by John Forester

(Originally in Bike World, March 1974)

home              bicycle engineering

Do you wonder how good your brakes are? I don't mean do you wish you had Campys or that the quick release was elsewhere; I mean do you worry that your brakes won't stop you in traffic or slow you on a hill? I made some experiments recently to find out just how safe rim brakes and coaster brakes are.

To explain what I did and why, I must start with some brake theory. Brakes have two different performance limits-deceleration and horsepower. Deceleration measures how quickly or in how little distance you can make one stop-like when the intersection ahead suddenly becomes blocked by cars.

Horsepower measures how big a hill and how fast you can safely descend. Brakes that are good in one way may not be good in the other.

There are three different kinds of brake: rim brakes, coaster brakes, and hub brakes. Rim brakes press rubber blocks against the rim, and are hand lever actuated. Coaster brakes force together metal parts inside the rear hub that are lubricated by grease or oil; these brakes are actuated by reverse pedal force. Hub brakes force asbestos brake lining against a drum or disc built outside the hub itself and outside its grease-lubricated bearings; these are actuated by hand levers.

Brakes may be fitted to either or both wheels (except, of course, coaster brakes which fit rear wheels only). Which wheel it brakes makes an enormous difference in deceleration but none, by itself, in horsepower.

Deceleration is measured in gravity units. If your brakes could persuade the road surface to push back at you with a force equal to the weight of you and your bike, you would decelerate at 1g.

What happens is that the front brake tends to lift the rear end of the bike. At about 0.678g, (depending on the wheel base and your position) the front brake actually lifts the whole rear of the bike and you roll on over.

Consideration of how much force it takes to skid tires shows that the rear brake force must be "subtracted" from the front brake force. The rear brake can only be effective if the front brake is used at less than full deceleration.

It sounds like all we need is a front brake. This is true, except for four things. First, the skidding rear wheel gives a warning when you're about to catapult over the handlebars. When using sewups on long hills, it's best to divide the horsepower between two rims to keep heat from softening the rim cement; in the rain it is good to have two brakes; and a reserve brake is always handy if a part should fail.

How does Uncle Sam propose to handle brake problems? First, all bicycles must have rear brakes, and some bikes must have front brakes as well. The dividing line is a gear of 84 inches. Bicycles with two brakes must develop 0.5g and those with rear brakes only 0.223g. Obviously, the government's idea is that if your gear is less than 84 inches you don't need to stop quickly.

Cyclists often ride in heavy traffic or down hills, or even race, on gears of less than 84 inches, and may need to stop quickly at any corner. The rear-brake-only bicycle cannot be as safe as two brakes or even the front-brake-only bike, and must be ridden with greater caution and foresight in traffic to compensate for its low stopping power.

Horsepower is entirely different. When you ride down a hill, your potential energy is directed three ways: 1) you speed up; 2) you move along the road; some of the energy fights wind and rolling resistance; 3) the rest of the energy, and on a steep hill this is nearly all of it, ends up as heat in the brakes.

The energy is measured in footpounds: the total altitude drop times the weight of you and your bike. The speed of your descent converts the energy into horsepower: Horsepower=(Pounds)x (Miles per hour)x(grade)x(0.00266). Since no brake ever built can store the energy of more than a low hill, the question is: how hot does the brake have to get to dissipate the horsepower heat into the air as fast as it is produced? Does that heat change the operation of the brake or permanently damage it? Coaster brakes with their small surface area have to run much hotter than rims, given the same downhill road.

Brake blocks, says the government standard, shall be tested by baking them for 30 minutes at 250 deg F, and they shall neither blister nor melt. Blistering is unimportant-it makes no difference-but melting? Would you like your blocks to melt on a hill? So the idea seems reasonable. According to government calculations (not experiments, but computations) this safely exceeds the condition of descending 7,950 feet in 30 minutes-quite a ride. That's 15 mph on a 20% grade, or a descent rate of 264 feet per minute.

Of course, the test is ridiculous on other grounds; brake blocks are made of rubber that is vulcanized at temperatures over 250 F, so they don't melt at 250 F. Most of the cheap, melting plastics are too slippery for brake blocks.

I decided to test both rim brakes and coaster brakes under the same conditions to see what would happen, and to check the government's calculations. Being unable to reach a 7,950-foot drop I could descend at 264 FPM as specified. I used a local hill of 2040 feet drop, and the fastest I could descend was 211 FPM. (That's a 4.0 mile run, 9.6% average grade, run in 9.7 minutes at 24.8 mph.) It was quite hairy enough, but I was able to control myself and not use the front brake for more than one second twice.

My weight (170 pounds complete) developed 1.09HP. At a reasonable total weight of 210 pounds that would have been 1.36HP, and the government conditions represent 1.67HP.

I first instrumented my rims and the coaster brake hub to measure how hot they ran. When descending the hill I kept alert for changes in brake operation. Afterwards I examined the brake parts to see what had happened.

With two rim brakes, the front rim exceeded 175 F but did not reach 200 F (rise above air temperature over 105 F but under 130 F) while the rear rim did not reach 150 F.

Using only the rear rim brake, the rim again did not reach 200 F. As an experienced cyclist would expect, if you have two rim brakes the front does most of the work unless the cyclist deliberately compensates by using the rear brake harder.

There was no appreciable wear on brake blocks, although there were rubber smears on the rims. I noticed no changes in brake effectiveness during the run. From comparison with the later coaster brake test it is almost certain that the rims reached peak temperatures early, and would not get any hotter if the run were longer.

The government formula is approximately right. Corrected for my conditions of horsepower (but not for cooling air velocity), it predicts 160 F as the rim temperature. And, of course, I used the brakes much harder on the steeper bits and before the corners than on the straights, so peak temperatures should be considerably higher than the average.

The coaster brake was destroyed in one run. It started smoking a short distance down the hill. At 700 feet down the smoke was streaming behind. Several times during the run its effectiveness changed, sometimes more, sometimes less. About 1500 feet down the brake refused to release fully, and it dragged for the remainder of the run. At the bottom I found my heat-measuring instrumentation had been burned off and the chrome plate was a white powder. On examining the inside afterwards I had to pry apart the stack of brake discs because the steel discs had softened and jammed onto the stationary mandrel.

The bronze discs that engage the hub shell had softened also and their engaging lugs were half torn off. If those had gone I would have had no brake at all.

No grease was left anywhere except a thin film inside the sprocket cone bearing-the rest was either melted out or burned into a hard carbonized coating over the other parts. Laboratory examination of the used parts and comparison with new parts showed that the outside bearing balls had reached 600 F, the outer brake discs 800-900 F, the inner brake discs 900 F.

There is every indication that the coaster brake was still getting hotter at the end of the run. In other words, even at that temperature it was not able to dissipate the heat produced by 1.09HP. Another 500 feet or so and the brake lugs would have been all torn off, resulting either in complete freewheeling or complete lock-up, depending where the pieces went.

So the government-and we hope they've changed their minds-planned to accept bicycles with the best brakes in the world and those with the worst. They wanted to force this requirement with an almost impossibly severe test of the best brakes (try to find an 8000-foot descent you can get down in 30 minutes!); while accepting untested the brakes which were destroyed 17% of the way through the test.

How about the beginners and those using bicycles bought for them by unknowing parents-the groups the government is trying to protect. They are the ones who believe that any bicycles the government passes is safe.

There is only one safe system: brakes that will produce high deceleration and absorb high power. Practically all rim brake systems meet these requirements hands down, and some hub brake systems probably do. No coaster brake alone meets either requirement, and the combination of coaster and front rim brake presents control problems at least as bad as the front brake alone.

Return to: John Forester's Home Page                                          Up: Bicycle Engineering