction to the moment. Hence, the drive side cup is left hand threaded for that reason. But not the crank fixing bolts which do not spin relative to the spindle.

Yes, granted, left hand crank issues are more frequent. But not for precession reasons. There are a number of reasons left hand crank issues are more problematic.

a) One legged Roy Rogers mounting and dismounting puts more stress on that side
b) the left hand crank arm has much less metal at the taper than the right side which has the crank spider
c) mechanics will take care to torque the drive side down properly because it has chainrings, but neglect the left

But I've personally never damaged a left crank arm due to improper installation or letting it get loose. It simply doesn't happen when properly installed. Certainly, we've had many issues in the past with European bikes with ISO taper getting replaced by JIS spindles and because there was less metal on the left side crank, the looseness and deformation that led to failure would happen primarily on the left. But if properly installed, it's never been an issue of the left crank arm requiring a left-threaded fixing bolt or nut. I've simply never had one come loose and warp if properly installed.

sreten

Hi,

See my above post regarding most of your post.

Yes. The left crank doesn't need a reverse thread because properly
installed there is no precession applied to the nut/bolt holding it.

Once things get loose and chewed up a bit that is no longer true.
Tightening the nut/bolt can't make the crank seat properly and
it will come loose quite (or very) quickly due to the precession.

The damaged crank taper ridges need to be filed back to flat
for it to seat properly and solidly on the BB square taper.

rgds, sreten.

gyozadude

Wrong. The picture is 2D, but sufficient to explain that it's all in the axis of rotation coming out of the diagram. Just answer the question what part has bearings and is spinning in the same axis relative to something else fixed or rotating at a different frequency in the same axis and this will help pinpoint where precession may occur. Crank fixing bolts do not undergo precession because the crank is rotating at the same angular velocity at the fixing bolt as the spindle. No precession. No moment. But BB cups relative to spindle? Yes! Pedal axle relative to crank arm? Yes! Which is why the drive side BB cup and left pedal are left hand threaded.

You are too caught up with "precession" as some moment. It is not. It's an artifact of observing the net motion of bearings moving when sandwiched between two loaded surfaces. The force and moment on the spindle is still the primary driver of motion, and the end reactionary surface to counter the moment are the BB cups and pedal axles. And for those reasons, the right BB cup is and left pedal axle are left hand threaded. They don't move once tightened because the precession tightens them and keeps them tight.

BTW, the highest pressure at some contact point is mostly radial on the BB cup with the bearing in between and the cone surface of the spindle. The precession is due to the bearing's ability to spin and cause a counter rotational moment (a second order torque of significantly lesser magnitude) on the cup. This torque can be overcome by higher friction of a tightened Italian BB which is right hand threaded on both sides. However, because of years of experience of these Italian BBs eventually loosening, newer English threaded BBs went to a left hand drive side cup which is self tightening. But make no mistake, the precession doesn't generate a massive moment. It generates a steady moment and over time, due to vibrations, it is likely to lead to loosening of any screw thread that tightens counter to the precession. The freewheel and freehub cover plate over the internal bearings and pawls is usually left hand threaded for exactly this reason, so that freewheeling (coasting without pedaling) self tightens the mechanism. Note that if the freewheel is cheap and made in place with low quality controls and is missing a ring of bearings, then very quickly, the locking plate spins off and the freewheel slides off the center and looses any other bearings, which I've observed twice in the field in recent years.

Sreten, you seem to understand basic concepts that point you in the right directions of pedals and BBs. But the application of left hand thread to counter the precession's tendency to unwind a right hand threaded fixing nut or bolt is specific to where precession will actually occur. There are equations that govern precession analysis and you can estimate the torque. However, they all depend on relative angular velocity. Where relative angular velocity is zero, there is no precession and therefore no moment/torque. It's that simple.

Retro Grouch 

Uh - so what chain lube do you guys recommend?

gyozadude

Once a left hand crank bolt comes loose, again, it's not really precession. It's more like racheting action. The downward stroke with left crank forward has outward force because most of us are duck footed. This means under greater stroke force, we push the left side out while pushing down. This means the crank arm, due to the taper is sliding down against the bolt and contacts the bolt. It rachets the bolt looser on the down stroke on the left side. As it comes around, there is less force and if the crank is loose and wobbles back a few degrees on the spindle, there is less outward force meaning the bolt will not equally tighten. So the left side will tend to get looser and looser. Solution: torque the left side to proper spec. The crank was never supposed to move relative to the spindle taper.

If the goal is to say you can fix damaged left side cranks, then fine. But it's not a fix that would probably work for a large rider like me, and it isn't likely to work if the metals of the BB versus the crank arm are very dissimilar with the BB being much harder and able to continue to dig and widen the much softer crank arm hole. But I'd be more than happy to take the crank arm nut on a BSO with steel crank arms and soft steel spindle and over tighten it a bit more and see that fix last a long time and also claim that I can fix crank arms that have come loose. But it won't work with every setup, and most higher end setups use aluminum alloy cranks or CF with alloy inserts and they must be torqued to spec, and notice, they don't even attach with tapered spindles any longer and for good engineering reasons. Some toasted cranks might be repairable. But most are not after a certain point. The whole thing here is to avoid damaging them in the first place.

gyozadude

ChainL? :-) Sorry. Just getting caught up in a terrestrial physics discussion is all. Haven't had one of those in a long time, since I graduated to doing computational physics on Thermonuclear detonations. My terrestrial mechanics is a bit rusty.

sreten

Hi,

No I'm not. I've always been good at physics not because I can spout
loads of text book stuff, it is because I understand it and can use it
to cut to the quick of what is actually going on, usually, and will be
the first to admit to an error given a more insightful analysis.

Contradiction and statements of alleged misunderstood facts isn't analysis.

rgds, sreten.

demoncyclist

sreten

Hi,

c) Hindsight is a fine thing, learn to check your cranks,
especially a new bike in the first few months. PIBTC.

a) and b) : My point was if b) appears to be case just by tightening the nut
/ bolt and it simply won't stay tight a) is maybe applicable, but not certain.

rgds, sreten.

I fixed my very toasted folders left crank.

onespeedbiker

BTW, I have been riding bikes for 40 years and never had a left crank come loose (I simply torque them to spec), but I have heard anecdotal information that the left crank does tend to loosen more than the right. Have you ever wondered why the bicycle industry has addressed precession with rotating junctures such as the bottom bracket and pedals with reverse threading, but have completely ignored your discovery that it also occurs with non-rotating junctures, such as the crank arm attached to a square taper spindle? I understand how precession works with counter rotating junctures as the bearings that are rolling against the threaded cup of a BB or the spindle of pedal in a direction that will unscrew them, but how would it turn a bolt when there is no bearings counter rotating against it to unscrew it?

sreten

Hi,

Its not a "discovery" and I 've already outined my take on the subject.

If you can pick holes in what I actually said be my guest.

Don't try to say I said something that is clearly nonsense.

But I'll try again :

If you don't care you have got the wrong end of stick as to where the
precession occurs and want to just insist I have that is up to you.

I'm happy that I know / understand what is going on.
And I worked it out without referencing JB at SB.com.

It has nothing to do with the bearings. It occurs between two threads.

rgds, sreten.

cny-bikeman

Let's just end this - Sreten is right and the rest of us who disagreed, as well as "JB" and the sources that say precession requires clearance between the interacting parts are wrong. I'm sure there are lots of folks out there who agree with him and are just afraid to be put down by a bunch of pompous, clueless people using the wrong end of the stick.

sreten

Hi,

"JB" completely agrees with what I've been saying.

Making up arguments along the lines that I'm saying
"so and so" when I'm not is pathetic. You are wrong
that I am wrong. Your arguing for arguments sake.

rgds, sreten.

onespeedbiker

Perhaps you missed the point that in the entire paragraph JB was talking about the junction of the pedal and crank arm, not the crank arm and axle. In the same article he discusses the bottom bracket and axle, but still no discussion of the crank arm and axle. https://sheldonbrown.com/brandt/left.html The issue here is not the forces cause that cause a left crank bolt to loosen, but does that force fall within the narrow confines of the definition of precession. JB defines precession by saying Again an object rolling inside a circular ring; where is the crank bolt rolling inside a circular ring? It should also be noted that JB is not the final authority on many bicycle related information. When he sates in the same article, he seems to particular ignore that Italian BB have been around for over 100 years, and while they are less forgiving than a reverse thread English BB, by simply torquing the BB cup to spec, they pretty much stay put. Also something as common as dirty bearings can put a drag on a pedal, cancelling out the the precession and unscrewing the reverse threads.

sreten, I know you are intelligent and analytical, I'm not being sarcastic here, obviously the forum disagrees with you, but we won't change your mind and you won't change ours. This needs to end. Hopefully a moderator will but this to bed, because it's not going to get any better.

Retro Grouch 

I love it!

If OP had asked the time, these guys would argue about how to build a watch.

sreten

Hi,

The bolt or nut in or on its thread is the surface the precession takes place.

Precession cannot occur if the crank is correctly mounted on the BB.
The above surface does not see any rotating torque. I suggested it
does if the pedal is loose on the crank. More to the point if the crank
has been loose and is damaged beyond a certain it will not seat properly
on the BB no matter how tight the nut bolt is set and will still be "loose".

I even more certain now than when I suggested it, under "loose"
conditions, its precession that undoes the tightened nut/bolt.

If you don't agree that is up to you, but don't make out I'm
saying things or implying things that I'm not, just to make
disagreeing appear to more reasonable than it is.

I'm happy to have the veracity of what I'm actually saying debated.

And happy to agree to disagree with those that don't see it the same way.

rgds, sreten.

sreten

Hi,

If you want to repair a busted watch, rather replace it, how to build it matters.

The debate started about the efficacy of repairing a "toasted" crank,
as such, understanding the issues of the mechanism is apposite.

rgds, sreten.

cny-bikeman

I thought physics was not much a matter of opinion. How about instead of debating the veracity of what you are saying, or you disagreeing with others, you provide one credible source that verifies precession as an artifact of the interaction of two threaded parts. I'm pretty sure you would not be the first person to discover that phenomenon.

onespeedbiker

Have you considered that when the left crank comes loose the crank bolt simply ratchets off as the crank arm rocks on the spindle?

Booger1

It has nothing to do with percession,nothings turning......There's an axle,a crank arm and a bolt...non of them turn in relationship to each other.

A bike pedal is totally different...you have the pedal,bearings and an axle...all turning in relationship with each other.....different animal.

The pedal is rocking on the taper....on the right it will tighten itself....on the left it will loosen itself.....As you pedal your bike,you put forward pressure on the crankarm...That's the way the bolt is threaded,so it comes loose......if you pedal backwards,you'll put rearward pressure on the crankarm and it will tighten.

If the bolt was left hand thread,we wouldn't have these discussions,because when you left it too loose or it didn't seat well,it would tighten itself and not screw up the taper in the first place.

You really should retorque anything mounted on a single taper,it doesn't always settle down perfectly,especially when new....there's a mating period between the 2 tapers until they are happy with each other.

A single taper is going to settle in,so you need enough torque on the bolt in the first place to have some left over after it settles.(or just retorque it)

You need to keep checking it until it holds torque(couple times) then it will stay tight.

You don't have to file it to try and fix it,just keep torqueing the arm down until it holds torque.It will broach the hole out all by itself.It's already done it once,it can do it again.It's going to be a race between getting it to hold torque and having the bolt bottom on the end of the axle and cracking in the corner of the tapers....whichever comes first will determine the outcome.

sreten

Hi,

Well if you can explain the ratcheting mechanism I'd consider it.
(I simply can't see one for a loose crank and a tight bolt/nut.)

Someone suggested it "ratchets off" and it was not "precession" with
nothing to back up the assertion other than it is a (poor) explanation.

From my my current perspective there is no ratchet, there is precession.
Its acts like a ratchet, not quite, but in effect, enough to confuse some.

rgds, sreten.

gyozadude

I thought the debate over.

But consider omega (w) the angular frequency that is relative to two or more co-axial components. Precession can happen if bearings are between them (and it must be rotating) such that the bearings counter spin while in the coaxial gap. The relative differential in angular velocity will follow a function fn(w) where the analysis of precession and its affects are well behaved (i.e. fn(w) is continous, mathematically linear (not to mean constant slope or scalar constant but as opposed to non-linear) and is mostly monotonic). So engineers can estimate, say, a cadence of 90 RPM with a 1.37 inch diameter fitting with bearings some certain diameter will generate precession, and with various friction factors will estimate some quantity (min and max) of torque that will exist long term over the life of that fitting. And given a long time, they can tell us what torque to specify to make sure it stays tight against the combination of constant vibration coupled with torque from the precession. (E.g. How long with Italian BB at 40 ft-lbs torque assuming some standard coefficient of friction? It can be done and Italian BBs were used for a long time).

Now try to describe precession as a function fn(w) for a crank arm at the tapered end of a BB spindle with the nut. It's not supposed to move. Well, okay, it's loose. Okay, so tighten it so it doesn't move. You're saying, no, let's assume it's loose and not tight. Okay, I'll ask again, what is the fn(w) that describes the precession? Well (w) oscillates and goes positive and negative. In fact, it may be discountinous. Okay, no monotonicity. No continuous function. Non linear. And dude, there is no bearing that is spinning in the gap between the two layers that cause the damn counter rotation which allows to the less than obvious precession. Yes, precession causes COUNTER rotation that is less than obvious to the average person. It's caused by bearings seemingly rotating in the gap between two concentric objects under load and friction with each other, causing a counter-rotation transfer of moment to the other surface beneath.

We're dealing with a nomenclature issue. Some of us are being sticklers for the use of "precession" to describe the crank loosening due to racheting. It is NOT precession. There is no bearing spinning causing counter rotating moment transfer between the loose crank arm and the bolt. Instead, go back and read my previous post on how it works and why it's racheting due to preferential outward force against the face of the bolt on the downstoke. Is it that hard to visualize what I mean by preferential outward force against the face of the bolt on the downstroke?

Step 0. Crank arm is loose on tapered spindle
Step 1. Person steps harder on the down pedal stroke. THUNK.
Step 2. Crank is loose and sliding out against the bolt still holding it in place and puts pressure against the face/head of that bolt
Step 3. Bolt works loose and attempts to turn WITH direction of downward force (not precession - this isn't counter rotation)
Step 4. THUNK BACK. Upstroke has less pressure as the loose crank slides back against spindle so it moves bolt face opposite direction but with LESS pressure.
Now go back to Step 1 and repeat.

Over time, racheting the prefers the direction of crank motion on the power stroke with tend to unwind a right hand threaded bolt on the left side crank. The key is that this isn't precession because we see no bearings with rotating between a gap in two coaxial components. So engineering-wise, there is no need to put a left-hand screw into the crank arm on the left side. Bolted on correctly to the right torque, the crank never moves. No thunking. It never comes up until we use a crank extractor.

The objection here is the use of the term "precession" in the exact sense. You can choose to use the word "precession" and then describe the Physics and we'll all know and agree, yes, what you said was what you described. But the term "precession" has a specific meaning (e.g. function of differential angular velocity between two coaxial components being well formed to cause bearings to rotate). Loose crank arm isn't precession by that definition. You can choose to rename it that way. Fine. But let's not make it confusing for the bookish who've studied classical mechanics. So from hence forward, can we agree, we'll have "classical precession" which is just real precession. And we can have "Sreten's Precession for Left Loose Crank Arms." And then I think we can all rest.

Will that work?

sreten

Hi,

That is an inane comment given this thread.

rgds, sreten.

sreten

Hi,

I'm completely fed up with the personal denigration i have been subjected to
in this thread. I am describing classical precession. Real Scientists / Engineers
might add something useful to this thread but so far its a bunch of pseudo
- academics arguing the toss about something they don't really understand.

A lot in the personal commenting style of all those who know not much.

rgds, sreten.

onespeedbiker

Assuming you are not just leading us all on, what is "classical precession" and how does it unscrew a bolt, with an none rotating joint?

Wikipedia defines precession as
The above paragraph defines two separate processes; precession and fretting. In the precession definition it involves the counter rotation of a joint with two rotating axis; the easiest example being the right bottom bracket cup, where the axle is turning clockwise while pedaling but the bearings are turning counter clockwise; it is the bearings contact with the cup that turn it counter clockwise, which would unscrew the cup if did not have reverse threads. Now your definition seems to be different because there is no counter rotation of crank arm; the motion of the crank bolt would be the same as the crank arm, either clockwise or counter clockwise depending on the direction that has the most torque. When I said ratcheting, what I meant was if the left crank arm/bolt is loose, every time the crank arm is pedaled forward there would be a small amount of movement of the crank arm on the square taper spindle turning the bolt along with it unscrewing the bolt counter clockwise. When you later coast with the left pedal to the rear, the crank arm would move clockwise (or down) returning to it's previous position but the bolt would not move with the crank arm due to the reduced torque or load on the crank arm and bolt head. Then the next time you pedaled forward the crank arm would again move forward and with the added torque or load, the bolt would once again move forward with the crank arm and unscrew the bolt again; this process of ratcheting would continue until the bolt had completely unscrewed itself.

So what you have is the difference of load and torque ratcheting the crank bolt and unscrewing it, rather than the above definition of precession that requires two orbiting axis to unscrew the threaded cup via the bearings that are in contact with the cup and turning it the same direction as bearing are moving. It seems to me your are calling the above ratcheting process precession, which it is not because there are not two axis contacting each other and rotating in opposite directions.