kyselad

My wife rides an old Peugeot mixte that she loves. No matter what I do, I cannot get the rear wheel to sit straight. It has horizontal dropouts, so no problem adjusting to get it centered between the chainstays, but the wheel is invariably out of center between the seatstays. The difference is big enough that it makes fender and brake adjustment very difficult, plus it looks bad. I've use a homemade dropout alignment tool to verify that the dropouts are parallel and aligned to each other, and I've installed the same wheel in both orientations to make sure it's not an issue with the wheel either.

I'm sort of stumped. All the info I find online seems more focused on lateral rather than vertical alignment issues. There's no apparent damage to the frame, but it's definitely a cheap mass-market model that may have had loose tolerances for tube lengths, welds, etc. Is there some way to fix this problem, or is it hopeless?

Iride01 

QR or nutted axle? Either way, it might just be that the axle is ever so slightly off center because of the little bit of gap between it and the gap in the dropout. Doesn't take but a tiny amount of the axle to be off to make wheel off by a lot at the tire.

However I'd really suspect a bent axle if this is a bike with a freewheel.

smd4

Is the wheel dished properly?

kyselad

It is a freewheel QR, but I tested a nutted fixed gear wheel and same problem in both orientations. :/

kyselad

Yes, problem is the same even with a fixed gear wheel from a different bike that I know is perfectly dished. Also it seems like a problem with dish would create a global shift right or left but not a discrepancy only at the seatstay?

smd4

Can you post some pictures?

Iride01 

Are you good at making your own fixtures or jigs? Then make some that will allow you to hold the frame with the dropouts aligned properly. If the bike frames vertical centerline isn't vertical, then maybe you can figure out what to bend.

However for any bike whether worthless or worth something, it'll be time consuming for someone to deal with if you have to take it to others. If there is a frame builder in your area, you might make friends with them. After a few rounds of beer and some good conversation they might just take a look at it for a nominal amount.

smd4

If one seat stay was too short or too long, then I doubt the situation can be easily rectified.

FBinNY 

Two likely possibilities that are fairly easy to confirm either way.

One dropout is higher causing the wheel to be non-vertical. You'll need a bubble level or tri-square, and s friend. Start by mounting the rear wheel with the bike on the floor to ensure the axle I'd resting against the top of both dropouts. Brace the rear wheel in a doorframe or corner as close to full diameter as possible so nothing touches except the tire at 11&7. With your friend keeping it braced, check the seat tube to confirm that it is also vertical. (Double check the doorframe itself in case it's off plumb). If necessary, dropout height can be corrected with a flat file.

Or it might be within the frame. Place a yardstick edgewise against the head and seatubes extending to the stay. Note where the end meets the stay. Repeat on the other side. There are more precise ways to do this measurement, but with your bike b so far off, a quick eyeball will be enough.

There are other possibilities, and possible corrections, but the above tests are easy enough and will narrow the possibilities.

79pmooney

Actually it can be rectified quite easily but the danger is making things worse. Say left chainstay is 1/8" longer than the right so the dropout sits close to an 1/8" lower. You could file a little less than 1/8" off the top of the left dropout, doing FB's checks as you go. Of course,if you fil so much off the dropout bends or breaks, you went too far. With brazing skills, additional material could be added for strength at the expense of paint.

I've never done this. I'd listen to those who have before trying. But I've heard of it being done. If this mixte is of the UO-8 variety with the plate steel cut dropout, you are hardly going to cause any issues until you weaken it until it bends and even that won't be catastrophic. (Says he who rode his UO-8 into the ground (and two car doors - crashing perhaps 2 to 4 dozen times on it - 5 crashes a winter was the norm - and bent his right dropout beyond the ability to run a derailleur on it early on. Mild steel, Very forgiving. At 19,000 miles the right chainstay broke. I rode it home and had a friend weld it. 22,000 miles was the second car door, trashed fork and I didn't need the reminder so it got retired.)

zandoval 

The question is how far out of alignment. I recently adjusted a fork/tire alignment by just using a file to file out about 1mm of material on the fork ends. I opted for filing out the fork rather then trying to adjust the bend on the fork leg.

What I am trying to say is consider the amount of adjustment you need. 1 mm of change in axle position can make a big difference in wheel position.

icemilkcoffee 

They sell these drop out alignment tools that lets you visualize exactly what is out of alignment and how:


You can easily make your own version of this with two threaded rods and some nuts and washers:

base2 

What are the odds that the seat stays are a different length or at least welded at the wrong place causing a length differential?

If it was a mixte: What if it had a seized seatpost at one time and whoever freed it, did so by twisting the seat tube there by pulling one seat-stay up/forward and the other down/back? Not that I've ever encountered this. But, but after freeing up a few seat posts by twisting the bike about in a vice, I can imagine on a mild steel mixte it could theoretically be possible.

It's pretty mild steel & there is no guarantee it was built to high quality specs/standards in the first place. I'd determine whether or not the seat tube & head tube are in alignment before continuing. That'll give you reasonable reference points to determine if/how the rear triangle is out of spec.

FBinNY 

BEFORE beginning any kind of remedy, take the time to properly qualify and quantify the problem.

Only after doing so, and getting the complete picture will you have the info you need to consider repair options.

For example, if you confirm it is dropout height, 1mm is easily remedied with a file, though more might call for other ideas.

My signature quotes from Admiral Hopper, and about an ounce of diagnosis..... are there for a reason. Take a short time to measure and think BEFORE starting something that may take a long time to undo.

Andrew R Stewart 

The issue with these types of tools is that they only indicate the faces of the drop outs WRT each other, not WRT the main frame or the fork. It is easy to bend the dropouts so they look right, according to these tools, but the wheel can still sit cockeyed if the stay lengths or main/rear triangle centers are off.

I hope that IMC already knows this but just didn't explain the tool use well. Andy

kyselad

Exactly. I already did a home brewed version of this tool that isn’t as fancy but does the job. The dropouts were already close to aligned, and I fixed the little bit that they were off, but that just means the wheel now sits perfectly in the dropouts, and it made no perceptible improvement to the wheel alignment.

maddog34

Check the FRAME ALIGNMENT first.. THEN align the rear dropouts.
visually line up the head tube with the seat tube of your bent frame... does it look ok? good.. move on the checking the Rear triangles....

see: Park Tool ***-2, Frame alignment gauge
you can make one from wood, nuts, washers, and some thread stock quite easily.
You-tube has vids on their proper use...

If one side of the rear is bent out, the other ends up bent in.. this can lower one dropout, and raise the other side.. causing the rear rim to set closer to one side of the Seat Stays....
this can be fairly common on bikes that were "cold set" to fit a wider rear hub, by people with no ***-2, or an equivalent gauge....
here's another symptom of a home-grown "cold set".. somewhat non-parallel dropouts...... "Exactly. I already did a home brewed version of this tool that isn’t as fancy but does the job. The dropouts were already close to aligned, and I fixed the little bit that they were off, but that just means the wheel now sits perfectly in the dropouts, and it made no perceptible improvement to the wheel alignment."

Any WEAR bad enough to cause a misalignment would be fairly obvious.
Check.. Frame, front to mid, then front/mid/rear.. THEN reset the Dropout width, THEN reset the dropout alignment, then re-check width and frame front/mid/rear... then check it all again, to be sure.

bent axles are easy t see.. spinthe axle and watchthe right end.. that's the end that always bends on an old freewheel hub...

one thing, one thing.... leads to ano-o-o-o-ther... hey hey hey . Props to "The Fixx".

Iride01 

If the bike was ever right before, then why would one be too short or long. If something got bent or twisted, then it can be unbent or untwisted or something else can be bent or twisted.

No?

smd4

Sure, but it does sound like a weird problem--centered between the chain stays, off-center between the seat stays? Not sure a "bend" that isn't (apparently) even visible would cause this. I'd love to see some pictures to make sure what's being described is accurate.

Iride01 

Well not really, it doesn't take a lot of movement at the axle to make a lot of difference at the tire. I don't really expect that the issue is the frame. But if the OP is handy at making stuff, then attaching the bike by the dropouts to something that is known to be horizontal to the ground will show if the frame is off kilter in the vertical. Which is one thing that would cause their issue. With a level, some threaded rod of the correct size, nuts and some 2x4's, all of which I have just laying around, I could make one pretty quick.

However I still think it's something else we just don't know about yet. Perhaps the drop out slot is just wallowed out. And if so, then a fixture wouldn't necessarily show that.

kyselad

I don't know whether this problem is relatively recent or was already there at the time of manufacture, as it was already there when I bought the bike. It didn't have a rack or fenders when I bought it, so it's entirely possible the bike was built this way and whoever owned it previously just didn't notice.

grumpus

Check the head tube and seat tube are parallel. Do the string test to ensure the dropouts are centered on the front triangle. If you have a hefty bench vise firmly fitted to a bench you can clamp the faces of the bottom bracket and tug on the stays until they pass the string test and are correctly spaced in. If not you can support the frame on wooden blocks on the floor and with an assistant stand on it to set the stays. Set the dropouts parallel. Now when you check the rear triangle with a properly dished wheel you know that any misalignment is in the dropouts rather than the stays. You're aiming to get the rear rim coplanar with the front triangle, rather than centred in the stays - any apparent misalignment on the stays is now only cosmetic.

maddog34

they're horizontal dropouts...

frankenmike 

I’d just true the wheel to the frame and call it good if it were me.

smd4

You are correct (I saw “vertical” in the headline and went with it).