Top: John S. Allen's home page
Up: Expert witness


2001-2003, John S. Allen

Bicycle component designs are proven reasonably safe -- or not --  through engineering analysis and testing -- or else  through failure in use. It is disturbing to discover components which look as though they should be reliable, but which have serious engineering flaws. Certain handlebar stems provide examples of this problem.

The forward extension of the traditional aluminum alloy handlebar stem is a solid shaft, with a transverse, cylindrical opening at the front for the handlebars. To allow the handlebars to be clamped securely, the handlebar opening is split at the bottom or at the front, where it is subject to relatively light loading. A bolt pulls the two sides of the split together to clamp the handlebars.

The forward extension of the flawed stems is split behind the handlebars. To the untrained eye, a split shaft would seem to be as strong as a solid one, but it is not. Consider the example (photo below) from a French-made child's bicycle. This stem does not need a separate bolt to clamp the handlebars in place, because the bolt that secures the frame to the fork tightens the stem around the handlebars as well.

It might seem that the top and bottom halves of the split forward extension of the stem are both securely connected to the quill (vertical part) of the stem -- after all, the lower half of the forward extension is continuous with the quill, and the upper half is connected through a rigid steel bolt.

Split handlebar stem extension on a child's bicycle

pivo stem.jpg (24543 bytes)

But rigidity is a matter of degree. Though the steel bolt might seem rigid in your hand, it is far less rigid than the larger-diameter aluminum tube into which it is inserted. Also, the bolt is firmly secured only at its lower end of the quill where it is threaded into the expander plug. Everywhere else, it is loose inside an oversize hole. As a result, the upper half of the split stem extension is comparatively free to move around. Only the lower half of the extension contributes significant strength and rigidity.

The second photographic example is of a stem which failed, causing serious injuries to the cyclist.

Stem that failed

3T stem and bars.jpg (24831 bytes)

This Modolo stem was split behind the handlebar opening, but the handlebar clamp bolt was separate from the expander bolt. The lower part of the split stem extension carried the full load, and was further severely weakened by the large, counterbored hole for the handlebar clamp bolt. The stem failed at the bolt hole. Progress to failure was speeded by the use of triathlon handlebars, whose long forward reach places greater than usual bending loads on the stem.

The photo below shows the fracture surfaces. The scalloped pattern of progressive cracking, indicative of fatigue failure, is clearly visible near the top middle of the picture, to the right of the bolt.. The stem remained rigid enough that the cyclist did not become aware of the problem until the catastrophic failure ensued. Similar Modolo stems are, incredibly, still being sold in 2006, 15 years after the crash, with model names such as X-tenos and X-serta.

Fracture surfaces of failed  stem

3T stem.jpg (16449 bytes)

The next example is of a stem whose forward extension has a split that twists from vertical, at the back, to horizontal, at the front. This design cleverly allows the pivoted halves of the extension to spread apart, so the handlebars can be inserted without removing brake levers or handlebar tape -- and also, when the two split halves are pivoted along with one another, they allow the angle of the extension to be adjusted. But with this stem, as with the others, the strength of the forward extension is seriously compromised. Even though both halves of the split extension are connected to the quill with equal rigidity, the stem is much weaker then one with a solid forward extension, because bending strength goes up as the square of the depth of a cross-section. In addition, a hole has been drilled behind the handlebars for the clamp bolt.  The bolt hole is the likely location for failure.

Stem with twisted split

killrstm.jpg (29596 bytes)

Any split handlebar stem extension is prone to early failure, unless made bulkier and heavier to compensate for its inherent weakness. Light weight is desirable in bicycle components, and so a split extension is inherently an inferior design. The stems shown were not bulked up, but instead were made to similar dimensions as other stems.

The attachment of the stem shown below to the handlebars is not quite as bad, though still questionable -- because the clamp is at approximately a 45-degree angle to the stem extension, the stem extension is not compromised as much by the bolt hole.

Stem with diagonal clamp

DSCF0008stem.jpg (24958 bytes)

The photo below is of the same stem. I am concerned by the angle adjustment. Notice the bolt, with its bolt hole, underneath the extension.

Amusingly, the stem is adjustable, but the label on it says that it should not be adjusted. This is the kind of thing that happens when lawyers get involved in bicycle design..

Stem weakened for elevation adjustment

DSCF0013stem.jpg (23137 bytes)

Below, as a counterexample, is a stem which largely avoids the hazard, even though it allows the handlebars to be removed and replaced without removing levers and handgrip. The rear half of the clamp is supported by the entire cross-section of the stem’s forward extension, and transmits the load from the handlebars directly rather than through the bolts.

Non-split stem

020_20stem.jpg (17494 bytes)

Also, this stem provides another good example of evolution by imitation.  This stem is of a novel, weight-saving design without a quill. Instead, the stem extension is clamped around the top of the steerer tube (at the left side of the photo). This design is indeed different from what has been traditional for many decades. But, and this is important, the clamp arrangement is of the same kind that has been used for many decades to attach the handlebars to the stem. This is a use of a tried and true design in a new application rather than use of an untried design.

Top: John S. Allen's home page
Up: Expert witness

Contents 2001-2003, John S. Allen
Last modified 17 August 2006