timing in the forward one and one half somersault with one twist 3m springboard dive

Timing in the Forward One and One Half Somersault with One Twist 3m Springboard Dive

Ross H Sanders School of Biomedical and Sport Science, Edith Cowan University,

Perth, Western Australia

Sanders, R.H. (1999) Timing in the Forward One and One Half Somersault with One Twist 3m Springboard Dive, Journal of Science and Medicine in Sport 2(1): 57-66

The timing of actions to initiate and stop twist is critical to successful performance of the twisting and somersault rotations in 3m springboard dives. An important indicator of timing differences among subjects is the timing of hip flexion and extension. The purpose of this study was to quantify the timing and magnitude of hip flexions and extensions in the forward one and one half twisting dive with one twist. The timing and magnitude of hip flexion and extension of ten divers ranging in ability from New Zealand National to elite International standard were quantified using three-dimensional videography and analysis techniques. A Spearman (rho) correlation with p<.05 required for significance was conducted to determine the relationship between each of the variables and ability. The results indicated that skilled divers had more flight time than less skilled divers (rho=-.79), less hip flexion at takeoff (rho=.66), less pre-twist flexion (rho=.86), greater post-twist hip flexion, and had more time from the time of maximum post-twist flexion to entry than less skilled divers (rho=-.81). It was concluded that divers who currently initiate twist from a 'kick out' should learn to initiate twist without a 'kick out' and reduce hip flexion at takeoff to increase height and flight time.

Introduction Until recently, the deta i l s of how ro ta t ions a b o u t the long axis c a n be genera ted in ac roba t ic act ivi t ies were no t well unde r s tood . Cons ide rab le advances in knowledge a n d u n d e r s t a n d i n g have been ga ined t h rough relat ively recen t m a t h e m a t i c a l model l ing a n d s imula t ion by Yeadon (1990, 1993a). Twist ing mot ion in s p r i n g b o a r d dives m a y be genera ted pr ior to flight b y app ly ing t u r n i n g forces a b o u t t he long axis of the body while in con tac t wi th the sp r ingboa rd (contact twist) or b y ' t rading ' some of the s o m e r s a u l t mot ion for twis t mot ion (Bat terman, 1974, Yeadon, 1993b, 1993c). To t r a d e the s o m e r s a u l t mot ion for twist mot ion the long axis of the body m u s t be t i l ted away from the ver t ical p lane of the original s o m e r s a u l t ro ta t ion. Var ious a s y m m e t r i c a l body m o v e m e n t s affect the ti l t of the b o d y inc lud ing a symmet r i ca l m o v e m e n t of the a r m s (Frohlich, 1979; Van Gheluwe, 1981), l a te ra l flexions of the h ips (Van Gheluwe, 1981), a n d ro ta t ion of t he ches t a n d shou lde r s wi th r e spe c t to the h ips (Yeadon, 1993c). The s imula t ion s t ud i e s by Yeadon have s h o w n t h a t the effect d e p e n d s on the t iming of these m o v e m e n t s (Yeadon, 1993c). Piking a t the h ips is a n effective way of

57

Timing in the Forward One and One Half Somersault...

stopping the twist (Yeadon, 1993a). Asymmetrical a rm action, chest rotation, a n d lateral flexion may also be used to remove tilt and stop the twisting motion.

The timing of actions to initiate and stop twist is critical to successful performance of the twisting and somersaul t rotations. Divers have some flexibility as to how to time their body movements while still completing the necessary rotations. That is, the required twist and somersaul t rotations may be achieved with different timing and magnitude of body actions (technique). Some of these techniques are more aesthetically pleasing than others. Therefore, it is necessary to quantify timing and magnitude of body actions to determine ways in which the aesthetics of the dive, and thereby the scores attained, can be improved.

Although the timing of the asymmetrical actions such as the a rm action, lateral flexions, and chest rotations have important effects on the rate of twist, it may be argued that the main indicator of timing differences among subjects is the timing of hip flexion and extension. The reasons why the timing of hip flexion and extension is so important are as follows: 1. The rate of somersaul t rotation depends greatly on the amount of hip flexion.

When the hips are extended (straight body position) somersaul t rotation is slow and when flexed (piked) the rate of somersaul t rotation is fast. Thus, divers control their rate of somersaul t rotation by the amount of hip flexion adopted during flight. A diver's somersaul t angular m o m e n t u m is determined at the instant of last contact with the springboard. From that instant the diver's only control over the rate of somersaul t rotation is by changing the moment of inertia by piking or extending. If the diver is too late in piking after takeoff and too early in extending prior to entry the dive will be 'short ' (insufficient rotation). If the diver is too early in piking and too late in extending the dive will be 'long' (too m u c h rotation). In both cases the aesthetics of the dive and the neatness of entry are adversely affected.

2. The grace and aesthetics of the dive (form) during flight are affected by the magnitude and timing of hip flexion and extension.

3. The initiation of twist in forward dives is often linked to hip extension. Extension from a pike in combination with asymmetrical body actions is a common means of initiating twist in forward somersaul t dives (Yeadon and Atha, 1983; Yeadon, 1993c).

4. The cessation of twist occurs when the body pikes. The timing of the piking action is important to stop the twist at the appropriate t ime and to align the body 'square' for entry. Following the cessation of twist there needs to be sufficient time to extend for entry and have the correct entry angle. Thus, the timing of piking and extending prior to entry is critical to good performance.

5. The twist rate depends partly on the hip angle. The straighter the diver is, the greater the twist rate. The straight position needs to be maintained long enough to complete the twists. The longer the time taken to complete the twists, the less time there is to adopt aesthetically pleasing postures and to prepare for entry. Thus, divers need to be effective in generating twist rotation and then maximise twist rate by adopting a straight posture.

6. The amount of hip flexion while still in contact with the springboard affects the height achieved in forward dives (Miller and Munro, 1984, Sanders and Wilson, 1988) and therefore the time in the air. The greater the hip flexion the less the height achieved. Thus, if one is overly concerned about completing the required rotations, there is a tendency to pike too much while still in contact

58


with the springboard. This reduces the time in the air and is actually counter to the diver's intention to complete the rotations with time for a Controlled entry. The grace and form of the dive are adversely affected as the diver appears to 'rush'.

The purpose of this study was to quantify the timing and magnitude of hip flexions and extensions in the forward one and one half twisting dive with one twist (5132D).

Methods Data Collection and Analysis

Data were collected on two occasions. On the first occasion, twisting dives of male and female competitors in the 1995 Swedish Flag International Competition were recorded. On the second occasion, twisting dives of male and female divers competing in the 1996 New Zealand National Championships were recorded. Because some divers competed from one board while others competed from a different board the data required for analysis were obtained immediately after competition with all divers performing from the same board. Divers were requested to perform twisting dives of their choice for analysis. The dive with the largest representation among those dives was the forward one and one half somersault with one twist in a free position (5132D). One dive of each of the 10 divers performing 5132D was digitised and analysed. In cases where a diver performed more than one 5132D their best dive was selected for analysis. Two divers competed in both the Swedish Flag Competition and National Championships. For each of these divers, the dive performed after the National Championship, that is, the more recent dive, was analysed. Five male and five female divers, evenly distributed in terms of ability, contributed dives for analysis.

The dives were videotaped with three SVHS cameras at 50 fields per second. Recordings from two of the cameras were used to obtain quantitative data for subsequent three-dimensional analysis. These cameras were positioned with their axes at approximately 90 degrees to each other to maximise the accuracy of the three dimensional calculations in accordance with the recommendations of Shapiro (1978). The third camera was positioned in line with the diving board to obtain data for qualitative analysis.

Prior to competition, the three dimensional space was calibrated by videotaping a three dimensional control object which has markers at known positions. The object was suspended from the 5m platform so that it enclosed the space in which the diver performed. The control points on the calibration frame and selected anatomical landmarks of the subjects were manually digitised using a Peak Motion Analysis System. The body landmarks digitised were the vertex of the head, both shoulders, elbows, wrists, hips, knees, and ankles. The digitised points of the control object and subject and the known coordinates of the control object were input to the Peak 3D analysis software which calculated the three dimensional coordinates of the body landmarks for each digitised frame. These data were then input to a FORTRAN analysis program to calculate the variables of interest. Variables analysed In accord with the rationale outlined in the introduction, the magnitude and timing of hip flexion and extension with respect to the instant of takeoff from the springboard were quantified. These variables were:

59


• - time at entry. • the hip angle at the instant of takeoff. • the hip angle at max imum flexion during the period prior to twisting termed

'maximum hip angle pre-twist'. • the t ime from the instant of takeoff to m ax imum flexion in forward dives. • the t ime when the body could be regarded as straight following the hip

extension ('kick out'). • the t ime of commencement of hip flexion to stop the twist rotation. • the difference in time between full extension and initiation of hip flexion to stop

the twist (this indicated the time that the straight position was held to complete the necessary twist rotation).

• the hip angle at max imum hip flexion during the period in which twist rotation was stopped prior to extending for entry termed 'maximum hip angle post- twist'.

• time of post-twist max imum hip flexion (this gave an indication of when the diver had completed the twisting part of the dive and was ready to extend to prepare for entry).

• the difference in time between m a x i m u m flexion to stop the twist and the time of entry (this indicated the time the diver had to prepare for entry).

• hip angle at entry. The hip angle was calculated using the mathematical model described by

Sanders (1995) and expressed with respect to the extended position. When the hips were extended the hip angle was 0 degrees. Flexion angles were positive and angles of hyperextension were negative. For the purpose of identifying the period of full extension, the diver was regarded as fully extended when the hip angle was less than 10 degrees.

Because the dives were performed out of competition, scores could not be used as the dependent variable. Instead, divers were ranked according to their placings in competition and assessments of twisting ability by judges and coaches present at the competition. A Spearman (rho) correlation was conducted to determine the relationship between each of the variables and ability.

Time-normalisation of temporal variables One would expect that successful performance in twisting somersaul t dives is partly dependent on the flight time, defined as the time from the instant of takeoff from the springboard to the instant the hands break the water. Divers who achieve good height from the sprtngboard have a long flight time to complete the necessary somersaul t and twist rotations in an unhurried and aesthetically pleasing manner. However, it m u s t be recognised that the bigger divers may have a longer flight time than smaller divers but may rotate more slowly due to greater moments of inertia about the somersaul t and twist axes. Thus, it was of interest to investigate the temporal measures as real t imes and expressed as percentages of the flight time. It is possible tha t gender of the divers has an effect on height achieved due to differences in strength. Thus, normalising the temporal variables in this way reduces the possible effects of gender.

ReSults and Discussion Tables 1 to 3 show the times of critical events with respect to the instant of last springboard contact, the times of critical events expressed as a percentage of the

60

Timing in the Forward One and One Half Somersault..,

Table 1: Times of critical events for the fow~ard one and one half somersault with one twist.

Rank Max Full ext Hip Max Entry Hip flex- Entry- flex flex flex I Full ext Max flex I

1 0.00 0.13 0,36 0.76 1.40 0,23 0.64 2 0.07 0.20 0,40 0.86 1.56 0.20 0.70 3 0.07 0,30 0.52 0.83 1,48 0.22 0.65 4 0.17 0.28 0.60 0.84 1.48 0.32 0.64 5 0.08 0.26 0.33 0,79 1,20 0.07 0.41 6 0.12 0.36 0,68 0.96 1.20 0.32 0.24 7 0.12 0.52 0.60 0.87 1,20 0.08 0.33 8 0.08 0.34 0.46 0.76 1.12 0.12 0.36 9 0.12 0.24 0.56 0.84 1.20 0.32 0.36 10 0.12 0.40 0.56 0.96 1.20 0.16 0.24

Mean 0.095 0.303 0.507 0.847 1.304 0,204 0.457 SD 0,046 0.11 0.115 0.071 0.158 0.096 0.181 Spearman's rho 0.62 0.62 0.42 0.39 -,79** -0.17 -81 **

* Signifcant to .05 ** Significant to ,01

Table 2: Times of critical events as a percentage of flight time for the forward one and one half somersault with one twist.

Rank Max Full ext Hip Max Hip flex- Entry- flex flex flex I Full ext Max flex I

1 0.00 9.3 25.7 54.3 16.4 45.7 2 4.5 12.8 25.6 55.1 12.8 44.9 3 4.7 20.3 35,1 56.1 14.9 43.9 4 11.5 18.9 40.5 56.8 21,6 43.2 5 6.7 21.7 27.5 65,8 5.8 34,2 6 10.0 30.0 56.7 80.0 26.7 20.0 7 10.0 43.3 50.0 72.5 6.7 27.5 8 7.1 30.4 41.1 67.9 10.7 32.1 9 1 0 20.0 46.7 70.0 26.7 30.0 10 10.0 33,3 46.7 80.0 13,3 20,0

Mean 7.45 24.00 39.56 65.85 15.56 34.16 SD 3.57 10.22 10.84 9.96 7.41 9.96 Spearman's rho .64* .75* .75* ,87** -0,02 -,87 **

* Significant to .05 ** Significant to .01

flight time {the time from the instant of last contact with the springboard to first contact with the water}, and the hip angles at key events during the forward one and one half somersaul t dive with one twist. Means and s tandard deviations and the Spearman correlation {rho} between the variable and diver r ank are presented at the foot of the table for each variable. Figures 1 and 2 show the durations of intervals between key events in real time and as a percentage of the flight respectively. Figures 3 and 4 show the orientations of the highest and lowest ranked divers respectively.

61

Timing in the Forward One and One Half Somersault,,.

Table 3: Hip angles at key events during the forward one and one half with one twist.

Rank Takeoff Max flex Max flex I Entry

1 39 39 160 19 2 36 42 125 17 3 46 67 126 14 4 28 67 128 12 5 60 76 114 34 6 72 114 123 12 7 58 98 86 0 8 49 88 90 80 9 65 100 100 30 10 70 100 90 62

Mean 52,3 79.1 114.2 28 SD 15,1 25.4 23,1 24,9 Spearman's rho .66* ,86** -.85** 0.37

* Significant t o .05 ** Sgnificant t o .01

Figure 1: Durations and timing of intervals between key events during 5132D.

10

9

8

7

~ 6

5

4

3

2 + ,

1tl 0.0

IIIIIIIII1111 m ...... - r , , , - - ' / / / / JJ

IIIIII ~ / / / / / / / 7 2 1

I I~J3JJf lJJ I I ~ --'////////J

I I I I I I I I I I I I I I I I l l l - - / / / / / / / A

IIIIIIIIIII ~ _ ~ / / 1 1 1 1

~1111111 . . . . . ; - ~ - .... ~ ~

IIIII ~ / I I I I I I / i l I I / - ? I ~

IIFFTTTFFFFF~ ~ I / I I I I I I I I I I I I / I

I I I I I I / I / / / / ~

IIII . . . . . ~ ~ ~ I I I / I I I I I I / I I I A I I [ 1 I I I I

0.2 0.4 0.6 0.8 1.O 1.2 1.4 1.6

Max Flex 1 to Entry

[ ] Hip Flex to Max Flex 1

[ ] Full Ext to Hip Flex

[ ] Max Flex to Full Ext

• Take Off to Max Flex

11.8 2!0

Time (s)

Figure 2: Durations and timing of intervals between key events during 5132£) expressed as a percentage of flight time.

10

9

8

7

~ 6

4

3

1 lr 0%

II]1 IIIIII1~1~

IIIII

IIIII IIIIIIII ~1[!

IIIII IIIII]11111111111

IIIII IIIIIIII

IIIII F F'~

II1

E l i IIII]

/ / / / / / / / 1 [ ] Max Flex 1 to Entw

/ / / / / / / / / / / J • Hip Flexto Max Flex 1

Fun Ext to Hip Flex

~ / / / / ' / / / / / ' / / / I [ ] Max Flex tO Full Ext

/ / / / / / / / / / J • Take Off to Max Flex

/ / / / / / / A

' l ' / / / / / / / / / / / , d "///////////////~

~'///////////////A

II I I I I ////////////////,I

I I I I / / / / / / / / / / / / / / / / A I [ I [ ] I 1 I I

10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Time (% o f F l i g h t )

62


Figure 3: Orientation of the highest ranked diver at equispaced time intervals between takeoff and ent~ The light side of the diver is in solid lines. The diver is viewed from the right side and from the same-height as the diver's shoulders throughout the dive (top) and as if the viewer is attached to the right side of the diver's chest, and is rotating with the diver (bottom).

o

Figure4: Orientation of the lowest ranked diver at equispaced time intetl/als between takeoff and entry. The right side of the diver is in solid lines. The diver is viewed from the right side and from the same height as the diver's shoulders throughout the dive (top) and as if the viewer is attached to the right side of the diver's chest, and is rotating with the diver (bottom).

Time at entry The results indicated that skilled divers had more flight time than less skilled divers (rho=-.79). Because the time of flight is dependent on the height achieved by the diver this indicated that the skilled divers were more effective in working the board to achieve height than less skilled divers.

63


Hip angle at the instant of takeoff There was a strong relationship between ability and the angle of takeoff (rho=.66). This means that skilled divers had less hip flexion at takeoff than less skilled divers. Hip flexion is known to be one variable tha t affects the height achieved and height in the air (Miller & Munro, 1984; Sanders & Wilson, 1988). In this s tudy the hip angle at takeoff was also strongly related to flight time (Pearson r=-.79). Hip angle at maximum flexion (pre-twist)

There was a very strong relationship between ability and pre-twist maMmum flexion angle (rho=.86). This means tha t skilled divers used very little hip flexion whereas less skilled divers used a large amount of hip flexion prior to 'kicking out' to start the twist. The five highest ranked divers did not use a large 'kick out' and commenced their asymmetrical action and twist immediately after takeoff from the board. Those ranked 6 to 10 all commenced their twist from a strong pike and 'kick out'. Those divers had a later s tar t to twisting than those who had small angles of hip flexion after takeoff.

The time to maximum flexion (pre-twist) Time to max imum flexion was modestly related to ability (rho=.62) and did not quite reach statistical significance at the p=.05 level (p=.054). It is interesting that the best performer had no further flexion after the instant of takeoff from the springboard.

Time to full extension There was a modest relationship between ability and time to attain a straight position (rho=.62) and did not quite reach statistical significance at the p=.05 level (p=.054).

Time of commencement of hip flexion to stop the twist rotation This variable was only weakly correlated with ability (rho=.42). This means that there was a very slight tendency for skilled divers in this sample to pike earlier in the flight than less skilled divers. The relationship did not reach statistical significance at the p=.05 level (p=.23). However, when the times were expressed as a percentage of flight time, the relationship between time of commencement of hip flexion and ability was strong (rho=.75). Given that the skirled divers had long flight times, this result means tha t the time of hip flexion of skilled divers was early in the dive relative to the total duration of flight.

Time from full extension to hip flexion There was no significant relationship between this time and ability. Those with very short periods in which the body could be regarded as 'straight' performed much of the twist outside that period. This was due partly to much of the twist being performed before the straight position was attained and partly to considerable twist being performed after commencement of the hip flexion to stop the twist. The latter was due to a slow and indecisive hip flexion. It is worth noting that the four highest ranked divers held a straight position for between 0.20s and 0.32s. Dunng that time the bulk of the twist was performed and the twist was then stopped by a decisive and rapid hip flexion.

Hip angle at maximum flexion (post-twist) Hip angle at max imum flexion post-twist (termed Max flex 1 in the tables) was very strongly related to ability (rho=-.85). This means that skilled divers piked more to stop the twist than less skilled divers. Because the less skilled divers were 64


successful in stopping the twist, it can be assumed tha t the skilled divers piked more than necessary to stop the twist.

Achieving a greater pike angle during this par t of the dive may have aesthetic advantages. Adopting a large pike angle creates the impression tha t the diver is not rushed in preparat ion for entry, and allows the diver to quickly complete the last of the required somersaul t rotation so that the body can be optimally extended at entry without landing the dive 'short'.

Time of maximum hip flexion (post-twist) The relationship between this time and ability was weak (rho=0.39). However, it is worth noting that the best diver had the shortest time from takeoff to max imum post-twist hip flexion and yet bad the greatest amoun t of hip flexion.

When expressed as a percentage of flight time, the relationship between max imum hip flexion and ability was very strong (rbo=.87). This means that the more skilled divers completed hip flexion earlier in the dive than the less skilled divers and so had more time to extend and prepare for entry.

Time from maximum flexion to entry The mean time from max imum flexion to entry was 0.46s and ranged from 0.24s to 0.70s. Thus, there were large differences among divers in the time available to extend and prepare for entry. This time was strongly related to ability (rho=-.81), that is, skilled divers had more time to extend and prepare for entry than the less skilled divers. When the times were expressed as a percentage of flight time, the relationship between the time from maximal flexion to entry and ability was even stronger (rho---.87).

Hip angle at entry Mean hip angle at entry was highly variable over the group as a whole bu t consistent among the highly ranked divers. The four highest ranked divers commenced their entries with hip angles between 12 degrees and 19 degrees. There was no significant relationship between hip angle at entry and ability due to the great variability among the less skilled divers. It is apparent tha t the skilled divers had the time to control their entry to achieve a suitable entry angle, that is, neither 'short ' nor 'long' as well as a suitable hip angle.

Summary and Implications for Divers The better divers were distinguished by completing the twist with sufficient time to adopt an aesthetically pleasing tight pike position prior to extending for an entry that is neither 'short ' nor 'long'. They had a suitable hip angle as entry began (between 12 and 19 degrees). The skilled divers had more time after completing the twist than the less skilled divers for the following reasons: 1. They had more time in the air due to effective board work prior to takeoff. This

accounted for most of the difference in entry preparation time between the skilled and less skilled divers.

2. Skilled divers did not rely on a large 'kick out' from a strongly piked position. Therefore, they started twisting sooner than less skilled divers who relied on a strong 'kick out' to produce twist. By starting the twist earlier, the skilled divers also finished the twist earlier than less skilled divers. However, this accounted for a relatively small amount of the difference in preparat ion time and could be regarded as a 'characteristic' of skilled divers rather than a main reason for their time advantage. Avoiding a large 'kick out' probably has

65


advantages in terms of the aesthetics of the dive. 3. Divers who use a strong 'kick out' are likely to leave the board with

considerable hip flexion in preparat ion for the 'kick out'. This has the effect of reducing height achieved and time in the air (Miller and Munro, 1984; Sanders and Wilson, 1988). Conversely, divers who do not rely on a strong 'kick out' are likely to have less hip flexion at takeoff and to achieve good height and time i n t h e air.

From this s tudy it may be concluded tha t divers who currently initiate twist from a 'kick out ' should: 1. Learn to initiate twist without a 'kick out'. 2. Reduce hip flexion at takeoff and strive to increase height

Acknowledgements The author gratefully acknowledges funding support from the New Zealand Hfllary Commission's Sport Science Research and Development Programme. Thanks to Tony Stevely and Diving New Zealand for their ass is tance, encouragement, and support of this project. Thanks to the divers and coaches for their cooperation and participation.

References Batterman, C. (1974). The techniques of springboard diving. Cambridge, MA: MIT Press. Frohlich, C. (1979). Do springboard divers violate angular momentum conservation? American

Journal of Physics 47: 583-592. Miller, D.I. & Munro, C,F. (1984). Body contributions to height achieved during the flight of a

springboard dive. Medicine and Science in Sports and Exercise 16(3): 234-242. Sanders, R.H.& Wilson, B.D. (1988). Factors contributing to maximum height of dives after takeoff

from the 3m springboard. International Journal of Sport Biomechanics 4: 231-259. Sanders, R.H. (1995). Effect of ability on twisting techniques in forward somersaults on the trampoline.

Journal of Applied Biomechanics 11(3): 267-287. Shapiro, R. (1978). Direct linear transformation method for three-dimensional cinematography.

Research Quarteriy 49: 197-205. Van Gheluwe, B. (1981). A biomechanical simulation model for airborne twist in backward

somersaults. Journal of Human Movement Studies 7: 1-22. Yeadon, M.R. & Atha, J. (1983). The production of a sustained aerial twist during a s o m e r s a u I t

without the use of asymmetrical arm action. In D.A. Winter, R.W. Norman, R.P. Wells, K.C. Hayes, & A.E. Patla (Eds.) Biomechanics IX-B (pp.395-400} Human Kinetics: Champaign, IL.

Yeadon, M.R. (1990). The simulation of aerial movement. {3 Parts). Journal of Biomechanics 23: 59- 83.

Yeadon, M.R. (1993a). The biomechanics of twisting somersaults: Part I. Rigid body motions. Journal of Sports Sciences II: 187-198.

Yeadon, M.R. (1993b). The biomechanics of twisting somersaults: Part II.Contact twist. Journal of Sports Sciences 11: 199-208.

Yeadon, M.R. (1993c). The biomechanics of twisting somersaults: Part Ill. Aerial twist. Journal of Sports Sciences II: 209-218.

66

timing in the forward one and one half somersault with one twist 3m springboard dive

Documents