This is yet another area of bike quality I don’t understand, at least in regards to how it’s written about in bike reviews.
People are always saying that X bike has great quality welds. How do they know? By looking at them? Is it just me, or is that a completely invalid method of determining weld quality?
great looking welds do not always equate to great welds…it does however usually indicate that the person doing the welding knew what they were doing…which usually IS a good indication of weld quality. Given the thickness of todays bicycle tubes it is very difficult to make good quality welds that look good…it really requires a steady hand (or a very well programmed robot). In the hands of a skilled welder really good welds look like a work of art. The ripples you see should be evenly spaced between each other (overlap if you will) to the point where you almost can’t tell there are ripples there and the outer edges of those ripples should be the same in thickness. It’s hard ot try and equate that to someone who doesn’t or hasn’t welded but if you take a fat marker and try and draw a completely even thickness line around an object the same size and shape as tubes on a bike you’ll see how hard it can be. Now try doing it will molten metal joining two pieces of cardboard thin material. It truly is an art to get it just right…and in the right hands…you really can tell good welds equal good weld quality. In that respect it is the only method of determining weld quality (aside from taking the bike to an xray lab to inspect them)
Quite a technical topic
You can tell quite a bit about a weld by visually inspecting it, but not everything. For one the weld should be visible, one most steel and aluminum bikes it is not. A painted over weld is impossible to inspect. A smooth painted over surface could be a horrible weld covered with filler.
Styrrell
“Disclaimer” I am not a welder yet I play one on slowtwitch
A sloppy (un-uniform) weld to me signifies sloppy craftmanship. A sloppy weld may be just as strong as a uniform perfect beaded weld,yet who knows unless you cut it apart and dysect it, or it breaks and then you realize that it was not as strong. A sloppy weld is kind of like turning in a report that is handwritten on different size paper as compared to a report generated by a computer printed on legal bond paper with the use of a lazer printer. Both may quite possibly have the same content and substance yet one certainly looks better.
A sloppy weld tells me that the rest of the bike may be sloppy also. My welds look horrible yet they hold, yet a friend of my with much more experience can weld with perfect beads (uniform) his welds will hold and I believe he is getting better pentration and equal heat across the connected pieces of metal. If you heat up one side more than the other you can pull it over, a jig should help hold it but nevertheless the material wants to pull that way so if the weld looks the same around the entire tube theres a good chance the material was pulled equally on all sides so it more than likely stayed “centered”. A good weld in a sloppy jig makes a bike that sucks but that’s a whole different story, yet a sloppy weld may be a clue to the use of a sloppy jig!
I worked as a parts runner in a very large combo machine shop, sheet metal shop, weld shop, plating lab. I remember one project where they engineered a jig to weld some complex shaped tubes together. I think they were titanium tubes for propellant aboard a satellite. Anyway, on the first pass, they put all the tubes in, welded it and realized there was no way to get the welded contraption out of the jig. Brilliant!
Unless the reveiwer X-rayed the welds, there is no true way of saying “bike X has great-quality welds”.
I have had things fail that had the most uniform, nickel-on-top-of-nickels stacked on each other welds fail, and I have seen stuff that looks like a babboon welded it last for a great bit longer than it supposedly should have.
Unfortunately, you just have to hope that the QC dep’t knows how to inspect their welds and they know how to give their welders a good test when hiring and looking over a sample batch.
Here is the shocker, what most consumers find a quality looking weld is actually a horrible weld from a technical perspective. Virtually all (and I mean over 99%) bike frame welds do not have proper weld penetration, it really is metal gluing, not welding. For a proper weld the base material should melt, in bike frames that is almost never the case. In fact, consumers complain if it is a proper weld, they don’t like to see weld blowthrough from for example the toptube weld into the seattube. Yet that is one of the tell tales of a proper weld (of course you then need to remove the blowthrough for proper seatpost insertion, but nonetheless).
Now, you can make a bike frame strong enough even with “metal gluing”, but proper welds they are not.
There are also many problems associated with ‘blow through’ like crystalization and contamination that can weaken the weld as much if not more than ‘shallow’ welding as you discribe.
I think to say 99% is a bit far fetched, the loads on some areas of bike frames are not as structural as others so cosmetically you could very well do a picture perfect weld and it would be fine. What is often worse for a frame and the alloy surrounding the weld is to over dress or torch the weld surface to smoothen it.
If you want to know about good welding speak to people in the aircraft or petrochemical industry where they actually certify their welders and work.
Are any welders in the bike AOI, or other industry standard, certifed?
You mentioned something I almost entirely had forgotten (metal “blowthrogh”), as it has been awhile since I have done any welding. If you did not have that in your welds where I briefly worked, you did not pass the test. I was a welding apprentice at one point. I had to “clean up” the areas that were welded on broken equipment in a mine. I quickly was tired of that work, then went onto being a hairdresser.
The 99% is definitely not an exageration, as in 99% does not have proper penetration. You are correct in that in most cases this doesn’t really matter and the weld is still more than strong enough even though it is not a proper weld, but the fact remains that weld penetration is very bad in most frames.
Here is the shocker, what most consumers find a quality looking weld is actually a horrible weld from a technical perspective. Virtually all (and I mean over 99%) bike frame welds do not have proper weld penetration, it really is metal gluing, not welding. For a proper weld the base material should melt, in bike frames that is almost never the case. In fact, consumers complain if it is a proper weld, they don’t like to see weld blowthrough from for example the toptube weld into the seattube. Yet that is one of the tell tales of a proper weld (of course you then need to remove the blowthrough for proper seatpost insertion, but nonetheless).
Now, you can make a bike frame strong enough even with “metal gluing”, but proper welds they are not.
What does that look like?
Well, it’s hard to explain but consumers really seem to like it when the weld is “on top of” the joint. But that is just a matter of simply using a lot of welding rod, it doesn’t say anything about whether or not the base material melted or not. Don’t get me wrong, the welding rod is part of the procedure and necessary in most cases, but unfortunately it can also be used to mask. As to what a good weld really looks like, it is unfortunately tough to see without cutting the frame or X-raying it.
I’ll try to put something on our website with cross sections of various welds to show what I mean.
As for the blow-through, you can sometimes see it on the inside of the seattube where the toptube is connected. Sometimes cleaning the seattube out removes all evidence though, so that is not a very good tell tale either, and as somebody mentioned, blowthrough is not per definition proof of a good weld.
“The penetration of the weld on a critical component must be assured. Too little heat means partial penetration which results in an incomplete bond. Too much heat may weaken base metal, increase thermal distortion, and even burn through the workpiece.”
Ref: Http://www.crms.uky.edu/c4m/pubs_archive/advman/advman11_3/advanced.html
Note the reference to “burn through”, it is not exactly correct to say burnthough is good for a weld but more so that the weld pool must penetrate through the material without breaking the rear of the alloy. As I said previously to much heat can do a lot of damage and cause premature failures in the tubing localized near the weld zone, very often not the wld itself.
OH, I’ll have to check your website out if you get that up. I recently had my Klein warrantied (not sure that’s a functional verb) because a weld where the downtube met the bottom bracket was failing, evidenced by a hairline fracture. Do welds fail because of the riders power output, like X cycles at Y watts, or does failure generally occur from hitting a small pothole doing 70+km/h, or something else I’m not able to imagine right now?
Both are possibilities, which is why a test like the EFBe is nice but not all-encompassing. You also need road-impact tests, frontal impact tests, etc. Downtube-bb area sounds like rider input rather than road input though. Probably a small nick in the weld somewhere.