What happened here?
 

Situation: dropouts already attached to the stays. I cut out the caliper mount holes and brazed them in yesterday. Put the stays back in the jig and I see the new gap at the BB (see picture below). Note that prior to brazing in the caliper mounts, the NDS stay was as tight as the DS stay as in the picture. It's as if the NDS stay shortened by about 1mm while I brazed in the mounts.

I'm trying to understand how that happened exactly. The mounts were tightly fit in their respective holes and as mentioned the miter at the BB shell was also very tight. Now I'll have to shorten the DS stay to match, which I suppose isn't a huge deal, but not something I wanted to do in the first place.

Maybe in the future I should braze in the caliper mounts after I attached the stays to the BB?

Thoughts are very welcome. Thanks.

Brandon

https://cimg5.ibsrv.net/gimg/bikefor...41ad27109f.jpg

I have to admit I don't really understand how steel moves like that due to a heat cycle. I would like a better understanding of what happens.

I think the people that have strong opinions about the order these operations should be in would say you are doing it in the right order and should just fix the mitering after the fact. I'm not sure, but the deformation might have been more problematic if you had done the operations in the reverse order. It's likely the dropout would have moved instead.

Probably a combination of the tube shortening from being heated and the bending slightly from the fillets around the brake mounts pulling the tube in that direction. My miters always open up sightly after brazing in bottle bosses and need to be tuned up before welding.

I'm a little worried about the Columbus flat mount stays. I'm thinking I should braze the mount first, then the dropouts, then the bb shell. I imagine there is gong to be some movement when I braze the brake mounts

Yeah I think for the next one I'll get the dropouts on, mounts on, internal hose hole done, and then cut the miters for the BB shell last. That seems like it would take all of the movement from the heat out of the stay until I put it on the shell.

I hope that those with more experience in the matter will chime on on this.

Yet another reason I like sliding rear dropouts, especially for thru-axles. But... For all my builds, I do dropouts before finishing the miter at the BB.

Thanks David. In the future I'll be finishing everything on the stays before I miter the BB. Hoping that will fix the issue.

So I was very curious about what happened regarding the NDS CS that I ended up building another set. This time I did everything in a different order and the same exact thing happened. First I attached the drop outs to the stays the cut the hole for the internal brake hose and then I drilled the holes for the flat mount caliper mounts. At this point everything was good. I then brazed in the caliper mounts and internal tube at the same time. Once everything cooled, I cut the miters at the BB.

When I pulled the stays out of the fixture, the NDS was shorter than the DS, and it had pulled inward at the BB as well. So I used a different order of operations, but got the same results. All of the miters were TIGHT. I'm still learning just how the heat affects the metal, but I still don't know how the NDS stay literally got shorter. That baffles me.

The bending you get with heat means something is getting shorter. I really would be interested in knowing how this happens, maybe I'll cut something up and polish it and look at the grain structure.

Since installing the bosses means you mainly heat the inside of the stay, have you heated the outside of the stay? Start by heating the outside of the stay and finish by heating the outside of the stay. Doesn't help much to get it red hot.

Yes I did actually do this as I wanted to bring the whole area up in temp. Maybe I didn't do enough afterwards? Not sure.

Something that I just thought of is if I push out the NDS stay at the BB to it's proper placement, will that in turn end up having the NDS dropout move inward? That would not be good.

It really doesn't matter how tight the miters are, the cooling filler material shrinks and pulls the tube with it. With the brake bosses both being on the same side of the tube, it will always pull like that. This is why some of higher volume builders are making machined brake mounts that replace the end of the stay. They get welded all the way around the tube, so it doesn't pull in any direction.

You can fix your stay by clamping the dropout in a vise and tweaking the stay in the direction it needs to go. It will take a few tries to get it just right, but the dropout will stay parallel to the other one.

Thanks DSAUL. That totally makes sense. Because I actually like the second set of stays more than the first I ended up putting the first ones in the vise and doing exactly what you suggest. I ended up breaking the joint on the stay at one of the caliper mounts. So either I put too much English on it or I didn't do a good job on it in the first place. Scares me a bit to try to bend the second set now if I'm honest.

Do you have any pics or links to the machined brake mounts that are welded all the way around the stay? I'm trying to look for something similar but can't find it.

I don't have any pics, but Moots and Sklar are doing their own 3d printed titanium dropouts with the brake mount integrated. Triton is doing a cnc machined part that is pre=mitered for a round dropout and welds onto the end of the stay. All of those are proprietary and not available for sale. Torch and file has a Long Shen cast part here https://www.torchandfile.com/LONG-SH...IR_p_1324.html

They also list a nice looking set from Reynolds, but they are out of stock.

The main problem is that flat mount was created with carbon frames in mind and adapting that to metal frames has been a challenge.

Another note on this subject: One of the framebuilding instructors from UBI commented on an Instagram post from another builder about this issue. He said that he welds the dropout to the stay with the bottom bracket end of the stay 20mm out from its intended location, so that the stay ends up in the correct location after welding in the brake mounts.

Thanks for this. Was that 20mm outward of the BB or 20mm in length of the stay?

I was able to VERY gently cold set and get everything to line up but it was nerve racking to say the least. I'd like to figure this out so I don't have to cold set anything if possible. I had thought that I could just tack the CS to the DO, cut and braze the caliper mounts in, then finish up the braze at the DO before mitering the BB. Maybe then I can control the movement a bit more.

It was 20mm outward from the intended spot on the BB.

Roger that. Thanks. I thought it was.

Thanks to this thread, I decided to get some DR1102 dropouts from Paragon because the brake mounts just bolt on. Unfortunately they don't have the ones with eyelets in stock. Web site says they are manufacturing again though.

They got me all excited because they told me the DR1102 were back in stock, but it turned out to be the titanium ones. 3-4 weeks for stainless.

Realistically, I probably wouldn't need them before that, so I'm going to be patient and not buy the ones without eyelets.

Trying to decide if I want to use their stay end plugs, but I think I'm just going to go with slots and possibly cover the end of the stay with a section of tubing at 90 degrees like Steve Potts does on fork ends.

I realized the DR1102 has an issue. The stays are about 25mm across where the NDS dropout would attach and the dropouts are only 16mm there. I suppose I could figure out a way to close the gap. Starting to think about using the dropouts I already have and accommodate the movement in the stay somehow.

Not sure where you are looking to get the 25mm but on the PDF of the dimensions I see 16.6 mm where the seat stays attach.

https://paragonmachineworks.com/file...ocs/DR1102.PDF

Yes, the problem is the chain stays are about 25mm across where the dropout would attach. Still thinking about it.

Crimp the end of the chainstay? I've got a road bike that I used straight 1" tubing for the chainstays. I crimped one end (actually crimped the tube then cut the tube at the center of the crimp) and attached that to the dropout.

Do you have any pictures?

Yeah, trying to think about how that would work. Also thinking about slicing the chainstay up. I'm going to try to bend up the standard s-bend stays. I did that before and dented one, but I think I have better ideas on how to do it now.