<< Back to Swimming

As the sport of competitive swimming developed globally throughout the mid 1900s the demand for knowledge about how one can move proficiently in the water increased. Acquiring information on propulsion has numerous benefits. For example, it provides insight into the technique used to execute the swimming skill and thus aids in the process of stroke analysis and improvement. By knowing the ‘mechanisms’ of propulsion coaches and teachers can be optimally effective in improving stroke technique and performance. The knowledge can be applied to help elite swimmers break existing records, recreational swimmers to be able to swim with less effort, and most importantly to become more efficient in the water. Additionally, instruction given to learning swimmers is more likely to be correct from the beginning and an economical technique developed before faults become ‘bad habits’ resistant to change.

Many ideas have been proposed in an attempt to describe how elite swimmers propel themselves through the water. In relatively recent times ‘long-standing’ and credible ideas have been discredited. For example, the idea that swimmers ‘scull’ to propel the body using lift in accordance with the Bernoulli Principle has been seriously questioned and rejected by many. However, inappropriate ideas are still common among swimming coaches and teachers. Alternative ideas based on new scientific knowledge are gradually being considered.

This article presents a brief review of past and present ideas regarding the possible mechanisms by which swimmers propel themselves.

Action-Reaction

Prior to the late1960s Newton’s 3rd law, ‘For every action there will be an equal and opposite reaction’, was a popular explanation of how one progressed in the water. Consequently, swimmers where instructed to pull the arm straight backwards following entry. It was reasoned that pushing the water back created a counterforce of equal magnitude that propelled the swimmer forward. This propulsive force was determined by the mass of water and the acceleration of the water in accordance with Newton’s equation:

F=ma

As well as the amount of water (mass) accelerated, the time over which it was accelerated was important because the change in motion (momentum) of a body is equal to force * time. To apply force for a relatively long time, it was recognised that swimmers had to push the water through a large distance. Newton’s 3rd law provided a very simple and rational description of propulsion in swimming.

However, it was later recognised that this explanation on its own was insufficient to guide the development of good technique. It was realised that once swimmers started a mass of water moving, they could no longer elicit the same reaction force from it without substantially increasing their hand speed. This is because if the water and body segment are moving in the same direction, i.e. the water is moving with the hand, the speed relative to the water is reduced resulting in a reduction in the propulsive force. Hence, stroking with a straight arm backwards would require the swimmer to do excessive ‘work’ on the water. Thus, the idea of pulling the arm straight backwards was recognised as not being the most efficient way to produce propulsion.

In the late 1960s, Dr James Counsilman used underwater video analysis and observed that top swimmers were not pulling straight backwards, but in somewhat curvilinear or ‘S’ shape patterns. Dr Counsilman proposed that swimmers stroked in this manner, to continuously find ‘still’ water that allowed the swimmer to achieve a greater reaction force than when they pulled directly backwards. Because the majority of elite swimmers were observed to be stroking in a curved hand path researchers sought to explain how and why curved paths were more effective than straight paths.

Rushall et al., 1984, commented that the swimming community was divided over the issue of explaining propulsion in swimming. Some, such as Charles Silvia, continued to cite Newton’s 3rd law and proposed that the arm and hand was still travelling predominantly backwards (Rushall et al., 1984). He suggested that by manipulating the shoulder and elbow joints, the swimmer could find ‘still’ water to accelerate throughout the stroke. Regardless, the majority of people sided with the influential Counsilman, who borrowed the Bernoulli Principle from aerodynamics to explain why swimmers used curved hand paths.

Propulsion from the Bernoulli Principle

When oncoming fluid contacts an object that has one side more curved than another (i.e. an airfoil) it travels faster over the more curved side than the less curved side (Fig 1). This is due to the flow over the more curved side being ‘squeezed’ through a smaller cross sectional area than the airflow passing along the less curved surface. From the Principle of Continuity, when flow is ‘squeezed’ into a smaller area, it travels faster.

Fig. 1. Illustration of fluid flow travelling faster over the more curved side
of the foil, compared to the less curved side. Also note that fluid flow
is ‘squeezed’ into a smaller area above the foil.

When a fluid increases its speed, its pressure is reduced. Therefore, as the fluid flows around a foil a pressure differential is created, whereby the pressure is less on the more curved side than on the less curved side. This difference in pressure results in a force in the direction from high to low pressure. Thus, there is a substantial force at right angles to the direction of fluid flow. This is called a ‘lift’ force (Fig 2).

Fig 2. Illustration of how lift force is created

Dr Counsilman proposed that the hand could behave like an airfoil to generate lift force. Since lift always acts perpendicular to the direction of fluid flow, the swimmers could produce lift that had a component in the desired direction of travel, by continuously adjusting the pitch of the hand in a sculling motion.

However, increasingly, the possibility that lift is generated by using the hand as a foil has been seriously questioned. Three main concerns have been raised in relation to the Bernoulli Principle:

• Lift forces only are considered for propulsion. This Principle cannot predict or explain the propulsive effect of drag forces (Rushall et al., 1994). Drag forces are those acting in the direction of flow (opposite the direction of motion of the object relative to the fluid). Regardless of lift being generated by the Bernoulli Principle, there is always a drag force present. Depending on the orientation of the hand with respect to the flow, this drag force may also contribute to propulsion.
• The shape of the hand and forearm need to be that of an asymmetrical wing. The forearm is known to play a considerable role in propulsion and is not a lifting surface (Rushall et al., 1994; Maglischo, 2003)
• The Bernoulli model assumes that energy in the fluid system is conserved. However, if turbulence is present or if the boundary layer is changed energy is dissipated (Bixler, 2002; Cabe, 2000). Therefore, an intact or attached boundary layer is essential for lift forces to be produced efficiently by the Bernoulli Principle (Maglischo, 2003). However, many studies have illustrated that the boundary layer separates from the hand (Valiant et al., 1982; Holt and Holt, 1989; Bixler, 1999, Toussaint et al., 2000, Bixler and Riewald, 2002) even at low speeds (Ferrell, 1991). Cabe (2000) suggested that the surface area of the hands and feet are neither large enough nor curved enough to produce the necessary lift to move a swimmer through the water.

The aforementioned problems do not mean that lift force is not important in contributing to propulsion in swimming, but it does highlight that the Bernoulli Principle on its own does not account adequately for the lift forces produced in swimming.

One can think of force generation in fluids in different ways. In considering the Bernoulli Principle it is common to visualise the lift forces as a result of pressure differences across the object. However one can also think of the forces as arising from acceleration of a mass of water. Thus quite recently, Newton's 2nd and 3rd laws of motion have gained credibility to explain how lift force is generated and productive in swimming. Sprigings and Koehler, 1990, illustrated that as a fluid moves around a foil, the flow is accelerated downward. From Newton’s 2nd law [F= ma] the force is directly proportional to the acceleration of the fluid. The force accelerating the fluid downward must be accompanied by an equal and opposite force (Newton’s 3rd law) pushing the airfoil upward (Fig. 3).

Fig 3. Newton’s 2nd & 3rd laws of motion to explain lift force

The main advantage of applying Newton’s laws as an explanation for lift, is that they can account also for drag forces. From a practical point of view, this is more powerful and useful since good swimming technique calls for maximising the total forward force comprising contributions from both lift and drag force components.

Realising that drag was a contributor to forward propulsion, Dr Counsilman appropriately modified the view that propulsion is due mostly to lift. He believed that it is important to understand the relative contributions of lift and drag when assessing the effectiveness of the propulsive movement .

Woods (1977) indicated that propulsion is a result of a combination of both lift and drag forces, with the relative contribution of lift and drag changing throughout the stroke. Woods’ research indicated that during times of acceleration and propulsion, drag makes the greater contribution due to the hand moving in a predominantly backwards direction. During some of the decelerative phases, lift played a greater role than drag.

Maglischo (1989) added that not every movement in the stroke is propulsive. However, lift forces are minor and overshadowed by drag forces during the times that a swimmer is accelerating. Indeed, the majority of previous and current research supports Maglischo’s view that both lift and drag are integrated within the swim stroke and interchange throughout the stroke depending on the stroke phase.

How Should a Coach Demonstrate the Freestyle Pulling Action?

In light of the above what is deemed a technically appropriate poolside demonstration of the freestyle arm action has changed. Play the video clips below and assess which you think is the most appropriate demonstration given current thinking about how propulsion is produced by good swimmers.

To download or play this video, click one of the following buttons:

A

In clip A, the instructor pulls the arm straight downwards. This is not unlike the pull of many beginning swimmers who use a straight arm technique. Below we can see what this looks like from the side when applied by a swimmer in the water. We can see that water would be pushed backwards for only a small part of the stroke. At the beginning of the stroke water is being pushed downwards.

To download or play this video, click one of the following buttons:

Staright Arm Pull

This will not contribute much to propulsion and will tend to rotate the body so that the legs ‘sink’, thereby increasing drag. At the end of the stroke the water is being pushed upwards rather than backwards. Again, this does not contribute much to propulsion. In short, the technique demonstrated in A is not appropriate.

To download or play this video, click one of the following buttons:

B

In movie B, the instructor is bending the elbow so that the hand and forearm are pushing backwards for a longer period than in movie A. By viewing a good swimmer from the side, as in the movie below, we can see that this action results in a ‘flatter path’ than when a straight arm is used.

To download or play this video, click one of the following buttons:

Pull of a Good Swimmer

However neither of the demonstrations shown in Movie A and Movie B take body roll into account. If we did the same action in B but incorporated the body roll that is normally used by swimmers then we find that the path of the hand is very curved rather than moving backwards (Movie C).

To download or play this video, click one of the following buttons:

C

The front view of a good swimmer shown in the movie below indicates that even though the body is rolling the path of the hand is mostly backwards rather than with large curved paths as was indicated by the demonstration in Movie C.

To download or play this video, click one of the following buttons:

Front View of a Good Swimmer

In demonstration D, the instructor incorporates the body roll but moves the hand and forearm in a path more like that used by good swimmers. This means that the movement compensates for the body roll so that the path of the hand is mostly backwards.

To download or play this video, click one of the following buttons:

D

Demonstration D is the most correct and appropriate demonstration of the four demonstrations shown.

Should the Path of the Hand be Completely Straight Backward?

The answer to this question is NO! Forces in the forward direction are generated by accelerating water backward. As discussed above, this can be achieved by lift as well as drag. The hand can be angled and moved slightly laterally as well as backwards to be effective in generating force by drag and some lift. This is actually better than pulling the arm and forearm directly backward because once the water is moving it becomes less efficient to keep accelerating it. It is like trying to walk up a sand hill and the sand keeps moving under your feet. You use a lot more energy and don’t move as far. Small deviations from a straight-line path allow the hand and forearm to find ‘still water’ to accelerate.

There are other reasons why slightly curved paths and changes of direction are useful. Some of these reasons are related to using the muscle and lever systems more effectively and economically and to aid continuity of motion (see for example the article on this site insert link). Another may be to generate and use vortices for propulsion.

Propulsion from Vortices

Quite recently Colwin (2002) stated that “instead of belabouring the lift versus drag argument, we need to move on and learn more about the way water reacts when we swim”. Many researchers currently share this view and have observed traces of water patterns left in the swimmer’s wake. In particular, efficient swimmers have been noted to ‘leave’ masses of circulating water (vortices) with axes of rotation perpendicular to the direction of travel of the swimmer.

A vortex forms as a reaction to the propulsive impulses generated by the swimmer (Colwin, 2002). Arellano, 2002, added that these flow formations carry a certain amount of momentum that is transferred from the body to the water. Researchers have put forward the ‘vortex theory’ as a means of explaining how lift forces can play a major role in swimming propulsion, even when the flow conditions are unsteady and when the boundary layer (layer of water molecules in contact with the surface of the object around which the water is moving) ‘separates’ from the swimmer’s limbs as they move through the water.

Much knowledge about vortices has come from research of marine animals shedding vortices in their wake as they swim. It has been illustrated that the vortices are continually created and shed in pulses as a fish travels through the water, as they change the direction, angle of its tail, or body alignment.

It is now known that birds and insects also employ mechanisms involving the rapid generating and shedding of vortices (Fig. 4), which enables them to generate lift more quickly than would be possible in steady airflow.

Fig.4 Illustration of how insect generates two vortices on each wing and then sheds one big vortex at the end of the propulsive action. Adapted from Colwin, 2002

The process of developing and shedding a vortex begins with the formation of a ‘starting vortex’, whereby water molecules roll up toward the more curved side of the foil because the water pressure is lower on that side than the other (Fig. 5).

Fig. 5 Creation of starting vortex, which in turn induces a bound vortex, with the purpose of increasing pressure differential. Adapted from Maglischo, 2003

Consequently, due to the rules of fluid dynamics and Newton’s 3rd law; the formation of a starting vortex induces a countervortex of equal strength that travels in the opposite direction to the starting vortex. Otherwise known as the ‘bound vortex’, this acts like a layer of fluid that circulates clockwise around the foil, hence in the same direction as the flow over the foil and opposite to the flow underneath the foil. As a result, the speed of flow over the top of the foil is increased, therefore decreasing the pressure above the foil. At the same time, the rate of flow under the foil is reduced, resulting in an area of increased pressure. Consequently, the net effect of the bound vortex is that the pressure differential, needed for lift force, between the top and bottom of the foil is greatly increased. Therefore, the role of the bound vortex continues to emphasise the pressure differential, and thus lift force, until the starting vortex is shed and ‘washed away’.

The process is complete when the vortex is shed, as this event indicates the end of each propulsive impulse, within a swimming stroke in a particular direction (Colwin, 2002). The swimmer's actions move water, thereby transferring kinetic energy to it. As a reaction, the swimmer recaptures this energy from their own vortex, which is essentially their propulsive force that thrusts them forwards in the water. Ungerechts et al. (1999) suggested that the vortex represents the transfer of momentum between water and body, and vice versa, resulting in body translation. Therefore, after the shed vortex is displaced behind the swimmer, it provides a momentum transfer to the body (Ungerechts et al 1999). The transfer of an impulse directed back to the body coming with the motion of water is said to occur ‘online’ (Ungerechts et al., 1999). In other words, the movements made by the swimmer are directly connected to the movements of the water. Thus when the swimmer moves their limbs in the water, it will directly act back upon the swimmer almost instantaneously. This phenomenon is also seen in aquatic animals, whereby the backward momentum of the vortex rings, corresponds to the forward momentum gained by the fish (Müller et al., 1997)

Therefore, shed vortices indicate a ‘history’ of the swimming stroke representing the swimmers propulsion (Colwin, 2002). Colwin (2002) commented that the size, shape, direction, velocity and placement in the flow field reveal characteristics of the swimmer’s propulsion. For example, efficient swimmers are believed to elicit vortices that are perpendicular to the direction of travel, rotate in the same place without displacement and are symmetrical, well rounded rings (Arellano, 2002).

Some authors, such as Colwin 2002, believe that the timing of vortex shedding is important in optimising technique. For example, incorrect timing of directional changes, or directional changes that are too sudden, may reduce the effectiveness of the swimmer's propulsion. Furthermore, common causes of vortices being shed prematurely are holding the hand too rigidly or too sudden a directional change combined with excessive acceleration and application of force (Arellano, 2002; Colwin, 2002).

Ungerechts et al. (1999) suggest that large wakes, like behind a paddling blade or hand moved perpendicular to its path, create high drag values (equivalent to low pressure) but waste large amounts of energy. Alternatively vortices rotating in the same place indicate least drag. This information may be of some value when giving feedback to swimmers. The swimmer should strive not to ‘push’ the vortex in a paddle type action, as this may lead to a premature shed vortex, and thus wasted energy. Rather the swimmer is recommended to control their limb movements throughout the stroke so that shed vortex is done so at the appropriate time. Therefore, theoretically, an ‘organised’ vortex requires less energy than ‘paddle type’ actions of hands. Hence, the aim when applying the vortex theory is to “control the energy, not spending it…. and not to execute the movements too forcefully” (Ungerechts et al 1999). Instead, swimmers are advised to carefully manipulate the hand throughout the stroke in terms of speed and direction that will result in the best propulsion (Colwin 2002, Ungerechts et al 1999, Maglischo 2003).

Fig 6 & 7 illustrate some examples of a swimmer’s pulling pattern, in relation to vortex generation and shedding. Note that in Fig 6 two vortices are shed, one in the middle of the stroke and the second at the end of the stoke; Colwin, 2002, suggested that this is a typical high speed pulling pattern. Fig 7 displays one shed vortex at the end of the stroke, which Colwin stated is more typical of a long distance stroke.

Fig. 6 High Speed pattern, two vortices shed. Adapted
Colwin, 2002

Fig 7. Slower pace, smoother stroke, suitable for longer distance. Adapted from Colwin, 2002

Therefore, the implications of the vortex theory in the context of coaching and teaching are that swimmers should not be told to stroke in a ‘S’ shape pattern in the water. Rather swimmers should be encouraged to feel pressure and differential pressure through the pressure sensitive cells and kinaesthetic proprioceptive system. By doing so, the swimmer becomes better at controlling the direction of resultant force as their ability to feel difference between pushing water and applying effective force is increased.

The main problem with the vortex theory is that it has never really been tested to show that these vortices in the water are associated with propulsive impulses and should thus be treated with caution.

Propulsion from Axial Flow

Another mechanism of propulsion proposed recently by Toussaint (2002) is propulsion from axial flow. Toussaint has dubbed his idea ‘pumped up propulsion’.

The concept and principle of this theory is very different from the previous theories presented. It is concerned with the rotation of the arm about the shoulder, which leads to a fluid velocity gradient along the arm, so that the (tangential) velocity near the hand will be higher than near the elbow. In turn a pressure gradient is induced, due to principles of Bernoulli’s Law, where local pressure close to the limb decreases in the direction of the fingertips. It is this pressure gradient that induces an axial flow of fluid along the arm and hand toward the fingertips. Thus, rotating the limb (e.g. the arm) leads to a pressure gradient ‘pumping’ fluid along the arm toward the hand on both the front and backsides of the arm (Fig. 8)

Fig 8. Illustration of pressure gradient along the arm as the swimmer rotates the arm
about the shoulder. Adapted from Toussaint et al., 2002

Furthermore, translation through the water results in a greater pressure on the palmar (leading) side of the hand and a lower pressure on the dorsal (trailing) side of the hand. The net interaction of these two events: rotation and translation of the arm, is that the pressure gradient on the trailing side is steeper than on the leading side. Therefore, during the outsweep of the stroke, ‘suction’ is boosted on the trailing side increasing the pressure differential across the propelling surface (the hand) thereby increasing the propulsive force (Fig. 9)

Fig. 9. As the swimmer rotates and translates the arm through the water, the ‘pumping’ effect increases the propulsive force. Adapted from Toussaint et al., 2002

Therefore, briefly, this theory illustrates that a rotating arm during the outsweep acts as a pump, driving water along the trailing side of the arm toward the hand, increasing the pressure difference and consequently propulsion.

Toussaint et al. presented evidence of a pressure gradient acting along the direction of the arm. It was also observed that as the swimmer stroked through the outsweep phase, a ‘V’ shaped convergence of the fluid flow on the trailing side of the forearm emerged, which suggests a contraction and acceleration of the flow along the arm towards the hand (Fig 10).

Fig 10. ‘V’ type convergence of tufts on arm, showing acceleration of fluid flow toward the hand. From Toussaint et al 2002 with permission.

Furthermore, as the flow approaches the wrist, it splits into 2 paths: towards the thumb and towards the fingertips aligning to the movement direction of the hand (Fig 11). This observation suggests that the local flow around the hand is influenced by the rotation of the forearm, such that fluid is pumped in the direction of the hand.

Fig 11. Tufts splitting into two paths when approaching wrist. From Toussaint et al 2002 with permission

The implications of the ‘pumped up propulsion’ concept for teaching and performance development are somewhat unclear at the present time. This is due to the fact that this relatively innovative concept requires further exploration to explore the effect of rotational motion about axes other than the shoulder joint, for example, the elbow joint. However, the fundamental logic and evidence provided for this unique explanation for propulsion is rather intriguing. Additionally, it might apply to the leg action, due to the rotation of the hip, knee and ankle in front crawl, back crawl and butterfly swimming.

Which Mechanism Do Swimmers Use Most?

In conclusion, it remains unclear which mechanism is most responsible for propulsion in human swimming. However, there is strong evidence that generation of lift by sculling the hand at small angles of attack to generate lift in accordance with the Bernoulli Principle is not the main mechanism of propulsion. Combinations of lift and drag forces, with drag being the more dominant throughout most of the stroke appear likely for front crawl, butterfly, and backstroke. One might expect that the role of lift in breaststroke is more substantial than in the other strokes. There is a growing body of evidence that production and shedding of vortices plays an important role in propulsion. That propulsive force is generated by axial flow in accord with the ‘pumped up propulsion’ hypothesis is an exciting new possibility. Indeed propulsion in swimming could be an interaction of various mechanisms. There may even be others awaiting discovery!

References:

• Arellano, R., Pardillo, S., Gavilán, (2002) Underwater undulatory swimming: kinematic characteristics, vortex generation and application during the start, turn and swimming strokes. Universidad de Granada: ISBS 2002 Proceedings
• Bixler, B. and Riewald, S. (2002) Analysis of a swimmer’s hand and arm in steady flow conditions using computational fluid dynamics. Journal of Biomechanics, 35: 713-717
• Cabe, D.K. (2000) Going through the motions. In Scientific American, 283(3): 64
• Colwin, C.M. (2002) Breakthrough Swimming. Human Kinetics
• Maglischo, E.W. (1989) The basic propulsive sweep in competitive swimming. Proceedings of the VIIth International Symposium of Biomechanics in Sports. Melbourne, Victoria, Australia.
• Maglischo, E.W. (2003) Swimming Fastest. Human Kinetics
• Müller, U.K., Van Den Heuvel, B.L.E., Stamhuis, E.J., and Videler J.J. (1997) Fish foot prints: morphology and energetics of the wake behind a continuously swimming mullet (Chelo labrosus risso). The Journal of Experimental Biology, 200, 2893-2906
• Rushall, B.S, Holt, L.E, Sprigings, E.J, & Cappaert, J.M. (1994) A Re-evaluation of Forces in Swimming. Journal of Swimming Research, 10: 6-30
• Sprigings, E.J, & Koehler, J.A. (1990) The choice between Bernoulli’s or Newton’s model in predicting dynamic lift. International Journal of Sports Biomechanics, 6:235-245.
• Toussaint, H.M., Van den Berg, C. and Beek, W.J. (2002) “Pumped-Up Propulsion” during front crawl swimming. Medicine and Science in Sports and Exercise, 34(2): 314-319
• Ungerechts, B.E., Persyn, U., and Colman, V. (1999) Application of vortex flow formation to self-propulsion in water. In K.L. Keskinen, P. Komi, and A.P. Hollander (Eds), Biomechanics and Medicine in Swimming VIII (pp. 95-100) Jyvaskla: Gummerus Printing House.