gregm

4,000 miles on this Phil Wood rear hub. Of course, their legendary warranty & service has come through without a hitch, so I'll be sending off the wheel for repair, no sweat. Sooo glad I bought Phil.

Meanwhile, enjoy the pictures!

-Greg

p.s.: no real story here, just ridin' the bike.

Phantoj

You certainly look on the sunny side of things!

specialist

4000 miles doesn't sound very good for the life of a freehub, especially for a phil wood, which I guess wasn't cheap.

TandemGeek

Something is amiss at Phil and it seems to be getting worse, not better. I've still got a set of Phil Wood hubs from '98 that have 12k or there abouts and they went back to Phil twice for rework. One of my builders has never had anything but trouble with Phil Wood hubs and only put them on bikes when a client really pushed for them and was unrelenting even when given his history of the hubs. In the off-road world, they have the worst record of all the tandem hubs, and failures extend to failed hub bodies... the most recent one I saw having 3 cracks radiating from the drive side shell opening into the center of body. What's disconcerting is that most of the Phil failures seem to involve a single pair of pawls failing which seems to suggest that engagement of all four pairs of pawls isn't happening uniformly. This was the problem with my hub(s) back in '98, and it's what we're also seeing in your hub.

It's good that Phil continues offer excellent customer service and warranty, but IMHO when it comes to durability / reliability they've been displaced by Chris King which are truly the Top Shelf tandem hub these days. I've been partial to White which has a nice product for a mid-range price of late. Shimano's HF08 are the price performers: they aren't the lightest or sexiest tandem hub, but they're very affordable and as durable as all but the Chris King and don't require nearly as much care and attention.

Just some thoughts....

zonatandem

Phil Wood was the hallmark for decades. First used Phil hubs, bottom brackets and pedals (yes, he made a great sealed bearing touring pedal) in 1977 on our custom Assenmacher tandem. Put 64,000 miles on the Phil components without any problems except for bashing up a pedal on a then new-fangled speed bump crossing the border from Arizona into old Mexico.
Next used Phil hubs and bottom brackets on our Colin Laing custom tandem. The stoker's BB got some play in it after 30,000 miles (pushy stokers can be tough on equipment!). Sent it back to Phil Wood with a note explaining problem and mileage. Got a new BB with note from Phil saying that after 30,000 miles we deserved a new one, NO CHARGE!
That is exceptional customer service and backing up a product.
On the next tandem, a Co-Motion, we specced Phil hubs only and used Syncross BBs. Never an issue with either for 57,000 miles.
Currently using Chris King hubs and headset . . . 18,000 trouble free miles so far.
Sorry to hear that the Phil freehub only lasted 4,000 miles.
Pedal on TWOgether!
Rudy and Kay/zonatandem

oldacura

We must be weak. About 10k miles on our rear hub with no issues.

I understand how manufacturers could intend to have multiple pawls share the load but in reality, it seems that only one at a time will really be taking the load.

TandemGeek

Assuming it's a good design and manufactured to proper specification, why would only one of three pair (White Industries) or one of four pair (Phil Wood) of pawls only be properly seated and engaged in the ratchet teeth?

A rear hub works like a spokes wheel in some respects: the entire purpose of having multiple sets of pawls or engagement points is to evenly distribute the loads around the hub, and the more engagement points you have the lower the stress becomes on any one point. Chris King, for example, uses 72 matched sets of engagement teeth in their main drive rings, which speaks to their distinctive freewheeling 'buzz' as well as their ability to handle massive amounts of torque.

Again, something's amiss if all the pawls are not engaging the ratchet rings with equal force and it will eventually lead to pawl, ratchet, or even hub shell failures, any of which makes for a long walk home.


Just my .02

specbill

Our moderately powered, 305 pound team had somewhere around 20K on our early 90's Phil hubs that never missed a beat. I just retired the rear to upgrade to 9sp compatible Phil hub, other wise I would have ridden it another 20K. ' quite surprised to see anyone with issues as they have been nothing but bullet for us.
Bill J.

TandemGeek

That's the rub... It's either hit or miss in some cases and the perceived quality based on past history makes it hard to believe that their could be any problems.

If I hadn't had my own personal problems with 3 out of four hubs that I've owned and been privy to failure reports accumulated by builders or dealers who have had a lot of come-backs, I would also think that problems were rare.

Thankfully, their customer support, warranty, and willingness to take care of long-time clients has been so darn good that it's still a good bet that you'll either have a hub that has no problems or, if it does, they'll resolve it in fine fashion.

However, problems are still a pain to deal with since they seem to happen at the worst possible time. Moreover, hearing that others have had 10,000, 20,000 or 30,000 trouble-free miles does nothing to soften the blow when one does fail.

gregm

A view from the outside of the same hub:



-Greg

TandemGeek

Looks about normal for a loose-mounted cogset where the driven cog bears all of the drive torque.

Coincidentally, since we're be discussing distribution of loads, the Campy MK 2 cassettes or Shimano models that use cogsets where the cogs are riveted together in clusters or, in some cases, mounted to a splined spider distribute the loads from the driven cog across a much larger surface area that eliminates those nasty looking intentations. They're a pain to clean compared to loose cogs, but they don't tear up the cassette carriers.

oldacura

I haven't closely inspected a freehub but my understanding is that multiple pawls sit in recesses in a hub that is fixed to the wheel. The outer end of the pawls engage teeth in the inside of the hub that is driven by the cassette. Unless all of the pawls are exactly the same length, the recesses in the wheel hub are at exactly the same radius and the teeth on the inside of the cassette hub are exactly the same depth, (e.g. zero tolerances), only one pawl will bear the load & the rest will not quite touch. The pawl that bears the load will be the longest one that engages the tooth with the smallest radius. This will change depending on the angular orientation of the two hubs.

Is my understanding incorrect?

TandemGeek

At least for hubs that use pawls, you've got it 180* around: The pawls are nested in base of the cassette carrier body and the ratchet teeth are usually installed as a subassembly that is pressed into the hub body. Even though Phil Wood will sometimes be cited as having 8 pawls, what it really has are four pawls that are split down the middle to accommodate the spring that give you four "pairs" of pawl teeth. There are 20 ratchet detents into which the four pair of pawl teeth engage. Assuming the hub's parts are produced to spec and the hub is assembled correctly, the pawls and ratches should be concentric. Once the rear hub is under load, all four pawls should engage the ratchet and share of the total load. If they're not, then one or more of the pawls is hanging up due to contamination, the parts aren't concentric as they should be, or there is an excessive amount of deflection occuring within the hub, which is also problematic.

Bottom Line: A single pair of pawls should never be bearing the entire drive train load; it should be evenly distributed around the hub body by all of the pawls or engagement teeth. If the loads aren't being evenly distributed, the hub won't last very long.

Phantoj

I don't really understand what he's saying here, but, apparently, you have to expect that only one pawl will end up bearing the entire load:

https://yarchive.net/bike/freewheels.html

swc7916

Why are freewheels/freehubs designed with pawls and ratchets? Hasn't anyone designed a one-way friction clutch mechanism that doesn't click? I would guess that it has something to do with weight, durability/reliability, simplicity, space, or price - but it seems that it would be doable.

TandemGeek

There was a typo as it should have been "If the loads aren't being evenly distributed, the hub won't last very long."

Instead of reading about loose bearing freewheel hubs (this would be the thread-on freewheel types of hubs to which Jobst refers), take apart a few current sealed bearing cassette hubs and see for yourself how the engagement systems work. They are very different designs and, yes, it was mountain bikes that seemed to force the design change to more robust engagement systems that have benefited tandem riders ever since, as have many bicycle component changes brought about by mountain bike technology.

In theory, all the pawls should be engaging the ratchets and sharing the load as a wheel rotates, even if one pawl momentarily shares a higher load at some point during each rotation of the wheel due to deflection or elasticity of the materials and parts.

prairie*boy

Mark/oldacura:

Mathematician versus engineer, the loads aren't exactly balanced...but close enough for practical purposes.

As old acura points out manfacturing has tolerances and this results in one pawl taking more load. But the tolerances should be kept small enough, proper assembly (put the short pawl in the small hole/long pawl in big hole - so be careful dis-assembling/re-assemling to put everything back exactly where it came from) and initial break-in wear serve to minimize the variances. Also, under load grease flows out faster from the higher pressure pawl hub resulting in a more even load.

So at the end of the day, one pawl might see 30% of the load, one - 27.5%, one 22.5% and the last 20%...But this is better than 100% on one. Greg's pictures show the components post break-in...the load has evened out by deforming/work hardening the casstte carrier and is now being carried by all the splines.

So if one's homework (e.g. I have designed the pawl/hub to take 60% of peak load) was done correctly lasts the expected life of the product (a couple hundred thousand fatigue cycles). But if you let it run without grease, take it apart and re-assemble it slightly different, ride more than expected, load it higher, have the manufacturing/assemblling process get out of spec, you will see failures.

My day job involves designing machinery - trading off: expected product lifetime with cost/weight, stress calculations, tolerance stack-up checks, assembly procedure writing, checking for patent infringement and failure analysis. The satisfacton of getting it right keeps me coming back...although watching the factory floor make mockery of my precise calculations has left me crying on occassion...

Greg:

How big is your team?; How much do you use granny/how low is it? Do you tend to spin or mash? Curiosity is getting the better of me.

Prairie*boy

mrfish

interesting discussion - gives rise to some interesting maintenance strategies varying from leaving things alone until they die through to regular meticulous disassembly and rebuild. Personally on modern components I am a fan of leaving things alone until they die then replacing the cartridge bearings, or alternatively getting older components with grease ports so that an occasional grease injection suffices.

TandemGeek

No, a mathematically dyslexic shade-tree mechanic who has owned and disassembled a lot of different freewheels (SunTour, Shimano, Miche, Campy, etc..) and cassttee hubs (Edco, Coda/Hugi, Phil Wood, White Ind, Shimano, Chris King, Campy, MAVIC and Asian-made-your-brand-goes-here) and who has seen quite a few different hubs that have failed for a variety of different reasons vs. some commonly held beliefs based on what others have heard said or read on the internet.

The point was, in theory, multiple pawls and engagement points are used to distribute loads around the hub and if that gets out of whack bad things usually begin to happen.

TandemGeek

Actually, most of these hubs are relatively easy to service, the Phil Woods in particular, and without undue risk from even a novice home wrench. They don't use cup and cone bearing systems that require just the right "touch" and have few if any adjustment points: you simply align the parts and bolt-up the axles. Even the highly complex Chris King hubs are user serviceable, but only to a point. Each brand of hub is a little bit different so don't assume that you've learned or read about any given hub is directly transferrable to another. However, the documentation provided with them and/or available on their Website is sufficient to allow most home mechanics to undertand how their hubs work what is required to keep them properly serviced. Therefore, if you put a lot of hard miles on your tandem you owe it to yourself to read the technical data that pertains to your hubs and wheels: it could save you a lot of trouble down the road.

The last thing I'll say on this matter is, when something's amiss with hubs there are usually audible or tactile indications... uneven sounding pawl action, drivetrain slip (noting that chains don't usually just jump one link... think hub), unexplained sqweaks, creaks, or grinding sounds particularly when you are climbing or driving through corners and putting lots of side loads on front or rear wheels. Every hub that's ever given me trouble -- or that was determined to need some type of service -- has foreshadowed the impending problem: Coda/Hugi, Phil Wood, White Ind, MAVIC and even the Chris Kings.

TandemGeek

Quick follow-up: I've conferred with a few more folks who are "in the know" on current hub durability and quality and Phil Woods have definitely be surpassed by several other of the aforementioned tandem hubsets. They may have been "the best" at one time, but that's a hard place to be in a very competitive and development-driven industry.

If you have them, be attentive to any new issues or sounds to preclude any on-the-road break downs vis-a-vis preventative maintenance. The customer support is still exceptional as is their warranty / product coverage, but if a pawl wasn't properly hardened or a hub is a bit out of spec those two things won't prevent a failure.