[jakeru]

Hey folks -

I have some aluminum round tubing, .065" thick that is mandrel bent, and of uncertain alloy. It is significantly more malleable than 5052 however, so maybe its something like 3003.

I have tried welding it to 5052 sheet and also to 6061 tubing, using 4043 filler rod, and in some cases I experience some small "hairline" cracks right through the weld bead, always right where the two materials join. The crack shows up almost immediately, upon cooling of the weld bead. The tiny crack shows up so quickly, you can't even hear the metal go "tink" like sometimes tack welds go when they pop apart.

I can re-melt the cracked bead, and while hot it seems like it flows together nicely, but the crack just comes back immediately. And it seems that trying to penetrate deeply does not solve the problem - the crack just comes right back, almost immediately, just the same.

Pre-heating seems to help some (OK actually, maybe quite a bit). As does it seems, going slower (maybe because that pre-heats the part) and not trying to penetrate as much. (OK maybe that is more because when going slower, I am worried about melt-through.) I noticed that the problem seems to be reduced after getting the parts hotter, so I think pre-heating seems to help. On the last tubing sections I joined that had this cracking behavior, I also went over it with a second pass, not deeply penetrating (IE: more arc length to spread the heat out) and added a more pronounced, raised bead for some extra strength. But I am not totally satisfied by this because the pronounced bead just looks a little ugly.

I should also mention that I am fitting the parts up tight. I am curious if I fitted the parts up with a gap, to be filled by filler rod, if that could eliminate the cracking. (IE: prevent the dissimilar parent metals from flowing together)

I have not experienced this cracking behavior before welding 6061 to 6061, to 5052, with 4043 or 5356 filler, etc. I've got more of these mandrel bends that I'd like to use on upcoming some upcoming welds/projects, so I'd like to learn to become proficient at welding them.

Is this a common problem for certain alloys, and what are the common solutions used?

[Gary]

I have had the same problem when just trying to fusion weld pieces together , I think it is a problem of dirt or inclusions in the weld puddle .not sure though , but when I run in to that problem and just run a bead over it it does not crack , just if I do a little fusion tacking is when that happens. But if you are running a bead over the crack and it is too high then you are not getting enough weld penitration. I know you said thin material but you have to find the right amps to get full penitration and get material in there for it to be a solid weld.
Just my 2 cent , I can be wrong but that is what I have found.
Gary

[jakeru]

Gary - thanks for the ideas, however "hot and fast" was the first thing I tried and it actually seemed to have the cracking problem really bad. I'd be surprised if this were a contamination issue because the molten puddle when it is a liquid seems to be flowing together just fine. (And there are no other signs of it being a contamination issue.) But I will double check I am using a clean ss brush and acetoning well on my next attempt.

Also, I have experienced the crack when fusion welding 6061 and 5052 together, and it just seems different than what I am getting. I recall the crack on the fusion 6061 and 5052 necks down and goes "tink" and kind of indicates that there are some stresses being generated from cooling that pulls the bead apart until it cracks. The cracking behavior I am seeing more recently however seems like it happens before the material has even cooled enough to generate any stress... there is no "necking down" of the material around the cracked area, and the crack is a tiny hairline crack through what you might think otherwise looks like a sound weld bead... it is pretty weird. Almost like it is forming when the aluminum is "slushy" (between liquid and solid.)

I did some research on cracking, and found lots of technical material on it actually. But I'll try and simplify it to what I think might be applicable for my case. Basically, certain alloying ingredients of aluminum at certain mixtures ("just the right" amounts... not too little... not too much...) are highly sensitive to hot cracking. By joining dissimilar metals, we are "mixing our own" alloys in the weld bead, which would be a combination of the two parent metals. We can control the alloy to some extent by filler rod selection, and also controlling how much % of filler rod we mix in ("diluting") to the parent metals composition.

Here are the popular aluminum alloying "ingredients" that can cause problems with crack sensitivity:


Perhaps the one that is giving me a problem is silicon? I can't really rule anything out since one tube I am welding I am really just guessing about. (I just know that it is pretty ductile...)

The way you control the filler rod dilution is supposed to be by joint prep... IE: by "veeing" the joint or leaving a gap instead of a tight fit-up. (Or perhaps, by the tall raised bead I tried, but I am not satisfied with that.) I haven't tried the gapped aluminum fit-up before, because I always hear as a general recommendation to "strive for as tight of a fit up in your parts as possible for best results", but maybe if its a uniform gap it will work OK? I think I am going to give it a go and report back here.

I am also kind of curious to see what welding on some 2024 aluminum I've got is like. Just to see how the hot cracking compares on a material that's supposed to be "unweldable".

[jakeru]

I tried another weld with everything the same (same parent metals, same surface preparation, same filler rod) except that I fit the parts up with a small gap (stood the pieces apart using a .035" mig wire), instead of fitted up tightly like I did at first. I also did no pre-heating, to keep as much the same process as before, to test only the effect of the gapped fit-up. Result: the gap made it weld nicely, without any tendency to crack at all!

By fitting the parts up with the gap, I was basically forced to add a good bit of filler rod to get the two pieces to flow together. Before with the tight fit up, the pieces flowed together with very little filler, so only a small amount of filler was added here and there. I guess I can't rule out the possibility that the filler rod cooling the bead had an effect, but my feeling is that the dilution of the alloy with additional filler rod is primarily what got rid of the hot cracking behavior.

So next time you hear someone say "to get good results TIG'ing aluminum, fit the parts up as tightly as possible without any gap", know that there are some exceptions! I found that a uniform gapped fit still allowed me to obtain a decent stack of dimes weld bead appearance.

[Trackmaster welder]

You figured thru it on your own my friend, .... good job! The problem you were having was exactly that, ..... not enough filler rod. On the tight fit up it is exactly as ya found. Very tempting, or compelling is maybe a better word, to be skimpy on the build up, and that is exactly what happens. The thin strange mixture will immediately not withstand the thermal stress and crack before you can flip your hood up! You also saw that it got better with better attention to preheating, but even so you have to get some material in there and a VERY slow reduction and torch swirl, adding another cap layer on that last bead if necessary. You've got good instincts. Looking at your weld, you just need more confidence in your abilities you have! Wet in, move along at a good but controlable clip, but finsh slow, take your time on that last crater fill. Nice bro.

[Trackmaster welder]

Oh yeah, one last thing if i may. Especially if your going to grind the finished weld, forget about the "stack of dimes" thing! Get your dips in there more frequently as the larger steps thin the material build up and by nature stretch it thin. Get in there with 2 dips or adds per second approx. and you can do tight fit ups on .062 wall with good penetration, and a nice rounded bead profile!

[jakeru]

Thanks Lou - I appreciate the thoughtful words. I like having things to think about and improve for my next welds. I think you are right, I am going slower on this than I could with more practice. I am finding it tricky to get the torch to "rotate" around the pipe! And I am probably a little bit intimidated by burn-through. (Or rather, "excessive penetration", leaving a big blob on the inside of the tube.) More experience on TIG'ing round tubing will probably help. But the dissimilar alloys I think were also making this really tricky.

I tried the stacked dimes just for practice. With my methods and tools available (hacksaw, etc) it takes me a while to prep each weld, so I figured why not try it. And I think you are onto something (thanks!) that faster travel speed and more frequnt filler rod adds would make a better looking bead appearance. I've laid down some really good looking bead sections before (memorable occasions ), on flat sections mostly where I had pretty fast travel speed and pretty small, frequent filler rod dips. I don't know why it can't be done more slowly, but faster travel with smaller dips seems to give a more evenly "glossy" (less grainy) bead appearance. If I had a better set up fixture to hold the tubing vertical (so I could weld horizontally rotating around it), I think it would be easier to rotate around the pipe and weld longer more consistent beads. These welds were mostly with the tubing laying close to flat on a table.

Anyhow, what I didn't picture before the ground/polished bead, was I took a second pass around it to smooth the "stacked dimes" down evenly before sanding/polishing. I have pictured that below. It left me without much material that needed to be removed. And there was no problem with any cracks appearing from any reheating (I was curious to see if there would be.)

On another butt weld tube joint I made, also of the same dissimilar alloys with the controlled .035" gapped setup, I was joining a straight tube to another that I cut in the middle of a mandrel bend. The bend left the wall thickness much thinner at the outside edge than it started. The inconsistent wall thickness made it considerably harder to get a good bead appearance, and the gapped fit up by itself adds quite a bit to the difficulty of needing to properly time and position the filler rod. The thin side really wanted to melt readily. Anyhow, the resulting bead appearance wasn't the greatest looking on that (kind in the section with one side being thin), so for an experiment, to try "cleaning it up", I went over it with another pass of the torch and no (or very little) filler, using a sort of weave pattern to even out the bead width and make it smoother. It definitely made an improvement in the bead appearance. The weave is kind of pretty up close, I wouldn't say as nice as a perfect stack of dimes to look at or anything, but its interesting. I am going to be OK putting this in a highly visible location in the middle of a vehicle engine compartment.

I also tried some .035" (4043) mig filler rod to fuse a small 5/16" OD pressure tapping tube to the main 2" tubing. Boy, there is a lot of potential for precision work with with that .035" filler rod! My main challenge on that joint was again, keeping the proper torch angle while progressing around the tubing joint. The joint shape didn't offer much of a consistent surface to rest the cup on. The .035" filler rod is very sensitive to melting from torch heat aimed the wrong way.

The end result is completed now; shown below.

[jakeru]

Here is a close up on that last "weave bead" I tried. (I posted this on another thread about how to get rid of the white etching.)

[Trackmaster welder]

Hey there my fushion brotha! I was just posting on the everlast site to someone having trouble with larger od tube to flat filet weld. It takes a lot of torch manipulation ( dexterity ) probably the most of any weld joint. You're juggling torch movement, torch angle, and hold off all at once. And sometimes having to direct the heat to one or the other for thickness mismatches! I was telling him to rehearse his movements, hood up dry runs, over and over till you find a way to not strain but also not have to stop every half inch or so. Whether you have to stand on your head, hold the torch upside down etc! Ive seen 6 different guy's do the same weld 6 difference ways, so it's like anything else. You'll find what works for you. I probably have 2 or 3 thousand hours of production welding of thin wall 4130 frame tubing, and those are challenging joints for sure. But you will find the movement thru sheer repitition! And then like i said, you will trust the "force" and just book! - Lou

[jakeru]

Hey trackmaster thanks for chiming in again and the moral support.

Been meaning to mention to you that in my childhood, my brother and I had two BMX bicicycles, one of which was a Redline! This probably would have been purchased in the '80s. Wonder if you might have welded it?

I remember it was very solidly built and looked good too. I remember it was a shiney, silvery color. (Pretty shiney now that I think about it! Perhaps chrome?)

I remember my brother set that bike up with "stunt pegs" on the rear axles. I don't think either of us really figured out how to use 'em though, except just to stand on them while the bike was rolling. As we got older, we moved on to larger mountain bikes and I can't remember what happened to the old BMX bikes.

I had a little Raleigh BMX (sort of a miniature BMX bike) that was sweet too, I was the younger one so it matched my smaller size. It was really sweet looking... a large factor being its deep, dark red paintjob with chrome accents.

[Trackmaster welder]

Hey bud! , Yep entirely likely if it was in the earlier part of the 80's. Kind lost track (no pun intended!) of the company after that, Later 80's i think they even started to have them welded overseas! Bummer. Was a really cool company with a really interesting background.

Oh yeah, nickle plated at least in the early day's, .... and for some reason my memory is failing on the chrome, but i think so also. And some powder coated too. We sent them out for all that, so mostly we just looked at the beautiful raw Chromoly all day long. That job is where I really cut my chops welding! Just so much hood time you really had a fast learning curve