Biomechanics and Medicine in Swimming XI

BiomechanicsandmedicineinswimmingXi
coordinates in space. Space mean reconstruction accuracy, calculated as
described by Kwon & Casebolt (2006), was 6.2 mm. Data were filtered
with a Butterworth II filter (Winter, 1990). Cut-off frequencies were
between 5 and 7 Hz. The landmark’s positions associated with Dempster’s
anthropometric data (1959) were used to determine the trajectory
of the centre of mass. Using the coordinates of the landmarks and the
centre of mass, the following variables were defined:
- time (t) of the centre of mass position,
- the horizontal velocity of the centre of mass (Vx g ) and hip (Vx h );
- the angle of attack of segments trunk (α tr ), thigh (α th ), leg (α le ),
foot (α fo ) which were defined as the angle between each segment
and the velocity vector of this proximal landmarks;
- the mean kick frequency (f ) of the underwater undulatory swimming
or the reverse period of the ankle motion;
- the mean kick amplitude (A) or the mean peak to peak amplitude
of the ankle during the underwater undulatory swimming;
- the phase time of the knee (P k ) and the ankle (P a ), that is, the
time between the beginning of the hip undulatory motion and
the respective distal joint (knee or ankle) undulatory motion.
These variables were calculated for each swimmer every 0.5 m between
5 m to 7.5 m. After 7.5 m, more than 40% (five swimmers) of the total
number of swimmers began arm propulsion. For the range of distance,
the normality of the variables was tested using Jarque-Bera Test. The
effect of the independent variables (α tr , α th , α le , α fo, f, A, P k, P a ) on the
dependent variables Vx g and Vx h at each normalized distance was analysed
using stepwise linear regression. The limit of significance was set as
p≤0.05. Time (t) at each distance was compared with Vx g and Vx h using
correlation coefficients.
results
Statistical results showed that there was a reverse correlation between
Vx g, Vx h and t at each distance. The regression equation of the parameters
that influence Vx g, Vx h at each distance are given in Table 1 and
Table 2 with corresponding R 2 and p values. These results showed that
different parameters influence Vx g and Vx h at different phases of the
underwater undulatory swimming except at 5 m. At 5.5 m, the stepwise
regression analysis showed that decrease of mean kick amplitude (A)
and attack of trunk (α tr ) are the selected variables to improve respectively
Vx g and Vx h At 5.5, 6 and 7.5 m, the parameters that influence the
horizontal velocity (Vx g and Vx h ) are different for the centre of mass and
the hip. Between 5.5 to 6.5 m, the stepwise regression analysis showed
that the decrease of angles of attack of different segments (α tr at 5.5 m,
α fo at 6 m, α th at 6.5 m) are selected variables to improve the horizontal
velocity Vx g and Vx h . Between 6 to 7.5 m increase of the phase time (P k
and P a ) was related to increasing horizontal velocity Vx g and Vx h . At
6.5 m the decrease of the angles of attack of thigh (α th ) and the increase
of the phase time of the knee (P k ) were related to improve horizontal
velocity Vx g and Vx h (R 2 =0.79 for Vx g and R 2 =0.79 for Vx h ). At 7.5 m
the increase of phase time of the knee (P k ) and mean kick frequency (f )
improved respectively Vx g and Vx h
Table 1: Regression equation with statistical coefficient and p values of
independent variables which present Vxg. Distance (m) Equation R2 5
P values
5.5 Vxg = -0.03 A 0.43 P=0.03
6 Vxg = -0.02 αfo 0.7 P=0.01
6.5 Vxg = -0.009 αth + 1.12 Pa 0.79 P=0.004
7 Vxg = 2.29 Pa 0.52 P=0.017
7.5 Vxg = 1.89 Pk 0.68 P=0.01
98
Table 2: Regression equation with statistical coefficient and p values of
independent variables which present Vxh. Distance (m) Equation R2 5
P values
5.5
6
Vxh = -0.02 αtr 0.56 P=0.01
6.5 Vxh = -0.019 αth + 1.12 Pa 0.89 P=0.0004
7 Vxh = 2.25 Pa 0.52 P=0.017
7.5 Vxh = 0.62 f 0.68 P=0.01
dIscussIon
The inverse correlation between time of the centre of mass position and
the horizontal velocity Vx g and Vx h confirm that horizontal velocity influences
directly the swimmer performance (Elipot et al., 2009). The most
important limitation of the result using stepwise linear regression analysis
is the low number (n) of subjects to predict independent variables (k) using
stepwise regression (Fonton et al., 1998). In the present study: n0.8, the
stepwise regression presents the independent variables as best predictors
and explains more that 50% of the event variability (Fonton et al., 1998).
The results of the present study indicated that the swimmer should stay
in a streamlined position and limit the underwater undulatory swimming
before he reaches 5.5 m. At this distance, mean velocities are Vx g = 2.18
± 0.21 m.s -1 and Vx h = 2.15 ± 0.27 m.s -1 . This is in agreement with the
previous works (Lyttle et al, 2000; Elipot et al., 2009). If the swimmer
would have begun its underwater undulatory swimming too early, hydrodynamic
resistances would have increased and limited the performance of
the underwater phase of the start (Elipot et al., 2009). Decreasing of the
angle of attack the trunk (α tr ) as selected variable of Vx h , confirmed that
the angles of attack directly influence drag and also lift coefficients of the
body. So it has an impact on the swimming propulsion efficiency (Rouboa
et al., 2006). Decreasing the angle of attack of the thigh (α th ) at 6.5 m and
increasing the phase time of the ankle (P a ) improves the horizontal velocity
Vx g and Vx h. This result is in agreement with studies realised on fish
(Loebbecke et al., 2008). For the dolphins, the propulsion minimizes the
displacement of the drag producing forward parts of the body, and maximizes
the displacement of the thrust producing fluke. The displacement
wave that travels the length of the body also has a small magnitude along
the torso, and reaches a maximum at the toes. After 7 m, the increase of the
phase time of the knee (P k ) and the ankle (P a ) improves horizontal velocity.
At 7.5 m, the mean velocities of the swimmers were Vx g = 1.76 ± 0.15 m.s -1
and Vx h = 1.81 ± 0.15 m.s -1 , mean kick frequency (f= 2.32 ± 0.21 Hz) and
mean amplitude (70.8 ± 6.04 cm) were higher than the velocity, frequency
and amplitude observed in the study of Gavilan et al. (2006). The result of
the stepwise regression and the comparison with the results of Gavillan
et al. (2006) confirmed that the swimmer can improve his velocity with
increasing frequency if he maintains large mean kick amplitude.
conclusIon
The stepwise regression enabled four principles to be proposed to improve
the underwater phase of the swimmer: i) to be streamline with
linear adjustment of the trunk and the lower body segments at the beginning
of the underwater gliding phase, ii) to start the dolphin kicking
after 5.5 m, with a high undulation’s frequency iii) to move like dolphins
using only foot and leg for propulsive segment in underwater undulatory
swimming, iv) to use an optimal phase time coordination of the lower
body segments to improve propulsive forces.