Disclaimer

Basis Systeme netzwerk (BSn) makes no representation or warranty, express or implied with respect to this publication or the programs or information described in this publication. In no event shall Edward C. Zimmermann, BSn, it employees or contractors be liable for specific, indirect, or consequential damages.

Trademarks other than those owned by BSn used throughout this article belong to the owners and are used here in a purely editorial fashion.

The selection of crank length is a theme that more resembles religion than bio-mechanics. While it is commonly accepted that too long cranks can create undue stress on the knee the issue of optimum length is little, if not at all, understood. Although the "170mm cranks fits-all" viewpoint, dominant though the 50s and into the 60s, is no longer widely accepted, what little empirical research has been conducted has delivered inconclusive and often contradictory results: some leading to very short (130 mm) and others to very long cranks. The empirical research in this area is far from conclusive and the literature is very thin...

Under http://www.bsn.com/Cycling/ergobike.html there is also a section on crank-lengths with a few of the models out there and some of the issues. More important than crank length is position, style and comfort. Despite a common view, people DO experiment (and have over many decades) with crank lengths. Lennard Zinn in his series of articles for VeloNews is just one of many "amateur scientists" (in the great cycling tradition) that have tried to approach the subject and he, as many others, have found little more than guess-work. The newest product to address the issue of crank length is a pedal axle (Trimatic) that lets one change the effective crank length-- even during use.

Trimatic System Description: Replacement pedal axle for pedals from Look, Time, Shimano (SPD) featuring an excentric mechanism to allow for the variation of effective crank length. Three (3) variants are available in ±2.5mm, ±5.0mm, ±7.5mm in 3, resp. 5, increments. Available also as a complete pedal system (eg. Campagnolo Record pedal, 642 DM). Vendor: Trimatic-Radtechnik (Germany) Tel: +49 (03327) 497-48 Fax: +49 (03327) 497-50

While the Trimatic System appears to be interesting, its utility, even for experimentation, is at best limited. The problem with using something like SRM cranks with these "pedals" to gather some empirical data is the "system" hystersis: it takes a significant time to get accustomed to the different crank lengths. In the book Hinault, B and Genzling, C. (1986), Cyclisme sur Route, Editions Robert Laffont, Paris, Genzling quotes Jacques Anquetil (176 cm height, 84 cm inseam) of needing 1 year for the (successfull) transition from 175mm to 177.5mm! Even if the adjustment time was only 1 month, the period, given training curves and condition, would apriori preclude any meaningfull tests. Different cranks lengths probably require slight modifications to position as well-- another *very* complex theme..

A few of the main points:

1. Gonzalez and Hull, for instance, found in their Journal of Biomechanics articles that for a rider of average anthropometry the cost function global minimum occurred with 140mm cranks and also stipulated a qualitative anthropometic correlation: tall people at longer crank arm length and lower cadences than short people.
2. Proportional models have been suggested by some, viewing crank length from the perspective of total angle of movement. The problem with these first order models-- magic formulas such as 28.4% of C (Hinault's measurement from knee to buttocks), 20.5% of inseam, 10% of height or the Palm factor of 21.5% of inseam--- is that they don't consider the distribution of force during a stroke, muscle usage, bio-mechanical stride preferences and, especially, the knee stress at the top of the stoke just when power is applied--- which via the lower, forward shifted seat to accomodate the longer cranks is acentuated. While some of these models might seem reasonable they all lead to exaggerated crank lengths for the longed limbed.
3. Looking at other sports, such as running: In contrast to cycling, at any given running speed, a runner can optimize stride length and step rate to minimize energy requirements. While one finds that longer legged runners often have longer natural strides, some have had exceptionally short. The tradeoff is between stride lengths and step rates. A direct linear correlation between limb segments and stride among elite runners does not seem to hold. There is, at best, the tendency to shorten stride length and increase step rate on climbing and to lengthen stride and decrease rate on decents. This seems, however, to be explainable via weight shifts since in running, lacking support, the foot must be placed under the center of gravity. It is of some interest to note that women tend to have a slighly shorter stides and higher average step cadences than men.
   While the relevance of running data to cycling is limited-- in running the speed is the product of step AND candence, in cycling its determined by cadence (and gearing) alone-- the lack of a linear correlation, however, does seem to substantiate the claim that a correlation of limb segment length to crank length is not possible.
4. While longer cranks increase leverage, cycling is not simply the maximal application of force. Efficient cycling and the selection of crank length is based upon other factors, most importantly the personal ability of the rider to maintain a smooth circular (repeating pattern of limb segment motion and force application) pedaling action at an optimal cadence. An increase of 2.5mm in crank length increases the distance of limb segment motion traveled in a cycle by 15.7 mm (5 mm increased step), whence a reduction in cadence. While increased length increases, through its larger range of movement, hip and knee joint excursion, it also reduces the mechanical pressure (or Effective Force) applied within a stroke (constant torque). According to Marsh in Cycling Science Summer 1996, What Determines The Optimal Cadence, a goal to reduce average pedal force per revolution also seems to account, via the linkage between cadence and muscle fiber recruitment, for higher cadence preferences. So by inference both shorter cranks and higher rpms and longer cranks with cadences near maximal gross efficiency efficiency seem both, although mutually exclusive, to be warranted-- explaining the often contradictory results of some tests.
5. The focus of competitive road cycling is on sustained, prolonged and changing rather than maximal (anaerobic) power. While longer cranks reduce force, they, on the other hand, reduce cadence and can tend to increase pounding over a fluid stroke and thus deteriorate biomechanical efficiency over the range of activities. Quick changes in speed typical of strategic cycling are also more difficult with long cranks. In this light, crank length depends upon not just skeletal length but a complex combination of personal (genetic) muscle fiber traits, (more trainable aerobic) condition and application.
6. Since a long distant tourist tends to have lower cadence preferences (60 RPM) and higher effective gearing (weight load), the longer cranks appear beneficial from the the perspective of effective force-- substantiated by preference revelation¹⁾ in the general trend of equipping these bicycles with longer cranks. [ed: Oval, biopace® and MaxiSport® chainrings seem to support the use of long crank arm lengths.]. For road racing it is less clear cut.
7. For time-trials and mountain stages is it not uncommon for elite cyclists to adopt longer cranks. Indurian, for example used 180 mm cranks in time trials instead of his 175 mm. Jacques Anqueteil, according to Genzling, used, with his 84 cm inseam, 175 mm cranks (and 175mm in TTs). Eddy Merckx with his longer femurs (and 91 inseam) too used 175s-- and often 177.5 mm in TTs and mountain stages. Bernard Hinault (long femured, 83 cm inseam), on the other hand, did not change cranks, using 172.5 mm cranks for all stages. Many shorter cyclists have been known, however, to use 180s. Marc Madiot used, despite 86 cm inseam, 180 mm cranks all year round.
8. The effective crank length is a factor, however, not just of the measured crank length but of the length of the foot segment. The longer the length of the foot from heel to ball, the longer the effective length and also (commonly) the higher the saddle. With size 47.0pt feet, the factor on saddle height is typ. approx. 3.7 cm.-- a single larger or smaller size (in pt.) contributing about 1 mm of difference. The impact of foot length, however, depends upon stoke style and the position of the foot during a stroke. This is often linked to, and dictated, by other position characteristics. The horizontal saddle position determines the stroke force distribution within a cycle. Cyclists with longer femurs tend to push the saddle back and this too increases the effective crank length-- and given the general correlation between foot and femur lengths, the same crank arm length is effectively longer (at relevant point of force application) for larger cyclists. A longer crank often means pushing the saddle forward, reducing the effective length.
9. While a larger rider, in particular one with longer femurs, will envitably be better suited to a longer crank, an increase in crank length on road bikes also increases the total movement of the knee to chest on a rider with an absolutely smaller relative frame. This might increase the tendency to "ankle".

   While it has been suggested that "ankling" (having the heel lowered when at the top center of the stroke and in a raised position at the bottom) would make it "easier to apply force", little empirical substantition has been found. Many elite competitive cyclists tend, by contrast, to raise their heels at the top of the stroke and hold it nearly level at the bottom--- Eddy Merckx even lowered his heel at the bottom of the stroke.

   See also Film (MPEG) 164k bytes.

10. Since the seat height is not limited by the high but by the low stroke (leg extension, upper/lower leg angle) a longer crank requires a lower seat position. On the other hand, to decrease the range of femur motion one wants to raise the saddle. One is constrained here by geometry.
11. Empirical testing is complicated by the time needed to adjust to a crank length. Even a "magical" optimal crank will, if different from ones current length, will require some time for the establishment of a new position. The body will require some time to adapt to the change in muscle use, and the brain will need adjust to a new set of experiences and controls for the determination and selection of cadence. This hystersis apriori precludes any possibility for a simple laboratory test.
12. One should also not under-estimate the impact of imagery: cyclists told that they have longer, resp. shorter, cranks often adopt lower, resp. higher, cadences.

The main reason that the 170mm crank length lasted so long was not due to lack of experimentation but because it does seem to provide a reasonable length. While longer femurs will be better suited to longer cranks these same individuals tend to also have longer feet, also contributing, with the typical raise heel style, to a longer effective crank length. A cyclist with 91 cm inseam and size 47-48 pt shoes will have, with the same foot position, an increase of around 7-8 mm over a 81 cm inseamed cyclist with size 40pt shoes-- the 175mm cranks of the former build a lever that is nearly 1 cm longer than the 172.5 mm cranks of the later.

Despite the above recomendations, many small cyclists have had great success with cranks larger than the sub-170 mm. in the table. These are, however, mainly "mountain fleas" where the usual suggestion to use 2.5mm longer cranks brings them up to 170mm. While tall cyclists, on the other hand, have large femurs to help them keep a round stoke up the hills they will usually be best at sprinting and given their smaller absolute frames relative to their size, to reduce hip excursion, they will often want to select slightly smaller cranks. Going to 180mm (or longer) cranks, should one be able to effectively use them, requires raising the bottom bracket, possibly a steeper seat-tube and longer top-tube and/or stem--- which is why this works best in time-trials using a variant of the Position Americain. The downsides on handling of shifting weight to the front combined with less bottom bracket drop (what the author has called a "cross between a tri and a cross bike") could well, in road competition, offset any depreciation in efficiency of using shorter cranks. So this brings us to:

From a pragmatic perspective, and in line with "shared experience", the longest cranks (170-180.0) that one can maintain a fluid "round" stroke is postulated as optimal. The assumption is that if different lengths (from these) made sense then over the history of cycling one would envitably have witnessed a trend towards them.

If you look at elite cyclists, despite an increasing number of long femured (tall) competitors, you will see that 175mm is pretty much the longest crank, other than for time trials, that is used. I would not, as has often been suggested put this to conservatism. On the contrary, the fierce competition and prizes of professional cycling are substantial and bring a willingness to quickly adopt what might appear to provide some form of advantage--- which more often than not is at best psychological. In a realm where high tech voodoo and strange science abound I can't accept the arguments to the contrary...

Questions? Post them to http://forums.bsn.com/VeloTech.html (BSn's Velo Technology Discussion Forum).

1. Burke, Edmund (1986). Effects of Saddle Height and Pedaling Cadence on Power Output and Efficiency. Science of Cycling. Human Kinetics Publishers, Champaign IL.
2. Burke, Edmund ed. (1996). High-Tech Cycling, Human Kinetics Human Kinetics Publishers, Champaign IL.
3. Cavanagh, P. R. and Sanderson, D. J. (1986). The biomechanics of cycling: Studies of the pedaling mechanics of elite pursuit riders. Science of Cycling. Human Kinetics Publishers, Champaign IL.
4. Hinault, B and Genzling, C. (1986), Cyclisme sur Route, Editions Robert Laffont, Paris
5. Hull, M. L. and Gonzalez, H. (1988) Bi-variate optimization of pedaling rate and crank arm length in cycling. 21:839-849.
6. Hull, M. L. and Gonzalez, H. (1989) Multivariable optimization of cycling biomechanics. J. Biomech. 22: 1151-1161
7. Hull, M. L., Gonzalez, H., and Redfield, R. (1988). Optimization of pedaling rate in cycling using a muscle stress-based objective function. Int. J. Sport Biomech. 4, 1-21.
8. Marsh, A. P. (1996). What Determines The Optimal Cadence?, Cycling Science, Summer 1996
9. Morris, D. M,, Londeree, B. R. (1997), The Effects of Bicycle Crank Arm Length on Oxygen Consumption. Canadian Journal of Applied Physiology v 22 n 5 p429.
10. Redfield, R. and Hull, M. L. (1986). On the relation between joint moments and pedalling rates at constant power in bicycling. J. Biomech. 19: 317-329.
11. Too, D. (1996). The effect of pedal crankarm length on joint kinematics and power production in upright cycle ergometry. The Research Quarterly for Exercise and Sport, 67 (1)(supplement), A22. (Abstract)
12. Yoshihuku, Yasuo et al. (1990). Optimal design parameters of the bicycle-rider system for maximal muscle power output. Journal of Biomechanics v 23 n 10 p1069-1079.
13. Yoshihuku, Yasuo et al. (1989). Maximal muscle power output in bicycling as a function of rider position, rate of pedaling and definition of muscle length. Abstracts of the XII Congress, International Society of Biomechanics, Los Angeles, CA, USA. Journal of Biomechanics v 22 n 10 p1104.