Sprintic Magazine

Subscribe to the Sprintic Magazine Newsletter and receive track and field news, articles, training descriptions, photo and video sections updates. We have 1800+ subscribers already!


Contributions of the ankle and knee muscles to sprint starting.


Starting is an important component of the sprint race, especially the 100 m dash, since any errors made will affect a significant proportion of the race. The sprint start has been extensively analyzed (e.g., Dickinson, 1934; Henry, 1952; Baumann, 1976; Gagnon, 1976; Desipres 1973) yet there is little information available as to which muscle groups contribute to the motion, how much and how does each contribute, and in what order. It was the purpose of this research to investigate these aspects of sprint starting using link segment indirect dynamics.

Review and theory

Link-segment indirect dynamics is a method for computing net forces and moments of force at the joints from body segment parameters and kinematics derived from filmed human motions. It is assumed that the computed moments of force are primarily caused by muscular contractions and, thus, inverse dynamics offers an excellent tool for evaluating muscle functioning. In actuality moments of force may also be caused by ligaments, joint capsules, bones, and other structures. The net moments of force represent the summed moments of force of all the structures that act across a joint, whether or not they cross more than one joint.

Unfortunately, solving for actual muscle functions is usually impossible because in human motion analyses most systems of equations are mathematically indeterminate. Thus, the concept of the “single equivalent muscle” is used to simplify the real situation. This concept assumes that there is a pair of “muscles” at each joint– one extensor and one flexor. It is also assumed that pairs of muscles act in reciprocal inhibition, that is, only one of the pair can be active at any given instant.

A single equivalent muscle can be said to contract concentrically when its action (i.e., moment of force) assists the joint’s angular motion (i.e., directions are the same). Conversely, the muscle can act eccentrically when its moment resists the joint’s motion. These characteristics may be determined by computing the powers produced by the moments of force. These powers are calculated by taking the product of moment of force times joint angular velocity. When the power is positive then the moment is said to be concentric. When the power is negative the moment is an eccentric one. Integration of the power over time yields the work done by the moment or the energy dissipated (Elftman, 1939; Winter, 1979; Robertson & Winter, 1980).


Two subjects who compete nationally in the sprints or sprint hurdles were filmed at 50 fps in a laboratory while starting from a Kistler force platform. Each subject was filmed starting with only one foot on the force plate. Three starts of each foot were filmed with the best trials of each subject being selected for linked- segment analysis.

The subjects ran with their regular running attire and spiked footwear on a material that is used for indoor running tracks. A cutout piece of the material was affixed to the force platform in such a way as to be isolated from the rest of the indoor track. This was necessary to obtain accurate horizontal forces. Previous tests have shown that as much as 50% of the horizontal force can be lost if the indoor track extends in one piece across the force platform.

The subjects used special indoor starting blocks that were separate for each foot. During the filming the athletes were allowed to prepare as is usual in a sprint race. The athletes were instructed to sprint as fast as possible for the first seven metres. As incentive their times over this distance were recorded and reported to them. The athletes, who regularly trained on the same track surface, were allowed sufficient practice trials to accommodate to the laboratory setting.

The film data after processing as projected onto a table equipped with a Numonics digitizer interfaced to microNova computer. Markers at the approximate joint centres of the shank and foot segments were digitized to an accuracy of 0.3 mm. These data were subsequently transferred along with the synchronized force recordings to a mainframe computer for processing by the computer software, BIOMECH (Kinesiology Dept., Univ. of Waterloo). This package performed the aforementioned inverse dynamics to yield the powers, work, moments of force, and joint angular velocity histories of the ankles and knees.

Result and discussion

Angular kinematics. The two athletes exhibited similar kinematic patterns at both joints and for both legs. Both athletes exhibited slowly increasing plantar flexion and knee extension peaking just prior to toe-off. The peak velocities of these joints occurred almost simultaneously in contrast with the notion that the sequence of joint movements should be knee first then ankle.

These movements were followed immediately by rapidly rising dorsiflexion and knee flexion. The most dramatic pattern being the knee flexion which reached velocities exceeding 15 rad/s.

Ankle kinetics. The two athletes exhibited different patterns at the ankle. One athlete had a single weak plantar flexor concentric contraction which produced only 12 J of work and a peak moment of 70 N.m. The other athlete produced a much larger plantar flexor moment which peaked at 125 N.m and yielded three work periods. An initial concentric period followed by an eccentric period and then another concentric period. The final concentric period had a peak power output of over 400 W compared with a peak of only 200 W by the other athlete.

The front leg of the two athletes begins with a much larger moment of force (100 N.m vs 0 for the back leg) since this leg is in a better position to carry the body’s weight. As with the back leg a plantar flexor moment dominates throughout the thrust period. The peak moment for both athletes was nearly 200 N.m. The functions of this moment were similar for both athletes. Initially, the moment is isometric, that is, its power output is nearly zero yet its moment of force is not zero. After about 0.3 s the moment becomes eccentric allowing the ankle to dorsiflex slightly before plantar flexing. This brief dorsiflexion may act as a counter-movement enabling a more power plantar contraction. The final plantar flexor concentric contractions prior to toe-off reached powers of between 500 and 700 W and produced between 50 and 100 J of work.

After toe-off the work and power output of the ankle muscles drops very low since the muscles are no longer acting to propel the body but are only needed to dorsiflex or plantar flex the foot.

Knee kinetics. The muscles of the back legs of the two athletes had quite different functions. In one athlete the knee extensors predominated acting eccentrically to prevent excessive knee flexion. In the other, the knee extensors acted concentrically causing extension and forward motion of the upper body. Finally, prior to toe-off the knee flexors dominated with an eccentric contraction preventing hyper- extension of the knee similar to the role that they play during the push-off phase of running (Robertson, in press).

The knee muscles of the front leg of both athletes had concentric periods of contraction of the knee extensors prior to toe-off as might be expected. These contractions had peak powers of 400 and 800 W and peak moments of, approximately, 100 N.m. The athlete that had the highest power also had a preceding knee flexor concentric contraction. This contraction caused the knee to flex and perhaps enabled a more forceful knee extensor contraction, however, the additional time required could be disadvantageous.

Additional time spent in the blocks means the athlete will be behind at the start and unless the athlete’s velocity is greater he/she will not be able to make up for the loss if other factors are equal. At this stage of the research the author is uncertain as to whether more time spent creating thrust and consequent velocity is more important than getting away earlier. Previous studies comparing the “bunch” and “elongated” starts showed that even though the bunch start gets the athlete out of the blocks earlier an elongated start gets the athlete out faster, however, over 5, 10, and 50 yd. the elongated start produced faster times (Henry, 1952).


1. Baumann, W. Biomech. V-B, 194-199, 1976.
2. Desipres, M. Biomech III, 364-369, 1973.
3. Dickinson, A.D. Res. Quart. 5(supp): 12-19, 1934.
4. Elftman, H. Amer. J. Physiol. 125:339-356, 1939.
5. Gagnon, M. Biomech. V-B, 46-50,1976.
6. Henry, F.M. Res. Quart. 23: 301-318, 1952.
7. Robertson, D.G.E. Biomech. X, in press.
8. Robertson, D.G.E. & Winter, D.A. J. Biomech. 13:845-854, 1980.
9. Winter, D.A. Biomechanics of Human Motion, Wiley: Toronto, 1979.


D. Gordon E. Robertson,
Kinanthropology Department
University of Ottawa

Complete Speed Training

The FIRST and ONLY All-Inclusive, Step by Step, Speed Development Program to Show You Exactly How to Make Your Athletes Faster and More a Athletic Than the Competition!

DVD #1: Pre Competition
DVD #2: Agility Training
DVD #3: Hardcore Conditioning
DVD #4: High Powered Training
DVD #5: Pure Speed Training

- Quick and easy methods for getting more done in less time so you can focus on the skills specific to your sport.

- Easy to understand and apply strategies for speed development.

- Clear progressions that can be used for beginner and advanced athletes at the same practice.

- Drills and exercises on video so you can see exactly how to perform and teach drills properly.

- Specific instructions detailing how, where and when to use each movement without having “to earn a degree in exercise science or biomechanics”

- Proven sample workouts and programs you can instantly bring to practice – the same day your program arrives

Order Now