The Effect of Fatigue on the Underwater Arm Stroke Motion in the 100-m Front Crawl

Transcription

1 Journal of Applied Biomechanics, 2008, 24, Human Kinetics, Inc. The Effect of Fatigue on the Underwater Arm Stroke Motion in the 100-m Front Crawl Hiroshi Suito, 1 Yasuo Ikegami, 1 Hiroyuki Nunome, 1 Shinya Sano, 2 Hironari Shinkai, 1 and Norio Tsujimoto 1 1 Nagoya University and 2 Gifu City Women s College The purpose of this study was to indicate the effect of fatigue on the underwater right arm stroke motion during the 100-m front crawl. The arm stroke motions of eight male competitive swimmers were captured three-dimensionally at 60 Hz in the positions of 15 m and 65 m from the start. The hand velocity, the arm angular velocities and the relative contribution of the arm angular velocities to the hand velocity were computed at each instant during the arm stroke motion. A significant decrease of the hand velocity and the peak angular velocity of shoulder adduction were observed in the second half than in the first half. The contribution of shoulder adduction was especially large in the pull phase and subsequently that of shoulder horizontal abduction became dominant in the push phase. However, in the second half, the contribution of shoulder adduction tended to decrease while that of shoulder internal rotation tended to increase. Thus, it is quite likely that the arm stroke motion of swimmers were driven to be influenced by induced fatigue and resulted in an increase in the contribution of shoulder internal rotation to compensate the decreased contribution of shoulder adduction angular velocity. Keywords: hand velocity, angular velocity, joint contributions In swimming, propulsive force is induced by motion of the arms and legs. During the front crawl, the propul- Suito, Shikai, and Tsujimoto are with the Graduate School of Education and Human Development, Nagoya University, Nagoya, Japan; Ikegami and Nunome are with the Research Center of Health, Physical Fitness and Sports, Nagoya University, Nagoya, Japan; and Sano is with the Department of International Culture Studies, Physical Fitness and Sports, Gifu City Women s College, Japan. sive force is known to be mainly generated by the arm stroke motion (Hollander et al., 1988; Deschodt et al., It is known that the hydrodynamic forces consist of two components: drag force (F D ) and lift force (F L ). Several studies have been made on the estimate of drag and lift forces acting on hand or hand and forearm (Berger et al., 1995; Bixler & Riewald, 2002; Rouboa et al., 2006; Schleihauf, 1979). The drag and lift forces were first experimentally measured by Schleihauf (1979) using a hand model. Berger et al. (1995) followed this and conducted a similar study using a model including the hand and forearm. Recently, Bixler & Riewald (2002) and Rouboa et al. (2006) estimated the drag and lift forces using computational fluid dynamics system. The result indicated that the drag coefficient was higher than the lift coefficient. Berger et al. (1995) reported also a lower value of the lift coefficient when the forearm was accounted into the model. This implies that the contribution of the forearm for the lift force is negligible. Generally, the drag and lift forces are proportional to the square of velocity, and projected area of the hand and forearm toward the moving direction. Of the main upper limb segments, the hand will reach the highest velocity during the underwater arm stroke motion and has the biggest projected area and drag and lift coefficients of the hand will also be larger because of its flat plate shape compared with cylindrical shape of the forearm. For these reasons, it is reasonable to state that the hand is the main generator of propulsive force and its velocity is one of the main factors when discussing the propulsive force for the front crawl. By contrast, Craig et al. (1985) showed a decrease in swimming velocity during the second half of competitive events. A number of studies in swimming have reported a strong relationship between swimming velocity and arm power (Sharp et al., 1982). Toussaint et al. (2006) reported that the decrease in swimming velocity throughout a 100-m front crawl race is the result of decreases in the power-producing capacity of a swimmer due to fatigue. Typically, swimmers find it difficult to retain stroke motion during competition due to muscle 316

2 Fatigue and the Underwater Arm Stroke 317 fatigue. It is likely that fatigue worsens some part of the arm stroke motion thereby causing a decreased swimming velocity in the final part of swimming. However, little is known about how fatigue affects the kinematics of arm stroke motion. The motions of the hand result from rotational motions of the arm segments, and it is theoretically possible to quantify the contributions of the rotations of the various parts of the arm segment to hand velocity. However, most of the previous studies focused on several simple kinematic and temporal parameters: e.g., the trajectory of the middle fingertip and wrist or stroke length and stroke rate (Craig & Pendergast, 1979; Pai et al., 1984; Craig et al., 1985; Kennedy et al., 1990; Keskinen & Komi, 1993; Arellano et al., 1994; Cappaert et al., 1995; Chollet et al., 1997; Cardelli et al., 1999). To date, detailed contributions of the various parts of the arm segments motion to hand velocity have never been investigated for arm stroke motion in swimming. Sprigings et al. (1994) established a kinematic method for determining anatomical angular velocities of racquetarm segments and the contribution of these rotations to racquet head speed. This method is applicable to determine the relative contributions of arm segments to hand velocity during underwater arm stroke motion and its change due to fatigue is of considerable interest to both swimmers and coaches. The present study, therefore, aims to investigate the effect of fatigue on underwater arm stroke motion during the performance of the full-exertion 100-m front crawl from the aspects of arm angular motion and the relative contributions of arm segments to hand velocity using the methods of Sprigings et al. (1994). Methods Eight experienced male collegiate swimmers (age = 20.5 ± 1.4 yr; height = ± 4.2 cm; mass = 63.1 ± 4.2 kg; the season s best time of 100 m = ± 2.72 s) volunteered to participate in this study. Participants were completely informed about the procedures and demands of the study and signed a written informed consent form. All swimmers performed the full-exertion 100-m front crawl in a 25-m pool (water temperature = 29.0 C). All swimmers started from the water, using a push-off start. Before the start of measurement, the swimmers conducted a self-selected warm-up for 30 min. Before the trials, half-spherical markers (diameter = 15 mm for hand, and 30 mm for wrist, elbow, shoulder and upper body) were fixed securely onto the lateral side of bony anatomical landmarks of the right arm, i.e., the 3rd fingertip (hand), 3rd knuckle (hand), 5th knuckle (hand), styloid process of the ulna (wrist), radial head (elbow), acromial process (shoulder), and substernal region (upper body). Two electrically synchronized video cameras (SONY Inc., CCD-IRIS) were used to capture the right upper limb motion at 60 Hz (exposure time 1/1000 s). Each camera was enclosed in a waterproof cylinder and fixed underwater at a depth of approximately 0.6 m from the surface. The cameras were positioned to the right front and the right rear of the participant (Figure 1). To calibrate the performance area, a calibration frame ( m) with 27 control points was videotaped before the trials. To compare the first and second halves of the fullexertion 100-m front crawl, the entire underwater stroke motion of the right arm was recorded at 15 m and 65 m using the same camera set-up. A digitizing system (DKH Inc., Frame-DIAS) was used to manually digitize the eight body landmarks. The upper body landmark encountered a difficulty in its visualization when the marker was obscured by the arm action. Such cases typically occurred in four or five consecutive fields. Cubic spline function was used to interpolate the hidden part of the upper body marker. Each trial was digitized from the entry of the right hand into the water to the release of the right hand from the water. The direct linear transformation method (Abdel Aziz & Karara, 1971) was used to obtain the threedimensional coordinates of each landmark. The threedimensional coordinates were expressed as a righthanded orthogonal reference frame fixed on the surface of the water (Z was vertical and pointed upward, Y was pointed in the swimming direction, X was perpendicular to Y and Z). The net root mean square error for the X, Y, and Z components in the performance area (X = 1.0 Y = 2.0 Z = 1.0 m) were 5 mm, 3 mm, and 2 mm, respectively. The swimming velocity was computed as the first derivative of displacement of the upper body landmark (substernal region) using central differentiation. The hand velocity was also computed in the same manner, as the first derivative of displacement of the hand s center of mass, which was defined by the fingertip and wrist markers. The absolute magnitude of hand velocity was calculated from the values of each velocity component (X, Y, and Z). For calculating the absolute angular velocity of each segment, we defined the segment coordinate system that consists of three orthogonal principal axes of inertia of the upper arm (RU), forearm (RF), and hand (RH) segments, respectively (Figure 2a). The term Z U was a vector directed from the right elbow to Figure 1 Schematic representation of the experimental set-up.

3 318 Suito et al. Figure 2 (a) Definition of the segment fixed coordinate systems for the upper body (RU), forearm (RF), and hand (RH). (b) Definition of the joint coordinate systems for the shoulder (RAS), elbow (RAE), and wrist (RAW) joints. the right shoulder, Z F was a vector directed from the right wrist to the right elbow, Z H was a vector directed from the right hand s 3rd knuckle to the right wrist. The term Y U was the vector product of Z U and Z F, Y F was the same vector with Y U, Y H was the vector product of Z H and a vector directed from the right hand s 3rd knuckle to the right hand s 5th knuckle. The terms X U, X F, and X H were the vector products of Z U, Z F, and Z H and Y U, Y F, and Y H, respectively. All reference frame vectors were normalized to unit lengths. The angular velocity ( i ) of each segment was calculated using the method similar to those reported by Feltner and Nelson (1996). For example, the absolute angular velocity of the upper arm ( U ) was computed by determining the angular velocity of each axis of RU. The transverse angular velocity component UX was calculated as the scalar product of Z U and the first derivative of Y U (d Y U / dt). The transverse angular velocity component UY was calculated as the scalar product of X U and dz U /dt. The transverse angular velocity component UZ was calculated as the scalar product of Y U and dx U /dt. Using methods similar to those reported by Sprigings et al. (1994), the relative angular velocity vector of each joint was calculated by subtracting its absolute angular velocity vector from that of the adjacent proximal segment. The angular velocity of the shoulder joint ( S ) was defined as the angular velocity of the upper arm ( U ) using the method similar to those reported by Sprigings et al. (1994). The angular velocity of the elbow joint ( E ) was defined as the angular velocity of the forearm ( F ) relative to the upper arm ( U ), and the angular velocity of the wrist joint ( W ) was defined as the angular velocity of the hand ( H ) relative to the forearm ( F ) and calculated as follows: E = U F (1) W = F H (2) As shown in Figure 2b, the relative angular velocity vectors were divided into orthogonal components using unit vectors included in anatomically relevant joint coordinate system defined at the shoulder (RAS), elbow (RAE), and wrist (RAW) joints; Z AS, Z AE, and Z AW were equal to Z U, Z F, and Z F, respectively. The term Y AS was the vector product of a vector directed from the substernal region to the acromial process and Z AS. The term X AS was the vector product of Y AS and Z AS. The term Y AE was the vector product of Z AE and Z AS. X AE was the vector product of Y AE and Z AE. X AW was the vector product of Y H and Z AE. Y AW was the vector product of X AW and Z AW. The shoulder angular velocity vector was separated into three components: horizontal adduction (+X AS ) / horizontal abduction ( X AS ), adduction (+Y AS ) / abduction ( Y AS ), internal rotation (+Z AS ) / external rotation ( Z AS ). The elbow angular velocity vector was one component: extension (+Y AE ) / flexion ( Y AE ). The wrist angular velocity vector was separated into two components: dorsal flexion (+X AW ) / palmar flexion ( X AW ), and radial flexion (+Y AW ) / ulnar flexion ( Y AW ). Sprigings et al. (1994) established a kinematic method for determining anatomical angular velocities of racquet-arm segments and the contribution of these rotations to racquet head speed. In the current study, the relative contributions of arm segments to the hand velocity were computed using the methods of Sprigings et al. (1994). According to Sprigings et al. (1994), the velocity of the hand s center of mass (V H ) relative to the shoulder was considered to be the sum of the linear velocities contributed by the angular velocities of the shoulder, elbow, and wrist joints ( S, E, and W, respectively): V H = S r H/S + E r H/E + W r H/W (3) where r H/S, r H/E, and r H/W were vectors pointing from the shoulder, elbow, and wrist to hand s center of mass,

4 Fatigue and the Underwater Arm Stroke 319 respectively. Each term in Equation 3 is computed as follows: S r H/S = ( SX *X AS + SY *Y AS + SZ *Z AS ) r H/S = ( SX *X AS r H/S ) + ( SY *Y AS r H/S ) + ( SZ *Z AS r H/S ) (4) where X AS, Y AS, and Z AS were the unit vectors that indicate each across vector of the joint coordinate system for the shoulder joint (x, y, and z). E r H/E = ( EX *X AE + EY *Y AE + EZ *Z AE ) r H/E = ( EX *X AE r H/E ) + ( EY *Y AE r H/E ) + ( EZ *Z AE r H/E ) where X AE, Y AE, and Z AE were the unit vectors that indicate each across vector of the joint coordinate system for the elbow joint (x, y, and z). From the degree of freedom of the elbow joint, EX = 0, EZ = 0, then E r H/E = ( EY *Y AE r H/E ) (5) W r H/W = ( WX *X AW + WY *Y AW + WZ *Z AW ) r H/W = ( WX *X AW r H/W ) + ( WY *Y AW r H/W ) + ( WZ *Z AW r H/W ) where X AW, Y AW, and Z AW were the unit vectors that indicate each across vector of the joint coordinate system for the wrist joint (x, y, and z). From the degree of freedom of the wrist joint, WZ = 0, then W r H/W = ( WX *X AW r H/W ) + ( WY *Y AW r H/W ) (6) where SX *X AS, SY *Y AS, and SZ *Z AS were angular velocity vectors of horizontal adduction / abduction, adduction / abduction, and internal rotation / external rotation; EY *Y AE was angular velocity vector of extension / flexion; and WX *X AW and WY *Y AW were angular velocity vectors of dorsal / palmar flexion and radial / ulnar flexion. Thus, the contribution of these anatomical rotations to the hand velocity can be determined by substituting Equations 4, 5, and 6 into Equation 3: V H = ( SX *X AS r H/S ) + ( SY *Y AS r H/S ) + ( SZ *Z AS r H/S ) + ( EY *Y AE r H/E ) + ( WX *X AW r H/W ) + ( WY *Y AW r H/W ) (7) The algorithm developed in this article relies on the following assumptions: (a) the constructed orthogonal axes for the upper limb segments closely approximate their anatomical axes and (b) the pronation / supination rotations at forearm do not largely affect the hand velocity. All parameters were digitally smoothed by a fourthorder Butterworth low-pass digital filter at 7 Hz (Winter, 2004). The propulsive phase when propulsive force was generated by arm stroke motion was divided into two phases by the three key moments shown in Figure 3, defined by Chollet et al. (2000). Propulsive force begins to be generated when the hand moves backward against the water. The pull phase was defined as a time period from the moment when the hand begins to move backward to the moment when the hand reaches the vertical line from the shoulder. The push phase followed and ended when the hand moved forward against the water at the end of the underwater arm stroke motion. In the current study, stroke motions were analyzed for propulsive phase, and the duration of propulsive phase was normalized to 100%. Mean and standard deviations were calculated for the average swimming velocity, average hand velocity, and the peak angular velocities of shoulder adduction, shoulder horizontal abduction, elbow flexion and elbow extension through the propulsive phase. Comparisons between the first and second halves of the full-exertion 100-m front crawl were made using one-paired Student s t tests with Bonferroni correction. Bonferroni correction was made by dividing alpha level by the number of tests (6) conducted in the current study. Probability values less than (alpha level = 0.05 / 6) were judged to be significant. Results The average swimming velocities of the first and second halves were 1.62 ± 0.06 m/s and 1.49 ± 0.06 m/s, respectively. The average hand velocities of the first and second halves were 2.54 ± 0.17 m/s and 2.29 ± 0.19 m/s, respectively. Both the average swimming velocity and hand velocity in the second half were significantly lower than those in the first half (p <.0083, respectively). The average changes of average angular velocity of the shoulder, elbow, and wrist in the first and second halves are presented in Figure 4. The general patterns of shoulder, elbow, and wrist angular velocity were similar in both halves. In the pull phase, shoulder adduction angular velocity was the largest and the peak angular velocity occurred at the end of the pull phase. In the push phase, shoulder horizontal abduction and elbow extension angular velocities began to be larger than those of the others. The peak magnitudes of shoulder adduction, shoulder horizontal abduction, elbow flexion, and elbow extension are summarized in Table 1. A significant difference was found between the two halves for the peak angular velocities of shoulder adduction in the pull phase (7.2 ± 0.7 rad/s vs 6.6 ± 0.8 rad/s, p <.0083). The average changes of contribution of effective segmental rotation to the hand velocity relative to the

5 320 Suito et al. Figure 3 Definition of the two phases of underwater stroke motion in front crawl swimming. Figure 4 Average changes in angular velocities of the shoulder (bottom), elbow (middle), and wrist (top) during the first (left) and second (right) halves of full-exertion 100-m front crawl swimming. In all graphs, time from backward movement of hand until forward movement of hand was normalized to 100%. shoulder are presented in Figure 5. Overall, these changes were similar between the first and second halves. Among the three upper limb joints (shoulder, elbow, and hand), the contribution of the angular veloc- ity generated by the shoulder joint was larger than those of the others. The average changes of relative contribution of the shoulder joint to the hand velocity are highlighted in

6 Fatigue and the Underwater Arm Stroke 321 Table 1 Selected Peak Angular Velocities During the First and Second Halves (rad/s) Angular Velocities 1st half (15 m) 2nd half (65 m) Mean ± SD Mean ± SD p value Shoulder Adduction 7.2 ± ± * Horizontal abduction 5.5 ± ± Elbow Flexion 5.4 ± ± Extension 5.2 ± ± *Significant difference after Bonferroni adjustment with six comparisons (p <.0083). Figure 5 Individual contributions to the hand velocity relative to the shoulder during underwater stroke motion during the first (left) and second (right) halves by the rotations of the upper arm, forearm, and hand. In all graphs, time from backward movement of hand until forward movement of hand was normalized to 100%. Figure 6. In the pull phase, there was a dominant contribution to hand velocity attributed to the shoulder adduction / abduction. However, in the push phase, a marked positive contribution was observed for shoulder horizontal adduction / abduction. In the second half of 100 m of the front crawl, the average contribution of the shoulder adduction / abduction tended to decrease at the beginning of the pull phase (five of the eight subjects decreased from 91.1% to 87.0%). In contrast, that of the shoulder internal rotation / external rotation tended to increase at the beginning of the pull phase (seven of the eight subjects increased from 8.5% to 14.7%). Discussion In the current study, the decrease of swimming velocity was approximately 7.7% between the first and second halves in the full-exertion 100-m front crawl. Chollet et al. (1997) reported a 5.7% decrease in swimming velocity during the second half of a long-course 100-m race. Pai et al. (1984) reported a difference in velocity of 6.3% between the first and second halves. The decrease of swimming velocity observed in the current study was slightly greater than those found in the literature. Craig et al. (1985) reported that higher skilled swimmers were greater in maintaining their swimming velocity than

7 322 Suito et al. Figure 6 Average changes in the percentages of contribution of hand velocity relative to the shoulder during underwater stroke motion during the first (solid lines) and second (dashed line) halves by the rotations of the upper arm (internal / external rotation [bottom], adduction /abduction [middle], and horizontal adduction / abduction [top]). In all graphs, time from backward movement of hand until forward movement of hand was normalized to 100%. lower skilled swimmers. Despite there being a difference for calculating procedure of the swimming velocity between the present and previous studies, the average swimming velocity of the current study (1.62 ± 0.06 m/s), was distinctively slower than the range of average values ( m/s) reported previously (Pai et al., 1984; Chollet et al., 1997). The slower velocities represented the skill level of the present subjects and may account for the larger decrease of swimming velocity observed in the current study. The propulsive forces by the arm stroke motion were strongly related to the hand orientations relative to water flow and proportional to the square of hand velocity (Schleihauf, 1979). Thus it can be considered that hand velocity is one of the main factors to predict swimming velocity in the front crawl. In the current study, the decline of hand velocity was significant (approximately 9.9%) between the first and second halves. It is reasonable to suggest that the induced fatigue decreased the hand velocity, thereby most likely leading to the reduced swimming velocity. In swimming, arm stroke motion biomechanics have been studied by several authors using three-dimensional cinematographic procedures (Aujouannet et al., 2006; Cappaert et al., 1995; Schleihauf, 1979). However, these studies solely reported fingertip trajectories as kinematic parameters, in which the angular velocity changes of the arm joints have never been clearly illustrated. In the current study, the angular velocity changes of the arm were clearly illustrated and the change between the first and second halves was examined in detail. In the pull phase, a significant difference was observed in the peak angular velocity of shoulder

8 Fatigue and the Underwater Arm Stroke 323 adduction. Rouard & Billat (1990) observed higher reciprocal contractions of biceps-triceps during the pull phase than during the push phase. Clarys & Rouard (1996) and Rouard et al. (1997) also showed that muscle activity during the pull phase was higher than that of the push phase. It can be considered that the pull phase requires a larger burst of muscle activity than the push phase, and thereby the stroke motion in the pull phase would be more susceptible to fatigue. Furthermore, the current study investigated the contributions of arm rotational motion to hand velocity relative to the shoulder velocity in the first half and second half of 100 m of the front crawl. Figure 7 shows the measured (solid line) and predicted (dashed line) hand velocities of one typical subject. It can be seen that there was a good agreement between the two curves throughout the arm stroke motion. To quantify the agreement of the two curves, the root mean square errors of the predicted velocity to the measured velocity was calculated for all subjects. The average error of 5.7 ± 1.6% (maximum = 8.4%, minimum = 3.2%, respectively) was acceptable. This guaranteed that the hand velocity was accurately predicted in the current study. However, there was a relatively larger gap between the two curves at the end of the push phase. In the end of push phase, it was observed for all subjects that the shoulder position became near the surface of the water. The accuracy of the three-dimensional coordinates of the shoulder will decline when the shoulder was near to the waved water surface. Thus, the relatively large error was most likely caused by the inaccurate shoulder position. However, in the current study, as we focused on the change in the pull phase, the main results were not affected by the above phenomena. From the aspects of the relative contribution, it was revealed that the shoulder adduction motion was the dominant contributor during the pull phase and the horizontal abduction motion of the shoulder become the dominant contributor during the consecutive push phase. In the current study, we indicated more dominant role of shoulder motion to hand velocity in the front crawl. Basically, these contributions to hand velocity were similar throughout the first and second halves of swimming. However, at the beginning of the pull phase, the average contribution of shoulder adduction tended to decrease in the second half ( %). In contrast, the average contribution of shoulder internal rotation at the beginning of the pull phase tended to increase in the second half ( %). Aujouannet et al. (2006) showed that the maximal isometric force of shoulder flexion (corresponds to shoulder adduction in the current study) was significantly decreased by fatigue induced during swimming. This finding is in line with the results of the current study, in particular when looking at the significant decrease of shoulder adduction angular velocity and its reduced contribution to hand velocity in the second half of swimming. Thus, it is quite likely that, for the swimmers of the current study, their arm stroke motion was influenced by fatigue and resulted in an increase in the contribution of shoulder internal rotation to compensate for the decreased shoulder adduction angular velocity. In summary, the current study succeeded in delineating the effect of fatigue on the aspects of arm angular Figure 7 The comparison of the measured and the predicted hand velocities calculated by the link segment model of the upper limb. Solid line: the measured hand velocity relative to the shoulder. Dashed line: the predicted hand velocity relative to the shoulder.

9 324 Suito et al. motion during the performance of full-exertion 100-m front crawl. As a consequence, the swimming velocity, hand velocity, and peak angular velocity of shoulder adduction were reduced significantly from the first half to the second half of full-exertion 100-m front crawl. As the hand is assumed to be the main generator of the propulsive force, it is quite likely that the decreased swimming velocity was caused by the reduced hand velocity. Moreover, from a kinematic procedure, it was revealed that the shoulder joint rotations accounted for most of the hand s velocity and these aspects were influenced by fatigue, in which an increased contribution of shoulder internal rotation and a reduced contribution of shoulder adduction were simultaneously observed at the beginning of the pull phase. These results suggest that the effect of fatigue appeared on the shoulder muscle, thereby inducing a decrease in hand velocity and change of the stroke maneuver. References Abdel-Aziz, Y.I., & Karara, H.M. (1971). Direct linear transformation: From comparator coordinates into object coordinates in close-range photogrammetry (ASP Symposium on Close-Range Photogrammetry). Falls Church, VA: American Society of Photogrammetry. Arellano, R., Brown, P., Cappaert, J., & Nelson, R.C. (1994). Analysis of 50-, 100-, and 200-m freestyle swimmers at the 1992 Olympic games. Journal of Applied Biomechanics, 10, Aujouannet, Y.A., Bonifazi, M., Hintzy, F., Vuillerme, N., & Rouard, A.H. (2006). Effects of a high-intensity swim test on kinematic parameters in high-level athletes. Applied Physiology, Nutrition, and Metabolism, 31(2), Berger, M.A.M., de Groot, G., & Hollander, A.P. (1995). Hydrodynamic drag and lift forces on human hand/arm models. Journal of Biomechanics, 28(2), Bixler, B., & Riewald, S. (2002). Analysis of a swimmer s hand and arm in steady flow conditions using computational fluid dynamics. Journal of Biomechanics, 35(5), Cappaert, J.M., Pease, D.L., & Troup, J.P. (1995). Threedimensional analysis of the men s 100-m freestyle during the 1992 Olympic games. Journal of Applied Biomechanics, 11, Cardelli, C., Chollet, D., & Lerda, R. (1999). Analysis of the 100-m front crawl as a function of skill level in nonexpert swimmers. Journal of Human Movement Studies, 36, Chollet, D., Chalies, S., & Chatard, J.C. (2000). A new index of coordination for the crawl: Description and usefulness. International Journal of Sports Medicine, 21(1), Chollet, D., Pelayo, P., Delaplace, C., Tourny, C., & Sidney, M. (1997). Stroking characteristic variations in the 100-m freestyle for male swimmers of differing skill. Perceptual and Motor Skills, 85(1), Clarys, J.P., & Rouard, A.H. (1996). The frontcrawl downsweep: shoulder protection and/or performance inhibition. Journal of Sports, Medicine, and Physical Fitness, 36(2), Craig, A.B., & Pendergast, D.R. (1979). Relationships of stroke rate, distance per stroke, and velocity in competitive swimming. Medicine and Science in Sports, 11(3), Craig, A.B., Skehan, P.L., Pawelczyk, J.A., & Boomer, W.L. (1985). Velocity, stroke rate, and distance per stroke during elite swimming competition. Medicine and Science in Sports and Exercise, 17(6), Deschodt, V.J., Arsac, L.M., & Rouard, A.H. (1999). Relative contribution of arms and legs in humans to propulsion in 25-m sprint front-crawl swimming. European Journal of Applied Physiology, 80(3), Feltner, M.E., & Nelson, S.T. (1996). Three-dimensional kinematics of the throwing arm during the penalty throw in water polo. Journal of Applied Biomechanics, 12, Hollander, A.P., de Groot, G., van Ingen Schenau, G.J., Kahman, R., & Toussaint, H.M. (1988). Contribution of the legs to propulsion in front crawl swimming. In B.E. Ungerechts, K. Wilke, & K. Reischle (Eds.), Swimming science V (pp ). Champaign, IL: Human Kinetics. Kennedy, P., Brown, P., Chengalur, S.N., & Nelson, R.C. (1990). Analysis of male and female Olympic swimmers in the 100-meter events. International Journal of Sport Biomechanics, 6, Keskinen, K.L., & Komi, P.V. (1993). Stroking characteristics of front crawl swimming during exercise. Journal of Applied Biomechanics, 9, Pai, Y.C., Hay, J.G., & Wilson, B.D. (1984). Stroking techniques of elite swimmers. Journal of Sports Sciences, 2, Rouard, A.H., & Billat, B.P. (1990). Influences of sex and level of performance on freestyle stroke: an electromyography and kinematic study. International Journal of Sports Medicine, 11(2), Rouard, A.H., Billat, R.P., Deschodt, V., & Clarys, J.P. (1997). Muscular activations during repetitions of sculling movements up to exhaustion in swimming. Archives of Physiology and Biochemistry, 105(7), Rouboa, A., Silva, A., Leal, L., Rocha, J., & Alves, F. (2006). The effect of swimmer s hand/forearm acceleration on propulsive forces generation using computational fluid dynamics. Journal of Biomechanics, 39(7), Sharp, R.L., Troup, J.P., & Costill, D.L. (1982). Relationship between power and sprint freestyle swimming. Medicine and Science in Sports and Exercise, 14(1), Schleihauf, R.E. (1979). A hydrodynamic analysis of swimming propulsion. In J. Terauds, & E.W. Bedingfield (Eds.), Swimming III: Proceedings of the Third International Symposium of Biomechanics in Swimming (pp ). Baltimore: University Park Press. Sprigings, E., Marshall, R., Elliott, B., & Jennings, L. (1994). A three-dimensional kinematic method for determining the effectiveness of arm segment rotations in producing racquet-head speed. Journal of Biomechanics, 27(3), Toussaint, H.M., Carol, A., Kranenborg, H., & Truijens, M.J. (2006). Effect of fatigue on stroking characteristics in an arms-only 100-m front-crawl race. Medicine and Science in Sports and Exercise, 38(9), Winter, D.A. (2004). Biomechanics and motor control of human movement (3rd ed.). New York: John Wiley & Sons.

10