Comparison of Auminium, Steel and Carbon forks?

"jim beam" wrote:
[...]
wrong end of the stick mike. it's a property of the material. you go
ahead and show how much hysteresis and therefore absorption /doesn't/
happen. [using a spell check wouldn't hurt you either.]

Funny that someone who never uses capital letters where they are
required should complain about the spelling of others.

--
Tom Sherman - Holstein-Friesland Bovinia
The weather is here, wish you were beautiful

On Apr 20, 11:33 am, Ben C wrote:
On 2008-04-20, wrote:
[...]

Right. Three forks of the same make and model should be exactly the
same (although variance is likely to go up substantially at the bottom
end of the market), but you can't just test one carbon fork next to
one aluminum fork and say that the difference is the material.
Whatever vibration parameter you're measuring is going to have a big
enough range across different models of carbon forks that it's going
to overlap with the range across different models of aluminum forks.

Sometimes different materials dictate a different shape.

For example, I think I'm right in saying that if you make an aluminium
fork as flexy as you can make a steel one the aluminium one will fatigue
badly. So you have to make it a bit thicker and/or fatter and stiffer.
Sure you _could_ make a fork just as stiff out of steel, but you don't
have to and might not.


Material dictates shape, but there's still a working range. If you
have a steel tube of a given flexural strength, you can match that
strength with aluminum by using an appropriate ratio of wall thickness
to tube diameter. The aluminum tubes are going to range from big with
thin walls (stiffer than the steel tube) to small with very thick
walls (less stiff than the steel). Where the stiffness of the steel
tube falls in the working range for aluminum will depend on how big
and how stiff it was to begin with.


So it's better to compare complete forks as sold and then say something
like "out of the 100 forks tested, the CF ones mostly absorbed vibration
better than the Al ones".

Exactly. I would even plot the distribution curves to see how much
overlap there was.

On Apr 20, 10:28 pm, wrote:
On Apr 20, 3:10 pm, jim beam wrote:



Ben C wrote:

snip for clarity

I would expect there to be much more likely to be non-negligible
differences between forks.

of course. krygowski, being an intelligent, informed engineering
professor with access to the correct instrumentation, already the owner
of a cross-section of different forks, a sound knowledge of the
principles, and with an open, inquiring mind, has already tested this
position and is simply waiting for an opportunity to publish his
results. or he's simply an idiot voicing underinformed opinion as fact
and who has no inclination to actually test any damned thing that could
possibly upset his comfy luddite little world.

False dichotomy, jim. One doesn't have to personally run tests and
publish results to recognize the bull**** component of the advertising
hype that pops up in bike magazine ads and articles - things like
"rigid, yet compliant," "superfoods that increase healing power,"
"sealing gaps at the molecular level reduces friction at racing
speeds." Yes, and "incomparable, magic ride quality."

And a careful reader will note that I was simply giving my speculation
on what a fork comparison test would show. While I'd be willing to
bet with my friends on the issue, I wouldn't testify in court unless
I'd performed the proper test, or seen results I judged worthwhile.

Again, the worthwhile test would be a blind, on-road comparison test
using multiple riders, where the other factors were held constant.
It's the only way to filter out the placebo effect.

- Frank Krygowski

An even easier test is to take a carbon fork fresh out of the box,
grab it by the steerer, and whack one of the dropouts against
something. It's going to hum like a tuning fork for several seconds.
It may seem clear to some that the polymer matrix in the composite is
soaking up significant amounts of energy, but there are two things
going against that theory. 1) Polymers dissipate energy proportional
to their volume and the amount of strain they're under. A carbon fork
does not contain enough epoxy under enough deformation to dissipate a
significant amount of energy. 2) Manufacturers are constantly pushing
to use less and less epoxy in their layups. It's dead weight, and
doesn't contribute to the strength and stiffness of the fork. In all
of their advertising, they somehow forget to explain what their latest
revolutionary process (it was nanotubes last I heard) to increase
fiber volume ratio for increased strength to weight was doing to the
fork's imaginary damping properties.

On 2008-04-22, wrote:
On Apr 20, 11:33 am, Ben C wrote:
On 2008-04-20, wrote:
[...]

Right. Three forks of the same make and model should be exactly the
same (although variance is likely to go up substantially at the bottom
end of the market), but you can't just test one carbon fork next to
one aluminum fork and say that the difference is the material.
Whatever vibration parameter you're measuring is going to have a big
enough range across different models of carbon forks that it's going
to overlap with the range across different models of aluminum forks.

Sometimes different materials dictate a different shape.

For example, I think I'm right in saying that if you make an aluminium
fork as flexy as you can make a steel one the aluminium one will fatigue
badly. So you have to make it a bit thicker and/or fatter and stiffer.
Sure you _could_ make a fork just as stiff out of steel, but you don't
have to and might not.


Material dictates shape, but there's still a working range. If you
have a steel tube of a given flexural strength, you can match that
strength with aluminum by using an appropriate ratio of wall thickness
to tube diameter. The aluminum tubes are going to range from big with
thin walls (stiffer than the steel tube) to small with very thick
walls (less stiff than the steel). Where the stiffness of the steel
tube falls in the working range for aluminum will depend on how big
and how stiff it was to begin with.

OK I have two questions.

Strength is one thing, fatigue life is another. Any component has to
have enough of both.

For fatigue life we have the S-N graph of mean stress against number of
cycles. You need to reduce the stress (by increasing the cross-section
for the sake of argument) to the point where the component will survive
plenty of cycles (I don't know, 10^6 or something).

The component also needs to be strong enough so as not to yield when you
hit a reasonably normal bump in the road and so forth.

An ideal component engineered for the lightest weight has just enough of
each of strength and fatigue life and no more.

So I'm making a tube of a given length out of a given volume of metal. I
have two parameters: wall thickness and tube diameter. If I'm just
pulling on the tube (loading it axially) all that matters I think is
cross-section, so it's moot whether I go for almost a solid rod or a
thin-walled "oversize" tube.

So let's say I'm bending it. It might be the top tube on a bike (that
might be simpler than the fork). Suppose it's strong enough, but its
fatigue life is too short. Would it help to make the tube fatter but
thinner-walled, or vice versa, or not?

My second question is: if I make two tubes that are both just right for
both yield strength and fatigue life for a particular application, one
out of steel and one out of aluminium, is it inevitable that one tube
will be stiffer than the other, or is there enough working range that
after satisfying the strength and fatigue requirements the designer can
also make them the same stiffness? (Of course they will have different
diameters and thicknesses).

On 4/22/2008 9:13 AM wrote:

On Apr 20, 11:33 am, Ben C wrote:

So it's better to compare complete forks as sold and then say something
like "out of the 100 forks tested, the CF ones mostly absorbed vibration
better than the Al ones".

Exactly. I would even plot the distribution curves to see how much
overlap there was.

Whoa. Too useful.

--
Mike "Rocket J Squirrel"

On Apr 22, 5:56 pm, Ben C wrote:
On 2008-04-22, wrote:



On Apr 20, 11:33 am, Ben C wrote:
On 2008-04-20, wrote:
[...]

Right. Three forks of the same make and model should be exactly the
same (although variance is likely to go up substantially at the bottom
end of the market), but you can't just test one carbon fork next to
one aluminum fork and say that the difference is the material.
Whatever vibration parameter you're measuring is going to have a big
enough range across different models of carbon forks that it's going
to overlap with the range across different models of aluminum forks.

Sometimes different materials dictate a different shape.

For example, I think I'm right in saying that if you make an aluminium
fork as flexy as you can make a steel one the aluminium one will fatigue
badly. So you have to make it a bit thicker and/or fatter and stiffer.
Sure you _could_ make a fork just as stiff out of steel, but you don't
have to and might not.

Material dictates shape, but there's still a working range. If you
have a steel tube of a given flexural strength, you can match that
strength with aluminum by using an appropriate ratio of wall thickness
to tube diameter. The aluminum tubes are going to range from big with
thin walls (stiffer than the steel tube) to small with very thick
walls (less stiff than the steel). Where the stiffness of the steel
tube falls in the working range for aluminum will depend on how big
and how stiff it was to begin with.

OK I have two questions.

Strength is one thing, fatigue life is another. Any component has to
have enough of both.

For fatigue life we have the S-N graph of mean stress against number of
cycles. You need to reduce the stress (by increasing the cross-section
for the sake of argument) to the point where the component will survive
plenty of cycles (I don't know, 10^6 or something).

The component also needs to be strong enough so as not to yield when you
hit a reasonably normal bump in the road and so forth.

An ideal component engineered for the lightest weight has just enough of
each of strength and fatigue life and no more.


Well, plus a factor of safety. Nobody trusts their analysis, their
material supplier, and their manufacturing tolerances enough to go
right to the line. Whether or not a single design satisfies both of
those constraints depends on what you want it to do. Only if the
single cycle load requirement is much larger than the average cyclic
load requirement or the fatigue life is very low will one design
exactly meet both requirements. Typically, one of them will dominate.



So I'm making a tube of a given length out of a given volume of metal. I
have two parameters: wall thickness and tube diameter. If I'm just
pulling on the tube (loading it axially) all that matters I think is
cross-section, so it's moot whether I go for almost a solid rod or a
thin-walled "oversize" tube.

So let's say I'm bending it. It might be the top tube on a bike (that
might be simpler than the fork). Suppose it's strong enough, but its
fatigue life is too short. Would it help to make the tube fatter but
thinner-walled, or vice versa, or not?

My second question is: if I make two tubes that are both just right for
both yield strength and fatigue life for a particular application, one
out of steel and one out of aluminium, is it inevitable that one tube
will be stiffer than the other, or is there enough working range that
after satisfying the strength and fatigue requirements the designer can
also make them the same stiffness? (Of course they will have different
diameters and thicknesses).

That's a good question, and to be honest it's more math than I feel
like doing. If you put three exact constraints (volume, yield
strength, and fatigue strength for n cycles) on a problem with three
dependent variables (diameter, thickness, and stiffness), any non-
trivial solution should be unique. So, there's no working range
left. The system would still need to be worked out for every possible
steel tube that fits your constraints to say that one is always
stiffer.

Ben C wrote:

So let's say I'm bending it. It might be the top tube on a bike (that
might be simpler than the fork). Suppose it's strong enough, but its
fatigue life is too short. Would it help to make the tube fatter but
thinner-walled, or vice versa, or not?

Yes, the larger cross section will have lower stress for the same load.


My second question is: if I make two tubes that are both just right for
both yield strength and fatigue life for a particular application, one
out of steel and one out of aluminium, is it inevitable that one tube
will be stiffer than the other, or is there enough working range that
after satisfying the strength and fatigue requirements the designer can
also make them the same stiffness? (Of course they will have different
diameters and thicknesses).

The two materials are different, so that makes direct comparison
problematic. Generally, to get comparable fatigue life, the aluminum
will use a bit more material than needed for comparable strength, so you
have a choice of similar fatigue, but aluminum stronger, or similar
strength with aluminum less durable. The limit for steel tubing diameter
is dent/crumple resistance. This has been all pretty well worked out
empirically in bike frames over the last couple of decades.

Peter Cole wrote:

So let's say I'm bending it. It might be the top tube on a bike
(that might be simpler than the fork). Suppose it's strong enough,
but its fatigue life is too short. Would it help to make the tube
fatter but thinner-walled, or vice versa, or not?

Yes, the larger cross section will have lower stress for the same
load.

Another aspect of this is that for the same deflection, stress is
higher in the larger cross section element. For this reason, I have
seen more wrinkled aluminum than steel tubes from bending. Before the
advent of aluminum frames, steel frames wrinkled their downtubes at
the transition from thick walled tube ends to thin wall. Forks in
contrast did not wrinkle, having no localized wall thickness
transition nor excess "beam height" as did aluminum forks.

My second question is: if I make two tubes that are both just
right for both yield strength and fatigue life for a particular
application, one out of steel and one out of aluminium, is it
inevitable that one tube will be stiffer than the other, or is
there enough working range that after satisfying the strength and
fatigue requirements the designer can also make them the same
stiffness? (Of course they will have different diameters and
thicknesses).

The two materials are different, so that makes direct comparison
problematic. Generally, to get comparable fatigue life, the
aluminum will use a bit more material than needed for comparable
strength, so you have a choice of similar fatigue, but aluminum
stronger, or similar strength with aluminum less durable. The limit
for steel tubing diameter is dent/crumple resistance. This has been
all pretty well worked out empirically in bike frames over the last
couple of decades.

To avoid the stiffness problem (that increases with the third power of
"beam height"), that causes failures elsewhere, some aluminum forks
were made to external steel dimensions but with greater wall
thickness. That was a dud and you probably cannot find such a frame
still in regular use. Weight weenies have done little good for the
durability and repairability of bicycles yet that is what sells.

Jobst Brandt

On 2008-04-25, Peter Cole wrote:
Ben C wrote:

So let's say I'm bending it. It might be the top tube on a bike (that
might be simpler than the fork). Suppose it's strong enough, but its
fatigue life is too short. Would it help to make the tube fatter but
thinner-walled, or vice versa, or not?

Yes, the larger cross section will have lower stress for the same load.


My second question is: if I make two tubes that are both just right for
both yield strength and fatigue life for a particular application, one
out of steel and one out of aluminium, is it inevitable that one tube
will be stiffer than the other, or is there enough working range that
after satisfying the strength and fatigue requirements the designer can
also make them the same stiffness? (Of course they will have different
diameters and thicknesses).

The two materials are different, so that makes direct comparison
problematic. Generally, to get comparable fatigue life, the aluminum
will use a bit more material than needed for comparable strength, so you
have a choice of similar fatigue, but aluminum stronger, or similar
strength with aluminum less durable. The limit for steel tubing diameter
is dent/crumple resistance. This has been all pretty well worked out
empirically in bike frames over the last couple of decades.

Thanks for the answer. I'm thinking then that it is quite likely that in
practice aluminium frames and forks _will_ be stiffer than steel ones
(when comparing similar styles of bike) as a consequence of designing
them for sufficient fatigue life?

But I still don't know really-- the aluminium frame may be stronger, but
aluminium has a lower modulus than steel, so it's not obvious that it's
going to be stiffer.

On 2008-04-25, wrote:
Peter Cole wrote:
[...]
My second question is: if I make two tubes that are both just
right for both yield strength and fatigue life for a particular
application, one out of steel and one out of aluminium, is it
inevitable that one tube will be stiffer than the other, or is
there enough working range that after satisfying the strength and
fatigue requirements the designer can also make them the same
stiffness? (Of course they will have different diameters and
thicknesses).

The two materials are different, so that makes direct comparison
problematic. Generally, to get comparable fatigue life, the
aluminum will use a bit more material than needed for comparable
strength, so you have a choice of similar fatigue, but aluminum
stronger, or similar strength with aluminum less durable. The limit
for steel tubing diameter is dent/crumple resistance. This has been
all pretty well worked out empirically in bike frames over the last
couple of decades.

To avoid the stiffness problem (that increases with the third power of
"beam height"), that causes failures elsewhere, some aluminum forks
were made to external steel dimensions but with greater wall
thickness.

This sounds interesting but I'm not sure I understand exactly what
you're saying. What's "beam height"-- is it tube diameter in this
context-- and what were the failures elsewhere?

That was a dud and you probably cannot find such a frame still in
regular use.

What was wrong with it?

Weight weenies have done little good for the durability and
repairability of bicycles yet that is what sells.

Well they are weight weenies not durability or repairability weenies.