I think you are thinking about this too hard.

Assuming that the drive chain is properly adjusted, we would logically expect cassette durability to be the inverse of chain wear. When there is repeated metal-to-metal contact, making one component more durable means that the other component would wear faster. See, e.g., brake rotors vs. brake pads. 8 speed chains are similarly inexpensive as these 3 cassette, so which do you want to replace more frequently?

As for rust resistance, as 70sSanO and alcjphil said, that is generally not an issue with reasonably well lubricated chains. But if you are against any maintenance at all, they you are essentially asking which finish (HG31 zinc phosphate, HG41 nickel, or HG51 chrome) lasts the longest, and I am not that there are any controlled measurements on this aspect.

Tiagra 4700 and Sora R3000 hubs both still appear to be in current production. Both are available in 28H rim brake models.

https://bike.shimano.com/en-EU/produ.../FH-RS300.html

https://bike.shimano.com/en-EU/produ.../FH-RS400.html

I believe that Shimano's WS-RS100 and WS-RS500 wheelsets also are still in current production (not sure about the latter). They're both rim brake and is 20/24 spoke (20 F, 24 R).

No, they're not 105-level or higher. But they are quality road products, albeit not on the "cutting edge", that are quite suitable for many riders and budgets.

Yes, I know wheelsets are available in lower spoke counts - mostly what I consider ridiculously low like 16/20.

Sora hubs are only 8/9/10 speed. Where do you see that Tiagra comes in 28 hole? I build my rim brake wheels 24F/32R anyway.

(sigh) Read the specifications part of the RS400 hub link I posted. It clearly states that the RS400 rim-brake hubs are available in 28, 32, and 36 hole versions.

Sora hubs are indeed limited to 10-speed road cassettes. However, I seem to remember that 11-speed mountain cassettes (e.g., those with a 34-tooth or larger largest cog) have a largest cog that is dished inwards - and reportedly will fit on a 10-speed Shimano road hub. I haven't tested this myself, so I may well be wrong about that last item.

Nevermind. I was looking in the right column because that's where options were listed.

​​​​​​There must be such studies, not just rust resistance but also durability and abrasiveness. Just need a source.

That's what I thought until I bought a set of Shimano WH-R560 wheels a few years ago at an "I can't pass this up" price. These are laced 16 radial front and 20 radial ds/1X nds with bladed spokes and modestly deep (24mm) rims. Yes, that's correct, they are radial laced on the drive side rear wheel.

I assumed they would be rather short lived due to the low spoke count but that was over 47,000 miles ago and they are still in great condition and in daily use on the rough roads in Western PA. They have never needed the attention of a spoke wrench either so I believe their durability has been well demonstrated.

I do not recall any 16/20 rim brake wheelsets off hand, but I have ridden my Mavic Ksyrium Elite 18/20 wheelset for years (at a chunky almost 200 lbs.) without issue, and plenty of people similarly have good experience with the Campagnolo Zonda 16/21 wheelset.

So this is the wheelset I could not recall above. Do you mean radial drive side and 2x on the non-drive side? If so, the Mavic Ksyrium Elite rear wheel is also laced in this manner, which is the opposite of the more traditional 2x drive side and radial non-drive side of the stock Mavic Aksium wheels on my bike.

Maybe straight pull hubs need fewer spokes?

Well I can't argue with 47K miles wo seeing a spoke wrench. You didn't say how much you weigh, but I'm guessing you're not heavy. A big concern for me is if you did break a spoke mid-ride on a wheel with this few spokes, you probably wouldn't be able to adjust other spokes enough to make this wheel stable enough to get home like you could on a 24 spoke or greater count.

Not to mention what is the advantage of a spoke count this low? Any reduced weight by using fewer spokes would have to be compensated for by using a heavier rim to handle the extra load on each spoke thereby making it a wash. Add to that if your tire clearance is already close, a higher spoke count wheel will flex less and be less likely to rub on stiff climbs.

Nope. This has nothing to do with how many spokes you need. The weakest link in a wheel build is the rim.

It's not so much that wear is less on larger gears because more teeth are sharing the load, but that each load-bearing tooth passes through the "wear zone" less frequently.

Assume $0.25 wholesale cost per spoke, $0.03 wholesale cost per nipple, 1 second per spoke for machine install, and a production run of 50,000.

A 16/20 spoke wheelset would use 36 spokes. A 32/32 spoke wheelset would use 64. That's a difference of $0.28 x 28 x 50,000 = $14k in materials alone - plus a time savings (at 1 sec per spoke for machine install) of a bit under 388.9 hours. Not to mention the savings in machine maintenance. Plus less QC time and labor cost.

If the resulting wheelset can be made durable enough to satisfy the consumer, then yes - there is a definite advantage to a manufacturer in producing a low-spoke-count wheelset.

And if they are durable enough and perform well, there's no downside for the consumer. Classic win-win.

Sound like a classic bean counter scenario. Someone probably got a bonus for suggesting this as an idea to save the company x dollars which is really pocket change compared to the company's total profits.

This is a false syllogism, as applied to chain drive. The fact is that while worn chains cause sprocket wear, sprockets don't cause chain wear.

Primary chain wear happens INSIDE the chain as links flex on the pins when winding onto and off the sprockets. This flex is the only wearing action.

Once the chain pitch increases (stretch), each link no longer hinges freely as intended, and the rollers sequencially roll on the backs of the adjacent tooth, causing the characteristic hooked profile.

FWIW all wear happens was at either end of the top loop, coming off the be cassette, and onto the chainring. No significant wear happens elsewear, either because there's either no movement, or no tension.

While I agree that chain wear primarily occurs at the interface between roller and pin, I am not sure that there is no wear at all between the roller and cog interface, especially if you take shifting into consideration.

In other words, would you answer the OP's original question by stating that chain wear would not be any different between any of the 3 Shimano cassettes?

Two points, the second first.

Yes, cassette choice makes zero difference as to chain wear. Zero, none, nada.

The other is that we don't quite agree on the nature of chain wear. You reference an interface of roller and pin, but no such interface exists.

Pin wear (stretch) happens where pins ride within formed journals on inner plates. The rollers ride on the outside of those journals, and their wear isn't measurable as stretch.

So there are two different kinds of wear, one correctly measured with a ruler, and both combined when measured with chain wear tools. That conflating of pin and roller wear accounts for the discrepancies between measuring methods.

"Classic bean counter scenario"? Perhaps.

But remember: commercial entities are not public works jobs programs. They exist to make money. If they don't turn a profit, they don't stay in business - at least, not for the long term.

Also: small increments of improvement add up. Multiply a similar "bean counter scenario" savings of $25k (materials plus machine time saved plus maintenance/QC costs avoided) by 100 different products, and you have $2.5M. At $125k per year for salary and benefits, that amount alone could fund 20 additional junior engineers in a company's R&D section. And I'd guess Shimano makes far more than 100 different products these days.

During SunTour's heyday, they focused on technically superior products. They also priced their items at "reasonable markup" rather than at whatever the market would bear. And Suntour's management made bad economic decisions - which you deride as "bean counting" - on top of that. In contrast, Shimano didn't. And Shimano also devoted approximately 20% of their profits to R&D.

Which company went bankrupt?

Point taken. Though you could also argue that if Shimano used a higher spoke count like 24/28, they could save money on alloy by making the rims lighter, no?

Regarding bean counting just for the sake of getting a bonus, IIRC, there was a guy who worked for one of the airlines who got an incentive for the suggestion that by removing the olive from the salads they serve on their flights, they saved $40K per year.

Yeah, they could. But it would be a false economy, for two reasons.

1. The consumer these days seems generally to want deeper, more aero rims. Deeper rims with reasonable strength means using more alloy. Not much way around that if you're going to produce a durable product. If they don't give the customer what he/she wants at a good price, someone else will offer it - and the customer can (and will) go elsewhere.

2. Alloy in bulk is fairly cheap; it's the fabrication that makes a rim expensive. Given today's CNC-driven production, I'd guess there's very little (if any) manufacturing cost difference between making a 16-hole and a 24- or 28-hole rim (the latter two might even be more expensive in terms of manufacturing time and cost, since they'd require more holes to be drilled). So the additional cost for more alloy is very likely outweighed by the additional spoke/nipple costs.

In theory, a 32/32 or 36/36 wheelset, properly built, should be damn near bulletproof. But if someone can make and sell a 16/20 or 20/24 set that lasts 40+k miles with zero problems and sell it at a reasonable price, then yes: as a consumer I'm damn well interested.

There is also the perception that fewer spokes are sexy. And as we all know, sex sells.

That too. Plus a small aero benefit.

And the pros use aero, ultralight, low-spoke-count wheels when they race. So everyone else should do the same - right? :)

I’m sure companies run tests and have internal data. I’m skeptical that there are external studies because you have no actual coating/plating specs, just descriptions.

The irony of this is that the actual steel wears with use. How in the world will mil or two of any coating have any impact on durability?

John

Yup. The sheeple will follow what the pros do in hopes they will ride faster.

Some coatings do reportedly provide abrasion resistance (the Ti-N coatings used on drill bits comes to mind). But I'm not sure any of the three listed above provide much abrasion protection. I think they're there mostly to prevent corrosion.