Trakhak

Tourist in MSN posted in another thread about a Surly frame that was overly flexible when pedaling with heavy loads. The excessive flexibility was judged by a frame builder to have been a consequence of improper welding technique (specifically, improper heat settings), but Surly refused to replace the frame under warranty. Is that (i.e., bad welding technique resulting in an overly flexible frame) a known phenomenon? I've never heard of it.

Andrew R Stewart 

I believe the question is whether the steel's modulus of elasticity lowers due to the heat of poorly done welding. I just did the unusual, for me, and tried to find the answer via surfing the internet. here's what I found. I've omitted the names and such for their privacy.

The mechanical properties of steel will certainly change with heat treatment and microstructure but this statement pertains to the properties that are governed by plasticity, e.g., strength, hardness, ductility, toughness, etc. In stark contrast, the elastic properties, such as the Young's modulus and Poisson's ratio, hardly change at all, or if they do, the change will be very small. The modulus is essentially insensitive to microstructure, heat treatment, strain rate, prior history and even small changes in composition, as it is a function of the elastic "flexing" of the atomic bonds. Thus, compare a low strength mild steel, with a yield strength of 300 MPa, with an alloyed maraging steel that could have a corresponding strength of 2000 MPa, and contrast this almost order magnitude increase in strength with the fact that the Young's modulus of the two steels is essentially unchanged.
Bottom line: The elastic properties such as the elastic modulus are essentially insensitive to heat treatment and microstructure.

This jives with what I seem to remember being told a long time ago by one of my mentors. Any heating then cooling of steel is heat treating to some degree. We tend to only use the term when talking about controlled temps, time and cool down processes which result in a pre decided spec. But bad welding is just not controlled heat treating too. I have more thoughts but I should go back and read the linked thread first. Andy

PS- Well after 122 posts and on page 5 of Trakhak's linked thread I found the relevant post about the story. I gleamed no new info for it though.

guy153

Thank goodness for the internet I remember the dark days before we had it and people went around believing weird myth and lore like this because some old guy in a bike shop told them. Another common misconception is that over time springs become less springy. Do I need to replace my derailleur because the spring has gone soft? Of course not, unless it's broken.

unterhausen

There is so much B.S. in the bike community that I really have to work to ignore it most of the time. This definitely includes opinions from framebuilders. There are no qualifications for becoming a framebuilder, and even the few that are engineers probably don't understand materials science to any degree beyond the one class they took as a junior, which they slept through. I'm probably the most qualified framebuilder on that front, and there are things about material science that I don't fully understand.

Leaf springs on trucks fail, but that's probably just buckling from overload. Or maybe rust. Or plastic deformation over time because they have lost their temper and the strength has decreased. I am pretty sure if the same thing happens on a bike, something is going to permanently bend and break.

dsaul

That's just not a thing. That alleged framebuilder was either making things up or repeating some fallacy that he heard. Cooking a weld joint, by heating it so much that it cant cool to a non-reactive temperature before the shielding gas stops flowing, can cause premature failure of the tubes. It doesn't make them more flexible.

I would think the 460mm chainstays were mostly responsible for the flexible feeling at the bottom bracket.

Nessism

This reminds me of the claim that high strength tubes are stiffer than non heat treated tubes. In reality, the modulus of elasticity doesn't care about strength. A cheap, low end, thick tube, will be stiffer than a high end tube that's thinner.

bulgie 

Even Reynolds used that old canard in their marketing materials for 753. Any engineers in the company probably winced, and/or hung their heads in shame.

unterhausen

As someone who broke 2 teledyne titans at the bb, it makes it real flexible when the tubes crack and the cracks are more than half way round the seat tube. But it isn't noticeable for far longer cracks than most people would imagine. Which is funny in the context of this thread, as if a little overheating is going to make a noticeable difference in stiffness.

This reminds me of my consulting deal. $20 and I'll tell you what you want to hear. Want someone to tell you your bike is getting old and less stiff so you need to buy a new one? $20 and I'll write you a letter signed by a Ph.D. mechanical engineer. I should raise my rates.

Andrew R Stewart 

Back in the 1970s when I was starting out in the LBS world we used the word "resiliency" to describe the advantage of high strength steel with thin walls compared to the more common gas pipe bikes. I would explain that many department store bikes actually had stiffer frames, although less strong and seriously hampered by weight and geometry. But that these more rigid bikes weren't what one wanted to ride for hours at a time.

Then Cannondale came onto the market and that with the 1984 Olympics road race, won by Alexi G and in Colorado yet, shifted the consumer's view of what was the cool ride. The whole touring portion of the market went away and all kinds of boy racers wannabees appeared and thus we have the crazy market we have now. Andy (repeating a rant that he has grumbled about ever since 1985)

unterhausen

I think you'll have to drop your rant, road racing outside of parking lot crits is dying in the U.S.

I see shops saying they can't sell low-end road bikes, just $5000 road bikes. That certainly seems to be true at the LBS I shop at, although they have always been mtb-oriented. I'm not sure what they have for road bikes right now. It's true that they sell a lot of gravel bikes now. Before they split with Trek, the only road bikes they sold were domanes for the most part. I'm not counting hybrids as road bikes, although that's where they are ridden.
On a side note, Trek kept harassing them to be a Trek only dealer so they quit selling Trek altogether. I bet Trek hasen't sold a Domane to someone that lives anywhere in a wide swath of Central PA since then. They used to sell them regularly. Win for spesh.

repechage

‘that above is a bit confusing to me, the ‘84 Olympic road race was won on what appeared as a multicolored Pinarello, with Dura-Ace AX cranks and pedals. I attended that race, near the top of one of the rises.
‘it was too bad that the Soviets did not attend and a few other Eastern Bloc countries, payback for 1980.
the course was based on some of the 1974 Southern California District road championships loop.

repechage

Regarding welding or brazing temperature and the feel of a frame’s stiffness, no hard facts but there was one Carlsbad Masi that was on the rack at the shop I worked for that just did not sell. I owned one and swapped wheels into it and test rode it rode it for an evaluation. It felt “dead” as if any power was just swallowed by the frame and never to be retrieved. Only one mind you, but was weird. Was eventually sold at a discount to a customer who more than likely just hung it up to display and boast. Who knows why.

many years later in a materials and processes class, there was presentation of time at temperature while joining, max temperature and rate of cooling, ambient air, the general temp of the space and various quenching approaches, oil, water and rate.
I thought back about that dead feeling bike.

Andrew R Stewart 

Sorry if my phrasing suggested I thought that Cannondale was a presence, or won, the 84 Olympic road race. More my point was that with the cycling fever created by a US Gold medal was the foundation of the rise of the road racing bike being The Way if one was an athletic person (or not, more often) Cannondale just happened to have their "new" frame coming to the market at the same time and the two complemented each other. Andy

unterhausen

I rode a trek 500 when they wanted to evaluate the arabesque Shimano components. I was the only person there that rode at the time, otherwise they wouldn't have asked for my opinion. It felt dead to me. In retrospect, I'm pretty sure it just didn't fit. In those days, none of their available sizes fit me and I should have sized down. I can make any of my bikes feel dead by raising the stem an inch. Something similar happened at the LBS. Somebody got a new custom bike, and they didn't like how it rode. Then the shop realized they hadn't transferred the person's measurements. Once they did that, the rider was happy.

Kontact 

See, that's funny. You are propagating the myth that springs can't wear out when they clearly do. Systems with springs that are kept inside their elastic limit are fine, but small springs pushed too far do take a set and lose force. As almost anyone with a well used Shimano derailleur can tell you about their B springs.

repechage

OK, just read weird.
the ‘84 games did get me out on the bike again.
riding along San Vicente Blvd in Santa Monica a few Olympians wanted to talk while riding.
they were hoping to sell their bikes before returning home. One answered his event was the points race. A Belgian, he declared he was going to win. And darn, I got to watch him do just that.

unterhausen

That's an interesting point. Bicycle parts are often built with questionable materials. I have wondered what would happen if someone decided to build bike parts with the best materials available. Maybe they do, sometimes, but I suspect that shimano cuts corners on those springs. I have also found a Mafac brake spring that was very weak, not sure how. The materials on those really seem questionable. I bent it back and it works just like new. I imagine in another 50 years someone is going to have to do it again.

Springs can be tricky to design because getting the desired travel may exceed plastic limits. I decided to make a tribute to the EVT dishing gauge. I couldn't find a stock spring that works for the plunger because it has so much travel. I am pretty sure I tried to see if someone would make one for me and it exceeded their limits, but I'm not positive about that. So I wound my own. I suspect it's going to wear out some day if it gets used too much. I would be interested in knowing if EVT buys springs. Of course, I imagine they take design a little more seriously than I did.

In my experience, however, worn out springs are just as springy in their new configuration. They just have less travel. Since bike frames that are "worn out" still have their original dimensions, it seems unlikely that they are really less springy.

As an aside, I asked Jan Heine if he would sell me some springs. He said there is no demand, send him $200 for some rebuild kits. So I'm going to bend my own when I run out of old ones.

guy153

The stiffness of the metal is the same. But yes if it's yielded (as opposed to actually broken) the geometry is different so the spring will behave a bit differently. But a B spring won't yield in normal use. You would need to overextend it somehow.

Kontact 

If B springs won't yield, why do worn Shimano derailleurs run out of B springs adjustment to clear the cassette? No, the chain is not too short. I've seen this hundreds of times as a shop mechanic

Just as plenty of people have seen worn out springs in firearms. The US military has a standard procedure to detect worn recoil springs in pistols by measuring their length.

bulgie 

I don't think anyone has said springs can't yield, they clearly can. If they yield in normal use, that's a design fail.

But I think it's important to remember that yielding only changes the spring preload, not its spring constant. The stiffness of the spring doesn't change at all, just its initial condition (preload).

And yielding almost never happens bit by bit, slowly over time. It's almost always a one-time overload.

It's been decades since I was a shop mechanic so my exposure to Shimano mechs with a weak B-spring is nil, but if you say it's common then I guess I just need to get out more. But I have at least one Shimano from the early '70s and its spring seems about the same as new, to my uncalibrated fingers.

Kontact 

See, that all sounds like what the textbooks imply, but if you have a spring that is pushed just past yield many times, it seems to degrade in stages and not really be correctable.

The most obvious example from my experience are magazine springs in pistols. Many of them seem to last forever, but the ones designed to maximize capacity (like a Glock that fits 17 rounds in the space traditionally designed for 15), the mag springs don't fail the first time loaded to full capacity or even the 100th time. But after enough cycles that start with spring pushed to its limit, the magazine quits feeding correctly. And when you try to correct by stretching the spring, it only works for a handful of cycles. This sort of thing is common knowledge in the military with contract Beretta magazines.

Shimano B springs behave exactly like this.

The other thing to keep in mind is that there are some steel alloys that make better springs than others. If springs had the kind of binary behavior people keep describing, there would be no reason for using chrome silicon in preference to 17-7. The spring thing has been oversimplified to the point of being inapplicable to real life.

unterhausen

I'm not sure why we're arguing springs though, it doesn't seem relevant to bike frames. Spring failure occurs after obvious plastic deformation across the full section of the spring. Bike frames don't see that except under unusual and very obvious circumstances. A spring in the late stages of failure is going to have physical properties that can't be applied to a bike frame. Those springs are no longer in the linear range of deformation. That's an issue of elastic limit being easy to reach under normal usage loads, not the stiffness. As I said in my previous reply, it's fairly easy to design a spring that will fail. Especially if they are made out of used chewing gum, like bike parts often seem to be. Most springs see deformations that a bike frame will never see.

Just exploring the "repeatedly pushing past yield" point again, first, bike frames don't see this. Second, the elastic behavior of spring steels isn't normally like mild steel, but as it hits the elastic limit over and over again it probably is. Just hits a load where it deforms instead of pushing back more. At some point, springs and paperclips do things that aren't seen in a bike frame.