Rear wheel alignment
 

I have a carbon frame road bike, and noticed that the rear wheel is out of alignment in the chain stay. It is quite closer to the left side, to the point that it is almost rubbing. The strange thing is that it is properly center in the frame above at the rear brakes. I have checked the wheel dish and eliminated that as the problem. Any ideas?

Was this wheel ever centered? Is it uniformly out of center throughout the sweep of the wheel? If so, redish it or change spacers so it's centered.

If the wheel dish is good, then it sounds like the dropouts were machined badly. Is there enough play in the dropout that you can make it centered ?
Is this a new frame with a warantee ??

There is a little play in the drop out and I can get it fairly close to center in the chainstay if I tweak it a little. However then it goes out of center at the seat stay. The frame is approx. 1 year old. I never noticed the mis alignment until recently, so I don't think that it has been like this since new

First of all re-mount the wheel by setting the bike on level ground, then opening the QR tapping the wheel so it settles home, and closing the QR. This is the wheels natural position and if you get into the habit of always mounting the wheel on the floor it'll always return to the same place.

Now check for centering in the chain and seat stays. If it's off center, especially in the seat stays, it could be a mis-aligned frame, or dropouts not perfectly positioned, or a bent axle.

Begin the process of elimination by opening the QR and rotating the axle in 90 degree increments and testing all 4 positions. If the wheel doesn't move the axle is fine, and since you checked the wheel's dish it might be the frame. Before going crazy, be aware that many frames don't have symmetrical chainstays so it might be perfectly fine. Also, double check for dish by mounting the wheel backwards (cassette on left). The rim should be in the same place, otherwise it is the wheel after all.

To check the frame, make a small loop in a piece of string and slip it over the QR skewer. Loop the string around the seat tube, and bring it back to the the other side. Pull it tight and close the QR trapping it. Measure from the string to the rim (if the tire touches remove it).

Repeat the process with the string just above the BB, half way up the seat tube trapped over a bottle boss, and at the top trapped above the top tube. All three pairs of measurements must be equal right and left (though not equal to other pairs) within 1-2mm. If not, than the wheel isn't centered within the rear triangle, something that can be fixed by filing either dropout, but you need to know what you're doing.

Note this test only determines the wheels position in the rear triangle, and if this is off it might indicate other frame alignment issues, so you might want to have someone check everything.

FB,
Thanks for all the good info. I performed all the steps that you recommended. Everything checked out except the measurement when the string was over the BB. The wheel is closer to the left side by about 4mm. It doesn't make sense to me that it is only closer at the bottom??

Double check all the measurements. Make sure that the string is only around the tube and not affected by any welds or lug flanges. If it's symmetrical at the top but not the bottom the wheel is out of plumb, ie. not vertical. There are many possibilities so you have to check everything.

If you own a bubble level and a repair stand you can do some more measurements. Mount the rear wheel, clamp the bike in the stand and orient the frame so both the seat tube and top tube are level per the bubble level.

This establishes the central plain as level to the earth. Now place a broomstick or any straightedge across the wheel on a secant so it touches the rim but not the spokes or tire. Since the wheel should be in the same palne as the bike, this should be level in all orientations. If it isn't you can figure out how to file the dropouts to correct.

BTW- If you lack a repair stand the same can be done on the floor using books, or other stuff to support the frame so it's level. To reduce the height needed remove at least on pedal, and turn or remove the handlebar.

This test can also be done in a vertical plane with a right angle bubble level, but front/back and top/bottom have to be analyzed separately, and it's a bit harder. If you have a good head for spacial relationships, all kinds of measurements are possible with the bubble level, string and straightedge.

Check the rear dropouts...using the "T-Handle" type tool with adjustable cups on the end....in addition to all of the above...

=8-)

Not sure what you mean by "T-Handle" type tool with adjustable cups

He's referring to tools like these, which are used in pairs to make sure that the dropouts are parallel with each other. You fit them into the dropouts and work them so that the two cups line up and are equidistant from each other all around the rim. Non-parallel dropouts can flex the axle and stays moving the wheel when the QR is tightened.

These are Campagnolo, Park makes a much less expensive pair.

ok thanks..I guess I'll have to talk with a mechanic that has these, rather than spend $90.00 for a one time use

If you have vertical dropouts and a bit of skill, you can check for dropout parallelism with a home made tool that'll cost you just a few bucks. Buy a 3/8" threaded rod about 8" long, and 2 nuts and washers for it. Attache the rod to one dropout so it just barely comes to the other. Tighten it and look through the other dropout. The end of the rod should be perfectly aligned with the dropout. Repeat from the other side. You won't be able to bend the dropouts using this because it isn't strong enough, but it'll serve as a decent gauge.

BTW- I don't think that non-parallel dropouts alone could cause your problem, is seems to be a more complex frame alignment issue. If you opt to have the LBS check it, make sure it's one used to dealing with high end road bikes with a mechanic that knows about frame alignment. You seem to have a carbon frame, so if the error can't be fixed by filing the dropout axle position, you're SOL.

ok FB thanks again for all the help