Lighter is not necessarily faster
Thu, February 21, 2013 
Here is an interesting article that bears out something I have been saying for years. That the weight of a bike has little bearing on the performance, even when climbing hills, even though simple physics tells us that a heavier bike should be slower.
Bicycle Quarterly did a controlled test with two riders climbing a hill, one on a 17 lb. titanium frame and the other on a 26 lb. steel randonneur bike. That is a whopping 9 lbs. difference and yet these bikes consistently climbed side by side at the same rate even with different riders.
The randonneur bike also has mudguards and a handlebar bag that must increase drag. So clearly the rider on the randonneur bike is putting out more power, enough to offset the huge weight difference. Is this possible?
Apparently so. In another test Bicycle Quarterly did they compared two bikes weighing the same but with different frame tubes.
In this test, with the same riders switching bikes, the same one bike was consistently faster. Power Tap meters were used to measure each rider’s output. On the faster bike one rider put out 5% more power than he did on the slower bike, and the other 2% more.
The author of the above linked article points out that it is difficult to reach maximum heart rate on a bike, unlike running which of course is much harder. He surmises that if the muscles can’t use all the oxygen that the heart is pumping to them, why should the heart beat any faster?
If one bike or bike frame leads to the rider becoming fatigued at a faster rate, the power output for that bike or frame will be less.
Back in the early 1980s when I was building my custom ‘dave moulton’ frames, I made a “Road” frame built with Columbus SL tubing, and a “Criterium” frame made from the heavier and stiffer Columbus SP tubing.
People who bought the Criterium frame told me all the time, “This bike climbs like you wouldn’t believe.” Many different people, of all different riding abilities, all said the same thing.
Another typical comment Criterium frame owners would make, “This bike kicks my ass… I can’t help but ride fast and I come back from a ride knowing that I have had one Hell of a work-out.” Clearly the stiffer, even though heavier Criterium frame caused the rider to put out more power.
All frames I built were stiffer than many production frames in the same material, because I put less heat into the frame. Many larger manufacturers used conveyer style production lines with gas flames to pre-heat the frame before it reached the person who would braze the joint.
Heating a tube in this way means it is glowing red hot several inches from the lug, thus removing much of the tube's inbuilt strength. Working quickly using a smaller but hotter flame I heated the tube barely a quarter of an inch from the lug.
Terry Shaw who was the owner of Shaw’s Lightweight Cycles, and a bike dealer who sold more Fuso frames than any other, coined a phrase that was part of his sales pitch. He always said "The Fuso climbs lighter than it actually is.”
So rather than looking for the lightest bike, people should be looking for the one that enables the rider to put out the most power. Because isn’t the whole object of a road bike to go faster if you are racing and to get more of a work out if you are training or riding for exercise.
Manufacturers of bikes for professional riders should be researching this. Although always it is the rider, not the bike that is faster; it seems highly probable that some bikes make better use of the rider’s output, and may even cause the rider to increase his output.
Reader Comments (17)
A lighter bike requires less energy to move against the force of gravity. Its simple physics - energy equals mass x distance x acceleration due to gravity. Power is energy divided by time .
If two identical riders ride up a mountain at 12mph, one on a light bike and the other on a heavy bike (both with equal stiffness), the time taken to get to the top stays the same but the mass is greater on the heavy bike so more energy is required to climb at the same speed. Having a heavy bike is fine for training as it forces you to ride harder to keep up with other people when climbing. but in a race the whole point is saving energy for the times when you need it . The lighter bike (if as stiff as the heavy one) saves more energy on climbs thus the rider will be less fatigued at the top.
If the two riders both climb at the same power output, the one on the lighter bike will reach the top quicker .
with all that said, it's easier to lose body weight than bike weight , andif you're not already super lean you can lose much more of it , so will gain more of an advantage from this as long as you don't lose power in the process.
There are a number of variables here, if the bicycle quarterly test (I also read the article) had been run in the same manner with two Dave Moulton frames which were unlabelled (so the name Criterium could have no psychological effect) and the fractionally lighter frame were weighted to eliminate the effects of gravity (or anyone weighing them to determine a difference) and neither rider knew which bike was which then the only variable would be frame stiffness. I also assumed that the frame angles are the same and other cycle parts are identical. It would be valid to conclude that this was the factor in climbing ability or not. Psychology has such a huge impact on performance that even the frame colour would have to be the same.
While I'd agree that the heavier tubing is stiffer (more material), I'd disagree with your statement that heating up the tubes more when brazing will affect stiffness. It will affect strength (how much load the material will take before permanent deformation or failure) so your frames should be stronger and last longer. Stiffness should remain essentially the same, I believe.
One more thing - Power Taps are hub-based power meters. They measure power after losses due to drivetrain friction, chainstay flex, BB flex etc.
Example: Two riders on different bikes, one stiff & with clean drivetrain, other one flexy and with dirty chain. Both riders riding at an actual power output of 300 watts. The powertap on the flexy bike with dirty chain will report 285 watts, the one on the stiff bike with clean chain will report 295 watts.
So unless you use a crank-based power meter instead, you can't really compare power output on two different bikes unless you account for these variables.
I agree with what you're saying though - and I know that practically there is a tradeoff between weight and stiffness... benefits of higher stiffness can outweigh drawbacks of higher weight... it's about finding the best compromise.
Jan Heine (the author of the linked article) has also written quite a bit about narrower and higher pressure tires not necessarily being faster than wider and lower pressure tires. His work and writing on frame flex, wind resistance, and frame geometries is also excellent. You may not always agree with his conclusions, because they so often fly in the face of "accepted wisdom", but Bicycle Quarterly is always worth reading. The website only has a fraction of the content available in the magazine.
If two riders are riding side by side up hill and one rider gives the other his water bottle, it means one rider has lost a pound and a half, the other has gained that amount. Is there any difference that either rider can feel?
Human bodies are not mechanical machines that put out a fixed amount of energy. We put out whatever is needed to get the job done. That is true whether we are riding a bike, pounding a nail in a piece of wood, or opening a screw top jar.
This last week on stage 5 of the tour of Oman I saw Contador, Rodriguez, and Froome go head to head. Froome appeared to be slightly stronger than the other two. I would speculate that if Froome had been on a steel frame weighing 1 lb. more the result would have been exactly the same.
The above scenario will never happen so it is a moot point.
Dave
"Simple physics" is not the way to analyze complex systems. Bicycles are relatively simple machines, but they're powered by relatively complex machines, and the two working in combination are inherently complex. One thing I appreciate about Heine's work is his attempts to quantify through testing the affect that that different variables have within this complex system. His results on the affect of tires, frame tubing, frame angles and the like on cycling performance seems very much worth reading and using as a springboard for further research. Nothing demonstrated experimentally should be dismissed out of hand just because it appears to violate "simple physics."
Good write-up. I've noticed, too, that my lightweight SOMA doesn't appear to climb any faster than my touring Americano (21 vs 33 pounds). Here's another question: My Bike Friday pocket rocket does appear to climb EASIER (and faster?) than a full sized bike. 451 vs 700 wheels . . .
Do you know if anyone has done research (real research - NOT BIke Friday) on whether or not small wheels really do result in a faster climb ride? This upcoming Furnace Creek 508 rider really would like to know.
Always a good and interesting read Dave. I believe stiffness is the critical factor in climbing and how efficiently the bike responds to rider input. My CF bike climbs better than my Ti bike, though they are within a pound or two of each other. I've always put this down to the increased stiffness of the Cf frame.
The weight debate is always an interesting one. I have two road bikes: Newer Ibis carbon fiber with Ultegra, nice wheels, etc. Never weighed it, would guess 17 - 18 pounds with cages and pedals. Bike number two is a steel '97 Ibis cyclocross bike set up for commuting - mixture of road and mountain bike parts, fenders, etc. Never weighed it, would guess 21 pounds or so - with cages, pedals, etc.
My commute is 17 miles each way, mostly flat, then a steep 2 mile climb home. And of course, reversed, a 2 mile downhill on the way to work.
I've commuted hundreds of times, timing myself - not to the second, but to the minute. After hundreds of commutes, pretty good way to average things out.
Which bike is faster? No difference. The carbon Ibis feels faster and snappier, but in fact, no real speed gain. My personal best time, going to work and riding home, is in fact on the older steel Ibis. And that's because at the time, I was fitter and riding well. The bikes appear to make little or no difference.
Interesting stuff and flies in the face of most bicycle marketing.
Dave,
I really hate it when you post simple truths, because they almost always point to simple facts. It's not the weight of the bike, but the rider. If you put Taylor Phinney on a Huffy that weighs 50 lbs, and a 'weekend warrior' on a team-replica bike (by any manufacturer), he'll be faster and win an uphill run every time.
I have found that no matter which bike I ride (my old Bottecchia, my new (lighter) custom Ti, or a 'superlight' carbon bike), my legs make the difference. When I am fit, I ride faster. When I am fatter, I'm slower. In the grand scheme of things, the few grams you save in frame weight make little to no difference in the overall weight hauled uphill. Lose body weight=become faster.
All I know is I bought a Bob Jackson Reynolds 853 frame/fork last year, outfitted it with Campy Chorus, Turbo saddle, and a Gimondi bar and that bike climbs and accelerates like greased lightning. I wanted something that would remind me of my venerable FUSO from the 1980's (Columbus SL). It's getting weirder and weirder out there in "Lighter is always Better" world. I absolutely hate these outboard-cup bottombracket setups that are also supposed to be so much lighter. They appear to be a solution for a problem that didn't even exist, until you started marketing plasticized-carbon fiber frames. The bearings are a little bigger than a BB (as in an air-gun; not 'BottomBracket') and the Campy Veloce version I tried on my previous frame started clicking after only a month of use... back to my Campy SR bottombracket from the 80's which probably has ~100,000 miles on it... with regular maintenance, I've had to changed out the bearings twice in all that time. That's what is missing these days... a balance amongst many factors... weight... durability... simplicity... etc.
I have two ally road bikes, one a 90s machine with a threaded steerer and square taper cranks that weighs 9.1kg, and the other a modern bike with a cheap heavy frame, but 31.8mm bars, Hollowtech II cranks, huge downtube etc. that weighs about 11kg.
I'm only a light guy (65kg), but the stiffness of the modern gear more than makes up for the extra weight; I'd much rather climb on the new bike.
So IMO stiffness matters a lot more than weight. Although of course, psychologically weight is magnified as a factor... I'd suggest it's best to minimise the perceived weight, by using lighter tyres, rims, seat, seatpost, bars and stem.
So are people disagreeing with the fact that
a) the heavier an object is, the more energy is required to move it a certain distance against gravity, and
b) the human body contains a limited amount of energy (particularly glycogen)?
"a) the heavier an object is, the more energy is required to move it a certain distance against gravity, "
Well it depends. It is easier riding a 30 pound bike uphill than dragging a 30 pound sledge uphill. It depends on how the energy is applied and friction and other efficiency losses.
Excellent post, as always, Dave!
On the subject of which power meter to use, Jan is limited in his options (DCRainmaker likely would have had at least 3). But PowerTap does the best job in this case, because one of the assertions is that flexible frames, which Jan prefers (at least in the top tube) transmit less power to the wheel. By using a powertap he's able to include this potential effect in his comparison. From what I've seen frame flex is mostly elastic and the energy going into flex is restored to propulsion when the frame de-flexes. So the issue then becomes which frame becomes the best interface between rider motion and bike propulsion. In Jan's case it's clear his familiar randoneurring bike does better than the LiteSpeed or Lynsky Ti frames he tested. I suspect you might get a different result if Peter Sagan was the rider. The point is valid, however: variability in power production is much greater than variability in power requirement due strictly to mass differences. Mass effects are easy to quantify, so they receive focus, while "interfacing" is extremely different to assess. Too often stiffness is assumed to result in better power transmission so companies advertise raw stiffness numbers. But Jan's point is stiffness is best tuned and optimized. In any case, I thought the article was excellent: open to scrutiny like any experiment, but food for thought and there are lessons which can be extracted from it.