Baumann et al , 1988, 3D analysis of Snatch

INTERNATIONAL JOURNAL OF SPORT BIOMECHANICS, 19B8. 4, 69-89

The Snatch Techniqueof World Class Weightlifters

at the 1985 World Championships

Wolfgang Baumann, Volker Gross, Karl Quade,Peter Galbierz, and Ansgar Schwirtz

The purpose of this study was (a) to describe the snatch technique in termsof kinematic and external and internal kinetic parameters, and (b) to com-pare the results for athletes of different groups and weight categories. Bymeans of three-dimensional film analysis and measurements of ground reac-tion forces during the 1985 World Championships in Sweden, it was pos-sible to analyze the spatial movements and to calculate joint moments of forcein each leg. Concerning the kinematics, a snatch technique starting with astrong pull toward the lifter could be established. The most interesting kinet-ic results are that the knee joint moments are relatively small (one third ofthe hip joint moments of force) and do not correlate very well with the totalload. The best lifters seem able to limit the knee joint moment by praisecontrol of the knee pwsition with respect to the ground reaction force. Al-together, the results concerning the internal kinetic parameters question theJogic of the classical division of the lifting technique into phases accordingto external kinetic parameters.

Until now, research in weightlifting has been predoTninantly concernedwith the kinematics of competitive techniques, particularly the two-dimensionalanalysis ofthe trajectory ofthe bar itself (Garhammer, 1979, 1981, 1985), occa-sioiwlly including selected parameters with respect to actual body movements(Enoka, 1979; Ono, Kubota, & Kato, 1969). The external kinetics of lifting tech-niques have been investigated in detail by Vorobyev (1978) and reported in histextbook, Weightlifting. Similarly, the work of Kauhanen, Hakkinen, and Komi(1984) concerned itself with selected kinematic and external kinetic parametersof training techniques. In most studies, however, the movements have been con-fined to the sagittal plane corresponding to the preferred observation perspectiveof the trainer. To date, no known studies have been conducted considering the

The authors are with the Institut fur Biomechanik, Deutsche Sporthochschule.Carl-Diem-Weg, D-5000. Koln 41, Federal Republic of Germany.

68

SNATCH TECHNIQUE OF WEICHTLIFTERS 69

features of (a) three-dimensional data collection, (b) measurements of ground reac-tion forces, and (c) movements under competitive conditions, that altogether wouldhave allowed the determination of joint moments of force of competitive tech-niques in weightlifting. Lacking, therefore, is essential information on the fac-tors governing specific techniques and conditioning practices. Through thisinvestigation, we attempted to overcome this lack of scientific information.

Methods

Subjects

At the 1985 World Weightlifting Championships in Sodertalje, Sweden, virtuallyall lifts in the snatch and in the clean and jerk in all weight categories were recordedusing video techniques. In addition, 20% of these were also filmed and groundreaction forces were measured for about 80% of the lifts.

Kinematic parameters of the movement of the bar were determined fromvideo recordings. With respect to these, two extreme groups were formed usingfour weight categories (i.e., 60, 75, 90, and 110+ kg), the first comprising the10 best lifts of the four first-place Group A athletes and the second comprisingthe 10 poorest lifts from Group B. In all, 82 lifts were studied.

The three-dimensional kinematic and kinetic analysis, including a calcu-lation of the joint moments, was obtained directly from film and the measure-ment of ground reaction force (GRF). Chosen were 17 attempts in three weightcategories (i.e., 60, 82.5, and 110+ kg). Among these were Shalamanov's worldrecord lift and the best lifts of gold medalists Vardanyan and Krastev. Six unsuc-cessful lifts were also included in the analysis.

Definition of Variables

The present analysis focused on the snatch technique from the beginning of themovement to the point at which the lifter dropped under the barbell. This phaseis considered to be the most important and technically most difficult part of thewhole movement and is treated accordingly in the literature. The initial choiceof parameters was based on a theoretical and practical approach by consideringthe kinematics of the movement of the bar and of the lower limb, and externalkinetics in the form of ground reaction forces. Then certain variables associatedwith muscular control, namely net joint moments of force in the lower extremi-ty, were investigated. Table 1 shows the selected variables.

Data Acquisition and Processing

Kinematics. In order to determine the kinematic parameters of the movementsof the bar and the lifter, video and film techniques were employed. Video waschosen because it provides a suitable, low cost means for collecting data on allthe lifts while providing adequate precision for quantitative analysis. The twocameras used were semiprofessional JVC triple-tube color models (PAL 50 Hz,625 lines) which were set up in the horizontal plane at an angle of 90° to eachother (see Figure 1) in order to allow the movement to be viewed from the sideand from the front. The pictures were recorded on U-matic recorders via a com-mon timer. The films were taken using two phase-locked synchronized 35-mmArritechno 150 cameras, which were placed at 45° angles to the frontal plane

70 BAUMANN, GROSS, QUADE, GALBIERZ, AND SCHWIRTZ

Table 1

Experimental Variables

Symbol/unit Description/definition

body massbody heightbarbell massgrip widthmaximum height of barbeilloss of height during drop under barbeli

m/s first peak vertical velocity of barbeilsecond peak vertical velocity of barbell ' 'time from lift-off to maximum velocity of barbelltime from iift-off to maximum height of barbeilAverage power during T2: Pav = (BBW + BBHmax) x 9.81m S'/T2duration of first pullduration of second pullfoot angle (ankle joint)knee angtehip angleknee anguiar veiocityhip anguiar velocityvertical component of GRFmaximum Fz during first pull iminimum Fzmaximum Fz during second pullhorizontal component of GRF. anterior-posteriorhorizontai component of GRF, lateralmaximum rate of change in vertical component of GRFnet muscular moment at ankle jointnet muscular moment at knee jointnet muscular moment at hip joint

of movement (see Figure 1) approximately 40 m from the middle of the competi-tion platform and 4 m above it. The field of view of both cameras encompasseda width of 7 m. The camera speed of 50 fps was used. Because of the need forfilm precision and reliability, the cameras were attached to concrete pillars with-in the competition hall to prevent vibration. They were not disturbed throughoutthe 10 days of competition.

The spatial coordinates of various points required were calculated fromthe films using the well established Direct Linear Transformation procedure(Abdel-Azis & Karara, 1971). This method allows the use of nonmetric cameras,though it does require a precise, three-dimensional reference system that encom-passes the space in which the movement occurs. The reference system used herewas a metal framework that formed a precise 4 x 2 x 1 meter rectangular cube,each joint of which was 1 meter (± 1 mm) from its immediate neighbors. At varioustimes between the competitions, the reference framework was mounted in a fixedorientation to the competition platform and force plates (see Figure 2), and was

BWBHBBWGWBBHmaxBBHLVmaxiVmax2T1T2PavTP1TP2aFaKaHavKavHFzFziFz2Fz3FxFyRFzMFMKMH

kgmkgmmmm/sm/sssWss0

0

0

o/so/sNNNNNNN/sNmNmNm

SNATCH TECHNIQUE OF WEiCHTLIFTERS 71

Figure I — Positioning of the video cameras VI and V2 and the cine cameras Cland C2 with respect to the weightiifters' stage.

1000 mm

2000 mm

Force Platforms

Weightlifters Platform

Figure 2 — Spatial reference cube aligned with the force plates.

72 BAUMANN, CROSS, QUADE, CALBIERZ, AND SCHWIRTZ

I f

Figure 3 — Body and barbell points to be d^tized.

recorded by both the video and film cameras. Figure 3 shows the points that weredigitized in the three-dimensional analysis from the films. Reduction of the videorecordings was completed using a video digitizer, with the reference frameworkproviding a suitable scale for the various sagittal planes involved in the movement.

Kinetics—Ground Reaction Forces. The measurement of ground reactionforces (GRF) in high level competition poses problems that are quickly obviousin weightlifting. Whereas external kinetics can be fully determined from the GRF,the mounting of force plates may be difficult for several reasons. The insertionof the plates into the competition platform must not alter its mechanical or visualcharacteristics. Furthermore, the precision of the measuring apparatus must bemaintained throughout the competition, and also, in this case, niust withstandover 1,200 consecutive impacts from falling weights.

Ground reaction forces were measured using two Kistler force plates (600X 1,000 mm) specially designed for this project. Particular attention was paidto precision, mechanical stability, and safety with respect to impact. The mount-ing frames were set in a four-tonne concrete block. The force plates were thenfirmly mounted on top and within precisely cut recesses in the competition plat-form. In order to avoid any mechanical contact between them, the 5-mm slitsthat separated the plates from the surrounding material of the platform were thensealed with a soft dustproof plastic. The output of al! 16 measuring channels wasfed into a Data General Computer through appropriate amplifiers. The data ac-quisition was triggered automatically as soon as a threshold of 300 N was reachedand included all measurements during the previous 3 seconds as well. The sam-pling frequency was 100 Hz and the total measurement time on each occasionwas 10 seconds. This system, using separate plates for right and left feet, per-mits measurement of the three GRF components, the coordinates of the pointof application of the force, and the free perpendicular moment.

Kinetics—Muscular Moments. An effective evaluation of technique anddevelopment of optimal methods of training a knowledge of the time historiesof the various muscular moments (joint moments) is of utmost importance. Fromexternal kinetic data and three-dimensional kinematic information concerning thejoint centers of the lower limbs, it is possible to calculate the net muscular mo-ments around various joints in all three planes. It is important to note here that


these are net moments and that the effects of antagonistic muscular activity havenot been considered.

The calculation undertaken here with respect to muscular moments aroundthe joints of the lower limb are based on careful determination of the positionof the joint center and ofthe GRF vector. Control tests showed that the momentscreated by the inertial forces ofthe individual segments ofthe body were consis-tently under 5 %. These inertia! forces were therefore not included in the calcula-tions. In the results, only those muscular moments acting in the principal planeof movement are reported. This plane of movement is defined by the relativeposition ofthe long axes of adjacent segments and is that which is associated withthe motion brought about by the major muscle groups involved.

Results and Discussion

Barbell Kinematics (Two-Dimensional)

Trajectory of the Barbell. The movement of the bar is the result of forces thelifter applies to it. The displacement-time and velocity-time relationships are oftenseen at a practical level as the most important criterion for assessing lifting tech-nique. A number of pathways ofthe barbell's movement are illustrated in Figure4 and correspond to lifters from different weight categories. At the beginning

Verticil

110+ kg 110+ k |

A

.5 m

Figure 4 — Barbell trajectories of different attempts. 0 = initial position, 1 = endof 1st pull, 2 = end of 2nd pull.


of the first pull the barbell is moved toward the lifter. With the lowering of theknees a small opposite movement occurs, during which the second pull and thedrop under the barbell again results in movement toward the lifter. In the caseof nearly all lifters in Group A, the barbell's pathway does not cross a verticalreference line projected upward from the initial position of the bar. This appliesto many Group B lifters too. As a rule, this movement ends with a jump back-ward in the drop under the barbell. Garhammer (1985) has already drawn atten-tion to this jump, which is considered by Vorobyev (1978) to be a fault.

The patterns of Group A and B lifters from various weight categories arepresented in Figure 5 for comparison. In particular the patterns show the maxi-mum changes in horizontal displacement during the various phases of the lift.It is clear from this analysis that the new technique of pulling the bar towardthe lifter is widely used, and the total variation in horizontal movements of GroupA lifters is noticeably less than for Group B. The extent of these movements in-dicates the degree of instability involved or, as the case may be, the degree ofcorrection needed to complete the lift. This parameter also serves as a measureof the additional acceleration and mechanical work that must be produced.

Tertictl

110-i- kg Class

Figure 5 — Horizontal variations of the barbell in the weight categories 60 kg and110+ kg. The horizontal bars represent the standard deviations of the horizontalexcursions of the barbell at their extreme left and right positions, respectively.


The maximum height attained by the barbell increases as the weights ofthe lifters increase. This variable essentially depends on body stature, as can beseen from Figure 6. With very little variation, the height to which the bar is lift-ed in snatch lifts corresponds to 60% of the lifter's stature.

Velocity of tbe Barbell. The velocity-time relationship of the bar, par-ticularly peak vertical velocity, is an important dimension as far as coaches areconcerned. Figure 7 illustrates two typical velocity curves. The one on the lefthas two velocity peaks while the one on the right shows a steady increase in ve-locity to a single maximum value. The latter is characteristic of better weight-lifters, and even in the case of a delayed rise in the velocity curve. Group A liftersseldom show any notable dip in velocity.

The absolute maximum velocity ofthe bar increases with increasing weightcategories. The bigger loads that can be lifted by Group A are generally relatedto smaller maximum velocity values. These and other descriptive results of thissection are summarized in Table 2.

As expected, the average power output (Pav) during the raising ofthe barup to its highest point {BBH max) shows significant differences between GroupsA and B, mainly due to the higher barbell weights lifted by Group A, and toa lesser extent, to the generally shorter duration of the lift.

The general trends allow the following differences to be identified with

1 . 2 5

1.0

.75

BfiHnuiz [ml

8

go AA A

8

0 " _ .A " " " B

D A A. AO A A ^

o o'

-1 L J L

1.5 1.6 1.7 1.8

oo

6075

9o110+

88

kg o

kg A

BHliQ]I

1 .9

Figure 6 — Correlation between maximum effective lift of barbell (BBH max) andbody height (BH).

76 BAUMANN, CROSS, QUADB, CALBIERZ. AND SCHWIRTZ

2.5

2.0

1.5

1.0

.5

0.0

Velocity Im/s]

time Is]

Figure 7 — Different types of barbell velocities.

respect to Group A: (a) the maximum barbell height attained is less; (b) the maxi-mum second peak is smaller; (c) the duration of the lift is shorter; and (d) bodystature is slightly shorter. These characteristics reflect the better technique of theGroup A lifters.

Kinematics of Body Motion (Three-Dimensional)

The classic phase stnicture of the snatch technique is based mainly on the changesin knee joint angle. In general the differences between first and second pull phasesare identified. The first pull begins with the extension of all joints of the lowerextremity (other definitions start the first pull with the lift-off of the barbell; Pietka& Spitz, 1978). The angle at the knee reaches a maximum and then decreasesbriefly, with the smallest angle reached marking the end of the first pull. Theextension that then begins proceeds to a maximum, and it is this movement thatis designated as the second pull. During this phase, maximum barbell velocityis attained, followed hy the drop under the barbell.

Two-dimensional analyses of competitive movements are generally facedwith certain problems. The first concerns the obstructed view of the knee behindthe weights themselves over a fairly wide range of movement, particularly at cer-tain critical points, which allows less than adequate precision in measurement.The second concerns the projection of body angles in a single plane, which maydistort the true values. The extent of error in the two-dimensional estimation ofknee angle, for example, is clearly illustrated in Figure 8.

One curve represents the true changing knee angle between the long axisof the thigh and lower leg calculated from three-dimensional data. The second,using the same basic data, represents the changing knee angle in two dimensions,that is, confined to the sagittal plane. The variation between the angles calculated

SNATCH TECHNIQUE OF WEIGHTLIFTERS 77

1

Parameter

BBW

BH

GW

BBHmax

BBHL

Vmaxi

\/max2

T1 " .

T2

BBHmax/BHPav

Unit

kg

m

m

m

m

m/s

m/s

s

s

w

Table 2

Barbell Kinematics

60kgA B

n=10 n = 11

130.87.51.570.04

0.830.05

0.880.05

0,100.02

1.310.05

1.650.08

0.620,04

0.890.03

0.561269

105.08.7

1.600.020.860.03

0.970.05

0.130.05

1.310.16

1.760.12

0.650.10

0.950.08

0.611052

Weight class/group

75kgA B

n = i o n=10

157.02,7

1.640.030.870,02

0.960.03

0.120.02

—

1.740.06

0.670.07

0.960.05

0.591541

130.75.7

1.670.050.900.03

0.990.05

0.120.02

—

1.760.12

0.700.08

1.000.08

0.591269

90kgA B

n = 10 n = 11

174.33.0

1.740.050.880.09

1.060.07

0.140.02

1.390.05

1,860,09

0.730.10

0,990,09

0,611830

147.54.7

1,720,030,850,04

1.050.04

0.110.05

1.390.13

1.850,08

0.700.05

0.990.06

0.611534

110+kgA B

n=10 n=10

184.05.4

1.850.021.000,05

1,130.060,140.05

1.430.091.800.12

0.740.06

1.030.06

0.611981

144.37.9

1.830.041,000,06

1,160.05

0,140,06

1.420.19

1.880.08

0.740.11

1.060.08

0.631549

KNEE JOINT ANGLE Io]

leo

foot-off

go

1.00 time [s]

Figure 8 — Comparison between three-dimensional and two-dimensional knee angle.

78 BAUMANN, CROSS, QUADE, CALBIERZ, AND SCHWIRTZ

is clear to see and is about 15' at full bend. Such differences can vary accordingto the size of the true angle involved and the orientation of the plane of visionto the plane of movement. Thus it appears that a two-dimensional estimation ofbody angles (i.e., with one camera only) offers limited scientific application. Asatisfactory solution to this problem is only possible through a three-dimensionalanalysis using two appropriately positioned cameras.

Figure 9 shows the characteristic curves of the angles of the lower limbas well as the velocity curve of the bar in Shalamanov's world record lift (143

Velocity

•

•

•

of btrbeU

/

[m/tl

L.

. — — ^

50

Sh&luiiuioT/60/U3/+

TOSLD BECOBD

GOLD MEDAL

V m t x 2r\1

\

\ .,00 N ime [

[s]

BODY ANGLSS loi foot-off

180

00 time [sj

Figure 9 — Velocity of barbell and ai^es of tbe lower extremity and angle betweentnink and horizonUl. TPl and TP2 are the times of 1st and 2nd pull, respectively.TPl = lm-off until min. of knee angle; TP2 = min. of knee angle until 2nd max.of knee angle (Pietka & Spitz, 1978).


kg barbell, 60 kg weight category). The knee angle increases to an initial maxi-mum even before the lift-off of the barbell, followed by a pronounced bend atthe knees until a minimum angle is reached (end of first pull). Then there is aclearly shorter intensive extension made up to the second and absolute maximumknee angle achieved (end of second pull). The changes in angle at the ankle aregenerally similar to those at the knee, but of much less magnitude. The angleat the hip increases steadily over the two phases to a maximum that coincideswell with the maxima ofthe other two joint angles. All three lower limb anglesin fact reach their maximum values within 0.04 sec of one another at the endof the second pull. The angle between the trunk and the horizontal decreases slight-ly at the beginning of the first pull, as the increase in knee angle at this pointis greater than that of the hip. Once the initial maximum knee angle has beenreached, the trunk angle increases steadily to a maximum that coincides approxi-mately with the foot-off, which precedes the action of dropping under the bar-bell. The maximum vertical velocity of the barbell is reached just before maximumextension of all the joints approximately 0.05 sec before foot-off.

Listed in Table 3 are the major kinematic results of 17 selected attemptsfrom various weight categories. These data are purely descriptive and provideinformation on successful and unsuccessful lifts by world class athletes, includ-ing Shalamanov's world record (143 kg) and the Gold Medal winning lifts of

Subject

Shalamanov

KritskyTrautmanSenetVardanyan

Erhard

TsJntsanls

Gunyashev

Krastev

BW

kg

60

B2.5

134

136

130

150

Kinematics

BBW

kg

135140143

140147.5

-150-175175177.5

-145145160

-165195

-202.5202.5

-205

Table

ofthe

3

Body Motion

Joint angles"

max1

129121123

146147145154152151

159157142143144153158155

Knee

min r

115111114

128126129130134137

136134131129132132138138

nax2

172167165

176166167169168167

172175168170172176174171

Mip

max

214212210

209187190186181182

193195204207209207205203

TP1

s

0.460.440.40

0.480.400.460.420.420.42

0.400.520.560.520.580.580.600.64

Max. an£TP2

s

0.200.240.26

0.140.140.180.160.140.14

0.120.160.160.180.160.160.160.14

]. vel.Knee

PI

321235223

258281361332350332

556229223258264287292264

P2

372361298

424384338235281223

361350327304315413344264

"/secHip

P2

418470504

527436436527407413

332344378556510481430453

80 BAUMANN, GROSS, QUADE, CALBIERZ, AND SCHWIRJZ

Vardanyan (177.5 kg) and Krastev (202.5 kg). The size of this sample, togetherwith the enormous differences in antbropometric characteristics and lifting tech-niques between the various weight categories, does not allow these particular datato provide any worthwhile characteristics of better perfonnances. Even when com-paring good and poor lifts by the same athlete, the kinematic data do not revealany particular trends. In fact, a comparison of the top lifters Shalamanov andVardanyan reveals remarkable differences in all the measured parameters in boththe first and second pulls.

As a rule, however, it appears that as the barbell weight (BBW) increases,the duration of the first pull also increases, whereas with the exception ofShalamanov the duration of the second pull remains approximately constant ataround 0.15 sec. This result agrees well with the findings of Vorobyev (1978).The tnaximum angtilar velocities around the knee in the second pull are also gener-ally larger than in the first pull, and the extension of the hip occurs faster thanat the knee.

Kinetics of the Body/Barbell System

The forces measured between the ground and the lifter do not represent the en-tire forces involved in the movements since, until lift-off, forces are being ap-plied directly on the barbell from the competition platform. As soon as lift-offoccurs, GRF with respect to the lifter accounts for all the forces acting on thesystem. Without careful kinematic analysis of the movements of the lifter andbarbell, the effects ofthe GRF on the various parts ofthe system cannot be iden-tified. The results with respect to external and internal kinetics are treated separate-ly in the following section.

External Kinetics (GRF). Ground reaction forc^ were recorded separatelyfor the right and left feet. Figure 10 shows an example of the forces recorded,giving not only an impression of the three orthogonal force components but alsothe degree of symmetry between right and left, though this aspect is not consid-ered bere. All three force components are of course included in the calculationof the muscular moments. In this section the results of the vertical componentofthe GRF during the technically most important phase (i.e., from the beginningof the lift to the drop under the barbell) will be presented. Figure 11 depicts atypical force curve combining right and left sides together with definitions of thespecific parameters used.

The first intersection between the force curve and the line representing thetotalmassof the system (i.e., body weight + barbell) defmes the point of lift-offofthe barbell. Table 4 presents the numerical results of 17 lifts. The differentmaximum force values (Fzl and Fz3) rise steadily as barbell weight increasesfrom 135 kg to 205 kg. The correlation coefficients between each of these valuesand the total mass of the system were both found to be highly significant (r >0.97). There would appear to be a relationship between rate of change of force(RFz) and total mass, but this was not the case as no evidence of a relationshipwas found.

Figures 12a through 12j show the vertical force-time curves for 10 selectedlifts from two weight categories (82.5 kg and 110+ kg). The scale used in eachis the same, and the horizontal line in each represents the respective individualcombined weight of the lifter and barbell.


6000

4000

2000

lme [a]

Figure 10 — Example of all registered force components during snatch.

6000IRASTET/+110/202.5kcA

4000

2000

2.00 time [s]

Figure 11 — Vertical component of GRF (Gold Medal in 110 + 1% weight cat^ory).


Table 4

Selected Parameters of Ground Reaction Forces

Subject

Shalamanov

KritskyTrautman n

SenetVardanyan

ErhardTsintsanis

Gunyashev

Krastev

BWkg

60

82.5

134136

130

ISO

BBWkg

135140143

140140145147.5

-150-175

175177.5

-145160

-165195

-202.5202.5

-205

RFzkN/s

17.528.520.0

12.07.06.56.4

24.014.512.515.0

14.59.0

12.011.011.022.028.0

FziN

257325322501

28572867292529073061335033103312

3743375737934060401943354290

Fz2N

172518111749

16681675178517731663221820852235

1701208025562312245321802151

Fz3N

251324932462

32503325339233353168364035713441

4335413341054510463349585028

internal Kinetics (Muscular Moments). The muscle moments, as pre-viously stated, are calculated from the carefully determined spatial coordinatesof the center of the joints of the lower extremity and the resultant GRF vector.Despite the limitation that only net muscular moments are considered, they stillrepresent a very important kinetic parameter that is closely related to the muscu-lar control of the movement. The muscular moments of one leg are shown inFigure 13. The chosen example is typical of lower weight categories. When acurve is positive the extensors must be active, whereas when it is negative theflexors must be active (mathematically speaking, the curve for the knee shouldin fact be inverted). As the barbell nears lift-off and up to the end of the firstpull, the moment about the hip joint becomes approximately constant while themoments about the other two joints decrease. The moment about the knee be-come negative, that is, the knee flexors become active. The relatively rapid flexingof the knees is only possible with the active support of the flexor muscles. Atthe beginning of the second pull all of the moments are once again positive, theknee and ankle moments reaching a second maximum during this phase beforedeclining rapidly. Immediately before the foot leaves the ground (foot-off) allthree moments reach zero.

These general patterns of the moment-time curves are very reproduciblefor individual athletes, though they change somewhat between athletes as a resultof performance differences. In the heavier weight categories there is an essen-tially similar pattern in the first pull, although variations do occur in the secondpull.

SNATCH TECHNIQUE Of WEIGHTLIFTERS 83

a)

b)

d)

e)

TSINTB*MI8/+110/tBOka/*

g)

, L

Figure 12 — Total GRF, vertical component for selected attempts in the weightcategories 82.5 kg (a-e), and 110 +kg (f-j), respectively. The data given arename/weight category/weight of barbell/ + = successful, - = unsuccessful.

84 BAUMANN, CROSS, QUADE, CALBIERZ, AND SCHWiRTZ

lBOfoot-off

400

200

-200

Figure 13 — Muscular moments acting in tbe joints of the leg from beginning untilfoot-off. Top = knee angle.

The numerical results of 15 individual attempts are presented in Table 5.It is interesting to note the relatively small knee and ankle moments. Rangingfrom the 60 kg weight category to the 110+ kg weight category, the maximumvalues for the knee extensors He between 65 Nm and 258 Nm. The knee flexorscreate maximum net moments of between 65 Nm and 161 Nm. From these figures,characteristic of competitive movements at the highest level, it is impossible toexplain problems reported from the apparent overloading of the joint. The hipextensors must be active for a significantly longer period of time, which includesthe first pull and adds up to between 0.4 sec and 0.6 sec. In addition they haveto compensate for net moments 2-4 times larger than in the case of the knee,that is, from 260 Nm to 660 Nm. These values indicate the dominant role of thehip extensors in weightlifting and provide evidence of the massive loads to whichthe muscles and joint structures of the hip are exposed.

There is a high correlation between maximum moments at the hip andthe total mass of the system (r = 0.95), which also means that increasing barbellweights leads unavoidably to increased loads on the hip (see Figure 14). The cor-responding values for the knee joint show that the moments do not increase propor-tionally with external loads. In Figure 14 the second peak maximum moment aboutthe knee (MK max2) and maximum negative moment in the opposite direction


Table 5

Extreme Values of Muscular Moments

Subject

Shalamanov

KritskyTrautmannSenetVardanyan

ErhardTsintsanis

GunyashevKrastev

BW

kg

60

82.5

134136

130150

BBW

kg

135140143

140147.5

-150-175175177 5

-145160

-165195202.5

-205

Max1

127150152

165137169200187142

221193202222214204

MF [Nm]

Min

276832

462534026

59618726102116

Max2

149134135

145168188204201151

222201191264192265

Max1

746582

129178126154183170

245220207222244221

MK [Nm]

Min

-73-85-85

-38-127-101-120-98-65

-161-94-115-84-159-132

Max2

6587100

160116954989—

178258227120187175

MH [Nm]

Max

260280300

320340370450420440

560550520640660560

. Masculu* Moments [Nm]6 0 0

4 0 0

2 0 0

o = H Eo = MKmaz2A = MEmin

Bf+BBf [k|]1

l a o 2 3 0 aao 3 3 0 3 8 0

Figure 14 — Plot of selected moments in the leg versus system mass (BW + BBW).


about the knee (MK min) are plotted against the total mass of the system. Thecorrelation coefficients for the extensors (r = 0.61) and the flexors (r = 0.57)were found to be relatively weak and can perhaps be explained through differ-ences in technique.

The moment at the knee joint is dependent upon the magnitude of theresultant GRF and the perpendicular distance between the line of action of thisforce and the joint center. Through changes in this moment ann, the momentabout the joint will vary. Figure 15 shows an example using the heaviest weightcategory, comparing Gold Medalist Krastev (BW 150 kg, BBW 202.5 kg) andGroup B lifter Tsintsanis (BW 136 kg, BBW 160 kg). The vertical component

sooo

IruteT

Tiintitnii

rifht foot

tltna Cs]

2 O O

time ts]

time [s]

-200

Figure 15 — Influence of the moment ann of GRF on the muscular moment at theknee.


of the GRF is larger for Krastev because of the greater mass involved. However,the maximum moment about the knee joint is quite the reverse, Krastev's in factbeing smaller. The explanation lies in the different moment arms of the GRF,Tsintsanis' being larger than Krastev's. The larger moment of the knee flexorsin Krastev's case simply underlines the forceful support offered by the knee flex-ors at the end of the first pull.

This comparison demonstrates two things: first, that knowledge of the GRFalone is insufficient for deducing anything about the muscular activity involved,and second, that the position of the knee joint with regard to the GRP directionappears to be an important technical factor in this phase of the snatch. This factorcan significantly influence the forces transmitted by the joint and surroundingmusculature. The position of knee joint is certainly as important as the time his-tory of the knee angle.

Summary and Conclusions

The resuits describe some important aspects of snatch technique. Since the origi-nal data come from world class athletes under the most demanding competitiveconditions, they can be used as reference data not only for coaches and athletesbut also with respect to future biomechanical research. Figure 16 summarizesthe picture of the most important characteristics in a lift.

Most of the results with respect to the kinematics of the barbell and move-ment of the body are in good agreement with the results reported in the litera-ture. The exception is the pathway of the barbell, which has clearly changed,coming more toward the lifter during the first pull. As a consequence there isa backward jump during the drop under the barbell, which Vorobyev consideredto be a fault in technique.

The most important results appear to be those concerning the internal kine-tics, namely the muscular moments. These parameters are closely related to themuscular control of the movement. In light of the present findings, it would ap-pear that the division of the snatch technique into its usual pull phases is no longerentirely logical. Such a structure is derived from the kinematic characteristicsof the movement of the barbell or change in knee angle (Vorobyev, 1978; Pietka& Spitz, 1978). An alternative to this has been suggested by Kauhanen et al.(1984), who based division of the technique into three phases on the sole crite-rion of minimum and maximum knee angles. According to this, the vertical com-ponent of the GRF could also be divided into phases of eccentric and concentricactivity of the knee extensors. However, this would be a mistake and might leadone to draw false conclusions. It is in fact not possible to infer muscular activityfrom the measurement of knee angle and vertical GRF alone. As has been shown,it is the time history of the muscular moment that in fact gives insight into thedetailed control of the movement. However, it must be noted that the momentscalculated here are only net moments of force since antagonistic muscle activityhas had to be neglected. Also, the problem of two-joint muscles has not beenaddressed, including for example the division of force between m. soleus andm. gastrocnemius. The problem remains unsolved, despite its obvious impor-tance, because the effects on the results with respect to the knee and even thehip joints are unknown. The addition of EMG measurements would be an impor-tant consideration in future studies.

68 BAUMANN, GROSS, QUAOE, CALBIERZ, AND SCHWIRTZ

IBO

90

200

-Velocity ofbarbell [m/<]

GROUND REACTION PORCE [N]

-200

Figure 16 — Complete set of parameters. From top to bottom: body angles, barbellvelocity, vertical component of GRF, muscular moments.

SNATCH TECHNIQUE OF WEIGHTLIFTERS . 89

Nevertheless, it seems reasonable to use muscular moments as suitablecriteria for the overall structural characterization and division of the lifting tech-nique. The biomechanical structure ofthe movement includes not only kinematicbut also internal and external kinetic dimensions. It is not the ease of measure-ment but the essential system parameters and their interrelationships that are ofprime importance when choosing appropriate scientific methodologies. If we areprepared to make such considerations in our analyses, we will probably increaseour understanding ofthe weighdifters' movements. And finally, there is the prob-lem of how the kinetic features of the movement can be translated into meaning-ful movement terms that can be understood and applied at the practical level.

References

Abdel-Aziz, Y.I., & Karara, H.M. (1971). Direct linear transformation from comparatorcoordinates into object space coordinates in close range photogrammetry. Proceedingsof ASP/UI Symposium on Close Range Photogrammetry, Illinois.

Enoka, R.M. (1979). The pull in Olympic weightlifting. Medicine and Science in Sports,11, 131-137.

Garhammer, J. (1979). Performance evaluation of Olympic weightlifters. Medicine andScience in Sports, 11, 284-287.

Garhammer, J. (1981). Biomechanica] characteristics of the 1978 world weightliftingchampions. In A. Morecki, K. Fidelus, K. Kedzior, & A. Wit (Eds.). BiomechanicsVII-B (pp. 300-304). Baltimore: University Park Press.

Garhammer, J. (1985). Biomechanical profiles of Olympic weightlifters. InternationalJournal of Sport Biomechanics, 1, 122-130.

Kauhanen, H., Hakkinen, K., & Komi, P. (1984). A biomechanical analysis ofthe snatchand clean & jerk techniques of Finnish elite and district level weightlifters. Scan-dinavian Journal of Sports Science, 6, 47-56.

Ono, M., Kubota, M., & Kato, K. (1969). The analysis of weightlifling movement atthree kinds of events for weightlifting participants ofthe Tokyo Olympic games.Journal of Sports Medicine, 9, 263-281.

Pietka, L., & Spitz, L. (1978). Kinematic—Bewegungsbeurteilung im Raum [Evaluationof movement in space]. In Bundesverband Deutscher Gewichtheber (Ed.), Lehr-beilage Gewichtheben, 4(1), 3-17.

Vorobyev, A.N. (1978). Weightlifting. Budapest: IWF.

Acknowledgments

This investigation was initiated and supported by the Subcommission for Bio-mechanics and Sport Physiology ofthe Medical Commission ofthe IOC (Chairman, PrinceAlexandre de Merode). We are grateful to the Federal Institute for Sport Science (BISP)who supported the analysis, to Kistler Instruments for providing special force plates, andto Data General for their computer assistance.

Thanks are also extended to The International Weightiifting Federation (President,G. Schodl; General Secretary, T. Ajan) and the Organizing Committee ofthe SwedishWeightlifting Association (C.-E. Hermansson, S. Johansson, B. Johansson, and T. Tor-stenson) for their excellent support of this project.