Velo News tests 4 aero bikes- Anyone get this yet?

Back by popular demand, the general all-things Road forum!

Moderator: robbosmans

naisan
Posts: 210
Joined: Thu Oct 22, 2009 9:28 pm

by naisan

After a few careful reads of the article, I liked it, and especially how there are nods to the subjective and objective evidence.

One question and one big criticism:

Question: The standard wheels chart vs 404 wheels chart lists a standard road frame with exactly the same figures. What did you mean by that? Did you mean that you left the standard wheels on it for both tests and those were *not* 404s on the bottom chart?

Criticism: The deflection test seems all but useless (unless you only ride your bike on a trainer!) as it's not even close to the stresses that are placed on a bike while on the road: Seems that VN should have clamped the bars (like I do in my hand) rather than clamping the forks.

The rebuttal probably goes like this: "Dude, if you involve the bars and stem, all *kinds* of things come into play, so we just measured the frame. . . I still say better to remove the stem, clamp the steerer on a fixture, and actually replicate the stress that an actual bike goes through, as layup on a carbon frame can easily resist one force in a very different way than a force going a different direction, as that's the basis for a comfortable frame that is stiff under power.

I suspect the authors know this, as they say that the stiffness tests didn't reflect how the bike felt. I guess that's all you need to hear to emphasize that poor testing is worse than no testing.

While I applaud VN for doing lab testing, some real thought should be put in place before a flawed approach is replicated over and over again on many bikes throughout the year.

If I'm totally off on this, please educate me, but that portion of the test seems backwards from what you should test to determine something about how a bike handles in a sharp corner, during a sudden out-of-the-saddle acceleration, or when hitting a bump (I think you're trying to test all 3 situations at one time with this test!).

As an aside, VNTech mentions protocol somewhere online for the tests? After a few minutes looking through the VN site, I can't find anything other than a high-level article by Ben about VN starting testing procedures. The magazine page links to another page with some pictures. . .

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



User avatar
djconnel
Posts: 7917
Joined: Wed Sep 26, 2007 1:57 pm
Location: San Francisco, CA
Contact:

by djconnel

For the wheels: my interpretation was the control bike was tested only with the Zipps. Since they could pick any bike they want as the "control", why not use the good stuff on the wheels? If anyone cares about aero they're going to be on good aero wheels anyway.

I think on the clamping: Newton's laws come into play. "For every force, there is an equal and opposite force". So if you clamp the fork and measure the handlebar deflection, but I clamp the handlebars and measure the fork deflection, we will get exactly the same answer (with opposite sign, obviously).

One change I'd make would be to standardize the handlebar setup. Using the same bars on all of the bikes would have been a fairer comparison, since nobody in their right mind chooses a frame based on the stock bar size, for example.

User avatar
Cateye
Posts: 16
Joined: Sun Jun 01, 2008 8:37 am

by Cateye

In the last chart showing drag by bike it show grams of drag normalized to 30mph. No problem as far as testing goes but I assume that not too many people buy a bike based on the drag at 30 mph because they spend very little time at that speed on any given ride. I would have liked to see the drag numbers at say 20-25 mph.

naisan
Posts: 210
Joined: Thu Oct 22, 2009 9:28 pm

by naisan

djconnel wrote:So if you clamp the fork and measure the handlebar deflection, but I clamp the handlebars and measure the fork deflection, we will get exactly the same answer (with opposite sign, obviously).


Are you certain about that? While the Newton's law is not under dispute, you can't simplistically apply it once to a complex system of points without being dreadfully wrong in most cases. In addition, carbon fiber is not a homogeneous material that responds to stresses in different places and directions in equal ways.

Thinking that newton's 3rd law will account for all of that seems problematic at best.

When undertaking finite element analysis, you have to make many assumptions about homogeneity of materials and response, as well as the geometry of each system. A simple example: if I apply a force to an isosceles triangle with 10cm legs, and apply a deformation to a similar triangle with 10m legs, there may be a slight difference in the deformation in the triangles right?

In addition, the force at the bars when riding is in a different direction than the force at the forks (i.e. your vector would be different) so if you drew out a force diagram even if we assumed that you could simply apply newton's law you will have to tell me how to best split up the opposite reactions as well, and that isn't account for by simply stating that if you clamp that it will show up there.

So by clamping the bar, you may find that the bb in fact swings more to the side, or that the seat tube/head tube junction does, or that the engineer who spent some time and attention thinking about how the rider might be pulling up on the bar, while pushing down on a pedal, while leaning the bike over a bit decided to place a few strategic plies here and there to combat one kind of torsion while purposely allowing other force vectors to deform the frame to ensure that the bike is comfortable.

Measuring that all at the same time just makes a muddle, as the testers found out: "Interestingly, the torsional test didn't correlate much with our overall opinion of a bike" (-p72 of the VN issue) I interpret this comment in context to mean that they knew the Ridley was stiff, as the tests point out, but stiffer didn't mean better to them.

I have a different take: while I agree that stiff does not equal better (it does not, and I love riding my old steel bikes, and I'm greatly doubtful that stiffness means more efficient either, until somebody can show me a study that demonstrates this, but that's a different discussion), doing a silly test of frame stiffness won't help understand the dynamics of the bike much, so why bother doing it at all?

Much better off using a slightly different design that mimics *one* scenario at a time, and perhaps being able to infer something from that.

I know these tests are never going to get as good as real life, but at least if we're going to be reductionist and test a certain factor, let's do that in a way that allows us all (and velonews most of all, if they want to make money) to actually be relevant.

IMHO it would be vastly superior to simply clamp the frame down sideways on a reference table, place weights in turn at each interesting place (BB, stem attachment, fork dropouts), and measure deflection than muddling around with pushing on one point and thinking that will tell us anything much worth knowing.

Somebody clever should actually think about the angles at which these forces are applied while riding and maybe tilt the table to account. And perhaps somebody more clever will make a bike riding dummy that can measure vibration and then we'll act (and spend) like the car industry ;-)

User avatar
giant man
Posts: 967
Joined: Fri Jan 14, 2005 3:39 pm
Location: Essex / Lincs UK
Contact:

by giant man

Good test just a shame the Canyon Aeroad wasn't in there.

User avatar
djconnel
Posts: 7917
Joined: Wed Sep 26, 2007 1:57 pm
Location: San Francisco, CA
Contact:

by djconnel

It got clobbered in the Tour test.

I'd have rather seen the LiteSpeed. But then if is water bottle optimized...

User avatar
funhog1
Posts: 229
Joined: Tue Jul 15, 2008 11:59 am

by funhog1

Totally. I think, given the cables, it is a better test than the Tour test.

I want to see the Litespeed tested as well as the big brands' *non* aero bikes with internal vs exposed cables... The Litespeed alone would be great for that as they already have that frame with both internal and exposed routing.

As far as clamping goes, I agree that it is a best guess. I think VN's clamping is reasonable, and I personally like it. It's not like they or anyone has the time to figure out the best way to clamp thousands of frames, of different sizes just to figure out a clamping metric before they test. I think the above points are valid. Yet ultimately 6 of 1 or 1/2 of a dozen. :noidea:

Still. Would be nice for VN to give us clarity on the wheel situation with the control bike...

404's or no 404's?

There's no difference in the drag data and no mention. :hmm: tsk tsk

Would the test would turn into a long term advertisement for a 404/aero wheelset if they did?

I don't think so.

Any body know the A2 WT guys in NC? I'm sure they'd know.
Last edited by funhog1 on Thu Mar 10, 2011 2:45 am, edited 5 times in total.
Buy it. & Ride it.
Only if it has a high margin of utility.

tanhalt
Posts: 410
Joined: Wed Dec 27, 2006 6:36 pm

by tanhalt

Cateye wrote:In the last chart showing drag by bike it show grams of drag normalized to 30mph. No problem as far as testing goes but I assume that not too many people buy a bike based on the drag at 30 mph because they spend very little time at that speed on any given ride. I would have liked to see the drag numbers at say 20-25 mph.


30 mph is "traditional" for bike wind tunnel testing as a means to increase the "signal" of the drag values. The applicability of those drag values to lower velocities occurs because they're still in the same Reynold's number regime. The drag merely scales to the ratio of the square of the velocities...in other words, if you want to know what the drag would be at 25 mph, just multiply the 30 mph value by (25^2/30^2) = .694

This is part of the reason why I think all drag values should be reported in terms of CdA. It takes wind velocity and air density out of the results. Besides, "grams" isn't even a force value :roll:

rruff
Shop Owner
Posts: 2196
Joined: Sun Jan 08, 2006 9:12 am
Location: Alto, NM

by rruff

tanhalt wrote:This is part of the reason why I think all drag values should be reported in terms of CdA. It takes wind velocity and air density out of the results. Besides, "grams" isn't even a force value :roll:


Damn straight!

VNTech
Posts: 195
Joined: Tue Jan 25, 2011 8:08 pm

by VNTech

Hey guys,
Thanks for all the responses, I'm glad you liked the review (for the most part ;)).

A few quick follow-ups:
re: the stiffness test - we found that the test correlated quite well to perceived stiffness, and thus was very successful. However, stiffer did not always equal "better". That's the point we were making in the Results section. We're confident that the test is a good representation of frame stiffness, as perceived on the road.

re: CdA versus grams - to put it simply, more readers can relate to grams than to CdA.

re: control wheels - the control bike was set up with standard Fulcrum wheels. It's a Masi Evoluzione, with very similar tube shapes (at least to the naked eye) to a Trek Madone.

As always, I invite more specific questions via email (cfretz@competitorgroup.com) or PM.

Cheers
VeloNews Magazine/VeloNews.com tech

User avatar
djconnel
Posts: 7917
Joined: Wed Sep 26, 2007 1:57 pm
Location: San Francisco, CA
Contact:

by djconnel

Actually, I related to Newtons a lot more than grams....

tanhalt
Posts: 410
Joined: Wed Dec 27, 2006 6:36 pm

by tanhalt

VNTech wrote:re: CdA versus grams - to put it simply, more readers can relate to grams than to CdA.

Only because of cycling's obsession with mass uber alles! (Oh wait, what forum is this? Oops :oops: )

Seriously though, most readers don't relate to grams of drag either. If I had a dollar for every time I answered the question "How much time does X grams of drag equate to?", I'd be...well, not rich...but, I'd have a LOT more dollars!

Give the results in CdA (with only a brief explanation), and then "translate" that into time savings per km at "typical race speeds" (and standard atmosphere) for the masses (no pun intended). Skip the grams step...it really doesn't mean much of anything to anyone...

VNTech wrote:re: control wheels - the control bike was set up with standard Fulcrum wheels. It's a Masi Evoluzione, with very similar tube shapes (at least to the naked eye) to a Trek Madone.


Thanks! Any reason why this wasn't mentioned in the article?

User avatar
Cateye
Posts: 16
Joined: Sun Jun 01, 2008 8:37 am

by Cateye

tanhalt wrote:
Cateye wrote:In the last chart showing drag by bike it show grams of drag normalized to 30mph. No problem as far as testing goes but I assume that not too many people buy a bike based on the drag at 30 mph because they spend very little time at that speed on any given ride. I would have liked to see the drag numbers at say 20-25 mph.


30 mph is "traditional" for bike wind tunnel testing as a means to increase the "signal" of the drag values. The applicability of those drag values to lower velocities occurs because they're still in the same Reynold's number regime. The drag merely scales to the ratio of the square of the velocities...in other words, if you want to know what the drag would be at 25 mph, just multiply the 30 mph value by (25^2/30^2) = .694

This is part of the reason why I think all drag values should be reported in terms of CdA. It takes wind velocity and air density out of the results. Besides, "grams" isn't even a force value :roll:


Cool....thanks for the explanation.

User avatar
djconnel
Posts: 7917
Joined: Wed Sep 26, 2007 1:57 pm
Location: San Francisco, CA
Contact:

by djconnel

P.S. It was not at all clear that the control bike in the "with Zipp 440 wheels" plot didn't have Zipp 440 wheels. Obvious, the point is to compare frames and not wheels, so to account for the difference in wheels, readers should subtract the improvement of the Ridley with the Zipps vs the Fulcrum from the control bike results in that plot. (I know that's a crude approximation, but better than nothing).

But looking @ the Masi vs Ridley in the "stock wheel" plot, there will still be a considerable advantage to the "aero" bikes.

Masi:
Image

Dan

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



User avatar
funhog1
Posts: 229
Joined: Tue Jul 15, 2008 11:59 am

by funhog1

VNTech wrote:Hey guys,
Thanks for all the responses, I'm glad you liked the review (for the most part ;)).



That's what I love about the the bike world. For anything to be considered as being any good it has to be cursed and spat upon....by the collective......before it may be fully assimilated. :D

I think that's what I liked the most about your guys's test. I'm perceiving you as getting that, as you did go completely subjective with the Ridley with Big Ben D. Which made it fun.

As scientific as *everyone* likes to pretend to be, I'm seeing everyone embrace bike-dork-geek-tests out of a sense of fun.... albeit.... a weird one at times.

Which is why I don't see the American Drama Queen Dope Opera....as being so much about doping (science), rather a man-princess-y girl fight over who gets to WIN

Ok rant over. Back to OT/OP. :wink:
Buy it. & Ride it.
Only if it has a high margin of utility.

Post Reply