• Increase font size
  • Default font size
  • Decrease font size
Home Reviews Stems Stem Test

Stem Test

E-mail Print PDF
Article Index
Stem Test
Ratings Development
All Pages

Stem Evaluation Methods 

Mass, stiffness, durability, and price were measured for each stem tested.


Every component on the stems was measured using the same digital scale.  The principle components of the asembly were the faceplate, faceplate screws, extension body, steer clamp screws, and the steer tube shim.


Stiffness is defined as the load to deflection ratio.    The loading condition was a combined mode, meaning that a +/- horizontal/vertical load was applied concurrent with a small bending moment (load was applied 85 mm off of the handlebar clamp centerline).  This combined mode is consistent with how stems are loaded during any out of the saddle effort such as climbing or sprinting.

The load was varied using an air cylinder in 5 psi increments as adjusted through an analog air pressure regulator (model).  Using specifications provided by the air cylinder manufacturer, the air pressure values could be converted to a load.  Deflections were measured along the same axis as the load was applied using a dial indicator.

At each stage of the durability testing, static stiffness measurements were made.  The many data points taken allows for a statistically better measurement as random errors associated with manually adjusting air pressure or reading the dial indicator were minimized.

Although the combined loading produces a rotational/torsional component of displacement, only the translational movement along the axis of the applied load was measured.  Consequently, the stiffness measurements that are reported are not entirely representative of the physical phenomena at work – it is a simplification that is extremely repeatable, however.

The raw data was averaged, plotted two-dimensionally, and then linearly regressed (a straight line was fit to the data).  The resulting slope of the load-deflection curve represented the stiffness of the stem specimen. 


Using the same two loading configurations in the static stiffness tests described above, the loads were cycled in the positive and negative directions according to the following schedule: 

Extreme Durability

In order to ensure that all stems were tested to failure, the following schedule was followed after the initial durability testing phase was completed.

 At the heart of this test is a fixture that uses some creative engineering to accomplish the task.  Below is a schematic of the fixture setup, which illustrates its key features. 

Air compressor

The compressor is the engine of the system.  An electric motor drives a piston, which charges the main cylinder up to 125 psi.  This reservoir then provides the necessary compressed air to supply the air cylinder on the test fixture.  Analog air regulators control the individual air pressure to each of the solenoid valves.

PC and parallel port control box

A friend of the cause designed the parallel port control system that allows the load cylinder to be cycled in both the positive and negative directions.  The control box houses the relays that drive the solenoids.  The relays interface with an old school IBM thinkpad laptop computer.  A simple program written in qbasic toggles an electric pulse to the relays – when the relay sees this signal it closes the circuit and sends a larger electric signal to the solenoid.


The solenoids are simply valves that are electrically controlled.  They can be toggled to be either open or closed.  In the closed position, the air supply from the compressor is cut off from the load cylinder and the compressed air that was previously in the cylinder is vented out to the atmosphere.

Load Cylinder

This cylinder has a 3.5 inch piston diameter and has air inlets on opposite sides of the piston.  Depending on which side is energized with compressed air, a load is generated in either the positive or negative direction.  The cylinder has pinned boundary conditions on both of its attachment points which ensures that the load is applied along one axis and it also helps prevent cylinder damage due to bending loads.

Rating Methodology

The stem ratings were determined on three basic design variables - performance, price, and durability.


Performance of stems is primarily based on mass, with some subjective improvements to be had by increased stiffness.  It could be assumed that once a stem has an "acceptable" stiffness, any additional improvements will have diminished effect.  Furthermore, at this point in time the relative contribution to overall steering column stiffness of the stem is unknown.  It could be speculated that the bar itself is the primary contributor to the perceived amount of flex in the overall steering system due to its length - stiffness being proportional to length cubed.  However, the stem does provide the boundary condition, or "fixity-ness", for the handlebar and is therefore an important part to evaluate.

Based on these two variables of stiffness and mass and the preceding logic, a relative weighting scheme was determined to help quantify stem performance.  Overall stem performance was rated using a weighting of 80% for the importance of mass and a 20% weighting for the importance of stiffness.


The range of the average of horizontal/vertical stem stiffness values (with a little buffer above and below what has already been measured) was broken up over a scale of 1 to 10 as illustrated below:


The range of stem mass was also divided up on a scale of 1 to 10 as indicated below:


The price rating was determined in a similar manner as the core stiffness and mass rating values.  Reasonable limits for pricing were assigned to ratings of 1 and 10.  The values in between were determined on a linear basis.


A minimum value of 1 was assigned to a stem that survives what one might consider to be a maximal loading condition (approximate 200 lb horizontal load) for 25,000 cycles.  It should be remembered that since this is a comparative rating system, this value of 1 does not necessarily mean that the product should be considered unsafe.  The author has been riding a stem that ranks in the 3-4 range for nearly 100,000 kilometers.  The upper end of the durability scale was determined by the capacity of my compressor/air cylinder system which is in the 900 lb range.

Pdf data sheets

Control Tech  - test protocol development sample

ITM Big One

ITM Millennium

3T Forgie

Ritchey WCS

Performance Forte


Deda Newton (31.8 mm clamp)

ITM The Stem 

TTT Zepp XL (31.8mm clamp)

Thumbnails of failures

ITM Big One

Control Tech

Deda Newton (31.8 mm clamp)

TTT Forgie

Performance Forte


ITM Millennium

Ritchey WCS

ITM The Stem

TTT - Zepp XL (31.8 mm clamp)

Analysis and Summary

Bicycle stems are essentially nothing more than a cantilever beam.  They are fixed (not allowed to rotate or translate) at the steer tube and are free to deflect/move at the handlebar clamp.  Cantilever beam deflection is described by the following equation...


Bicycle stems are essentially nothing more than a cantilever beam.  They are fixed (not allowed to rotate or translate) at the steer tube and are free to deflect/move at the handlebar clamp.  Cantilever beam deflection is described by the following equation:

deflection =  PL3/(16EI)

Rearranging the terms it is possible to define stiffness as load divided by deflection (k):


For the purposes of this discussion, it can be assumed that the cross section of the stem extension is circular.  It can be shown, then, that the section inertia, I, is proportional to pi*(Ro-Ri)4 – where Ro is the outer extension radius and Ri- is the inner radius of the extension.  It should be clear that the most important factors in determining stem stiffness are the outer radius of the extension at the steer tube clamp, wall thickness, and the overall length of the stem.  As the length of the stem increases, the stiffness decreases; and as the radius of the extension increases (for a given wall thickness), the stiffness will also increase. 

To a lesser extent, the stiffness of the stem will depend on the material used (modulus of elasticity, E).  For the stems tested, the only sample that was made of a significantly different material was the ITM The Stem, which was constructed out of a magnesium alloy. Typically, magnesium alloys have a modulus that is 30% less than aluminum alloys.  In order for the lower modulus materials to have equivalent stiffness, they must have a larger extension radius, or a thicker wall.  If neither of these design variables are changed, the stem will lose stiffness.  The ITM stem did not significantly change its geometry, and as a result had a lower stiffness than the rest of the stems tested.

This previous stiffness equation will adequately describe the performance in the vertical test configuration. Looking at the measured stiffness and the predicted stiffness we can see the following relationship:

There is a very good relationship; most notably, the Forgie stands out as the stiffest.  Of course, it is the stiffest because it has the largest extension tube radius by nearly 20%.  Note that the oversized stems and the alternate material stem (magnesium) are omitted from the preceding plot.

It should also be clear that bar bolt pattern has no material affect on stiffness.  For example, it is a common belief that a four bolt pattern will be stiffer than a one or two-bolt pattern.  The TTT Forgie lays this notion to bed.  The Forgie uses a two bolt pattern, yet has the highest stiffness measured.   Of course, the high stiffness of the Forgie is a function of the large diameter of the extension tube.

The following table summarizes the stiffness of all the stems in both the horizontal and vertical direction.

 The average stiffness are accurately captured in the final stiffness ratings as well:

In summary, stem stiffness can be reasonably well predicted by the following variables:

-Length (shorter is stiffer)

-Extension tube diameter (larger is stiffer)

-Wall thickness (usually pretty constant, and thicker is stiffer)


The manufacturing community has gravitated to mass as a key selling point.  This is probably due to the fact that it is the most tangible concept that consumers can understand and even measure.  There is foundation to the argument and mass does affect ultimate cycling performance, but usually only in the very specific circumstance of climbing steep hills.  Mass is generally a second order affect (nearly ten times less significant than other performance variables) on overall cycling performance; however, it does play a role and was therefore, measured and compared.

The range of weights for all stems covered a mere 90 grams, with the ITM The Stem coming in at 120 grams and the Forgie at 208 grams.  It is interesting to note that most of the lightweight stems try to save weight in the hardware, mainly in the bolts.  Slight mass savings were had in the faceplates and the extension tubes.  The following graph shows the ratings of all the samples.


Two basic designs were used for the stems evaluated.  One style consisted of separate components welded together at the steer tube clamp and handlebar clamp areas.  The other design combined the clamps and the extension tube into a single continuous piece.  This single piece was generally made by either machining (cutting/removing) the shape out of a block of material, or forging (forming the material under temperature and pressure) into its final shape.

There were distinctly different failure modes and durability performance characteristics for these two construction techniques.  The welded stems failed sooner and at or near the weld beads.  This type of failure is to be expected, since welding does not guarantee full penetration/fusion of both components and also because no geometrical reinforcement is provided for when using constant thickness extension tubes.  On the other hand, the one piece body stems allow for a thickening of the walls at the juncture of the clamps and the extension tube – this is a good thing and is represented by the superior performance of this type of stem.

The one piece stems also failed in similar manners amongst themselves.  The primary cause of failure in the standard bar clamp size stems was due to stress concentrations.  The concentrations observed included thickness transition points and cutting tool/forming tool marks.  The best example of a failure at a stress concentration point can be seen on the failure surface of the ITM The Stem. 

Interestingly, the oversized stems performed the worst in the durability testing.  This is the opposite trend that has been observed with oversized steer tube forks.  In the case of forks, the larger steer tube contributes to a measurable improvement in their strength (assuming they have been manufactured well).  The stems, on the other hand, failed quickly – though it should be mentioned that the failure mode was not extension tube failure but instead hardware failure.

The TTT Zepp XL tested was identical to the style recalled by Cannondale late in 2002. This stem was not recalled by the manufacturer, though, and is still available from at least one national mail order company.  The reason for the Cannondale recall was that the thread engagement length for the faceplate bolts was insufficient.  A general rule of thumb is that the thread engagement length for a fastener should be at least 1.5 times the diameter of the bolt.  Thus, for the M5 bolt used on the Zepp XL, the engagement should be at least 7.5mm.  The engagement on the stem tested was only 5.9 mm.  The M5 steer tube clamp bolt had a measured 7.5mm engagement, and the other non-oversized stems all included hardware that abided by the 1.5 times the screw diameter engagement rule of thumb. 

The recall by Cannondale does seem to be justified, since this fastener stripped the threads in the stem extension body during a fixture changeover.  It was also measured that the minimum diameter of the bar bolt was 4.66 mm - this dimension is well under the specified minimum for a M5 bolt which is 4.826mm.  For these reasons, it is possible that under a very large load, the thread could have failed causing the bars to rotate causing a subsequent loss of control.  The durability testing stopped as a result of the thread failure.  It should be recommended that owners of this stem abide by the recommendations set forth in the CPSC recall and stop using the stem until proper length screws have been installed.

Another stem failed due to thread manufacturing reasons.  The Deda Newton stripped out the thread in the steer tube clamp region.  The problems with this fastener were also twofold.  First, the thread engagement for the M5 titanium bolt was only 5.75 mm instead of the recommended 7.5 mm.  Second, the CNC turned screw was not round and had a minimum diameter of 4.72mm – according to specifications a M5 bolt should have a minimum diameter of 4.826 mm.  It should be clear that the turned bolt was undersized.  The combination of an undersized bolt and insufficient thread engagement length led to the failure of this stem.

For all stems, the fasteners were tightened using a torque wrench set to the manufacturers recommended installation torque values.  All bolt threads were greased prior to installation.  The threads that failed did so at torque values less than the recommend installation values - this is undesirable and is primarily due to the reasons outlined above.

What is stiffnes even worth measuring?  Often times, stiffness correlates grossly with strength and maybe even fatigue strength.  The following graph shows that there is some mild correlation between horizontal stiffness and fatigue resistance.  Major outliers can be seen with the the oversized stems due to their early failure due to thread stripping.

Altogether it should be clear that for this testing protocol, the one piece forged/machined extension stems performed better than the welded stems.  The primary failure location for the welded stems was at or near the weld at the steer tube clamp.  The failure for the other stems was due to stress concentrators such as thickness transitions or tooling/machining marks.  The two oversized bar clamp stems failed due to improper thread engagement and undersized bolt thread diameters.

A final disclaimer must be made:  these durability results are based on a comparative test methodology that may or may not directly reflect actual usage.  It was the best test that could be done with the available hardware.  It is also the best look at stem durability that has been published in the United States to our knowledge.

The final durability ratings are shown below:


There is not a whole lot that can be said about pricing, other than manufacturers will charge what the market will bear - meaning that there is not a whole lot of objective reasons for pricing.  One might reasonably expect that price might correlate well with increased performance (higher price for lower mass).  This is not necessarily the case, since the lower price stems and the higher price stems have overlapping mass values.

The highest price stem, ITM The Stem, does claim the lowest mass by a scant 25 grams.  It also claims another feature that manufacturers like to promote - a new material.  The Stem was the only stem reviewed that was constructed out of a magnesium alloy.  The exclusivity of owning this product and its slightly lighter weight lead the manufacturer to demand the excessive price tag of approximately $200.  It is of note that the Performance Forte stem had similar durability results and a higher stiffness, but is available at a  fraction of the cost.  Is the extra $160 for a 60 gram weight savings worth it?  That decision is up to you, the consumer.

Most often, there is a distinct relationship between time on the market, and price.  This relationship is seen every year when last years “inferior” product is closed out at bargain prices in order to clear the shelves for the new and improved product of this year.  It is often times hard to resist the temptation to buy the new stuff, but great deals are often had by buying last years closeout models.


Evaluating products based on different weightings of the three primary design variables of price, performance, and durability allows consumers to choose what matters to them.  Curmudgeonly retro-grouches will praise the old-school designs that favor price and durability.  Racer types only care about performance.  So, peruse the data sheets and the summary graphs presented below and decide for yourself.  BikeTech Review provides the data and leaves the ultimate purchasing decision up to you.

Price/Durability Weighted Consumer Ratings (45% price/durability and 10% performance)

Performance Weighted (90% performance, 5% price/durability)

Even Weighted (33.33% price/performance/durability)

Last Updated on Sunday, 24 January 2010 01:07