Are bike frames soldered or brazed?

[Keith Bontrager]

jim beam <[email protected]> wrote in message news:<%[email protected]>...
> > Did you read the Keith Bontrager article posted to this thread? That should answer any other
> > questions you might have.
>
> with all due respect to keith, his metallurgy was a little bit askew. shot peening does /not/
> reduce residal stress for example - i'm sure he meant to say it reduces fatigue. similarly
> comments about phase changes and annealing - should not occur to any significant degree during
> normal brazing - if it /is/ happening, it's being done wrong. finally, his tig welding h.a.z.
> comments do not acknowledge the role of air-hardening tube in this arena - it's specifically
> designed to negate the effects he describes. great comment otherwise.
>
> jb

Hi Jim,

Maybe I should clarify.

Shot peening DOES reduce the residual stress in the surface of the material. The residual tensile
stresses in the surface of the tubes near welds can be pretty high after the heating/cooling cycle
during welding or brazing. Peening reduces these. The affect of peening on the fatigue strength of
the welded joint is very evident in test results, especially on a steel or Ti weldment.

All brazing with brass alloys would be done above the transformation temperature of steel.

Most silver brazing operations performed on steel will bring the temperature of the base metal above
the transformation temperature. The liquidus of some of the cadmium bearing silver alloys is
slightly below that temperature, so a skilled framebuilder might be able to avoid anealing the
tubing if he was willing to risk the exposure to cadmium. But the tubing would still be tempered
significantly, and the hardness tests I've done indicated that the reduction in strength due to the
tempering is even greater than that incurred during annealing (where some self quench or air
hardening is possible).

Air hardening alloys came along as I was winding down on steel, so my experience with these is
limited. But, given that the temperature gradient is continuous from the puddle to the material that
is left at ambient temps in a welded joint, some point on the tube, away from the weld area, gets
very hot, but does not reach the critical temperature, so it cannot air harden. The material that is
just under that temperature would be tempered though, and softened considerably in that process. So
the metalurgical notch at that point in the tube would be a problem. It might be far enough from the
welded joint to minimize the reduction in fatigue strength that would be obvious. It would tend to
be a region that yielded in a crash or impact, and that type of failure (yielding at a narrow band
of material away from the weld) would reduce the fatigue strength of the joint. I never investigated
it in detail, so I am not sure of magnitudes.

Mr. Askew Metallurgy (aka KB, or Keith Bontrager).

 

[Matt O'Toole]

"Keith Bontrager" <[email protected]> wrote in message
news:[email protected]...

> Air hardening alloys came along as I was winding down on steel, so my experience with these is
> limited. But, given that the temperature gradient is continuous from the puddle to the material
> that is left at ambient temps in a welded joint, some point on the tube, away from the weld area,
> gets very hot, but does not reach the critical temperature, so it cannot air harden. The material
> that is just under that temperature would be tempered though, and softened considerably in that
> process. So the metalurgical notch at that point in the tube would be a problem. It might be far
> enough from the welded joint to minimize the reduction in fatigue strength that would be obvious.
> It would tend to be a region that yielded in a crash or impact, and that type of failure (yielding
> at a narrow band of material away from the weld) would reduce the fatigue strength of the joint. I
> never investigated it in detail, so I am not sure of magnitudes.

Interesting.

It just so happens that I had a frame a couple of weeks ago, with crash damage like that. The top
tube was buckled about 1cm behind the head tube. The down tube was cracked about halfway through,
from the bottom, in the same place -- probably right where this softened point that you describe is.
What's really interesting is that the buckling is so much neater than with older frames.

I guess the trick is to make the butted ends long enough and thick enough to compensate. Since this
frame showed some barely perceptible buckling at the shift bosses too, the middle of the tube and
the heat-affected ends were probably close in strength. So I guess the engineers at
Lemond/Trek/Reynolds did their job pretty well. Imagine that.

In case you're wondering about this frame -- I bought the bike on eBay, and it arrived broken like
that. To me it looked like someone drove it into a garage on a roof rack. I sent it back, and the
lame-o who sold it to me refunded my money.

I may still have the pictures in my camera. I could post them tomorrow if anyone's interested.

Moral of the story -- inspect all frames of unknown history very carefully.

Cheers,

Matt O.

 

[Jim Beam]

hey keith!

> Shot peening DOES reduce the residual stress in the surface of the material.

yes, it reduces tensile at the surface, but creates compressive residual stress instead. i'm no good
at this ascii stuff, but it's something like this: [use monospace font like courier]

+ | - residual stress

_______________________surface \ | \ | \|
| \
| \
| |
| |
| |

because the surface is in compressive residual stress, it reduces fatigue - most fatigue starts at
surfaces - hence your improved tests, but it does not mean residual tensile is removed across the
whole piece.

> All brazing with brass alloys would be done above the transformation temperature of steel.

you mean austenitic transformation? to be honest, i'm rusty on my brazing stuff, but iirc, you're
looking at ~800C + for that transition and for tubes that are heavily cold worked, you can/should
braze below that temperature or you're going to lose tube strength. and if you /do/ have an
austenitic transition and cool too rapidly, you get untempered martensite - brittle!

i know a lot of brazing, just as you say /is/ done at higher temperatures, and austenitic transition
does take place, but for high precision work, as you may imagine, it's not necessarily a good thing
and is best done on either lower end or air hardening tube.

> Most silver brazing operations performed on steel will bring the temperature of the base metal
> above the transformation temperature.

silver brazing is below 600C iirc, hence below transition.

> The liquidus of some of the cadmium bearing silver alloys is slightly below that temperature, so a
> skilled framebuilder might be able to avoid anealing the tubing if he was willing to risk the
> exposure to cadmium.

not much cadmium about these days! most silver brazing materials for steel, iirc, are ~450C - ~650C
working range.

> But the tubing would still be tempered significantly, and the hardness tests I've done indicated
> that the reduction in strength due to the tempering is even greater than that incurred during
> annealing (where some self quench or air hardening is possible).

talking about "tempering" makes me uncomfortable. to me, that implies martensite and the process
used to control its properties - unlikely if we're talking silver brazing. you get softening of the
tube on excess heating, for sure, but that's more akin to ageing & annealing processes. imo, if you
were getting getting significant softening, then the brazing temperature was too high and/or the
piece was heated for too long.

> Air hardening alloys came along as I was winding down on steel, so my experience with these is
> limited. But, given that the temperature gradient is continuous from the puddle to the material
> that is left at ambient temps in a welded joint, some point on the tube, away from the weld area,
> gets very hot, but does not reach the critical temperature, so it cannot air harden.

true.

> The material that is just under that temperature would be tempered though, and softened
> considerably in that process.

depends on the alloy. air hardened tube is highly complex stuff and intrinsically quite hard. i'm
not saying you're wrong, but i'd be reluctant to state that it's as susceptible to softening as a
tube that's more extensively cold worked.

> So the metalurgical notch at that point in the tube would be a problem. It might be far enough
> from the welded joint to minimize the reduction in fatigue strength that would be obvious. It
> would tend to be a region that yielded in a crash or impact, and that type of failure (yielding at
> a narrow band of material away from the weld) would reduce the fatigue strength of the joint.

correct, but as it's within the butted portion of the tube, the effect in service is debatable. most
cracks i've seen are well within the weld zone, not outside the h.a.z., the area you describe.

I never investigated it in detail, so I am
> not sure of magnitudes.

unfortunately, me neither. if someone knows though, it would be good to learn.

>
> Mr. Askew Metallurgy (aka KB, or Keith Bontrager).

great hearing from you, sir!

jb

 

[Keith Bontrager]

jim beam <[email protected]> wrote in message news:<[email protected]>...
> hey keith!
>
> > Shot peening DOES reduce the residual stress in the surface of the material.
>
> yes, it reduces tensile at the surface, but creates compressive residual stress instead. i'm no
> good at this ascii stuff, but it's something like this: [use monospace font like courier]
>
> + | - residual stress
>
> _______________________surface \ | \ | \|
> | \
> | \
> | |
> | |
> | |
>
> because the surface is in compressive residual stress, it reduces fatigue - most fatigue starts at
> surfaces - hence your improved tests, but it does not mean residual tensile is removed across the
> whole piece.

You are right. But the change at the surface is important with regards to fatigue strength,
especially in a welded part, and it works without question in many (but not all) situations
encountered on a bike frame.

> > All brazing with brass alloys would be done above the transformation temperature of steel.
>
> you mean austenitic transformation? to be honest, i'm rusty on my brazing stuff, but iirc, you're
> looking at ~800C + for that transition and for tubes that are heavily cold worked, you can/should
> braze below that temperature or you're going to lose tube strength. and if you /do/ have an
> austenitic transition and cool too rapidly, you get untempered martensite - brittle!

1) Check brazing alloys out. The only ones that will let you keep the temps below the transformation
temp are cad bearing, like BAG 1. I used to use that alloy because it worked so well. But it is
not common (or safe). Most brazed frames are tempered in the HAZ.

And (this is the important bit) even if you do use a brazing alloy that has a low enough liquidus
temp, and even if you are skilled enough with the torch to keep the steel under the transformation
temp, you temper the hell out of the tubes in the process. And, as a result, they are very weak in
that area. I measured it, so have many others.

The traditions passed on about brazing temperatures and alloys are not too accurate when it comes to
the "why you should do it this or that way". One of the things I was never able to resolve, but that
I suspect strongly, is that the higher brazing temperatures for brass and the non-cad bearing silver
alloys allowed the brazing ally to diffuse into the base metal much more, and that reduced the
fatigue strength of the joint. I think that is the best reason (if it's true) for a framebuilder to
use the low temperature silver alloys, if they are going to braze the frame at all.

> i know a lot of brazing, just as you say /is/ done at higher temperatures, and austenitic
> transition does take place, but for high precision work, as you may imagine, it's not necessarily
> a good thing and is best done on either lower end or air hardening tube.

I agree with th eformer, but the latter is not obvious to me. No sweat. I am not too keen on knowing
it anymore.

> > Most silver brazing operations performed on steel will bring the temperature of the base metal
> > above the transformation temperature.
>
> silver brazing is below 600C iirc, hence below transition.

Check the liquidus listed for common alloys.

I don't recall all of the numbers, but it is all available.

> > The liquidus of some of the cadmium bearing silver alloys is slightly below that temperature, so
> > a skilled framebuilder might be able to avoid anealing the tubing if he was willing to risk the
> > exposure to cadmium.
>
> not much cadmium about these days! most silver brazing materials for steel, iirc, are ~450C -
> ~650C working range.

Those numbers are too low.

http://www.turbobraze.co.uk/alloyssbacf.htm

http://www.turbobraze.co.uk/alloyssbacb.htm

http://www.sra-solder.com/brazing_wire.htm

I recall the lower transformation temp for steel is in the mid 1300s F. You can sort it out. I do
not know of a brazing alloy that will let you join steel tubes comfortably below the lower critical
temp other than BAG-1.

> > But the tubing would still be tempered significantly, and the hardness tests I've done indicated
> > that the reduction in strength due to the tempering is even greater than that incurred during
> > annealing (where some self quench or air hardening is possible).
>
> talking about "tempering" makes me uncomfortable. to me, that implies martensite and the process
> used to control its properties - unlikely if we're talking silver brazing. you get softening of
> the tube on excess heating, for sure, but that's more akin to ageing & annealing processes. imo,
> if you were getting getting significant softening, then the brazing temperature was too high
> and/or the piece was heated for too long.

There might be some slop in the terms that is causing trouble here too. The temperatures you
would use to braze a frame with the lowest temp silver are above common tempering temperatures.
It might be better to call it sub-critical annealing. Dunno. But the numbers are there no matter
what you call it.

The strengthening by cold working is not stable at high tempering/ low annealing temperatures, those
you would bring the tubes up to even if you were brazing with BAG-1. Doesn't matter who is holding
the torch. It is just the way it works.

Do the hardness tests. They are convincing.

> > Air hardening alloys came along as I was winding down on steel, so my experience with these is
> > limited. But, given that the temperature gradient is continuous from the puddle to the material
> > that is left at ambient temps in a welded joint, some point on the tube, away from the weld
> > area, gets very hot, but does not reach the critical temperature, so it cannot air harden.
>
> true.
>
> > The material that is just under that temperature would be tempered though, and softened
> > considerably in that process.
>
> depends on the alloy. air hardened tube is highly complex stuff and intrinsically quite hard. i'm
> not saying you're wrong, but i'd be reluctant to state that it's as susceptible to softening as a
> tube that's more extensively cold worked.

Doesn't matter. At some point away from the joint it is at the lower transformation temp for quite a
while. And the material next to it is cooler than that. It won't harden from that temp. Then the
"sub-critical annealing" bit applies.

> > So the metalurgical notch at that point in the tube would be a problem. It might be far enough
> > from the welded joint to minimize the reduction in fatigue strength that would be obvious. It
> > would tend to be a region that yielded in a crash or impact, and that type of failure (yielding
> > at a narrow band of material away from the weld) would reduce the fatigue strength of the joint.
>
> correct, but as it's within the butted portion of the tube, the effect in service is debatable.
> most cracks i've seen are well within the weld zone, not outside the h.a.z., the area you
> describe.

This notch is not outside the HAZ.

My experience indicates that the initial crack in a fatigue failure often starts in that notch after
crash damage and some local yielding. A good study with a prcise hardness tester will show you
what's happening. Without that this is all too vague to be useful. I did a lot of that in college
but it's been 25 years since and the numbers are gone now.

But the fundamentals are what was behind the development of tig welded, reinforced frames. That
approach solved the strength issues we encountered with brazed or welded (but unreinforced) steel
frames. That method has stood up well over time.

> I never investigated it in detail, so I am
> > not sure of magnitudes.
>
> unfortunately, me neither. if someone knows though, it would be good to learn.

let fly someone...

> > Mr. Askew Metallurgy (aka KB, or Keith Bontrager).
>
> great hearing from you, sir!

I vowed not to do this ever again.

I am pretty sure I won't again for some time. but it was fun.

> jb

I am off to Costa Rica and La Ruta to do a hardness test on my legs (and butt) now.

;-)

KB