bulgie 

I'm open to suggestions, but most ways of holding a braze-on are at least a little less versatile, as in ability to work anywhere. Different tube diameters, cramped spaces... oh and did I mention speed? My way after a little practice is about 20 seconds total. I don't add a fillet like in the Chapman video though.

I'm OK with calling it the Chapman method. I was doing it in the '70s, but I probably didn't invent it. Anyway it's kinda obvious, so I'm sure it's been re-invented a thousand times.

My way is slightly different than Chapman's though, and I think maybe a little better. I put the dab of filler on the BO, not on the tube. I think this can make for a quicker final braze, because the flame can easily heat the BO, and the thermal conductivity between the BO and the filler blob is, uh, optimal? maximal? I'll think about how to word that better later, but my point is, getting the little blob to melt might be quicker if the blob is on the BO versus on the tube. Plus the thermal mass of the BO + blob stays hotter while you bring the contraption over to the tube, so it's closer to brazing temperature. The tube + blob has higher self-quench, the tendency to cool off from heat conducting to surrounding steel. I get the whole BO somewhat hot, so almost no self-quench -- the blob is ready to melt again as soon as the thin tube comes up to liquidus. Boom, done. Zero spatter or blobs, zero cleanup.

One other minor advantage: while I'm putting the blob on the BO, it is upside-down, so the melted flux is running up the shaft ("up" relative to the orientation of the BO later when I stick it to the tube). So that reservoir of molten flux is headed back down towards the joint while I'm attaching it to the tube.

These are minor differences between me and Chapman. Hey, I said "maybe a little better", right? Not making grandiose claims here. If anyone knows any reason why the Chapman way (starting with the blob on the tube) is better, let me know. I think our ways are substantially the same though.

Mark B in Seattle

Tandem Tom

So a quick question. I am going to install a new peg with brass. Should I mitre it or leave the bottom flat to allow brass to flow underneath?

unterhausen

I thought of some advantages right as I hit post, but since I'm not coordinated enough to do it that way they don't apply to me. It's a bit of a pain to hold a pump peg in place, they are awkward. But if you put the peg on and don't like the way it is sitting, or have to add filler, then some way of holding it is best.

Tom, you want to have it mitered. You should be able to flow filler under it.

Tandem Tom

Thanks!
I was practicing with brass this morning but I think I will use Fillet Pro for the second go round.

duanedr 

Congratulations Doug. A very good reason to put the torch down!

Kuromori

Results don't seem that surprising to me. If there's a small gap, the strength of the braze joint can exceed that of the base material, something that can't be said for a big fillet laid on top. Especially if you use silver on un-heat-treated tubes because it undoes the cold work, weakening the tubes.

calstar 

I don't understand this, will you explain it to me? Do you mean the cold working of the tube forming process?

thanks, Brian

Kuromori

Yes, if you look at spec sheets, most cycle tube manufacturers list a higher UTS than typical of 4130. Maybe some of it is due to faster rates of cooling when normalizing something so thin, but another part is the amount of cold work and associated work hardening which gives the grains an elongated form. When you start heating the steel, the ferrite starts recrystallizing and replacing the work hardened elongated grains, returning the steel a state similar to before cold working, which is weaker. This happens very quickly as you approach silver brazing temperatures.

It also happens when you heat the steel for brass brazing temperatures, but the higher temperature means steel actually raised to brazing temperature is austenitized, so the microstructure can be manipulated with rate of cooling. I think it's reasonable to expect some gain in strength in a brass joint which is where the stress riser is, although there's the other issues of brass like increased distortion etc. There's still a ring of weaker steel around the somewhat hardened area, but ideally it's not where there are any stress risers so there's less stress for the weaker steel to deal with.

It's different for heat-treated tubes as I'm pretty sure they're heat treated post-cold-drawing, and end up with a martensitic microstructure.

If you want a more in depth explanation, I have one written up here https://kuromori.home.blog/making-se...rame-building/ under "Annealing and Recrystallization" but it's still a work in progress and there's several things I need to fix or add.

Andrew R Stewart 

In the 1980s one of the bike mags (Bicycling or Bike Tech?, can't remember) did a comparison of joining tubes with a silvered lug, brass fillet and steel weld. This was in the day of very little "heat treated" tubing being available or in use. I believe the tubes used for the test were the more common 4130 (from Tange?). The joints were tested for ultimate failure from bending forces and surface hardness along the tube. The silvered lugged joints failed/crimped just at/past the lug tips. The fillet joints at the tube butt and the TIGed ones very close to the weld. It was stated that this suggested where the HAZ transitioned to a non affected state of the tube. (and the lugs acted as an additional and really thick butt). (Now with K's very nice description I also see that the silvered joint's temp level was also a factor). The hardness tests showed the lugged tube lost hardness approaching and including the joint. The fillet showed the hardness loss away from the joint (matching the HAZ transition) and, I think, a lessened hardness along the thick butt to the joint. The welded joint had a hardness loss very close to the weld bead then a rapid increase at the bead.

The takeaway that I remember was that each method of joining had it's failure mode. The beer canning occurred at different places but all types of joining would still fail. This test didn't address fatigue life, joining issues (distortion/loss of wall thickness) or the tube material having any intended thermal responses specific to any joining method.

K's descriptions do a better job of explaining what's going on then I can and pretty much is what I've read and been told by smarter then I others. Andy

bulgie 

Wow that's a really good article, thank you! Some of it went over my head, but the parts that I knew from my two courses in Strength of Materials in engineering school were all correct as far as I could tell, nothing to critique. The parts that went further into the weeds than my college courses, I will have to read again to try to absorb.

I was already a bike frame builder before I went to engineering school, so I was highly motivated to learn Strength of Materials, I ate that sh¡t up! The lab course was awesome -- we got to cold-work and heat-treat, test hardness, then section, polish and etch, and look at the resulting grain structure under a microscope. My favorite part was breaking stuff though. Tensile tests are interesting but the Charpy test for toughness is the best. It really makes concepts like "the area under the stress-strain curve" come to life, for this visual learner.

It was immediately useful in my frame building, helping explain things like why silver-brazed frames bent right at the edge of the lug when crashed, but brass-brazed frames bent some distance away from the lug. Reading Mario Emiliani's articles on his testing of bike frame steels (flawed but still worthwhile) also added to my understanding. Check them out if interested, available in archived Bike Tech magazines here and there on the web. Let em know if you can't find them, I know a place...

Anyway, thanks for reminding me of that, a fun trip down memory lane. But I am boring everyone now so I'll shut up.

Mark B