guy153

Having made a new frame and fork for my Trek 520 Tourer (actually few components are original to the Trek anyway) I thought I'd have a closer look at the welds on the old one. None of this is "dangerous" or anything in my opinion and I think the bike is fine (I rode it for only a few years, as it was a warranty replacement for an earlier 520 I owned, made in USA, on which the frame actually broke) but it's interesting to see now that I have more of an "eye" for TIG welds.

First the frame and fork next to the new ones:



And now some of the interesting features on the Chinese one. Exhibit A, this looks like the TT burned back and our welder had to fill in a bit of a hole:



Exhibit B. This is underneath the DT (where the original USA-made frame failed actually). You can see that the weld has actually been filed down. I wonder what horror show this was "fixing":



Exhibit C, a little bit of undercut on the side of the DT:


Exhibit D. Possibly the ST burned back a bit here. Or it may be just that a sudden loud noise caused our welder to let the TIG torch stray a bit:


Exhibit E, the inside the SS where they join to the dropouts. Just a bit ugly is all:


Exhibit F, undercut on the fork:


And finally, the underside of the fork. What's happened here is the bottom of the sleeve has melted into the weld. It would be better to weld first, with a bit more overhang, and then run a hole-saw through for that curved effect that people like (I actually don't mitre the bottom of the steerer at all and put a cap on it, for better or worse):

Andrew R Stewart 

I agree that with our better trained eyes we can see joining (or forming too) "issues" that most don't even look for. I suspect a Trek rep might say that the frame hadn't broken is proof of how well the frame is made, regardless of the looks. Andy

guy153

Yes I would call this good enough. I don't think any of these issues would cause an actual failure any time soon. Someone paying top dollar for a custom frame would expect better but as you say most people would never even notice.

pwyg

It is production welding, i have seen the same thing on custom built frames and many production frames including Bridgestones. Quality control is a time based and therefore a factor in the final cost if the product. Therefore you see a lot of this. Paint choice can also hide a lot too, this is why powder coating is great as it is heavier on average when compared to wet paint. But then I have stripped a cycle art painted frame that had an unbelievably thick primer/filler layer under the paint. There is a lot to the term what you see and do not see. Since frames are painted your assumptions of quality is based on that aspect of the frame not the actual important bits that natter more. Since I an of an earlier generation of cyclists and the same for my indoctrination in to building regardless of the appearance of a Tig joint I view it negatively, since the nature of Tig is a shortcut joining process.

guy153

Update to this. I just measured the alignment. The string goes from the dropouts around the HT in the usual way. The gap from ST to string is 4.5mm larger on one side than the other! These calipers are set to measure the gap on the other side so they should just touch the string on this side. I never noticed this when riding it of course but it's still pretty awful.

Andrew R Stewart 

Many bike brands are very nebulous about alignment tolerances. For good reason as the common production bike frame hasn't had much time allowed for regarding its alignment. As bikes have become more a commodity then a trades product this should be of little surprise. Andy

unterhausen

My lbs said that trek told them they wouldn't warranty something that was within 5mm. If a frame fails the string test by 4.5mm, that means it's only off less than 2.5mm from centerline. So that's within tolerance. Now imagine that if it was off 9.5mm and being told it's within tolerance.

guy153

The part I don't understand is that in my experience, with 6 TIG frames in a row now, is that if my jigging is right to start with then things hardly move during welding. Certainly nothing like that far. And in the case of this Trek, the HT and ST are still in parallel planes (which is what matters much more). If the cause were welding distortion then it would be all over the place. So the most likely cause is just poor jigging. But you would think that in a production environment the easiest part would be to give everyone a decent jig where you just throw the tubes in and they're automatically in the right places.

Andrew R Stewart 

I think there are some assumptions said that are not as true as often as you think. From the jig tooling condition, the amount a frame can move with welding and the motivations/skills (beyond laying a bead) of the workers. Having seen a number of vids taken in Asia factories (and presumably showing only what the company wanted to show, their best practices I would think...) there are a number of scenes where I take issues with methods. Andy