hazzak

]I picked up a old steel road bike a few weeks ago as a project and I think the seat/chainstays are out of alignment. It was impossible to get the rear brake to work because with the rear wheel aligned in the chainstays the rim was about 4mm out at the brake bridge. It didn't help that the wheel had a broken spoke and the dish was out so I tried Sheldon's string method.

It was about 2mm out on one side at the seat tube according to this (obviously the sheldon method assume the rest of the frame is straight!
I also tried flipping it around to rule out the dish of the wheel being the issue as well as using another correctly dished wheel. The alignment is better with the wheel flipped and with the other wheel, but the rear triangle is definitely out of alignment.

It's hard to illustrate but here's a photo of the bike - is it worth fixing/spending money on? Or should I just sell the components?

[

FBinNY 

There are a number of possible causes for the error, but start with using the string method correctly. The string must be run symmetrically to both dropouts. The loop and knot you have on the right side can throw off results. Repeat that measurement threading the string as you did on the left for both sides (through dropout and tied to opposite side. Then measure carefully.

Also confirm the centering at the bridge using a known to be correctly dished with a straight axle. Be sure both ends of the axle are against the top of the dropout. Confirm my flipping the wheel, and reconfirm by rotating the axle with a cone wrench on the left side (a bent axle will push hte wheel to one side).

Now understand what you're measuring. The string method is a quick (but non-difinitive) diagnostic that the dropouts are at equal distance from the central plane. Even if they are, it's still possible for the wheel to be off center at the brake bridge. For example, the stays could be slightly curved, or one dropout may one slightly higher than the other. So go slow before "correcting" anything until you have a good sense of what the problem really is.

ultraman6970

Wonder if this kid has the wrong brakeset in there, long vs short brakes.

Second detail. the wheels you are putting in there are like 120 mm wide vs 130 or 127 of that frame??? the rear end in the yellow bike looks really bad and pretty much is bent in. You can do +- 2 mm but if you are putting a 120 mm wheel in that bike then clearly you will have problems. So far i can see you have the wrong hub spacing in that bike.

Whats the spacing in the frame versus the spacing in the hub?

Good luck with this one, IMO you have the wrong wheels in that frame to start with, thats the root of your problem. (Whats the spacing in the frame versus the spacing in the hub?)

hazzak

I have tested it with another wheel which I know to be dished properly and by flipping the original wheel. The dish of the original wheel isn't great, but there is definitely an issue since all of the above combinations weren't aligned properly. So in addition to the string method, what else can I do at home?

The rear wheel (which originally came with the bike when it was built in the 70s) has a 120mm hub, but the dropout spacing appears to be a couple of mm wider than this (not as wide as 126mm). I presume the rear spacing is supposed to be 120mm since it was a custom frame and would have fit the wheels, so I guess at some point in it's life the previous owner had an accident and bent the rear end?

In the picture I posted, I have a spare wheel from another bike on there. It is a 126mm hub and obviously doesn't fit properly. However, I knew this wheel was correctly dished and was just using it to see how much the incorrect dish of the original wheel was affecting the alignment measurement. As you can see even with a correctly dished wheel it is still out. The reach of the brakes appears to be fine for the 27" wheel it came with ( I think they're Weinmann 500s)

FBinNY 

To answer the original question, it's probably worth straightening. But you need to do an analysis to make sure you've correctly identified the error(s) and the shortest route to a straight frame.

Draw yourself a sketch with the measured deflections and see if they're consistent, or seem contradictory. Once you have a clear picture of what needs to be done, then it's fairly easy to go about resolving it. But you don't want to be working this back and forth trial and error. Steel is forgiving of cold working, but even steel has finite patience.

onespeedbiker

You don't have to spend any money on this, just follow Sheldon Brown's Cold setting directions.https://sheldonbrown.com/frame-spacing.html#spreading; while this procedure is advertised to spread the stays for wider hubs, it can also be used to straighten a frame; something I recently did with my old Bontrager; I grabbed my Bonty to ride with my son and noticed the rear wheel was canted toward the left chain stay (I can't imagine how it occurred unless it was damage from a crash about 10 years ago, but you'd think I would notice it before now). Anyway I flipped the tire and indeed it was still canted toward the left chain stay. This means there is a dish issue also because if it was just the stays flipping the wheel would make no difference. Anyway I used the string method to double check and indeed the stays were out of alignment.

For this/my method, you need a wheel that you know is straight (you can also use the string method but this is faster). Your bicycle looks like the wheel is canted to the left looking at the brake center. This means the stays are canted to the left. You'll want to bend the stays to the right. Thread a 2x4 through the stays per Sheldon Brown. Since you want to move the stays to the left, you will want to spread the right stay and narrow the left stay. As an example, spread the left stay about 2mm (difference between the present measurement and spread measurement). Then narrow the right stay the same distance until your hub can slide in (not to wide, not to narrow). Here's were the straight wheel comes in; mount the wheel and see if the stays are now straight or they need more bending. Once the frame is to a point where the wheel looks straight in the stays, re-check with the string method. This is what I did and it worked great, just make sure you remove the wheels and either remove the rear derailleur or place a piece of 2x4 (or similar) adjacent to the derailleur to protect it during the bending process.

Andrew R Stewart 

I agree with FB that this might not be an issue of the rear triangle being offset to one side. If that was the only missalignment then cold setting would be a good soultion. But there are a few other ways a rear end can be off and cold setting will often only shield the true problem. If one only looks at one point of alignment then that's the only frame of reference that they will talk about. There are three planes of alignment that a rear wheel needs to agree with to be "in line". Cold setting will only adress one.

As has been talked about here recently with both front and rear alignments, one needs to fully understand the various factors and how to make them better. many people who don't build bikes or spend time with full capicity measuring devices (note I did not say any one specific tool) and have used them enough to really know the give and take of aligning will only talk about what they do know. While not wrong in the partial world they discribe the full picture is more involved then the simple one plane understanding they speak within. Andy.

Andrew R Stewart 

I should mention that if you include the wheel as a frame alignment issue there's at least two more planes of alignment to consider. It does get complicated if you don't break down each aspect well. Andy.

JohnDThompson 

Are you sure the brake bridge is mitered evenly? From the picture it appears the left side of the bridge may be shorter than the right. If so, this is a cosmetic issue. A frame that shows 2mm difference in the "string test" is only 1mm out of alignment; well within manufacturing tolerance.

FBinNY 

Actually the string test reads roughly 1/2 the actual error. So if it reads a 2mm (high to low) difference, then each side is 1mm off at the seat tube, or roughly 2mm at the dropout. (This is a highly rounded analysis, the actual multiplier depends on the relative distance of the seat tube and dropouts to the head tube).

Either way. The string method is not gospel. It's more analogous to a screening test in medicine, where a positive (error) would indicate the need for more definitive testing.

Over the years, I seen too many perfectly good frames "corrected" into scrap metal, so I'm a strong advocate of a measure twice, cut once approach to frame straightening.

cny-bikeman

Agreed on the string test. I posted a no-tools evaluation at https://www.bikeforums.net/showthread...1#post14960754 that can provide a fairly clear picture of what is going on with a bike's overall alignment, as well as helping to isolate the problem area(s). It does require two properly dished wheels and should be done with tires mounted. I used this method many times in combination with measurementg to assist in evaluating frames without the need for disassembly or expensive gauges.

One thing I had not thought of is that with a digital camera and tripod one no longer needs quite the vision this used to require.

neurocop

I understand that you tried a different wheel and had a similar allignment problem, but I'm not sure the rear stays are "bent." Looking at your picture it seems that the wheel needs to be offset a few mms to the non-drive side. Couldn't you effect this by installing sdifferent axle spacers (thinner on the non-drive side and thicker on the drive side)?

Andrew R Stewart 

One aspect of rear end alignment is whether the wheel sits vertically or not (One of the planes of alignment I mentioned before). A well centered wheel will show this. Assuming that the wheels the OP has tried are properly dished and the rim sits well centered between the chain stays then the drop outs/seat stays are not aligned. One drop out is higher then the other (or one seat stay is shorter then the other). Playing with axle spacers to get the rim to center between the seat stays will only transfer the "offness" to the chainstays and further make the wheel's contact point with the ground off line.

Like I said before if you don't fully understand frame alignment issues you need to listen to some one who does. Andy.

Captain Blight

If you have a 4-foot level and a dial caliper, you can get a more accurate measurement using the same technique as the string method. I recently did this to my Trek 616, my long-suff'ring all-rounder. It never had ridden right since it'd been in a crash with me a year and a half ago. I'd thought that I'd re-aligned it fairly well, but... not really. SO I wrapped a piece of tape around the head tube as a reference point and another one around the seat tube so I could get repeatable measurements. I then clamped the level to the head tube and the dropout, measured the distance between the seat tube and the level, then flopped the arrangement and repeated the procedure. Turns out I'd been off by a little over 5mm!!

So Ai sat down with pencil and paper and committed arithmetic for a while, came up with how far each side had to move to center the rear triangle and maintain my 132.5mm spacing, and went to work with a piece of plywood (on the seat tube, to spread the load) and a 4' length of handrailing. The actual coldsetting only took about 5 minutes.

So now the frame is aligned perfectly side-to-side, within a millimeter. The difference is startling. Before, I had a shimmy that would start up at about 15 mph; and riding no-handed required me to shift my whole self off to the left a good little bit. Now I can stay centered on the saddle and run downhill no-handed without any shimmy at all. I run out of courage before the bike gets unstable.

If you don't have a long level or other solid straightedge, I might possibly use a fluorescent tube and duct tape. They are the most common really straight thing around, with tolerances of something like .0001 per foot. *I* might do this, but probably wouldn't: Mercury vapor is A Very Bad Thing. You shouldn't even think about it.

You can get good results with the string method; but you've got to be able to get the string really, really tight, and it takes some care and thought to do this without pulling the dropouts in or out.

Hazzak, what is the make and model of the bike? It doesn't look to be of dreadful quality in your photo. Got any more pics?

Andrew R Stewart 

Let's see what's wrong with this.

"If you don't have a long level or other solid straightedge, you could probably use a fluorescent tube and duct tape. They are the most common really straight thing around, with tolerances of something like .0001 per foot. Just be careful, mercury vapor is A Very Bad Thing."

Fragile and dangerous device with point contacts against hard material. Really! Can you be any more irresponsible? Oh yes, You did let yourself off the hook by mentioning it's "a very bad thing". Now I feel soo much at ease.

I really try to stay away from making my posts personal. But this one takes the cake. IT'S NOT ACCEPTABLE FOR THIS KIND OF ADVICE TO BE POSTED.

Captian Blight- Please don't go any further.

I'm out of this one. Andy.

Captain Blight

What part of "be careful" escaped your attention?

Be careful. Pay attention. Proceed cautiously. Move deliberately. Think.

Use your head.



Now, that being said, I wouldn't say that this should be your go-to tool. I forget that many, even most, people aren't as careful or deliberate as I am. Good tight string would more than suffice.

FBinNY 

I'm with Andrew here. Not that I'm a great practitioner of caution, but using a risky method carefully isn't a good choice when the hazard is easily avoided in the first place. I simply can't imagine that using a fluorescent bulb as a straight edge could ever make sense, no matter how careful you think you are. There are just too many alternatives around.

Besides, what's straighter than a string under tension.

Captain Blight

Uhhh... A fluorescent light bulb. Geez, the more I think about it, the worse the advice seems! Because it's round, you'd have a heck of a time getting an accurate reading off its edge. Kind of a bad idea all the way around.

Kimmo 

Well, I for one find the fluoro tube straightedge to be a good tip.

It might never come up, but then again one day it might just save the day.

As for safety concerns, duh - it's a fluoro tube! Handle with care.

dedhed

So you're say'n we shouldn't have all those sword fights out in the barn when we were kids with those 4 footers?