Whenever such a discussion becomes sufficiently widespread, a mathematician seems to appear in the thread and set everyone straight, quantifying the exact impact of weight, drag, friction etc. on speed. Then the veteran cyclists appear with their usual: It's not the bike that determines speed, it's the engine. And then you wonder why people spend thousands of dollars on a bike if none of it matters. Why not just go buy a $20 Huffy at a garage sale?
All the advice out there is enough to make your head spin, instead of your wheels! So how do bike weight, tire size and aerodynamics (just to name a few) really impact speed (if they do at all)?
Because I, like most self-respecting cyclists, have a stable of different bike types, I decided to put it to the test and get some real numbers for myself.
And, like any good statistician, let me lay out my assumptions first and foremost.
ASSUMPTIONS
1) I am taking it for granted that extremely knobby tires are slower than slick tires. Not too many people would argue that. If you want to argue it, just put an a pair of DH heavy duty mud tires and go on a double century ride and let me know how it goes (see you in 9 hours--even if you are pretty fit! Haha!). So, not a single bike I used in the experiment has any significant amount of tread on the tires.2) Shocks will slow you down! Once again, if you are thinking that putting shocks on a fully rigid MTB might help you speed up, I am sending you back on that double century again (let me know how it goes). Jumping on a trampoline while trying to move forward is not a good way to conserve energy that can be dedicated to forward momentum. So, shocks were not used in this experiment either. One bike (a 29er XC bike) has a front shock with full lockout, which was employed for all tests.
Now you may say, "I bet you my full suspension MTB with knobbies will be faster screaming down a mountainside in Virgin, UT compared to your road bike with 23mm road tires!" Well, you are probably right. Which brings me to the third assumption:
3) We are talking about road and paved trail speed here. No XC, CX or DH. No mountainsides. Just roads, and varying-degrees-of -smooth bike paths.
And, technically this is pseudo-science! The bikes were not chosen at random and I didn't really do anything to control how hard I was riding, or to record how much headwind there was on a given day. I basically picked a bike and tried to just ride how I usually would on that day without giving it too much extra thought. The one thing that does sort of account for the possibility that I was subconsciously favoring a certain bike is that I did control for wattage estimates on each ride. Nothing special, just the ones provided by Strava that may or may not be very accurate...
All distances, speeds and wattages were recording using a Strava app and an Android phone. Strava uses "moving time". So, if you slow down to 0, it discounts that from your total time. However, I never used the "pause" function on any ride. This means that the information on speed is somewhat "deflated"by time spent slowing down to stop at traffic lights or fix a mechanical problem or whatever.
So there are the assumptions. All laid out. And now, on to the rest of the experiment.
First, let's meet the bikes--
- 2011 Specialized Allez Sport road bike--mostly stock but SPD pedals and a 23mm Continental Grand Prix 4000s II.
- 2013 Specialized Carve Comp--swapped out pedals for SPDs and put Nashbar 35mm "road tires" that are actually more like CX tires...
- The Trek named "Conversion"--a 1991 Trek 820 frame and fork with a LOT of mods: 1 x 8 setup with a 42t up front and 11-32 in back, SPD pedals, Geax Evolution II tires, SRAM x4 shifter and RD, and some "aero bars" made out of bar ends installed toward the center of the bar. {picture coming}
- A 1 x 7 orphan bike named "Scratch". In just over an hour, it emerged from scratch like the monster of Frankenstein from my spare parts bin and a cheap aluminum Schwinn road bike frame one evening when I was needing to let off a lot of stress...{picture coming}
The advantage here is that we have a little of everything...Again, not highly experimentally controlled, but just happens to be a little of everything, but especially in terms of aerodynamic setup, tire width, and weight. Look at the bikes one-by-one
AERODYNAMIC OPTIONS
- Drop bars, pretty narrow profile, skinny 23 mm tires
- Very upright position, no drops, no aero option on the handlebars--just a very wide flat bar
- The improved aero bars actually do a pretty good job. The bike profile is not especially narrow, but made back when tubes were chromo steel and pretty thin. The Trek named Conversion boasts a wide tire profile though at 50mm...
- Scratch also has drop bars and a pretty narrow road bike profile. The tire is a little wider at 35mm which might create a little more drag...
TIRE WIDTH
- 700 x 23. About as narrow as it gets unless you go to 20s or something...
- For this experiment, the Carve is wearing 35 mm semi-slick CX tires.
- 50mm Geax Evolution IIs--the widest tires in this experiment
- 35mm Origin 8 CX tires--semi-slick (something like a Herringbone pattern in the center)
While there is really no debate about whether knobbies or slick roll faster on roads (it's slicks by the way) there is a lot of debate about what kind of slick tires roll the fastest. Some say that wider slicks decrease resistance, but there is greater drag as well that comes with the wider profile.
There is also an issue of weight with the tires. It is common to hear that a heavier tire has a bigger impact on the overall weight of the bike than if that same weight were in, say, the frame instead. The argument is that rotational mass has a bigger penalty than other weight. However, the counter argument is that heavier tires may conserve momentum more than lighter tires. So you work harder to get them going, but they stay rolling easier. Additionally, wider tires seem to handle better on rougher terrain because they flex more over uneven surfaces and reduce wasted energy in the form of bouncing across terrain (instead of rolling smoothly).
So what is the deal? Ultimately, which tire is faster? Well, in practice, we need only look at the Tour de France to realize that NO ONE is climbing the Alpe d'Huez on 50mm semi slicks. So, in practice, what the best riders are using would suggest that the competition is not really between 20mm and 50s. The battle for speed is happening somewhere between 20s and 28s. However, in this experiment we have quite a range: 23mm up to 50mm. With all of the theory behind why wider tires should roll faster, why not have a 50 in the competition just to make a nice stark comparison?
WEIGHT
Over and over I hear veteran cyclists tell people: weight doesn't matter than much. Considering that you can spend THOUSANDS of dollars shaving grams, they have a point. But, again, I have never seen anyone rolling across the finish line in 1st place at the the Tour de France on a 30 pound steel bike.
The issue, though, is that in the Tour, seconds matter. So they spend thousands on their bikes just to shave off seconds.
Additionally, steel has seen something of a resurgence lately. Maybe the roads you cruise down are as smooth as an indoor track, but, living in a very rainy city with NASTY roads, I can really appreciate the way steel soaks up vibrations.
 |
| This reminds me of Michigan potholes.... |
Here are how the 4 bikes "size up" in terms of weight:
- 22 lbs. The lightest of the four....
- 23.5lbs. Only an extra pound and a half and $1000 less than the MSRP on the Allez! BLAST!
- 27lbs. Not too bad for a 29er MTB...
- 28.6lbs of sheer steel--OK, steel alloy...(Chromoly). For most of the rides, this bike weighed in at 30.4lbs, but a switch-over on the cranks chopped off nearly 2 lbs. The appropriate weights were used on this bike in the statistical analysis depending on how much it weighed on any given ride.
THE EXPERIMENT
Ok--pseudo-experiment. For 14 rides between August 2014 and the present, I switched up which bike I took on my solo training rides. I did not include any group rides because of the complexity of accounting for how much time I spend pulling vs. drafting and other group dynamics.
For each ride, I used Strava, and afterwards recorded the distance of the ride, the average speed, average wattage, and I also made note of the route I took because the conditions are very different on the very bumpy/mixed terrain river trail, compared to the smooth/recently refinished old highway.
Here are the preliminary results:
Bike
|
Gearing
|
Weight
(lbs)
|
Tires
|
Handlebars
|
Avg.
speed
|
Avg.
distance
|
Avg.
diff. from route
|
Allez
|
2 x 9
(50/36 x 12/28)
|
22.0
|
700 x
23
|
Drop
|
17.9
|
17.9
|
0.2
|
Scratch
|
1 x 7
(42 x 14/34)
|
23.5
|
700 x
35 CX
|
Drop
|
16.0
|
15.6
|
-0.4
|
Carve w/ front lockout
|
2 x 10
(36/22 x 11/36)
|
27.0
|
29 x
1.4
|
Flat
bar
|
16.1
|
8.7
|
-0.8
|
Trek MTB
|
1 x 8
(42 x 11/34)
|
28.6
|
26 x
1.95 semi slick
|
Flat with
center aero bars
|
17.3
|
17.6
|
0.3
|
So far, the mathematicians get to pat themselves on the back and chalk up a point!
However, the results from the other road bike and the other MTB don't seem to "add up" in the same way. Why would a 30 lb steel bike be 1.3 mph faster on average than the 23.5 lb road bike, for example?
Well to really see what is going on, we need to step up our analysis a little bit and statistically control for different factors that may be at work here.
SPEED COMPARED TO ROUTE AVERAGE
But first, we can take care of one part pretty easily. It turns out that the different routes have different average speeds regardless of which bike was being used. Check out this table:So instead of talking about average speed from now on, we will talk about how much faster the bike is than what we expect it to be. That way bikes aren't penalized for being taken on slower, bumpier, curvier, hillier routes.
USING MULTIVARIATE REGRESSION
To make the other adjustments, we turn to a form of statistical analysis called "multivariate regression", but you don't have to know multivariate regression to understand the results...
In fact, I am not even going into the details of it, other than to say that we can now predict how much faster you will be compared to average based on 5 factors: RIDE DISTANCE, BIKE WEIGHT, TIRE WIDTH, AERODYNAMIC HANDLEBARS, and WATTAGE.
There is an equation that expresses how these things work together to predict speed:
SPEED
DIFF=-0.055*DIST -0.123*WEIGHT -0.0495*TIRES +1.983*AERO +0.02237*WATTS -0.2053.
It works like this. If you keep all the other factors the same:
Each additional mile you add to the total ride distance is
predicted to slow you down .055mph
|
Each additional pound of bike weight slows you down 0.12mph
|
Each additional mm of tire width slows you down .05mph
|
Having an aero position on the bars speeds you up by 1.98mph
|
CONCLUSIONS:
So who is the grand winner? What impacts speed the most? Aerodynamics, Tire Size or Weight? We put all three on trial, and, when keeping everything else the same, having an aerodynamic position on the handlebars made the biggest difference.
Specifically, having a more aero position on the handlebars means you can basically count on a 2 mph increase in your speed (at least I can!!).
The next biggest factor of the three was weight. Keeping everything else constant, each extra pound will probably slow you down by a MIND BOGGLING 0.12 mph. OK, so it doesn't sound like much, and basically it isn't. But it is something. The extra 8 lbs you carry on your average chromo steel bike compared just to aluminum means you are about 1 mph slower! That is a difference you are pretty sure to notice.
Compared to a pretty nice carbon bike, your chromo steel ride is costing you almost 2 miles per hour! That puts some perspective into the argument. Now the next time someone says not to worry to much about shaving weight of your bike, and someone else says, "If you have the money, go for it!" you can decide exactly how much it is worth to you. Just multiply the weight savings in pounds by 0.12 and that is the difference you may be able to expect on your speed.
So, if you can pay $100 to drop a pound off your bike, you can expect an increase of 0.12 mph in your speed. Only you know if it is worth it to you, but now you have hard numbers to use in making a decision, but with $100 to burn on a bike upgrade, I would probably drop that money on a pair of aero bars if you don't already have them based on these numbers, and get almost 20 times the speed boost you would get from dropping a pound of weight!
As far as tire width, it looks like the pros are on to something with their skinny tires--usually between 23 and 28mm. I like have 50mm tires on my the Trek named Conversion though because it completes the great "all-arounder" feel on that bike. It is a bike that can hold its own on the roads, but is not afraid to jump of onto some fire roads, gravel or even some XC- or CX-like terrains. So what are those 50mm tires costing me in terms of speed? 0.05 mph for each millimeter. Because I am about 25mm over the average road bike, they are costing me about 1.25 mph. So that tiny multiplier of 0.05 mph lost per millimeter of tire width can really add up! So, tires can make a big difference. Not to mention that dropping to skinnier tires would probably drop a little weight off the bike as well.
PUTTING IT ALL IN PERSPECTIVE
My beloved chromo-steel Trek costs me compared to my Allez Sport:
- The extra 8 lbs cost me about 1 mph, and
- the 50 mm tires cost me about 1.25 mph, compared to the average road bike tires
- If I do not use my makeshift aero bars, it costs me almost another 2 mph
In total, a rigid steel MTB at around 30 lbs, without an aero bar setup and with 2" slick tires is about 4.25 mph slower than the typical aluminum entry-level road bike!
Bad news if you love your steel bike? Well, for me, I make it work by putting a little more effort (WATTAGE!) into my rides on the steel bike--probably because I am having so much fun destroying those potholed roads of Michigan. If you love it, it can work for you too. But this pseudo experiment backs up what we already know watching the Tour de France about the best bike setup if nothing but speed matters:
- Get the lightest bike you can legally ride (14 lbs)
- Go as skinny as you can afford to go on tires before the ride gets too harsh
- Have an aero option on your handlebars (biggest impact!)
But, for those of us not in the Tour, now you have enough information to tell you exactly how your favorite ride will probably impact your speed!
IMPACT OF AERODYNAMICS, TIRE SIZE AND WEIGHT CALCULATOR
Multiply
|
by
|
||||
Distance of ride
|
-0.055
|
||||
Bike weight
|
-0.12
|
||||
Tire width
|
-0.05
|
||||
Aero bars (y=1/n=0)
|
1.98
|
||||
Avg watt
|
0.02
|
||||
Then add it all up and subtract .21
|
|||||
That will be your predicted speed improvement (or decrease if
negative)
|
|||||
No comments:
Post a Comment