Monday, October 12, 2020

Easton R4 freehub service

Yes, you can. No, you shouldn't.

I have 2 sets of ~2014 Easton EA90 wheels with R4 hubs. They're cheap, generally reliable, easy to service, and when they're working well, will spin for DAYS. Both sets came to me with 10-speed R4 freehubs, and both of them are running 11-speed upgraded freehubs.

One set of these wheels came with my first (used) road bike, and the other I picked up at Performance Bicycle before they imploded for $300 because of a broken axle. That broken axle gave me a chance to learn all about how these wheels are put together, and where/how to source parts for Easton R4 hubs, which is important because these wheels love to eat bearings (no joke--I've spent more in bearings than I have on the actual purchase of the wheels).

At the actual axle interfaces it's not a big deal: 6901s & 6902s are pretty easy to source, and you don't even need to go with expensive stuff: it's not going to last more than a couple thousand miles anyway--why bother with ceramic or higher ABEC certifications?

But the freehubs? Man, what a pain. Easton designed these to be unserviceable. And expensive. At $80 a pop, these things are the most expensive component on my bike by size and service interval, with each lasting only about 2500 miles. Across 2 wheelsets, I've gone through 5. Those $300 wheelsets aren't so inexpensive now.

So after slapping on a brand new chain and heading out for my first ride in a light rain, what should happen but the chain goes slack while freewheeling. Classic Easton freehub failure. I came home and tried to order freehub #6, and lo and behold they're no longer made. Classic Easton parts availability.

I had tried once before to figure out the inner workings of these things, but to no avail: I couldn't press out the cartridge bearings, and bashing them with a dead-blow hammer just resulted in a pile of little balls on the bench.

But my options were now limited by availability of parts: try again to figure out how to service these bearings or buy new wheels. Option B is a non-starter because I'm not buying more rim-brake wheels, which means it's either fix these wheels or buy a new bike. Once again, option B is probably not a realistic choice at the moment.

Literal cut-away
What I do have, though, is the original 10-speed freehub and a cutoff wheel. I put that old bugger in the vice and sliced it in half to reveal the inner construction, and the reason why I'd been unsuccessful the first time around: the dreaded circlip.

Long the bane of auto mechanics the world over for its uncanny ability to shoot across the room to never be seen again, this is the first time I've encountered their ilk on a bicycle. Luckily I still have my circlip pliers and good lighting.

Bisecting the old freehub also revealed that the spacer between the inner & outer bearings should be easy enough to push aside to gain purchase on the outer bearing.

You're gonna break something here. Be sure to break the thing that costs the least to replace.

Unfortunately the only real way to remove that outer bearing is with striking force on the inner race, which is pretty much guaranteed to destroy the bearing.


So if your goal was to just replace the inner, you're probably out of luck. Do be careful not to apply the striking force to the spacer because you will damage it. Once you've smashed the outer bearing out and dropped the spacer under your workbench, you can get that horrid circlip out. It's a beefy one in a deep-set groove, so make sure your tools are in good working order.

Then it's time, once again, to smash out a bearing. And I say smash because, once again, you're only going to be able to apply force to the inner race. It's a horrible design. And this bearing you get to smash twice because it's the same size as the outer one, so you have to align it to the outer's shell and force it through there, too.

At this point it might be convenient to have a quick discussion about bearing anatomy. Here's a pic:

Sirs "Not Appearing in This Film" are the seals that sit on either face to keep the grease in and the dust & water out. Water is the sworn enemy of a precision bearing. In an ideal situation, pressure is applied principally either from the outside-in or from the inside-out, but always (always ALWAYS) vertically. Much care must be taken to ensure the load on the bearing aligns with its intended purpose. There are, for instance, angular-contact bearings that beef up the shoulders for transverse loads. These are not they. There are also needle bearings which replace the balls with roller pins (or needles--hence the name) for precision loads. These seem like they'd be ideal for bicycles, but no load on a bicycle is truly precisely aligned (Hambini explains this amazingly well in his anti-GXP rant), so ball bearings are the best bet.


Who TF thought this was a good idea??
Once you've made it this far it's time to go online and order a pair of 15267 bearings. I went to Amazon, but be warned: there aren't a ton of 15267's out there (mcmaster doesn't have anything, which is rare). Also be warned: what you get may be utter shit.

I pressed in the Amazon bearings, reloaded the circlip, re-aligned the inner sleeve, and mounted it up, and lo and behold...the freehub wouldn't turn with the wheel. Well crap. Pulled it back apart and soaked the pawls in bar & chain oil just to be sure there wasn't some weird resistance in the actual freehub interface, but no: nothing. It would turn, but only by hand.

So I smashed those shitty bearings back out and gambled on the amount of damage done to the originals during disassembly, because dammit I want my bike back, and amazingly, the damage wasn't *too* bad (there are 2 little crunches per wheel-rev, but a good firm twist of the wheel still takes about 30 seconds to spin down).

The trickiest part to reassembling these is that there is no preload adjustment between the inner & outer, but the sleeve needs to sit firmly in place between them, with nothing "sleeving" the outer races. Go too light and the sleeve flops around; go too tight and it applies lateral force to the shoulders, which will eat the races. That means Easton presses these things in with a precision instrument that I do not have, and I ended up pressing & testing about 10 times before I got to a point where I dared not add any more pressure.

The "right" pressure is functionally unknowable, but you'll certainly know when you exceed it (bearings won't turn), and at that point the only thing you can do is press the outer back out and start over.

At this point I'm not interested in trying to source additional 15267's. Better ones would cost more than half the price of the whole unit (even though that's no longer available). I think I'm done chasing this wheel. It doesn't help that the aluminum end caps on these axles are butter-soft aluminum and now ALSO discontinued.

Long story short? You can do this! I believe in you! But don't bother, because it's a waste of time. So now I guess it's time to start shopping for a new bike while these bearings slowly eat themselves.

Friday, September 11, 2020

WTRL TTT - New team, same great taste!

 Last night saw my return to Zwift racing after a multi-month respite. I've been focusing on the great outdoors and back-filling Zwift rides as recovery efforts, so I haven't really been doing the kinds of efforts required in a team time trial, and I hadn't really cared much about it until I got a surprise invite on Tuesday to race with the CRYO-GEN Eagles.

That's not a lot of time to prepare, and I have been enjoying a LOT of cookies this summer.

I was able to shed 2 lbs by race day--not back to spring fighting weight, but enough to make the watts-per-kilogram numbers a little less upsetting, especially since the target power was 5.5 W/kg. That's a big number, and well beyond my threshold. The plan also had each rider taking 30-second pulls, and we were running a full crew of 8 at the start, so there should have, at least in theory, been plenty of time to draft below threshold between pulls.

But the plan had a harsh reality to overcome: the circuit was my 2nd least favorite Zwift course: Harrogate. Blech. Gross. Literally only the jungle is worse. Gimme a drag race up Ventoux, or boring laps of Tempus Fugit. But Harrogate? It's just false flats all over the place, and that long drag up Otley Rd is so punishing. The downhills are barely long enough to recover, and the climb to the finish is really tricky to time right. I don't care for it.

Anyway.

We got out of the pens and organized far more quickly than with my previous team, and rotations settled in early. One rider was performing team car duties excellently, calling up rotations and keeping overall calm. Otley Rd claimed an early victim, but our remaining team of 7 soldiered on to the first climb at a crushing 4.55W/kg average. I was on the ragged edge for most of the race, struggling to find consistency on the front, and struggling to get my heart-rate under control OFF the front.

By the time we'd done Otley Rd on the 2nd lap, we were down to 5. With a minimum of 4 riders needed for a team result, we were determined not to lose anyone else, just in case someone had a network/power/vmechanical issue. That meant big efforts and a couple of skipped pulls, which really sucked to have to do. We gave up 22 seconds on the 2nd lap, which is pretty daggone respectable given that we lost 2 riders early. But the way we gave that time up suggests the strategy could use some refinement.

Let me be clear: this is a FAST team with some very strong riders. Everybody on the squad was in A, with 2 A+ riders in the group. Looking through Zwiftpower's sign-up list, everyone had recently held at least 4.3W/kg. And our lap times showed that: we were over 2 minutes faster in our first lap than the Vultures were back in February. So power was NOT an issue here.

But looking at the lap-over-lap analysis, it seems trying to hold a consistent power may not have been the best approach to this course.

The first lap saw an average wattage of 296, with a stupid high peak of 560, and an average heart rate 1 beat into Z5 for me @ 175.

What's fascinating to me is what the 30-second intervals did to my body compared to the climbs. From the turn onto Otley Rd to the sharp drop on Pot Bank, my heart never comes out of Z5. I wasn't frankly sure I'd survive the first lap, and even with a brief reprieve into the very tippy top of Z4 on the descent, I was pushing on max HR before we even hit the first KOM run. This is not a recipe for success for me.

But then things get interesting. I was able to settle into exactly 5.5W/kg for the first half of the KOM, letting off the pressure just slightly on the flatter middle plateau to 5W/kg to bring the group back together, and while my HR soared to 183 for the 2:40 climb, it actually recovered very well on the descent, hitting its lowest number for the rest of the lap @ 168.

The group was, at this point, disorganized and gasping for breath, but we held a consistent 280W--still slightly over 4W/kg--to the end of the lap. Looking at the top chart above, you can see we lost no time lap-over-lap after the KOM, and in fact my average power and heart rate were identical between them, again with 2 fewer riders.

To me that suggests we were letting the terrain be a greater influence on our output than choreographed power targets, and the data in the visual above proves it out: when the ground goes up, so does my heart rate.

Because we allowed that natural use of terrain only after getting shattered on the climb, it further suggests that approach could have been employed to greater effect on the front half of the course.

For instance, our first run up Otley Rd took 7 riders 2:54 with my average power sitting at 317W. The 2nd run, gasping for breath and down 2 more riders, was completed 7 seconds slower @ 3:01 & 306W. 11 watts, smaller draft: 7 second penalty. That trend continued as we pushed through the lap, except I was having to skip pulls to stay with the group. 

By the 2nd false flat (finishing Otley & onto Pot bank), we were down to 3 or 4 riders pulling, with a 4-second loss and an overall power loss of 31W compared to the first lap. That's a big drop-off, with almost identical losses on the 3rd lump.

Conversely, we were losing almost nothing on the descents, irrespective of power output.

And that's where our race was decided: on those 3 false flats in the first half of the lap. Big crazy power numbers on the first lap, deflated riders on the 2nd.

But then we hit the KOM again and pretty much stopped losing time, and looking through my data, it seems that while I was off from the 5.5W/kg on the first half of the first run, I wasn't off by much. A 22W loss brought me to 5.1W/kg, again backing off slightly to identical first-lap numbers after the mid-point. Both 5+W/kg intervals were just over 70 seconds, suggesting that's my sweet spot for bigger numbers. And just as with the first lap, the following descent saw my lowest heart rate and best recovery of the lap.

Next time I think I'll advocate for longer turns on the front with target variations for the terrain, and be happy to get either.

All in all I was extremely pleased with our performance, and thrilled to be able to lay down the watts after a 3 month break from Zwift races. I need to add it back into my regimen, but figure out how to do it without turning back into a 1-hour specialist, because outdoor races WILL come back at some point (please please please).

Total: 39:26 / 26.5 mph / 290W (4.46W/kg) / 175bpm
Lap 1: 19:22 / 26.8 mph / 296W (4.55W/kg) / 175bpm
Lap 2: 19:43 / 26.3 mph / 285W (4.38W/kg) / 175bpm

Otley Rd Climb (1 mile @ 3%)
Lap 1: 2:54 / 20.7 mph / 317W / 172bpm / 7 riders
Lap 2: 3:01 / 19.9 mph / 306W / 179bpm / 5 riders

Otley False Flat (1.1 miles @ 1%)
Lap 1: 2:23 / 27.8 mph / 313W / 177bpm / 7 riders
Lap 2: 2:27 / 27.1 mph / 282W / 172bpm / 5 riders (no pulls)

Yorkshire KOM (0.74 miles @ 5%)
Lap 1: 2:40 / 17.2 mph / 331W (357 / 309) / 178bpm
Lap 2: 2:42 / 17.0 mph / 321W (331 / 309) / 178bpm

KOM top to finish (3.1 miles @ -0.7%)
Lap 1: 6:23 / 29.2 mph / 281W / 174bpm / 7 riders
Lap 2: 6:19 / 29.5 mph / 282W / 175bpm / 5 riders (back in rotation, sprint finish)