Paul Lew on wheel dynamics

[mvrider]

Summary here:
http://road.cc/content/news/114442-paul-lew-talks-wheel-dynamics

Original articles here:
http://www.reynoldscycling.com/reynolds/news/Understanding-Wheel-Dynamics-Wheel-Stiffness
http://www.reynoldscycling.com/reynolds/news/Understanding-Wheel-Dynamics-Spoke-Crossing-&-Rider-Comfort
http://www.reynoldscycling.com/reynolds/news/Understanding-Wheel-Dynamics:-Carbon-Fiber-Variability

http://road.cc/sites/default/files/imagecache/galleria_600/images/Paul%20Lew/Paul%20Lew%201.jpg

[redir]

Interesting about the spoke patterns.

[Mark McM]

Interesting about the spoke patterns.

Yes, Lew writes some good stuff on the influence of spoke patterns. Of course, this has been known (and previously published) decades ago, but wheel designers have to keep "re-inventing" the wheel every few years, so they need to forget what's already known about lacing patterns.

My one qualm with the articles is with the part one of the series, which while correctly stating that the number and thickness of the spokes affects the vertical stiffness of the wheel, it implies that this has a meaningful affect on ride compliance. But part two of the series clears this up by pointing out that tire has a much, much larger compliance than even the most compliant spokes, rendering the affect of the spokes on compliance effectively meaningless.

[MadRocketSci]

As I've said before, when it comes to comfort, i want to see data on the dynamic behavior of the structure, ie, frequency domain stuff like power spectral density, mode types, etc, which are more meaningful to me than just static deflection.

[ergott]

I'm not sure, but I think there's a type-o in this paragraph.

"Unlike the hub, which is two-dimensional and can contribute favorably to both lateral and vertical performance, the spokes are one-dimensional. So, a stiff spoke (one that resists bending) decreases side-to-side deflection and at the same time decreases vertical deflection, whereas a flexible spoke (one that is easily bent) decreases side-to-side deflection and increases vertical deflection."

[Louis]

I'm not sure, but I think there's a type-o in this paragraph.

There are several statements in that quote that don't make sense. Is Paul a structural engineer? If so, he needs to learn to write/speak. If not, he should stay away from that sort of topic.

Edit:

I Googled him and found the quote below. He should have known better than to make the statement quoted above. (Unless the quote was garbled.)

BIKERUMOR: What’s your educational background?

LEW: In 1985, I got a bachelors of science in mechanical engineering from the US Naval Academy. In 1989 I finished a bachelors of Architecture. The formal education in fluid dynamics, which is a broader category than aerodynamics, pertains to air. For example, when I use the CFD program to design bicycle wheels, I have to choose the medium (air, water, oil, etc.). As far as architecture, prior to learning how to put a structure together, there’s a period where you focus solely on design. The first two years of the program are highly focused on design. We take art, sculpting and drawing classes. We design a lot of things other than buildings. Things like cases for small items, bicycles, medical equipment, etc. Then, that’s integrated into how bricks, steel and wood go together, stress loading and the nuts and bolts of it. It’s a nice way to complement my engineering education. It really helped me with design.

[beeatnik]

They're just wheels, yo.

[oldpotatoe]

There are several statements in that quote that don't make sense. Is Paul a structural engineer? If so, he needs to learn to write/speak. If not, he should stay away from that sort of topic.

Edit:

I Googled him and found the quote below. He should have known better than to make the statement quoted above. (Unless the quote was garbled.)

Tried to find out but..what did he do in the USN? Class of 1985, 'ring knocker'...

[Charles M]

Yes, Lew writes some good stuff on the influence of spoke patterns. Of course, this has been known (and previously published) decades ago, but wheel designers have to keep "re-inventing" the wheel every few years, so they need to forget what's already known about lacing patterns.

My one qualm with the articles is with the part one of the series, which while correctly stating that the number and thickness of the spokes affects the vertical stiffness of the wheel, it implies that this has a meaningful affect on ride compliance. But part two of the series clears this up by pointing out that tire has a much, much larger compliance than even the most compliant spokes, rendering the affect of the spokes on compliance effectively meaningless.

No... That's just not correct.


That's as inaccurate as the common theme that small aerodynamic gains don't matter because big aerodynamic gains are big... And or a few grams don't matter because rider weight is greater...


Grams weigh what they weigh... Grams of drag are what they are...

Relative to wheels or saddle rails or frames or anything else, deflection / vibe damping / compliance is what it is. Generally speaking, they add up to a given ride quality rather than eliminating the importance of each other.


All that said, as it stands, I won't ride a Lew related product after what he did with his company and how he handled it's closing. And there are too many reputable sources for wheel info and development for me to bother reading what he writes (or regurgitates wrapped as original thinking)

[oldpotatoe]

All that said, as it stands, I won't ride a Lew related product after what he did with his company and how he handled it's closing. And there are too many reputable sources for wheel info and development for me to bother reading what he writes (or regurgitates wrapped as original thinking)

I thought he sold to Reynolds, guess not....what did he do?

[wasfast]

He went to work at Reynolds. There was not "selling to" that I know of.

Lew Racing had outstanding, prepaid orders that were never filled. No refunds were issued for the most part either. There was another guy that was on the business side (can't remember his name) but the mess that was left from Lew Racing was exactly that, a mess.

Despite the years that have passed, like Charles, it's not easy to forget in a small industry like this.

[Charles M]

The primary guys at Lew were Paul Lew and Lee Vacaro.

The thought was that Paul sold "intellectual property" to Reynolds and also took a job with them...

It seemed like Reynolds and Paul Lew thought that it was the ethical thing to do to take the IP and leave behind all of the liability, bad wheels, warranty issues AND PREPAID PRE-ORDERS...

Some may also remember the litigation involving EDGE/ENVE near that time as well...

A huge amount of pressure was on Paul and Reynolds (for enabling his bull···· company dump) and some folks were eventually taken care of(at least partially). But it was only after a HUGE amount of bad press toward Reynolds and Lew...

I have no idea how Lee wound up. It's fairly plain that the Reynolds RZR products were/are Lew Racing products and it seemed like Lee was left holding the bag (though he may have partly deserved to be holding that bag as well).

[Mark McM]

No... That's just not correct.


That's as inaccurate as the common theme that small aerodynamic gains don't matter because big aerodynamic gains are big... And or a few grams don't matter because rider weight is greater...


Grams weigh what they weigh... Grams of drag are what they are...

Relative to wheels or saddle rails or frames or anything else, deflection / vibe damping / compliance is what it is. Generally speaking, they add up to a given ride quality rather than eliminating the importance of each other.

That may be true from a theoretical standpoint, but is the compliance/vibration/damping of a wheel large enough to matter from a practical standpoint? Given how little (vertical) deflection there is in a wheel, there should be little vibrational energy in the wheel, and little damping as well - especially as compared to the compliance/damping of the tire.

Zipp has done blind testing of wheels, to see if riders can really tell the difference between wheels. Here's what Josh Poertner, former Zipp Engineer, has written in a post on the SlowTwitch.com message board (http://forum.slowtwitch.com/forum/Slowtwitch_Forums_C1/Triathlon_Forum_F1/Road_bike_recommendation_from_Cervelo_RS_lovers_pl ease_P4944984/):

We ran blind wheel tests a couple of times a year at Zipp to benchmark competitive wheels and our own prototypes, and we also found that blinded riders were generally unable to tell the difference between stiffness and inertia, had no reliable feedback on weight, lateral stiffness, or comfort in general, and in the end were generally only able to pick out the aero wheels because they were riding laps around a closed park environment using power, so the more observant ones would notice speed differences. In the end, we sort of determined that when riders didn't know what they 'should' feel, they really struggled to find differences in stiffness, compliance and weight between frames or wheels.



There's more about the relative affects of stiff components on total ride compliance in an interview of Josh Poertner on the NY Velocity web page: http://nyvelocity.com/content/interviews/2014/josh-poertner

Josh's experiences shouldn't be surprising, given how very stiff and non-damping wheels are compared to other things like the tires (and therefore how little difference there is in relative stiffness and damping between wheels). If I had instruments sensitive enough, I could measure differences in bike/frame stiffnesses due to type of water bottle cages that were bolted on. But does anyone believe that a minor difference due to the water bottles has any meaninful affect?

[R2D2]

all that said, as it stands, i won't ride a lew related product after what he did with his company and how he handled it's closing. And there are too many reputable sources for wheel info and development for me to bother reading what he writes (or regurgitates wrapped as original thinking)

+1

[Charles M]

That may be true from a theoretical standpoint,

Zipp has done blind testing of wheels, to see if riders can really tell the difference between wheels. Here's what Josh Poertner...

But does anyone believe that a minor difference due to the water bottles has any meaninful affect?

It's not theory... It's data. And I know Josh... I spoke to him a couple days ago about the new stuff coming from Silca (and it's pretty great stuff).

I was also (an incredibly insignificant) part of some of that testing and had a chunk of the data from both the cobble testing and more standard road testing.

And you're the only person saying people could feel the vibe damping of water bottles...


What Josh will tell you (if you kept the conversation in context) is that the cumulative effect of improvements, be they aero, vibe damping and ,in his new job, tolerances (relative to pump performance durability) are important and do have an impact.

If you asked him if there is a difference in wheel feel from the most compliant thing he's tested and the least, he will absolutely tell you the difference is substantial. His comments were about wheels that were similarly purpose built (albeit from different materials) rather being about how substantially different wheels can perform.

Staying in context of this conversation... What you said was

"tire has a much, much larger compliance than even the most compliant spokes, rendering the affect of the spokes on compliance effectively meaningless."


One form of compliance doesn't render others meaningless. They all still have X effect and combine to have a total result.

That total resulting thing (be it aero, compliance, etc) can be substantial.


I get what you mean and you're right. There are far more important things.


What I'm saying is that the lesser things are not meaningless...

[thirdgenbird]

It's not theory... It's data. And I know Josh... I spoke to him a couple days ago about the new stuff coming from Silca (and it's pretty great stuff).

Anything on the frame or mini pump front? I'm about to buy one but may wait if there is some new stuff on the way.

[happycampyer]

Anything on the frame or mini pump front? I'm about to buy one but may wait if there is some new stuff on the way.Perhaps a vibration-damping frame pump?

[Charles M]

Honestly, a mass damper would work... You may not appreciate carrying it around though, but then that's exactly why some folks like the ride of relatively heavy bike frames :)




Anything on the frame or mini pump front? I'm about to buy one but may wait if there is some new stuff on the way.

BOTH!

Not sure what's postable... There's a new floor pump as well.

They're all made like products you'll have for decades. A couple of replaceable parts but the hardware is a whole other level from the plastic crap dominating the market right now.

[thirdgenbird]

Gah...

I want it now!!!

The specialized road pro is my favorite mini/frame pump going. Mostly aluminum, simple presta only head, and clean looks. A full size frame pump along the same lines would have me won over.

[MadRocketSci]

Deflections in a wheel might be small, but accelerations are not. And in my experience that's what influences my perception of comfort the most. This was illustrated to me when I was test riding a time vxr proteam down at colorado cyclist. They're located in an office park with some typically moderately crappy chip-seal type parking lot surfaces to test in. I really wanted to like that bike but i hated the ride...it was buzzy and nasty. Then i realized that the bike had the common denominator for every bike i've ever hated on a test ride - Ksyriums. After I got them to swap out those for some Easton EA90 wheels they had lying around, the ride improved a lot. Still not enough for me to buy the thing, but definitely noticeable.

So the deflection of the vibrational modes in the Ksyriums or Ksyrii are small...but since they have those stiff/light AL spokes, they have some nice high frequency modes. High frequency modes = high acceleration/low displacement, since acceleration in vibration scales with the square of frequency. That's definitely something my ass and hands can feel in a blind test. The tire is a nice filter and definitely improves the ride over bare rims though, i'll agree with that.

[Mark McM]

I was also (an incredibly insignificant) part of some of that testing and had a chunk of the data from both the cobble testing and more standard road testing.

Would you be able to share it? I'd be interested in seeing it.

And you're the only person saying people could feel the vibe damping of water bottles...

No, I didn't say that. Bolting a water bottle cage to the frame was just an example of another thing that that made only a very small (but real) change in compliance. Do you think that there are some compliance differences that are too small to be significant? If so, where do you draw the line on changes in compliance that become significant?



If you asked him if there is a difference in wheel feel from the most compliant thing he's tested and the least, he will absolutely tell you the difference is substantial. His comments were about wheels that were similarly purpose built (albeit from different materials) rather being about how substantially different wheels can perform.

Actually, Josh was talking about how undetectable differences in wheels are. In the next paragraph, he wrote:

During the 303 Roubaix development, we broke the entire bicycle system into component stiffnesses to try and determine the relative contributions of things to ride quality. We were shocked to find that the entire difference between the special, longer wheelbase, layup optimized, cobble specific frame, and the standard road frame (which everybody could tell you was WAY too stiff to ride on the cobbles) equated to a whopping 4psi of tire pressure. Similar result with wheels, turns out that the 32 spoke box section Amrosio rims the teams swore by had about 50% higher vertical stiffness than the 303…yet the riders swore up and down that the 303 was 'too stiff' until we blinded them, at which point the opinions became nearly random with a slight bias toward the carbon wheel as more comfortable, yet with the riders preferring the carbon wheel only willing to proclaim it because they are all certain that it was the aluminum one.

So, two wheels, completely different in design, materials and purpose, which riders "knew" had significantly different compliance, and yet they couldn't be distinguished from each other during blind tests.

And it appears that the teams have gotten the message, as indicated by this Velonews article about the nearly complete disappearance of shallow aluminum rims from the Paris-Roubiax race (http://velonews.competitor.com/2013/04/bikes-and-tech/the-torqued-wrench-ambrosio-nemesis-wheels-expire-in-the-pro-peloton_281545).


One form of compliance doesn't render others meaningless. They all still have X effect and combine to have a total result.

That total resulting thing (be it aero, compliance, etc) can be substantial.


I get what you mean and you're right. There are far more important things.


What I'm saying is that the lesser things are not meaningless...

I agree that everything adds up, and the resulting compliance/comfort is the net sum of all the parts. But it is also possible to be penny wise and pound foolish, and over-emphasize the less important parts. Personally, when I look at the factors relating to comfort/compliance on a bike, I look at the rider fit/position, the tires, the saddle, the handlebars and stem, the seatpost and even the fork (all of which have measurably more compliance) before I look at the wheels.

[Mark McM]

Deflections in a wheel might be small, but accelerations are not. And in my experience that's what influences my perception of comfort the most.

But if the deflections in the wheel are small, the wheels will pass the shock motions/accelerations directly to the frame with little attenuation, and so should have only a small affect on perception of comfort. And what little affect they have will be swamped by the deflections/attenuations in the other components, so the small differences in compliance due to the wheels may be barely noticeable (or more likely unnoticeable). As we've seen from Josh Poertner's tests, the differences in compliance between wheels might be affectively the same as a few psi difference in the tires. Can you notice that a tire is inflated to 98 psi instead of 100 psi? If so, would a "rigid" wheel with a 98 psi tire be just as comfortable as a "compliant" wheel with a 100 psi tire?


This was illustrated to me when I was test riding a time vxr proteam down at colorado cyclist. They're located in an office park with some typically moderately crappy chip-seal type parking lot surfaces to test in. I really wanted to like that bike but i hated the ride...it was buzzy and nasty. Then i realized that the bike had the common denominator for every bike i've ever hated on a test ride - Ksyriums. After I got them to swap out those for some Easton EA90 wheels they had lying around, the ride improved a lot.

So, professional racing teams can't tell different wheels apart in a blind test (as per Josh Poertner's post), but you can feel a significant difference between EA90 and Ksyrium wheels?

So the deflection of the vibrational modes in
the Ksyriums or Ksyrii are small...but since they have those stiff/light AL spokes, they have some nice high frequency modes. High frequency modes = high acceleration/low displacement, since acceleration in vibration scales with the square of frequency. That's definitely something my ass and hands can feel in a blind test. The tire is a nice filter and definitely improves the ride over bare rims though, i'll agree with that.

Try again. The aluminum spokes in the Ksyrium are actually about the same stiffness as steel spokes (they have 3 times the cross sectional area but 1/3 the modulus). Plus, neither steel nor aluminum has much vibration damping, so they'll pass vibration nearly equally well. Your theory doesn't work.

Have you really done a true double blind test? If not, maybe you're just experiencing what you expect to experience.

Personally, I can often tell the difference between different tires, and I can sometimes feel wheel lateral stiffness if a wheel is particularly flexy, but I can't tell the distinguish differences in wheel vertical stiffness. However, I can hear differences in wheels, which might lead to a perception of wheel smoothness/compliance, but it comes down to it, I can't actually feel any differences in vertical compliance/comfort between wheels.

[MadRocketSci]

I don't know why we keep talking past each other on this topic. You insist on bringing up deflection, which is just one tree in the forest. I keep talking about acceleration of that deflection...again, even though deflections are small, the 2nd derivative (ie change in the rate of change in the rate of) of that deflection is NOT small at high frequencies. Take a 2nd derivative of that deflection. You do agree that if you force the wheel it will ring, right???? So those deflections might be pretty small, but the change in the velocity of those displacements is large because it scales with frequency squared. That acceleration is felt as a general "harshness or buzz." You are focused on deflection but in a dynamic sense, with elastic structures vibrating, that is just one aspect out of many. I don't know how better to explain this. So please, if you respond to this, please stop talking just about static deflection. As you point out, these things have little damping apart from the tires/air so they have dynamic behavior and vibrational modes just like anything else made of metal.

I don't know any pro racers and have no idea what they can and can't feel. They might be very accustomed and insensitive to the stiffness and harshness of their super stiff and light bikes. I don't ride a 14lb massively oversized tube bike with super stiff wheels on a regular basis. I do know what my ass feels and I'm guessing others around here have felt a harsh bike despite the displacements of whatever component being small.

But to be clear, yes, i can feel the difference between a Ksyrium and EA90 wheelset.

I probably can't tell between 2 psi of wheel pressure. I doubt such a small change in wheel forcing function results in that big a difference in modal behavior.

Try again. Nice. Steel and Aluminum behave differently in vibration. You can match overall stiffness but the densities are different. That's why my Peg 830am doesn't ride exactly like a steel peg. You can try to match each tube's stiffness but it aint gonna ring the same. Go here:

http://ocw.mit.edu/courses/mechanical-engineering/2-002-mechanics-and-materials-ii-spring-2004/lecture-notes/lec2.pdf

vibrational modes depend on E, I, and density. The density is in the denominator. Less dense materials will thus generally vibrate at HIGHER frequencies. [Edit: the stuff in bold is wrong in this case, hasty math on my part...for the case we are discussing E/(density*cross_sectional_area) is roughly a constant so the difference in vibrational (bending) frequency is due to the Area moment of Inertia (I) term] Besides, if the spoke is 3 times thicker then you are not matching I either...which goes at about thickness cubed, an even greater effect, more than the linear effect from the difference in E between steel and AL. This is why you can make a bike out of Al with similar stiffness to steel but lighter weight.

This is more than just statically hanging loads onto wheels/bikes/structures and measuring deflections. This is higher level....

But, the way we've discussed this in the past, you aren't going to convince me and i'm not going to convince you. So, I can try to clarify specific points, but if not, have the last word.

[madcow]

Maybe I'm a little off point here, but to me it just seems like the debate is more about quantity. Sure small changes may be undetectable, but make enough small changes and at some point they do become something you notice.

Damon did an interesting study on JND, I'm sure it's probably been posted and discussed already, but for those that haven't seen it you can find it here: http://www.cervelo.com/en/engineering/ask-the-engineers/just-noticeable-difference.html

Lastly on the topic of why Lew is at Reynolds, the story I got is that his employment there was part of a settlement stemming from a suit over ownership of the Lew name. Not sure if it's true or not, but did come from a reliable source.

[MadRocketSci]

Ok, I realize now why there is a disconnect here between Mark and I. Mark, you are talking about deflections in the plane of the wheel and longitudinally in the spokes. I agree there that those types of deflections are mostly insignificant - the wheel is strong/rigid in that direction. What I am talking about is all the bending type modes in the wheel and spokes....which is where the wheel/spokes are less stiff. These modes will definitely exist, and be excited while riding along a chippy-sealy road. So, I'll agree to your points but i think in regards to comfort you should consider those bending type vibrations that i'm referring to.

Also, I think taking static measurements of wheels under constant load and then extrapolating to determine "comfort" is overly simplistic. Wheels in riding experience dynamic loads and periodic (vibrational) dynamic responses. Static testing might give you some stiffness data to play with but there's a lot more that goes into the bike/wheels dynamic behavior and perceived comfort.

Regarding that cervelo study, I must fall into the more sensitive group, though i never considered myself that sensitive :) I've said before on this forum that I have an unnatural hatred for Ksyrium wheels. They've screwed up plenty of test rides - A BMC Road Machine, a Sierra Nevada Ottrott, a Merlin Agilis and the Time VXR Proteam before I realized that I hated the wheels, not the frame.