Why the dealer tendency to chose sprockets with even number of teeth?

[silva]

The number of beers you have drunk?

I've read an unclear sentence and I asked for clarity.
If the sentence was unclear due to beer or whatever I don't know, that's a question to its writer.

Btw my (drinking) menu is short - the liquid part is just water from distribution. Since some decades.

 

[silva]

I think it's the case that a 47t chainring is an infrequent choice across the cycling world, and confess I couldn't understand why a 46t or a 48t would not suffice. Most people are not that picky ("my usage standards") and how would a 48t or a 46t "void the frame"?

I don't get it, can you explain it more clearly/detail?

When I was on the lookout for a new bike, I wanted the same gear as on my previous (52/16) and the dealer of a possible good for me - bike told me that the frame of the bike didn't give enough room for such size chainring, and that 48t was the max. After weighting pros and cons I decided okay - not less than ratio 3, the minimum I had decided.
I have now (again) new cranks and the previous replacement the chainring mount sat closer to the frame - the dealer had to replace the axle with a longer one to compensate.
The (other) dealer now thought, but wasn't sure since he recommended a tryout first, the 1 tooth smaller 47t would not hit
the frame.
If you order an adult bike from a dealer, and the dealer sells you a kid bike, are you then "picky" if you don't agree?

 

[silva]

According to Sheldon Browns information legacy on the web, chain elongation only takes place on the inner links, the outer ones are "bridged" by the outer plates and plates don't "stretch", with the elongation occurring under tension, due to worn-off material between the link parts contact surfaces.

According to several dealers I asked, elongation is due to plates stretching so permanent deformation of the steel grade, which would require a force stretching it beyond its elastic range.

Sheldon Brown calls an even number of teeth a benefit (on the wear aspect, see https://www.sheldonbrown.com/chain-life.html ) based on aboves first - the chains elongation is every other link, and in the case of an even number of teeth this allows teeth and link to wear-in so properly mesh with eachother, while in the case of an odd number of teeth there is, in a degree, a random engagement which causes improper meshing and rough running, which would thus decrease chain life.

Both statements lead to the same, whether it is the worn off material, or stretched plates, there is an elongation every other link and thus an argument for even numbered teeth sprockets.
However, in a case of a combination both, in regards to the wear related question even/odd number of teeth, one would compensate the other and even/odd would cease to matter.

An integer gear ratio, alike that 48/16 (which is according to what I've read on forums a quite common one) I had, proved itself as a by far bigger cause of wear than aboves. Those are even numbers of teeth, but 45/15 or 51/17 is also integer and should result in a similar wear concentration.

My gear choices are limited due to chainring not having more frame clearance due to bike design / accomodation for 62 mm tyres in combination with a bottom bracket located eccenter to tension the chain.
My gear options are:
chainring/rear cog/chain length (links)/eccenter remaining range at mount of new chain:
- 47/16/106.58/10.6mm - gear 2.94
- 48/17/107,6/10.8mm - gear 2.82
- 49/16/107,6/10.8mm - gear 3.06
- 49/17/108,1/7.5mm - gear 2.88
- 50/16/108,2/8.3mm - gear 3.125
- 50/17/108,6 /10.8mm - gear 2.94
The higher the remaining range, the better, the eccenter is specified as a 1/2" or 12.7mm range.
Any situation with a low remaining range means that I have to take out a link couple during the chains normal lifecycle.
The 48/16 I had isn't listed since no option for a variety of reasons including a near-worst-case (at the front) of this initial position.
Odd numbers links are also no option since that would require halflink chain usage which has drawbacks of weaker, harder to use joint links and much more expensive. At least that's what I have read around.

So this all explains my 47/16 choice. I just don't have any better option, smaller is unacceptable, and achieving it along a smaller rear cog means faster wear but also problem with mounting it on the disc brake mount - the bolts would void the chain running over the cog, and only along some special provisions (countersunk bolt heads) possible. Less than 15 is just not possible.

The only somewhat possible alternative would be a 50/17 with chainring and cog (replacing the initial) mounted along a chain that already wore enough to give the chainring the frame clearance it lacked at a new chain. But frankly, lol, I have other things to do than wasting my time on such poormans ersatz-solutions.

 

[Heltor Chasca]

Nope just nope. Odd numbers on anything would cause anyone with any slight OCD tendencies to spontaneously combust.

 

[silva]

In the land of the blind, the one-eyed man is king.

 

[Sharky]

53-39 is referred to as a ‘standard chain set’ 50-34 as a ‘compact’, 52-36 as a ‘semi’ or ‘Pro compact’ other ‘oddities’ don’t have these terms applied.

52-42 seemed to be the norm when I first joined a club, coupled with a 5 speed block (14-18).

 

[silva]

Was it YellowSaddle explaining it somewhere? Only wear on inner (or outer?) links would be responsible for chain elongation, which causes a wearing-in (which causes a lower further wear) every other link, with that wearing-in being voided by uneven numbers of teeth.
So 1 option I now consider is to further reduce the bikes gear ratio from 47/16 to 46/16. But only an option if the bottom bracket eccenter chain tensioners' initial position doesn't move forward, inflicting me again need to take away a link pair during the chains lifetime. A smaller sprocket means either initial position moving forward either chain 2 links shorter (so an entire pitch 12.7mm being also the eccenter range). My current chain 'space' was measured as 106,25 links (so 106 links) so there is a little free room for the eccenter to be moved backward - the opposite. If I'm not wrong somewhere. A 46T would then bring back that other problem. But if a 46T would further reduce chain / drivetrain wear, I could live with having to throw a chain in the bin before its lifetime end.

 

[wafter]

The pins, ramps and gates on chainrings are timed for changes to happen most easily when the cranks are vertical, when there's least tension on the chain.
If there's an even number of teeth, the pins etc ate identically positioned on each side of the chainring, so manufacture is a matter of turning the ring 180 degrees and repeating the same operation(s). With odd numbers of teeth there is no such standard angle, so, as rrk says, costs go up, and a small increase in manufacturing cost is a considerably larger increase in sale price, so competitiveness suffers.

As for 47T chainrings, TA make them (Spa Cycles in the UK)

An interesting an insigtful explanation, however surely all the big boys would be running CNC / automation now so such issues shouldn't really present a problem any longer? Also, it seems that the RRP of chainsets of different tooth-counts within the same model range are the same; suggesting that the cost price would reflect this and as such offer no incentive for dealers to stock one over the other.

I think it's more down to people's irrational love of even numbers..