OldTryGuy

You supply it, I'll try it.

practical

I had a Cannondale CAAD8 (aluminum) that weighed about 22 pounds. I also had an old Nishiki Seral touring bike that weighed about 32 pounds. The Cannondale was very sweet going up hills compared to the Nishiki but going down or on flats, it was no faster. My point is that weight matters most in climbs and if you're doing gaps, then CF would matter a lot. If you're not climbing mountains or ride in very hilly terrain, it probably won't much matter. Go with the bike that feels the most responsive to each pedal stroke.

rpenmanparker 

When I said above that I wished it were that easy, I meant it. Here is the very complex type experiment one would have to conduct to get any meaningful results:

First you would have to pick four bikes, steel, Al, Ti, and CF. Ideally they would be of representative stiffness for that particular frame material, or better yet, representative stiffness/comfort balance. That is no easy task, but has to be done. Personally I would choose Specialized's top Tarmac for the CF, but that is just my prejudice. Every bike would have to be outfitted exactly the same, all same parts brands and models, same wheels, same tires, same inflation, same bottle cages, etc. The only differences allowed could be the steerer length, stem length, and number of spacers to make the drop and reach identical among the bikes. Any bike that would not allow that would have to be discarded in favor of a brand and model that could be made to have identical fit to all the other bikes. Then I would handicap the bike weights like in horseracing, make sure every single bike was precisely the same weight by adding the necessary weight to the lighter ones.

Then you would need to assemble a minimum of 5 skilled riders. If they were not all exactly the same size (yeah, right!), you would need one of each type bike for each of them. Whew this is getting a bit sticky.

A course would have to be laid out that would be long enough to give a statistically valid test, but not so long that all four types could not be ridden once each riders within about a two hour period, say 9 AM to 1 AM to minimize weather differences like wind. The course would have to have straights, curves, flats, and hills. Multiple laps of a smaller course would be allowed. No traffic complications or traffic control signals could be tolerated so the course might have to closed.

The tests would continue for five weeks, Monday through Friday. One of each type bike would be tested each day with a different rider. No rider would ride more than once on any day. The rider, ride order, and bike assignments would be randomly determined. The goal would be to have every rider ride every bike type five times. The riders would ride totally alone as in a time trial. In the case of inclement weather, make up tests would be conducted in the afternoons and/or on the weekends.

No on-bike computers or other speed measuring devices would be allowed. Riders would not be informed of their progress and performance during the tests.

At the end of the five week period, the results would be processed as follows: the course times for each rider on each bike would be separately averaged for the five experimental rides. The averages for the five rides on each bike type for each rider would be averaged. Then the results for all four bikes would be separately averaged for each rider. The difference between the time for a bike and the average time for all four bikes would be recorded as a percentage of the average of the four bikes. The percentage deviation from the means for all five riders would be averaged for each bike. Comparing those percentage deviations from the mean course time would determine the order of speed of the four bike types.

This would be a pretty good experiment, but there would still be significant potential sources of error like wind differences, lack of consistent rider output, etc. Oh well, no experiment is perfect, and you have to start somewhere.

See what I mean when I say I wish it were as easy as your friends telling you you're riding faster on your new bike!

wphamilton

You're testing the wrong thing! OP asked about bikes, whether someone with an aluminum bike could expect to go faster if he wen't to a CF bike. You're trying to answer a different question, such as "would I go faster if I bought a CF frame with the exact same weight and geometry as my current bike?" A valid question, but the wrong one.

"There's a lot of people who have steel or aluminum bikes that would like to know if they would be faster on a good cf bike"

Lazyass

Umm. Nevermind, I'll just keep my mouth shut on this one

Voodoo76

I spent many years until recently on a very stiff, fast Davidson steel frame. In May I started riding a Ridley Excal. I can't give a specific number as to how much faster it is, however over the long haul there is one significant advantage that I don't often hear discussed.

That advantage is ride quality. With apologies to the "steel is real" crowd there is no comparison between the ride of a truly stiff steel bike and a modern CF frame. There are rough sections of road on a couple of routes I frequent (bad chip seal) that had me hanging on for my life or slowing down on my steel ride while the CF bike blows right through them. Old CF (I raced on OCLV Treks in the 90's) had that same soft ride but were also soft in the BB. The difference I feel in newer CF frames is that the old trade off between a soft ride and BB stiffness no longer applies.

Bacciagalupe

I haven't noticed a big speed difference moving from an aluminum race-style bike (Allez) to a CF endurance-style bike (Roubaix).

I can't think of any reason why there would be a big difference based on frame material. Even the weight difference might be too small for Joe Average to accurately quantify, unless they live in a very hilly area.

I used the same wheels, tires and pressure on both bikes. The Roubaix has a substantially better ride feel in many conditions (e.g. chipseal, rough roads). It's a bigger difference than changing from 23c to 25c tires.

However, it isn't just the frame material that's different, it's also the frame design. In my test rides, the difference between the Secteur and Roubaix were small, whereas the difference between the Allez and the Secteur/Roubaix was noticeable right away.

rpenmanparker 

No, I don't think you are right. You have to answer the CF question in isolation first. What is the point of buying a CF bike to go faster if you don't know what about it makes it faster? You might buy a CF bike that isn't faster at all. The first question to answer is whether CF and its current level of stiffness makes a bike faster EVERYTHING ELSE AS EQUAL AS POSSIBLE. There is no upside to wandering around in the middle of 15 or 20 variables claiming you magically know what is important and what isn't.

How am I testing the wrong thing? I am proposing to test bikes, not specimens of the materials in a laboratory. It is necessary to first know whether CF is doing something, then you can move on to the other variables. I did this stuff for about 40 years and I tell you, garbage in, garbage out. If you don't properly design your experiment, your results are worthless. Sure there are multivariate experimental methods, usually known as statistical design of experiments, but there is just not enough background in hand on this topic yet to set one of them up.

mercator

My steel bike weighs 23 pounds. It's fast and comfortable. I've ridden it in hammerfests and double centuries and it has been great.
My carbon bike weighs 16.5 pounds. It's also fast and comfortable. You shouldn't be surprised to learn that I am faster going uphill on it, but there is not much difference otherwise.

I think the whole stiffness question is important when sprinting or climbing, but for most of the riding I do it doesn't make that much difference. I regularly ride an 80km loop with about 1000m of climbing on it. Looking at my garmin logs, I'd be hard pressed to tell which bike I rode over the past couple of years. The greatest variation appears to be caused by the wind, followed by my level of training.

wphamilton

It's the going, not the knowing.

You said it: you want to test one variable, the material. That won't tell OP whether he'll be faster if he replaces his old aluminum bike with a nice carbon fiber bike. In fact, as we both probably suspect, it would give him the wrong answer. He will be faster on the nice carbon bike, but the carbon itself with literally everything else equal will not make him faster.

Homebrew01

That question is too vague to be answerable. Basically, which ever bike is lighter will be "faster". How much faster ? Depends on the weight difference and the amount of climbing.

OP's Question:
rpenmanparker gives a pretty good overview of how you would have to test to answer the OP's question. You make all the variables the same, except 1, test, change 1 other variable, test etc ..... scientific method.

rpenmanparker 

So tell me WHY will be be faster on the CF bike? Could he get this same effect with a different material if it isn't the material?

rpenmanparker 

You have no idea just how much I appreciate your comment. Thanks.

Whizzer283

I had the same experience, though it was a 17 lbs CAAD 10 and a 25 lbs Nishiki. Not much difference in the flats and descents but climbs were much harder on the Nishiki.

wphamilton

Geometry, positioning, components, wheels, placebo.

e_guevara

I agree with Robert's approach on isolating the frame material in determining which is "faster".

Comparing "bikes" ridden over different riding scenarios is useless. I could have a CF bike with a climbing setup (50/34 cranks, 12-32 cassette) versus an aluminum bike spec'd with 53/39 cranks and 11-23 cassette. Given the same rider on a flat course, the CF bike is "undergeared".

How about weight? I have a mate with an old 90's Kestrel CF bike which was ~9 kg (20 lbs). Compare that with a lightweight steel bike (you can make a steel bike with modern components - not DT shifters) that weighs around 7.5 kg (16.5 lbs). The steel bike would feel "faster".

I understand that a new bike will make you want to ride harder (i.e. faster). That is true, whether you are going from steel to aluminum, aluminum to titanium, or titanium to CF. The effect is psychological (placebo) and doesn't tell anything about the real-world advantages of the material.

e_guevara

This can be achieved on any bike, be it steel, aluminum, titanium, or CF.

Carbonfiberboy 

Yes, except that it's impossible to do that, and even if it were, no one's going to do it, which is the reason we have threads like this.

My take: some bikes are faster than others. Maybe the material matters, maybe it doesn't: depends on the user and the use. What you want is the bike that goes for you, that when you pedal hard, you just keep going up through the gears seemingly without limit. That you can push into a corner and it just goes around. The doesn't beat the crap out of you if that is even of any concern. The stronger you are, the more it's going to matter and the bigger the differences you'll notice. I'm saying there are huge differences between bikes, whatever the material. The fact that you have selected bikes of different materials and they all ride about the same is about you and not the material.

All that said, CF is more tunable than any other material. The way it works is that a design is commissioned by a marketing department to sell well in a particular market segment, not to be the "best" bike one can build, whatever that might mean, which is nothing. So it's designed and engineered as well as possible, then they build a frame and a pro frame tester takes it out and rides the crap out of it, then comes back and makes suggestions about how it could be tuned to better suit the target market segment. Repeat. So that's how it might be possible to buy a CF bike that is faster than your stock steel rig. Or not. You have to do substantial test rides to see. The parking lot will tell you nothing.

coasting 

i am definitely faster on my carbon bike than my steel bike. some of you feel the need to say it must be other factors other than material. But by the same logic it could be material and not other factors or a bit of everything. why all the carbon scepticism?

Homebrew01

Empirical v conjectural

wphamilton

Buy these are generally not, particularly on stock bikes. Additionally, shape and certain structural characteristics of carbon fiber frames are difficult to achieve with aluminum and steel. Keep in mind also that for some reason, top end aluminum bikes just aren't spec'd as well as carbon fiber bikes.

He didn't ask if he could upgrade all of his components, buy better wheels and tires, and commission a hydroformed aluminum frame, would that make him faster. He asked if someone having an aluminum framed bike, not cognizant of all of the details that make a bike and rider faster, could expect more speed if he bought a "nice carbon bike".

rpenmanparker's suggested testing method would work, is valid scientific method in my opinion, but wouldn't answer the question. If you really want to know the answer, take a number of representative bikes, put power meters on them, and run the tests. Or else ask the cyclists who have experience with upgrading from an aluminum bike, as OP did.

Homebrew01

Ummm, Doing rigorous, controlled tests will not answer the question, but asking cyclists opinions will answer the question ?!?! hmmmmmm

I hypothesize, without testing, that frame weight, not material, will make the most difference in speed, especially if significant climbing is involved. There may be factors other than weighy that make 1 material "faster" than another. But without testing of some kind, we can banter back & forth all day with really answering the question.

wphamilton

When it's testing the wrong question, yes ...

And don't you think it's silly to completely ignore first hand experience?



Pardon me for saying, but I think it's obvious that the material itself, when all other factors (including the surface finish, stiffness in each direction, shape and area, etc) except weight are made identical, would make no difference in speed at all.

tonski 

Honestly they'd all fall from a helicopter at the same rate, buy what you like and ride it as hard or long as you'd like. Anecdotal evidence in this thread should be enough to tell you there is no difference. The guy who said find a new LBS if yours won't let you ride CF for a distance is right, so are the people saying find better tires. Tubulars > Tubeless > Clinchers and nearly any decent tire > Continental bone jarring tires. The Evo Corsa SC is far and away the most comfortable tire I've ever ridden, Tufo S33 is what I normally ride and I ride them until there are holes worn in and the tube is poking through (and usually pops and I strand myself somewhere riding a flat home because I don't carry a spare tubular and the goop I carry won't fix the tube)

Homebrew01

Wrong question ? Please explain.

I don't trust first hand experience in trying to determine a difference in speed from 1 frame to another. One may "feel faster" than another, but that doesn't make it faster.