9:Zero:7 Chain Rub Problem Solved

This is a particularly long and nerdy post, so if you're not all that interested and wanna blow it off, I totally understand. Catch ya next time. If, on the other hand, you're needing a good nerd fix, read on . . .

I haven't talked much about it on this blog yet, but the process of buying my fatbike has been long, problematic and frustrating. By "process", I mean the period of time between placing the order (Oct 19th) and having the bike in my possession, assembled and working correctly to the point that I feel it's really ready to go get busy, which is today (Dec 18).

During a bike hang a few weeks ago, I was talking to Joe2 and I mentioned that I was having some problems and he told me that he thought I really should let people know, because based on what he had read on this blog, he had assumed that everything was smooth as silk and that if he or someone was interested in a 9:Zero:7, I was doing them a disservice by not telling the whole story. I think he made a good point, but I also knew that I was still processing it all and wanted some more perspective before I'd be ready to write about it. I guess I got ready enough this past week, because a new thread that appeared on the fat bike forum at MTRB.com prompted me to reply with the details of my experience. If you're interested, you can read about it here. I'm not going to re-hash those issues in this post. What I am going to talk about is the chain rub issue that I have had, and what I found out in the process of correcting it.

Chain rub is a common issue with fatbikes, because you're trying to cram a lot of tire width into a space that's limited by the chainline, which is constrained by the rear wheel's OLD and a reasonable Q Factor. That's a bit of an oversimplification, but for purposes of this discussion, it's a sufficient statement.

So within the relatively new world of fat (the Surly Pugsley, which was the first version of what we currently know as fatbikes, came out in 2006), some workable standards have developed. The standard I opted to base my purchase on was a 170mm OLD rear spacing, combined with a 100mm bottom bracket and an FSA Alpha Drive, and Surly's Endomorph rear tire. I thought that this was a sufficiently tried-and-tested setup that I wouldn't have any chain rub worries. So I was disappointed to find that I did. And oh yeah, it bugged the hell outta me. So I set out yesterday to find out what was causing the problem and to see what I could do about it. A lot of the detail in here is included for the benefit of those who might be coming over from MTBR.com because they're experiencing similar problems - I had a lot of trouble finding this kind of data.

One of the things I wanted to know was whether any part of the problem was caused by frame misalingment. In order to find out, I needed first get some good, accurate baseline measurements.

This photo shows the problem - in the lowest gear, the chain rubs on the tire both above and below the chainstay.
You can see the wear mark developing on the tire after just the few miles I've ridden the bike so far.

I checked clearances in some of the other gears, just to get a sense of how much space I would need to gain back in order to eliminate the rub and have a little clearance.

Upper chain clearances:
0 mm - Largest cog
2 mm - 2nd cog
3.5 mm - 3rd cog
5 mm - 4th cog

Lower chain clearances:
0 mm - Largest cog
1 mm - 2nd cog
2.5 mm - 3rd cog
4 mm - 4th cog

Based on these measurements, it looked like the upper was just barely touching and that the lower was interfering by .5 to 1 mm. Ideally, I'd like to have at least a couple mm clearance, so that would mean moving the chainline over by 2.5 to 3 mm at the tire.

I also wanted to check and see how well the tire was centered in the frame.  Just eyeballing it, I could tell it was close to being centered.  I found piece of steel to use as a spacer that was about the right thickness and used it as a gauge.  It was slightly tighter going in on the drive side and I used feeler gauges to estimate that the  difference in gaps was .5mm or maybe just a little less.  So this means that the tire might be sitting .25 mm closer to the chain than if it were perfectly centered.  So not much of issue here.

Checking disc-side clearance between the tire and chainstay

Checking clearance on the drive side, with the same spacer.  It was slightly tighter on this side.

I also wanted to get an accurate width measurement of tires at different pressures, so that I would know if running at higher pressure contributed to the problem, as I have heard account of fat tires growing quite a bit with pressure.

The 27 TPI Surly Endomorph on the rear measured as follows:
20 psi - 100 mm
15 psi - 100 mm
10 psi - 100 mm
5 psi - 99 mm

It has nothing to do with the chain rub issue, but I measured the front tire, a 27 TPI Surly Larry, while I was at it:
20 psi - 101 mm
15 psi - 101 mm
10 psi - 101 mm
5 psi - 101 mm

It surprised me that the width of these tires was so consistent over this pressure range.

I measured between the chain-to-chainring and chain-to-cog points of contact to get an idea of the geometry involved with moving the chainline out at the chainrings vs moving it out at the cassette. This distance is about 18 inches, and the point of contact between the chain and tire is about 6 inches from the chainring end. What this means is that for every mm that you move the chain out at the cassette, you only move it off the tire by 1/3 mm. But for every mm you move it out at the chainring, you move it off the tire by 2/3 mm.

Next, was something I was really interested in - whether there appeared to be any frame misalignment. I clamped a straightedge to the rim, rested it on the bottom bracket shell and then measured from the edge of the straightedge to the edge of the shell, on both disc and drive sides, with these results:

8.3 mm - drive side
8.7 mm - disc side
0.4 mm - difference

This slight amount of difference was in the same direction and approximately the same amount as the difference in gap between the tire and chainstays.  This makes me think that the wheel dish is off by this small amount (I forgot to check this with a dishing gauge when the tire was off the wheel).  All in all, my conclusion was that the frame was very well aligned and that frame misalignment had very little if any contribution to the chain rub problem.

At this point, I wanted to focus on solving the problem by spacing the chainrings farther out from the center of the bike, rather than spacing the cassette out - the common method of doing this involves removing one of the smaller cogs and adding a spacer behind the largest cog. The effect is that you lose a gear, but move your largest cog away from the centerline of the bike by a distance equal to the distance between cogs.

The bottom bracket is an ISIS type and I thought I had the right splined tool to remove it, but the inside diameter of the tool wasn't large enough to fit over the spindle.  Luckily, Glen was in his shop and hooked me up with the right tool, along with a spacer, which I thought I was going to have to make on the lathe.  Awesome.

Difference between the ID's of the tool I had (L) and the tool that Glen gave me (R).

Drive side of the bottom bracket before removal.

The black piece below is the bracket for the e-type derailleur and it clamps between the bottom bracket shell and the flange on the bottom bracket cup. The plan is to add another spacer adjacent to the bracket, thereby moving the whole bottom bracket and crankset away from the centerline of the bike on the drive side.

Disc side of the bottom bracket before removal.

When I removed the bottom bracket, I left the disc side cup in place, so that I could make sure it would have enough adjustment to accommodate for the new shim. By eyeballing it, I could tell that it would, but in hindsight, I wish I would have bottomed it out on the threads and taken some measurements so that would know how much adjustment is available in case I ever want to add more shim for whatever reason.

Looking into the bottom bracket shell from the drive side.

Disc side bottom bracket cup.

Below is the drive side shot of the bottom bracket after installation of the shim. The shim thickness is 2.7 mm, so this should move the chain outboard at the tire by 2/3 x 2.7 = 1.8 mm. Initially, I was looking for 1.5 to 2 mm, so this should do the trick.

Disc side of the bottom bracket after installation of the shim.  If  you compare to the picture before removal,
you can see how much farther inside the shell the cup sits now.

I needed a spacer of approximately equal thickness for the derailleur upper mount point and ended up using three 5 mm washers.

Not sure why blogger decided to rotate this picture, but you get the idea.

Ghost auto-rotation in effect here as well.  This shows the upper chain-to-tire clearance
after adding the shim - it's about 2 mm.

This shows the lower clearance, which is about 1.5 mm.

Back in business, sans the rub.

BTW, I measured the offset of the cranks with respect to bike centerline before adding the shim and it was .6mm to the disc side. With the addition of the 2.7 mm shim, the offset is now 2.1 mm to the drive side.

It took a lot of time to go through the bike and figure out what was happening and then apply the fix, but I'm really glad I did - it feels right now. Hopefully this is the last of the hassles with the initial build and setup of this bike and I can go put some miles on it now.