Folding Bikes - BF, Birdy, Swift owners: a completely unscientific test

Do any performance folder riders own full sized bikes with similar tires? If so, it might be fun to use the folder in place of your big bike for 10 trials or so and report your average speed on each. Anyone up for the task?

You would need to:
1) Use the average speed function on your computer.
2) Use similar tires on both bikes.
3) Consider weight differences a simple disadvantage of a folder.
4) Make sure that the route conditions are similar or do at least 30 trials.
5) Try not to draft other riders! (This puts the folder at an advantage and adds a variable.)

This will still be biased by the placebo effect, slight inaccuracies in calculated wheel circumfrence (0.5 cm adds up over 20Km!), variations in conditions (the more trials, the better).

We'll see if Dr. Moulton's laboratory tests hold in real world riding conditions.

Dr Moulton's ideas vs real world conditions... Great idea for a study.

Most modern folders don't have suspension, like Moultons did. I find that small wheels without suspension suffer quite badly on poor roads, so only the Birdy is close.

At least, that's how it feels here in SF on my Swift, where the roads are $h!t. London roads weren't great, but where smoothed out enormously by my Moulton F-frame. YMMV depending on local surface conditions..

It will be interesting to see the results. My full size bike is a recumbent, so I don't think any meaningful comparison can be drawn..

Some times I think about testing efficiency of various bikes by starting on top of a hill at full stop and seeing how far the bike will roll with out pedaling.

Anything more complicated than that would have to control for the amount of energy you put into the bike. It might be possible to do with a heart rate monitor using pulse as a proxy for energy output.

Great idea !! Purely subjectively, I've always felt a 700c with 100+psi makes most road rides easier. Although we have a huge amount to thank Alex Moulton for, on road bikes the best suspension, (and most adjustable :) ) is that long 'air-shock' that wraps around the rim. He came from a wealthy rubber making family and had expertise in rubber suspension systems in cars, so he had a 'stakeholding' in doing the same on bikes.

Suspension has its place, off road and on rough roads. My old moulton F frame was brilliant at smoothing out the rough - but I felt cheated by good ol 'Bob' when climbing, unless spinning really smooth. Favourite off road, down-hill runs would be impossible on non-'full-sus' bikes, at least at similar speeds, but then where is the fun.

I had plans for such trials, and outlined them in the "Light Swift" thread. At the time, my Swift was a fixed gear, and I have a 700c fixed gear.

I've since put a Nexus 8 premium hub on the Swift, so there are no immediate plans to do the original comparison, though I still hope to at some point.

However, I did just ride both bikes on a loop that's a good TT course: no traffic, fairly good road surface. Though the drivetrains are very different, and my aerodynamic position is bound to be a factor, the results were interesting.

Both bikes were within a second or two of each other--total elapsed time 14 minutes. I wasn't watching the clock during the runs, so I don't believe I was in any way aiming for a similar time. Unfortunately, I don't have a heart rate monitor or power meter, which would obviously be useful tools for such comparisons.

While the course has no major climbs/descents, it's not perfectly flat, so the Nexus was an advantage on the Swift. On the other hand, the Nexus certainly has more drag than my fixed drivetrain. The 700c bike has a Pasela 28mm on the front wheel, and a Pasela 32 on the rear; the Swift has IRC metro 1.25's front and back. I subjectively think I'm more aero on the 700c--flatter back, mainly.

Obviously, the above leaves a lot to be desired in terms of a test, not least because it was just one trial. I do hope to ride it occasionally on all of my bikes, to track fitness as well as compare sleds.

One final point--with the IRC Metros (or I assume any tire of equal or greater width), I don't feel I'd gain much by having a suspension. Run at 80 psi, they seem to soak up about the same amount of road irregularites as narrower tires at higher pressure.

Jack

I dunno, sounds awfully time-consuming for something you can figure out from the comfort of your desk. :D

In most cases, the performance impact of the stiffer folders will be pretty minimal and comes from a few things:

- more relaxed geometry (not applicable to some of the BF's)
- simplified gearing (i.e. harder to get optimized cadence; smaller overall range; also not applicable to some folders)
- more rolling resistance
- better aerodynamics of the small wheels = better performance at higher speeds (15+ mph?)
- smaller tires = rougher ride
- any suspension = performance penalty

So basically, if you get a Pocket Rocket Pro with a steel frame, drop bars and a double, I don't think you'd see any performance penalty. Birdy and Swift will have a minor performance hit from the geometry and gearing; the Swift will also have a harsher ride, whereas the Birdy will lose out from the suspension.


If you are going to pursue this anyway, you would really want to match not just the tires, but rider position and handlebar setup. Also....


You would need to:
1) Use the average speed function on your computer.
Unfortunately, this is just about the worst possible means of evaluating efficiency. To get something more accurate than ye olde armchair guesses, you'd have to use a power meter and see how fast you go, in the exact same conditions (including wind and temperature), with the same wattage output.



3) Consider weight differences a simple disadvantage of a folder.
Actually weight makes almost no difference in terms of performance. You probably wouldn't pick up on a 5 lb difference via subjective means.

Some times I think about testing efficiency of various bikes by starting on top of a hill at full stop and seeing how far the bike will roll with out pedaling.

Anything more complicated than that would have to control for the amount of energy you put into the bike. It might be possible to do with a heart rate monitor using pulse as a proxy for energy output.
That measure is biased by the weight of the bike because it takes more energy to get a heavy bike uphill in the first place. Sounds like a good idea if you divide by the weight of the bike. Although it still doesn't include drive train efficiency and losses due to impedence mismatching (loading effects on the body which account for the difference in the body's ability to hoist 1 pound 1 foot up, 1/100 of a pound 100 feet up, or 1000 pounds 1/1000 of a foot up), a serious omission as these factors probably account for the majority of difference between scooters, road bikes, kick bikes, mountain bikes, fixed gear bikes, etc.


Actually weight makes almost no difference in terms of performance. You probably wouldn't pick up on a 5 lb difference via subjective means.
This is often stated, but I seriously doubt it's validity. The human body is extremely sensitive. You'd notice the difference walking if your shoes were 5 lbs heavier, why should a bike be any different?

This is often stated, but I seriously doubt it's validity. The human body is extremely sensitive. You'd notice the difference walking if your shoes were 5 lbs heavier, why should a bike be any different?

Because you're accelerating/decelerating that weight with every step when walking, but once you've expended the energy to get the bike weight moving, you're done (uphills + sprints excepted, of course).

Serious analysis allowing you to explore most all of the various factors in bike speed is available at: http://www.analyticcycling.com/ It takes some time to explore, but is enlightening.

Jack

Because you're accelerating/decelerating that weight with every step when walking, but once you've expended the energy to get the bike weight moving, you're done (uphills + sprints excepted, of course).

Serious analysis allowing you to explore most all of the various factors in bike speed is available at: http://www.analyticcycling.com/ It takes some time to explore, but is enlightening.

Jack

Hmmm, I had written a post explaining that the acceleration/deceleration of your feet have little to do with it, as a 5 pound overcoat would certainly be noticeable despite the fact that your upper body does not accelerate/decelerate with your feet. I then proceeded to use the fact that the energy wasted by the muscles when walking is not independent of weight to question the assumption that it is independent of weight when cycling and can, therefore, be ignored when comparing cycling on a heavy bike to cycling on a light bike (where only the mechanical energy required to propel the bike is often considered).

However, I soon realized that the energy wasted by the muscles is independent of weight when cycling with the right choice of gearing. Here's why:
Assuming you're going fast enough that the bike essentially balances itself (independent of bike weight and speed) and assuming the affects of bumps and wind resistance on your muscles are negligible (or at least relatively independent of weight and speed), the only contact your body has with the outside word is through the seat, the handlebars, and the pedals. The affect of the seat and handlebars on your muscles is obviously independent of bike weight and speed (they're just supporting your body and can't discern the weight of the bike underneith them anymore than they can discern the weight of the asfalt). Therefore, the only possible avenue for the weight of the bike to affect your muscles is the pedals. The only feedback property the pedals possess is resistance. Resistance can be selected via gearing, with the only consequence being a change in the speed of the bike, which by assumption does not affect any of the other muscle feedback mechanisms (balance, bumps, wind resistance, and seat/handlebar forces). So with the right choice of gearing, the pedals' affect on the muscles is constant and, thus, also independent of bike weight. Therefore, all muscle contact with the outside world is independent of bike weight, which means that the energy wasted by the muscles is independent of bike weight.

The key insights here are:
1. Sitting on the weight as opposed to the weight sitting on you. Contrary to the simple rules of mechanical energy, muscles waste chemical energy when producing a constant force (this energy goes into heat, as the sustaining the force itself requires no mechanical energy). So sitting on top of a weight is not the same as a weight sitting on top of you (as it would be in a purely mechanical system). When you carry something while walking, the weight is sitting on top of you.
2. Mechanical leverage. The muscles do not always waste the same amount of energy for a given power output. If you had to adjust the cadence and force of your legs while pedaling (as you would with an ungeared bike), then the energy wasted by the muscles would depend heavily on the weight of the bike. So although bike weight would not affect the performance of a perfectly geared bike, it would affect the performance of a single speed bike or a scooter.

Of course, this analysis neglects the direct affects of balancing, bumps, wind, and imperfect gearing on muscle efficiency, but these things are so heavily influenced by the design of the bike in question that they are probably more appropriately attributed to bike design than bike weight.

So there you have it. Sorry for lecturing you all, but this was twisting my brain and I thought it might help one of you ease your own cerebral torsion. :)

Yangmusa, I think you would blow any upright away on your recumbant, so that's out.

Upright positions do make a huge difference. Birdy has a stem that puts you about as far forward as a road bike (a bit more upright), and many Swift owners have dialed in their cockpits from the looks of the photos here.

Probably not too many people have power meters, Baccia. But your armchair hypothesizing has some problems. Foremost, we don't really know whether small or large wheels have more rolling resistance. Most tests show smaller wheels have less on perfect surfaces. (The original Brompton tire at 120 had 1/3 that of a 700cc in one test.) Second, we don't really know what energy sapping effects suspension has. Moulton argues that it reduces rolling resistance. It certainly doesn't take much force to bounce a shock. Ditto with weight. Third, hey, this is for fun anyway! The results won't mean much.

I do know that my journey time to work on my bike is about 40," but was over an hour on a 30# commuter with Kenda Kwest tires. That's a big difference!

This is often stated, but I seriously doubt it's validity. The human body is extremely sensitive. You'd notice the difference walking if your shoes were 5 lbs heavier, why should a bike be any different?
It's pretty well known that for walking / running, weight on the feet is far worse than weight above the waist, by the way. Kind of a bad example. ;)

As to your theories, I'm not sure they're correct.... Weight matters when accelerating (which is fairly brief on the flats) and when resisting gravity (climbing). And if you're going more than 10 mph or so, your biggest enemy isn't weight -- it's drag. This is why position is so critical to performance.

Humans may be sensitive, but our brains are not equipped with gauges, calorimiters, thermometers, speedometers or HRM's. ;) One of the first things you learn about endurance training is that we are rather bad at judging expenditures of effort, which is why HRM's or power meters are de rigeur for effective training.

Thus, 5 lbs on the bike frame makes virtually no difference at all for performance on the flats. It may change the ride feel, but not the actual speed. The place where weight makes a noticeable difference is on the climbs.

And how big is that difference? Chris Carmichael discusses this in a recent issue of Bicycling Rag --er, Magazine. Basically, dropping 5 lbs from the rider's body will cut about 40 seconds off of a 5k, 8% grade climb. (Boosting your power output by 20w, in comparison, will save you 2+ minutes.)

I don't think you're going to get very good measurements, let alone awareness, of this via subjective means. Unless you're doing Cat3 or higher racing, the performance differences in bike weight, especially if it's 5 lbs or so, is pretty much negligible; and with folding bikes, other factors like geometry and rolling resistance will have a much bigger impact anyway....


pm124: I agree that no one will put a power meter on a Dahon. ;) But unscientific tests produce unreliable results.

I'm pretty confident that a pair of Stelvios will have more rolling resistance in 406 than 700 x 23c, and I don't see how suspension could provide an enhancement to performance. (Besides, if it did then the pro racer's bikes would have suspension for more races than just Paris-Roubaix....)

As to your example, without knowing anything about the bikes in question I can't possibly theorize about the possible causes for a performance difference.

Although it's a bit of a "love-fest" for Dr Moulton, this video is quite interesting:
http://video.google.com/videoplay?docid=-8522870086389552343&q=alex+moulton&total=30&start=0&num=10&so=0&type=search&plindex=0

It's about the development of the Moulton New Series - but it also talks about the original Moulton bikes, and all the research and testing that went into it.

pm124: I agree that no one will put a power meter on a Dahon. ;) But unscientific tests produce unreliable results.

I'm pretty confident that a pair of Stelvios will have more rolling resistance in 406 than 700 x 23c, and I don't see how suspension could provide an enhancement to performance. (Besides, if it did then the pro racer's bikes would have suspension for more races than just Paris-Roubaix....)

As to your example, without knowing anything about the bikes in question I can't possibly theorize about the possible causes for a performance difference.

Can't agree with you more on the first point. So, do we do a randomized controlled trial of mammography versus no mammography and then wait 10 years to see whether women denied the test have shorter lives? (OK, an extreme example, but if you can't get the data, it's slightly informative and fun to get input. We already have one data point!!)

Regarding the Stelvios, check this out: http://www.legslarry.beerdrinkers.co.uk/tech/GS.htm. OK, anyone with beerdrinkers in their URL is suspect.

Regarding suspension, the argument only applies to rough surfaces, where small objects push back. These are the two sides of the problem. Big wheels will have less rolling resistance on rough road precisely because they have a bigger contact patch. Small wheels get push back from pebbles, etc. Suspension allows the wheel to roll over such objects. But, yes, your point is well taken. In fact, the endurance of the diamond frame and large wheeled bikes probably has some reason beyond racing regulations. (Yes, recumbants are the most efficient machine out there and have done 150 MPH, but that doesn't do you any good if you are sucking exhaust.)

So, enough of this mental ************ ladies and gents. Start your odometers!

Actually, I have done this test, many times, on each of my 3 bikes while riding to work. I have not relied on the cyclometer, but have simply measured the time for the commute. There is signficant variance for any given ride (depending on wind and whether the traffic light break my way) but the long term average is the same for all bikes.

The bikes:
Friday NWT with 406 wheels, Sachs 3x7 internal hub.
Bianchi Volpe tourer, lightened up with a carbon fork and Easton Circuit wheels.
Friday Pocket Rocket Pro.

Though the 3 bikes can be ridden in the same time over the 17 miles to work, they are really very different bikes. The PRP is my favorite. It is way lighter than the other two. I hate it on hills because it's geared too high, but it accelerates much faster than the others. The NWT is a tank, but it's my favorite bike for hills despite its weight, because it's geared low, and the fit is ideal of out-of-the-saddle climbing. The Bianchi is intermediate between the two. Mostly, it doesn't fit me well (which is why I like the Fridays).

The only big difference I've noticed between these bikes is when I swapped the tires on the NWT from slick Comets to cushy Conti Top Touring. That alone was enough to add 5 minutes to the trip time.

It's pretty well known that for walking / running, weight on the feet is far worse than weight above the waist, by the way. Kind of a bad example. ;)

So it was.


As to your theories, I'm not sure they're correct.... Weight matters when accelerating (which is fairly brief on the flats) and when resisting gravity (climbing). And if you're going more than 10 mph or so, your biggest enemy isn't weight -- it's drag. This is why position is so critical to performance.

The theory I gave was justification why bike mass is equivalent to body mass. What you mentioned here only relates to bike mass's contribution to total mass, the importance of which has already been covered ad nauseum (acceleration is brief, gravity is conservative, bike weight is small percentage of overall weight, etc).

Bianchi Volpe tourer, lightened up with a carbon fork and Easton Circuit wheels.


How much did this lighten your bike up? I may end up doing this myself if I can't sell the bike.

How much did this lighten your bike up? I may end up doing this myself if I can't sell the bike.

Quite a bit. The wheels alone were a pound and a half lighter than the stock Bianchi wheels, and the whole surgery took over 3 pounds away.

If I were to do this again, I would simply get the wheels.

The only big difference I've noticed between these bikes is when I swapped the tires on the NWT from slick Comets to cushy Conti Top Touring. That alone was enough to add 5 minutes to the trip time.

To me, the tire difference is the most helpful item in this thread, and also implies that different tires (at least footprint and psi) are a major barrier to a meaningful test.

To me, the tire difference is the most helpful item in this thread, and also implies that different tires (at least footprint and psi) are a major barrier to a meaningful test.

Indeed. It's important to run the same tires on both bikes.

So, two data points, and two votes for no difference between 20" and 26+" wheels, but one vote for differences in tires...

Do any performance folder riders own full sized bikes with similar tires? If so, it might be fun to use the folder in place of your big bike for 10 trials or so and report your average speed on each. Anyone up for the task?

You would need to:
1) Use the average speed function on your computer.
2) Use similar tires on both bikes.
3) Consider weight differences a simple disadvantage of a folder.
4) Make sure that the route conditions are similar or do at least 30 trials.
5) Try not to draft other riders! (This puts the folder at an advantage and adds a variable.)

This will still be biased by the placebo effect, slight inaccuracies in calculated wheel circumfrence (0.5 cm adds up over 20Km!), variations in conditions (the more trials, the better).

We'll see if Dr. Moulton's laboratory tests hold in real world riding conditions.

I like science. I have a Bike Friday NWT and a 1984 vintage Guerciotti.

2) The Guerciotti has 25 mm high pressure tires. The NWT has 32 mm 100 PSI tires. The volume of air is not too different.

3) The NWT weighs 25 lbs, the Guerciotti 23 lbs.

4) I don't know that I can do 30 trials. I'll do as many as I can.

5) I will probably do this solo, but why would you expect drafting to be more of an advantage to the folder?

I would expect the Guerciotti to be a bit faster on average. There is the two pound difference, but I would expect that to be peanuts. The NWT gears are lower then the Guerciotti. That would be an advantage in a hilly area, but near home there are no hills significant enough to challenge the Guerciotti. The top gear of the Guerciotti is enough taller that it may prove to be a significant advantage on some long gradual downhills on my standard route. Biggest impact is that when I really push the NWT there is noticeable flex in the seat post and stem. The Guerciotti is far more solid in that regard.

(Edit) Both bikes have drop bars, and my positions on the bikes are similar.(end Edit)

Speedo

There have been several comments questioning how meaningful this test will be. I completely agree that it will be difficult, if not impossible, to extend the results of one persons side by side test to other riders and bikes. On the other hand, my side by side test is, at the very least, interesting and meaningful to me.

As noted in the post above I think that I know how this will turn out. It would be very interesting if the folder were either much faster than I thought, or much slower than I thought.

Speedo

There have been several comments questioning how meaningful this test will be. I completely agree that it will be difficult, if not impossible, to extend the results of one persons side by side test to other riders and bikes. On the other hand, my side by side test is, at the very least, interesting and meaningful to me.

As noted in the post above I think that I know how this will turn out. It would be very interesting if the folder were either much faster than I thought, or much slower than I thought.

Speedo

Great! It's all for fun, not publication. And the heavily qualified title of this thread is meant to be taken literally. Still, even if the error is +/- 1 MPH and generally systematic, it's an interesting comparison. I'm not sure why folders are supposed to be better drafters. Part of the logic is that the bike itself has very little drag as it sits squarely (low) in the wake of the bike ahead of you.

A folder can suck wheel better because of the shorter total length. So you can be closer to the body in front of you.