gsa103

I think its mainly a matter of interpretation and wording. Wikipedia is actually citing the same NHTSA study that I was recalling from memory. In straight line braking ABS outperformed "Best Effort" by 5-20% in all conditions except Grass and Loose Gravel (snow/ice were not tested). I definitely consider Grass a very low traction surface, and similar with loose gravel. They're vastly different than most city conditions.

One thing I would really like to see is how that ABS test compares with newer cars. The biggest limitation on ABS is the cycle time and electronics have improved significantly since that test was conducted. I suspect the gap between "Best Effort" and ABS has gotten bigger.


The reaction time is an interesting question especially in light of the above discussion on braking technique. The ideal test would be a car and a bike side by side doing a reverse drag race. The light turns red and both try to stop as fast as possible. I suspect the cyclist will initially react faster, since their hands are on the levers. At the same time, the car will achieve maximum braking force much sooner, especially since many modern cars have proactive braking that will apply maximum force if they detect a quick brake application, before the brake is actually fully depressed. Maybe have two cyclist tests, on the hoods vs drops.

Or you could do a more realistic test....have the car driver a cell phone, except the distance would be infinite.

dougmc

That's a pretty convincing test. He's obviously trying hard to maximize his braking and beginning to endo in his picture.

I wonder if the folk who calculated the point that the endo happens didn't move the CoG back and down as much as he did for their "butt off the seat" calculations.

I'm surprised that he found so much variation between the brakes. I've certainly found that some levers/brakes/pads/rims need to be squeezed harder to get a given effect than others, but in general I've found that I can squeeze most of them hard enough to reach the endo point, even with my butt off the back of the seat. (I say most because some of the really weak ones just bend more and more rather than braking better. Sounds like their weakest ones were like that, based on the descriptions.)

dougmc

This is probably what you were referring to, or very similar. I guess I have to concede the point, it seems that for most conditions ABS beats a skilled driver by a small but significant margin. (But it does blow it pretty badly for gravel, at least in 1999.)

Though they should test snow, ice and wet ice -- having grown up in Anchorage, *that* is what I'd call very, very low traction. From what I can find, it seems to slow braking in snow similar to how it does on gravel, and I'm not sure about ice. (No matter what you do, you're not stopping quickly on ice unless you crash into something. Though studded snow tires certainly do help.)

phoebeisis

Great find!
The numbers are just fine as comparison numbers-which brake is the best.
They aren't so good as absolute numbers-distances- deceleration- because the rider almost certainly anticipated the mark-and braked early.
You have to dial the human out of the equation- or just hook him up to sensors that will actually tell you just where he hit the brakes.
As we are all aware- a .9g deceleration-means the stop averaged .9g- peak deceleration would be more at times.
Great stuff!

phoebeisis

dougmc
I ride a mtb framed bike-but in bolt upright posture-with 1950's type upright handlebars-bad neck age-gotta see where I am going and speed is of no consequence.
I could probably get low and straighten my arms pretty quickly- but I would have a heck of a time dealing with even 120lbs slamming me forward.

Riders who ride in the drops-they would have straighten their arms--very quickly- shifting backwards in their saddle-and pull upward with their clipped in shoes-pulling their butts into the saddle-
athletic folks who practice-MIGHT be able to manage a .9g stop- if they were hyper alert-and the surface was perfect.

As you say-ABS- doesn't "make" more adhesion-but normal drivers do much better with it than without

rekmeyata

This is very argumentative, especially since reaction time is a very complicated behavior and not all humans respond the same therefore a pat answer of 2 seconds is wildly inaccurate. I've also read that the average response time is 1.5 seconds which I think is a bit more accurate than 2...for the average person, however a above average response time is actually as fast as 0.7 seconds! But what dictates above average response time? another complicated behavior, most of that is due to an alert driver, some of it's due to a person having faster reaction times either due to they way their wired or age. But a average reaction time of a non alert driver would be between 1.5 and 2 seconds and more. And there is a slew of other variables that alter reaction time such as nature of the signal, lighting, vision, complexity of the situation, gender, age, cognitive factors, perceived urgency, and maybe others. But the best estimates of average reaction time and average awareness is 1.5 seconds. I would think on a bike that time could be less by at least a 1/4th a second because you don't have to move your foot off the throttle onto the brake, your hand is already on the hoods, so I could see about a 1/2 second improvement in reaction times on a bike.

howsteepisit

this thread is approaching Monte Pythonish levels of absurdity. Anybody know what kind of swallows best apply brakes on drop bar bikes as opposed to uprights?

3alarmer 

....if anyone still cares at this point, it's in here from about p237-p261.
The short answer is, "The physics of this problem say you are in error."

RobertHurst

People will be repeating the braking-related nonsense in that book and in Forester's book for decades I guess, even though the claims have been long disproven beyond a shadow of a doubt.

The formula they used in that book to determine max deceleration of an upright bicycle assumes a seated, static rider. That's about the worst way to stop a bike.

If they had bothered to check their formula with a simple experiment the way we did in 5th grade, we wouldn't be having this discussion.

3alarmer 

...the nonsense here is in the OP. Even the best way to stop a bike cannot overcome the
basic problem of your weight distribution as a unit with the bike versus your average automobile.

But hey, experiment all you want to, friend. My suggestion would be to follow a car at about one
car length or less and have the driver do a panic stop in front of you...if you coordinate it you
can both start braking at the same time and thus minimize reaction time as a variable.

I would love to have your "disproven beyond a shadow of a doubt" material, because I always learn stuff here.

Darth Lefty 

The bike can stop in one quarter of a tire rotation if you don't care if you keep going without it.

dougmc

The "beyond a shadow of a doubt" comes from his claim that upright bikes can stop at around 0.95 g's, with no supporting data beyond "do it yourself and you'll see".

Now, gsa103 *did* provide some data (ironically, he did a better job of proving RobertHurst's point than RobertHurst did), though it was a magazine article comparing brakes and it didn't really go into much detail about the experimental procedure, raw data or math used. As given, it certainly suggests that upright brakes are capable of braking at over 0.9 g's (basically by simply saying "we measured it!"), but falls far short of proving it "beyond a shadow of a doubt".

The fact that they found so much variation between the various brakes tested and yet showed a picture of somebody with the rear wheel off the ground (i.e. for that test, the tendency to endo should be the limiting factor, not the brakes themselves) with brakes that don't appear to the be same as the one that got the best results in their test makes me wonder -- is that picture a stock photo or a photo of their real test or what? I'd want to see all their data and procedures (or just repeat the tests myself, then share with some others for peer review -- and to do this properly isn't just a few minute thing) before I really trusted them.

And even if bikes can do 0.95 g's ... cars can do more. Cars have the advantages of no tendency to endo, ABS brakes and lower tire pressures (30 vs. 60+ psi -- so more rubber on the road per pound -- rubber on concrete does deviate considerably from the "fixed coefficient of friction" ideal, and so "more is better" where if it followed the rule exactly it wouldn't matter.)

I-Like-To-Bike

Better yet would be a test where a speedy cyclist is following a car by one car length, both going the same speed, no advance warning to the cyclist that the driver will make a sudden stop, nor any advance notice to the cyclist that his braking skill will be measured or is even to be tested. Car driver slams on brakes. Measure bicyclist stopping results.

I suspect that most cyclists will crash into the rear of the car before they even apply the brakes, let alone start some previously practiced body contortions to rearrange their CG in coordination with deft application of front brakes and rear brakes.

B. Carfree

My that's a familiar scenario. I had a car pull just less than one car length in front of me from an adjacent lane and proceed to lock his wheels up (thirteen foot skid mark). Yes, this was a purposeful brake check. I did manage to apply my front (only) brake, such as it was on my old Raleigh upright, but did not successfully stop the bike before my front wheel impacted his rear bumper. The impact did stop the bike (as well as separate the down tube from the head tube) and, for the one and only time in my many, many miles of cycling, I came unseated. To my credit, I did stick the landing next to the driver's door after a full flip with a one and three-quarter twist.

I think I could have slowed enough to avoid impact, likely by swerving to the side, if I would have been on a bike with a proper set of brakes. Since I'm not keen to repeat that episode, I'll have to accept the results as they were which match quite well with your prediction except that I did manage to apply the brake and did indeed manage to shift my weight behind the seat. (The weight transfer turned out to be critical to the successful dismount.)

dougmc

The problem with that test ... is that a car following the car would probably fail it as well. Even at speeds as low as 20 mph. 20 mph works out to about 30 feet per second -- about two car lengths every second, and good reaction times are around a second.

A car could go from 20 to 0 mph in just under a second. If you take the best possible reaction time -- 0.7 seconds from what I've heard -- and multiply that times 20 mph, you get 20 feet. With an average car length of 15 feet or so, even with the best possible reaction time, he's still going to hit the car in front of him. And it gets worse as speeds increase.

Ultimately, if you follow the car in front of you at 15 feet at any non-trivial speed, the only thing keeping you from slamming into the back of him if he brakes hard is seeing what caused him to brake hard and braking hard at the same time as him -- if you instead have to react to him braking hard, you're likely to hit him if the speeds are 20 mph or higher.

RobertHurst

The claim that has been disproven, waaaaay beyond a shadow of a doubt, is that bikes will endo at .6 -.7 g maximum. You don't even need a physical experiment to prove that. (As Joe Riel showed, just by moving the body weight behind the seat the same formula spits out .83 g.)

I've never made the claim that bikes can stop at .95 g myself, but it sure seems likely to be possible, considering I can personally stop my ill-mainained Waterford at .85-plus (~.85 g on a slight downhill slope), and have personally measured other riders with calibrated speedometers closing in on .9 g as well.

To someone who performed this very simple experiment multiple times, the continued lecturing about how bikes will endo above .6 is pretty funny. You'd have to be a really awful bike handler to endo your bike at .7 g (meaning, anything less than 20 feet from 20 mph).

My suggestion is that people spend more time reading books with claims that have actually been checked, when possible, by the authors.

RobertHurst

Why don't you read the rest of the thread, friend.

3alarmer 

...I reviewed it page by page prior to posting. You are talking through your hat.

rekmeyata

this is great point.

spare_wheel

As long as their brake lights work it's not a problem. The downhill portion of my commute is on a major ambulance route.

spare_wheel

Gah! A skid and controlled fall is a far better option than a full on hit of a stationary object.

RobertHurst

Okay then.

The "laws of physics" dictate that you will endo your bicycle when attempting a .7-g stop.

Anybody who can stop shorter than 20 feet from 20 mph must be a witch.

RobertHurst

This thread is a frightening reminder of the way humans cling to their beliefs even after they've been proven wrong.

RobertHurst

Some can, some can't. Many can't. My '70 Chevy pickup certainly can't come close to that.

So -- can a bicyclist stop faster than a car? Yes, it is possible, but it depends on the car and depends on the bicyclist.

This is not that hard.

I-Like-To-Bike

If that is meant to be a convincing argument, I remain unconvinced by it.

I am also unconvinced by other posters who cite experiments and measurements of individuals selected specifically for their ability/skill to maximize the results as being relevant to a discussion of bike braking distances for typical bicyclists. It is as relevant as citing the stopping ability of Formula 1 racing machines driven by Formula 1 drivers.