Introduction

As you may all know, gymnastics is a unique sport placing high demands on competitors. Male gymnasts are required to compete on six apparatuses, while female gymnasts are competing on four. With the exception of vaulting, which requires the execution of a single skill, competitors on all apparatuses perform routines composed of a series of individual skills. It has been estimated that several hundreds and possibly thousands of skills and skill combinations already exist, and the number is always increasing with the addition of new ones. Although a good number of these skills share common principles and therefore can be grouped together, the number of groups is still quite large making it extremely difficult, if not impossible, for anyone to examine and study all gymnastic skills and identify specific principles applicable to "the sport of gymnastics".

For comparison purposes, human gait, a single physical activity, has been the subject of over 1000 biomechanical research studies, and, predictably, it will be the subject of many more. How many studies are needed then to be able to "understand" gymnastics? The answer is obvious and intimidating. In light of the numbers' reality, attempts have been made for grouping gymnastics skills into a few categories comprised of "tricks" that share common elements, making thus the study more manageable. The most recent classification was made by Brüggemann (1994a) who, building upon Hochmuth and Marhold (1987, as reported by Brüggemann, 1994a), grouped gymnastic skills into the following five categories:

1. Takeoff and pushoff from solid or elastic surface;
2. Rotations in vertical plane about a fixed or flexible horizontal axis of rotation;
3. Rotations in a vertical plane about a vertical axis of rotation;
4. Airborne rotations, and;
5. Landings.
   

There is a lot to be learned by reviewing previous work. And, I believe that re-inventing the wheel is wasteful. Therefore, in reviewing biomechanical research in gymnastics, I have drawn upon previous work including reports by Brüggemann (1994a) and Prassas (1995a). Space limitations makes it impossible for this review to be exhaustive. For this reason, only a selective number of gymnastics research has been included. The studies mentioned here should not necessarily be viewed as been more significant than work that has been omitted. Availability and language of the published studies (mainly English) were some of the criteria albeit important ones for inclusion. Research on some apparatuses such as the pommel horse and most of the women's events is very limited and so is related literature presented here.

Concepts: Listed below are some commonly found concepts in biomechanical research as applied to gymnastics:

Angular momentum: Describes the quantity of angular motion possessed by the gymnast. It is made up of the sum of the angular momenta of the body's segments. Quantities influencing angular momentum are the rotational speed of the gymnast, the point about which the gymnast is rotating, and the gymnast's body configuration. In airborne activities such as dismounts and somersaults, the angular momentum is constant - "conserved". As a result, when body configuration changes, the angular speed changes. For example, the gymnast slows down when he/she "opens up" before landing. Or, when a body part slows down, another body part speeds up, or vice versa he last been referred as transfer of (angular) momentum.
Moment of inertia: Describes the level of resistance to changes in rotational speed. It depends on the mass of the gymnast and how that mass is configured about the point of rotation (axis). A gymnast's moment of inertia, for example, progressively increases as he/she goes from a tucked to a piked to a layout position during somersaulting.
Torque: Describes the rotational effect of a force. It depends on the (magnitude of) force and its distance from the point of rotation (axis). Whereas, for example, the gravitational force (the weight) of a gymnast doing a giant swing is the same throughout the swing, the corresponding torque increases as the weight moves away from the bar and decreases as it comes closer.
Kinetic energy: Describe the amount of energy a gymnast has because of his/her linear and/or angular motion. The faster he/she moves, the more energy he/she possesses.
Computer simulation: Describes the (re)production of a movement by computers. It offers the advantage of trying a skill repetitively and/or under different conditions. Caution should be exercised to ensure that the "different" conditions are realistic, i.e. they represent what can be done by real subjects.

What is Done

Floor Exercises: The great majority of floor exercises consist of jumping/rotating elements interconnected by simpler transitional skills. Understandably then, most research in floor exercises examines the takeoff and (on occasion) landing characteristics of various types of somersaults, mostly backward. Hwang, Seo and Liu (1990) investigated takeoff mechanics of three different types of backward somersaults performed at the 1988 Seoul Olympic games including the contribution of the different body parts to the total angular momentum. , i.e. the required "spin". It was found that, in all cases, the legs' contribution to the total angular momentum was dominant. Similar takeoff mechanics were found by Kerwin, Webb & Yeadon (1998) who investigated the production of angular momentum in double backward somersaults performed during the 1996 Olympics. Angular momentum and center of mass (CM) kinematics of single and double backward somersaults were investigated by Brüggemann (1983). Knoll (1993) examined the same parameters when studying implications for round-off and flic-flac techniques concluding that maximum height and takeoff angular momentum must be optimized. Most recently, takeoff and landing characteristics of double back somersaults on the floor performed at the 1994 World gymnastics championship were studied by Geiblinger, Morrison and McLaughlin (1995a; 1995b); the (kinematic) results presented are in agreement with previous literature. Forward somersaults have received less attention. The Russian one, favored by the majority of gymnasts, has been studied by Knight, Wilson and Hay (1978) who concentrated mainly on the action of the arms. Ground reaction forces for the Russian type of somersaults were also examined by Miller and Nissinen (1987) in order to investigate their characteristics in relation to performance. In summary, there is a wealth of information and good understanding of somersaults' takeoff requirements. Landings, however, have not been studied as much and, consequently, they are not as well understood. In addition, there is a lack of information on the extremely high loads placed on the muscle/tendon system during the short contact time in both takeoffs and landings. These loads are augmented when combinations such as backward somersaults immediately followed by forward ones are performed.

Vaulting: Vaulting is the only apparatus involving a single movement. Partially for this reason, it might be the apparatus most researched (at least in proportion to the number of skills performed on it) and best understood. Studies by Bajin (1979), Dainis (1979), Brüggemann (1984), Takei (1989; 1990; 1991a; 1991b; 1992; 1996; 1998), Takei and Kim (1992), Li (1998), and Krug, Knoll, Koethe, and Zocher (1998) have examined springboard parameters, parameters while in contact with the horse, and/or landing parameters. In addition, the correlation between mechanical variables and the scores given to the vaults has been investigated. As a result, it is generally accepted that, in vaulting, running approach horizontal velocity and takeoff springboard linear and angular parameters are more important than parameters during horse contact that means that it is very difficult to compensate for errors made during takeoff, while in contact with the horse. It is also generally accepted that the initial (takeoff) angular momentum is invariably reduced during contact with the horse and converted to vertical velocity. A model for gymnastics vaulting developed by Dainis (1981) for the airborne and horse-support phases of vaulting may be one worth the effort for every coach to study it and understand it. The model establishes some of the aforementioned relationships, clearly showing that initial (springboard) "takeoff velocity and distance from the horse to be the principle variables affecting the outcome of the vault".

Horizontal Bar: Research on the horizontal bar has focused mostly on dismount takeoff requirements and the mechanics of associated giant swings. Some transitional techniques and an ever increasing number of release-regrasp skills have also been investigated. George (1968) offered some of the first descriptive data for four different types of giant swings. Yeadon (1997), Yeadon, Lee and Kerwin (1990), and Kerwin, Yeadon and Lee (1990) utilized data obtained at the 1988 Seoul Olympic Games to determine the contributions of contact and aerial techniques in twisting techniques used in high bar dismounts and to examine the necessary modifications in body configurations and angular momentum needed in multiple somersault dismounts. It was found that twisting techniques relate to the timing of the twist within the two somersaults and that the tilt angle relates to the body configuration and number of twists. Takei, Nohara and Kamimura (1992) found significant correlations between vertical release velocity, height above the bar and total time in the air (which are, of course, inter-related) and successfully performed double somersault dismounts. Kinematic release data for double layout and triple somersault dismounts were presented by Park and Prassas (1995). Additional kinematic, kinetic and EMG data for giant swings have been reported by a number of investigators (Boone, 1977a; Cheetham, 1984; Prassas and Kelley, 1985; Okamoto, Samurai, Ikegami and Yabe, 1989; Yamashita, Kumamoto and Okamoto, 1979) and the transition to the inverted giant swing (the "stoop-in") was studied by Prassas, Terauds and Russel (1988). In order to establish profiles for the different dismounts and release-regrasp skills and to identify differences between the techniques studied, Brüggemann, Cheetham, Alp and Arampatzis (1994b) studied the mechanics of seventy dismounts and release-regrasp skills performed at the 1992 Barcelona Olympic Games. The seventy movements were divided into 10 groups and, among them, three groups were found to be significantly different in terms of maximum values and timing of a variety of kinematic and kinetic variables. Release-regrasp techniques have been studied by Prassas and Terauds (1986), Prassas (1990), Gervais and Talley (1993), Brüggemann et al. (1994b), and Cuk (1995a). The energetics of high bar giants have been studied by Okamoto, Sakurai, Ikegami, & Yabe (1989) and most recently by Natta (1988) and Arampatzis and Brüggemann (1998). In summary, there is a good understanding of the mechanics of giant swings and a number of release-regrasp skills and dismounts in the horizontal bar.

Rings & Parallel Bars: The skill level and their type in these apparatuses has rapidly changed over the last decade with swinging skills comprising currently a major part of gymnasts' routines. Research, however, has not progressed equally. With regard to the rings, Nissinen and Brüggemann (reported by Brüggemann, 1987) presented kinematic and kinetic profiles of straight arms giant swings contradicting coaching opinions. Geiblinger, McLaughlin and Morrisson (1995c) reported kinematics of a case study of the "stretched double feldge backward to forward swing in hang" the so called "O' Neil". Yeadon (1994) studied twisting techniques used in dismounts at the 1992 Olympic games, concluding that the majority of gymnasts use asymetrical use of the arms to initiate twists. Research on the parallel bars is also not extensive. The feldge (or beach basket) has been studied by Boone (1977b) and Takei, Dunn, Nohara, and Kamimura (1995) who compared the (traditional) inner and (newer) outer grip techniques in the feldge to handstand mount. It was concluded that the outer grip has advantages over the inner by elevating the body's CG more at regrasp. Liu and Liu presented a case study on swings in the extended hang (Liu and Liu, reported by Brüggemann, 1994a). A quasi-static movement, the press handstand, was studied by Prassas, Kelley and Pike (1986) and Prassas (1988; 1991). Prassas reported also on the techniques of two basic skills, the back toss (1994) and the backward somersault dismount (1995c). The dynamics of both skills have been investigated by Prassas and Papadopoulos (1998). Differences in vertical and horizontal forces during the upward, pushoff phase were found and these differences were related to the greater height attained in the back toss and the need for different horizontal flight displacements. Lastly, a case study of the double back somersault dismount was presented by Manoni and DeLeva (1993a) who also reported on different types of forward somersaults (1993b).

Miscellaneous Research: As it was said previously, women's gymnastics research is limited. Among the few studies conducted, Brown, Witten, Espinoza, & Witten (1996) investigated ground reaction forces in two relatively simple dismounts from the balance beam, which were found to be over 10 times body weight. In a follow up study, for more difficult (somersault) dismounts, the forces were found to up to 13 times body weight (Brown, Witten, Weise, Espinoza, Wisner, Learman, & Wilson, 1996). As a result, they suggested that, at least in practice and possibly in competition, gymnasts should be allowed to roll out of various dismounts a suggestion highly unlikely to be adapted by gymnastics' governing bodies. Knoll (1996) found that gymnasts employ the same biomechanical mechanisms in the performance of acrobatic tumbling exercises on floor and balance beam, i.e. trade-off between take-off angular momentum and take-off linear velocities. Research in the uneven bars is limited to studies of the overgrip giant swing (Witten, Brown, Witten, & Wells, 1996), overgrip and undergrip dismount giants (Prassas, Papadopoulos, & Krug, 1998), and uneven bar dismounts (Prassas, 1996). In general, similarities between the mechanics of uneven bars and high bar dismount giants result in similarities in some of the take-off dismount conditions. However, differences in the beat action through the bottom of the swing, differences in the physical characteristics, design and construction of the apparatuses and anthropometric differences between male and female gymnasts may explain some of the velocity and related parameter differences between the two apparatuses. Whereas the pommel horse is considered one of the most difficult apparatuses and relative research could be of extra value to practitioners, research is limited to a case study comparing the Thomas flaire spindle and the Magyar spindle (Cuk, 1995b). It was concluded that, although the (kinematic) results suggested that the former may be more difficult, the fact that gymnasts perform the Magyar spindle less frequently suggest that it is more difficult "they (the gymnasts) know best how difficult an element is".

Summary

What is Needed

The volume of scientific research in gymnastics is considerable and ever increasing. With few exemptions, related research has attempted to answer questions related to the 1) Jumping, 2) Twisting/somersaulting, and/or 3) Swinging requirements of the sport. Subsequent papers in this volume deal in depth with some of these questions. The majority of the existing research efforts have been descriptive in nature offering limited information to scientists and to practitioners. Within the sport's uniqueness and multifaceted approach, however, biomechanics is uniquely positioned to assist with regard to:

1. understanding of already existing techniques,.
2. new skill development,
3. increase in safety, and
4. equipment design and/or modification.

Questions such as: what it takes to do a quadruple somersault? how many twists are possible? how flexible the bar(s) should be? or how springy a floor, or a spring board should be?...are legitimate questions and biomechanics may assist in finding proper answers. For that purpose, descriptive studies of specific skills should continuously be undertaken for description is the first step in understanding. Scientific efforts, however, that attempt to develop principles applicable to an ever larger number of gymnastic techniques would be more valuable. The ultimate success would be the development of a set of principles applicable to all new and existing skills that would have the ability to "explain the sport of gymnastics". It is very optimistic to predict that such a set of principles would be developed in the near future, but without dreaming, it will never occur. Lastly, in being consistent with ISBS's pledge to "bridge the gap between the researcher and the practitioner", it will be of tremendous value if scientists find a way to trickle a greater portion of the existing and new information down to the practitioners, the coaches and athletes. This information should be presented in a meaningful and understandable to the practitioners form.