Many people have used this test to draw conclusions regarding regarding suitability of various materials for bicycle frames. In particular, since aluminum frames faired better in the tests than steel frames, this has led many to believe that aluminum is an all-around better frame material.
Unfortunately, the testing method does not fairly represent real-world riding stresses on frames, and the conclusions ignore some very basic information.
The test fixture is designed to mimic riding out of the saddle, which is generally believed to apply the most stress to a bike frame. The test applies 268 pounds of force to the cranks for 100,000 cycles, and then 291 pounds for another 100,000 cycles. The force is applied to the pedals at an angle that should be consistent with out-of-the-saddle riding.
Let's suppose that this is a valid test. Their test fixture caused failures in nine of fourteen frames tested. How could this be? At a cadence of 60 RPMs, 200,000 cycles is 55 hours of riding. The first frame failed in under 57,000 cycles -- a good rider at 90 RPMs would reach this point in about ten and a half hours of riding (out of the saddle). In real world terms, riders can expect their brand new very expensive bikes to fail during his first couple of weeks riding in the mountains. Right away we should know that something is wrong with this test.
The actual measure for the endurance limit is in stress, not force. Stress is force per square inch. This means that if something is stressed above its endurance limit, you can always make it last longer just by making your metal thicker. You may even make it strong enough that you drive the stress below the endurance limit, and then it will never fail (under normal usage).
For aluminum though, the endurance limit is always asumed to be zero (if it is non-zero, it is too small to be useful). You can never get below the endurance limit, so eventually, every aluminum bike frame will fail. The reason consumers accept this is that (hopefully) aluminum frames are made with the metal thick enough that the fatigue failure doesn't occur until peak cycles get up into the hundreds of millions, or higher.
For steel, however, the endurance limit is comfortably distant from zero. Practical steel supports can be designed that should never fail during normal operation. The endurance limit for steel typically comes into play on the order of one million cycles; in other words if something has survived a million cycles of stress, it should never fail (due to that level of stress). For a bicycle at a rather slow cadence of 60 RPMs, this is less than 280 hours of riding.
There are certainly many riders capable of accumulating this many hours of out-of-the-saddle riding time in a single year, and a few who could get in a million cycles (out of saddle) in just a couple of months. And most of these riders have frames which have lasted them much longer than this. This is simply not the time scale in which frames should be failing. From this I conclude that steel-framed bikes are generally designed with all normal usage falling below the endurance limit.
One of the things this means is that if you do have a steel frame failure, it is due to one of three things: You exceeded normal usage (e.g. crashed), or the frame had a manufacturing defect, or the frame was poorly designed. Note that crashing may not cause immediate failure, but it can lead to a failure later on. For example, a crack in a frame will redistribute stress to the areas around the crack. These areas will see significantly more stress than they saw before there was a crack, and hence may be prone to an endurance limit failure.
The point to all of this is that they managed to break most frames (including all of the steel frames), in far less than a million cycles. This means that somehow, they've subjected these frames to stresses far in excess of normal riding limits. This puts any conclusions about frame materials in doubt. Steel frames failed in the test that may never have failed in real life. Aluminum frames lasted longer in the test, possibly because they have to be stronger than steel frames in order to delay their fatigue failure.
The linkage from
the pedal to rear wheel is approximated by connecting an extension
from the pedal directly to the rear dropout. This applies a force,
Fb as shown in the diagram, directly to the rear dropout. This force exactly
matches the "chain tension" being modelled.
On a real bike, the
chain pulls on the rear wheel, which applies force to the rear dropout
only by levering against the frame from the ground.
This can be modelled as a simple lever arm, shown as the red line
in the diagram. The forces, Fc, Fb, and Fg, are the forces which this
lever applies to the world around it (yes, this is backwards from traditional
free-body diagrams, but I think it is clearer for those with less mechancal
engineering backround, like myself). Fc is the reaction force that the
freewheel cog applies to the chain, as a result of the chain tension.
Fb is the force being applied to the bike at the rear dropout, and Fg is
the force being applied to the ground.
We can calculate the relationship between Fc and Fb, by considering the
torques applied by our red line, around the road contact point. Torques
must balance, so Fc times its distance from the ground must equal
Fb times its distance from the ground. Fb must be a stronger force,
in order to compensate for its shorter torque arm. The difference depends
on the actual sizes. If we assume a 13 inch radius tire, and a big
hill-climbing cog, with a radius of 2.5 inches, then Fc*15.5 must
equal Fb*13, therefore, Fb=1.19*Fc. In other words, the force applied
to the bike is almost 20% higher than that found in the chain.
The EFBe test actually underestimates the force applied to the dropout by anywhere from five to twenty percent (depending on the size of the rear cog being simulated). During climbing out of the saddle, the rider is more likely on a larger cog, hence the underestimate is likely closer to twenty percent. Clearly, I'm on the wrong track here.
One area for further analysis is a three-dimensional accounting of all of the
forces. You can get a better view of the linkages in these pictures.
I haven't analyzed these in three dimentions, however it does seem entirely
possible that there will be significant differences in resulting forces,
and especially in torques, because of the differences in where the force
is applied.
It's entirely possible that their linkage can introduce
forces which can't occur in the real world.
Okay, so the linkage to the rear dropouts may still be a problem, but to the extent that I'm able to analyze it, it is actually too gentle. What about the assumptions of force? Remember, 268 pounds of force, and later 291 pounds. A top-performing athlete could probably not achieve these forces for any length of time, because the rider will probably weigh less than 150 pounds. Such a rider would have to fully unweight the opposite pedal, and add an addtional 118 (or more, if lighter) pounds of upward force on the handlebar, for long periods of time.
A heavier rider could more easily come close to this, although a heavier rider is usually not going to be fit enough and skilled enough to fully or mostly unweight the opposite pedal while climbing out of the saddle.
Still, there are a few riders who are heavy, and fit, and have good technique. So I won't agree that these forces are typical, but I will allow that they may be attainable by a some small minority of riders.
Still not a slam dunk. It certainly doesn't seem like this should be enough to account for such a high failure rate.
The big clue to this problem lies with noting that two thirds of the failed frames failed in the head tube or down tube, even with the force being applied to the bottom bracket and rear dropouts. How can this be? A close examination of their test setup reveals what they did wrong: the front fork is mounted into a rigid fixture.
This introduces forces into the frame which will never be seen on the road. A bike can never support front-to-back forces at the front fork dropouts, except under braking or in a collision. Of course, braking is a tiny percentage of riding time, and cyclists generally don't continue to pedal out of the saddle while they brake.
The real-life front forks also only support minimal side-to-side forces, because the wheel rolls away from any force applied there. The test fixture can introduce significant side-to-side forces at the front fork.
It is however a realistic frame test for someone who is going to bolt their frame into a trainer, and then ride hard on the trainer out of the saddle for long periods of time. But I think most serious cyclists who use trainers with fork mounts ought to know that they aren't supposed to ride them out of the saddle.
It seems likely that this test setup is going to dramatically exceed design limits around the head tube, which is exactly where most of the failures occured. And the additional forces introduced at the front fork would probably have effects throughout the frame, although I haven't done any sort of analysis to see exactly what these forces might affect.
This is all particularly frustrating, because they came very close to doing something really useful. It would be very easy to fix their testing: redesign the front fork mount; redesign the linkage to the rear dropout; lower the input forces somewhat; increase the cycles to a million, before increasing the force.
Instead, all we get is more misinformation, and EFBe gets a nice profit for it all, too.
| My Home | Professional Home | TomSaraZac Home | Work Email |