Frame Verification and Alignment Surface

For the rear that is very common, to the point where the most commonly used dummy axels are 2 mm wider to compensate. Next time add a 1 mm washer on each side.


I’ll echo everyone else on the dropouts, the seat stays always pull them in. Part of your final alignment should be setting the dropout faces from center so that will correct them.

Definitely work on being consistent. Did you start tig welding just to build bikes? Have you practiced on thicker material? If not, I suggest getting some .125 or thicker material for practice. It slows everything down and lets you learn what you are looking for in the puddle.


Interesting, in my case when I jumped to TIG welding, alignment was much better than filet brazing. Much easier to pull one side or another and much more consistent results. Still need to be smooth all around the weld but it’s a long learning process.
I know how hard it can be when you have to think about everything at the same time during the process, checking alignment, staying focus on your weld, checking again.Keep at it and don’t worry about the numbers, if it look straight it will ride straight. Take note of your process and see what could have been improve for the next one. good luck!


The old tradition was that you were supposed to make your own welding table once you had the basics down. Welding thicker stuff is a great way, as @Neuhaus_Metalworks points out, to get a sense of how liquid metal behaves/learn to see the puddle well/etc without the stress of worrying about blowing holes in expensive bike tubing.



Do you happen to know which BQ this was in?

I agree with @Meriwether that head tube is a good datum. I have an Alex Meade set of tube blocks and one part is a headtube fixture that holds the HT on cones level above my 3’x4’ granite table. This tool is a bit fiddly and it doesn’t work well with tapered HT and some of the new standards but the approach is right. Nice bench centers are the next measuring tools I’d like to invest in.

I also use a BB post for rear triangle alignment and measurement on my surface plate. Documenting your measurements is important. I print and keep my BikeCAD drawings in a binder for each frameset and take lots of notes.

1 Like

Maybe BQ 15 (Spring 2006)? See BQ 15 (Spring 2006) – Rene Herse Cycles

I have an alignment conundrum. I have scenario where when the frame is alignment HT to dropouts (according to Park tool and to center-line check with height gauge on alignment table) the rear wheel is skewed 3- 4 mm from frame in one direction. That is, the wheel is parallel to the frame line but off-center to one side 3-4 mm. The center-line on my DA is at the same height as the center line of ST and HT. AND the thru-axle is both perpendicular and parallel to the frame at BB and ST within 0.3 degrees. The dropouts are also parallel to the rest of frame within 0.2 deg. as best my angle finder can do.

Is 0.3 degrees enough to make 3 - 4 mm difference at the outer edge of a 29’er rim?

I’ve been at this all day, totally baffled as the cause. I don’t know what to chop, re-do, fix, without knowing the cause…

Anyway, I guess I’m hoping someone has a “Once when I… the rear end was wonky and I found that… fixed it”… or, “Hey, moron, did you remember to check… compare the … with the … obviously”

either is fine :smiley: And NO I don’t have another wheel to try on hand. I could maybe find one…

Have you flipped the wheel and checked?


Moved this question to the alignment thread

L = R*sin(\Theta)

29er rim is ~ 640mm diameter

320*sin(.3*\pi / 180) = 1.68mm

Your logic checks out. As @DEVLINCC suggests, If you haven’t already, try flipping the wheel

Are you measuring this 3-4mm wheel offset relative to your ST?

@RHughs that’s weird. I would say - especially if you’re fairly new to framebuilding or you haven’t got too much invested in this particular frame - to maybe just live with it and try to understand where it went wrong so you don’t make the same mistake again (rather than trying to fix it and risk ruining it all). If you have issues with tyre clearance then you could dish the wheel over by 2-3mm (chainline and/or spoke tension might be awkward if the dropouts are offset to the non-drive side though). Getting a frame straight and true is really tough, but to get it straight but not true is even tougher!

1 Like

Thanks for thread move, didn’t see this thread.

@DEVLINCC Yes, I flipped the wheel and same measurements.

@Daniel_Y I am measuring the offset at ST, yes. Using Park frame alignment tool thingy.

I did a further experiment last night and mis-aligned the frame front to back to see if I could get the wheel straight with frame. So it’s mis-aligned from HT to dropout by 6mm and the wheel is off at CS by 1 mm and off relative to SS 2 mm +/- and on the non-drive side. Visually now it’s better but still offset from ST just under 2 mm. Slightly visually offset at SS but looks even with CS visually… until you really look closely at wheel to ST line…

Front end is aligned with itself.

@JoeNation I am really thinking hard along that line! I am new, yes, this is my 8th or 9th frame.


Silver braze a 3-4mm washer inside the offending dropout and call it done. :slight_smile:


Interesting topic! I have extremely poor heat control while welding and I use flame straightening on my frames (aka magic wanding). I began after reading the NJS builder Level using it in Bicycle Quarterly. It’s super effective even on frames, that won’t cold set jumping on the head tube.

This AGA video is my bible:


Thanks for posting this, It may have saved me quite a few hours. I have a misalignment that I need to fix and this might just be the ticket.

1 Like

I’m sure you’ll watch the whole AGA video, BUT, you must always have something to block the movement of the part to the incorrect direction, only then will the contraction of the metal pull your tube towards the direction you apply the heat from.

1 Like

@normaali that is a great reference! Amazing. Thanks for posting.


What’s that thing I’ve read about where the NASCAR guys pour some kind of epoxy in a pan and it settles and dries to a flat tolerance of like .00001 flatness or something?

I’ve heard that it’s cheap and super accurate, and I’ve scoured the usual, forums, instagrams, flickrs, etc. but have never found a framebuilder that has applied this tech to our needs. Anybody ever actually done this or know somebody who has? I would try something like that as a guinea pig, but I have no idea on sourcing materials and methods to actually do it.

1 Like

Its just an epoxy resin. Its great for making a very flat surface but is not tough ie. its damaged easy. Nothing beats getting a cast iron or granite surface plate and its worrh hunting one down. But remember we are making bikes, not precision scientific instruments.


Thanks everyone for sharing. I’ve always been curious about allowable tolerances for pro builders. I’ve spent time stressing around a calibrated granite block and height gauge but now just prefer to get a sense of alignment from my jig and have fun riding bikes.
A few years ago in my stress around a granite block phase I was wondering how much deformation we’re getting just by clamping the bike sideways by the bb shell and ran a quick FEA. It looks like 0.2mm displacement at the headtube. This is a large frame with Columbus Life tubing.

We also used to check rear alignment on the plate with a wheel installed…
Sure this will be symmetric when you flip it over, assumes perfect bbfaces and fixture stiffness, etc. I’m not telling you this is good or bad. Just thought it might be interesting to folks on this thread.