The myth of "stiff frames" and "stiff wheels"

[Arty50]

Is this a picture of phUnk?

Originally posted by jayfrizzo

[jayfrizzo]

That's a picture of what I'm going to shove up his ass after I club him with it.

[Plakespear]

Originally posted by jayfrizzo
That's a picture of what I'm going to shove up his ass after I club him with it.

Dude, that's not going to help dispel the notions of your homosexuality.

[Spats]

You can argue with conclusions, but you can't argue with good data. So it is incumbent upon *you*, defenders of the "frames/wheels ride rough/smooth" hypothesis, to explain how variations of approximately 1/1000 of an inch are detectable when traveling at speeds averaging 20 MPH.

I'm taking the simplest explanation: they aren't, and you're sensing something else.

I have no trouble believing that those bikes feel different to you, or to anyone else, including me. Therefore, I look for alternative explanations. Let's take the example of freshies' Ti and Al road bikes.

Did they have identical forks? The article measured fork flex as the second largest deflection, behind tire flex.

Did they have the exact same seat? Seat rail flex and padding are at least as significant as fork flex.

Did they have the exact same stem and handlebars and grip tape and shifters/brake hoods? The material your hands are resting on is way more important to any vibrational "feel" than four inches of head tube that's doing very little besides keeping the fork vertical.

Is the geometry of the two frames exactly the same?

Do they make the same noises when they hit a bump? (See my earlier post about sound influencing perceived picture quality.)

I'm going to bet that *every single one* of these things was different between your two bikes, with the possible exception of the seat.

This seems a much more reasonable explanation for the different "feel" of different bikes.

[lph]

spats, in my case, i had a trek aluminum 2200 frame that got bent. it came with the carbon fiber fork, which also got bent, i bought a new trek 5200 frame (which came with the same fork) and simply transfered everything over to the new frame.

the difference was immediately noticeable. you have been on my street, it is mess, cracks, pot holes, filled in potholes. the amount of vibration i 'felt' seemed to me to be hugely different to me. I will grant you, as a somewhat of scientist that it may simply be the power of suggestion at work, people tell me that the carbon fiber frame would suck up more of the hits on the road than the aluminum. I believe that is entirely possible. placebo effect.

whatever, the case, i like my new frame alot more!!

[Arty50]

Originally posted by Spats
You can argue with conclusions, but you can't argue with good data.

That's just it. It's not good data.

Static loading and dynamic loading are two completely different ball games. Static forces occur along a singular plane at a specific frequency. Dynamic forces not only occur on multiple planes, but also at varying frequencies.

When you're out on the road your hitting surface features that present dynamic forces on the bike. In other words, that pothole you just hit creates forces with multiple vectors over a very short time. Dropping into it initially causes the fork to flex outward slightly, the middle of the pothole causes the bike to bottom out, and the final cusp of it will cause an upward jolt along with flexing the fork backward (note that I'm even simplifying this). All of this is of course flexing the headtube which is then spread over the entire triangle in various ways.

So a bike is actually a complicated system that is designed to spread forces occuring in various directions and with various magnitudes and frequencies. How this occurs is naturally affected by the type of materials used. And some materials handle certain forces better than others. So you have to ask yourself what type of forces confront the system of a bike: the weight of a rider, singular large shocks, smaller yet frequent bumps, etc.

The tests done by these "scientists" were static; as they occur only on the verticle plane and without any frequency to speak of. Weighting the bike like this would only be indicative of the force created by a riders weight. It completely neglects all of the forces that are caused by the system's interaction with the road. These are forces that cannot be replicated by a simple weight.

This is why I tend to pay little attention to stuff like this. All it does it cause stupid discussions like this. Bob Bundy should stick to psychology.

[Crinkle]

THIS JUST IN......Einstein was wrong.

[jayfrizzo]

Originally posted by Plakespear
Dude, that's not going to help dispel the notions of your homosexuality.

My hot wife does that for me. So I'm left free to lay on the homo-humor.

[phUnk]

Originally posted by lph
personal experience

Yeah, my shiny new bike (32 pounds) was a lot faster than my old, dirty beat-up bike (28 pounds) too. But then it got dirty and slowed down.

[Spats]

Arty's argument is the best one I've heard so far. I would caution that there's nothing magic about dynamic forces: a force over a short time must bend the frame less or equal to an identical force applied for a long time.

I've found some interesting papers, but none that produce a direct conclusion. There's some interesting stuff here:

http://www.sheldonbrown.com/rinard/fea.htm

The FEA-modeled strain percentage charts point out that axial deformation and in-plane bending are very small relative to out-of-plane bending and torsion -- but this is under hard pedaling, not under pavement irregularities.

Here's a list of interesting bicycle tech-related articles:
http://www.sheldonbrown.com/rinard/

[Plakespear]

I remember about 12 years ago Bicycle Guide decided to test the differences between various Columbus tubing. They had a custom builder make several bikes with identical dimensions, and spec'ed the bikes with the same component gruppos. The test riders had no idea which frame was made out of which tubing. The testers were able to feel differences between the various bikes, even though they didn't know which they were riding.

[Arty50]

Originally posted by Spats
Arty's argument is the best one I've heard so far. I would caution that there's nothing magic about dynamic forces: a force over a short time must bend the frame less or equal to an identical force applied for a long time.

Yes and no. But there are different types of forces; and that's why materials undergo various types of tests, one of which is stress testing. If you've ever done a stress/strain test on a material, you'd know that the amount of force and duration of the force are two important factors. Cyclical loading comes into play also. You're seeking to refute my argument on one small point. I'm saying that multiple factors, not just the amount of force, come into play. For instance, a huge one is the direction of the force. Let's examine stress testing though.

The test you listed above was done with light weight in a 100% vertical fashion. Also, while we're talking about stress/strain tests. When undertaking them, you soon learn that materials do not act in a purely linear fashion according to the stress/strain placed upon them. Essentially the test is done by placing a bar of the material in an apparatus that exerts an increasing level of linear force on the bar. In the case of metals, they all have a initial period where they exhibit linear elasticity. In this case the level of deformation is linear to the force applied. Past that point (at a certain level of force) the metal ceases to return to it's initial demensions and begins to deform. Past that it will eventually fail catastophically. As I recall, the rate at which the forces are increased can also change the results of the test. And cyclical loading can also cause the material to deform/fail more rapidly.

Now I know that we're not talking about failure of components here. But we are talking about how various materials interact to similar forces. And I'm saying that putting a light weight on a bike in a purely linear fashion is a grossly inadequate test of the materials being used. All sorts of things are being neglected in this sort of test: cyclical loading, the frequency of the cyclical loading, the direction of the forces created by the uneven features found on a road, and the ability of a given material to react or avoid deformation in these instances.

This test neglects all of these things. There's a reason people feel a difference in the different frame materials. I can't say exactly why without conducting a battery of tests with equipment I no longer have access to. But suffice it to say, this overly simplistic test essentially renders our dear psychologists conclusions as a load of shite.

Edit: Forgive the last sentence, but I think it's important. This tester is a psychologist, NOT a materials engineer/scientist. Heck, I'm just a hack who took mechanical engineering for a couple of years in college. But I know enough to say that this test is extraordinarily limited in scope and thus shouldn't be used to validate his conclusion.

[Lunch]

I read the article and the only conclusion that can be drawn is that psychologist should not be conducting engineering experiments. The only reason I read the whole article is because he mentions Jobst Brandt who is very vocal on the boards and who does have some knowledge about bicycle technology. At least enough to write a book about wheel building.

Anyway, I was just going to ignore this whole post because the test was obviously not grounded in science and it would be a waste of time to comment on it. Then Arty50 did the footwork so all I have to say is” I basically agree with Arty50”. In fact, Arty50 only begins to mention all the inadequacies of this garage test. A static 80# vertical force applied to a few points along the main triangle doesn’t even begin to mimic the forces acting on a bike being ridden. The psychologist that conducted and posted this experiment even seems to concede this point. I just want to quickly comment what is meant by how a bike or frame “feels”.

How a bike “feels” is probably a combination of many things including “stiffness” (i.e., how a frame deforms and behaves under various loads). Bike stiffness is a function of many frame factors including geometry (e.g., tube length), material properties, tube diameter & wall thickness, fork design, etc. AND many non-frame components like the bars, seat, cranks, etc. In general, stiff is good because it’s efficient. Less energy is wasted to deformation. Since not all tube components of a bike experience the same loading, you can trade-off stiffness for weight reduction. (I have a frame made up of 4 different types of steel.)

But how a bike “feels” is probably much much more including vibration transfer and material properties which has obviously been discussed here. Geometry also contributes to the “feel” (“quick”, “responsive”, etc.). Bike “feel” seems to be very subjective. Many simple things obviously could change what a rider experiences (e.g., a different saddle). Most people probably can’t discern what exactly is most responsible for a different “feel”. I’m sure there is (can be) a strong psychological component to feel and lots of exaggeration. Advertisement is very effective. On the other hand, I don’t discount what riders describe. I don’t dare speculate about how much change is being detected and how much is in a person’s head. I wouldn’t be surprised if Lance Armstrong could “feel” a chip in his frame paint. On the other hand, I’m sure that millions of people brag about ghost frame attributes.

[Plakespear]

Originally posted by Lunch
I wouldn’t be surprised if Lance Armstrong could “feel” a chip in his frame paint.

Even though he couldn't "feel" that he was riding a cracked frame on an alpine Tour de France stage?

[Oarhead]

The only thing that that 'scientific test' tested (?that was a fucked up intro) was how different bikes feel when you are at the start line completely motionless. Of course nobody can tell the difference between frames, NOTHING IS HAPPENING!
When you do this:


you are putting all sorts of forces and moments on the bike. Yes extra components make a difference but vibrational dampening and torsional resistance are critical to the 'feel' as well as the performance of the bike.

I agree there is a lot of hype with bike stuff, but this 'test' is such a pile of garbage. Thanks to arty for the leg work and articulating it nicely.

[InspectorGadget]

Originally posted by Oarhead
Thanks to arty for the leg work and articulating it nicely.

Seconded.



Recently I was thinking about buying a new road frame; I spent quite a bit of time reading the reviews on the Road Bike Review site. There were numerous reports about dramatic changes in the feel of the bike in question when users switched from a metallic seatpost to a carbon variety. The reviewers claimed their very stiff frame was made significantly more comfortable by this simple component exchange.

[L7]

Found this on an unrelated search. I think Arty sums up much of it but to me the most glaring obvious ommission is the 80 lbs. As Arty states forces of hitting a pothole are non linear and dispersed through the frame. What he doesn't mention is that sometime during that impact the force is FAR greater than the weight of an 150 -200 lbs rider. Measurable force at somepoint may hit 5 or even 10 times that amount and this guy uses an 80 lbs. weight statically and then says 'there I told you'.

About all this proves to me is when you decide on the answer before you design the experiment you can usually come up with an experiment to confirm that answer. He should do a study on why people consider this sort of 'research' as valid or worthwhile.

[Arty50]

Originally posted by L7
What he doesn't mention is that sometime during that impact the force is FAR greater than the weight of an 150 -200 lbs rider. Measurable force at somepoint may hit 5 or even 10 times that amount and this guy uses an 80 lbs. weight statically and then says 'there I told you'.

Excellent point.

[Spats]

Originally posted by L7
What he doesn't mention is that sometime during that impact the force is FAR greater than the weight of an 150 -200 lbs rider.

That doesn't make the relative contribution of the frame any more significant. The relative magnitudes of the deflections won't change no matter how much force you apply (up to the point of material failure.)

Example: the flex measured at the seat tube for all the bikes was approximately 75/1000ths of an inch, and was 0 at the head tube for all bikes. The largest difference between two frame materials at any point was 1/1000th, or roughly 1/75th of the total difference.

So let's say you put a force of 800 pounds on the bike. Assuming a linear increase, the total deflection would be 75/100ths, or 3/4 of an inch. This is quite significant, and probably about what you'd get when hitting a really big pothole. But you're still only getting 1/75th of that deflection from the frame! The rest is coming from the fork and tires.

Arty: If the frame is getting into nonlinear deflection, it's permanently deforming or failing catastrophically (as you pointed out), and we can therefore ignore it. We're not addressing the crashing or breaking case, we're addressing the "road feel" case.

Oarhead: "When you do this you are putting all sorts of forces and moments on the bike." True but irrelevant. My point is limited to this assertion:

"The perception that different frame materials or different wheel constructions cause a bicycle to feel perceptibly 'rougher' or 'smoother' is not supported by the available data."

Here's what the data does support:
Data supports the assertion that different frame materials can feel different to the rider under hard acceleration (measurable flex) and under hard cornering on bumpy roads (measurable flex).

Data also supports the assertion that changing fork material and dimensions can feel different to the rider.

Data supports the assertion that tire inflation and construction can feel different to the rider.

Data supports the assertion that differences in seat construction can feel very different to the rider.

---

It's strange to see everyone get so worked up about this...like I've attacked everyone's religion. Not a *single person* has stood up and said "You know, maybe there's something to that. Maybe road bikes haven't fundamentally changed in 80 years, so we're reduced to inflating insignificant differences to the magnitude of holy war in order to differentiate our products." And I think it's telling that the arguments have all been either
a) anecdotal
or b) "well, I know this is wrong but I don't have the time or equipment to prove it."

At some point, someone's going to weld accelerometers to a seatpost and handlebars and capture a bunch of data. Until then, I'm chalking this up to the DVD effect I mentioned earlier: people are bad observers, and they really want to believe that the toy they just paid $2000 for is much better than their old one or the one their friend has.

[Spats]

This article should lay the topic to rest.

"Frame Materials for the Touring Cyclist"
http://sheldonbrown.com/frame-materials.html

Does anyone else have anything further to add?

[L7]

I'm not sure how worked up you think I got but I do tend to be annoyed when someone takes really bad data and presents it as valid hard science. I deal with Parks Canada on a daily basis for christ sakes I don't need more of that crap in my life.

With his tiny weight this guy is measuring comparitive deflection of different materials. There is NO REASON to ASSUME that this will remain constant under much bigger loads. He offers no proof that it will and yet extrapolates results assuming that this will remain constant. What's more the load is so light and measureable differences so small that even the margin of error could account for a different feel at real life load values that he has not come remotely close to.

Afterall a 100lb person's stiff bike is a 200 pounders wet noodle that may well be exacerbated with different materials. Maybe not, we certainly don't know with this 'research'.

I don't get worked up at people trying to change or redesign bicycles but having someone trying to foist bad science on me and others is a little irritating.

[Viva]

...and they really want to believe that the toy they just paid $2000 for is much better than their old one or the one their friend has.

Actually, no. At least not for everyone. In fact, at this very moment, I'm attempting to buy a 15 year old steel road frame. Why? It was built by one of the finest frame builders to ever come out of SoCal. Ride matters to me and who builds the bike is absolutely the most important consideration. I'm also going to look into to buying back my old Medici Pro Strada, one of the best bikes I have ever ridden.

[Arty50]

Originally posted by Spats
Example: the flex measured at the seat tube for all the bikes was approximately 75/1000ths of an inch, and was 0 at the head tube for all bikes. The largest difference between two frame materials at any point was 1/1000th, or roughly 1/75th of the total difference.

So let's say you put a force of 800 pounds on the bike. Assuming a linear increase, the total deflection would be 75/100ths, or 3/4 of an inch. This is quite significant, and probably about what you'd get when hitting a really big pothole. But you're still only getting 1/75th of that deflection from the frame! The rest is coming from the fork and tires.

Arty: If the frame is getting into nonlinear deflection, it's permanently deforming or failing catastrophically (as you pointed out), and we can therefore ignore it. We're not addressing the crashing or breaking case, we're addressing the "road feel" case.

What you've presented once again is static loading. That tells us nothing. As I pointed out before, when you're riding a bike the forces it experiences are very dynamic. Magnitude, duration, direction, cyclical loading, frequency, etc. All of these come into play and will generate the "feel" that everyone is talking about. Sure other components may play a part in this. I don't discount that. But the frame will still play a part. But basic engineering principles (some of which I've noted above) tell us that this test is poorly conceived and thus its results are largely meaningless.

For a good individual example, look at Galloping Gertie. A specific frequency range of wind would cause the bridge to oscillate wildly. Finally the bridge collapsed from the stress. Now I know we're not talking about failure nor a bridge. But Galloping Gertie is a classic example of the things people overlook when they are designing or analysing a system. The things we build live in a dynamic universe and thus we must consider the multitude of dynamic actions that can and will occur.

With this in mind, hanging weights from a non-moving bike doesn't really strike me as a very comprehensive test.

Look, I'm not exactly god's gift to engineering. I studied ME for my first 2 years in college. So I won't claim to have all the answers. But I do know enough to say that this test doesn't even begin to cover all that is occuring in a dynamic system such as a bike.

[Ogre]

This thread is starting to PISS ME OFF!!!!!


GGGGRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRR!!!!!!

[L7]

While I don't mind this thread dieing at all I must say I'd take a room full of nerds over one chickenshit trying to be a censor while hiding behind an alias.

When you think about it what's the worst kind of nerd? One that doesn't have the strength of conviction to identify himself with his own beliefs. Those are the chickenshit nerds that create viruses and terrorizes internet boards while remaining anonymous.

Ogre, your mama called she wants you to go back to Powder and play with the other nerds now.

If it's just a matter of comprehension I'm glad to help you with the tricky bits. The easiest thing to comprehend though, is IF YOU DON"T LIKE IT DON"T READ IT. Easy concept, you should try it then you don't have to be anonymous chickenshit boy.