35..
‘1: .
a. 33.19..
.1 “It? I

Jrfi
n. ‘Dnnuw.

try.

1:55,
‘ r I.
‘ Z. :1

wu‘sxmufi....

i... 5

i hymn" m3?§ 5%
Law}: .

:xzai.
.‘\‘ 0‘, V kl 3““
£25.... :3.an
A... uzhllsil. .Iavl .
3., :5...“
. .5 x. . ..

 

Alt; 4L ‘ »?I n
r?“ 31.?th xvii". 1 34.51
..,. 1...... {if}

”6! xmofic

. 53TH?

L.
1 . 24 Ylu‘tu \ 41.19.. t

 

 

THESIS

250;

This is to certify that the

thesis entitled

THREE-DIMENSIONAL BIOMECHANICAL ANALYSIS OF
LANDING FROM GRAND JETE: THE EFFECT OF
BALLET FOOTWEAR ON SELECTED KINETIC AND

KINEMATIC VARIABLES
presented by

Ethel Ruth Leslie

has been accepted towards fulfillment
of the requirements for

_M.._S.._degree in Kinesialogy

me'

Major professor

Date ATAJS‘AS-f lqlazfldl

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

W 00;? tags

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c-JCIRCJDatoDua.p65-p.15

 

THREE-DIMENSIONAL BIOMECHANICAL ANALYSIS OF LANDING
FROM GRAND JETE: THE EFFECT OF BALLET FOOTWEAR ON
SELECTED KINETIC AND KINEMATIC VARIABLES

By

Ethel Ruth Leslie

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Kinesiology

2002

ABSTRACT
THREE-DIMENSIONAL BIOMECHANICAL ANALYSIS OF LANDING
FROM GRAND JETE: THE EFFECT OF BALLET FOOTWEAR ON
SELECTED KINETIC AND KINEMATIC VARIABLES
By

Ethel Ruth Leslie

This project was a case study which examined the effect of ballet footwear on the
three-dimensional sagittal plane kinematics and three-dimensional kinetics of landing
from grand jeté. The following variables were measured: (1) angular displacement,
velocity, and acceleration of the ankle and knee joints, (2) peak vertical ground reaction
force (GRF) and rate of loading, and (3) force-time characteristics of GRF components.
One ballet student was filmed landing onto a force plate under three foot conditions:
pointe shoes, ballet slippers, and barefoot There was no apparent difference between
conditions for the pattern or magnitude of angular displacement of the knee and angular
velocity and acceleration of knee flexion and ankle dorsi-flexion. Average dorsi-flexion
angular displacement was lower for the barefoot condition. Peak vertical GRF ranged
from 3.06 to 5.12 BW. On average, all three components of GRF and peak rate of

loading vertical GRF were highest for the pointe shoe condition.

C0pyfight by
ETHEL RUTH LESLIE
2002

“Dance is the only art of which we ourselves are the stuff of which it is made.”
Ted Shawn

“Think of the magic of that foot, comparatively small, upon which your whole weight
rests. It’s a miracle, and the dance is a celebration of that miracle.”
Martha Graham

iv

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. V. Dianne Ulibarri and the members of my
thesis committee — Dr. Claudia Angeli, Dr. Dixie Durr, and Dr. Robert Malina — for their
guidance. 1 thank fellow student Miguel Narvaez for his help throughout this project and
for his friendship throughout my time at Michigan State University. Thank you to
Madeline and Joe Anthony for their hospitality and friendship during the process of
completing my degree. I extend gratitude to my best friend, Tommy Parlon, for his
relentless and unwavering support. Finally, I thank my family, especially my parents,
Drs. James and Louisa Leslie — their endless reserves of love and encouragement made

this degree possible.

TABLE OF CONTENTS

LIST OF TABLES ..................................................... viii
LIST OF FIGURES ..................................................... ix
CHAPTER 1
INTRODUCTION ....................................................... 1
CHAPTER 2
REVIEW OF LITERATURE ............................................... 4
Description of the Pointe Shoe ....................................... 4
Kinetic Analysis of Dancing in the Pointe Shoe .......................... 4
Kinematic and Kinetic Studies of Aerial Movement in Dance .............. 11
Joint Moments ................................................... 18
Explanation of Grand Jeté .......................................... l9
Biomechanical Analysis of Grand Jeté ................................ 20
Kinematics and Kinetics of Landing from Grand J eté ..................... 24
Problem Statement ................................................ 27
Hypotheses ...................................................... 28
CHAPTER 3
METHODOLOGY ...................................................... 29
Instrumentation .................................................. 29
Subject .................. 29
Calibration Space ................................................. 31
Subject Preparation ............................................... 33
Identification of Rigid Bodies in Space ................................ 36
Data Analysis .................................................... 37
CHAPTER 4
RESULTS AND DISCUSSION ............................................ 38
Kinematics ...................................................... 39
Flexion/extension angular displacement. ........................ 39
Flexion/extension angular velocity. ............................ 45
Flexion/extension angular acceleration. ........................ 51
Kinetics ........................................................ 57
Ground Reaction Forces. ..................................... 5 7
Vertical ground reaction force (FZ). ..................... 57
Braking/acceleration ground reaction force (FY). ........... 63
Medial/lateral ground reaction force (FX). ................ 66
Rate of Loading Vertical Ground Reaction Force. ................ 68

vi

CHAPTER 5

SUMMARY AND CONCLUSION ......................................... 7O
Findings ........................................................ 70
Conclusion ...................................................... 71

APPENDICES ......................................................... 75
Appendix A: Glossary of Terms ..................................... 76
Appendix B: Detailed Description of the Pointe Shoe .................... 78

REFERENCES .............. . .......................................... 82

vii

LIST OF TABLES

Table 1. Peak angle of knee flexion and ankle dorsi-flexion with time at peak. ...... 43

Table 2. Maximum and minimum knee and ankle angular velocities with time at
occurrence. ...................................................... 49

Table 3. Maximum and minimum knee and ankle angular accelerations with time at

occurrence. ...................................................... 55
Table 4. Peaks of ground reaction forces with time at peaks. .................... 61
Table 5. Loading rate of peak vertical GRF ................................... 68

viii

LIST OF FIGURES

Figure 1. Labanotation of grand jete’ pas de chat from Gail Grant dictiom of classical
ballet in Labanotation (p. 72), by A. Miles, 1976, New York: Dance Notation
Bureau. Copyright 1976 by Dance Notation Bureau, Inc.. Reprinted with

permission. ...................................................... 19
Figure 2. Calibration set-up and camera position for data collection. .............. 32
Figure 3. Placement of reflective markers. .................................. 35

Figure 4. Knee flexion/extension angular displacement for the three barefoot trials.
............................................................... 40

Figure 5. Knee flexion/extension angular displacement for the three slipper trials. . . . 40

Figure 6. Knee flexion/extension angular displacement for the two pointe shoe trials.

............................................................... 41
Figure 7. Ankle plantar-ldorsi-flexion angular displacement for the three barefoot trials.
............................................................... 41
Figure 8. Ankle plantar-ldorsi-flexion angular displacement for the three slipper trials.
............................................................... 42
Figure 9. Ankle plantar-/dorsi-flexion angular displacement for the two pointe Shoe
trials. .......................................................... 42
Figure 10. Knee flexion/extension angular velocity for the three barefoot trials. ..... 46
Figure 11. Knee flexion/extension angular velocity for the three slipper trials ........ 46
Figure 12. Knee flexion/extension angular velocity for the two pointe shoe trials. . . . . 47
Figure 13. Ankle plantar-/dorsi-flexion angular velocity for the three barefoot trials.
............................................................... 47

Figure 14. Ankle plantar-ldorsi-flexion angular velocity for the three slipper trials. . . . 48

Figure 15. Ankle plantar-/dorsi-flexion angular velocity for the two pointe shoe trials.
............................................................... 48

Figure 16. Knee flexion/extension angular acceleration for the three barefoot trials.
............................................................... 52

ix

O D

Figure 28.

. Knee flexion/extension angular acceleration for the three Slipper trials. . . . 52

. Knee flexion/extension angular acceleration for the two pointe shoe trials.

............................................................. 53
. Ankle plantar-/dorSi-flexion angular acceleration for the three barefoot trials.
............................................................. 53
. Ankle plantar-ldorsi-flexion angular acceleration for the three slipper trials.
............................................................. 54
. Ankle plantar-ldorsi-flexion angular acceleration for the pointe shoe trials.
............................................................. 54
. Vertical grormd reaction force for the three barefoot trials. ............. 58
. Vertical ground reaction force for the three slipper trials. .............. 58
. Vertical ground reaction force for the two pointe shoe trials ............. 59
. Braking/acceleration GRF for trial PT4. ............................ 64
. Medial/lateral GRF for trial PT4. ................................. 67
. Five positions of the feet in ballet. ................................ 77

The pointe shoe with main components identified‘ (A) the box, (B) the outer

shank, (C) the inner shank. Ribbons and elastic are not attached as they most
likely would be for rehearsal and performance. (Photograph by Miguel Narvaez.)

............................................................. 79

Chapter 1
Introduction

Ballet is a challenging form of movement which has its roots in the court dances
performed by European royalty during the Renaissance (Hammond, 1974). The physical
and aesthetic requirements of ballet technique make it one of the most demanding forms
of physical activity, yet there has been relatively little scientific study of actual ballet
movements.

Probably the most widely recognized and intriguing characteristic of ballet is the
use of special shoes that allow the ballerina to dance on her toes, or “en pointe”. While
female ballet dancers traditionally perform en pointe, male dancers have been known to
dance in pointe shoes as part of a character role or as training to strengthen their feet
Since feet and toes were not designed to bear weight in this fashion, dancing en pointe
may contribute to a number of injuries and pathologies seen in ballet dancers. Sprained
ankles, bunions, calluses, and bruised toenails are common among ballet dancers
(Kravitz et al., 1986; Milan, 1984; Quirk, 1994; Sammarco, 1982; Weiker, 1988) as are
structural adaptations of the bones in the feet (Kravitz, Fink, Huber, Bohanske &
Cicilioni, 1985; Sammarco, 1982; Schneider, King, Bronson & Miller, 1974). Tissue
damage that results in injury and/or structural adaptation can be caused by one or more of
the following: 1) a single maximum load, 2) repetitive loading, and 3) the rate of loading
(Nordin & Frankel, 1989). The injuries and adaptations experienced by a dancer’s body
are in response to the extreme demands inherent to dance movement and the repetitive
nature of dance practice.

One of the more frequently performed movements in ballet, and dance in general,

is grand jeté. Grand jeté is a spectacular leap that demands the dancer jump high into the
air and form a 180 degree sagittal split of the legs. Upon landing, grand jeté has been
known to generate peak vertical ground reaction forces of at least three times body
weight (Becker, 1984; Miller, Paulos, Parker & Fishell, 1990). Take-off and landing
kinematics have been given some study (Murgia, 1995; Ryman, 1976) as has the effect of
body configuration on the illusion of suspension (Midgett, O’Bryant, Stone & Johnson,
1993; Rasmussen & Hay, 1993). Miller et al. ( 1990) found pressure more evenly
distributed upon landing from grand jeté wearing the ballet slipper as compared to
landing barefoot; however, they offered no explanation to account for this finding. The
influence of pointe shoes on landing from jumping movements, such as grand jete’, has
not yet been investigated.

To date, scientific study of ballet movements, and dance movement in general,
has been scarce. Minimal attention has been given to the investigation of dance shoes
and the influence of the shoes on movements dancers perform daily, as well as the effects
of these movements on joint mechanics and stressors. While most athletes have specially
designed shoes developed to protect their body from physical stress, ballet dancers have
used the same footwear design for over a century (Barringer & Schlesinger, 1998). Made
of little more than fabric and leather, ballet slippers and pointe shoes provide the dancer
with little, if any, protection from potential injury. The current study was necessary in
order to better understand landing movements from aerial movements in dance and the
effect of dance footwear on the performance of these movements. Increased
understanding of movements such as grand jeté may lead to the development of footwear

and other equipment that can provide protection and support for dancers’ feet.

The purpose of this study was to examine the effect of ballet footwear on landing
from grand jeté using three-dimensional sagittal plane kinematics and three-dimensional
kinetics. For the purposes of this study, landing was defined by the time the dancer was
in contact with the force plate, from toe down to toe off. One advanced level ballet
student performed grand jeté under three foot conditions: 1) barefoot, 2) wearing pointe
shoes, and 3) wearing ballet slippers. Three-dimensional high-speed videographic and
force plate data were used to investigate the following: (1) angular displacement of the
knee and ankle joints, (2) angular velocity of the knee and ankle joints, (3) angular
acceleration of the knee and ankle joints, (4) peak vertical ground reaction force and rate
of loading, and (5) force-time characteristics of ground reaction force components (FX,

FY, FZ).

Chapter 2
Review of Literature

Descripg'on of the Pointe Shoe

Marie Taglioni commonly is credited with being the first major ballerina to dance
on her toes as an essential choreographic element by perfomring the title role in the
ballet La Sylmidg in 1832 (Barringer & Schlesinger, 1998). To dance en pointe,l the
ends of her shoes were darned and she probably padded the ends of her ballet slippers
with cotton wool. As ballerinas were asked to perform more difficult feats en pointe, the
toes of ballet slippers were darned and stiffened with glue and the insole was reinforced
with light cardboard to provide support (Barringer & Schlesinger). Pointe shoes today
are heavier than in the past (Lawson, 1983), yet are still little more than satin, glue, and
leather even as the choreographic demands of ballet become more and more challenging.
A more complete description of the pointe shoe is provided in Appendix B.
Kinetic Analysis of Dancing in the Pointe Shoe

A handful of studies (Albers, Hu, McPoil & Cornwall, 1992; Dozzi & Winter,
1993; Galea & Norman, 1985; Kravitz et al., 1986; Teitz, Harrington & Wiley, 1985 ;
Torba & Rice, 1993; Tuckman, Werner & Bayley, 1991) have investigated various
kinetic aspects of dancing in the pointe shoe. Several of the authors of these studies
explored occurrences inside the pointe shoe, while other authors utilized information of
external forces and pressure distribution to examine the mechanical demands on the foot
when en pointe. Pressure distribution inside pointe shoes have been investigated

quantitatively using transducers (Kravitz et al., 1986; Teitz et al., 1985) and qualitatively

 

1Definitions of dance terms are provided in Appendix A.

using a molding technique (Tuckman et al., 1991 ). Torba and Rice ( 1993) combined the
use of quantitative (force sensing resistors) and qualitative (molding and pressure
sensitive film) techniques in an effort to assess pressure distribution of the foot while en
pointe. External force and pressure data have been used in the exploration of bone-on-
bone forces at the ankle (Galea & Norman, 1985), plantar pressures (Albers et al., 1992),
and energetics of the foot and ankle (Dozzi & Winter, 1993) during a rise onto pointe.
Teitz et al. (1985) studied the pressure distribution on the foot in pointe shoes
using Kulite pressure transducers. Transducers were placed on the tips of the first and
second toes and on the medial aspect of the metatarsophalangeal (MTP) joint. All 13
participants wore a new pair of the same brand and style of pointe shoe. Each performed
plié—relevé two different ways. First they did relevé properly, rising through the center of
the foot, keeping their weight equally distributed over all the toes as the rise onto pointe
was completed. The movement was then done in an everted position, shifting most of
the dancer’s weight forward onto the big toe at the completion of relevé. Dancing with
the foot everted is thought to be a contributing factor to bunion formation. Although
eversion was found to increase the pressure on the first MTP joint for most subjects, the
authors of this study were not able to document the pressure’s contribution to the
formation of bunions. Those participants whose second toe was shorter tlmn their great
toe, placed a cap onto their second toe and performed an additional set of relevés. From
thedata,theauthors revealedthatthefirsttoetookatleastasmuch, ifnot more, pressure
than the second toe regardless of relative toe length or whether the second toe was
capped This finding contradicted a common belief among dancers and their doctors that

the best foot shape for dancing en pointe is one with the first two toes of equal length to

provide better pressure distribution (Barringer & Schlesinger, 1998; Contompasis, 1986;
Kravitz et al., 1985; Sarnmarco, 1982; Weiker, 1988). Dancers with shorter second toes
frequemly put a cap on it to conform to this assumption.

Pressure transducers also were used in a pilot study by Kravitz et al. (1986) to
examine the pressures at several locations on the foot inside the pointe shoe. Sensors
were placed on ten landmarks of the foot: seven on and around the first and second toes,
one on the plantar aspect of the first MTP joint, and two on the calcaneus. Sixteen
dancers participated in this study and the researchers used an electrodynogram computer
system to analyze the forces produced on the foot during releve' onto one and two feet.
Because the electrodynogram was able to record pressures over time, relative pressures
during the movement could be examined. The greatest percentages of peak force and
peak shock occurred on the plantar aspect of the first MTP joint during the transitional
period of the rise onto pointe. The authors did not indicate how peak shock and pressure
were calculated Pressures were greater when the dancers rose to a position on one foot
than to the releve’ on both feet. This finding was not surprising since the dancers would
be transferring full body weight onto one foot rather than dividing their body weight, as
occurred when rising to balance on two feet. Forces were generated through the medial
aspect of the forefoot as the dancer reached full pointe and settled into position. This
pattern of force during the rise onto pointe confirmed the visual observation made by
these researchers in preparation for the study: that as the dancer rose onto pointe, the foot
supinated, then pronated slightly as the final position was reached. When the dancers
balanced in their final position on one foot, the three highest pressures were recorded by

transducers located on the distal aspect of the hallux, the medial aspect of the first MTP

joint, and the dorsal aspect of the first interphalangeal joint, respectively. In the final
position on two feet, the three highest pressures, in order, were recorded by transducers
placed on the distal aspect of the hallux, the dorsal aspect of the first interphalangeal
joint, and the lateral tubercle. As was shown in comparison, the highest pressure for both
the one foot and two feet conditions were the same, while the second and third highest
pressures were completely different. As would be expected, as the dancers balanced on
one foot, the pressure recorded at the distal aspect of the hallux was greater than when
the dancers balanced on two feet. It was not made clear in the report whether the dancers
were: 1) tested wearing their own pointe shoes, 2) if the pointe shoes were standardized
in some way, or 3) whether the shoes were broken in These factors are important
because of their influence on the stiffness of the pointe shoe box. Stiffness of the box
would influence the measurements taken, especially on the first MTP joint, and should
be considered when interpreting the data.

Qualitative assessment of the forefoot while en pointe was done by Tuckman et
al. (1991) using a molding technique. An alginate mixture was poured into the box of
the pointe shoe and the dancer balanced en pointe for about one minute with minimal
support as the mixture solidified leaving a negative impression of the forefoot. Plaster
wasthenpouredintotheshoeand,oncehardened,theshoewascutawayleavinga
positive impression of the positioning of the toes en pointe. The nine student dancers
who participated were given their preference of new pointe shoes and were asked to
break in their shoes prior to testing. Pressure on the medial aspect of the foot was
evident because of hallux valgus and flattening of soft tissue, both attributed to the

unrelenting shape of the box. All molds except one demonstrated pressure on more than

one toe and about half of the molds showed toenail deformations. There did not appear
to be any relationship between the configurations of the toes en pointe and relative toe
length, pain while en pointe, or toenail problems. This finding was in agreement with
other research (Ogilvie-Harris, Carr & Fleming, 1995; Teitz et al., 1985) which found no
relationship between toe length and performance en pointe. Based on the results of this
study, the authors suggested that the shape of the foot outside the pointe shoe may not be
as important as the position of the foot inside the shoe. Tuckman and colleagues
hypothesized that the pressures on the forefoot may best be distributed if the toes are
aligned while en pointe. These researchers recommended further study to investigate this
assertion. Although the method used in this study was qualitative in nature, it presented
visual evidence of pressure distribution over the entire forefoot that may provide
direction for quantitative studies.

Torba and Rice (1993) used the alginate molding technique developed by
Tuckman et al. (1991) along with pressure sensitive film and force sensing resistors to
examine pressure distribution on the foot while en pointe. Neither the number of
subjects nor the technical ability of the subjects was provided; however, the brand and
style of pointe shoe worn by the subject(s) was noted. To assess the contribution of
friction to supporting body weight, the foot was covered with plastic food wrap, which
acted as a lubricant, and pressure was measured at bony prominences. Upon examination
of the plaster moldings, Torba and Rice (1993) found that flattening of soft tissue and
other adaptations were necessary to allow the foot to fit inside the shoe, similar to the
findings of Tuckman et al. (1991). Although pressure distribution varied among

subjects, higher peak pressures and higher readings from the force sensor resistors were

consistently recorded where the foot was flattened or touched the side of the shoe. The
pressure data of one subject from sampling with a force sensing resistor showed the first
toe took greater pressure than the second toe and was in general agreement with the
results of Teitz et al. (1985). Torba and Rice (1993) determined that 85% of the dancer’s
weight was accounted for by normal forces. The remaining percentage of body weight,
15%, was attributed to friction between the foot and shoe because lubrication of the foot
was found to increase normal forces at all bony prominences examined. The findings of
this study make sense because a normal force is defined as the force holding two objects
together. The findings also make sense when it is considered that a normal force is
proportional and perpendicular to friction.

Measurement of pressures inside the pointe shoe has led to a better understanding
of what happens inside the shoe. Pressure transducers provided measurement of the
pressures at a particular location, but they were not capable of recording the pressure
distribution across the entire forefoot or expressing values for the force components that
comprised pressure. The molding technique allowed a better understanding of the
formation of the foot while in the shoe. Combining these quantitative and qualitative
methods provided a more complete picture of occurrences inside the pointe shoe. The
authors of two of these studies (Teitz et al., 1986; Tuckman et al., 1991) have suggested
and agreed that the relative length of the toes may not be as important as the
configuration of the foot inside the shoe.

Galea and Norman (1985) used force plate and surface electromyographic (EMG)
data to estimate the bone-on-bone forces at the ankle joint during two footed relevé onto

full pointe. Bone-on-bone forces were the result of joint reaction force as well as muscle

and ligament forces acting at the joint. Bone-on-bone forces were calculated using a
model that combined the joint reaction force and the force of contraction of the muscles
crossing the ankle joint. Horizontal and vertical ground reaction forces, as measured by a
force plate, were used to determine joint reaction force. EMG data recorded during
relevé served as a measure of muscle activity. Muscle force and EMG outputs of a
maximum isometric contraction were measured for the muscles around the ankle. These
measurements provided a standard from which relative muscle forces during the
movement were determined The model also took into account changes in muscle length
and velocity of muscle contraction. Peak bone-on-bone forces were calculated to be as
high as ten times body weight and usually occurred when the dancer was on full pointe.
The authors pointed out this amount of force alone is not necessarily destructive, but
rather the repetitive loading of the joint surface. It should be added that the rate of
loading would affect the potential for injury (Nordin & Frankel, 1989).

The relative contribution of the ankle extensor and metatarsal-phalangeal flexor
muscle groups to the energetics of rising onto pointe were investigated by Dozzi and
Winter (1993 ). Two female professioml dancers, performers with a world class ballet
company, performed eleve' onto pointe. Time-history of the rise onto pointe showed that
during the first part of elevé, the foot segment raised while the phalangeal segment
remained stationary then, as the movement was completed, the phalangeal segment
moved upward A distinct pause associated with the transition from raising the foot
segment to raising the phalangeal segment was noted. Calculations of the mechanical
work done by both subjects showed that 22% and 33% of work was done at the

metatarsal-phalangeal joint. It should be considered that the mathematical equations

10

used to calculate the amount of work done at the two joints took into account vertical
ground reaction force but not medial-lateral or anterior-posterior reaction forces (Winter
& Robertson, 1978). Nonetheless, the percentage of work is noteworthy because, based
on cross-sectional area, the muscles crossing the metatarsal-phalangeal joint are much
smaller than the muscles crossing the ankle joint and therefore were estimated to work
2.5 to 3 times harder.

Using a pressure platform, Albers et al. (1992) compared plantar pressures during
walking barefoot and in pointe shoes, peak pressures of relevé and eleve', peak pressures
of walking in the pointe shoe and the two ways of rising onto pointe. These researchers
showed that plantar pressure was higher while walking in pointe shoes compared to
walking barefoot. This higher pressure was due to the constrictive nature of the pointe
shoe which physically limits the plantar surface permitted to contact the ground Not
surprisingly, these researchers also found that peak pressure en pointe was greater than
peak pressure while walking in the pointe shoe. Greater peak pressure en pointe was to
be expected given that the surface area in contact with the ground while en pointe was
greatly reduced when compared to the plantar surface contacting the ground while
walking, for the same relative body weight. The researchers found no difference in the
plantar pressures during relevé and elevé.

Kinematic and Kinetic Studies of Aerial Movement in Dance

Of all the movements in dance, aerial movement has received the most attention
and is of interest because of the high ground reaction forces upon landing due to the
effect of gravitational acceleration Simple dance jumps were studied by Clarkson,

Kennedy and Flanagan (1984) as well as McNitt-Gray, Koff and Hall (1992). These

11

researchers compared the effect of training on the performance of saute'. Clarkson et al.
(1984) compared how dance students and non-dancers performed saute with the legs
externally rotated from the hip and heels together in first position. Measuring knee angle
with an electrogoniometer, these researchers found that the dance students used greater
knee flexion than non-dancers in the demi-pliés that preceded and followed the jump.
Furthermore, they found that those with dance training used a similar degree of knee
flexion to perform relevé. Non-dancers, on the other hand, varied their degree of knee
flexion with the different requirements of each movement. The authors attributed this
finding to a learned motor pattern that sets the depth of the demi-plié. A reasonable
explanation given was that demi-plié is learned on the first day of ballet class and is used
thereafter in the performance of a variety of movements.

McNitt-Gray et al. (1992) used two-dimensional videographic and force plate
analysis to compare dancers, dance students, and non-dancers in the performance of
sauté with the feet in first and parallel positions. Because kinematic data for first
position were deemed unacceptable, only data from the parallel condition were
examined Dancers and dance students used more knee flexion upon landing in demi-
plié than non-dancers, in agreement with the findings of Clarkson et al. (1984); however,
no difference in minimum ankle dorsi-flexion between groups was found Dancers took
longer than non-dancers to reach maximum knee flexion and maximum dorsi-flexion
upon landing from the jump. Vertical ground reaction force peak magnitudes at impact
ranged from 2.81 to 3.82 times body weight and no significant differences were found
between subject groups or foot positions. These peak values were similar to those of

Becker (1984) who found impact peaks of 3.36 and 3.51 for male and female dancers,

12

respectively. While not statistically significant, dancers tended to delay time to peak
vertical reaction force; and dancers, as well as dance students, tended to have longer
landing phase times and took longer to reduce the center of gravity’s vertical velocity to
are than non-dancers. Dancers are trained to land from jumps softly, with a great deal
of control and Becker (1984) suggested that adaptations learned as part of dance training
allow impact forces to be attenuated upon landing from a jump.

Ballet dancers are easily identified by their erect posture and regal carriage of the
upper body. “Pulling-up” the abdominal muscles, or pull-up, as it is referred to in dance
vernacular, is a concept used in ballet training which allows the dancer to achieve the
correct postural alignment. F u et al. (1994) investigated the effect of pull-up on the
performance of saute using two-dimensional motion analysis and force plate data. To
compare the pull-up and non-pull-up conditions, six female advanced level students
performed saute in first position five consecutive times under both conditions. The
dancers were instructed to jump maximally. Upon analysis of the data, two factors were
reported prior to initiation of the first jump: 1) the seventh cervical vertebrae (C7) was
higher when the subject was pulled-up and 2) the height of the second sacral vertebrae
(82) showed no difference between the two conditions. In the air, the vertical
displacements of C7 and 82 were higher with pull-up than without. The dancers jmnped
higher with pull-up and suspension time increased under the pulled-up condition.
Because the subjects were advanced dancers, they were probably more familiar with the
pull-up condition and may have felt more secure executing a maximal jump in that
condition. Additionally, the pull-up jumps were performed after the non-pull-up jumps

and practice effect may have influenced the height of the jrnnps as the dancers became

13

warm and they became more comfortable in the test space. F u et al. (1994) found no
significant differences in the landing force or maximal ground reaction force between the
two conditions, although the dancers felt they landed more softly while jumping with
pull-up. Again, this finding may have had to do with their familiarity with the pull-up
condition.

A series of studies by Simpson and colleagues (1996, 1997, 1997) utilized two-
dimensional inverse dynamics analysis to evaluate the effect of jump distance on the
forces acting at the knee and ankle upon landing from a simple dance leap (Simpson,
Jameson & Odunr, 1996; Simpson & Kanter, 1997; Simpson & Pettit, 1997). The
protocol was the same for all three studies and the data for the same six dancers were
used in the analyses. The leap began from a stationary position, the subject extended a
leg forward, and leapt onto it. With the use of a metronome, flight time was held
constant as the dancers leapt horizontal distances equal to 30, 60, and 90 percent of their
predetermined maximum leap. The magnitude of peak anterior/posterior ground reaction
force increased as leap distance increased Maximal peak vertical grermd reaction force
was nearly three times body weight upon landing from the longest leap, similar to the
impact peak magnitudes measured by McNitt-Gray et al. (1992) in the performance of
sauté. The difference between a sauté and a leap is that with a saute', the dancer lands on
two feet so both legs share the forces of landing, while with a leap, the dancer lands on
one foot and one leg must accept and attenuate the forces.

In their investigation of occurrences at the knee during a simple leap, Simpson et
al. ( 1996) analyzed maximum magnitudes and derivatives for quadriceps force,

patellofemoral compressive force, and patellofemoral contact pressure. With increased

14

leap distance, quadriceps force, peak compressive force, and maximum patellofemoral
pressure increased At landing, knee angular displacement increased as leap distance
increased, allowing the distribution of force through a greater number of patellofemoral
contact regions. Additionally, at longer leap distances, peak knee flexion velocity also
increased, allowing pressure to shift rapidly from one contact region to another. Thus, no
one patellar or femoral region was subjected to high loads for an extensive period of
time. As leap distance increased, the rates of loading compressive force and
patellofemoral pressure increased and the corresponding time to peak values decreased.
The rate of force application is important because high rates of loading are known to
contribute to tissue damage (Nordin & Frankel, 1989).

The effect of distance on axial (Simpson and Kanter, 1997) and shear (Simpson
and Pettit, 1997) forces upon landing from a simplified dance leap also have been
investigated. Joint reaction forces and muscle axial forces of the knee and ankle,
quadriceps shear force, and ankle shear force increased at longer leap distances. Rates of
loading the shear force at the ankle and the axial forces at the knee and ankle increased at
longer leap distances. However, the effect of leap distance on the rate of shear loading at
the knee was not consistent among the subjects. Muscle axial forces demonstrated a
greater influence on the magnitude and rate of applying compressive forces at both
joints, than did joint axial reaction forces. Maximum quadriceps shear force was greater
than maximum knee joint reaction force for all participants. At longer distances,
quadriceps shear force increased; however, the influence of increased quadriceps shear
force on knee shear force was inconsistent since knee shear force increased for only half

of the participants. Values for gastrocnemius and triceps surae shear force and knee

15

shear force were inconsistent among the sample. Greater leap distances did not result in
greater peak values for gastrocnemius and triceps surae shear force and knee shear force
for all participants.

The authors failed to mention in any of the articles whether the dancers were
asked to maintain knee flexion in the sagittal plane upon landing or if they were allowed
to externally rotate their leg as they may be asked to do in a dance class. If the dancers
were landing with their leg externally rotated, perspective one error resulting from the
use of two-dimensional analysis would have influenced the kinematic data used to derive
various muscle and joint force magnitudes.

Although these studies provide a good foundation for comparison with other
aerial movements in dance, the leap studied by Simpson and her colleagues (1996, 1997,
1997) cannot be generalized to all dance leaps. Rarely does a choreographer require a
dancer to perform one movement in isolation, especially an airborne traveling
movement, like a leap. Usually there are steps that lead into traveling aerial movement
allowing the dancer to gain horizontal momentum, some of which is converted to vertical
momentum to launch the body off the ground. Likewise, steps follow traveling aerial
movement as the body continues to move after landing.

Additionally, it is not possible to generalize the findings of the studies by
Simpson and colleagues (1996, 1997, 1997) to all dancers. The amount of dance training
for the participants in these studies was not specified in any of the articles and
inconsistencies existed in the brief descriptions provided about the length of time the
participants had been dancing. In the earliest article, examining patellofemoral forces

during dance landings, the dancers had a minimum of one year of training at the

16

University of Georgia (Simpson et al., 1996). The participants of the other two articles
were identified as six skilled modern dancers (Simpson & Kanter, 1997; Simpson &
Pettit, 1997). It was not made clear how much dance training, if any, the participants had
prior to their training at the University of Georgia It is possible that some of the subjects
may have begun their dance training at the collegiate level and, therefore, may have had
one year of training It is doubtful that one year of training in any activity qualifies one
as “skilled”.

The amount of training is important for two reasons. First of all, those dancers
with more training may be expected to jump further than those with less training This
longer jump distance, in turn, would influence the forces acting on the knee and ankle
due to the increased horizontal force necessary to complete the jump as specified
Secondly, dance training has been shown to change the way forces are attenuated upon
landing from jumps (McNitt-Gray et al., 1992). The preferred aesthetic in dance is to
land toe-heel, with training this movement becomes second nature to the dancer.
McNitt-Gray et al. found that the landing mechanics of professional dancers and dance
students differed from the landing mechanics of non-dancers: those with dance training
used greater maximum knee flexion and took longer to reach maximum knee flexion and
maximum dorsi-flexion than non-dancers. These differences in how the forces were
attenuated upon landing would affect the forces acting at the knee and ankle. Individuals
with less training may not have been able to use this landing technique to its full

advantage.

l7

J Qint Moments

A diffrcult variation of saute, entrechat six, was used by Ravn et al. (1999) in a
study of jump take-off strategy. Entrechat six was filmed as the dancers faced the
camera because external rotation of the hips was required to properly perform the jump.
The vertical height of the jumps, performed by three male dancers from the Royal Danish
Ballet, ranged from 0.27 m to 0.33 m. It was determined that the dancers used a
simultaneous strategy to perform this jump and that the knee and ankle were the
dominant joints. A jump was considered simultaneous if: 1) the two dominant joints
began extension at the same time and 2) the net joint moment of the two dominant joints
peaked at the same time. Ravn et al. referred to flexion-extension moments; however,
the terminology is incorrect because movements performed in the frontal plane are
referred to as abduction-adduction; hence abduction-adduction moments. Correct
movement terms are indicated in parentheses for the Ravn et al. findings. Throughout
the extension phase of the jump, beginning with the lowest point of the center of gravity
and ending at toe off, the knee and ankle joints demonstrated extension (adduction)
moments, while the hip joint showed a flexion (abduction) moment for all dancers. Ravn
et al. ( 1999) could not explain the flexion (abduction) moment at the hip and suggested
that extension (adduction) of the hip in the frontal plane occurred because of extension at
the knee and ankle joints.

Simpson and Kanter (1997), reported that during the first 50-100 ms of landing
from a simple dance jump, some subjects demonstrated a dorsi-flexion moment and a
knee flexion moment. For the remainder of landing, extensor moments were produced at

the knee and ankle of all subjects. Furthermore, the magnitude of peak extensor

18

moments at the knee and ankle joints increased with longer jump distances as the
corresponding time to peak decreased.
Explanation of Grand J ete’

In dance, a leap is defined as an aerial movement that takes off from one foot and
lands on the other (Hammond, 1974). Leaps may be large or small, traveling or
stationary. A frequently performed traveling leap in classical ballet technique is called
grand jeté pas de chat by the Russian ballet school. Figure 1 shows the movement as it is
described in Labanotation, a system of recording dance movement devised by Rudolf
Laban (Miles, 1976). Outside the realm of classical ballet, grand jeté pas de chat may be
called a split or stag leap; however, it will be referred to simply as grand jeté for the
purposes of this review. This movement may be performed by men or women and is not
the sole property of classical ballet. Grand jeté is often performed as choreography
within modern dance, jazz dance, and gymnastics floor and beam exercises. While the
arm and head positions may differ from classical ballet, the actions of the legs are the

531116.

 

 

 

Figure l. Labanotation of grand jeté pas de chat from @'1 Gm gh'ctigrm of classical
ballet in Labanotation (p. 72), by A. Miles, 1976, New York: Dance Notation Bureau.

Copyright 1976 by Dance Notation Bureau, Inc.. Reprinted with permission.

l9

As the name implies, grand jeté is a large leap. Preparatory steps, such as a run or
glissade (a gliding step) must precede grand jete' (Grant, 1982, Vaganova, 1969) and
serve to produce momentum to launch the body into the air (Vaganova, 1969). As the
preparatory movement is completed, the leg that will be positioned anteriorly, pushes off
the ground and is raised to 90 degrees of hip flexion with the knee flexed. This front leg
is extended quickly at the knee as the back leg pushes against the ground to send the
dancer into the air. In dance training, the performer is continually reminded to push off
the ground with the leg that will be anterior in grand jeté. The timing of this initial push
is critical to successful performance of this leap. The dancer is expected to jump high
and cover a great deal of distance. Ideally, at peak height of grand jeté the legs are fully
extended at the knee anteriorly and posteriorly to form a 180 degree split.

Landing is made on the front leg as the back leg continues moving forward into
the next choreographed movement (Grant, 1982). Upon landing from any aerial
movement, the landing should be soft and silent. According to Vaganova (1969), the toe
should touch the floor first, the heel is lowered, and then the knee bends into demi-plie'.
Because movement in dance is continuous, demi-plié upon landing is not only the end of
the jrrrnp, but is also preparation for the following movement (Hammond, 1974).
Bigmeghanigj Analfiis of m Jete’

Various biomechanical aspects of grand jeté have been explored. Kinematic
analysis of take-off and flight phases was done by Ryman (1976) and Murgia (1995),
who also measured ground reaction forces of take-off. The effect of body configuration
on hang time and the illusion of suspension at the height of the leap was investigated by

Midgett et al. (1993) and Rasmussen and Hay (1993). The forces at landing were

20

measured by Becker (1984) as well as Miller et al. (1990). Becker (1984) examined
kinetic and kinematic parameters of landing from grand jeté while Miller et al. (1990)
used grand jeté to compare the effect of modified ballet shoes on the force and pressure
distribution on the foot upon landing.

Ryman (1976) and Murgia (1995) biomechanically examined take—off and flight
of grand jete'. With one highly regarded professional ballerina as her subject, Ryman
(1976) used one camera for a two-dimensional analysis to compare the ballet definition
with the actual performance of six aerial movements, one of which was grand jeté.
Additionally, Ryman examined velocity and acceleration of the center of gravity during
the push off and time of flight of these movements. Descriptive (kinematic) analysis
revealed that grand jeté was performed as it is described in the ballet literature, with the
exceptionthatthedancerwasunabletodefygravityandhangintheairattheheightof
the aerial movements. Ryman observed that as body position changed at the peak of
each aerial movement, the top of the dancer’s head maintained its vertical height Thus,
Ryman reached the same conclusion as two more recent studies (Midgett et al., 1993;
Rasmussen & Hay, 1993), the ability to hang in the air was an illusion determined by the
configuration of the body throughout the movement. Kinematic analysis revealed that of
the six movements studied, grand jeté turd the highest vertical velocity at take-off.
Furthermore, when ranked with the other five aerial movements, grand jeté had the
second highest horizontal velocity at take-off and the second largest vertical
displacement as measured from the depth of the preparatory demi-plié to height of the
movement It was not surprising that grand jete' had the longest flight time and the

second longest horizontal distance traveled of the movements studied because a

21

projectile’s horizontal displacement, peak height, and flight time are due to gravity, angle
of projection, initial velocity, and height of take off.

Murgia (1995) used 15 participants with various degrees of dance experience to
explore a number of biomechanical variables for three aerial movements commonly
performed in dance. The three movements were grand jeté pas de chat, grand jeté
performed with the front leg fully extended at take-off, and a jazz leap that required the
dancer to change legs in the air and land on the take-off leg. The subjects were allowed a
two step preparation for each movement and were instructed to gain as much height and
distance as possible. Three-dimensional kinematic and kinetic data were collected and
analyzed in this study. No significant differences were found between the aerial
movements for many of the variables examined, including magnitude and angle of
application of the ground reaction force. Both styles of grand jeté were found to have
greater peak angular velocity at the knee than the jazz leap and grand jeté pas de chat
demonstrated longer flight time tlmn the other two jumps. The findings of this study
were reported as results of statistical analyses. Murgia presented the results of ANOVAS,
and when necessary, the results of Scheffé posthoc tests, for each biomechanical variable
examined. The age of the subjects ranged from 12 to 34 years and their dance experience
ranged from four months to 15 years. Neither the age of the participants nor their level
of training was considered in the evaluation of the data. The wide range in ages and skill
levels may have contributed to the lack of significant differences for some of the
variables.

In biomechanics, extreme skill groups commonly are used for comparison and

when comparing low vs. high skills, the levels must be defined Rarely will one see low

22

vs. intermediate or intermediate vs. high skill comparisons due to the fact that very gross
measures are used to try to find subtle differences. Fine measurements are inherently
problematic in biomechanics. Additionally, it should be considered that when trying to
perform statistical analysis on biomechanical data, the journey (i.e. path of information)
is often more revealing than the mean (average) of the path — or any part of the path (V.
D. Ulibarri, KIN 830, Fall, 1998).

Midgett et al. (1993) and Rasmussen and Hay (1993) utilized two-dimensional
video analysis to investigate the effect of body configuration on hang time in the
performance of grand jeté. In an effort to quantify hang time, Midgett et al. (1993)
examined the effect of arm position on the relationship between the vertical
displacements of specific points on the body and the center of mass. For the purposes of
their study, Midgett et al. defined “hang time” as “horizontal movement of the head and
trunk during the peak of a leap, with little or no vertical displacement, for a relatively
greater period of time than the center of mass” (p. 4). Eleven female advanced-level
collegiate dancers performed grand jete' using two different classical ballet arm motions
and a modified arm motion The classical arm motions maintained or raised the arm
position during flight and maintained the arm position upon landing, while the modified
motion had the dancer lower her arms immediately after reaching the peak height of the
leap. The leaps were normalized by calculating the top quarter of the leap, based on the
total vertical displacement of the dancer’s center of mass, and the relative time each
point stayed in this top quarter was determined. AS expected, the center of gravity of the
body followed a parabolic path from take-off to landing regardless of arm movement

during flight. Hang time was found to be longer for the modified motion compared to

23

the standard ballet arm positions. An explanation for a longer hang time with the
modified motion lies within the definition of this motion. Ranney (1988) suggested that
as the arms were dropped, the head was raised relative to the center of mass. For this
study, hang time was defined by head and trunk movement relative to the center of mass.

Rasmussen and Hay (1993) studied hang time and the contribution of all limbs to
the illusion of suspension in the performance of several aerial movements, including
grand jeté pas de chat. The movements were performed by one female advanced level
collegiate dancer. For this investigation, “hang time” was defined as the time the center
of gravity of the head, neck, and trunk stayed at a constant elevation within a determined
margin of error. Hang time of grand jeté pas de chat was 0.16 seconds, the longest of any
of the movements studied. Rasmussen and Hay found that the arms contributed less to
the illusion of hanging in the air than the legs and trunk The higher contribution of the
legs and trunk to the illusion of hanging in the air was not surprising, given that the mass
of these body segments was greater than the mass of the arms. The researchers suggested
that timing of body movements, particularly the legs, was important to insure the illusion
of suspension and that the illusion was more effective when hang time occurred during
the descent rather than during the ascent.

Kinematics and Kinetics of Landing from Grand Jete'

In an attempt to identify parameters of landing that reveal the most information
about impact forces, Becker (1984) investigated kinetic and kinematic factors of the
landing technique characteristically used by dancers. He explored the relationship of two
descent variables to peak impact force upon landing. These two descent variables were

1) descent to maximum knee flexion and 2) descent to full heel compression. F ifty-six

24

female and nine male dancers volunteered for this study, all of them were considered
highly proficient in the performance of aerial movements. The dancers were given a
choice between an approach run or a glissade as preparatory steps to the leap. The
dancerslandedin arabesque supportedontheleg fintherfromthecamerasuchthata
medial view of the landing leg was visible. Given that the dancers were asked to
complete the movement by landing in arabesque, the leap performed was most likely one
of several variations of grand jete’; however, a complete description of the leap chosen
for study was not provided Mean peak vertical ground reaction forces upon landing
were 5.13 and 4.47 times body weight for females and males, respectively. These values
were higher than the peak vertical ground reaction forces of three times body weight
reported by Simpson and her colleagues in the performance of a simple dance leap
(Simpson et al., 1996; Simpson & Kanter, 1997; Simpson & Pettit, 1997). The main
difference between the simple leap studied by Simpson and colleagues and the leap
studied by Becker (1984) was the use of preparatory steps which allow the dancer to gain
horizontal velocity some of which was converted to vertical velocity upon take-off.
Other things being equal, more vertical velocity upon take-off would allow the dancer to
attain greater vertical height, possibly increasing the magnitude of the ground reaction
force upon landing Descent to full heel compression coincided with peak vertical force
and descent to maximum knee flexion represented completion of landing. However,
based on the low positive correlation between the peak force and the two descent
variables, neither descent variable was formd to be a good predictor of the magnitude of
peak landing force.

To determine if pressure across the foot can be more evenly distributed upon

25

landing, one male dancer performed grand jeté with a single step preparation (Miller et
al., 1990). Pressure distribution across the foot and vertical force were examined under
13 conditions: barefoot, wearing a standard ballet slipper, and wearing 11 different
ballet slippers modified with various materials used for orthotics. Pressure plate analysis
was performed for all 13 conditions while force plate analysis was performed for only
four conditions: barefoot, in the unmodified ballet slipper, and two of the modified
shoes. Force plate data were used to plot the dancer’s center of gravity on an outline of
his foot Winter (1990) defines center of gravity as the location of the body’s center of
mass in the vertical direction and Shimba (1984) and Soutas-Little (1987) define center
of pressure as the intersection of the screw axis force system resultant and the horizontal
surface of the force plate. Given that force plate data were used in the analysis, the
authors may have plotted the path of the center of pressure (kinetic) rather than center of
gravity (kinematic) as expressed in the article. Pressure distribution differed under
barefoot and ballet slipper conditions with pressure distributed more evenly with the
ballet slipper. All 11 of the modified shoes demonstrated pressure distribution across the
foot, away from the first and second toes. The modified shoe that most successfully
distributed pressure across the foot was also comfortable for the dancer; an indication
that ballet shoes could provide protection and not interfere with performance of
movement. Peak vertical reaction force under all conditions was determined to be
approximately three times body weight This finding was lower than the findings from
Becker (1984), but consistent with more recent studies of dance leaps (Simpson et al.,
1996; Simpson & Kanter, 1997; Simpson & Pettit, 1997). Upon examination of the

center of gravity outlines, the authors found that the center of gravity was consistently

26

over the first and second metatarsal heads. This may be the only study to date to
examine the effect of shoes on external force and pressure distribution in the
performance of dance movements.

Problem Statement

Study of aerial movements in dance is essential because they are performed
repetitively on a daily basis by both dancers and gymnasts. To the knowledge of this
researcher, there has been no three-dimensional analysis of landing from any dance leap
reported in the literature, nor has any study investigated the influence of pointe shoes on
landing from aerial movements. Because tissue injury can be caused by 1) one
maximum load, 2) repetitive loading, and 3) the rate of loading (Nordin & Frankel,
1989), grand jeté is of particular interest. This leap requires the performer to leap high
into the air and results in vertical ground reaction forces over three times body weight
upon landing (Becker, 1984; Miller et al., 1990). Grande jeté is a common movement,
performed numerous times as part of dance training and over the course of a dance
career.

The purpose of this case study was to examine the effect of ballet footwear on the
three-dimensional sagittal plane kinematics and three-dimensional kinetics of landing
from grand jeté. The phrase “sagittal plane analysis” was used when discussing the
three-dimensional sagittal plane analysis.

The following variables were measured at landing from grand jeté: (1) angular
displacement of the ankle and knee joints, (2) angular velocity of the ankle and knee
joints, (3) angular acceleration of the ankle and knee joints, (4) peak vertical ground

reaction force and rate of loading, and (5) force-time characteristics of ground reaction

27

force components (FX, FY, FZ). Foot conditions examined were pointe shoes, ballet
slipmrs, and barefoot.
Hymtheses

At landing from grand jete':

(1) Angular displacement of the knee and ankle joints will differ with foot
condition

(2) Angular velocity of the knee and ankle joints will differ with foot condition.

(3) Angular acceleration of the knee and ankle joints will differ with foot
condition.

(4) Peak vertical ground reaction forces will not differ with foot condition.

(5) Peak vertical ground reaction force rate of loading will differ with foot
condition.

(6) F orce-time characteristics of ground reaction force components (FX, FY, FZ)

will differ with foot condition.

28

Chapter 3
Methodology
Instrumentation

Data were collected in the Biomechanics Research Station of the Department of
Kinesiology at Michigan State University. Force data were collected with an AMT I force
plate (model OR6-5-2000) and video data were collected with two Panasonic Super VHS
video cameras. Timing lights were used to time match the video data from each camera
and to coordinate the force data with video data. Round markers with reflective tape
were attached to the subject’s skin and shoes with non-allergenic tape and were used to
define the ankle and knee joints in a coordinate space. A still camera was used to
document such details as marker configuration, shoe conditions, calibration set-up for
three-dimensional analysis, and data-collection set-up.

Because the floor of the Research Station is of a different material (wood) than
the force plate (metal), rosin was spread evenly over the entire surface with which the
dancer had contact, including the force plate. Ideally, a section of linoleum flooring or
marley would have been used; however, placing these materials over the force plate
would have dampened the data recorded from the force plate.

Video data were recorded at 60 Hz, then digitized using the Ariel Performance
Analysis System (APAS). The kinematic data was then downloaded from the APAS
system for analysis. Forces were recorded at 1000 Hz and downloaded from the APAS
system for analysis. Rate of loading was calculated from the force data
511—13116:

One female ballet student who attends advanced level ballet classes at a local

29

ballet school volunteered to participate in this study. The original proposal called for the
participation of three to five female professional ballet dancers. No professional dancers
were available to participate, therefore the criteria were changed to allow the
participation of advanced level ballet students. Only one dancer was available at the
time of data collection and the prospects of getting other dancers was not encouraging.
Traditionally, researchers in biomechanics have used a low number of subjects due to the
time intensive nature of data reduction via digitization. Since it is uncommon for male
ballet dancers to wear pointe shoes, no attempt was made to recruit male dancers.

The dancer who participated was 25 years old. She had approximately 15 years
of training in ballet and had danced en pointe for approximately eight years.

The dancer wore a leotard, but not tights. The dancer’s legs were bare from the
greater trochanter to the ankle to prevent any movement of the markers that might have
occurred due to movement of the tights over the leg. Movement due to the tights would
have compounded any movement of the markers due to soft tissue movement.

The dancer was asked to wear her own ballet slippers and pointe shoes. Both
pairs of shoes were prepared as they would have been prepared for rehearsal or
performance. Preparation of the ballet slippers included attachment of elastic which
helped hold the shoes on the dancer’s feet Preparation of the pointe shoes included
attachment of ribbons and elastic which helped hold the shoes on the dancer’s feet.
Pointe shoes were broken-in, as they would have been for use in rehearsal or
performance. Because each dancer has their own preference for the brand and condition
oftheshoestheywearinrehearsalandperformance,thebrandandamountoftimethe

shoes had been worn was not controlled, but was noted at data collection. The brand of

30

ballet slippers worn by the dancer was Sansha. The brand and make of pointe shoes was
Capezio Aerial that had approximately 20 hours of wear.
Calibration Smce

The optimal size of the calibration space was determined by the expected height
of the jump, the position of the force plate, and how far the movement was expected to
travel anteriorly. The space was large enough to contain the movement, yet small
enough that the recorded image was as large as possible.

To determine the area of floor space, a frame was laid on the floor around the
force plate such that each corner formed a right angle. Masking tape was placed under
each of the comers and the comer point was marked. The sides were measured for
accuracy and, to insure that each corner was perpendicular, the diagonals were measured.
The space was deemed acceptable when opposite sides were equal to one another in
length and the diagonal lengths were equal.

Next, the 16-point calibration frame was erected. A calibration stand was placed
outside each corner of the measured space. From the arm extending from the top of each
calibration stand, a string of four ping pong balls, each covered with retro-reflective tape,
serving as space markers, was suspended. The surveyors’ cord was adjusted such that the
surveyor’s plumb bob point at the end of the cord hovered over the comer marking on the
masking tape. The heights of the markers on the first cord were adjusted, using toggles
located beneath each marker, so that the markers were approximately equidistant from
one another. The heights of the markers on one string matched the heights of the
markers on the other cords. The height of each marker from the ground was measured

and recorded The set up of the calibration frame, indicating the height of each marker in

31

centimeters along with the length and width of the calibration space, is shown in Figure

2. Each marker’s X, Y, Z coordinates were entered into APAS to define the calibrated

space. Movement of the markers attached to the dancer’s leg within the calibrated space

then could be accurately located

The cameras were positioned so that the entire frame was visible. One camera

recorded grand jeté from the side and the other camera recorded the movement from a

corner view (Figure 2). Each camera was focused in the following manner. An

individual assisting with data collection stood in the calibration frame and the camera

operator focused on the individual’s watch. Once the watch was in focus, the operator

pulled the field of view out so that all 16 points of the calibration frame were visible.

4 105.0)

3 78.0)
16 105.0) 2 45.0) 12 105.0)
15 78.0) 1 15.0) 11 78.0)
14 45.0) 10 45.0)
13 15.5) 9 15.0)

Len h X-axi : 123.4 m

 

i

 

 

 

Side Camera

8 105.0)
7 78.0)
6 45.0)
5 15.0)

Width jZ-axis): 93.3 cm

Q

Comer Camera

Figure 2. Calibration set-up and camera position for data collection.

32

A fixed point was established by placing a reflective marker on the ground out of
the way of the movement The calibration frame was allowed to settle until no
movement was detected. The calibration frame was videotaped simultaneously by both
cameras and removed from the area.

A pair of synchronized timing lights was used to coordinate digitization of the
two views. One box of timing lights was visible to the side camera; the other, visible to
the comer camera. Both boxes of timing lights were placed well out of the way of the
dancer and did not affect or hinder her ability to perform the movement.

Subjeg Premtion

The dancer was responsible for her own wann-up. Once warmed-up, the dancer
practiced grand jeté pas de chat so that at landing the entire foot was on the force plate.
Practice of the movements allowed the dancer to get a feel for the space and to determine
her starting point The dancer performed grand jeté landing on her right leg, the lateral
side of which was visible to both cameras. The sequence of movements was as follows:
temps levé en arabesque, tombé, glissade, grand jeté (see Appendix A for a description of
these steps). These steps are commonly performed in sequence and were familiar to the
dancer. After landing from the grand jeté, the dancer ran, as if exiting the stage. These
movements were performed to the first sixteen counts of “Grand Allegro,” a selection
from “01 1” Music for Ballet Class by Olga Evreinoff and Lynn Stanford Since this is
a compact disc commonly used by ballet teachers, the music was familiar to the dancer.

Order of foot conditions (barefoot, ballet slipper, and pointe shoe) was
randomized to offset the influence of practice and fatigue. Each of the foot conditions

was written on a separate card and placed into an opaque bag. Prior to filming, the

33

dancer drew the cards out of the bag and the order the cards were drawn determined the

sequence of foot conditions. The order of foot conditions was barefoot, ballet slipper,

and pointe shoe. The dancer rested between each trial and the length of the period of rest

between trials was determined by the dancer, except when markers fell off and had to be

replaced. A trial was recorded and considered good if the entire landing foot were on the

force plate, the grand jete' were done in time to the music, all markers stayed attached to

the dancer, and the dancer felt the grand jeté was representative of her ability. Ideally,

three good trials would have been recorded for each condition. However, due to camera

malfunction on the last recorded good trial, only two pointe shoe trials were suitable for

analysis. Three good trials were recorded for the barefoot and slipper conditions.

Reflective markers were placed on the right leg as shown in Figure 3. Water

soluble, non-allergenic ink was used to place a dot on the skin and shoe of the dancer at

each specified site of marker placement. These dots insured accurate replacement of

markers, when they fell off during data collection The anatomical landmarks for the

markers were as follows:

1.

2.

Greater trochanter

Anterior thigh, in area with minimal muscle movement
Femoral lateral epicondyle

Superior anterior tibia, under the knee (on tibial crest)
Inferior anterior tibia, above the ankle (on tibial crest)
Lateral malleolus

Middle of the top of the rearfoot

Fifth metatarsal-phalangeal joint

34

 

 

Figure 3. Placement of reflective markers.

The points on the leg defined the leg segments in space and from their locations,
displacements, velocities, and accelerations of the knee and ankle joints could be
determined. The thigh segment was defined by markers one and three; the shank
segment, by markers three and six; and the foot segment, by markers six and eight The
knee joint was defined by the thigh and shank segments and the ankle joint was defined
by the shank and foot segments.

Pointe shoes and ballet slippers did not seem to pose a problem for marker
placement on the foot Both types of shoes were tight fitting and identifying the bony
landmarks of the foot was not difficult. Because ballet shoes are designed to conform to
the foot during movement, there should have been little, if any, movement of the markers

on the shoe, over the foot. Prior to data collection, the body weight of the dancer was

35

recorded. The dancer weighed 47.61 kg at the time of data collection. The dancer was
also filmed standing on the force plate to assure proper placement of markers and to
establish a standing file from which relative position was measured. For this standing
file, additional markers were placed on the medial condyle and medial malleolus of the
right leg. These additional markers were necessary to estimate joint centers of the knee
and ankle and would have been required for the calculation of moments about the
respective joints.

Identification of Rigjd Bodies in Sm

Three non—colinear points define a rigid body in three-dimensional space. Points
1, 2, and 3 defmedthe thigh; 3,4, and 5, the shank; and 6, 7, and 8, the foot The APAS
software defines angular displacement relative to the angle formed between the
horizontal and a counter-clockwise rotation to the two-dimensiomrl line representing the
segment. The thigh segment was defined as the line between the points 1 and 3; the
shank segment, points 3 and 6; and the foot segment, points 6 and 8 (Figure 3).

Knee flexion/extension was determined by the angle between the thigh and shank
segments. Knee flexion/extension was calculated by subtracting the complementary
angle between the two segments, as determined by APAS, from 180 degrees. Zero
degrees represented full knee extension and the angle increased as the knee flexed during
landing. Ankle plantar-ldorsi-flexion was determined by the angle between the shank
and foot segments. Ankle plantar-/dorsi-flexion was calculated by subtracting the angle
between the two segments from 180 degrees. Zero degrees was full plantar-flexion and
the angle increased as the ankle dorsi-flexed during landing. Neutral ankle position for

the subject was 60 degrees and it is shown on the ankle angular displacement graphs as a

36

darker line.
9% Analysis

The three good trials were recorded for the barefoot and slipper foot conditions.
However, only two good trials of the pointe shoe condition were available for analysis
due to camera malfunction on the final good trial done by the dancer. Once the trials
were digitized, transformed, and smoothed with a cubic spline filter using APAS, change
in angular displacement of the knee and ankle joints was plotted against time and was
compared and contrasted among foot conditions. The paths of angular velocity and
acceleration across time also were compared and contrasted among foot conditions.
Special attention was paid to any sudden changes of direction that may have occurred,
because sudden changes of direction in displacement, velocity or acceleration may be an
indication of an increased risk of injury.

Kinetically, the peak vertical force and force-time characteristics of ground
reaction force components (FX, FY, FZ) at landing, as well as the rate at which loading
occurred were compared and contrasted among foot conditions. Vertical (FZ) data were
not smoothed; however, FX and FY data were smoothed further due to noise in the
system. The shape of the graphs were maintained Difficulties were encountered in
calibrating the force plate to record accurate force plate moments. Calculation of the
path of the center of pressure and knee and ankle joint flexion/extension moments
required force plate moments for calculation and therefore could not be evaluated for this
study. Noteworthy kinetic events were coordinated with the corresponding kinematic

data by matching the times as recorded by force and video methods, respectively.

37

Chapter 4
Results and Discussion

A trial was considered good if the entire landing foot was on the force plate, the
grand jeté was done in time to the music, all markers stayed attached to the dancer, and
the dancer felt the grand jeté was representative of her ability. The order of foot
conditions was: 1) barefoot, 2) ballet slipper, and 3) pointe shoe. Three good trials were
chosen for the barefoot and slipper foot conditions. Due to camera malfunction on the
last potentially good trial done by the dancer, only two good trials for the pointe shoe
condition were available for analysis. The trials chosen for analysis were Barefoot trials
8, 10, 11 (BF8, BF 10, BF 1 1); Slipper trials 6, 7, 8 (SL6, SL7, SL8); and Pointe Shoe trials
1, 4 (PTl, PT4).

For the purposes of this study, landing was defined as contact with the force plate
fiomtoedowrtthedancer’s firstcontactwiththeforceplate,totoeoff,whenthedancer
lost contact with the force plate. The graphs of landing for kinematic data began with the
first frame of video in which the dancer was in contact with the force plate. Knee
flexion/extension was determined by the angle between the thigh and shank segments as
explained earlier. The thigh segment was defined as the line between the markers at the
greater trochanter and lateral epicondyle (points 1 and 3); the shank segment was defined
by the line between the markers of the lateral epicondyle and the lateral malleolus (points
3 and 6) (Figure 3). Knee flexion/extension was calculated by subtracting the
complementary angle between the two segments, as determined by APAS, from 180
degrees. Therefore, zero degrees was full knee extension and the angle increased as the

knee flexed during landing Ankle plantar-ldorsi-flexion was determined by the angle

38

between the shank and foot segments. The foot segment was defined as the line between
the markers of the lateral malleolus and the fifth metatarsal—phalangeal joint (points 6
and 8). Zero degrees was full plantar—flexion and the angle increased as the ankle dorsi-
flexed during landing Neutral ankle position for the subject was 60 degrees, which is
shown on the ankle angular displacement graphs as a darker line. Knee flexion and ankle
dorsi-flexion were considered positive; knee extension and ankle plantar-flexion,
negative.

Kinetic data were recorded during landing from the dancer’s first recorded
contact with the force plate until the foot left the plate. Relative to the force plate
coordinate system, the dancer performed grand jeté in the positive Y direction.
Medial/lateral reaction force was dependent upon which leg the dancer used in the
landing. In this study, the dancer landed on the foot of her right leg; therefore, medial
movements were considered positive; lateral movements, negative.

Kinematics

Kinematic information describes movement. Linear and angular displacements,
velocities, and accelerations are used in these descriptions, as well as temporal
components. Kinematic information is of interest because sudden changes in direction of
displacement, velocity, or acceleration data may indicate an increased risk for injury at
the time of the change.

FIexion/extension angular displacement Knee flexion/extension angular displacement
for all trials can be seen in Figures 4-6. Ankle dorsi-lplantar-flexion angular

displacement for all trials is depicted in Figures 7-9.

39

Knee FIexlonExbnelon Angular Dbpheement
Barefoot T rials

 

 

 

 

-——-—BF8
——--BF10
------ BF11

 

 

 

 

 

 

 

tO.OOO~~———Aa—#i—~— ~M1mwr
o.mVYYYTTIIrT
ors .- N0 chvc- IDNO
86B8§§9$§§¢28N§888§§
0000000000000 000000
Them

Figure 4. Knee flexion/extension angular displacement for the three barefoot trials.

Knee FlexionlExtension Angus Displacement

 

 

 

 

 

   

 

 

 

 

 

 

SlipperTrids
eonoo - ~“ ~
50.000
40000 —
gamma
20M+
10.000 ~ *7 we
°'°°° a's'a’t's‘t'aia'a'a'ssteamers8
oooooooooo’o’ooooooooo
Trme(s)

Figure 5. Knee flexion/extension angular displacement for the three slipper trials.

40

Knee FlexionlExtension Angular Displacement
Pointe Shoe Trials

 

+ Bill

 

 

 

 

10000 LL-__,._._,LL,._L,LL ——-~F—Ai—~ e~~«~r——-—— —~-—fl ~~ ——m——w~»r—»~—»-——-ee 7«
O.WTTIIYTTTTFIIITTITTT
°~Sr§msmaaearreaam§a
'- ‘-
00000000000000000000
Time(s)

Figure 6. Knee flexion/extension angular displacement for the two pointe shoe trials.

Ankle Plantar-lDorsi—flexion Angular Displacement
Barefoot Trials

120.CX)0 .

 

 

 

 

 

 

 

73::BF8_
_ ----BF10
-vv-8F11
—l
L_N¢‘_‘IPU
4ooaa« _______L. —f “LL, ALL,L_ ——
20000 « «—~»~—*AAA-- _-., -———ALA—w——~_—4—-k#~44 4e
o.mrTTTrTIVTTrTTYTTT fl
ONVFO Nm‘DMONVv-QIDNOJCDM
8588882.:eeeeaaaaaaaa
OOOOOOOOOOOOOOOOOOOO
Trme(s)

Figure 7. Ankle plantar-ldorsi-flexion angular displacement for the three barefoot trials.

41

Ankle PIantar—IDorsi—flexion Angular Displacement

 

 

 

 

 

 

 

 

 

 

 

SlipperTn‘als
120.000~
100.000»
80.000 .
——SL6
gm” . """ SL7
g ----SL8
—Neural
40.000 ~——M~A~w~weA#ee; fi-WAP— m—~
20.000 L
o.m 7 T T I I I 1 T fl T I Y T ‘I T I T I I I
0 Is 1- to N O) n O N 0- O 10 N (‘3
83§8§892§923§fi8§tfi§§8§
0 0 O 0 0 O O 0 O O 0 O O O O O O O O
Trme(s)

Figure 8. Ankle plantar-ldorsi-flexion angular displacement for the three slipper trials.

Ankle PW-JDorsi—flexion Amulet Dispboemerrt

 

 

 

 

 

 

 

 

 

 

 

 

PoirteShoeTrials
120.000“
100.000
80.000
8
" 60.000
8-
0
40.000 -_. ~ ~ < , ~ ~¥~ ~
20.000 — _L_
om T 77 r a r Y r r 1
N V v- N O, ‘D o N '- 10 N a) ('3
§sas§§222322§n§aaa§a
O O O O O O O O O O O O O O O O O O O O
Trme(a)

Figure 9. Ankle plantar-ldorsi-flexion angular displacement for the two pointe shoe
trials.

42

Table 1.

Peak an e of knee fl xion and ankle dorsi-flexion with time at peak.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. Peak knee Time at peak Peak ankle Time at peak

Trial . . dorsr-flexron . .

flexron (deg) flexron (s) dorsr-flexron (5)
(deg)

BF8 52.5 0.136 105.44 0.119
BF10 37.44 0.1 19 94.94 0.119
BFll 55.18 0.17 100.68 0.153

SL6 35.96 0.136 98.34 0.136

SL7 50.01 0.153 1 10.52 0.153

SL8 54.84 0.153 108.48 0.136

PT 1 50.32 0.136 106.53 0.153

PT4 50.31 0.136 105.58 0.153

 

Generally, the pattern and magnitude of knee flexion/extension were similar for

all three conditions. Maximum knee flexion ranged from 35.96 to 55. 18 degrees of

flexion. However, for two trials, BFlO and SL6, the magnitude of maximum knee

flexion was lower than any of the other trials (Table 1). When the two lowest values are

excluded, the range of values for maximum knee flexion becomes 50.01 to 55.18

degrees. The time to maximum knee flexion after toe down, ranged from 0.119 to 0.170

seconds, with no evidence of any difference between foot conditions.

The apparent outliers in knee displacement, BF 10 and SL6, may have occurred

because in these trials, compared to the other trials, the dancer brought her leg further

underneath herself to a more vertical position as she descended from peak height of the

leap. If the dancer brought her leg more underneath herself dming descent, she would

not have been able to take full advantage of knee flexion to dampen the force at landing

43

 

because the momentum of her body weight moving forward would not have given her
time to reach a greater angle of knee flexion The relationship between the markers at
the greater trochanter and lateral malleolus along the X-axis seemed to confirm this
supposition At contact, during trials BF 10 and SL6, there was less difference between
the two markers than during the other trials, indicating that the markers were closer to
being aligned along the vertical Y-axis.

The pattern and magnitude of ankle dorsi-flexion were similar for all three
conditions (Table 1). Of the three foot conditions, the condition that demonstrated, on
average, the least amount of dorsi-flexion was the barefoot condition. Average
maximum dorsi-flexion for the barefoot condition was 100.35 degrees, while the
averages for the slipper and pointe shoe conditions were 105.78 degrees and 106.06
degrees, respectively. While not markedly lower than the other trials, the two lowest
magnitudes for maximum dorsi-flexion occurred during trials BF 10 and SL6, the trials
with the lowest angle of knee flexion. The lower angles of dorsi-flexion seem to support
the speculation that the dancer’s weight was further forward in relation to her leg and the
horizontal momentum of the movement kept her moving forward and did not allow her
time to reach a greater angle of dorsi-flexion to dampen the force of landing. Time to
maximum dorsi-flexion ranged from 0.119 to 0.153 seconds with no apparent difference
between foot conditions.

Throughout landing, knee flexion and dorsi-flexion increased In each trial,
regardless of foot condition, maximum knee flexion and maximum dorsi-flexion were
reached within one frame (0.017 s) of each other, however, the order in which these two

events occurred differed with foot condition. Peak dorsi-flexion during the barefoot and

44

ballet slipper conditions, occurred before or at the same time as peak knee flexion.
During the two pointe shoe trials, however, peak dorsi-flexion occurred one frame after
peak knee flexion. The angle of knee flexion at foot contact was higher for the pointe
shoe condition than for most of the other trials; the exception being BF8. A higher knee
flexion angle at contact would have given the knee joint a head start reaching peak
angular displacement compared to the ankle joint. The difference in timing under the
pointe shoe condition may also be due to the relatively heavy leather shank of the shoes
and/or the elastic and ribbons crossed at the dancer’s ankle that held the shoe on her foot
during movement. In particular, the shank of the shoe, and also the elastic and ribbons
added resistance that the dancer had to overcome during landing.

Flexionlextensign angular velocity. Angular velocity is the rate at which angular
displacement occurs. Knee flexion/extension angular velocity for all eight trials can be
seen in Figures 10-12 while ankle dorsi-/plantar-flexion angular velocity for all trials is

depicted in Figures 13-15.

45

4(1)!!!) ~

311“!)

230W

Deon-alas

0111)

-‘l (DIX!)

-2(X).(X)O

$0030

Knee F IexionlExtension Angula Velocity
Barefoot Trials

 

 

 

 

‘ 1:338 El
--—-BF10
»« L;_;BEL‘_1'

 

 

0.323

 

 

 

 

Figure 10. Knee flexion/extension angular velocity for the three barefoot trials.

o§§§§
§§§§§

Deg recs I s

400.000 4

-200.(X)0

Knee FIexionExbneion Angina Velocity
prer Trids

 

 

 

 

 

0000

0.017

0034 A
05

o 255
I 0.272

 

 

 

 

 

 

 

 

Time (s)

Figure 11. Knee flexion/extension angular velocity for the three slipper trials.

46

Knee Halon/Extension Angular Velocity
Pointe Shoe Trials

 

 

 

Dogma/a
0
§

 

0000
0017i
0034
0051
0088
0085
0102
0119
0136
o
170
0487
0204
0221 .-
025
0.272 "
02891
0306
032%

 

 

 

‘1 m.” V
-200.000
000.000 4

Time (a)

Figure 12. Knee flexion/extension angular velocity for the two pointe shoe trials.

Alida Phil—Dom Nutheloeity

 

 

 

 

 

 

 

 

BarefootTrids
400.000 .
200.000 -_
a ,
a .__m 1
3 0.000 T - a a 1 "I. r - firn-BHO’
N V '- ID 0‘ 0 \ ('0 Q h- v- 0 IO N a, (D (V)
g B588§89$¥29RN88R888 .2511
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
-200.000 ‘ \ 2.
400.000 -,
000000 J
Trme(s)

Figure 13. Ankle plantar-ldorsi-flexion angular velocity for the three barefoot trials.

47

Ankle PIantar—IDorsi—flexion Angular Velocity
Slipper Trials

600.11” ~

 

 

 

 

 

 

200010 *1,
0 \'~.
\ \-
8 \-.,
a a.” I Tri I T—T R2,! I—T—T WTT I r I
g nvv—wrooror "'-.or\ .— «or
o S58888932r.t9§8‘1§u88§
000000066 "9000 coo/’0
S 4
(

 

\__.—-

 

400.000

 

Trme(s)

Figure 14. Ankle plantar-ldorsi-flexion angular velocity for the three slipper trials.

Ankle Plain-IDorsi-flexion Amuhr Velocity
Pointe Shoe Trids

 

 

 

 

 

 

400.000

 

 

 

 

 

 

 

 

 

 

0

§ 0000* ,__ _ _l—Pr1
E - N V " 00 m.“ or (O o is v .— 00 In myth in """PT4
o §sassaeze~2=°gaanaaa§a
400.000 _ - _ 4

Time (s)

Figure 15. Ankle plantar-ldorsi-flexion angular velocity for the two pointe shoe trials.

48

Maximum angular velocity of the knee occurred at, or within three frames (0.051
s) after, toe down for all trials (Table 2). Peak knee velocity for all trials, ranged from
222.27 to 328.98 deg/s with no apparent difference between shoe conditions. The trials
with the highest and lowest magnitudes of peak knee velocity were BFll and BF 10,
respectively. The two lowest values for peak knee velocity were recorded for BF10 and
SL6, the trials with the least amount of knee flexion displacement. Knee flexion velocity
for these two outlying trials was lower than the remaining trials because the knee covered
less angular displacement in a similar amount of time. The rate of knee flexion
decreased steadily until reaching zero velocity to coincide with maximum knee flexion.
Minimum angular velocity of the knee occurred 10 to 12 frames (0.170 to 0.204 s) after
peak angular velocity and ranged from -27.99 to -194.82 deg/s. This range included the

two highest values for minimum angular velocity, those for BF10 and SL6.

Table 2

Maximum and minimum knee and ankle angglar velocities with time at occurrence.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Trial Knee Time from toe Ankle Time from toe
‘ Max/mm (deg/s) down(s) Max/mm (deg/s) down (5)
BF8 315.15 / -175.61 0.017 / 0.204 531.95 /-38l.27 0.017 / 0.204
BF 10 222.27/ -27.99 0.00 / 0.153 471.92 / -283.60 0.017/0.187
BFll 328.98 / -108.52 0.051 /0.238 461.31 /-310.56 0.051 /0.238
SL6 240.97 / -40.50 0.00 / 0.153 529.88 / -248.29 0.017 / 0.187
SL7 301.77 / -150.81 0.017 / 0.221 561.71 /-347.43 0.034 / 0.238
SL8 328.09 / -194.82 0.034 / 0.221 552.94 / -377.08 0.034 / 0.221
PTl 306.55 / -124.19 0.017 / 0.204 574.97 / -293.54 0.034 / 0.221
PT 4 308.40 / -66.73 0.00 / 0.187 528.51 / -243.67 0.034 / 0.221

 

49

 

For all trials, maximum ankle velocity occurred within 0.051 seconds (three
frames) after toe down and ranged from 461.31 to 574.97 deg/s with the barefoot
condition showing a lower average of dorsi-flexion velocity compared to the other
conditions. The two trials with the lowest magnitude of dorsi-flexion velocity occurred
under the barefoot condition. These two trials also had two of the three lowest
magnitudes of dorsi-flexion angular displacement The velocity of ankle dorsi-flexion
was lower for those trials with less angular displacement because the ankle did not have
to cover as angular displacement in a similar amount of time. After reaching maximum
ankle velocity, angular velocity continued to decrease as dorsi-flexion increased.
Minimum angular velocities ranged from ~243.67 to -381.27 deg/s. As was the case with
minimum knee angular velocity, minimum ankle angular velocity occurred 10 to 12
frames (0.170 to 0.204 s) after maximum ankle angular velocity.

Peak knee angular velocity occurred at the same time as peak ankle angular
velocity in three of the trials — BF8, BFl 1, SL8. In the remaining trials, peak knee
angular velocity occurred before peak ankle angular velocity. Because two of the trials
with simultaneous peak knee velocity and peak ankle velocity were in the barefoot
condition, it can be reasoned that the footwear influenced the angular velocity of the
knee and ankle joints during landing Further study may be warranted to examine the
effect of footwear on angular velocity of the knee and ankle joints. The ballet slippers
and pointe shoes may have provided resistance that hindered the dancer’s ability to
manipulate as quickly through her metatarsal joints and delayed peak ankle angular

velocity of the five non-simultaneous trials.

50

Flexion_/extensign gngglar gecelgrgtion. Angular acceleration is the rate of change of

angular velocity. Knee flexion/extension angular acceleration for all of the trials can be
seen in Figures 16-18; ankle dorsi-/plantar-flexion angular acceleration, in Figures 19-21.

The knee angular acceleration pattern demonstrated two peaks. For all trials, the
first peak of knee acceleration occurred within four frames (0.119 5) prior to toe down
and is not shown on the graphs. The pre-landing peak is not of concern while the
individual is in the air but might be of concern as to how it affected the landing and the
movement after landing Further work needs to be done to examine the effect, if any, of
this pre-landing peak of knee acceleration. The values for the first peak of knee angular
acceleration ranged from 1753.39 to 2994.19 deg/s2 (Table 3). The first peak may
indicate the dancer initiated knee flexion in preparation to absorb the impact of landing.
The second peak of knee angular acceleration, ranging in value from 3204.33 to 4050.51
deg/$2, occurred as the dancer continued to move forward after maximum knee flexion.
The second peak is shown on the graphs for all trials with the exception of the graph for
BF8, in which the second peak occurred one frame after toe off. The second peak of
knee acceleration was higher than the first peak for all trials because the dancer was
preparing for toe off which required an increase in knee flexion as the leg came forward.
The pattern of knee angular acceleration and magnitudes of the two peaks in knee
angular acceleration were similar for all three foot conditions.

Knee angular deceleration showed one peak Peak angular deceleration of the
knee occurred between 0.085 and 0.153 seconds after toe down and ranged in magnitude
from -2749.37 to 4369.43 deg/52. There were no apparent differences between foot

conditions in peak angular deceleration of the knee.

51

Knee FlexionlExtension Angula Acceleration
Barefoot Trials

 

 

 

 

 

 

Degrees/02

 

 

 

 

 

 

Time (a)

Figure 16. Knee flexion/extension angular acceleration for the three barefoot trials.

Knee F Iera‘onlExension Anglia Acceleraion
Sl'pper Trids
5000.000 ~

1000.000 C; — / 2’“.

 

 

 

 

a . - ——
\ / r l:-—SLsi
a ,4" i l
g r r , r y r r r fl 1 ‘w‘ SL7

m N v ’ IO N or m
g /n co 0 ,5 In rx no N L—---SL8.
'- "' N ‘, N. ._ ,,_L._ 01—521”. 0! _-.,E. ('2 “ "— ”J

o o" o o o o o o o

." I,
I

 

 

 

 

 

 

Time (s)

Figure 17. Knee flexion/extension angular acceleration for the three slipper trials.

52

Knee F lexionlExtenslon Anglia Acceleration
Pointe Shoe Trials

 

 

 

 

-———Pr1
' ~-----PT4

 

 

 

 

 

 

 

 

 

 

Figure 18. Knee flexion/extension angular acceleration for the two pointe shoe trials.

Ankle Plain-Domination Moth Aooelerfion

 

 

 

 

 

BaetootTrids
6000.000 —
’1,
\\\ ,1 .'
(3 a.” r \ \ 1"“. r r r— . ! r T if, 7 1* f-L r r 1 Y T
[N v-‘- IO N 03 (D 0‘) O ,IN F3<D IO N O) ('3 T—““
~ §5\§\8~§89:1’£tlg/§815Q8RR§%r—BFB'l
§ 4000.000 ci (3' o d (3'41 ci :5 d d or ' o' r: o' d a a a o‘ i-——-BF1O
8 ’1' / iL. .AHBF11
o . , __
4000.000 ’

 

 

 

 

40000000 ‘
Trme(s)

Figure 19. Ankle plantar-/dorsi-flexion angular acceleration for the three barefoot trials.

53

Anlde PIarrta-lDorsl—flexion Angular Acceleration
Slipper Trials

 

 

 

 

Degree: [.2

 

 

 

 

 

Time (s)

Figure 20. Ankle plantar-/dorsi-flexion angular acceleration for the three slipper trials.

Anlde Plan-JDorsi-flexion Aruuar Acceleraion
Pom Shoe Trials

60(0000 ~

 

 

 

Degrees 1:2

 

 

 

 

 

 

40000000 .

Time (s)

Figure 21. Ankle plantar-/dorsi-flexion angular acceleration for the pointe shoe trials.

54

Table 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Maxim minim kne anklean c lerati us with tim t urren e.
. Knee T1me from Ankle Time from
Tnal 2 toe down 2 toe down
(dfig/s ) (5) (deg/5 ) (S)
1" 2994.19 -0.068 Max 1‘ 5423.56 -0.068
BF8 M” 2'“ 3519.66 0.272 294373.54 0.255
Min 4257.78 0.119 Min -8394.87 0.119
181753.39 -0.068 Max 1‘404822 .0051
BF10 M” 2'“I 3330.39 0.221 2'“ 3799.76 0.238
Min -274937 0.085 Min -735004 0.119
M 1‘288838 .0051 Max 1‘ 3640.92 .0034
BFll 3" 2"d 3245.62 0.289 2"d 4422.90 0.289
Min -3777.8 0.153 Min -6694.56 0.153
1" 2406.06 -0.068 1“ 4883.66 0051
SL6 “a" 2"d 3801.48 0.221 Max 2"d 4086.88 0.238
Min -303081 0.085 Min -7101.07 0.102
1“ 2830.06 -0.068 P497181 -0.068
SL7 “a" 2'“ 3204.33 0.289 Max 2'“ 4484.65 0.289
Min -3536.84 0.136 Min -6989.69 0.136
1“ 2926.57 .0051 1“ 4822.65 -0.051
SL8 M” 2*"1 4050.51 0.289 M” 2'" 4997.22 0.289
Min 4369.43 0.136 Min -7063.61 0.136
11 1* 2534.22 -0.068 1“ 5026.11 0.051
PT] 2"d 3520.15 0.272 M3" 2'"1 3857.48 0.289
Min -355907 0.119 Min -696l.93 0.119
Max 1“ 2692.90 -0.068 1*429300 .0051
1714 2ml 3561.76 0.238 M” 2"d 3874.28 0.255
Min -3488.97 0.102 Min -6761.59 0.136

 

 

 

 

 

 

 

 

55

 

As with maximum knee angular acceleration, there were two peaks of ankle
angular acceleration. The first peak, ranging in magnitude from 3640.92 to 5423.56
deg/52, occurred within 0.051 seconds prior to toe down for all trials and is not shown on
thegraphs. Aswiththefirstpeakofkneeaccelemfiomthe firstpeakofankle
acceleration may indicate the dancer initiating dorsi-flexion as preparation for absorbing
the impact of landing The second peak in ankle acceleration occurred as the dancer
moved forward afier completion of landing and ranged from 3799.76 to 4997.22 deg/32.
The second peak of ankle acceleration was comparable to the first peak for all trials. The
magnitude of the second peak of ankle acceleration was similar among the foot
conditions.

Maximum angular deceleration of the ankle occurred between 0. 102 and 0.153
seconds afier toe down and ranged in magnitude from -6694.56 to -8394.87 deg/$2.

There were no apparent differences among the three conditions in peak angular
deceleration.

As indicated by the results for angular acceleration, the knee and ankle joints
acted in a coordinated manner throughout landing Peak knee angular acceleration
occurred within 0.017 seconds of peak ankle acceleration for all trials. Peak knee and
ankle accelerations occurred simultaneously for three trials: BF] 1, SL7, SL8. Likewise,
minimum knee angular acceleration occurred within 0.017 seconds of minimum ankle
acceleration for all trials. Minimum knee and ankle angular accelerations coincided for
five trials, BF8, BF] 1, SL7, SL8, and PT], and minimum ankle acceleration occurred one

frame(0.017 s) after minimum knee acceleration for the remaining trials.

56

Elm

ground Reaction Forces. Ground reaction forces (GRF) are the forces that act on the
body in reaction to the forces exerted by the body on the ground In this study, GRF
acted on the dancer as she landed on her right foot. GRF are vectors of equal magnitude
and opposite direction to the force exerted on the ground The three directional
components of GRF are superior/inferior (vertical), anterior/posterior, and medial/lateral,
identified by F2, FY, and FX, respectively, in accordance with the force plate coordinate
system. GRF were measured by the force plate as separate components and recorded by
APAS.

Vertical ground reaction force (Ell. Vertical ground reaction force for all trials can be
seen in Figures 22-24.

With the exception of BF 10 and the two pointe shoe trials, two distinct spikes
were evident on the charts for vertical GRF. The second spike was higher and
represented peak GRF. Becker (1984) associated the first spike with full metatarsal head
contact and peak GRF was associated with firll heel compression. This association
between the two spikes of vertical GRF and foot kinematics could not be confirmed in
the present study. The patterns of vertical force for the two pointe shoe trials did not
show a distinct first peak. The lack of a distinct first peak for the pointe shoe trials may
have been due to the heavy leather shank of the pointe shoes which may have prevented

the dancer from articulating through the forefoot.

57

 

391
7
J

. E '
L.

31.
—-—SL_8

-—-—BF8

 

M 05.0

 

Vertical Ground Reaction Force
Barefoot Trials

H l w... 0 8.0 M

08.0

00nd
00N.0
0R0
08.0
08.0
9N0
00nd
0NN.0
0 30
00nd
0000
00 F0
0h v.0
00 rd
00 _..0
9 F0
on v.0
ON '0
0 w v.0
00rd
0000
000.0
0N0.0
08.0
000.0
03.0
000.0
0No.0
0 5.0

M 88
m 83
. 83
End
83
83
93
83
ammo
93
8nd
. 83
" 00—..0
2.3
. 83
83
93
88
83
o . to
83
83
o8.o
. 29o
08¢
88
Sod
89°
08.0
05.0

 

 

 

 

T1me(s)
Vertical Ground Reaction Force
Slipper Trials

 

 

 

 

 

 

 

 

 

 

 

 

 

08.0

 

Figure 22. Vertical ground reaction force for the three barefoot trials.

Tune (8)
58

Figure 23. Vertical ground reaction force for the three slipper trials.

Vertical Ground Reaction Force

Pointe Shoe Trials

 

_' -21

P
:«- PT

wr-

 

 

 

 

 

 

 

 

6.0m

 

 

05.0

000.0
80.0
05.0
80.0

Time (s)

Figure 24. Vertical ground reaction force for the two pointe shoe trials.

59

The pattern of vertical force for BF 10 showed a distinct first spike, similar to the
other barefoot trials and all of the slipper trials; however, there was not a distinct second
spike. After the first spike for BF10, the magnitude of vertical force increased
curvilinearly rather than a sharp spike and rapid decline. The different pattern of vertical
force for BF 10 was due to the dancer not reaching full heel compression in this trial as
she did in the other trials. Because she did not reach full heel compression, the dancer
bore more vertical force for a longer length of time on her forefoot, than she did in any of
the other trials. Landing on the forefoot without full heel compression could increase the
dancer’s risk of injury, because the forefoot is comprised of bones that are smaller and
lighter, compared to the bones in the rearfoot which are better designed to withstand the
stress of peak vertical GRF.

The first spikes in GRF for the barefoot and slipper trials were similar in
magnitude, ranging from 1.95 to 3.72 BW (Table 4). For those trials with two distinct
spikes in vertical GRF, the first spike occurred between 0.043 and 0.059 seconds prior to
peak vertical GRF. The time interval between the initial spike and peak GRF was similar

among foot conditions.

60

Table 4

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P of undr infor withtim t

V rti l Braking/ Medial/

Trial 53;; Time (s) Acceleration Time (s) lateral Time (5)

(BW) (13W)

BF8 1‘: 3.72 0.013 13: 2.70 0.015 L: 1.15 0.074
2'“: 4.56 0.056 A- 0.78 0.196 M: 0.61 0.056

BF10 1‘: 2.03 0.016 B: 0.99 0.019 1.: 0.39 0.025
2":3.06 0.105 A: 1.36 0.173 M: 0.59 0.072

BF“ 1‘: 1.95 0.017 B: 1.44 0.014 L: 0.52 0.028
2'": 3.38 0.065 A‘ 083 0.198 M: 0.51 0.062

SL6 1“; 2.24 0.015 13: 1.46 0.017 L: 0.28 0.024
2'": 3.74 0.078 A: 1.04 0.178 M: 0.75 0.072

SL7 1“: 2.61 0.015 13: 1.78 0.017 L: 0.36 0.024
2'": 3.53 0.074 A: 0.83 0.199 M- 0.59 0.072

SL8 1“: 2.69 0.014 B: 2.15 0.016 1.: 0.60 0.030
2": 4.40 0.061 A: 0.88 0.207 M: 0.56 0.062

PT] 1*. n/a -0.069 B: 1.70 0.014 L: 1.00 0.025
2": 3.90 A: 0.65 0.207 M: 0.64 0.029

PT4 1‘2n/a 0.059 B: 2.28 0.014 L: 0.60 0.024
2“: 5.12 A: 0.91 0.018 M: 1.27 0.058

 

 

For all trials, peak GRF ranged from 3.06 to 5.12 BW. In comparison, the

magnitudeofpeakGRFforrunningatspeedsof4.5and6.0m/slmsbeenreportedtobe

approximately 2.70 BW (Cavanagh & Lafortune, 1980; Nilsson & Thorstensson, 1989).

There was a trend toward higher GRF when the dancer was wearing pointe shoes.

Average peak GRF was 4.51 BW for the pointe shoe trials, 3.67 BW for the barefoot

t1ials, and 3.89 BW for the slipper trials. It is not clear whether the higher average peak

GRF in the pointe shoe condition was due to the pointe shoes or the influence of practice

61

 

effect since the pointe shoe condition was the last condition recorded. The pointe shoes
may have contributed to the higher average vertical GRF because the relatively heavy
shank did not allow the dancer to attenuate any of the initial force of landing through the
forefoot Upon landing while barefoot and wearing ballet slippers, the dancer was able
to articulate clearly through the forefoot and, as such, absorb some of the force of
landing Practice effect may also lmve contributed to the higher vertical GRF in the
pointe shoe condition. As the dancer became more familiar and comfortable in the
space, her leaps got higher, based on the height of the marker at her greater trochanter,
leading to an increased vertical GRF.

The range of magnitudes of peak vertical GRF for this study was comparable to
the findings of Becker (1984) who found mean peak GRF upon landing of 5.13 and 4.47
BW for females and males, respectively. The values for peak GRF for this study and that
of Becker were higher than the peak GRF of approximately three times body weight
reported by other investigators (Miller, Paulos, Parker & Fishell, 1990; Simpson &
Kanter, 1997). The higher values of GRF for this investigation and that of Becker (1984)
may be explained by the use ofpreparatory steps. The dancer in the current study
performed a series of preparatory movements leading to grand jeté. The entire sequence
of movements was as follows: temps levé en arabesque, tombé, glissade, grand jeté.
Preparatory steps allowed the dancer to gain horizontal velocity, some of which was
converted to vertical velocity upon take-off. More vertical velocity upon take-ofl‘ would
allow the dancer to jump higher, possibly increasing the magnitude of vertical GRF upon
landing The subjects who participated in the study done by Simpson and Kanter (1997)

were not given any preparatory steps, while the participant in the study by Miller et al.

62

(1990) was given a one step preparation.
WW Interpretation of FY GRF is
dependent upon the relationship between the coordinate system of the force plate and the
direction the subject is moving, as well as the direction of the force applied to the force
plate. If the person is moving along the Y axis, when their foot hits the plate, the ground
reaction force acts as a braking force. When the person pushes against the plate to propel
themself forward, the ground reaction force acts as an acceleration force. Since the
dancer performed grand jeté moving in the positive Y direction, braking GRF force is
positive, while acceleration GRF force is negative. On the FY graphs, crossover is shown
as the point where the X axis is crossed Ifthe areas under the braking and acceleration
curves are equal, the person is moving across the force plate with uniform velocity.
Figure 25 shows braking/accelerating GRF for a representative trial, PT4. Peak values
for braking and acceleration GRF for all trials are shown in Table 4.

Peak braking GRF ranged from 0.99 to 2.70 BW. Peak braking force of landing
from grand jeté was two to six times higher than the peak braking GRF of 0.45 BW for
running at 4.5 m/s (Cavanagh & Lafortune, 1980). The magnitudes of peak braking GRF
reported for grand jeté in this study were higher than those reported for a dance leap by
Simpson and Kanter (1997). Those investigators reported maximum braking GRF
magnitudes ranging from 0.21 to 0.98 BW. As with vertical GRF, the preparatory steps
performed by the dancer in the current study may have contributed to higher magnitudes
of peak braking GRF upon landing. The horizontal velocity attained by the dancer as a
result of the preparatory steps would result in higher braking forces at landing compared

to a leap performed from a standing position as in the study by Simpson and Kanter.

63

Bracing/Acceleration Gromd Reaction Force
Pointe Shoe Trial 4

3.000, -

2.500 - . - - . . - 2-, m--.

 

3: ____-..
i ‘—

g!
at

m\

8‘;

30

u.’
E 1"
3 0000
0 00°F
3 «1910".
< 0000

 

 

 

 

Figure 25. Braking/acceleration GRF for trial PT4.

The two lowest values for peak braking GRF, 0.99 and 1.44 BW, occurred in the
barefoot condition. It should be noted that the barefoot condition also recorded the
greatest magnitude of peak braking GRF. The wide range in magnitudes of peak braking
GRF in the barefoot condition may have been the result of the dancer’s lack of
familiarity with performing grand jeté barefoot Dancers who normally wear shoes while
dancing may be uncomfortable dancing without footwear.

Peak acceleration GRF ranged in magnitude from 0.83 to 1.36 BW. In
comparison, the magnitude of peak acceleration GRF for running at speeds of 4.5 and 6.0
m/s has been reported to be approximately 0.50 BW (Cavanagh & Lafortune, 1980;
Nilsson & Thorstensson, 1989). In gait analysis, the size of the area under the curve on
the FY graph indicates the velocity at which the person carries the force over the foot.
Depending on which area is greater, the person is decelerating (seen as positive in this

study), accelerating (seen as negative) or, if the areas are equal, moving at a constant

64

velocity during foot contact. Upon completion of grand jeté, the dancer ran as if exiting
the stage. Had she been asked to perform a series of grand jetés, the magnitude of
acceleration GRF would probably be expected to increase.

Average peak braking GRF was somewhat higher for the pointe shoe condition
than for the other conditions. The average peak magnitude of braking GRF for barefoot,
slipper, and pointe shoe conditions was 1.71, 1.80, and 1.99 BW, respectively. As
previously stated, the dancer leaped higher on the grand jetés performed for the pointe
shoe condition, based on the height of the marker at her greater trochanter. A more
vertical leap would have lead to higher magnitudes for braking GRF because the dancer
would have to control of the effect of increased gravitational acceleration upon landing.
Average peak acceleration GRF was somewhat lower for the pointe shoe condition. The
average peak acceleration GRF was 0.99, 0.91, and 0.78 BW for the barefoot, slipper,
and pointe shoe conditions, respectively. Peak acceleration GRF may have been less for
the pointe shoe condition because the dancer had to overcome the resistance of the
relatively heavy shank of the pointe shoe as well as the resistance added by the ribbons
and elastic that were attached to the shoe.

Peak braking GRF occurred before peak acceleration GRF for all trials. With the
exception of PT 4, peak acceleration GRF occurred at the end of landing For all trials,
crossover fiom braking to acceleration GRF occurred between 0.063 and 0.127 seconds
after contact The two trials with the earliest time at crossover to acceleration GRF were
BF 10 and SL6, whose crossovers occurred at 0.063 and 0.069 seconds, respectively.
These trials were the two trials with the lowest magnitudes of knee flexion and ankle

dorsi-flexion compared to the other trials. Earlier crossover time compared to the other

65

trials supports the previous discussion that the dancer brought her leg underneath herself
more during descent fi'om the leap’s peak height With her leg more vertically oriented
beneath her, the dancer’s weight was further forward in relation to her leg at contact than
it was for the six remaining trials. The horizontal momentum of the movement
continued to can-y her weight forward as she landed and she spent less time braking.
Medial/lateral may ream'on fgrce (EX), Medial/lateral force for a representative
trial, PT4, is shown in Figures 26. On the graph for medial/lateral GRF, lateral GRF is
graphed as positive, while the medial GRF is negative. Peak values for lateral and
medial GRF for all trials are shown in Table 4.

The magnitude of peak medial GRF ranged from 0.51 to 1.27 BW. The
magnitude of peak lateral GRF ranged from 0.28 to 1.15 BW. Consistent with the
literature for running and walking, the magnitudes of FX are relatively small compared
with F2 and FY. Peak to peak amplitude of medial/lateral GRF for individual trials
ranged from 0.95 to 1.87 BW, approximately three to five times higher than the peak to
peak amplitude reported for running with a forefoot strike (Cavanagh & Lafortune, 1980;
Nilsson & Thorstensson, 1989). The average peak to peak amplitude for the pointe shoe
condition was greater than for the other two foot conditions. Average peak to peak
amplitude for the pointe shoe condition was 1.76 BW, compared to 1.26 and 1.05 BW for
the barefoot and slipper conditions, respectively. Since medial/lateral force is an
indicator of stability, the greater average peak to peak amplitude for the pointe shoe
condition may be evidence of the influence of the shank of the pointe shoe on landing.
Theshankofapointeshoecanbeasthickas 1/4 inchandthedancerhastostabilize

herself over this shank since the shank is smaller than the dancer’s foot.

66

Medial/Latera Gromd Reaction Force
Pointe Shoe Trial 4

1.500

’5
I
|
1
1
l
1
l
l
l
l
l
l

 

 

Force (8W)
Model H I Lateral (+)
0

‘8’

. 0.310 i

 

 

Time (s)

Figure 26. Medial/lateral GRF for trial PT4.

Similar patterns of medial/lateral GRF was seen in the graphs for all three
conditions. At toe down, a lateral GRF occurred. The force vacillated medially and
laterally with a general increase in the medial direction as the dancer stabilized herself
over her foot. After peak medial GRF, indicated by a distinct spike for most trials, the
force increased for a short period in the lateral direction before dipping medially, and
i then increasing slightly to hover about zero as the dancer continued to move after the
completion of landing. In seven of the eight t1ials, mak lateral GRF occurred before
peak medial GRF. Furthermore, peak medial force coincided temporally with peak
vertical force for six of the eight trials. Although not confirmed by the present study,
Becker (1984) associated full heel compression with peak vertical GRF. Had peak
vertical force coincided with full heel compression, more force would have been applied

to the medial side of the dancer’s heel compared to the lateral side, for those trials in

67

which peak vertical force coincided with peak medial force.

flte of LAME Vemg' 1 Ground Reaction Force. Rate of loading GRF is the smd at
which force is applied and is of importance because higher rates of loading are known to
cause tissue damage. Tissues are known to be stronger and stiffer under faster loading
than under slower loading (Nordin & Frankel, 1989). Rate of loading peak vertical GRF
was calculated by dividing the peak vertical GRF by the time to peak vertical GRF.

Loading rates for peak vertical GRF are shown in Table 5.

Table 5.

Loadin rat f verti RF

 

 

 

 

 

 

 

 

 

Trial Vertical (BW/s)
BF8 81.4
BF10 29.1
BF] 1 52.0
SL6 47.9
SL7 47.7
SL8 72.1
PT] 56.5
PT4 86.7

 

 

 

 

Rate of loading peak vertical GRF ranged from 29.1 to 86.7 BW/s. These values
were equivalent to 1385 to 4128 kg/s, for the dancer in this study. Rate of loading was
lowest for trial BF10 because peak vertical GRF was lowest and occurred later for this

trial than for any other trial. As previously discussed, the pattern of vertical GRF for

68

BF 10 differed from the other trials since vertical GRF increased curvilinearly, possibly
because the dancer did not reach fill] heel compression during landing. Landing without
full heel compression may have increased the dancer’s risk of injury because the loading
rate of peak vertical GRF was borne by the smaller, lighter bones of the forefoot rather
than the larger, heavier bones of the rearfoot, which are better designed to withstand the
stress of loading When the lowest value for loading rate is excluded, the range for all
trials becomes 47.7 - 86.7 BW/s.

The range of loading rate of vertical GRF for the pointe shoe trials was 56.5-86.7
BW/s, while the ranges for the barefoot and slipper trials were 47.7-72.1 and 29.1-81.4
BW/s, respectively. Loading rate tended to be higher for the pointe shoe condition,
compared to the other two conditions, as the pointe shoe trials had two of the four
highest loading rates. The higher loading rates for the pointe shoe condition may be due
to the resistance of the relatively heavy shank of the pointe shoe. The shank may have
prevented the dancer from articulating through her forefoot which would have delayed
time to peak vertical GRF and allowed her to attenuate the initial force of landing. In
general, there was little difi‘erence in the rate of loading between the barefoot and slipper

conditions.

69

Chapter 5
Summary and Conclusion

The purpose of this case study was to examine the effect of ballet footwear on
landing from grand jete’ using three-dimensional sagittal plane kinematics and three-
dimensional kinetics. For the purposes of this study, landing was defined by the time the
dancer was in contact with the force plate, from toe down to toe off. One advanced level
ballet student performed grand jeté under three foot conditions: barefoot, wearing pointe
shoes, and wearing ballet slippers. Three-dimensional high-speed videographic and force
plate data were used to investigate the following: (1) angular displacement of the ankle
and knee joints, (2) angular velocity of the ankle and knee joints, (3) angular acceleration
of the ankle and knee joints, (4) peak vertical ground reaction force and rate of loading,
and (5) force-time characteristics of ground reaction force components (FX, FY, FZ).

Findings

]. Hypothesis 1 stated: Angular displacement of the knee and ankle joints will differ
with foot condition. This researcher found the pattern and magnitude of angular
displacement of the knee were similar for all foot conditions. The pattern of
ankle dorsi-flexion was also similar for all foot conditions; however, the average
magnitude of peak dorsi-flexion was lowest for the barefoot condition.

2. Hypothesis 2 stated' Angular velocity of the knee and ankle joints will differ with
foot condition. This researcher found the pattern and magnitude of angular
velocity of the knee and ankle joints were similar for all foot conditions.

3. Hypothesis 3 stated: Angular acceleration of the knee and ankle joints will differ

with foot condition This researcher found the pattern and magnitude of angular

7O

acceleration of the knee and ankle joints were similar for all foot conditions.
Hypothesis 4 stated: Peak vertical GRF will not differ with foot condition. This
researcher found the average peak of vertical GRF was higher for the pointe shoe
condition compared to the other two foot conditions.

Hypothesis 5 stated: Peak vertical GRF rate of loading will differ with foot
condition. This researcher found the loading rate of peak vertical GRF tended to
be higher for the pointe shoe condition compared to the other two foot conditions.
Hypothesis 6 stated: Force-time characteristics of GRF components will differ
with foot condition (FX, FY, FZ). This researcher found the force-time pattern of
vertical (FZ) GRF was different for the pointe shoe condition Two distinct peaks
were evident for most of the barefoot and all of the ballet slipper trials; however,
there was no distinct first peak in the force-time pattern of vertical GRF for the
pointe shoe trials. This researcher found the force-time pattern of
braking/acceleration (FY) GRF was similar for the three foot conditions. The
force-time pattern of medial/lateral (FX) GRF was similar for the three foot

conditions.

Conclusion

There were similarities among foot conditions for the following variables:

angular displacement of the knee, velocity and acceleration of the knee and ankle joints,

force-time patterns of braking/acceleration and medial/lateral GRF. Several findings

suggest dance footwear influences landing from ariel movement The effect of footwear

was evident in the difference in force-time pattern of vertical GRF for the pointe shoe

condition. The difference in this pattern indicates that the design of pointe shoes as well

7]

as the attachment of ribbons and elastic may affect landing from ariel movement. The
influence may also be evident in the increased magnitude of all three components of GRF
even though it is not clear whether the increase in magnitude can be fully attributed to
the pointe shoes.

The results of this study have implications for teaching dancers and gymnasts
how to land more safely from aerial movements. The importance of utilizing the entire
foot to attenuate the force of landing was demonstrated by the higher vertical GRF and
higher rate of loading GRF lmder the pointe shoe condition When the dancer was not
able to use her entire foot during landing, vertical GRF and loading rate increased.
Because aerial movements are integral to dance and gymnastics, the high GRF and high
rates of loading that result will take a physical toll on the performer. Learning to land
with correct technique early in training may serve to lengthen the performer’s career.

The results of this study have implications for teaching ballet students how to
dance wearing pointe shoes. While it is important that ballet students learn to dance en
pointe, ballet teachers need to recognize the effect of pointe shoes on common
movements such as jumping Ballet teachers need to be aware that while learning to
dance in pointe shoes, students will be exposed to higher forces and rates of loading than
when barefoot or wearing ballet slippers beeause as the student adapts to doing the
movement in pointe shoes, the student will be less likely to use good landing technique.
The intensity and frequency of performing aerial movements while wearing pointe shoes
should be increased gradually to lessen the likelihood of injury by allowing the student’s
bodytoadapttotheincreasedstress.

As a result of the findings in this study, several suggestions for future research

72

have emerged:

]) Further study needs to be done on the resistance provided by ballet slippers
and pointe shoes that hinders a dancer’s ability to manipulate through the metatarsal
joints and delays peak ankle dorsi-flexion angular velocity. Additionally, the resistance
contributed by the ribbons and elastic which hold the shoes on the dancer’s feet should
be examined.

2) Further study should be done to examine the effect, if any, on the pre-landing
acceleration peak While this peak is not of concern while the individual is in the air and
in an unloaded condition, it might be of concern because of its affect on landing and the
movement afler landing.

3) Further study needs to be done to examine the effect of the pointe shoes’
heavy shank on the magnitude of vertical GRF upon landing While barefoot and
wearing ballet slippers, the dancer was able to articulate clearly through the forefoot and
absorb some of the force of landing. However, the heavy shank of the pointe shoe may
have contributed to the higher average vertical GRF because it did not allow the dancer
to attenuate any of the initial force of landing through the forefoot.

4) Further study should be done to examine the effect of the thickness of the
shank of the pointe shoe as well as the size of the shank in relation to the size of the
dancer’s foot. The shank of a pointe shoe can be as thick as 1/4 inch and the dancer has
to stabilize herself over this shank because the shank is smaller than the dancer’s foot.
The average peak to peak amplitude of medial/lateral GRF was greatest for the pointe
shoe condition and may be evidence of the influence of the shank of the pointe shoe on

landing.

73

5) Further study needs to be done to examine how the rate of loading vertical
GRF is influenced by the idiosyncracies of the pointe shoe. Maximum rate of loading
took place almost immediately upon landing in the pointe shoe condition which may
indicate that the hardness of the box area of the pointe shoe prevented the dancer from
attenuating any of the initial force of landing. The timing of maximum loading rate for
the pointe shoe condition could potentially be injurious to the dancer because the loading
rate would affect the phalanges, a part of the foot not designed for high rates of impact.

6) Further study of landing from aerial movements in dance should also include
kinematic analysis of the hip, knee, and ankle joints in the frontal and transverse planes

as well as examination of the moments at the hip, knee, and ankle joints during landing.

74

APPENDICES

75

Appendix A

Glossary of Terms

Arabesque: a position of the body in which the weight is supported on one leg and the
other leg is extended behind the body (Grant, 1982).

Demi-plié: bend of one or both knees. In ballet technique, tum-out of the legs should
be maintained, the knee should be over the toes, and the entire foot should
remain on the floor. All aerial movements begin and end with demi-plié
(Grant).

Demi-pointe: indicates the dancer is on the balls of the feet, as opposed to en pointe, in
which the dancer is on the toes (Grant).

Elevé: a rise onto pointe or demi-pointe without a preceding demi-plie'. This
movement may be done in all five positions of the feet (Hammond, 1974).

En pointe: indicates the dancer is on the tips of the toes (Grant, 1982).

Entrechat six: a beating jump that begins and ends in fifth position. In the air, the legs
cross and uncross three times before landing with the opposite foot in
front (Grant).

Glissade: a gliding step used as a transition between other steps. There are several
variations of glissade, all of which begin and end in demi-plié (Grant).

Grand jeté: a large leap in which the legs are thrown to 90 degrees and are split in the
air. Grand jeté may be performed forward, sideways, or backwards and is
always preceded by a preparatory movement such as glissade (Grant).

Grand jeté pas de chat: a variation of grand jeté in which the leading leg is thrown

through a bent knee position to full extension anteriorly at

76

the same time as the back leg opens posteriorly to form a

split in the air (Grant).
Positions of the feet : see Figure 27.
Releve': a rise to demi-pointe or pointe preceded by a preparatory demi-plié. This

movement may be done in all five positions of the feet (Hammond, 1974).

Saute': a jump which takes-off of both feet and ends in the same foot position
(Hammond).
Temp levé arabesque: a hop beginning and ending in arabesque (Grant, 1982).

 

lst position 2nd position 3rd position

 

   

4th position 5th position

Figure 27. Five positions of the feet in ballet.

77

Appendix B
Detailed Description of the Pointe Shoe

There are at least 18 pointe shoe manufacturers worldwide and each has a
different method for constructing pointe shoes (Barringer & Schlesinger, 1998). Because
feet have different shapes, strengths, and degrees of flexibility, each manufacturer has
developed a variety of styles. While some pointe shoe manufacturers use automated
machinery to make their shoes, most companies employ many “makers” who construct
the shoes by hand Needless to say, handcrafiing pointe shoes results in a great deal of
variability among shoes, even between shoes in the same pair.

The most important parts of the pointe shoe are the box and the shank which
support the foot while en pointe (Figure 28). The box of the pointe shoe, which extends
over the toes and supports the forefoot, is typically constructed of paper mache or many
layers of fabric and glue. The box has a flattened tip on which the dancer balances while
en pointe. The box tapers slightly on either side of the foot, not only to emulate the
foot’s shape, but also to create an aesthetically pleasing line of extension from the leg to
the foot (Barringer & Schlesinger, 1998). The tapering of the pointe shoe box forces the
great toe into abduction and the lateral toes into adduction, forcing body weight to be
borne on the first and second toes (Sammarco, 1982). Some investigators have attributed
the usually high prevalence of hallux valgus, or bunions, found in ballet dancers to the
combination of the shape of the box and the force distribution en pointe (Sammarco,

1982; Tuckman et al., 199]; Weiker, 1988).

78

 

Figure 28. The pointe shoe with main components identified: (A) the box, (B) the outer
shank, (C) the inner shank. Ribbons and elastic are not attached as they most likely
would be for rehearsal and performance. (Photograph by Miguel Narvaez.)

The shank is comprised of the inner and outer soles. The inner sole is made of
heavy cardboard or a combination of leather and fiber to support the foot while bearing
weight in fill] plantar-flexion. Additional support is attached beneath the inner sole.
This additional support may be one of any number of materials, including wood or steel.
Regardless of the material used, it must be pliable enough to follow the line of the foot in
plantar-flexion, yet strong enough to support the dancer while performing intricate
movements en pointe. The outer sole, on the bottom of the shoe, is leather which
provides some traction as the dancer moves. To accommodate the changes in the foot

that occur with plantar- and dorsi-flexions, the outer sole does not come to the outer edge

of the heel as on a regular street shoe (Barringer & Schlesinger, 1998).

79

Except for the outer sole, the entire shoe is covered with a cotton lining and pink
satin (Barringer & Schlesinger, 1998). It has been erroneously reported in the literature
that the satin provides traction (Milan, 1994; Novella, 1987). In fact, satin is a slick
fabric that some dancers remove from the tip of the shoe to prevent slipping and falling
(Barringer & Schlesinger, 1998). The shoes are held onto the foot with ribbons sewn on
by the dancer. The ribbons are crossed at the front of the ankle, wrapped around the
ankle just above the line of the medial and lateral malleoli, and tied into a knot on the
medial side of the ankle slightly above the medial malleoli. The dancer also may sew a
band of elastic to the back of the shoe to keep the heel of the shoe secured to the foot
during movement. It is of note that right and left shoes are not designated and that the
dancer may choose to alternate her shoes, or assign right and left.

Cunningham, DiStefano, Kirjanov, Levine and Schon (1998) examined the
mechanical properties of five different brands of shoes. Static and fatigue materials
testings were performed using a servohydraulic device under axial and vertical
conditions. Axial conditions simulated the forces applied to the tip on the shoe while en
pointe and vertical conditions simulated forces applied to the plantar aspect of the box
that would occur in activities such as take-offs and landings in running and jumping. All
shoes were found to be stronger and stiffer under static axial conditions, compared to
static vertical conditions. Predominant failure under static conditions in all shoes was
buckling of the pointe shoe box in on itself, similar to that which occurs with wear and
tear in the dance setting. When fatigue testing was conducted in the axial condition, one
brand of pointe shoes demonstrated a fatigue range approximately ]0 times greater than

that of the other four brands. The brand was an innovative shoe whose box and shank

80

were formed from an elastomeric, or rubber-like, material layered with shock-absorbing
foam. The elastomeric material is used in these shoes because it is more durable and
requires less time to break in than conventional pointe shoe materials (Minden, 2000).
While the results of the fatigue testing indicated this brand was more durable, these
results should not be interpreted to mean that these are the best pointe shoes for every
dancer. The brand and style of pointe shoe a dancer chooses to wear is an individual
decision made by taking a number of factors into account including fit, comfort, and

appearance (Barringer & Schlesinger, 1998).

8]

REFERENCES

82

REFERENCES

Albers, D., H11, R, McPoil, T., & Cornwall, MW. (1992). Comparison of foot

plantar pressures during walking and en pointe. Kinesiology and Medicine for Dance,
_]_§(1), 25-32.

Barringer, J ., & Schlesinger, S. (1998). The peinte book: Shoes, training, and
m (Rev. ed). Hightstown, NJ: Princeton.

Becker, T. J. (1984). Kinetic and kinematic parameters of landing impact forces
in the dance jump and leap (Doctoral dissertation, Indiana University, 1984).
Dissertation Abstracts International 45(09), 28668.

 

Cavanagh, P. R & Lafortune, M. A (1980). Ground reaction forces in distance
running. 10ml of Biomghanics, 13, 397-406.

Clarkson, P.M., Kennedy, T.E., & Flanagan, J. (1984). A study of three

movements in classical ballet. Research Qu_a_rterly fer Exereige Ed Spert, 55(2), 175-
179.

Contompasis, J. P. (1986). The classical ballet shoe. Smrts Mflicine, 3(4), 63]-
647.

Cunningham, B. W., DiStefano, A. F., Kirjanov, N. A., Levine, S. E., & Schon, L.
C. (1998). A comparative mechanical analysis of the pointe shoe toe box. American

Journal of Spegte Medicine, 26(4), 555-561.

Dozzi, PA & Winter, DA (1993). Biomechanical analysis of the foot during
rises to fill] pointe: Implications for injuries to the metatarsal-phalangeal joints and shoe

redesign. Kinesiology and Medicine for Dance, 16(1), 1-1 ].

Fu, A, Chan, K M., Chierl, P., Mak, A., Hong, Y. L., Li, C., K., Lam, P.,
Melwani, J ., & Wong, P. (1994). A computer assisted investigation into the effect of
pull-up in ballet sautés - a pilot study. In Technology for Health. Proceeding of BME

6

94, International Conference on Biomedical Engineering 1994, Hong Kong Convention

and Exhibition Centre, Heng Keng (pp. 363-366). Hong Kong: Biomedical Division,
Hong Kong Institution of Engineers.

Galea, V. & Norman, R W. (1985). Bone-on-bone forces at the ankle joint
during a rapid dynamic movement. Biomechanics IX(A), 71-76.

Grant, G. (1982). Technite manual and dictionary of classical ballet (3rd ed).
New York: Dover.

83

Hammond, S. N. (1974). Ballet basics. Palo Alto, CA: Mayfield.

Kravitz, S. R, Fink, K. L., Huber, S., Bohanske, L., & Cicilioni, S. (1985).
Osseous changes in the second ray of classical ballet dancers. Journal ef the Amerign
Podiatric Medical Association, 75(7), 346-348.

Kravitz, S. R, Murgia, C. J ., Huber, S., Fink, K., Shaffer, M., & Varela, L.
(1986). Bunion deformity and the forced generated around the great toe: A
biomechanical approach to analysis of pointe dance, classical ballet In C. G. Shell (Ed),

The 1984 Olmpic scientific conggess proceedings: Vol. 8. The dancer as athlete (pp.
213-225). Champaign, IL: Human Kinetics.

Lawson, J. (1983). Shoes and injuries. Dancing Times, 73(876), 946-947.

McNitt-Gray, J. L., Koff, S. R., & Hall, B. L. ( 1992). The influence of dance
training and foot position on landing mechanics. Medical Problems ef Performing
Artists 7(3), 87-91.

 

Midgett, v. D., O’Bryant, r1 8., Stone, M. n, & Johnson, a L. (1993). Effects

of arm position on hang time during a grand jeté. Kinesiology Ed Medieine for Dance,
1_5_(2), 3-22.

Milan, K. R. ( 1994). Injury in ballet: A review of relevant topics for the physical
therapist. Journal Of Orthemeg'c And Smrts Physical Therapy, 19(2), 121-129.

Miles, A (1976). The Gail Grant dictidnag of classical ballet in Labanetation.
New York: Dance Notation Bureau.

Miller, C. D., Paulos, L. B, Parker, R. D., & Fishell, M. (1990). The ballet
technique shoe: A preliminary study of eleven differently modified ballet technique shoes
using force and pressure plates. Foot and Ankle ] 1(2), 97-100.

 

Minden, E. (2000). Gaynor Minden explained. Available:
http://www.dancer.com/GIVlExplained.html.

Murgia, C. J. (1995). Relationships among selected kinetic and kinematic
parameters of three types of dance leaps (Doctoral dissertation, Temple University,
1995). Dis tion A In 1 1 56(04), 1288A.

Nilsson, J. & Thorstensson, A. (1989). Ground reaction forces at different speeds
of human walking and running. Aega Physiolegiea xfldinavia, 136, 217-227.

Nordin, M & Frankel, V. H (1989). Basic biomechanics of the musculoskeletal
system (2"d ed). Philadelphia: Lea & Febiger.

84

Novella, T. M. (1987). Dancers’ shoes and foot care. In A. J. Ryan & R. E.

Stephens (Eds), Dance medicine: A eomprehenaive guide (pp. 139-176). Chicago:
Pluribus Press.

Ogilvie-Harris, D. J., Carr, M M., & Fleming, P. J. (1995). The foot in ballet
dancers: The importance of second toe length. Fmt ar_ld Ankle Intematienal, 16(3), 144-
147.

Quirk, R. (1994). Common foot and ankle injuries in dance. Orthomdic Clinics
of Nerth Ameriea, 25(1), 123-133.

Ranney, D. (1988). Biomechanics of dance. In P.M. Clarkson & M Skrinar
(Eds), Scieng of elm Qining (pp. 125-144). Champaign, IL: Human Kinetics.

Rasmussen, C. C. & Hay, J. G. (1993). The illusion of suspension in selected
dance leaps. Kinesiology and Medicine for Dance, 15(2), 23-41.

Ravn, S., Voigt, M., Simonsen, E. B., Alkaer, T., Bojsen-Maller, F., & Klausen,
K. (1999). Choice of jumping strategy in two standard jumps, squat and
counterrnovement jump -- effect of training background or inherited preference?

Scandinavian Journal of Medicine and Science in Sperts, 9(4), 201-208.

Ryman, R S. (1976). A kinematic an_d descrim've galysis of selected classical
ballet skills. Unpublished master’s thesis, York University, Toronto, Ontario, Canada.

Sammarco, G. J. (1982). The foot and ankle in classical ballet and modern dance.
In M. H. Jahss (Ed), Disorders of the foot: V91. 2 (pp. 1626-1659). Philadelphia: W. B.
Saunders.

Schneider, H. J., King, T. Y., Bronson, J. L., & Miller, E. H (1974). Stress
injuries and developmental change of lower extremities of ballet dancers. Melegy,
l_1_3, 627-632.

Shimba, T. (1984). An estimation of center of gravity from force platform data.
Journal of Biomechanica, 17(1), 53-60.

Simpson, K J., Jameson, E. G., & Odum, S. (1996). Estimated patellofemoral
compressive forces and contact pressures during dance landings. Journal pf Applied
Biomgmiea, 12(1), 1-14.

Simpson, K. J ., & Kanter, L. (1997). Jump distance of dance landings influencing

internal joint forces: 1. axial forces. Medicine gld Science in Smrts Ed Exercise, 29(7),
916-927.

85

Simpson, K. J ., & Pettit, M. (1997). Jump distance of dance landings influencing
internal joint forces: 11. Shear forces. Medicine and Science in Smrts and Exercise,
22(7), 928-936.

Soutas-Little, R W. (1987). Center of pressure plots for clinical uses. In J. L.
Stein (Ed), Biomechanics of Normal and Prosthetie Gait (pp. 69-75). New York:
American Society of Mechanical Engineers.

Teitz, C. C., Harrington, R. M., & Wiley, H. (1985). Pressures on the foot in
pointe shoes. Fmt and Ankle, 5(5), 216-221.

Torba, R G. and Rice, D. A. (1993). Pressure analysis of the ballet foot while en

pointe. Prggeegngs of the Twelfth gathern Biemedigl Engineering Confereng:
Talane University. New Orleans, Louisiana, April 2-44 1993 (pp. 48-50). Piscataway, NJ:

Institute of Electrical and Electronics Engineers.

Tuckman, A 8., Werner, F. W., & Bayley, J. C. (199]). Analysis of the forefoot
on pointe in the ballet dancer. Fo_qt and Ankle, 12(3), 144-148.

Vaganova, A. (1969). Basic principles of classical ballet: Russian ballet
technique (A. Chujoy, Trans). New York: Dover. (Original work published 1934)

Weiker, G. G. (1988). Dance injuries: The knee, ankle, and foot In PM.
Clarkson & M. Skrinar (Eds), Science of dance training (pp. 147-192). Champaign, IL:
Human Kinetics.

Winter, D. A. (1990). Biomechanics and motor control of human movement (2"d
ed). New York: John Wiley and Sons.

Winter, D. A & Robertson, D. G. E. (1978). Joint torque and energy patterns in
normal gait. Biological gametics, 29, 137-142.

86