Arm cranking, in particular, standing arm cranking, has become increasingly popular as a means of assessing upper limb performance (2,6,17). However, the optimal configurations for power production during arm cranking have not been determined. In cycling, the manipulation of joint angles, through changes in the structure of bicycle components, has been shown to influence performance (5,10,21). For example, changes in seat height and cycle crank lengths directly affect hip and knee joint angles, range of motion and angular velocity of the joints, and thus cycling performance (21). The optimal crank length for maximum power production has been reported to be 20% of leg length (10), and the optimal seat height appears to be 109% of inseam length (ischium to ground) (5). It seems highly likely therefore that changes to the configuration of arm-crank ergometry, specifically crank length and crank-axle height, could also affect performance. Given the angle-torque and torque-velocity relationships of human muscle function, there is a clear rationale for how interventions that affect upper extremity joint range of motion and angular velocities may influence arm-cranking performance.
Big-boat yacht racing is one of the only able-bodied sports where arm cranking is the primary physical activity. In the majority of professional big-boat yacht racing classes, maneuvers are performed manually, without the assistance of stored energy, and arm cranking ("grinding") is used to drive the winches, which in turn controls the sails and the mast (23). The average duration of grinding bouts during racing is approximately 6 s with an exercise:rest ratio of approximately 1:6 (13). International America's Cup Class version 5 yachts typically have four arm-crank stations ("grinding pedestals"), each manned by one or two athletes ("grinders"). The typical crank length and crank-axle height are 250 and approximately 850 mm, respectively, but there seems to be no scientific rationale for these settings. The height of the grinding crank-axle appears to have been largely determined by other aspects of yacht design, such as aerodynamics, related to boom and grinding pedestal heights, without an understanding or consideration of the effects of crank-axle height on "grinding" performance. Crank-axle height may influence posture, particularly hip flexion, as well as the relative weighting of the feet and the hands during grinding.
No studies have examined the effect of changes in crank length and crank-axle height on power production during standing arm-crank ergometry. The identification of optimal crank lengths and crank-axle heights would further our knowledge of standing arm-crank ergometry and may enhance the performance of America's Cup grinders. Therefore, the aim of this study was to assess the effects of different crank lengths and crank-axle heights on standing arm-cranking power and to determine the optimal crank length and crank-axle height for maximum power production. Ground reaction forces (GRF) beneath each foot and sagittal plane video were recorded to consider the biomechanical mechanisms for any changes in power with crank-axle height and crank length.
Nine elite professional male America's Cup yacht racing grinders (mean ± SEM age = 36 ± 2 yr) volunteered to participate in this study. Their physical characteristics are shown in Table 1. The athletes had all represented teams that competed in the 32nd America's Cup, with their collective experience including 28 America's Cup campaigns and 13 World Championship titles. Written informed consent was provided by all the athletes, and the study was approved by the Loughborough University Ethical Advisory Committee.
All anthropometric measurements were taken in accordance with the prescribed methods of the International Society for the Advancement of Kinanthropometry (9). Nude body mass was measured to the nearest 0.1 kg using a calibrated digital scale (Metler Toledo KcC 150, Leicester, UK), and height was measured to the nearest 0.001 m using a stadiometer (Seca 222, Hamburg, Germany). Skinfold thickness was measured in duplicate at seven sites (triceps, biceps, subscapular, abdominal, supraspinal, thigh, and medial calf) using calibrated skinfold calipers (Harpenden, Baty International, West Sussex, UK). Percentage body fat was calculated from the sum of seven skinfolds (7,18). Arm span was measured with the athlete standing back against a wall, heels together, and arms stretched out horizontally, and the distance between the tips of the furthest fingers on each hand was recorded (9).
All tests were conducted between 0900 and 1200 h. Anthropometric measurements were taken on arrival at the laboratory. After an initial 10-min self-paced warm-up, athletes performed eight maximum-effort sprints at predetermined combinations of crank length and crank-axle height in a randomized order. Because of the limited time available to test the athletes, not all possible combinations of crank length and crank-axle height were explored. The protocol included four variable crank lengths selected throughout the range available on the ergometer (162, 199, 236, and 273 mm) and encompassing 170 mm (standard leg-cycle and arm ergometer cranks) and 250 mm (grinding pedestal cranks) at a constant crank-axle height of 1050 mm (determined as the optima during a previous pilot study); and four variable crank-axle heights (850, 950, 1050, and 1150 mm) with a constant crank length of 250 mm (the standard length on America's Cup yachts). Each sprint was 6 s in duration with a 10-min rest interval between sprints to ensure complete recovery. A 5-s countdown was given before the start of each sprint during which time a crank velocity of approximately 50 rpm was maintained. Verbal encouragement was given throughout the test. Torque and angular velocity (crank velocity) were recorded throughout each sprint and analyzed off-line. Power was determined as the product of torque (T) in newton-meters and crank velocity (ω) expressed in radians per second.
All tests were conducted on an adjustable standing arm-crank ergometer (Technogym Top Excite, Gambettola, Italy), which was secured to the ground while remaining clear of the force plates (Fig. 1). The athletes were all familiar with the ergometer used in this study, which is commonly used in America's Cup training. The resistance software of the ergometer was upgraded (Technogym Excite version SW50.22.7) to provide increased resistance and set at level 30 (∼92 N according to the manufacturer's specifications). The crank handles were 0.52 m apart (mediolateral displacement), which was slightly wider than standard grinding pedestals (0.48 m). An SRM power crank (Schoberer Rad Messtechnik, Science Powermeter V, Jülich, Germany) was fitted to the center axle of the ergometer. Torque was recorded continuously at 200 Hz (SRM torque software) and averaged over 360°. Crank velocity was measured every revolution (SRM torque software). The SRM power crank was calibrated before each test protocol, according to the manufacturer's guidelines.
To assess differences in unilateral weight shifting due to the independent variables, GRF beneath each foot were measured using two calibrated force platforms (9253A2 (right) and 9281CA (left); Kistler Instrument AG, Winterthur, Switzerland) with a sampling rate of 200 Hz and 16-bit analogue to digital conversion (ADC). The arm-crank ergometer was positioned centrally over the two force platforms to allow the athletes to stand with one foot on each platform. The force platforms were calibrated according to the manufacturer's guidelines and zeroed before each sprint. The GRF were analyzed for each sprint over 5 s, beginning 1 s after the start of the sprint. For each force plate, the magnitude of the resultant GRF and its direction in the x-z plane (toward the ergometer) were calculated from the x, y, and z components. For both plates, the z-axis was vertical, the y-axis pointed perpendicular to a line joining the ergometer to the center of the two force plates, and the x-axis was perpendicular to z and y. The variation in force amplitude and direction was calculated as the SD of the resultant GRF and its direction.
Reflective markers were attached to the right side of each athlete at the following anatomical positions: iliac crest, greater trochanter and lateral epicondyle of the femur, and lateral malleolus of the fibula. All sprints were recorded by a video camera (NV-DS99EG mini DV; Panasonic, Tokyo, Japan) at 25 Hz, which was positioned perpendicular to the sagittal plane of the ergometer, 7 m from the athletes and at a fixed height of 1 m. Four 500-W lamps were projected onto the athlete to provide additional lighting. Hip joint angle between the iliac crest, the greater trochanter, and the knee, knee joint angle, and foot-to-floor angle were measured after 3 s of maximal grinding, on the right side of the body, when the right knee was at maximum extension. Determination of joint angles was performed using a digital software program (SiliconCOACH PRO, Dunedin, New Zealand).
Peak power, joint angles, resultant GRF, and GRF direction for the different crank lengths and crank-axle heights were compared with one-way repeated-measures ANOVA for each independent variable. Bonferroni post hoc tests were used to determine where any differences lay. Pearson product-moment correlation coefficients were calculated to assess bivariate relationships. Analyses were performed using SPSS for Windows (version 15.0; SPSS Inc., Chicago, IL). The group data describing the relationships between crank-axle height and crank length with peak power were fitted with quadratic (parabolic) relationships using SPSS, which generated theoretical optimum values for each parameter. Significance was defined as P ≤ 0.05, and all data are presented as mean ± SEM.
The athletes were characterized as having a high fat-free mass (Table 1). There was a significant difference in peak power between crank lengths (P = 0.006), with a lower peak power for 162 mm than all other crank lengths (P < 0.03; Table 2). When crank length was normalized for arm span, the relationship between maximum power and crank length (CL) was parabolic and fitted by the following equation: power = −11.127(CL)2 + 274.7(CL) − 361.2; r = 1.0 (Fig. 2). From this relationship, the highest theoretical peak power occurred at a crank length of 12.3% of arm span, which, in these grinders, equated to 241 ± 9 mm.
Peak power was significantly less for the crank-axle height of 850 mm compared with 1150 mm (P = 0.01; Table 2). Before normalizing crank-axle height to stature, two athletes were excluded from the data analysis because they exhibited significant differences in technique from all other athletes. This included substantial ankle plantarflexion, measured by foot-to-floor angle (53° ± 4° and 42° ± 5° for the excluded athletes vs 4° ± 2° for the remaining athletes; ANOVA, P < 0.001), thereby confounding their true stature. The maximum power and crank-axle height relationship revealed similar high values between 50% and 60% of stature (Fig. 3), which was 950-1150 mm in this cohort of grinders. When the data between maximum power and crank-axle height (CAH) were modeled with a parabolic relationship (power = −0.5251(CAH)2 + 60.096(CAH) - 382.76; r = 0.97), the theoretical optimum for crank-axle height for the group (n = 7) was 57.3% of stature. Hip joint angle during arm cranking was influenced by crank-axle height (P = 0.001; Fig. 4), with significantly greater hip flexion at 850 mm than at 1050 and 1150 mm (127° ± 3° vs 142° ± 3° and 146° ± 3°, respectively, P < 0.01, n = 7) but similar to 950 mm (136° ± 3°). Knee joint angle was unaffected by crank-axle height.
The mean resultant GRF was significantly different between crank-axle heights (P < 0.001; Table 3), with 850 mm less than all other heights (P < 0.01) and 950 mm less than 1050 and 1150 mm, respectively (P < 0.01), whereas the mean resultant GRF was not influenced by crank length. The mean GRF direction was unaffected by crank length or crank-axle height. When comparing differences between legs, resultant GRF was greater for the dominant leg than for the nondominant leg for crank lengths of 162, 236, and 273 mm and crank-axle height of 850 mm. When comparing the variability in amplitude and direction between legs (Table 4), the dominant leg had significantly greater amplitude at crank-axle heights of 850 and 950 mm and greater variability in direction at crank lengths of 162, 199, and 236 mm, respectively.
The main findings of this study were that changes in crank length and crank-axle height influenced performance during maximal standing arm-crank ergometry. The parabolic curves observed for peak power with increasing crank lengths and crank-axle heights can be attributed to interactions between torque, crank velocity, and posture. The design and the configuration of arm-crank ergometers and grinding pedestals should recognize the importance of these variables to performance, and we suggest an optimum crank length of 12%-12.5% of arm span and an optimum crank-axle height of 50%-60% of stature (∼241 and ∼1087 mm, respectively, in this cohort of grinders).
The peak power values reported in this study (range = 1203-1691 W; Table 2) are among the highest reported during arm cranking. Similar impressive peak power values have been reported previously in America's Cup grinders (15). To our knowledge, America's Cup grinders are the only athletes to have reported arm-crank peak power values greater than 1000 W. This is likely due to the unique nature of this cohort of athletes, who are selected and trained for the specific activity of standing arm cranking (12,15), and the use of more favorable arm-crank configurations than those that have often been used. When the data were normalized for the substantial stature of the athletes, the relative peak power (13.8 W·kg−1; range = 11.5 - 16.9 W·kg−1) was still greater than other trained upper body athletes (swimmers = 11.5 W·kg−1 , elite gymnasts = 10.6 W·kg−1 , elite wrestlers = 9.6 W·kg−1 , and javelin throwers = 8.5 W·kg−1 ). The reliability of peak power data during seated arm cranking in untrained individuals has been determined to have a coefficient of variation of <5% (19), and the athletes in this study were all highly trained and familiar with the ergometer used.
The selection of optimal arm-crank length for maximal power may be of interest to big-boat grinders and to studies using standing arm-crank ergometry as a measure of performance. Our data demonstrate that the optimal crank length for maximal power production was approximately 12.3% of arm span, which equated to approximately 241 mm (range = 228-252 mm) in this cohort of athletes (Fig. 2). This optimum crank length for the cohort was within 4% (9 mm) of the standard crank length used on America's Cup racing yachts (250 mm or 12.8% of arm span in this cohort). From the crank length-power relationship (Fig. 2), this equates to a reduction in power of <0.2% for the standard grinding pedestal crank length compared with the optimum we have found. Essentially, the standard crank length used in big-boat sailing facilitates very close to optimal power production for these elite grinders who have a wide arm span. Overall, the 68% variation in crank lengths used in this study elicited a 16% variation in maximum arm-cranking power, which is considerably greater than the variation reported in cycling (10). The optimal crank length in cycling has been determined to be 20% of leg length, which equates to approximately 170 mm for the average population (10). Standard leg cycling cranks (170-180 mm) have also been used extensively for arm-crank ergometry studies, with the pedals replaced by handles (2,19,20,22). This length of crank appears to be substantially less than the optimum we have found in this cohort and, on the basis of our data, would have resulted in a 13% reduction in power production, although individual differences in physique might attenuate the magnitude of this decrement. It is important to note that these findings relate specifically to standing arm cranking and may differ when seated.
The influence of crank length on performance is due to the complex interaction of force, torque, and velocity (4). Shorter crank lengths tend to decrease torque and elevate crank velocity (10). Therefore, grinding at different crank lengths involves changing the contractile conditions across the range of the power-velocity relationship. Because the power crank-velocity relationship is parabolic (3,10), it is not surprising therefore that the power crank-length relationship is also parabolic. Hence, peak performance (power) occurs at an optimum combination of torque and velocity.
To remove differences in anthropometry and to provide a normalized measurement, crank-axle height was calculated relative to stature. However, two athletes had an obviously different technique to the rest of the cohort, which included substantial plantarflexion of the ankle that changed their effective stature, and they were therefore excluded from the relative height data.
Contrary to that found in cycling (5), the parabolic curve for crank-axle height was a poor fit with no clear optimum and resembled more of a plateau. There was little difference in performance between crank-axle heights of 50%-60% of stature; however, peak power was significantly reduced when the crank-axle height was less than 50% of stature. These results indicate that the typical height of the grinding pedestals used on America's Cup yachts (∼850 mm) would reduce peak power by as much as 7% for the athletes tested in this study. It is unfortunate that the highest crank-axle height investigated in this study was 60.7% of stature because it would have been interesting to determine the effect of greater heights. It should be noted that it is possible that an interaction exists between crank-axle height and crank arm length and that some other combination of crank-axle height and crank arm length may have produced even higher (optimal) peak power outputs than observed in this study.
In cycling, seat-to-pedal height distance influences performance because of changes in joint angle at the hip and knee (5). Although it could be assumed that crank-axle height may have a similar effect on standing arm-crank performance, the joints with the greatest range of motion during arm cranking, the elbow and the shoulder, are probably not directly affected by the crank-axle height. Rather, it is the hip angle that had the greatest change according to the crank-axle height. The increased hip flexion at lower crank-axle heights resulted in an increased portion of the athlete's body mass being supported by the ergometer, as shown by the decreased resultant GRF through the feet, presumably due to the athlete's center of gravity shifting forward. The additional weight bearing of the upper limbs in this off-balance position results in an increase in energy required to move the limbs and to stabilize the upper body. This internal work is likely to be greater when the athlete is in an unbalanced posture, reducing the energy available at the crank, and may explain the attenuated peak power at the lowest crank-axle height. Another consideration is that with the increased hip flexion at lower crank-axle heights (Fig. 4), the load on the lower back would be expected to be greater. The incidence of lumbar spine injuries in America's Cup yacht racing is high (1,14,16), and this has been previously attributed to the forward flexed and rotated position of the spine during grinding at the standard grinding pedestal height (1,16). The use of EMG and three-dimensional kinematic analysis is recommended to investigate optimal body posture and joint angles on the technique of grinding performance, specifically the changes in upper limb joint angles and the relationship between upper and lower limbs.
When comparing the GRF between legs, the greater resultant force magnitude of the dominant leg suggests a degree of asymmetry force production (Table 3). This finding echoes observations by Smith et al. (20), who used nonspecifically trained participants during seated, submaximal arm-crank ergometry. Using torque data, these authors demonstrated that the propulsive forces produced by the dominant arm were greater than those of the contralateral limb and suggested that this might place an overreliance on the dominant arm to produce effective forces. Taken together, these findings clearly indicate that there is a bilateral asymmetry in the forces generated during arm cranking. In addition, the variation in the resultant GRF was greater for the dominant leg at lower crank-axle heights, indicating greater amplitude of the GRF during each crank revolution in this leg at these positions, whereas there was little difference between legs with varying crank lengths (Table 4). However, the dominant leg had greater variability in direction as the crank length shortened. Taken together, this seems to indicate that in the less mechanically efficient arm-crank configurations, the contribution of the dominant side increases. Further investigation is clearly warranted to identify the etiology of asymmetry in force generation during arm cranking.
Crank length and crank-axle height influence performance in standing arm-crank ergometry. The optimal crank length for maximal power was 12%-12.5% for arm span or 235-245 mm for the cohort in this study. These results suggest that standard leg-cycle crank lengths (typically used for arm-cranking) are inappropriate for maximal standing arm-cranking performance. Optimal crank-axle height was between 50% and 60% of stature (950-1150 mm in this study), and a crank-axle height of <50% of stature, which is typically used in America's Cup sailing, may result in reduced performance. The design and configuration of arm-crank ergometers and grinding pedestals should use these findings in optimizing performance. Future research should investigate the contribution of the lower limbs to standing arm-crank performance.
The authors are grateful to the athletes who traveled from various parts of the world to participate in this study. The authors' gratitude is extended to the Sports Technology Institute (www.sports-technology.com) for providing the laboratory facilities and to Technogym (http://www.technogym.com) for supplying the arm-crank ergometer used in this study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
Funding disclosure: This study was privately funded.
Conflict of interest: The authors are not aware of any conflict of interest. The arm-crank ergometer used in this study was provided by Technogym.
Ethical standards: To the best of our knowledge, the experiments comply with current laws.
1. Allen JB. Sports medicine injuries in the America's Cup 2000. NZ J Sports Med
2. Bouhlel E, Chelly MS, Tabka Z, Shephard R. Relationships between maximal anaerobic power of the arms and legs and javelin performance. J Sports Med Phys Fitness
3. Dorel S, Hautier CA, Rambaud O, et al. Torque and power-velocity relationships in cycling: relevance to track sprint performance in world-class cyclists. Int J Sports Med
4. Gardner AS, Martin JC, Martin DT, Barras M, Jenkins DG. Maximal torque- and power-pedaling rate relationships for elite sprint cyclists in laboratory and field tests. Eur J Appl Physiol
5. Hamley EJ, Thomas V. Physiological and postural factors in the calibration of the bicycle ergometer. J Physiol
6. Hubner-Wozniak E, Kosmol A, Lutoslawska G, Bem EZ. Anaerobic performance of arms and legs in male and female free style wrestlers. J Sci Med Sport
7. Jackson AS, Pollock ML. Generalized equations for predicting body density of men. Br J Nutr
8. Jemni M, Sands WA, Friemel F, Stone MH, Cooke CB. Any effect of gymnastics training on upper-body and lower-body aerobic and power components in national and international male gymnasts? J Strength Cond Res
9. Marfell-Jones M, Olds T, Stewart A, Carter L. International Standards for Anthropometric Assessment
. Potchefstroom (South Africa): ISAK; 2006. p. 32-89.
10. Martin JC, Spirduso WW. Determinants of maximal cycling power: crank length, pedaling rate and pedal speed. Eur J Appl Physiol
11. Mercier B, Granier P, Mercier J, Trouquet J, Prefaut C. Anaerobic and aerobic components during arm-crank exercise in sprint and middle-distance swimmers. Eur J Appl Physiol Occup Physiol
12. Neville V. America's Cup yacht racing is not just about the boat. Sport Exerc Sci
13. Neville V. The physiological demands of America's Cup yacht racing [dissertation]. Loughborough (UK): Loughborough University, School of Sports and Exercise Sciences; 2008. p. 67-71.
14. Neville V, Folland JP. The epidemiology and aetiology of injuries in sailing. Sports Med
15. Neville V, Pain MT, Folland JP. Aerobic power and peak power of elite America's Cup sailors. Eur J Appl Physiol
16. Neville VJ, Molloy J, Brooks JH, Speedy DB, Atkinson G. Epidemiology of injuries and illnesses in America's Cup yacht racing. Br J Sports Med
. 2006;40(4):304-11; discussion 311-2.
17. Pearson S, Hume P, Slyfield D, Cronin J. External work and peak power are reliable measures of ergometer grinding performance when tested under load, deck heel, and grinding direction conditions. Sports Biomech
18. Siri WE. Body composition from fluid space and density. In: Brozek J, Hanschel A, editors. Techniques for Measuring Body Composition
. Washington (DC): National Academy of Science; 1961. p. 223-44.
19. Smith PM, Price MJ, Davison RC, Scott D, Balmer J. Reproducibility of power production during sprint arm ergometry. J Strength Cond Res
20. Smith PM, Chapman ML, Hazlehurst KE, Goss-Sampson MA. The influence of crank pattern on muscle activity and torque production during arm crank ergometry. J Electromyogr Kinesiol
21. Too D, Landwer GE. The effect of pedal crank arm length on joint angle and power production in upright cycle ergometry. J Sports Sci
22. Vandewalle H, Peres G, Sourabie B, Stouvenel O, Monod H. Force-velocity relationship and maximal anaerobic power during cranking exercise in young swimmers. Int J Sports Med
23. Whiting P. The 32nd America's Cup: A Simple Guide
. Chichester (England): Wiley; 2007. p. 72.