REVIEW ARTICLE

Sports Medicine 13 (I): 8-24, 1992

0112-1642/92/0001-0008/$08.50/0
© Adis International Limited. All rights reserved.
SP0183
159619

Biomechanics of Competitive Front Crawl Swimming

Huub M. Toussaint and Peter J. Beek
Department of Health Science, and Department of Psychology, Faculty of Human Movement Sciences
Vrije Universiteit and Universiteit van Amsterdam, Amsterdam, The Netherlands

Contents
8 Summary
9 1. The Problem
9 2. Drag
12 3. Propulsion
12 3.1 Mechanics of Swimming Propulsion
13 3.2 Hydrodynamics of Swimming Propulsion
13 3.3 Pulling Patterns of Highly Skilled Swimmers
14 3.4 Energetics of Swimming Propulsion
16 3.5 Propelling Efficiency and Vorticity
16 4. Performance-Determining Factors
17 4.1 Propelling Efficiency
17 4.1.1 Propelling Efficiency and Hand Surface Area
18 4.1.2 Distance per Stroke
19 4.2 Power Output
20 5. The Partitioning of Time, Space and Energy in a Swimming Event
21 6. Conclusions and Implications for Training

Summary Essential performance-determining factors in front crawl swimming can be analysed within a
biomechanical framework, in reference to the physiological basis of performance. These factors
include: active drag forces, effective propulsive forces, propelling efficiency and power output.
The success of a swimmer is determined by the ability to generate propulsive force, while
reducing the resistance to forward motion. Although for a given competitive stroke a range of
optimal stroking styles may be expected across a sample of swimmers, a common element of
technique related to a high performance level is the use of complex sculling motions ofthe hands
to generate especially lift forces. By changing the orientation of the hand the propulsive force
acting on the hand is aimed successfully in the direction of motion. Furthermore, the swimming
velocity (v) is related to drag (A), power input (Pi, the rate of energy liberation via the aerobic/
anaerobic metabolism), the gross efficiency (eg), propelling efficiency (e p ), and power output (Po)
according to:
,.------
egO e • Pi
p
v=
A
Based on the research available at present it is concluded that: (a) drag in groups of elite swimmers
homogeneous with respect to swimming technique is determined by anthropometric dimensions;
Biomechanics of Front Crawl Swimming 9

(b) total mechanical power output (Po) is important since improvement in performance is related
to increased Po. Furthermore, it shows dramatic changes with training and possibly reflects the
size of the 'swimming engine'; (c) propelling efficiency seems to be important since it is much
higher in elite swimmers (6\ %) than in triathletes (44%); and (d) distance per stroke gives a fairly
good indication of propelling efficiency and may be used to evaluate individual progress in tech-
nical ability.

To swim as fast as possible over a certain dis- At the beginning of the race the acceleration is large
tance, which output variables should we optimise? because the push-off is made from the platform or
What factors determine the performance? To what side of the pool while drag is still small, resulting
degree does the elite swimmer's success depend on in a large net positive force. As long as the swim-
technique and coordination, as opposed to pure mer's velocity is relatively low, the propulsive force
anthropomorphic measures (such as body length (Fp) greatly exceeds the velocity-dependent resist-
and the size of the hands), sheer strength and aero- ant drag force (Fd). As the swimmer's velocity in-
bic capacities? How do the elite and average swim- creases, so does the drag force. Because the pro-
mer differ? How can the average swimmer im- pulsive force also depends on the swimmer's
prove beyond his or her colleagues? velocity (Martin et al. 1981), the basic balance of
Generally, swimming performance is deter- forces is
mined by the swimmer's physiology (capacity for (Eq. 2)
energy output, metabolic processes), morphology
(size of the hands, frontal area, determining active A constant velocity is reached when Fd = Fp
drag), neuromuscular properties (capacity for gen- and the acceleration is zero. Thus, the success of a
erating force, doing work, delivering power, all of swimmer is determined by the ability to generate
which relate to technique and coordination) and propulsive force while reducing the resistance to
psychological profile (arousal, anxiety, motivation, forward motion.
etc.). The present article attempts to bring together
the currently available, but often scattered and 2. Drag
piecemeal, information on the first 3 types of per-
formance-determining factors in front crawl swim- Several authors have suggested that drag forces
ming and discuss them in the context of a bio- may be diminished by improving the swimming
mechanical analysis of swimming. Special attention technique (Bober & Czabanski 1975; Counsilman
will be given to the role of the arms, since it is 1968; Maglischo 1982), which, if true, would lead
generally agreed that the arms provide more than to the hypothesis that elite swimmers have lower
85% of the total thrust in the crawl stroke (Bucher resistance at high swimming velocities than poorer
1975; Faulkner 1966; Hollander et al. 1988; Wat- ones. To verify this hypothesis it is necessary to
kins & Gordon 1983). determine the resistive forces to which the swim-
mer is exposed while stroking. For the measure-
1. The Problem ment of this active drag various methods have been
used (see Clarys 1979; Hay 1988; Miller 1975; Vaart
A competitive swimmer's success may be as- et al. 1987 for reviews and discussions on these
sessed by Newton's Second Law: methods). Recently a measurement system for ac-
~F = mv (Eq. 1)
tive drag, the MAD system, has been introduced
(see Hollander et al. 1986; Toussaint et al. 1988b
where ~F represents the various forces acting on a for an extensive description). Although the appli-
swimmer of mass m, and v his or her acceleration. cation and possible limitations of the MAD system
10 Sports Medicine /3 (1) 1992

ment in performance. These observations war-

Women Men
68 58 ranted further research into the relation between
~ • • •
~ • drag and morphology, although the first extensive
• • $.
•
$.
(I)
E
64 (I)
• • studies in this area by Clarys (1979) were not very
:;:::;
• § 53 • •• •
E 60 • • promising. In cooperation with Clarys and with the
••••
0 E
0
~

• ~
0
0 • help of the MAD system, the relationship between
56 48 morphology and active drag was re-evaluated
40 60 80 60 90 120
Drag (N) Drag (N) (Huijing et al. 1988). This time, significant corre-
lations were found between selected anthropome-
Fig. 1. Performance expressed as 100m time is presented as
tric variables and active drag. Particularly notable
a function of drag measured at a velocity of 1.63 m/sec for
women (n = 12) and 1.86 m/sec for men (n = 12). was the high degree of correlation of the maximal
body cross-section (r = 0.87) and active drag. In
another study using the MAD system, the drag val-
to the measurement of active drag in front crawl ues for male and female swimmers were compared
swimming are still open to discussion (Hay 1988), (Toussaint et al. 1988b). The higher drag values in
it seems at present the best of several alternatives the men (Fd = 30v2 versus 24v 2 for the women)
available for the estimation of drag force during could be related to a larger body cross-section
front crawl swimming. This drag (Fd in N) is re- (O.091m 2 versus 0.075m 2). In that study, however,
lated to the square of the swimming velocity (v in the observed groups were homogeneous with re-
m/sec) [Clarys 1979; Karpovich 1933] according
spect to height. Therefore, a longitudinal study (2.5
to: years) [Toussaint et al. 1990b] of a group of child-
Fd = A· v 2 (Eq. 3) ren was undertaken (mean age at the start: 12.9
where A is a proportionality constant. years) to test the hypothesis that the increase in
In one of the first experiments with the MAD body dimensions (especially the body cross-sec-
system, the relationship between drag and maxi- tional area) during growth results in a correspond-
mal swimming performance was investigated (Hol- ing increase of drag during swimming. However,
lander et al. 1985). Active drag was determined in while the body cross-sectional area of the children
12 female and 12 male elite swimmers at high ve- increased in size by 16%, no differences in total
locities (1.63 and 1.86 m/sec, respectively). As no drag were found (1985: 30.1 ± 2.37N versus 1988:
significant correlations were found (males r = 30.8 ± 4.50N).
-0.27, females r = 0.07) [fig. 1], it was concluded This unexpected result was explained by the fact
that drag per se is not a determining factor of the that the total drag force (Fd) in swimming is not
maximal swimming speed. only determined by the pressure drag (Fpd), but by
In the literature, however, it has often been 2 other components as well: friction drag (Ff) and
claimed that good swimmers have an advanta- wave-making resistance (Fw). Hence:
geous body build with respect to drag. For in- Fd = Fpd + Ff + Fw (Eq. 4)
stance, Cureton (1975) reported that the tall, slim
type has been shown to glide better through the The first component (Fpd), which dominates at
water. In addition, several others noted that swim- the prevailing high Reynold's number (Re) of 2.2
mers are taller than the mean population, suggest- X 106 to 2.5 X 106 (Clarys 1979), is determined
ing a performance advantage related to height (An- mainly by the body cross-sectional area.
drew et al. 1972; Faulkner 1968; Marconnet et al. The friction drag (Ff) depends on the friction
1978; Nomura 1983). Furthermore, Kunski et al. between skin and water and will theoretically in-
(1988) observed in a longitudinal study that a re- crease somewhat since the total skin surface will
lation exists between growth rate and improve- increase (but see also Toussaint et al. 1989). Hence
Biomechanics of Front Crawl Swimming 11

this component cannot explain the constant total crawl) swiming performance. Furthermore, form
drag value. indices derived from shipbuilding technology re-
The wave-making resistance (Fw) is the result vealed changes that indicated a more 'streamlined'
of the deformation of the water surface. The water body. It was concluded, therefore, that during
tends to pile up in front of the swimmer and to growth a complex process takes place in which dif-
form hollows behind, thus creating a wave system. ferent drag-determining factors, such as height, body
At higher velocities a significant bow wave devel- shape, and body cross-sectional area, change in di-
ops (Alley 1952), which could indicate that wave rections that have opposite effects on drag.
drag contributes considerably to total drag (Miller To these conclusions should perhaps be added
1975). This seems especially the case when swim- that an improved technique may result in lower
ming near the 'hull speed', a concept used in ship- drag values. Support for this idea was obtained in
building that was introduced into the literature on an experiment in which triathletes were compared
human swimming by Miller (1975). The hull speed to swimmers (Toussaint 1990). A considerable dif-
is the speed reached when the wavelength of the ference in performance was found which was at-
bow wave and the water-line length of the hull are tributed, among other factors, to a difference in the
equal. A longer hull (increased height) would make coefficient A (see equation 3), which was 30.5 for
a higher swimming speed possible. the competitive swimmers and 41.6 for the triath-
The suggested relation between height and wave
drag finds its expression in the Froude number (Fr),
I
which depends in human swimming on the height 80 I
I
I
of the swimmer as shown in equation 5. ---1988 I
-1985 ,I
Fr = v/v(g· L)
I
(Eq. 5) 60 ,t

where v represents swiming velocity, L the height

g
~ 40
of the swimmer, and g the acceleration due to grav- Cl
ety (9.81 m/sec2 ).
20
This implies that a decrease in Fr will lower the
wave-making resistance. Thus, in the children the
increase of height (from 1.52 to 1.69m) resulted in 0
a decrease of Fr (from 0.324 to 0.308 at a swim- 0.0 0.5 1.0 1.5 2.0
Velocity (m/sec)
ming velocity of 1.25 m/sec) and hence in a de-
crease of wave-making resistance. In figure 2 the
I
mean drag curves for the entire group are related 80
I
I
I

to speed. Again it is clear that the drag has not /

I

changed. When the same drag data are presented 60 I

/
I

with respect to the Froude number (fig. 2), hence g /

/
correcting for the change in height, it appears that ~ 40
the drag had increased. If the drag is calculated for Cl

a Froude number of 0.324 (1985: v = 1.25 m/sec) 20

it gives a drag of 30.1 N in 1985 and a value of
34.6N in 1988. The increase of 15% is about the
increase in the body cross-sectional area. The sug- O~~n>. .~. .rrnoTrn!~~
0.0 0.1 0.2 0.3 0.4 0.5
gestion that the increase in pressure drag was com-
Froude number
pensated by a decrease in wave-making resistance
is in line with the common notion that taller swim- Fig. 2. The relationship between drag and the Froude num-
mers seem to have an advantage for a good (front ber.
12 Sports Medicine 13 (1) 1992

judgments. Counsilman (1968) was the first to ap-

ply biomechanics to develop theories of hydrody-
namic propulsion. He reasoned that pushing water
backward would create a high pressure zone on the
palm, and a low pressure zone on the back of the
d hand. The resulting propulsive drag force would
propel swimmers forward. This simple theory was
c
quickly modified to explain the sculling move-
d
Side view ments in the arm pull observed in underwater cin-
ematography. An example of the 3-dimensional
Front view Bottom view
movement of the left hand through the water is
given in figure 3.
Fig.3. Front crawl stroke pattern in 3-dimensional space (after
Svec 1982). Abbreviations: a-b = entry; b-c = entry scull; c- Counsilman (1969) proposed that instead of
d = inward pull; d-e = outward pull; e-f = exit. pushing water on a straight line backwards, the
hand follows a curvilinear path to continuously find
letes. This could be interpreted as being a result of still water to push against and thus gain more re-
the poorer technique of the triathletes causing su- sistance than it would by pushing against water that
perfluous body movements in the vertical and hor- had already been accelerated. On the basis of fur-
izontal plane perpendicular to the swimming ve- ther hydrodynamic analyses, Counsilman also drew
locity (Bober & Czabanski 1975; Counsilman 1968; attention to the importance of the lift force in
Maglischo 1982). In other words, the competitive swimming (Counsilman 1971). This force, similar
swimmers were able to use a more symmetrical to that experienced by an aircraft wing, is related
stroking pattern which gives a more stable body to a difference in velocity of water passing above
position and hence a smaller drag coefficient. and below the surface of the hand. The higher ve-
Whether drag is a major performance-determin- locity associated with lower pressure (principle of
ing factor cannot be entirely resolved on the basis Bernouilli) and vice versa set up a pressure differ-
of the research presently available. It seems as if ential resulting in a lift force directed at right an-
drag in groups of elite swimmers that are homo- gles to the line of motion of the propelling surface
geneous with respect to swimming technique is de- (fig. 4). Subsequently, numerous papers were pub-
termined by anthropometric dimensions (e.g. body
cross-sectional area and height). A small reduction
in drag can probably be achieved by stretching the
arm in the glide phase of the stroke, as was sug-
gested by Holmer (1979). The body cross-sectional
area is especially reduced when the shoulder is
'stretched behind the arm'. Furthermore, stretch-
ing 'elongates the body', possibly resulting in re-
duced wave resistance (Larsen et al. 1981).

3. Propulsion Fig. 4. Movement through water is resisted by the water im-

3.1 Mechanics of Swimming Propulsion mediately in front and around the hydrofoil. This water re-
sistance is called drag. Drag force always acts in a direction
opposite to the direction of motion. Due to the form the
The mechanics involved in the generation of water speed above the hydrofoil is greater than below. The
propulsive forces received scant attention until the pressure above the hydrofoil is less than below and there is
late 1960s, Before that time descriptions of stroke thus a net upward force called dynamic lift. Lift force always
mechanics were based exclusively on empirical acts perpendicular to drag.
Biomechanics of Front Crawl Swimming 13

Lift
force component and to steer the resultant pro-
pulsive force straight ahead. This implies that in-

tL::o~g stead of a single optimal line of motion straight

backwards, abundant solutions exist to combine the
lift and the drag force components to provide pro-
Direction of motion
4 pulsion in a straight forward direction (fig. 7). Thus,
the pitch of the hand must be continuously ad-
justed to everchanging directions of pull (Schlei-
hauf 1974). In order to choose the angle of attack
which gives the best combination of drag and lift
forces at every moment of a pull, the swimmer has
Fig. 5. a denotes angle of attack (after Svec '982). to have a 'feel' for water. This feel for water is cer-
tainly associated with talent. Given the quantita-
lished (Barthels & Adrian 1974; Hay 1973; Rack- tive information on hand C, and hand Cd values,
ham 1975; Reischle 1979; Schleihauf 1974; Unger- it is possible to estimate hand propulsive force
echts (979) to support the notion that the total based upon 3-dimensional cinematographic infor-
propelling force acting on the hands is determined mation, which is necessary to determine the ve-
by both a lift and drag component. locity of the hand relative to the water. Using this
technique, Schleihauf has presented detailed anal-
3.2 Hydrodynamics of Swimming Propulsion yses of stroke mechanics and hydrodynamic forces
of elite swimmers representing all stroke tech-
The suggested hydrofoil behaviour of the hand niques (Schleihauf 1979; Schleihauf et al. (988).
was investigated in depth by Schleihauf (Schleihauf
1979, 1986; Schleihauf et al. 1983, (988). Accord- 3.3 Pulling Patterns of Highly
ing to hydrodynamic theory the drag and lift force Skilled Swimmers
can be derived using the following equations:
L = IhpVh2C,S (Eq. 6) The analyses reveal that swimmers use complex
sculling motions of the hands (fig. 3). By changing
(Eq. 7)
the angle of pitch of the hand they successfully
where L = lift force, D = drag force, p = density combine the lift and drag force in a resultant aim-
of water, Vh = hand velocity, C, = coefficient of ing in the direction of motion (fig. 7). Further, it
lift, Cd = coefficient of drag, and S = propelling
surface of the hand. Schleihauf measured forces on
hand models in the fluid lab and was able to quan-
tify the force-producing characteristics of the hand
11'=45° 11'=135°
in terms of coefficient of lift (C,) and coefficient of
drag (Cd) dependent on the angle of attack and the '\. ,/
sweep back angle. The angle of attack is the angle
formed by the inclination of the propelling surface + - - 11'=180°
11'=0°- +
to its direction of motion (fig. 5). The sweep back
angle defines the leading edge of the hand (fig. 6)
It appeared that the magnitude of the lift and
drag force component are extremely dependent on
/
11'=315° ~2250
the angle of attack. By varying the angle of attack
(pitch of the hand) given a direction of motion of
the hand, it is possible to vary the lift and drag Fig. 6. Sweepback angle convention (after Schleihauf '979).
14 Sports Medicine 13 (1) 1992

was demonstrated that the highly skilled swimmers 125

utilised the most rapid hand actions (high veloci- --- Total propulsion

ties v, implying high propulsive forces; see equa- 100 - - Effective propulsion
g
tions 6 and 7) in the side-to-side and up-and-down 75
dimensions of motion. On average the highest pro- ~
.2
pulsive arm speeds are aimed at about 63° to their Q)
>
50
'iii
bodies' line of motion (Schleihauf et al. 1983) [fig. '3
a. 25
3 phase d-e). The highest propulsive forces are pro- e
0..
duced near the end of the arm stroke (fig. 8). In a 0

group of elite swimmers the largest hand forces oc-

Time (sec)
curred on average at 78% of the total stroke time.
An impulsive force distribution seems character-
Fig. 8. The total and effective propulsive force of the hand
istic of skilled stroke technique. during a stroke. The effective force is the component of the
The expected use of lift force to generate the total force aiming in the straight forward direction.
high propulsive hand forces was underlined by high It seems logical to suppose, as we did before,
lift-drag ratios of 1.0 on average, indicating that the that a swimmer's success will be determined by the
key propulsive forces produced by highly skilled ability to generate propulsive force while reducing
swimmers involve equal combinations of lift and the resistance to forward motion. Although there
drag forces. Furthermore, at a swimming speed of is a direct relationship between the mean propul-
1.66 m/sec, on average 48N of effective propulsive sive force and swimming speed, swimming, and
force was delivered by the hands. The effective particularly swimming fast, is not just a matter of
forearm propulsion was about 24N. Hence, the total generating high propulsive forces. As most other
propulsive force ofthe arms was 72N. Given some sport activities, swimming is a form of exercise in
additional propulsive force by the feet (15%) this which the aerobic and anaerobic capacities code-
compares very well with the expected 83N termine success. Thus, the question is not simply
(30· 1.66 2 ) necessary to overcome drag (see equa- how to maximise the propulsive force, but rather
tion 3). how to achieve a high propulsive force with finite
It should also be noted that, for a given com- metabolic capacities.
petitive stroke, a range of optimal stroking styles
may be expected across a sample of skilled swim-
mers. Depending upon the individual's strength, 3.4 Energetics of Swimming Propulsion
flexibility, training background, etc., 'ideal' pat-
terns vary from person to person. When analysing the energetics of swimming it
is tempting to think that the work (J) a swimmer
has to deliver is simply the drag force (Fd) times
the distance (d) covered, or, if we consider the time
derivative of work, that the mechanical power out-
put of the swimmer has to be equal to the drag
force times the swimming velocity. Contrary to
most on-land activities, however, where the push-
off is performed against the earth, which is not ac-
celerated, the push-off in swimming is made against
water, which is accelerated (van Ingen Schenau &
Fig. 7. The direction of the hand velocity changes during the Cavanagh 1990). This principle is illustrated in fig-
pull. The angle of attack is continuously adapted to direct ure 9.
the propulsive force Fp forward. To generate a propulsive force, the swimmer
Biomechanics of Front Crawl Swimming 15

gives a mass of water 2:mi a velocity change .:lvi in

time T. Thus, the mean propulsive force equals:

Fp = -
1 It 1
Fp dt = - Lmi· .:lvi (Eq. 8)
ToT
where T equals the cycle time. The masses of water
(~mi) pushed away have acquired a kinetic energy Fig. 10. Free body diagram of a swimmer at time t. Fp rep-
resents the propulsive force acting on the hand. Fd is the drag
(Ek) in time T equal to:
force acting on the swimmer. Vh and Vb are the respective
(Eq. 9) velocities of the point of application of the forces. See text
for details.
This energy is transferred from the swimmer to
the water. Thus, part of the mechanical work the
swimmer delivers during push-off is dissipated in to the total power output has been defined as the
moving water (Charbonnier et al. 1975; Holmer propelling efficiency (ep) [Alexander & Goldspink
1975; Miyashita 1974) and only a proportion is used 1977; Groot & van Ingen Schenau 1988; Huijing
effectively to overcome body drag. As was pointed et al. 1983; van Ingen Schenau & Cavanagh 1990;
out in the literature on the swimming offish (Alex- Martin et al. 1981; Toussaint et al. 1988a, 1990a):
ander & Goldspink 1977; Bone 1975; Webb ep = Pd/Po = Pd/(Pd + Pk) (Eq. 11)
1971a,b) and the design of swim fins (Lewis &
Lorch 1979; McMurray 1977), the energy trans- Another approach to elucidate this phenom-
ferred from the swimmer to the accelerated water enon is to examine a free body diagram of a swim-
should be included in the power 'book-keeping'. mer (fig. 10). A swimmer has to produce mechan-
Hence, the total mechanical power (Po) produced ical power to overcome the power lost due to drag,
by the swimmer (assuming a constant swimming which equals:
velocity) equals the power to overcome drag (Pd) Pd = Fd· Vb • cos (180°) (Eq. 12)
plus power expended in giving masses of water
where Vb represents the velocity of the body, and
pushed away a kinetic energy change (Pk):
cos (180°) accounts for the opposite directions of
Po = Pd + Pk (Eq. 10) the force and velocity vector. Since cos (180°) =
The ratio of the useful power (to overcome drag) -1, Pd will be negative, indicating the loss of power
to be compensated for by the swimmer. The same
principle holds for the propulsive force: again it
can be observed that the force and velocity vector
have approximately opposite directions. Hence, the

- -- swimmer also loses power to the water during the

push-off (as explained before):
Pk = Fp • Vh • cos (a) (Eq. 13)
\
\ where Vh is the velocity of the hand and a the angle
between the force and velocity vector. It is inter-
E=0.5m~v2 \ esting to note that when the push-off is made using
\
'drag propulsion' only, a will be 180° and hence
cos (a) will be -1. The more 'lift propulsion' is
Fig. 9. In swimming a fixed push-off point is not available. used, the more a will approach 90°, resulting in a
The propulsive force is generated by pushing off against water
decreasing cos (a). This implies that more lift pro-
and will equal the impulse (mAv) of the mass of water pushed
away. During the push-off, energy ('hmAv2) is transferred from pulsion should lead to decreasing values of Pk,
the swimmer to the water. hence increasing the propelling efficiency. This lat-
16 Sports Medicine 13 (1) 1992

ter observation was substantiated by Groot and van isation of flow in swimming propulsion might shed
Ingen Schenau (1988). Using data from the liter- more light on this interesting issue, as was previ-
ature they roughly estimated the propelling effi- ously suggested by Ungerechts (1987, 1988).
ciency and demonstrated that, depending on the
amount of lift force used, the propelling efficiency 4. Performance-Determining Factors
increases form 0.37 (only drag propulsion) to 0.56
when the lift/drag ratio is 1.4. This compares to an When we confine our analysis of front crawl
efficiency of 25% calculated by Martin et al. (1981) swimming to the balance of the propulsive and the
in a model study in which the hand and arm only drag force, the performance-determining factor will
produce drag forces. Again, this is in accordance be the capacity to generate high propulsive forces
with the observation that the capacity to generate while reducing resistance. If the mechanical power
lift forces seems to differentiate between the better output is also taken into consideration, swimming
and the poorer swimmers (Reischle 1979; Schlei- performance seems to depend on several addi-
hauf 1986; Schleihauf et al. 1988). tional factors.
The mechanical power output (Po) of the swim-
3.5 Propelling Efficiency and Vorticity mer depends on the rate at which aerobic and an-
aerobic metabolisms liberate energy from the stor-
The propulsive force is caused by momentum age compounds. The latter is the power input (Pi)
flow (equation 8). This fluid momentum can only of the 'swimming engine'.
be convected by vortex action. Irrespective of the If the total power input is fixed then the swim-
kind of propulsion (i.e. lift or drag), the propulsive ming velocity will depend on 3 major factors:
system generates wake vortices. Wake vortices can 1. The rate of energy liberation (Pi) necessary
be described by a distribution of vortex lines lying to generate mechanical power output (Po). This can
perpendicular to the local plane of rotation. The be expressed in terms of a ratio, which is the gross
vortex strength is determined by the rate of rota- efficiency (eg):
tion about the vortex line. Vortex lines must form (Eq. 14)
closed loops and, as explained above, vortex loops
2. The effective apportionment of this power
transport momentum. The magnitude of the mo-
output to power to overcome drag (the propelling
mentum carried by the loop is proportional to the
efficiency):
product of its area and the vortex strength. The
energy required to generate Ek is proportional to (Eq. 15)
the strength squared divided by the square root of 3. The relation between the swimming velocity
the area. Thus, the most efficient vortex with given (v) and the power required to overcome drag (Pd):
momentum is large in size and has low vortex Pd = Fd' v = A· v 3 (Eq. 16)
strength. According to Lighthill (1969) vortices of

=~
approximately circular shape carry a large amount
v=
of momentum in relation to their energy. These
types of vortex are created by propelling surfaces
(Eq. 17)
(tails) with a special lunate shape, and although the
comparison with the human arm in this context As a result of this analysis it seems as if the swim-
seems far fetched, it can be hypothesised that the ming velocity is determined by 5 major factors: drag
hand plus underarm do fulfil the requirement of a (A), power input (Pi) [aerobic/anaerobic metabol-
large span but reduced area for lift propulsion ism], gross efficiency (eg), propelling efficiency (ep )
(Rayner· 1985). This suggests that the key to pro- and power output (Po), For a detailed discussion
ducing effective propulsion at a low energy cost lies on the first 3 factors the reader is referred to TOlis-
in creating vortices of a special form. The visual- saint (1991).
Biomechanics of Front Crawl Swimming 17

4.1 Propelling Efficiency factor in swimming. However, before this conclu-

sion can be drawn with certainty, the study should
The first attempt to evaluate the significance of be replicated with a more homogeneous group of
propelling efficiency in swimming was made by competitive swimmers. Furthermore, research must
Toussaint (1990). Two highly endurance-trained focus on the relation between technique and pro-
groups - competitive swimmers (n = 6) and triath- pelling efficiency, where special attention should
letes (n = 5) - were compared. With the aid of be paid to the notion that the power wasted in gen-
individually determined regression equations the erating lift forces is considerably less than the power
groups were compared at equal levels of energy ex- in generating drag forces (Alexander & Goldspink
penditure representing a power input of 1000W ("'" 1977; Groot & van Ingen Schenau 1988; van Ingen
2.86L 02/min); thus every subject in both groups Schenau & Cavanagh 1990). The development of
provided an equal amount of power input to the training programmes to improve propelling effi-
'swimming engine' (table I). ciency seems to have great practical relevance.
The results in table I show that at an equal level Given the high correlation between propelling ef-
of power input (oxygen uptake) the triathletes swam ficiency and distance per stroke, it seems prudent
at a velocity of 0.95 m/sec, while the competitive to direct such training programmes at an increase
swimmers swam at 1.17 m/sec, 23% faster. This of distance per stroke, while maintaining the same
difference could not be explained by a difference speed.
in gross efficiency, stroke frequency or work per
stroke. But there was a difference in distance per 4.1.1 Propelling Efficiency and Hand
stroke (2.46 versus 1.84m). This was due to a dif- Surface Area
ference in propelling efficiency; 44% for the triath- The principle of aquatic propulsion is to give
letes and 61 % for the swimmers. In other words, an impulse to the aquatic surrounding (fig. 8)
the competitive swimmers used 61 % of the work [equation 8]. From equation 9 it can be derived
per stroke to overcome drag, while only 39% was that the same propulsive force can be generated at
converted into kinetic energy of water during push- different energy cost: giving a small amount of water
off. Therefore, the competitive swimmers were able a large velocity change costs more kinetic energy
to cover a greater distance per stroke than the than giving a large amount of water a small veloc-
triathletes, who in contrast wasted 56% of the ity change. This suggests that large propelling sur-
available energy per stroke in moving water. faces (a large hand area) implies a performance ad-
These results underscore the importance of pro- vantage in swimming. This can be derived
pelling efficiency as a performance-determining mathematically, since:

Table I. Competitive swimmers compared with triathletes when both exercise at a power input level of 1000W ('" 2.86 L D2/min)
[see Toussaint 1990]. Values are mean ± SD

Parameter Competitive Triathletes t-value p

swimmers

Power input (W) 1000 1000

Swimming velocity (m/sec) 1.17 ± 0.08 0.95 ± 0.11 3.87 0.004
Gross efficiency (%) 8.13 ± 0.94 8.04 ± 1.45 0.13 ns
Stroke frequency (sec· 1) 0.48 ± 0.09 0.53 ± 0.13 -0.82 ns
Work per stroke (J) 172 ± 38.2 152 ± 28.4 0.96 ns
Distance per stroke (m) 2.46 ± 0.42 1.84 ± 0.46 2.33 0.045
Propelling efficiency (%) 60.8 ± 6.1 43.6 ± 3.4 5.61 0.00

Abbreviations: ns = not significant.

18 Sports Medicine 13 (J) 1992

ep = Pct!(Pd + Pk) = (Fd' V)/(Fd • v + Fp' u) this sense paddle swimming might be a rather spe-
(Eq. 18) cific form of strength training.
and the drag forces (Fd) acting on the body will
depend on the square of the swimming speed (v) 4.1.2 Distance Per Stroke
according to: It is interesting to elaborate the close relation
between propelling efficiency and distance per
Fd = Ihpv2CdbAp (Eq. 19)
stroke observed in the previous study. At a given
with p the density of water, Cdb the coefficient of swimming velocity the distance per stroke (d) will
drag, and Ap the frontal area of the whole body. depend on the work per stroke (W) available to
Fp is related to the square of the hand speed (u) overcome drag (e p ' W) and the magnitude' of the
relative to the water according to (equations 6 drag force (Fd):
and 7) d = (e p • W)/Fd (Eq. 22)
Fp = Ihpu2SCh (Eq. 20)
The notion that the better swimmer distin-
then equation 19 works out into guishes him- or herself from the poorer one by a
(lhpv3CdbAp)/(lhpv3Cd~p +1f2pu3SCh)
greater distance per stroke is widely accepted, in-
dicating the effect of technique (e p) [Costill et al.
1985; Craig et al. 1979, 1985; Craig & Pendergast
1979, 1980; East 1971; Hay et al. 1983; Reischle
= (Eq. 21)
1979; Swaine & Reilly 1983; Toussaint et al. 1983].
1 + The possible effect of the capacity to deliver work
is illustrated in a study by Grimston & Hay (1986)
In this formula the effect of an increased pro- in which the physique of swimmers was related to
pelling surface size S is directly related to the pro- distance per stroke and stroke frequency. They
pelling efficiency ep. The latter was tested by meas- showed that the axilla cross-sectional area had the
uring the propelling efficiency before and after largest influence on distance per stroke, accounting
artificially enlarging the hands of competitive for 57% of the variance in the stroke length. The
swimmers with swim paddles (Toussaint et al. muscles used during the propulsive phase of swim-
1991). At the same average velocity the effect of ming include the shoulder extension muscles, pec-
swimming with paddles was an increase of ep of toralis major, latissimus dorsi and teres major.
8%. This is in line with other studies that showed These muscles are all situated in the axilla region
that swimmers of a high performance level have a and, acting together, might be expected to contrib-
significantly larger hand and arm surface than ute substantially to the observed distance per stroke
swimmers of lower performance level (Grimston of a swimmer. It is well known that there is a cor-
& Hay 1986; Toussaint et al. 1983). This finding responding increase in muscle strength with muscle
has some practical implications for swim training. hypertrophy. It seems reasonable to assume that an
Although for obvious reasons the use of hand pad- increase in size of the muscles involved in the pro-
dles in competition is not allowed, they may be pulsive phase of swimming should lead to greater
useful during training. Swimming with paddles at work generation during that phase and thus to a
the same speed forces the swimmer to deliver more longer distance per stroke.
work per stroke cycle. If maximal performance is In addition, the difference in speed between male
taken into account, at an equal power output a and female swimmers is mainly due to a larger dis-
higher swimming velocity can be attained, due to tance per stroke (East 1971; Letzelter & Freitag
the higher propelling efficiency. Hence, a higher 1983; Pai et al. 1984). Furthermore, the increase in
propulsive force must be applied since the drag speed with increasing age depends mainly on the
forces will increase at higher swimming speed. In increase in the distance covered during each stroke
Biomechanics of Front Crawl Swimming 19

(Miyashita 1975; Saito 1982). As pointed out ear- 200 - 0 - Power 50

lier by Groot & van Ingen Schenau (1988), both --*- Training
190 40
propelling efficiency and distance per stroke im- ~
CD
CD
prove when propulsion is mainly based on 'lift pro- 180
pulsion'. This is in line with the already observed
[
...CD
30 ~~
170 C)
experimental evidence that lift is the preferred ~
0 c
20 ·c
0.. 160 Ul
source of propulsion (Barthels & Adrian 1974; iij
c
0 10 ~
Schleihauf 1974, 1979; Schleihauf et al. 1983, 1988), 150 ~
while Reischle (1979) has reported that better z
140 0
swimmers actually rely more on lift propulsion. Al-
though the exact relations between propelling ef-
200 -0-- Power 4.7
ficiency, distance per stroke and lift propulsion re-
main to be determined, it seems reasonable to
conclude that distance per stroke gives a fairly good [
190
180
-----
Ul
iij
c
0
V02max
4.6
"2
·E
indication of propelling efficiency and may be used ~
z 4.5 ~

•
lii 170
to evaluate individual progress in technical ability. ~
0
t "E
x

160 4.4 .~'"

0..
Ul
iij
c
4.2 Power Output 150 0
~
z
140 4.3
The power output delivered by the swimmer is June December June
produced by the propelling muscles. In these pro-
pelling muscles the (aerobic and anaerobic) power Fig. 11. Power output of the arms measured during a full
year of training in relation to the training volume and VOZmax
input is converted into mechanical power output.
measured while swimming.
The importance of the power output of the arms
is emphasised in studies in which improvements
in swimming performance were associated with in- cated' in the trunk and arm muscles. This could
creased power output (Charbonnier et al. 1975; imply that the maximum capacity (,,"02max) of the
Costil1 et al. 1983, 1985; Sharp et al. 1982; Tous- specific 'swimming combustion engine' is not de-
saint & Vervoorn 1990). Although in this section termined by central circulatory factors, as may be
power output is discussed, it is interesting to come the case in treadmill running, but rather by the
back to the discussion concerning ,,"02max in the amount and quality of the muscle mass used in
context of competitive swiming. In some studies it swimming. Conversely, this could imply that the
was observed that the ,,"02max when swimming limited factor of the production of mechanical
with arms only did change with training, whereas power output is the quantity and quality of the
,,"02max determined in the same subjects using 'propelling muscles' rather than the capacity to
conventional methods (treadmill, bicycle) re- transport oxygen to the working muscles (Holmer
mained the same (Holmer 1972, 1974). Gergley et 1974).
al. (1984) compared the effects of swim exercise The results of a study concerning the effects of
versus treadmill exercise. The results support the a year's training on different performance-deter-
specificity of aerobic improvement with training mining factors might substantiate this idea (fig. II).
and suggest that local adaptations contribute sig- In a group of 11 elite swimmers (table II) the effect
nificantly to improvements in peak ,,"02. As of training was evaluated approximately every 6
pointed out earlier, the major part of the propul- weeks (June to June). Maximal oxygen uptake while
sion in front crawl swimming is produced by the swimming, maximal power output (W) of the arms
arms. Therefore, it is reasonable to assume that the (MAD system), and stroke characteristics (indicat-
'combustion engine' of the swimmer is mainly '10- ing propelling efficiency) were measured.
20 Sports Medicine 13 (1) 1992

Table II. Elite swimmers participating in a study of the effects 5. The Partitioning of Time, Space, and
of a year's training on various performance-determining factors Energy in a Swimming Event
(see fig. 11)

Parameter Males (n = 6) Females (n = 5) As intimated before, a competitive swimmer's

objective is to swim, without violating the swim-
Height (m) 1.85 ± 0.08 1.85 ± 0.02
ming rules, the full distance of his or her event in
Weight (kg) 79.6 ± 9.8 72.6 ± 8.6
V02max (L/min) 4.83 ± 0.5 4.01 ± 0.2
the least time possible. Hay (1988) suggested par-
100m time (sec) 53.4 ± 1.4 56.7 ± 1.3 titioning the time that a swimmer takes to com-
plete an event, i.e. the event time, into a starting
time, a turning time and a stroking time. This par-
titioning is useful in that it not only draws atten-
The maximal power output of the arms showed tion to the fact that a swimming event consists of
significant (p < 0.05) changes during the season, a number of subactivities, each of which may pro-
which seemed to be related to the training volume. vide opportunities to gain time, but it also opens
The overall increase in power output was 18% up the possibility of placing the local biomechan-
(p < 0.01). The maximal oxygen uptake increased ical analyses discussed so far in the more global
4% (p < 0.05) over the period December to June. context of the complete swimming event. Now that
The stroke characteristics showed no change. the factors that differentiate between failure and
Expressed as a percentage, the power output of success have become so subtle, the time is ripe to
the arms demonstrated the greatest response to begin to develop an awareness of the constraints
training: 18% increase in 1 year. In this group of that are placed by the type of swimming event on
highly trained swimmers (among them 3 Olympic the swimmer's activity. Swimming performance
medal winners) this was a surprising result. It may be improved considerably by an appropriate
underlines the importance of the capacity of the distribution of the amount of power produced dur-
arms to generate power output for swimming per- ing a race over the total distance of the event. The
decrease of stroke length in the course of races other
formance.
than sprints may be a reflection of a diminishing
The design of the experiment was such that it
capacity to deliver power output and, perhaps, also
was not possible to relate the increase in ,,"02max
of non optimal power output distribution. Simu-
while swimming directly to the size and quality of
lations of optimal power output have been made
the arm musculature (the propelling engine). How- for other cyclical activities such as skating on ice
ever, it was noticeable that during the 'aerobic (van Ingen Schenau et al. 1990) and running (Per-
phase' of the training (September to December) onnet & Thibault 1989), but to date no such models
,,"02max did not change significantly, and had a are available for speed swimming. Furthermore,
tendency to decrease. In the training period in swimming performance may be enhanced by fine
which more high intensity work was done (January tuning the integration of the subactivities of start-
to March), and especially in the period after the ing, turning and stroking in the total swimming
short course Nationals (April to June) during which event.
the training intensity was strongly increased at the Two problems in this context that have received
expense of training volume, both ,,"02max and virtually no attention in the literature may be dis-
power output increased significantly. This is in line tinguished: the allocation problem and the linkage
with Neufer et al. (1987) who observed increased problem. The allocation problem, which is related
values for ,,"02max after 5 months of training which to power distribution, involves the question of how
were no longer elevated when training was re- much space or time should be allocated to each of
duced. This was associated with a reduction in swim the component activities of swimming to achieve
power and distance per stroke. an optimal result. For instance, in sprint swim-
Biomechanics of Front Crawl Swimming 21

ming events it may be essential to cover the dis- of maximising the propulsion speed with reduced
tance to be swum with a dive of a certain length propelling surfaces, while stroking with 1 arm cre-
followed by an integer number of strokes of a cer- ated the opportunity to focus on the diagonality of
tain length such that no new stroke has to be ini- the stroking movements. The effectiveness of this
tiated just before the final touch. The linkage prob- approach may be evaluated by keeping track of
lem is the problem of the braiding together of the changes in the distance per stroke per individual.
component activities of swimming in a fluent man- After all, as observed earlier by Craig & Pendergast
ner, and requires a high degree of coordination to (1979), the ability to achieve a high swimming ve-
be solved seamlessly. locity is directly related to the maximal distance
per stroke. The individuals who covered the long-
6. Conclusions and Implications est distance per stroke at slow stroking frequency
for Training had the greatest maximal swimming velocity.
Therefore, Craig & Pendergast (1979) suggest that
The performance-determining factors of front swimming performance may be improved by prac-
crawl swimming reviewed in this article fall into 2 tising with slow stroke frequencies in order to
categories; those related to 'technique' and those achieve a longer distance per stroke. If a swim-
related to 'physical performance capacity'. In par- mer's distance per stroke is short, there is less lat-
ticular, the former includes the generation of high itude for shortening and a greater dependence on
effective propelling forces while reducing the as- stroke frequency to swim fast.
sociated energy cost (leading to a high propelling Swimmers can be challenged to swim at their
efficiency). The relation between the movements greatest distance per stroke possible by introducing
of the hands and the generation of propelling forces a little game. While swimming a rather short dis-
was studied by Schleihauf, who was unable to iden- tance (e.g. 25m), the time is taken and the number
tify an ideal stroking pattern: what is 'ideal' proved of strokes is counted. The sum of the 2 must be
very much a function of the individual. For train- smaller than before on each consecutive lap.
ing practice this implies that the evaluation of an Since the distance per stroke not only relates to
individual stroke technique is not a straightfor- propelling efficiency, but also to work per stroke
ward matter, as a universal standard applicable to (equation 22), it is not surprising that during a race
every individual is absent. In order to improve the distance per stroke decreases (Hay et al. 1983;
stroke technique it is therefore more important to Letzelter & Freitag 1983; Pai et al. 1984). The on-
have swimmers develop a 'feel' for the water rather set of fatigue probably reduces the capacity to swim
than attempt to imitate the stroking pattern of the at a high propelling efficiency as well as the ca-
world record holder. To enhance this 'feel' for the pacity to deliver a high amount of work per stroke.
water, one may opt for the variability of practice In training, therefore, emphasis should be placed
advocated by Schmidt (1988) in the context of mo- on the importance of maintaining a good tech-
tor skill acquisition. The 'motor problem' of swim- nique even when fatigue sets in. This, of course, is
ming, to use Berstein's (1976) apt term, may be easier to accomplish when the capacity of the
made transparent to the swimmer by introducing swimming engine is enlarged.
'deviations' from the normal situation, thus con- The importance of improving the power output
fronting him or her with new challenges and forc- of the swimming engine is illustrated in the study
ing them to look at the problem from another an- by Costill et al. (1986) in which swim power was
gle. The basic idea behind these manipulations is measured before and after 8 weeks of swimming
to introduce variations in the work space that al- and strength training. During this period front crawl
low the swimmer to discover the invariant prop- velocity increased 4%, whereas the mean power in-
erties of the water to the swimmer. For example, creased 9.6%, emphasising the contribution of
swimming with closed fists provides the challenge power output to swimming performance. To im-
22 Sports Medicine 13 (1) 1992

prove this power output the training programme conversely, from the scientist that the acquired
for elite competitive swimmers often includes dry- knowledge is packaged in an accessible format, and
land exercises. In comparison to the load during conversely, from the practitioner that the available
actual swimming these exercises should provide a knowledge is being applied conscientiously Hope-
greater resistance to the working muscles and hence fully, we have done our job.
increase maximal power output more effectively.
However, the movement patterns of the different References
swimming strokes are difficult to reproduce out-
side the water (Olbrecht & Clarys 1983; Schleihauf Alexander RM, Goldspink G. Mechanics and energetics of ani-
mal locomotion, Chapman and Hall, London, 1977
1983) and thus any training effect may only par- Alley LE. An analysis of water resistance and propulsion in swim-
tially, if at all, carry over to the competitive swim- ming the crawl stroke. Research Quarterly 23: 253-270, 1952
Andrew GM, Becklake MR, Guleria JS, et al. Heart and lung
ming performance. A new water-based training de- functions in swimmers and nonathletes during growth. Journal
vice derived from the MAD system providing fixed of Applied Physiology 32: 245-251, 1972
push-off points in the water for swimming the front Barthels K, Adrian MJ. Three dimensional spatial hand patterns
of skilled butterfly swimmers. In Clarys & Lewillie (Eds)
crawl was therefore used to specifically improve Swimming II, pp. 154-160, University Park Press, Baltimore,
power output. The effects of training on this device 1974
Bernstein NA. The co-ordination and regulation of movement,
were determined by comparing the increase in per- Pergamon Press, London, 1976
formance of a training group to a control group Bober T, Czabanski B. Changes in breaststroke techniques under
different speed conditions. In Claryss & Lewillie (Eds) Swim-
(Toussaint & Vervoorn 1990). Despite the fact that ming II, pp. 188-193, University Park Press, Baltimore, 1975
training time and volume were equal, the training Bone Q. Muscular and energetic aspects offish swimming. In Wu
et al. (Eds) Swimming and flying in nature, pp. 493-528, Plenum
group showed a significantly greater improvement Press, New York, 1975
in force (from 91 to 94N, 3.3%) velocity (from 1.75 Bucher W. The influence of the leg kick and the arm stroke on
the total speed during the crawl stroke. In Clarys & Lewillie
to l.81 mlsec, 3.4%), and power (from 160 to 172W, (Eds) Swimming II, pp. 180-187, University Park Press, Bal-
7%) as measured on the MAD system, and an in- timore, 1975
Charbonnier JP, Lacour JR, Riffat J, et al. Experimental study
crease in distance per stroke in free swimming. The of the performance of competition swimmers. European Jour-
training group showed a significant improvement ani of Applied Physiology 34: 157-167, 1975
Clarys JP. Human morphology and hydrodynamics. In Terauds
in race times for 50m (from 27.2 to 26.6 sec), 100m & Bedingfield (Eds) Swimimng III, pp. 3-43, University Park
(from 59.3 to 57.4 sec), and 200m (from 129.6 to Press, Baltimore, 1979
Costill DL, King OS, Holdren A, et al. Sprint speed vs swimming
127.3 sec). Related to the increase of power output power. Swimming Technique 20: 20-22, 1983
is an increase of work per stroke, which can also Costill DL, Kovaleski J, Porter 0, et al. Energy expenditure dur-
ing front crawl swimming; predicting success in middle-dis-
be achieved with less equipment-demanding train- tance events. International Journal of Sports Medicine 6: 266-
ing techniques, such as swimming with paddles and 270, 1985
Costill DL, Rayfield F, Kirwan J, et al. A computer based system
sprint swimming with long rest intervals. for the measurement of force and power during front crawl
In summary, it is clear from these examples that swimming. Journal of Swimming Research 2: 16-19, 1986
Counsilman JE. Science of swimming, Prentice-Hall, Englewood
the present understanding of the key mechanical Cliffs, New Jersey, 1968
and energetic variables in (competitive front crawl) Counsilman JE, The role of sculling movements in the arm pull.
Swimming World 10: 6-7, 1969
swimming is such that reasonably solid recom- Counsilman JER. The application of Bernoulli's principle to hu-
mendations for training practice can be made. They man propulsion in water. In Lewillie & Clarys (Eds) Swim-
ming I, pp. 59-71, Universite Libre de Bruxelles, Brussels, 1971
can be substantiated both by theoretical consider- Craig AB, Boomer WL. Gibbons JF. Use of stroke rate, distance
ations and by empirical data. Nevertheless, a word per stroke, and velocity relationships during training for com-
petitive swimming. In Terauds & Bedingfield (Eds) Swimming
of caution is always in order in such matters. Sci- III, pp. 265-274, University Park Press, Baltimore, 1979
entific knowledge can never prescribe a particular Craig AB, Pendergast DR. Relationships of stroke rate, distance
per stroke, and velocity in competitive swimming. Medicine
practice, but rather should be seen as a resource to and Science in Sports and Exercises II: 278-283, 1979
draw from in enhancing the practice in question. Craig AB, Pendergast DR. Relationships of stroke rate, distance
per stroke and velocity in competitive swimming. Swimming
This requires from the scientist that the acquired Technique 17: 23-29, 1980 •
knowledge is packaged in an accessible format, and, Craig AB, Skehan PL, Pawelczyk JA, et al. Velocity, stroke rate,
Biomechanics of Front Crawl Swimming 23

and distance per stroke during elite swimming competition. ming, pp. 315-322, Human Kinetics Publishers, Champaign,
Medicine and Science in Sports and Exercise 17: 625-634, 1985 III., 1983
Cureton TK. Factors governing success in competitive swim- Lewis ER, Lorch 0, Swim fin design utilizing principles of marine
ming: A brief review of related studies. In Clarys & Lewillie animal locomotion. In Terauds & Bedingfield (Eds) Swimming
(Eds) Swimming II, pp. 9-39, University Park Press, Balti- III, pp. 289-297, University Park Press, Baltimore, 1979
more, 1975 Lighthill MJ. Hydromechanics of aquatic animal propulsion. An-
de Groot G, van Ingen Schenau GJ. Fundamental mechanics ap- nual Review of Fiuid Mechanics I: 413-445,1969
pliled to swimming: Technique and propelling efficiency. in Maglischo EW, Swimming faster, Mayfield Publishing Cy, Palo
Ungerechts et aL (Eds) Swimming science V, pp. 17-30, Hu- Alto, 1982
man Kinetics Books, Champaign, Illinois 1988 Marconnet P, Spinel W, Gastaud M, et aL Evaluation of some
East OJ. Stroke frequency, length and performance. Swimming physiological parameters in swimming school students during
Technique 8: 68-73, 1971 a two year period. In Eriksson & Furberg (Eds) Swimming
Faulkner JA. Physiology of swimming. Research Quarterly 37: medicine IV, pp. 161-169, University Park Press, Baltimore,
41-54, 1966 1978
Faulkner JA. Physiology of swimming and diving. In Falls (Ed.) Martin RB, Yeater RA, White MK. A simple analytical model
Exercise physiology, pp. 415-446, Academic Press, new York, for the crawl stroke. Journal of Biomechanics 14: 539-548,1981
1968 McMurray RG. Comparative efficiencies of conventional and
Gergley n, McArdle WD, Dejesus P, et aL Specificity of arm super-swimfin designs. Human Factors 19: 495-501, 1977
training on aerobic power during swimming and running. Miller DI. Biomechanics of swimming. In Wilmore & Keogh (Eds)
Medicine and Science in Sports and Exercise 16: 349-354, 1984 Exercise and sport sciences reviews, pp. 219-248, Academic
Grimstone SK, Hay JG. Relationships among anthropometric and Press, New York, 1975
stroking characteristics of college swimmers. Medicine and Sci- Miyashita M. Method of calculating mechanical power in swim-
nece in Sports and Exercise 18: 60-68, 1986 ming the breast sroke. Research Quarterly 45: 128-137, 1974
Hay JG. The biomechanics of sports techniques, Prentice Hall, Miyashita M. Arm action in the crawl stroke. In Clarys & Lewillie
Inc., Englewood Cliffs, New Jersey, 1973 (Eds) Swimming II, pp. 167-173, UniversityPark Press, Bal-
Hay JG. The status of research on the biomechanics of swim- timore, 1975
ming. In Ungerechts et aL (Eds) Swimming science V, pp. 3- Neufer PO, Costill DL, Fielding RA, et aL Effect of reduced train-
14, Human Kinetics Books, Champaign, Illinois, 1988 ing on muscular strength and endurance in competitive swim-
Hay JG, Guimaraes ACS, Grimston SK. A quantitative look at mers. Medicine and Science in Sports and Exercise 19: 486-
swimming biomechanics. Swimming Technique 20: 11-17, 1983 490, 1987
Hollander AP, de Groot G, van Ingen Schenau GJ, et aL Con- Nomura T. The influence of training and age on "'02max during
tribution of the legs in front crawl swimming. in Ungerechts swimming in Japanese elite age group and Olympic swimmers.
et aL (Eds) Swimming V, pp. 39-43, Human Kinetics Publish- In Hollander et aL (Eds) Biomechanics and medicine in swim-
ers, Inc., Champaign, Illinois, 1988 ming, pp. 251-257, Human Kinetics Publishers, Champaign,
Hollander AP, de Groot G, van Ingen Schenau GJ, et aL Meas- III, 1983
urement of active drag forces during swimming. Journal of Olbrecht J, Clarys JP. EMG of specific dry land training for the
Sports Sciences 4: 21-30, 1986 frontcrawL In Hollander et aL (Eds) Biomechanics and med-
Hollander AP, Toussaint HM, de Groot G, et al. Active drag and icine in swimming, pp. 136-141, Human Kinetics, Champaign,
swimming performance. New Zealand Journal of Sports Med- Ill, 1983
icine 13: 110-113, 1985 Pai YC, Hay JG, Wilson BD. Stroking techniques of elite swim-
Holmer I. Oxygen uptake during swimming in man. Journal of mers. Journal of Sport Sciences 2: 225-239, 1984
Applied Physiology 33: 502-509, 1972 Peronnet F, Thibault G. Mathematical anlaysis of running per-
Holmer I. Physiology of swimimng man. Acta Physiologica Scan- formance and world running records. Journal of Applied Phys-
dinavica (SuppL): 407c, 1974 iology 67: 453-465, 1989
Holmer I. Efficiency of breaststroke and freestyle swimming. In Rackham GW. An analysis of arm propulsion in swimming. In
Clarys & Lewillie (Eds) Swimming II, pp. 130-136, University Clarys & Lewillie (Eds) Swimming II, pp. 174-179, University
Park Press, Baltimore, 1975 Park Press, Baltimore, 1975
Holmer I. Analysis of acceleration as a measure of swimming Rayner JMV. Vorticity and propulsion mechanics in swimming
proficiency. In Terauds & Bedingfield (Eds) Swimming III, pp. and flying vertebrates. Sonderforschungsbereich 230, Heft 4:
118-125, University Park Press, Baltimore, 1979 Natiirliche Konstruktionen Leichtbau in Architektur und Na-
Huijing PA, Hollander AP, de Groot G. Efficiency and specificity tur, Konstruktionsprinzipien lebender und ausgestorbener
of training in swimming: an editorial. In Hollander et aL (Eds) Reptilien (Principles of construction in fossil and recent rep-
Biomechanics and medicine in swimming, pp. 1-6, Human Ki- tiles), Universitiit Stuttgart and Tiibingen, 1985
netics Publishers, Champaign, III, 1983 Reischle K. A kinematic investigation of movement patterns in
Huijing PA, Toussaint HM, Clarys JP, et aL Active drag related swimming with photo-optical methods. In Terauds & Beding-
to body dimensions. In Ungerechts et aL (Eds) Swimming sci- field (Eds) Swimming III, pp. 127-136, University Park Press,
ence V, pp. 31-37, Human Kinetics Books, Champaign, III., Baltimore, 1979
1988 Saito M. The effect of training on the relationships among ve-
Kunski H, Jegier A, Maslankiewics A, et aL The relationship of locity, stroke rate and distance per stroke in untrained subjects
biological factors to swimming performance in top Polish jun- swimming the breaststroke. Research Quarterly 53: 323-329,
ior swimmers aged 12 to 14 years. In Ungerechts et aL (Eds) 1982
Swimming science V, pp. 109-113, Human Kinetics Books, Schleihauf RE. A biomechanical analysis of freestyle. Swimming
Champaign, III., 1988 Technique II: 89-96, 1974
Larsen OW, Yancher RP, Bear CLH. Boat design and swimming Schleihauf RE. A hydrodynamic analysis of swimming propul-
performance. Swimming Technique 18: 38-44, 1981 sion. In Terauds & Bedingfield (Eds) Swimming III, pp. 70-
Letzelter H, Freitag W. Stroke length and stroke frequency vari- 109, University Park Press, Baltimore, 1979
ations in men's and women's lOO-m freestyle swimming. in Schleihauf RE. Specificity of strength training in swimming: A
Hollander et al. (Eds) Biomechanics and medicine in swim- biomechanical viewpoint. in Hollander et aL (Eds) Biome-
24 Sports Medicine 13 (1) 1992

chanics and medicine in swimming, pp. 184-191, Human Ki- size on the mechanics and energetics of front crawl swimming.
netics, Champaign, Ill, 1983 Journal of Biomechanics, 24: 205-211, 1991
SchleihaufRE. Swimming skill: a review of basic theory. Journal Toussaint HM, Knops W, de Groot G, et al. The mechanical
of Swimming Research 2: 11-20, 1986 efficiency of front crawl swimming. Medicine and Science in
Schleihauf RE, Gray L, DeRose J. Three-dimensional analysis of Sports and Exercises 22: 402-408, 1990a
swimming propulsion in the sprint front crawlstroke. In Hol- Toussaint HM, de Looze M, van Rossem B, et al. The effect of
lander et al. (Eds) Biomechanics and medicine in swimming, growth on drag in young swimmers. Journal of Sport Bio-
pp. 173-184, Human Kinetics Publishers, Champaign, Ill., 1983 mechanics 6: 18-28, 1990b
Toussaint HM, Vervoorn K. Effects of specific high resistance
Schleihauf RE, Higgins JR, Hinrichs R, et al. Propulsive tech-
training in the water on competitive swimmers. International
niques: front crawl stroke, butterfly. backstroke, and breast-
Journal of Sports Medicine II: 228-233, 1990
stroke. In Ungerecths et al. (Eds) Swimming science V, pp. 53-
Ungerechts BE. Optimizing propulsion in swimming by rotation
59, Human Kinetics Books, Champaign, Ill, 1988 of the hands. In Terauds & Bedingfield (Eds) Swimming III,
Schmidt RA. Motor control and learning, Human Kinetics, pp. 55-61, University Park Press, Baltimore, 1979
Champaign. Ill, 1988 U ngerechts BE. On the relevance of rotating water flow for the
Sharp RL, Troup JP, Costill DL. Relationship between power and propulsion in swimming. In Jonsson (Ed.) Biomechanics X-B.
sprint freestyle swimming. Medicine and Science in Sports and pp. 713-716, Human Kinetics Publishers, Champaign, Ill, 1987
Exercise 14: 53-56, 1982 Ungerechts BE. The relation of peak body acceleration to phases
Svec O. Biofeedback for pulling efficiency. Swimming Technique of movements in swimming. In Ungerechts et al. (Eds) Swim-
19: 38-46, 1982 ming science V, pp. 61-66, Human Kinetics Books, Cham-
Swaine I, Reilly T. The freely-chosen swimming stroke rate in a paign, Ill., 1988
maximal swim and on a biokinetic swim bench. Medicine and Vaart AJM van de, Savel berg HHCM, de Groot G. et al. An
Science in Sports and Exercise 15: 370-375, 1983 estimation of active drag in front crawl swimming. Journal of
Toussaint HM. Differences in propelling efficiency between com- Biomechanics 20: 543-546, 1987
petitive and triathlon swimmers. Medicine and Science in Sports van Ingen Schenau GJ, Cavanagh PRo Power equations in en-
and Exercises 22: 409-415, 1990 durance sports. Journal of Biomechanics 23: 865-881, 1990
van Ingen Schenau GJ, de Koning JJ, de Groot G. A simulation
Toussaint HM. Performance determining factors of front crawl
of speed skating performances based on a power equation.
swimming. In Maclaren et al. (Eds) Biomechanics and med-
Medicine and Science in Sports and Exercise 22: 718-728, 1990
icine in swimming, E&FN Spon, London, in press, 1991 Watkins J, Gordon AT. The effect of leg action on performance
Toussaint HM, Beelen A. Rodenburg A, et al. Propelling effi- in the sprint front crawl stroke. In Hollander et al. (Eds) Bio-
ciency offront crawl swimming. Journal of Applied Physiology mechanics and medicine in swimming, pp. 310-314, Human
65: 2506-2512, 1988a Kinetics Publishers, Champaign, Ill., 1983
Toussaint HM, Bruinink L, Coster R, et al. Effect of a triathlon Webb PW. The swimming energetics of trout I: thrust and power
wet suit on drag during swimming. Medicine and Science in output at cruising speeds. Journal of Experimental Biology 55:
Sports and Exercises 21: 325-328, 1989 489-520, 1971a
Toussaint HM, de Groot G, Savelberg HHCM, et al. Active drag Webb PW. The swimming energetics of trout II: oxygen con-
related to velocity in male and female swimmers. Journal of sumption and swimming efficiency. Journal of Experimental
Biomechanics 21: 435-438, 1988b Biology 55: 521-540, 1971b
Toussaint HM, Helm FCT van der, Elzerman JR, et al. A power
balance applied to swimming. In Hollander et al. (Eds) Bio-
mechanics and medicine in swimming, pp. 165-172, Human Correspondence and reprints: Huub M. Toussaint, Department
Kinetics Publishers. Champaign, Ill., 1983 of Health Science, Faculty of Human Movement Sciences, vd
Toussaint HM, Janssen T, K1uft M. Effect of propelling surfaces Boechorststraat 9, Amsterdam, The Netherlands.