Frame Material Engineering
09-28-09, 07:11 AM
#76
Senior Member
Join Date: Apr 2005
Location: Ann Arbor, MI
Posts: 16,880
Bikes: 1980 Masi, 1984 Mondonico, 1984 Trek 610, 1980 Woodrup Giro, 2005 Mondonico Futura Leggera ELOS, 1967 PX10E, 1971 Peugeot UO-8
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 1858 Post(s)
Liked 664 Times
in
506 Posts
Knowing that you are an engineer means we can discuss this somewhat differently that if you weren't.
Vitus frames, which I know only in passing, didn't break due to excessive 'flex'. If they broke, they broke due to excessive stress. Stress relates to load and the amount of material that is made to take that load. Strain is a measure of the deflection, flex in your terminology, produced by the load. Stress and strain are not interchangeable, so you muddy the water when use them as you have.
Easton and Columbus offer large diameter tubing to reduce deflection and that is why higher quality frames are less prone to breaking you say? With respect to what? If we are comparing to a steel frame, then the modulus of elasticity of aluminum is roughly 1/3 that of steel. To produce an aluminum frame with the same deflections for a certain load, then the respective moments of inertia need to be increased to compensate. Keep in mind, this does not mean that we have to use 3 times the material as a steel frame. Aluminum race frames are generally a bit lighter than similar quality steel race frames. Depending on how we've chosen our materials, our aluminum frame may very well operate at higher stress levels (if not in absolute terms, then certainly as a percentage of yield) in localized areas than a steel frame even though it's overall deflection (flex) is less. The deflection is related to the stresses, but as a byproduct. Flex or the lack thereof doesn't keep our frame in one piece or cause failures.
And now we get to the oft-revered endurance limit. I don't know what stress levels steel frames are built to, but from what I've seen, steel bikes are built to see stress levels _higher_ than the materials endurance limit. This is a bike, not a bridge. While an endurance limit is a quality that steel has, it's only of importance if a designer keeps stresses low. If a designer chooses tubing that is hefty enough to keep them that low, then the bike turns out too heavy. Heavy bikes, specifically heavy road bikes, don't sell well. So, designers keep the tubing thin, the stresses a bit higher (but below yield) and they get the opportunity to actually sell the bikes they are designing.
Now this is all statics. This is the easy stuff.
The fun part is when we think about how these big tubes act when we start pumping road vibrations through them. Big diameter, thin-walled AL tubes have high natural frequencies. When road input transmits the right vibrations into the frame, those tubes start to dance around and the infamous 'road buzz' rears it's head. Skinny, dense steel tubes have a lower natural frequency. They get excited at lower frequencies that aren't as objectionable to many riders. Steel doesn't 'damp' anything, it just resonates lower and people assume that the higher stuff is damped. No, it's just not transmitted as efficiently.
Figuring out the particularities of these issues sounds like a hell of a fun geek-fest to me. I'd really enjoy doing it. So far I haven't had any takers.
Vitus frames, which I know only in passing, didn't break due to excessive 'flex'. If they broke, they broke due to excessive stress. Stress relates to load and the amount of material that is made to take that load. Strain is a measure of the deflection, flex in your terminology, produced by the load. Stress and strain are not interchangeable, so you muddy the water when use them as you have.
Easton and Columbus offer large diameter tubing to reduce deflection and that is why higher quality frames are less prone to breaking you say? With respect to what? If we are comparing to a steel frame, then the modulus of elasticity of aluminum is roughly 1/3 that of steel. To produce an aluminum frame with the same deflections for a certain load, then the respective moments of inertia need to be increased to compensate. Keep in mind, this does not mean that we have to use 3 times the material as a steel frame. Aluminum race frames are generally a bit lighter than similar quality steel race frames. Depending on how we've chosen our materials, our aluminum frame may very well operate at higher stress levels (if not in absolute terms, then certainly as a percentage of yield) in localized areas than a steel frame even though it's overall deflection (flex) is less. The deflection is related to the stresses, but as a byproduct. Flex or the lack thereof doesn't keep our frame in one piece or cause failures.
And now we get to the oft-revered endurance limit. I don't know what stress levels steel frames are built to, but from what I've seen, steel bikes are built to see stress levels _higher_ than the materials endurance limit. This is a bike, not a bridge. While an endurance limit is a quality that steel has, it's only of importance if a designer keeps stresses low. If a designer chooses tubing that is hefty enough to keep them that low, then the bike turns out too heavy. Heavy bikes, specifically heavy road bikes, don't sell well. So, designers keep the tubing thin, the stresses a bit higher (but below yield) and they get the opportunity to actually sell the bikes they are designing.
Now this is all statics. This is the easy stuff.
The fun part is when we think about how these big tubes act when we start pumping road vibrations through them. Big diameter, thin-walled AL tubes have high natural frequencies. When road input transmits the right vibrations into the frame, those tubes start to dance around and the infamous 'road buzz' rears it's head. Skinny, dense steel tubes have a lower natural frequency. They get excited at lower frequencies that aren't as objectionable to many riders. Steel doesn't 'damp' anything, it just resonates lower and people assume that the higher stuff is damped. No, it's just not transmitted as efficiently.
Figuring out the particularities of these issues sounds like a hell of a fun geek-fest to me. I'd really enjoy doing it. So far I haven't had any takers.
09-28-09, 07:12 AM
#77
Senior Member
Join Date: May 2005
Posts: 1,485
Bikes: Trek, Giant, PoS
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Being on BF i have realized one thing, many like to voice their opinions, yet there is typically a lack of proof or data to back up claims.
The long held debate between the vast selection of frame materials and ride comfort could easily be "tested" yet i have not seen hard facts surrounding this.
To end this debate, all it requires is a simple Design of experiments, a few strategically placed high frequency accelerometers and some analysis. This would quantify your 'ride quality' and go a long way in putting logic behind the debate.
The other debate about component groups is equally lacking true engineering information, but that would require a much larger set of metrics... i'll leave that for another day.
The long held debate between the vast selection of frame materials and ride comfort could easily be "tested" yet i have not seen hard facts surrounding this.
To end this debate, all it requires is a simple Design of experiments, a few strategically placed high frequency accelerometers and some analysis. This would quantify your 'ride quality' and go a long way in putting logic behind the debate.
The other debate about component groups is equally lacking true engineering information, but that would require a much larger set of metrics... i'll leave that for another day.
09-28-09, 08:12 AM
#78
Certifiable Bike "Expert"
Join Date: May 2005
Posts: 5,647
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I think the OP is on the right track. Put the accelerometer on the seat. Measure data from a few hundred miles. Stick the same instrumented seat on a different frame, but same wheels, tires, pressure. Ride a few hundred miles on that one.
Tranform the data into the frequency domain, and look at the magnitudes and frequencies where most of the vibration is happening. It would be interesting.
I don't think it would sway the true believers. They can't be swayed. How many golden-ear audiophile subjectivists have been swayed by science?
Tranform the data into the frequency domain, and look at the magnitudes and frequencies where most of the vibration is happening. It would be interesting.
I don't think it would sway the true believers. They can't be swayed. How many golden-ear audiophile subjectivists have been swayed by science?
09-28-09, 09:53 AM
#79
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
09-28-09, 09:59 AM
#80
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
From all the quantitative and qualitative customer testing that I've participated in, it isn't so simple. People's perception of sensations isn't very accurate - for example, if they're holding objects of different weight in each hand, they need a 5-7% difference to perceive anything.
If you're trained to pick out different sounds and feelings, they are not all that tough to notice.
09-28-09, 10:16 AM
#82
Senior Member
Join Date: Feb 2009
Location: Long Island NY
Posts: 772
Bikes: Panasonic 500
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
+1 (Wish I had remembered that)
Plus, if you are going to define the bike's ride in numbers, you have to have something in numbers to compare it to. I would like to see some one attempt to define "ride comfort" with a math formula. If it could be done, then the mattress and recliner people are going to want you very badly.
Plus, if you are going to define the bike's ride in numbers, you have to have something in numbers to compare it to. I would like to see some one attempt to define "ride comfort" with a math formula. If it could be done, then the mattress and recliner people are going to want you very badly.
09-28-09, 10:49 AM
#83
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
To the deniers:
Do you believe or do you not believe that this exact process is being done every day by every major automotive manufacturer?
Boys, this ain't rocket science.
Do you believe or do you not believe that this exact process is being done every day by every major automotive manufacturer?
Boys, this ain't rocket science.
09-28-09, 11:48 AM
#84
Senior Member
Join Date: Feb 2009
Location: Long Island NY
Posts: 772
Bikes: Panasonic 500
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
NOT.
09-28-09, 12:08 PM
#85
Senior Member
Just as a person from the general public isn't plopped in the seat of a test car from a auto manufacturer, you wouldn't drop joe-blow on some bike to test. You'd use a rider who could accurately detect small changes. These people exist. I'm not saying I'm one of them, but there are guys, specifically ones who are trained this way, who can discern 5 psi in tire pressure and/or other subtle changes. He may or may not be a pro racer. That certainly wouldn't be a requirement.
If you're trained to pick out different sounds and feelings, they are not all that tough to notice.
If you're trained to pick out different sounds and feelings, they are not all that tough to notice.
PS my position has a lot more nuance than 'denyer'.
09-28-09, 12:14 PM
#86
Certifiable Bike "Expert"
Join Date: May 2005
Posts: 5,647
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
So you are saying that there are professional bicycle test riders... they possess specially calibrated Golden Bottoms for evaluating ride quality? I find this hard to believe.
09-28-09, 01:07 PM
#87
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Keep in mind, these people do not necessarily define what is 'good' or 'bad'. What they do is identify a certain quality that the designer is shooting for or trying to avoid. Their opinion (in the case of cars) relates back to the human component of a NVH (Noise, Vibration, Harshness) issue. It is not done strictly by a 'math equation'. Of course, even using this terminology displays a complete lack of understanding of what is happening.
Often 'good' targets are set by marketing or product development and defined by the competition. Let's say that the marketing department of XYZ bike company says that they like how the Specialized Roubaix rides, but they want it to look different and have other options. They don't want to simply recreate the Roubaix, they just want their bike to have that type of ride. Engineering then has to come up with various way to hit these design bogeys. The first part of hitting a goal is defining it, so they go get a Roubaix, ride it, instrument it and make a good definition of their target.
Let's comment on what exRunner said. Does the auto industry have it all figured out? No, because there is not any one answer. BMW is a company that, IMO, does a really good job of the ride, handling and NVH compromise. A lot of people agree. Having said that, they do it at a price. When Hyundai was building their new sports sedan you can bet your ass they had several BMW's that they kept right next to their test cars. I know this for sure, because I know one of the Hyundai development engineers. Did they completely recreate the BMW? No. Product strategy people said we want to hit targets A, B and C and we want to do it for 40% of what a 5 series costs. Most people seem to think they've done a fairly good job of it. You may or may not agree.
Is any of this getting through or are you guys being obtuse for the fun of it?
09-28-09, 01:42 PM
#88
Certifiable Bike "Expert"
Join Date: May 2005
Posts: 5,647
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
No, what you actually said is that professional bike test riders exist. Apparently, that's not what you meant.
09-28-09, 02:16 PM
#89
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Just as a person from the general public isn't plopped in the seat of a test car from a auto manufacturer, you wouldn't drop joe-blow on some bike to test. You'd use a rider who could accurately detect small changes. These people exist. I'm not saying I'm one of them, but there are guys, specifically ones who are trained this way, who can discern 5 psi in tire pressure and/or other subtle changes. He may or may not be a pro racer. That certainly wouldn't be a requirement.
If you're trained to pick out different sounds and feelings, they are not all that tough to notice.
If you're trained to pick out different sounds and feelings, they are not all that tough to notice.
I do not know if such a position actually exists in bike world, but I'm confident that people exist who could reliably detect the changes I've described. Honestly, I'd be surprised if there isn't at least a few bike manufacturers that have this position even if only as a part of a larger role.
09-29-09, 10:47 AM
#90
Senior Member
Join Date: Aug 2009
Location: San Diego
Posts: 418
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked 0 Times
in
0 Posts
09-29-09, 11:01 AM
#91
Senior Member
I am astonished that people like you continue to believe that common beliefs about frame materials and their intrinsic effect upon ride are the result of insufficient data.
09-29-09, 11:02 AM
#92
Senior Member
Join Date: Aug 2009
Location: San Diego
Posts: 418
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked 0 Times
in
0 Posts
well if you were just testing the frame material you would just need to see what material dampens vibrations better.
https://www.sheldonbrown.com/frame-materials.html
after reading this i have come to the conclusion that it comes down to geometry and not material, although some materials may help, it would not be substantial enough to change your mind. the geometry would more likely cause the "feel"
and yes, i am an arm chair engineer because i quit my Mechanical Engineering degree. But sheldon browns page makes alot of sense.
https://www.sheldonbrown.com/frame-materials.html
after reading this i have come to the conclusion that it comes down to geometry and not material, although some materials may help, it would not be substantial enough to change your mind. the geometry would more likely cause the "feel"
and yes, i am an arm chair engineer because i quit my Mechanical Engineering degree. But sheldon browns page makes alot of sense.
09-29-09, 11:15 AM
#93
Senior Member
I think the OP is on the right track. Put the accelerometer on the seat. Measure data from a few hundred miles. Stick the same instrumented seat on a different frame, but same wheels, tires, pressure. Ride a few hundred miles on that one.
Tranform the data into the frequency domain, and look at the magnitudes and frequencies where most of the vibration is happening. It would be interesting.
Tranform the data into the frequency domain, and look at the magnitudes and frequencies where most of the vibration is happening. It would be interesting.
09-29-09, 11:33 AM
#94
Certifiable Bike "Expert"
Join Date: May 2005
Posts: 5,647
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I'm a "ride quality" skeptic. So if I had data from a few hundred bikes ridden from a few hundred miles, and if I saw that bikes with material A tended to have a characteristic vibration spectrum, and bikes with material C tended to have a different, but distinct vibration spectrum, that would at least prove to me that there is a significant difference between A and C... something I suspect is not true.
It might be possible, with another study, to equate vibration data to comfort. The only way I could think of doing this would be some kind of blind frame test - disguised and instrumented frames. After the rider rides for a few hundred miles, he rates the ride quality of the frame without knowing its brand or material. Then, the scientist can attempt to correlate the data to the subjective ride quality...
If you're doing blind frame tests, you could skip the instrumentation, but then you'd miss out on an opportunity to learn something.
09-29-09, 12:36 PM
#95
Senior Member
Join Date: Jul 2009
Posts: 51
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Your accelerometers will just prefer whippy frames.
People always compare bikes with different wheels and tires, and claim a difference.
Give 100 men a sugar pill shaped like a blue diamond, ...30 of them will get a woody.
09-29-09, 12:46 PM
#96
Senior Member
The long held debate between the vast selection of frame materials and ride comfort could easily be "tested" yet i have not seen hard facts surrounding this.
To end this debate, all it requires is a simple Design of experiments, a few strategically placed high frequency accelerometers and some analysis. This would quantify your 'ride quality' and go a long way in putting logic behind the debate.
To end this debate, all it requires is a simple Design of experiments, a few strategically placed high frequency accelerometers and some analysis. This would quantify your 'ride quality' and go a long way in putting logic behind the debate.
09-29-09, 01:55 PM
#97
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
09-29-09, 01:58 PM
#98
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
09-29-09, 02:03 PM
#99
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Keep in mind, I've stated that measuring the exact influence of material alone would be a waste of time. Measuring the influence of ride quality based on frame design as a whole would be a relatively straight-forward project and certainly less complicated than the exact same job on a motorcycle or car. Having said that, motorcycles and cars undergo this type of testing regularly.
09-29-09, 02:16 PM
#100
Wheelsuck
Join Date: Jun 2007
Posts: 6,158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
well if you were just testing the frame material you would just need to see what material dampens vibrations better.
https://www.sheldonbrown.com/frame-materials.html
after reading this i have come to the conclusion that it comes down to geometry and not material, although some materials may help, it would not be substantial enough to change your mind. the geometry would more likely cause the "feel"
and yes, i am an arm chair engineer because i quit my Mechanical Engineering degree. But sheldon browns page makes alot of sense.
https://www.sheldonbrown.com/frame-materials.html
after reading this i have come to the conclusion that it comes down to geometry and not material, although some materials may help, it would not be substantial enough to change your mind. the geometry would more likely cause the "feel"
and yes, i am an arm chair engineer because i quit my Mechanical Engineering degree. But sheldon browns page makes alot of sense.
Geometry plays a huge role in ride quality. That's exactly the point. Things like seat and head tube angles are primary players. How far do we go with the geometric influence, though? I would argue that the geometry of each individual tube is important and the material that is chosen has a large influence on tubing geometry. How these components work together as a single cohesive piece is key to understanding what produces the best ride qualities. Taking measurements to quantify what is preferred is a logical progression of this thinking.