When steel frames fail, and why
Mon, December 14, 2015 My last article drew the following comment that gave me food for thought:
I have seen a frame fail at the downtube shifter braze-on. The down tube was installed backwards, the short butt was at the top so this braze-on was installed in the thin center part of the tube. It weakened it enough that it buckled there.
First let me explain that the down tube in this case was not necessarily installed “Upside Down.” Double butted tubes are so-called because they are butted at either end. The butt at one end is longer. That is the end that is cut shorter when building smaller frames.
So yes, if the framebuilder was building a large frame and using the full length of the tube, he could put the long butt at the top, but this would not necessarily be normal practice.
In a production setting, which included small batch production like the Fuso, the down tubes would be pre-mitered at the top end, because that angle was the same across a range of sizes. The long butt was left at the bottom bracket end so it could be cut later to accommodate whatever size frame I was building.
Incidentally, a simple way to tell which end has the long butt, is to balance the tube on a finger in the center. The longer butted end is obviously heavier. Tubes are also usually marked with the maker’s name or trade mark on the end that is not cut, but these marks can get stamped on the wrong end, so best to double check.
The short butt is usually 3 or 4 inches of the heaviest wall thickness, then it tapers down gradually for 3 or 4 more inches to the thinnest part in the middle of the tube. So the gear lever boss would not normally be on the absolute thinnest part of the tube, and even if it was, under normal circumstances it shouldn’t fail.
The commenter mentioned that the tube buckled. This usually means the bike had a front end shunt at some point, and it does not have to be a serious crash. I remember one time in the UK, I built myself a brand new cyclo-cross frame. The first time out I dropped my front wheel in a mud hole, and did a spectacular vertical stand on my front wheel.
I did not go over the top, but simply fell over sideways. Later I noticed the down tube was buckled right behind the bottom head lug. Once a tube is rippled, it will crack and eventually fail.
Barring such accidents a good steel frame will last fifty years or more. Ridden hard enough and long enough metal fatigue will eventually cause it to fail. But how many frames are ridden that hard and that long? Although sometimes a tiny crack can happen during building.
Metal like wood has a grain. Actually nothing like wood, but the only reason I draw that parallel is to remind me that wood will crack or split along the grain, whereas metal will usually crack across the grain.
When metal cools from its molten state, it forms a crystalline structure. Steel is then often cold rolled into bars or sheets. Wire and tubes are drawn though dies. Either process crushes and elongates the crystals in the metal forming a grain that runs along the length of the bar or tube. This actually strengthens the metal. (See above picture.)
I found from experience that damage can be done to the very thin bicycle tubes, not only by overheating, but more often than not, by heating cold metal too quickly. Overheating while brazing causes the brass to flow in between the crystals of the steel, thus weakening it greatly. However, for this to occur the metal would have to be white hot and in the verge of melting.
More common is heating too quickly and this often happens when tacking a frame together. The metal is cold, and the framebuilder comes in with a small, hot flame to put a little blob of brass to hold the tubes in a lug, or a part like a brake or chainstay bridge.
Metal expands when it is heated, but if the metal is heated in one tiny spot, the surrounding cold metal will not expand and a minute crack can form, often so tiny it cannot be seen with the naked eye. The crack can fill with brass and may not fail until many years down the road.
My advice. Preheat the area first, and always tack at a point where the grain in the tube is a 90 degrees to the component part you are tacking. Not parallel with the grain. (See picture above.) Follow this simple rule and there will be less chance of a tube cracking.
This really applies to the initial tacks when the metal is cold. After two or three tacks and the metal warms up, others can be safely added. And don’t forget when fully brazing later and the frame is cold again. Start in a safe place at right angles to the grain, although not necessarily the same place or you will melt the original tack.
Earlier I mentioned a front end shunt, or crash. When this happens either the down tube ripples, or the front fork bends, occasionally both will happen. If the down tube ripples, it will break eventually, and so needs replacing. It will not fail suddenly, a crack will appear first.
If the front fork gets bent, don’t replace it unless the fork blades are rippled. It can be safely straightened. Let’s face it, the fork blade was first rolled into a round tube. This was done while the tube was in a cold state.
Next it was rolled into a taper and during this operation the wall thickness increased at the thin end. The excess metal has to go somewhere, right. The top end was pressed from a round to an oval shape. All these operations where done while the tube was in a cold state, no heat was required. Cold working actually strengthened the steel.
Finally the framebuilder cold bends the fork blade into a curve. So if the fork is bent slightly in a front end crash, and re-straightened (Cold.) by a skilled person with the right tools and know-how. Why should that compromise the integrity of the fork?
Of course I am not advocating you bend and re-straighten a fork more than once, but that is the beauty of steel. It will rarely fail suddenly, and when it does it can be fixed quite easily.
Reader Comments (12)
" A good steel frame will last fifty years or more". I'm hoping on much more as my 1970s bikes are earmarked for my youngest sons. Yes I have pampered a few, ridden them lightly and only take them out for easy Sunday rides in nice weather.
These frame lessons are great, keep them coming.
Your comments regarding metal grains and cracking hit home. As a former Navy pilot my engines relied on turbine blades that were (literally) grown from a single crystal of steel. Monocrystalline blades were incredibly strong with minimal thermal creep and the ability to last tens of thousands of hours bathed in exceptionally hot gases. Think how amazing it would be if we could "grow" steel tubes that were composed of a single crystal of steel. Welding with possibly overheating? Never a problem.
Technology marches on.
One of the beauties of steel is that the underlying crystal structure is cubic, and with each grain having a random orientation this results in a materiel with the same properties in all directions, even when the grains are elongated in one direction (this is not true with Ti, but that is another discussion).
Dave's observation is partially correct, in that you don't want to stress the most highly loaded areas of a joint. The abrupt and concentrated heating to tack is tough on the metal, and you also risk not having the best joint there. Sometimes it is difficult to reflow the tack braze and you end up with a less than prefect joint.
As Dave points out the bottom side where the down tube meets the head tube is the most highly stressed point in a frame, and it doesn't take much impact to show it.
The frame in question in the original quote did suffer an impact, though not an extreme one. It had been designed to have the bossed in the heavy sections and the down tube was installed backwards (there is one tube company that marks the end that you cut).
This whole discussion points out one key reason why we like steel frames, damage tolerance. Their combination of fatigue resistance and impact toughness result in a frame that can tolerate and absorb damage without catastrophic failure. And that gives me great comfort.
Insightful! Thanks!
I purchased a 1984 Trek 760 and found that the fork had been replaced with a unicrown marked with Giant. The original fork was gone for whatever reason, I don't' know. The TT and DT were check and found to be straight. Trying to find a 760 fork is like looking for a needle in the haystack. Found a 1985 fork, but it was bent back. It was your article and comments that encouraged me to bend it back, gently.
What is unique about this fork, other than the "Trident" crown, is the off set of 38 cm, a track fork off set. I have a friend who has built his own frames, Last one all RS tubes and lugs, with a jig to check the dimensions of my fork post adjustment.
With a repaint, this frame and fork will look paired. Yes, not only was the fork from a 760 but from the same frame size so the steerer was the right length. The steerer was straight too, only the blades bent.
One often hears of the purported long service life of steel frames and the frailty of carbon fiber composite frames and how they are one or two season throw-aways. I would like to offer my perspective from being an amateur bicycle racer of twenty years experience and a professional bicycle mechanic with more than thirty years experience. I have had the following failures with my steel frames: cracked seattube, cracked chainstay (3x), cracked chainstay bridge, cracked headtube. I have owned three carbon fiber composite frames over the last twenty-five years that have had the following failures: none. OK, that's a small sample; however, in my work I have seen far too many steel frame failures to be able to even guess at the number. I don't have to guess at the number of carbon fiber composite frame failures I have seen. That number is zero. I realize that this is all anecdotal so clearly not "scientific" data, but it makes me wonder why it is so contrary to "common knowledge". Help me here.
On having Carbon Fiber bikes for the past 25 yeas:
In 1990, there weren’t any. Indurain won the 1992 Tour on a steel bike. In 1998, Pantani won the Tour on an AL framed Bianchi with carbon forks. It wasn’t until after 2000 that full carbon bikes became more common.
That’s a 15-year history of readily available carbon frames.
I think one will find failures with any frame material, if you bother looking. But what is the point?
Paul T's comments regarding carbon fiber are also relevant to former Navy pilots. Not only were F-18 wings made of carbon fiber but most of the helicopter blades used today are likewise.
Wear and tear? They seem to be lasting nearly forever.
But I think the bottom line is this: Any material can be pushed to failure, even bamboo!
The reason that composite helicopter blades last is the high strength 17-4PH stainless steel leading edge that protects them.
Composite frames are fine, provided that they are suitably over built (until recently a given) and if there is any visible damage you stop using them.
Most (OK, All in my case) steel frame failures that I have seen showed themselves long before the actual failure. People felt over confident and continued to ride visibly damaged frames. Sometimes they got away with it for years.
At the same time there are many cases of frames being ridden hard for decades. Are they over built, sure. Would making them lighter improve their function, usually not.
Dave, just a curiosity (and a moot point today):
Why weren’t steerer tube threads rolled, rather than cut, which removes metal? The tread would have been stronger, but more importantly, it may not have weakened the threaded area. Thus retaining its strength.
Sorry for the thread hijack, but I feel my credibility has been questioned.
@Steve: My first carbon fiber composite bicycle was a 1990 Trek 2300. Yes, this was not constructed of only carbon fiber composite and I made no representation that it was. You could say the same of my present bicycle, a 2014 S-works Tarmac. That's not the point.
My question is, based on my rather extensive experience with both steel and carbon fiber composite frames, why are steel frames reputed to be so much more durable?
Thank you.
Backing up Paul Thober - CF frames go back to 1975, if you count the Graftek, which was made of tubes sandwiching CF and AL. In 1985, Vitus began producing its Carbone frame, which was CF tubes bonded to AL lugs. The first all-carbon frame was the Kestrel in 1988.
Carbon, like Boron, has to be joined with a bonding agent to be useful for bike frames. The agent also provides added resilience. (Klein glued Boron strips inside AL fork blades for a time.)
So your carbon fiber frames are as much glue. Otherwise, it would be even lighter.
Dave is merely pointing out How Steel frames can fail. He isn’t selling anyone on steel, nor comparing carbon frames. If anyone needs to be sold, just use Google, and one will find plenty of support (and counter) for whatever material your bike is made from.
But do you really need to?