3... tr. :....L
t9.“ 9. ...........9.
.9999.

.7‘
.
.5?

. .1.
x ..
.... 99. new

.

r. «.9... mafia

I .
n 20¢".

«'35? 5%.
.. .
fig“?
.99....93
5
{AI

 

a. VP.

9.99

. 'VI.
:5 7..
E... .u r

.9. .

2:“ .
:xv.

, .uu
1 ,fiuwfi .99. 9......

3.. f

!

A:

9
9 . .9 f»;

a

It .9. f
A!!&’.{ .
(.9... :8...

9.11%.... ..
v93 I‘t...‘
.947.‘. \|

9 :7}...
. 5:
by
. .v . ..

.33?

{Lamas . . . u}
2.5.3.... pl..._y.r...... . . . . . . . . . . . .99...
. . 9... 959.939.. .. . . . . . .

: . . . . . . ......u...n...... . . ,

.r 1..

 

tum.

 

THESIS .

2

-r' Log LIBRARY
Mlcblgan State
9 Unlvorsity

 

 

 

 

This is to certify that the

dissertation entitled

POSTURE AND FORCE MEASURES OF MID-SIZED
MEN IN SEATED POSITIONS

presented by
Tamara Reid Bush

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Engineering Mechanics

  

 

Major professor 1

 

9 28.2229

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

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

Li; 4 I" '1
. 7‘

 

 

050109
3%204129999

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 cJCIRC/DateOtanS-p.“

 

_ -9.“-

POSTURE AND FORCE MEASURES OF MID-SIZED MEN IN SEATED
POSITIONS

By

Tamara Reid Bush

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Materials Science and Mechanics

2000

ABSTRACT

POSTURE AND FORCE MEASURES OF MID-SIZED MEN IN SEATED
POSITIONS

By

Tamara Reid Bush

To better understand seating biomechanics, the normal and shear forces,
chair positions, and seated postures of twenty-three mid-sized male subjects
were measured and analyzed. These data were examined in relation to one
another and the affects of chair position on both body posture and support forces
were determined. Force data were collected and analyzed for both static and
dynamic conditions using six multi-axis force plates mounted to an experimental
chair under the thighs, under the buttocks, behind the pelvis, behind the thorax,
and to a steering wheel and foot support. Postural data and chair movement
data were collected using a motion measurement system.

These data were analyzed to determine what degree of motion was
necessary in the experimental chair, in terms of recline and back support
articulation, to produce statistically significant postural and loading changes for
the mid-sized men. Significant differences were found across the subject sample
in both the static and dynamic test trials.

The support force data in conjunction with subject anthropometry were
also used to estimate the internal loading in the lumbar spine and evaluate that

loading for one subject as his posture was varied.

Currently, the data presented for this dissertation are the only data of their
kind. It is expected that the availability of these data will have a significant
impact on seating design, seating evaluation and the tools used for design and

evaluation.

Copyright by
Tamara Reid Bush
2000

Dedication

I would like to dedicate this dissertation to my family including my husband, Neil,
daughter Brittany, parents Richard and Pamela, brother Paul and grandparents
Don, Jean, Helen and Harlan.

My family has given me the determination, the strength, and the support to
continue onward with this challenging, and seemingly endless endeavor.

Acknowledgements

I would like to thank my committee members, Robert Hubbard, Roger
Calantone, Joseph Vorro, Dahsin Liu and Roger Haut for their advice and
diligence in working with me during the development of this dissertation.

I would especially like to thank my major advisor, mentor and friend
Robert Hubbard for an exceptional experience as a doctoral student. You have
assisted me in every way possible and have played a large part in the success of
my career.

To Phaedra Gutowski, I would like to thank you for your friendship and all
of your hard work during the testing of the subjects and tracking of the data.

Cliff Beckett, our technical wizard also deserves special thanks for
keeping the equipment running and the computers virus free.

Lastly, I’d like to thank all of the students who helped with various aspects
of my work from data collection to moral support: Akram Ali, Heidi Bogard, Emily

Daligga, Sarah Gregg, Jocelyn Roy, and Richard Setyabudhy.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ xi
LIST OF FIGURES ............................................................................................ xvi
Introduction ......................................................................................................... 1
Scope of Research .............................................................................................. 5
Data ................................................................................................................... 5
Usage of Data ................................................................................................... 5
therature Survey .............................................................................................. 10
Spinal Column ................................................................................................. 10
Posture ............................................................................................................ 13
Physiological Effects from Postural Changes .................................................. 18
Literature Related to Support Force Data ........................................................ 23
Comfort Related Literature .............................................................................. 29
Testing Methods ............................................................................................... 33
Equipment Description .................................................................................... 33
Lab Chair ..................................................................................................... 33
Force Plates ................................................................................................ 36
Motion Measurement System ...................................................................... 37
Subject Recruitment and UCHRIS Approval (University Committee on
Research Involving Human Subjects) ............................................................. 37
Methods for Testing ......................................................................................... 39
Application of Targets .................................................................................. 39
Test Conditions and Setup .......................................................................... 42
Test Descriptions ............................................................................................. 45

vii

Reference File ............................................................................................. 45

Static Tests .................................................................................................. 45
Preferred Pastures ....................................................................................... 4 7
Dynamic Tests ............................................................................................. 48
Hard Seat .................................................................................................... 49
Analysis Methods ............................................................................................. 51
Motion Data ..................................................................................................... 51
Custom Data Analysis ..................................................................................... 52
Basic Vector Calculations ................................................................................ 52
Posture Measures - Subjects .......................................................................... 56
Measurement of Pelvis Orientation ................................................................. 57
Measurement of Thorax Orientation ................................................................ 59
Openness Angle .............................................................................................. 60
Body Recline ................................................................................................... 61
Head Position .................................................................................................. 63
Chair Position .................................................................................................. 64
Force Data ...................................................................................................... 67
Location of Resultant Forces ........................................................................... 70
Internal Joint Estimation .................................................................................. 72
Calculation of Segment Mass as a Percent of Body Weight (%BW) ............... 76
Calculation of Location of Center of Mass within Each Segment .................... 77
Results, Analysis and Discussions of Static Data ......................................... 89
Test Chair Repeatability .................................................................................. 89
Force Data for Static Tests .............................................................................. 93
Recline Angle Comparison for Primary Forces ............................................ 93

viii

Recline Angle Comparisons for Secondary Forces ................................... 100

Recline Angle Comparison for Lateral Forces ........................................... 107
Chair Articulations: Primary Forces ........................................................... 109
Chair Articulations: Secondary Forces ....................................................... 1 15
Support Forces Calculated as Percent Body Weight ................................. 120
Static Equilibrium ....................................................................................... 124
Estimation of Internal Joint Forces ............................................................. 129
Segment Motions for Static Trials .................................................................. 135
Thorax and Pelvis Angles .......................................................................... 135
Motion of Individual Segments: Thorax ...................................................... 138
Motion of Individual Segments: Pelvis ....................................................... 142
Limb Angles: Elbow and Knee ................................................................... 145
Results, Analysis and Discussion of Dynamic Data .................................... 151
Chair Motions ................................................................................................ 151
Recline Angle ............................................................................................. 151
Torso Articulations ..................................................................................... 153
Force Data for Dynamic Tests ....................................................................... 155
Primary Forces .......................................................................................... 155
Secondary Forces ...................................................................................... 164
Lateral Forces ............................................................................................ 171
Overall Comments on Dynamic Force Data .............................................. 173
Resultant Forces ........................................................................................ 175
Equilibrium ................................................................................................. 180
Torso and Limb Motion for Dynamic Tests .................................................... 185
Thorax ....................................................................................................... 185

Pelvis ......................................................................................................... 186

Chair vs. Body: Openness ........................................................................ 187
Chair vs. Body: Recline ............................................................................. 189
Discussion: .................................................................................................... 189
Conclusions .................................................................................................... 189
I. General Conclusions: ................................................................................. 189

ll. Static Data Conclusions: .......................................................................... 189
Ill. Dynamic Data Conclusions .................................................................... 189
Future Work .................................................................................................. 189
References ...................................................................................................... 189

LIST OF TABLES

Table 1: Conditions in an experimental chair tested by Andersson26 while

monitoring disc pressure and muscle activity. .................................................. 19
Table 2: Comparison of seat orientation based on low disc pressure

(Andersson33 and Nachemson3 4) and minimum discomfort (Grandjean2 ) ......... 30
Table 3: Recommended body angle ranges to avoid discomfort”. .......................... 30
Table 4: Mid-Male Anthropometry ............................................................................ 38
Table 5: Targeted Body Landmarks ......................................................................... 40
Table 6: Chair Targets ............................................................................................. 40
Table 7: Static Test Postures ................................................................................... 46
Table 8: Comfort Postures ....................................................................................... 48
Table 9: Dynamic Test Conditions ........................................................................... 48
Table 10: Percentage distribution of total body weight according to different

segmentation plans (from Webb Associates, 1978)9 ...................................... 77
Table 11 Estimates of mass distributions (kg) for various male and female

percentiles (a). .............................................................................................. 77
Table 12: Determining link lengths of arm”. ............................................................ 78
Table 13: Distances to segment center of mass in centimeters”. ........................... 81

Table 14: Locations of body landmarks from mid- male crash test dummy
specifications50 .................................................................................................. 85

Table 15: Translation coordinates between the dummy points and the target
points. ............................................................................................................... 85

Table 16: Average recline measurement of the chair in degrees for each
condition (n=92). The calibration of the recline angle relative to vertical was
computed by subtracting the offset of the recline bar. ...................................... 90

Table 17: Average angle in degrees measurement of the thorax support and the
pelvis support for each recline angle (negative sign means angle was
rearward of lab vertical relative to a right-handed coordinate system with x
anterior and z superior) ..................................................................................... 91

xi

Table 18: Average angle measurement of the thorax support and the pelvis
support for each recline angle after removal of one outlier in the shaded
boxes. ............................................................................................................... 92

Table 19: Primary forces for recline 1, 2 and 3 for each support plate. (Wheel is
shear load, Fx, while all others are normal loads, Fz) ....................................... 94

Table 20: Testing for a significant statistical difference between the primary
support forces at given recline angles. Normal forces (Fz), except wheel
force, which is (Fx) ............................................................................................ 95

Table 21: Secondary forces in Newtons. All support plates were the shear forces
(Fx) except for the wheel, which was a normal force (Fz) ............................... 102

Table 22: Testing for a significant statistical difference between the secondary
support forces at given recline angles. Shear forces (Fx), except wheel,
(Fz). ................................................................................................................ 104

Table 23: Average lateral forces (Fy) and standard deviations. Each recline is
the average of all test conditions (slumped, neutral, erect and super erect)... 107

Table 24: Testing for a significant statistical difference between the lateral
support forces at given recline angles. Lateral shear forces, Fy. .................... 108

Table 25: Primary forces in newtons for recline 1, 2 and 3 at all support plate
locations. ........................................................................................................ 1 10

Table 26: Comparison of primary support forces between chair articulations.
Slumped (8), Neutral (N), Erect (E), Super Erect (SE). .................................. 112

Table 27: Secondary forces in Nevvtons (Fx) for the thigh, buttocks, pelvis, thorax
and foot and F2 for the wheel. ........................................................................ 116

Table 28: Comparison secondary support forces between chair articulations.
Slumped (8), Neutral (N), Erect (E), Super Erect (SE). .................................. 119

Table 29: Primary forces represented in percent of total body weight (%BW) for
each condition. All forces are normal forces (Fz), except for the wheel which
is a shear force (Fx). ....................................................................................... 121

Table 30: Primary forces represented in percent of total body weight (%BW) for
each condition. All forces are normal forces (F2), except for the wheel, which
is a shear force (Fx). ....................................................................................... 122

Table 31: Horizontal forces summed for recline 1, showing data for all subjects... 126

Table 32: Average horizontal forces summed for recline 1, 2 and 3. ..................... 126

xii

Table 33: Vertical forces summed for Recline 1, showing data for all subjects ...... 127
Table 34: Average horizontal forces summed for Recline 1, 2 and 3. .................... 127

Table 35: Estimations of internal joint loads at the approximate L5/S1 joint center.
Subject 17, recline 1. Forces are converted to orientation of pelvis. Fx is
shear loading (positive anterior) and F2 is compressive loading (positive

supenon. ........................................................................................................ 129
Table 36: Conditions tested by Andersson58 while monitoring disc pressure and

muscle activity. ............................................................................................... 130
Table 37: Bush dissertation data compared to Andersson58 data. ......................... 132
Table 38: Average Openness of body (thorax plus pelvis orientation). ................. 136
Table 39: Average Total Support Angle (TSA) of test chair in degrees .................. 137
Table 40: Significant difference testing for body openness and chair TSA. ........... 138

Table 41: Subjects’ thorax angles with respect to vertical (wrtv) for Recline 1.... 139

Table 42: Average thorax angles for all subjects across all three recline angles
with respect to vertical. Negative angle is rearward of vertical. ..................... 140

Table 43: Average angle change of all subjects from the slumped to the super
erect condition for the thorax and pelvis. ........................................................ 140

Table 44: Testing for significant differences between the thorax angles of the
subjects ........................................................................................................... 140

Table 45: Pelvis angles for each subject at each chair articulation in recline 1.
The pelvis angle is measured with respect the horizontal lab (wrth) with the
negative sign meaning the vector is above horizontal. ................................... 143

Table 46: Average pelvis angles for all subjects across all three recline angles,
with respect to lab horizontal. ......................................................................... 144

Table 47: Testing for significant differences between the subjects’ pelvis angles. 144

Table 48: Elbow Angles for all the subjects at all the chair articulations for recline
1. Note the small deviation within a subject across conditions. ..................... 146

Table 49: Summary of Elbow Angles for Recline 1, 2 and 3. ................................. 147

Table 50: Testing for significant differences of the elbow angle between the
various chair articulations. .............................................................................. 147

xiii

Table 51: Testing for significant differences of the elbow angle between the

recline positions. ............................................................................................. 147
Table 52: Knee angles for all the subjects at all the chair articulations for recline

1. Note the small deviation within a subject across conditions. ..................... 148
Table 53: Summary of knee angles for recline 1, 2 and 3. ..................................... 149

Table 54: Testing for significant differences of the knee angle between the
various chair articulations. .............................................................................. 149

Table 55: Testing for significant differences of the elbow angle between the

recline positions. ............................................................................................. 149
Table 56: Total Recline Angle (TRA) for the chair during both dynamic and static
testing. All angles in degrees. ......................................................................... 152
Table 57: Statistical analysis of Total Recline Angle during static and dynamic
testing. ............................................................................................................ 153
Table 58: Total Support Angle (TSA) for the chair during both dynamic and static
testing. All angles in degrees. ......................................................................... 153
Table 59: Statistical analysis of Total Support Angle during static and dynamic
testing. ............................................................................................................ 155
Table 60: Average primary forces for both the Static and Dynamic Tests
(Newtons). ...................................................................................................... 156
Table 61: TSA where the maximum and minimum average primary forces were
exerted for static tests ..................................................................................... 159
Table 62: Sample of data from which Table 63 data was obtained ........................ 159

Table 63: The most common TSA position of the chair at the maximum and
minimum primary force values for Recline 2 dynamic tests. ........................... 159

Table 64: Statistically significant difference in static tests between slumped and
super erect conditions for the primary forces. ................................................. 160

Table 65: Statistically significant differences of primary forces for the dynamic
tests between the maximum and minimum values. ........................................ 160

Table 66: Secondary forces for both the static and dynamic tests in newtons. ...... 165

Table 67: Statistically significant differences of secondary forces for the static
tests between the slumped and super erect chair positions. .......................... 166

xiv

Table 68: Statistically significant differences of secondary forces for the dynamic
tests between the maximum and minimum values. ........................................ 166

Table 69: Lateral Forces for both the static and dynamic tests in newtons. ........... 171

Table 70: Statistically significant differences of lateral forces for the static tests
between all articulations. (Y) means statically significant differences found,
(-) means no statistical difference found. ........................................................ 172

Table 71: Statistically significant differences of lateral forces for the dynamic tests
between the maximum and minimum values. ................................................. 172

Table 72: Location of the resultant force with respect to the center of the force
plate in the X direction. X direction is positive inferior behind the thorax and
pelvis, and positive anterior under the buttocks and thighs. ........................... 178

Table 73: Location of the resultant force with respect to the center of the force
plate in the Y direction. The Y direction is positive toward the left side of the
body while seated in the chair. ....................................................................... 179

Table 74: Averages and standard deviations of best fit line slope and intercept for
the equilibrium horizontal and vertical force data. ........................................... 184

Table 75: Measure of Thoracic Angle in degrees during dynamic motion from a
slumped position to a super erect position. All values are negative, meaning
the thorax is rearward of laboratory vertical. (Maximum equals the maximum
deviation from vertical.) ................................................................................... 186

Table 76: Measure of Pelvic Angle in degrees during dynamic motion from a
slumped position to a super erect position. ..................................................... 187

Table 77: Average data from best fit line through a cross plot of the chair Total
Support Angle (TSA) vs the body Openness. r2 (Pearson product moment
correlation coefficient) is a measure of linear relationship between the two. .. 189

Table 78: Measure of body Openness and chair articulation (Total Support Angle,
TSA) in degrees during dynamic motion from a slumped position to a super
erect position. ................................................................................................. 189

Table 79: Measurement of the Body Recline Angle and Chair Total Recline
Angle in degrees for three recline angles. (Maximum angle equals the
maximum deviation from vertical) ................................................................... 189

Table 80: Offset between the BRA and TRA average maximum and minimum
measures. ....................................................................................................... 189

LIST OF FIGURES

Figure 1: SAE Oscar Manikin ..................................................................................... 7
Figure 2: ASPECT Manikin Prototype ........................................................................ 7
Figure 3: A two-dimensional representation of an interference model to estimate
deformation into an automotive seat. .................................................................. 9
Figure 4: Sections of a human spine ........................................................................ 11
Figure 5: The vertebral column from a side view, demonstrating kyphosis and
lordosis. ............................................................................................................ 12
Figure 6: Method for intervertebral discs obtaining nutrients. .................................. 13
Figure 7: Posture of the spine and pelvis in standing, slumped, neutral and erect
sitting conditions. .............................................................................................. 14
Figure 8: Lumbar region in a slouched posture, excessive amount of pressure
placed on the anterior portion of the disc. ......................................................... 14
Figure 9: Experimental chair used by Andersson25 for posture studies. .................. 18

Figure 10: Experimental chair used by Anderssonz6 allowing recline, thoracic
support, lumbar adjustments from +4 to —2 cm and pan tilt. ............................. 20

Figure 11 : Automotive seat used by Andersson”, where 1) seat back recline, 2)
lumbar support adjustment, 3) seat inclination,4) distance to dashboard, and

5) clutch pedal force. ........................................................................................ 21
Figure 12: From Andersson 26, eight different postures with normalized disc

pressure, refer to Table 1 for list of test conditions. .......................................... 22
Figure 13: Movement patterns of the thoracic region of the spine differ from

movement patterns of the lumbar spine10 ........................................................ 24
Figure 14: Preferred lumbar forces. Thoracic forceos necessary to lift a person

from a reclined position to an upright position". ............................................... 25
Figure 15: Forward Flexion from Hobson3. .............................................................. 27
Figure 16: Full Body Tilt from Hobson3. ................................................................... 28
Figure 17: Identification of the comfort ranges by Judic32 ........................................ 31
Figure 18: Pressure distribution from Thakurta’s 3‘ work ........................................... 32

Figure 19: Specially built laboratory chair used for testing. ...................................... 34
Figure 20: BAC chair articulation. ............................................................................ 35
Figure 21 : Anterior target sites. ................................................................................ 41
Figure 22: Posterior target sites. ............................................................................. 41
Figure 23: Leg targets. ............................................................................................. 42
Figure 24: Set-up conditions for the BAC ................................................................. 43
Figure 25: Slumped, neutral and erect positions of the test chair. ........................... 47
Figure 26: Hard seat reference file ........................................................................... 50
Figure 27: Computer image of subject being tracked, side view. ............................. 53
Figure 28: JOHN model with TRA measure. ............................................................ 55
Figure 29: Pelvis and thorax rotation about the upper and lower lumbar joints ........ 56
Figure 30: Vector used to determine pelvis orientation. ........................................... 59
Figure 31 : Sign convention of the thorax orientation ................................................ 60
Figure 32: JOHN model used to display the openness angle. ................................. 61
Figure 33: Body Recline Angle (BRA). ..................................................................... 62
Figure 34: Calculation of Head CG when only one head target is visible ................. 64
Figure 35: Measurement of wheel angle. ................................................................. 65
Figure 36: Angle measurements for the test chair, a larger angle is associated

with a more erect chair position. ....................................................................... 66
Figure 37: The Total Support Angle (TSA) of the test chair. A larger angle is

associated with a more erect chair position. ..................................................... 66
Figure 38: Force plate orientations. ......................................................................... 68
Figure 39: A comparison of the thorax and pelvis supports in a reference position

and a trial position ............................................................................................. 68
Figure 40: Rotation of the plate to recalculate the initial reference forces prior to

subtraction from the trial data. .......................................................................... 69

xvii

Figure 41 : Example of resultant force travel on the buttock plate for a dynamic
condition, top view. ........................................................................................... 71

Figure 42: Sectioning of body to determine the center of masses of each section. . 76

Figure 43: Body segment lengths expresses as a proportion of body stature by

Drillis and Contini (Roebuck, Kroemer, and Thomson, 1975)49 ....................... 79
Figure 44: Link boundaries and 9mass center locations as a percentage of link

lengths (Dempster, 1955)9 .............................................................................. 80
Figure 45: Calculation of CM for the upper arm, forearm and hand. ........................ 81

Figure 46: Subject 17, recline 1, neutral. Centers of Mass computed from the C7
skin point and from the St skin point. ............................................................... 86

Figure 47: Primary forces represent the normal forces under all of the force
plates except for the steering wheel where the shear force, Fx, was the
maximum force. ................................................................................................ 93

Figure 48: Primary support forces for the slumped condition at three recline
angles. .............................................................................................................. 96

Figure 49: Primary support forces for the neutral condition at three recline angles. 96
Figure 50: Primary support forces for the erect condition at three recline angles. 97

Figure 51: Primary support forces for the super erect condition at three recline
angles. .............................................................................................................. 97

Figure 52. Thoracic forces necessary to lift a person from a reclined position to
an upright position”. ......................................................................................... 99

Figure 53: Subject data for thorax support at three recline angles. Recline 1:
20°, Recline 2: 24°, Recline 3: 28° .................................................................. 100

Figure 54: Secondary forces measured under the buttocks, thighs and feet and
behind the thorax and pelvis Fx and the wheel Fz (positive is in the direction
of the arrows). ................................................................................................. 101

Figure 55: Shear forces measured were those acting on the support plates. The
forces acting on the body were equal in magnitude, but opposite in direction. 1 01

Figure 56: Secondary forces for the slumped condition comparing the three
reclines. .......................................................................................................... 102

Figure 57: Secondary forces for the neutral condition comparing the three
reclines. .......................................................................................................... 103

xviii

Figure 58: Secondary forces for the erect condition comparing the three reclines. 103

Figure 59: Secondary forces for the super erect condition comparing the three
reclines. .......................................................................................................... 104

Figure 60: Shear forces on the buttocks and pelvis region increase in magnitude
(negative direction) as the recline angle increases. ........................................ 106

Figure 61: Shear forces on the buttocks and pelvis region for the super erect
condition, decrease in magnitude (positive direction) as recline increases ..... 106

Figure 62: Primary seated support forces in Newtons. Comparison of chair
articulations .................................................................................................... 1 13

Figure 63: A comparison of the shear force directions between the Super Erect
and Slumped chair positions. .......................................................................... 117

Figure 64: Primary support forces in terms of Percent Body Weight (%BW). ........ 123

Figure 65: Visual display of the direction of loading on estimated L5/S1 joint
center .............................................................................................................. 130

Figure 66: From Andersson 58, eight different postures with normalized disc
pressure, refer to Table 1 for list of test conditions. ........................................ 131

Figure 67: Disc pressure measured in an office chair during simulated work
activities“. ...................................................................................................... 132

Figure 68: Comparison between Andersson58 and Bush (shaded) data for loading
of the lumbar spine. ........................................................................................ 133

Figure 70: Comparison of predicted L5/S1 forces during lifting of a large object... 134
Figure 71 : Openness as measured on the subjects. .............................................. 136

Figure 72: Total Support Angle, TSA is the thorax support angle plus the pelvic
support angle. ................................................................................................. 137

Figure 73: Primary forces represent the normal forces under all of the force
plates except for the steering wheel where the shear force, Fx, was the
maximum force. .............................................................................................. 156

Figure 74: Primary forces (Newtons) from Subject 16, in recline 2, dynamic
movement from neutral to super erect to slumped back toward super erect.
All force data is displayed with respect to the chair TSA (degrees). ............... 161

Figure 75: Primary forces from Subject 9, in recline 3, dynamic movement from
neutral to super erect to slumped back toward super erect. All force data is
displayed with respect to the chair TSA (degrees) .......................................... 162

xix

Figure 76: Primary forces from Subject 23, in recline 2, dynamic movement from
neutral to super erect to slumped back toward super erect. All force data is
displayed with respect to the chair TSA (degrees) .......................................... 163

Figure 77: Secondary forces measured under the buttocks, thighs and feet and
behind the thorax and pelvis Fx and the wheel Fz (positive is in the direction
of the arrows). ................................................................................................. 164

Figure 78: Secondary forces from Subject 9, in recline 2, dynamic movement
from neutral to super erect to slumped back toward super erect. All force
data is displayed with respect to the chair TSA (degrees). ............................. 168

Figure 79: Secondary forces from Subject 11, in recline 2, dynamic movement
from neutral to super erect to slumped back toward super erect. All force
data is displayed with respect to the chair TSA (degrees). ............................. 169

Figure 80: Secondary forces from Subject 13, in recline 2, dynamic movement
from neutral to slumped to super erect back toward slumped. All force data
is displayed with respect to the chair TSA (degrees). ..................................... 170

Figure 81: Location of resultant force for the thorax plate during dynamic chair
movement from slumped to erect. .................................................................. 176

Figure 82: Location of resultant force for the pelvis plate during dynamic chair
movement from slumped to erect. .................................................................. 177

Figure 83: Location of resultant force for the thigh plate during dynamic chair
movement from slumped to erect. .................................................................. 177

Figure 84: Location of resultant force for the buttocks plate during dynamic chair
movement from slumped to erect. .................................................................. 178

Figure 85: Equilibrium analysis of subject 11, recline 3 trial 2 depicting the
vertical and horizontal summed force data, along with the information from a
best fit line analysis ......................................................................................... 182

Figure 86: Equilibrium analysis of subject 16, recline 2 trial 1 depicting the
vertical and horizontal summed force data, along with the information from a

best fit line analysis ......................................................................................... 183
Figure 87: Chair TSA vs Body Openness for subject 11 in recline 3 ..................... 188
Figure 88: Chair TSA vs Body Openness for subject 20 in recline 2. .................... 189

Figure 89: A cross plot of the body recline angle (BRA) vs. the chair recline
angle. The chair recline was held constant while the subject was moved
from a slumped to an erect position. Subject 13, recline 2. ............................ 189

XX

Figure 90: A cross plot of the body recline angle (BRA) vs. the chair recline
angle. The chair recline was held constant while the subject was moved
from a slumped to an erect position. Subject 20, recline 2. ............................ 189

xxi

Introduction

With the transition of the work place into an information based facility, a
trend currently exists to replace heavy work load tasks with seated, light work, or
mental tasks‘. Many people spend not only the work day seated, but they are
also seated in automobiles, trains or planes on the way to and from work, and
once home they may even spend the evening seated in front of the television, or
at a theaterz. Due to the changing work environments and extended commuting
time, there is an increase in the number of people seated during the workday
thus increasing the duration they assume seated postures.

With people in seated positions for prolonged periods of time, it would be
beneficial to better understand the seated body and interactions occurring
between the body and the seat. The factors that have been proven to affect how
people sit and their interactions with the seat include: general health, age, activity
level of the individual, shear and normal forces, and the distribution of these
forces at the seat-body interface, temperature and humidity, and body posture3.
Hobson also included psychological deficits as a factor attributing to the seated
position, especially when referring to individuals spending long time periods in
wheelchairs.

One approach to studying seated postures, has been the use of
epidemiological evaluations on the relationship between occupation and back
pain. In a study of over 3000 individuals across eight different occupations,

Magora4' 5 investigated the relationship between low back pain (LBP) and

occupation. In the 1970 Magora4 study, data analysis was concentrated on
comparing the factors of age, sex, ethnic group and occupation to LBP. In terms
of LBP and its relationship to occupations, Magora4 found that the highest
incidence of LBP was in nurses and heavy industry workers; occupations which
required high physical effort and movement. The lowest incidence of LBP was
found in policemen, who spent the majority of their time either standing, or
walking. Magora4 also found an increasing linear relationship between age and
LBP; in comparisons between men and women, women tended to have a higher
incidence of LBP in occupations requiring high amounts of physical effort. When
the data were analyzed in terms of ethnicity, the Asian ethnic group recorded the
lowest LBP incidence while individuals in the western hemisphere had the
highest.

A second paper in 1972 by Magora5 divided each occupation into three
different sub-categories relating to time sitting, time standing, and weight lifted.
The subjects were placed into each category based on the amount of time they
spent standing, sitting or lifting an object (with a minimum mass of 5 kg). With
this more detailed analysis, Magora5 found a significant relationship between
LBP and the individuals who predominantly stood or were in a predominantly
seated posture throughout the work day. Weight lifting or the frequency of lifting
an object did not play a significant role in the causation or triggering of low back
pain. LBP occurred in a higher percentage of those who never lifted a 5 kg

object during their work as compared to those who performed lifting taskss.

The difference between the outcomes of the Magora 1970 4 and 1972 5
results were clearly due to the incorporation of the time factor into the statistical
analysis. These results were supported by Pope6 and Zacharkow7, who both
stated that uninterrupted sitting and uninterrupted standing had a profound
causal effect on musculoskeletal disease, specifically, low back pain.

Magora5 noted that the discrepancy between his findings and other
studies that also concluded the lifting of a weight to be a cause of LBP was
primarily due to the fact that other studies were performed in a retrospective
fashion on subjects who had already developed LBP. Magora used a group of
controls along with those experiencing LBP. Because of lack of standardization
in terms of either the measurements used or the diagnostic classifications
chosen by the medical staff, Troup8 also agreed that retrospective studies tended
to be unreliable.

One author8 felt that there was a link between driving and back problems,
or sore backs. A study investigating the relationship between herniated
intervertebral lumbar discs and driving found that individuals who spent more
than half of their working hours driving were three times more likely to suffer
lumbar disc herniation than those who did not8. The risks for development of a
herniated lumbar disc were even greater for the professional truck drivers than
for other drivers; the truck drivers also displayed an increased number of bone
deformities as compared to other drivers”.

Low back pain has been an extensive problem, affecting a large portion of

the population; studies6 cited that 70% of the United States population

experienced low back pain at one time in their life. Researchers such as
Magora4'5 and Troup8 have performed work attempting to identify causes of back
problems, mainly trying to link LBP to occupations.

Another approach to identifying the cause of back problems, aside from
the collection of epidemiological data, involved the study of the biomechanical
factors of sitting; including seated posturesg'w, muscle activity", chair design ‘2
and pressure or force distribution”. The approach of evaluating the interaction of
the body with a physical device, in terms of biomechanical factors, was the
approach taken for the work discussed in the following dissertation.

To better understand seated postures, research conducted for this
dissertation included data collection and analysis of the following biomechanical
factors: shear and normal forces, chair position, and seated postures. Currently,
the data presented in this dissertation are the only data of their kind. It is
expected that the availability of these data will have a significant impact on
seating design, seating evaluation and the tools used for seating design and
evaluation.

First, a brief scope of the research will be presented, followed by an
overview of the related literature. Following the literature review, is a detailed
discussion of the equipment and methods used for testing. The final sections
summarize the results and findings, interpret the meaning of these data and

discuss the relevance of these findings.

Scope of Research

Data
One facet of this research was to experimentally measure how mid-sized

male subjects loaded an experimental test seat that was placed in an
automotive-like environment. The seat and surrounding package, including the
steering wheel and foot rest, were moved through various configurations to
promote an array of postures that would be representative of positions in a mid-
sized automobile interior package.:One set of tests involved static measures
where the position of the seat was set, then posture and support force data were
collected simultaneously on this single position. In a second set of trials, the
subjects were dynamically moved through several positions while taking quasi-
static force and posture measures. With these data, comparisons were made
between the different seated postures and the transfer of seated support loads
during postural change;

3/

Because of the application of the support force and posture data to the
development of computer models, a second objective of this research was to
develop a mathematical model for determining the approximate joint forces and
moments in a section of the lumbar spine. Once the mathematical steps were
developed, a set of data from one subject was used to test the methodology and

evaluate the results.

Usage of Data
Two research projects have already incorporated the results of this

investigation. The first use for this data was through a project entitled ASPECT

(Automotive Seat and Package Evaluation Tools)”. This research was funded
by eleven automotive manufactures and seat suppliers (Johnson Controls,
Magna Interior Systems, Lear Corporation, General Motors, Chrysler, Ford,
Volkswagen, Volvo, PSA Peugeot - Citroen, BMW and Toyota) through the
Society of Automotive Engineers. The Biomechanical Design Research
Laboratory at Michigan State University and the University of Michigan
Transportation Research Institute were the two organizations that performed the
research. The primary task of this project was to develop a new three-
dimensional, full size, automotive seat evaluation and design tool to replace the
current Society of Engineers (SAE) Recommended Practice J826 that uses the
three dimensional SAE manikin (Figure 1)“. The new ASPECT manikin (Figure
2) was a more human-like representation of a mid-size male seated occupant
based upon the JOHN computer model developed at Michigan State University‘s.
(Details of the JOHN model are located in the Methods Section).

The main design changes in the ASPECT manikin included an articulating
torso, with additional joints in the torso to allow changes of lumbar curvature, legs
made of lightweight aluminum (not containing significant mass as in the current

SAE legs)”, and human-like body contours and mass distribution“ ‘3.

Since the ASPECT manikin without legs, arms, or a head, should load the
seat like a human (deforming the seat back and seat pan like a human),
obtaining the appropriate loading for each of the body segments was an
important portion of the new manikin development. Data collected by Bush13
determined the appropriate mass distribution for a more human-like loading into

the seat by the newly developed ASPECT manikin.

 

Figure 2: ASPECT Manikin Prototype

A second project currently underway, incorporates the data from this
dissertation into a computer model used to predict seated posture and support
forces. The initial phases of this simulation have already begun by Bush”, and
incorporate a human body representation in a seated environment. The human
body representation has a linkage like the JOHN model and the new ASPECT
manikin, and rigid segments with an articulating torso. Each segment has been
assigned a human-like amount of mass, and the human model interacts with a
computer representation of an automotive seat.

A more sophisticated model, to be developed in conjunction with a
software company (TecMath), will allow key descriptors of the automotive interior
package to be entered into the computer, and the posture of the human model to
be predicted. With the predicted posture, the loading on the seat pan for the
human model can be estimated. This loading estimate will be based on the
posture and support force relationships measured for this research. Even though
these models will be treated as rigid shapes, the back and buttocks will have a
deformed contour and the amount of seat deformation can be estimated by the
amount of interference between the human model and the seat. A basic two-
dimensional representation of this interference can be seen in Figure 3.

The position of the human model in the seat is valuable to seat designers
and to automotive interior package designers; knowing the occupant location
allows the designer to identify the placement of the occupant restraints, estimate

vision restrictions and develop appropriate foam distances from seat structures.

 

+ arm
+ torso

—o— leg

— thorax
—— pelvis

o acromion ‘ '
—seat contour /
3,4

 

 

 

Figure 3: A two-dimensional representation of an interference model to
estimate deformation into an automotive seat.

The final phase of the incorporation of these data into a computer model
will be accomplished with TecMath, a German modeling company that supplies
more that half of the automotive companies world-wide with automotive interior

package and human comfort analysis.

Literature Survey

Spinal Column
To appreciate the importance of seated postures, it is first necessary to

understand the anatomy and physiology of the spinal column and intervertebral
discs”. The spinal column is typically composed of 33 bones called vertebrae.
These bones are stacked on top of one another with discs between the
vertebrae. Each vertebra is named with respect to the region of the body in
which the bone is located. There are 5 different sections of the spinal column:
cervical, thoracic, lumbar, sacral, and coccygeal region, Figure 4 . The cervical
region contains seven vertebra, named from top to bottom as C1-C7. The
thoracic region, which is sometimes termed the thorax, contains the rib cage and
twelve vertebrae, named T1 -T12 from top to bottom. The lumbar region contains
5 vertebrae and is referred to as L1 -L5. The sacral region is part of the pelvis
and also has five vertebrae, but unlike the previously mentioned sections, these
vertebrae are fused together at birth and referred to as the sacrum. The final
region, the coccygeal region also has 4 fused vertebrae and is termed the
coccyx.

The spinal column has several purposes including support, mobility, and
protections. It is the framework for the underlying structures and supports the
upper body including the rib cage, head and arms. The spinal column allows for
twisting, and bending of the body; this flexibility is provided by the shape of the
individual vertebrae and the intervertebral discs. Lastly, the series of vertebrae

that form the column provide protection for the spinal cord and nerves.

10

In the natural erect standing posture, the spinal column is normally straight
from the anterior/posterior perspective, but from the medial/lateral perspective, it
has sections of curvature. The cervical and lumbar regions have a lordosis, or
curvature inward toward the front of the body (concave anterior), Figure 5, and

the thoracic region has kyphosis or a curvature away from the front of the body

1 Cervical

Thoracic

(concave posterior), Figure 5.

Lumbar

:l Sacral

Figure 4: Sections of a human spine.

 

 

Between each pair of the vertebral bodies, there is an intervertebral disc.
The disc is primarily composed of two structures, the annulus fibrosus and
nucleus pulposus. The outer-most portion of the disc is called the annulus
fibrosus and is composed of rings of collagenous fibers. Each fibrous layer
contains a different fiber orientation and the lamination of these layers provides
the strength of the disc structures. The center of the intervertebral disc, called the

nucleus pulpous, is a softer, highly elastic, fluid filled section”, which Is

11

contained by the annulus. The nucleus is 80% water and gradually desiccates

with ageszo.

 

 

Lordosis ’ Cervical

Kyphosis 3.1“: Thoracic
Lordosis Lumbar
Sacral

1 Coccyx

Figure 5: The vertebral column from a slde view, demonstrating kyphosis
and lordosis.

Since the discs of adults are avascular, the supply of nutrients depends on
the diffusion of substances from surrounding areas," ‘9 rather than directly from
the supply of blood. A relationship between hydrostatic pressure and osmotic
pressure drive the nutrient absorption, Figure 6. As the pressure inside the disc
is increased due to the weight of the torso, or movement of the torso, the disc
bulges and some of the fluid content in the disc is pushed out into surrounding
fluid. As the pressure is released, the disc reabsorbs nutrients from the
surrounding fluids. This method of nutrient distribution has also been referred to

as a ‘bellows” or “pump” action”.

12

   

Figure 6: Method for intervertebral discs obtaining nutrients.

Posture
The relative positions of the vertebrae define a person’s torso posture.

Compared to normal standing posture, slouching, or having a slumped posture,
causes the upper part of the pelvis to be rotated rearward, Figure 7, which results
in lumbar spine flattening. Lumbar spine flattening removes the lordotic
curvature found in the lumbar spine during standing and produces a flat or
straight lumbar spinee' 2'. When this occurs, additional pressure is placed on the
anterior sides of intervertebral discs while the posterior portions of the annuli
fibrosi are stretched, Figure 8. Maintaining this slumped position for prolonged
periods of time, without frequent movement, increases the risk of back
problems”. As a result, repeated compression of the discs in the flat lumbar
spine condition can cause the nucleus pulposus to protrude rearward through the
annulus placing pressure on spinal nerves; this condition is called a herniated

disc.

13

Standing Slumped Neutral

   

Figure 7: Posture of the spine and pelvis in standing, slumped, neutral and
erect sitting conditions.

 

Figure 8: Lumbar region in a slouched posture, excessive amount of
pressure placed on the anterior portion of the disc.

Verifying that a slouched posture increased disc pressure, Headman23
measured the compressive loads in discs by placing a thin load cell and pressure
sensitive film between lumbar vertebrae. Headman23 tested 12 lumbo-sacral
(L1-S1) spines in saline solution, he flexed and extended the spines while
measuring force between the vertebrae. The force sensing material was placed
under both the anterior and posterior portions of the vertebrae (between the
posterior facets of L3-L4 and L4-L5 and in the anterior, superior portion of L4 and

L5 vertebra). He concluded that high anterior disc pressures resulted when the

14

spine was in a flexed position, corresponding to a flat lumbar spine; while the
anterior disc forces were low for the extended spinal position, corresponding to
lumbar lordosis.

Not only is posture important in preventing spinal injuries, but posture also
plays an important role in keeping intervertebral discs healthy. Since,
intervertebral discs rely on pressure changes to maintain a supply of nutrition,
good posture and frequent changes of posture help keep the discs healthye' 22* 23.
Headman23 cited a case study, which agreed with Magoras, where people who
rarely sat had the highest incidence of back pain followed by those who
predominantly sat. Those who sat frequently for brief periods of time had
virtually no back problems”. Headman concluded that prolonged standing and
uninterrupted sitting should be avoided and that the integrity of the spine could
be best preserved by frequent changes in posture.

Seymour22 also agreed that the spine needed to be moved. Seymourz2
theorized that back problems endured in people who spent long periods driving,
not necessarily because of poor seat designs, but because of the long periods
spent motionless in a seated position with the back subjected to repeated jolts
and vibrations.

Bonney24 used a method entitled "shrinkage" to relate posture to disc
pressure. Some people stay seated the entire work day and by the end of the
day, they may shrink or shorten between 15 to 20 mm due to the changes in fluid
absorption by the intervertebral discs22' 24. Bonney used the shrinkage method to

measure the positions of spinous processes (in a sagittal plane) during an initial

15

seated condition and again, after a period of time, in a second seated test
position. If there was a decreased length between these spinous process
measures, then this was considered to be spinal shrinkage. Bonney related
measures of vertebral column shrinkage, or postural changes to disc pressures:
increased shrinkage was linked to increased disc pressures.

Bonney24 also found that a 90° (upright) back angle of the test seat
caused and average of 1.3 mm spinal shrinkage, while a test seat with a 110°
back angle (reclined 20 degrees from vertical) caused 0.8 mm of lengthening.
Therefore, according to Bonney an increased recline angle caused a decrease in
disc compression.

[Andersson25 evaluated the influence of physical characteristics of seats
such as the magnitude of lumbar prominence, the height of lumbar prominence
and the back rest angle also known as recline angle, on posture, specifically
changes in lumbar lordosis. To monitor changes in spinal articulation,
radiographs were taken subjects in the sagittal plane and were interpreted for
changes in lumbar lordosis}

The first of the four tests evaluated differences in lumbar lordosis from
standing to unsupported sitting. For these tests, Andersson” found an average
decrease in total lumbar curvature of 38 degrees; 28 of the 38 degrees occurred
due to the rearward rotation of the pelvis.

”The second set of tests evaluated the effects of changes in the chair’s
recline angle. The largest effect, caused by changing the recline angle, was an

increase in pelvis rotation relative to a standing position. The rotation was

16

measured by drawing a line from the sacrum to a point on the acetabulum and
comparing the angle of that line relative to a horizontal line25. The pelvis angle

measured 63.8° in standing and rotated to 53° in the 80° recline and then to 22°

in the 110° recline. ‘4

”The third parameter Andersson25 changed was the magnitude of the
lumbar support. The lumbar support prominence was varied (Figure 9) from -2.0
cm (reanivard from the back support) to +4.0 cm (protruding forward from the
back support). The change in lumbar support magnitude had a significant effect
on lumbar lordosis. The total lumbar angle increased 37 degrees from the -2.0
cm condition to the +4.0 cm condition; 20 of the 37 degrees was caused by pelvis
derotation25.

No significant change was found in lumbar lordosis with the change in
lumbar height2°.:\,

It should be noted that knee flexion was not a controlled variable in the
study by Andersson”, and knee flexion is known to have an affect on the ability
of people to rotate their pelvis. As the knees extend, the hamstring muscles,
which attach to the bottom of the pelvis become taut and do not allow the top of
the pelvis to rotate forward, which inhibit rather than facilitate lumbar lordosis. As

the knees flex, the opposite also becomes true. it should also be noted that the

subject size was not a controlled variable either.

17

80° 100°

1 1

+4 cm -2 cm
Recline

V’

 

A
l

T

\

 

 

 

 

 

 

 

 

 

 

 

 

J u

l. ..

Figure 9: Experimental chair used by Andersson” for posture studies.

Physiological Effects from Postural Changes
Research has been conducted to examine the correlation of physiological

changes in the body with posture, specifically muscle activity and fatigue. One
study by Bush11 examined the fatigue rate of back muscles for subjects seated in
two different automotive seats. Subjects were tested at 80% maximal voluntary
contraction and showed a slower rate of fatigue when seated in a firm automotive
seat containing a prominent lumbar support, as compared to a faster fatigue rate
in a soft seat with little lumbar prominence.\:he Bush11 study showed a
correlation between a change in posture and a change in the body’s physiology
defined as fatigue rate. It was concluded that the firm seat, with a prominent
lumbar support, provided a mechanical advantage for the muscles, which
exceeded that for the soft, flat seat. The seat with the prominent lumbar support
placed the spine in a position of increasing lumbar lordosis, moving the spine

"”3

closer to the natural amount of lumbar lordosis exhibited during standing. .

18

According to another researcher7, in an erect posture, the ribs, chest and
diaphragm are raised, facilitating movements of the diaphragm and thus
breathing is promoted. If breathing is promoted, the oxygen content in the blood
is higher, thus, muscles fatigue at a slower rate. This correlates with the case for
the firm seat with prominent lumbar support discussed in the Bush11 study.

Qt, additional studies by Anderssonze' 27, muscle activity and disc pressure
data were collected as various adjustments were made to different seating
apparatus. Using an experimental chair26 the back rest, lumbar support, thoracic
support angle, and seat pan angle were adjusted (Figure 10), while in the

automotive seat27

the recline angle, seat pan angle and the lumbar support were
adjusted.
In the experimental chair, Andersson26 tested four adults in eight different

positions (Table 1).

Table 1: Conditions in an experimental chair tested by Andersson26 while
monitoring disc pressure and muscle activity.

 

Standing at ease.

 

Relaxed (no back support) sitting with arms at the sides of the body.

 

Relaxed sitting with the arms supported.

 

Relaxed sitting with the feet unswported.

 

Straight (erect) sitting.

 

Relaxed anterior sittinflqk.

 

Straight anterior sitting.

53".\‘."’S":"“.“°!\3:"L

 

 

RelaxedJmsterior sitting.

 

19

 

90° 10°
100°
800 l" 1 1 00 f

\
120°
{4 ,.

 

V

 

 

 

 

 

 

 

 

L

 

 

L.

Figure 10: Experimental chair used by Andersson26 allowing recline,
thoracic support, lumbar adjustments from +4 to —2 cm and pan tilt.

The same four adults that were tested in Andersson’s experimental chair26
were also tested in an automotive seat”. The following seat parameters were
evaluated in the automotive study: four recline angles, 90°,100°,110°,120°; two
seat pan angles 10°, 14°; and five lumbar prominences, 0 to 5 cm, in 1 cm
increments, Figure 111

Needle electrodes were placed 3 centimeters lateral of the midline on both
left and right sides at the C4, T5 and L1 vertebral levels and in the psoas
muscles. Additional electrodes were placed at T8, T10 and L3 on the left side,
for two of the subjects. The sensor for disc pressure was inserted into the center
of the third lumbar disc, which was defined as the disc between the third and

fourth lumbar vertebrae.

20

 

 

 

 

 

 

 

A
v—f

Figure 11: Automotive seat used by Andersson”, where 1) seat back
recline, 2) lumbar support adjustment, 3) seat inclination,4) distance to
dashboard, and 5) clutch pedal force.

Anderssonze' 27 found that an increase in recline angle and an increase in
the lumbar prominence reduced the pressure measured in the L3 disc for both
the experimental chair and the automobile seat. The decrease in disc pressure
was most pronounced when the lumbar support was adjusted from the 0 cm
position to the +2 cm position in the experimental laboratory chair and +5 cm in
the automotive seat. The lowest disc pressure was found with the automotive
seat was reclined to 120° with the seat pan angle adjusted to 14°.

In the experimental chair“, the rotation of the top of thorax support
forward (Figure 10) increased disc pressure. The highest amount of disc
pressure was observed when the subjects were in relaxed anterior sitting
(condition 6 in Table 1) and the lowest overall disc pressure was when the

subjects were standing.

21

0.2 .4

0.1 -I
. f

0.0 1+ 1, 4.“,
in ,
)3

ti
#-
1

til] i”
-o.1 _. ..‘f, ‘

l

-o.2:r‘—I

 

. v" I.”

 

Normalized Disc Pressure KPa

Q—r-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 12: From Andersson 26, eight different postures with normalized disc
pressure, refer to Table 1 for list of test conditions.

[In the experimental chair26 the largest decrease in muscle activity
occurred in the lumbar region with the back rest reclined. Changes in lumbar
support, thoracic support, and in the seat pan angle only had minor affects on
muscle activity.

In Andersson’s automotive seat”, a decrease in muscle activity was found
relative to an increased recline angle, increased pan angle and increased lumbar
support prominence. In terms of muscle activity, not fatigue rate, low muscle
activity was found with a recline angle of 120° and a pan angle at 14° and the

. f1.
lumbar prominence at 5 cm. I.

_—//

22

Literature Related to Support Force Data
Not only does posture affect what happens to the internal structures of the

body such as diaphragm position, nutrient and blood flow, but the posture of a
person also affects how the body interacts with the external environment. The
postures of an individual affect their vision and reach zones, both important
factors when designing for a work or driving environment. In any type of seated
environment, a person’s posture also affects how their body contacts the seat
surface, and how support forces are transmitted between body and seat.

The findings of an extensive search in the area of seating revealed that
little research has been performed on measuring or documenting the support
forces of the body in the seated position. One paper by Faiks and Reinecke10
discussed research that can be compared to a portion of the data collected for
this dissertation. Otherwise, the most closely related research has been
performed in the medical and automotive industries and involved the
measurement of pressures under the buttocks and behind the torso.

Faiks 4° studied the movement patterns of people’s spines during
unsupported sitting tasks. He also evaluated the preferred lumbar forces
(subject chosen force values) produced during movement from a reclined
position to an upright position and the seat back support forces generated by

“lifting” a person from a reclined position to an upright position.

23

 

 

L

Figure 13: Movement patterns of the thoracic region of the spine differ
from movement patterns of the lumbar spine”.

Faiks concluded that the path and rate of motion of the lumbar spine was
independent of the path and rate of motion of the thoracic spine, Figure 13.
Faiks1o also determined subjects preferred forces in the lumbar region, and the
force necessary to lift the body from a reclined position to an upright position in
the thoracic region, Figure 14. These data showed that the subjects (n=21)
required an increasing amount of force on the thoracic region during a twenty
degree recline, while the lumbar support force maintained a nearly constant level
during recline. He found that as the amount of lumbar support was increased,
the amount of force needed in the thoracic region was decreased and the
opposite also held true.

‘\ Data broken down by gender showed that for men, the support levels for
both the thoracic and lumbar region were at the same magnitude in the upright

position and the thoracic force increased while lumbar stayed constant as they

24

reclined. For the women subjects, the magnitude of thoracic support was lower
than the lumbar support level in the upright position and increased during recline,

but stayed below the desired level of the lumbar suppo/rt]

Thoracic Support

Lumbar

    

40
30 30

539
-
20 max
a
In

Force (lbs)

10 10

     

0 5 2o 0
Degrees

Figure 14: Preferred lumbar forces. Thoracic forces necessary to lift a
person from a reclined position to an upright position‘ .

In terms of force or pressure distribution, two philosophies exist for
determining what is most comfortable or desirable for people. The first thought is
that an even distribution is the best for the body; the second thought is that the
forces should be distributed such that the higher loads are concentrated around
the stronger structures of the body, for example under the ischial tuberosities
when seated”. Goonetilleke28 provided reasoning that supported each of these
philosophies, but also offered a third suggestion, that higher localized forces,
rather than an even distribution may be desirable, such as the pressures that
would be induced by the use of a beaded seat cover. The beaded seat cover
produced high localized pressures, above the recommended levels for the
prevention of ischemia (less than 4.0-4.7 kPa) 28, yet was still found to be

desirable by many people. As of yet, no single method for pressure distribution

25

has been adopted by researchers or industry, the relationship of the pressure
and force distribution to comfort is still determined on an individual basis.

Researchers” ” have provided data on the amounts of pressure seen in
regions of the body during sitting. Philippe”, used pressure mapping to
determine the contact area under the region of the thighs and buttocks and
estimated the forces under the left and right side of the buttocks and thighs.
Philippe’s ” research measured pressures in three different conditions for a man
1.75 m tall and having a mass of 85 kg: 1) in a seat with no vehicle controls (no
steering wheel or accelerator pedal), 2) in a seat while using controls, and 3) in a
moving automobile. Philippe” estimated the forces under the buttocks and
thighs in the second condition and reported the approximate value of 196 N (20
kg) under the left buttock, 215 N (22 kg) under the right buttock, 88 N (9 kg)
under the right thigh and 78 N (8 kg) under the left thigh, or terms of percent of
body weight (%BW) 2426 %BW under each buttock and 9-10 °/oBW under each
thigh. Goonetilleke” listed the following measurements of pressures of a person
in a seated position in terms of percent body weight; 18 °/oBW under the ischial
tuberosities, 21 °/oBW under each thigh and 5 °/oBW under the sacrum. Large
differences in °/oBW were given between Phillipe” and Goonetilleke”, most likely
because each had a different definition for the region used to convert pressure to
force. Neither researcher provided these definitions.

The pressures at the body-seat interface are important for the disabled
population, especially to those confined to a wheelchair. Without movement of

the body, long periods of high pressures in localized regions can cause decubitus

26

ulcers3'”, also known as pressure sores. Also affecting the development of
pressure sores are temperature and humidity at the body-seat interface and the
posture of the person. A study by Hobson3 compared pressures produced on
the seat pan by spinal cord injured (SCI) subjects to non-disabled (ND) subjects
in different postures. These postural changes were induced by different
configurations of a wheelchair. Hobson3 found that the SCI subjects consistently
produced higher pressures in all postures as compared to the ND subjects.
Higher pressures would be expected because of the loss of muscle mass
producing a more prominent skeleton in most SCI individuals.

From the Hobson3 data, postures of a fonIvard flexion of 50 degrees
produced the lowest pressures on the seat pan for both subject groups. FonIvard
flexion of 30 degrees also showed low pressures for the ND group. The ND
subject group produced equally low pressures as in the 30 degree forward flexed
condition in the full body tilt of 20 degrees with 100 degrees of back rest recline.

SCI subjects also produced lower pressures in the full body tilt position.

50° Forward Flexion

 

 

 

 

I} 1 I}: Illllllll’lll’lllll

 

Figure 15: Forward Flexion from Hobson".

27

a 100° Recline
W’ 209 Tilt

‘5‘: '-

§

.\ MQ\K
i‘t‘iaa/‘ /’

G? , , .0

(I'll4 Ill'l'i'lllllF’II’I'I‘

Figure 16: Full Body Tilt from Hobsona.

In the Results section of this dissertation, discussions will include
comparisons to the force measures reported by Faiks”. Discussions will also
address dynamic data that shows possible methods of reducing or shifting the

loading under the buttocks to decrease, or prevent, the occurrence of decubitus

ulcers.

28

Comfort Related Literature

Since there is a wide variation in the human anthropometry, human body
behavior and human preference, it has been difficult to interpret and define
“comfort”, or to define what makes a person comfortable and keeps them
comfortable over time. Many researchers have tried to define what is

comfortable for a persona 23' 30.31

. However, difficulties arose in defining the term
comfort. Some researchers stated that comfort was the absence of discomfort”,
or that comfort was associated with feelings of relaxation and well-being”.
Judic32 stated that in the automotive environment there were several factors
which attributed to a person’s comfort, including visual comfort, initial touch,
postural comfort, dynamic comfort, sonic comfort and thermal comfort.

Designing a seat today requires that the seat pass several standards and
tests, including fatigue tests, wear tests, and restraint testing; however a comfort
test has not yet been developed. If a link could be established between the
objective measures and the subjective ratings, or a series of links could be
developed between several objective and subjective factors, a comfort test could
be designed. This test would provide basic guidelines for designing comfortable
seats and would address seat factors that affect a person’s overall comfort.

Grandjean (1964)2 established some basic seat positioning guidelines for
reduced discomfort in seated postures, Table 2. Grandjean found seat
inclination for minimum discomfort should be between 16 and 30 degrees.

Backrest inclination should be between 125 and 138 degrees from horizontal, the

seat height should be between 34 and 50 cm and the seat depth should be

29

between 41 and 55 cm. Table 2 compares the comfort guidelines established by

Grandjean to th@r positions at produced low intervertebral disc pressures

as found by Andersson and NaChemson (1970 & 1974) ”'34.

Table 2: Comparison of seat orientation based on low disc pressure
(Anderssona3 and Nachemson“) and minimum discomfort (Grandjeanz)

 

 

 

 

 

 

 

Low Disc Pressure Minimum Discomfort
(Andersson and (Grandjean)
Nachemson)
Seat Pan Inclination
(with respect to 14° 16-30°
horizontal)
// Backrest Angle 1 10-130° 125-138°
\" Height of Seat 34-50 cm
Depth of Seat 41 -55 cm

 

 

Table 3: Recommended body angle ranges to avoid discomfort”.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Definition Figure Ref. Minimum Maximum

(Figure 17) (degrees) (degrees)
Torso Axis — vertical a1 20 30
Torso axis — Thigh axis a2 95 120
Thigh axis — Lower legaxis a3 95 135
Leg axis — Pedal plane a4 78 105
Torso axis — Upper arm axis a5 0 50
Ujper arm — Forearm axis a6 80 170
F/orearm axis — Hand axis a7 170 190
\g/Pedal plane — Floor pan a8 40 70
filigh axis — Horizontal a9 12 25

 

3O

 

 

 

Figure 17: Identification of the comfort ranges by Judic‘"2

Judic32 also listed a set of data from Rebiffe’s35 research, Table 3,
describing the ranges of body positions that should be assumed to avoid
discomfort. Rebiffe developed an eight-segment body linkage system and used
this system in conjunction with experimentation of subject preference in
adjustable and unadjustable automotive seats to establish vision zones and
steering wheel zones for subjects of the smaller (5‘h percentile) and larger stature
(95th percentile).

Thakurtaa1 performed research that evaluated short and long term
comfort. Thakurta measured pressure distribution and collected subjective data
on 36 subjects at the beginning and end of an 80 mile drive. Each subject was
tested in five automobiles. The average distribution of pressure is displayed in
Figure 18. The objective measures were broken into the following pressure

zones: ischial, thigh, shoulder, tail burn on the cushion, tail burn on the back,

31

lateral cushion, lateral back, and lumbar. The subjective questions also

addressed these zones. Thakurta tried to establish a linear correlation between
the subjective questions addressing discomfort of body zones with the objective
pressure readings. Thakurta concluded that comfort was most likely a complex

non-linear relationship of many measures.

:1 Shoulder
Cush. Tail Burn 9 Thigh
Back Tail Burn 6% Shoulder I lschial
5% 10%
D Lun'bar
I Back Lat.
51 Cushion Lat.
I: Back Tail Burn
E Cush. Tail Burn

     

Cushion La giiii.

Th'gh
15°/o ‘

Schial
45%

Figure 18: Pressure distribution from Thakurta’s31 work.

An overview of the primary literature relating to the anatomy and
physiology of the spinal column, posture, the physiological effects of postural
changes and comfort has been presented. The next section discusses the
methods used to measure and define the posture of a seated person, the support

forces and the estimation of internal joint forces.

32

Testing Methods

The following section discusses the equipment used for testing, the
subject selection criteria, subject targeting and a description of the various test

postures.

Equipment Description

Lab Chair
The specially built chair used in this study was termed the

Biomechanically Articulating Chair (BAC) 36, it was an experimental apparatus
that was designed to move with people through postural changes. The BAC
(Figure 19) consisted of rigid sections providing support behind the thorax,
behind the sacral region, under the buttocks and under the thighsEl’ he rigid
supports were covered with a foam-backed fabric of approximately 15 mm in
thickness (13 mm of foam and 2 mm of fabric). The chair allowed two main
movements, one being the recline of the body and the second being the
increasing or decreasing of lordotic curvature in lumbar spine. Each of these
movements was powered by a small motor (adapted from a powered automotive
seat) which used a rotating screw to move the chair.

File chair was designed to move with the body and to recline about a point
located (approximately) under the ischial tuberosities of a seated person. This
recline point was chosen based on the assumption that the location of the
person’s ischial tuberosities did not change while in a seated position, therefore

the person would rotate about the ischial tuberosities while reclining. Allowing

33

the chair to pivot about this point, in theory, would reduce the shear forces on the
back of the chair and the back of the person as the seat reclined)

The BAC also had a unique method for inducing lordotic postures,
Figure 20. The chair did not have a typical lumbar support, such as bladder or
paddle devices commonly found in automotive seats. Instead, the chair
articulated the thoracic support and pelvic support in equal and opposite
directions, thus a lordotic curve in the lumbar spine was induced without using a
lumbar support. The buttocks section of the BAC cradled the pelvis, including
the sacrum and the buttocks region, slightly forward of the ischial tuberosities.
The thigh support did not move with the pelvis portion of the chair, rather it
rotated about an axis just forward of the ischial tuberosities and was fixed at an

angle 15° above horizontal.

. Thorax

Chair
Targets Support

. Pelvis
Buttocks Support
Support

’ Thigh Support
Force
Transducer

 

Figure 19: Specially built laboratory chair used for testing.

34

  

    

 
 
  

Buttocks and

B”"°°k° and Thigh Thioh Pivot

PeIVIs Support Support

Figure 20: BAC chair articulation.

@e BAC was used instead of an automotive seat because not only could
the BAC accommodate a large range of anthropometry and provide a large range
of movement, but it also could be positioned in a repeatable fashion with minimal
effort. Another reason for using the BAC rather than a conventional automotive
seat was the open seat back design that allowed easy access to different body
landmarks during testing.

It is necessary to make a small note pertaining to the BAC design.
Although this design was not a conventional practice in the automotive field, the
idea of stabilizing the pelvis during sitting and inducing spinal articulation by
moving the pelvis has been a common idea among the medical profession.
Zacharkows7 stated that a conventional lumbar support (such as a lumbar roll)

did not assure the proper pelvic alignment relative to the spine and that without

35

proper pelvic support continuous oscillations of the pelvis occurred. Zacharkow
believed that rather than providing support to the spine as in a lumbar support,
the pelvis, specifically the sacrum and posterior iliac crests, should be well

supported and positioned to maintain lordosis in a seated posturef.\

/

Force Plates
A second type of equipment used in conjunction with the BAC was a set of

multi-channel load cells (also termed force plates) located behind each of the
support plates (shown in Figure 19) of the chair. The force transducers
measured support forces in three directions, Fx, Fy and F2 and three moments

Mx, My and M2 relative to the center of the force plate. Plates were mounted
behind the thorax, pelvis, and steering wheel, and under the buttocks, thighs, and
feet. The force plates were mounted between two rigid boards, one board was
attached to the chair mechanism and the other board (which was foam covered)
came into contact with the subject at the various anatomical regions. With this
type of attachment, all the forces and moments from the seated subject were
transmitted through the force plates.

The capacities of the force plates were 1112 N (250 lb) (behind the thorax
and pelvis) and 4450 N (1000 lb) (buttocks, wheel, thigh, feet). The force plates
were Advanced Mechanical Technology, Inc. (AMTI) MC3A series plates. The
small size of the plates was the primary reason for the choice; these plates were
7.62 cm (3 inches) in height, width and depth.

The force plates were used in conjunction with AMTI amplifiers. The amplifiers

were set to a gain of 1000 and a filter of 10.5 Hz. Each force plate came with

36

factory measured sensitivities for each channel. These sensitivities were

incorporated into the calibration file for data processing.

Motion Measurement System
A five camera 60 hertz Oualisys” motion measurement system was used

to capture three dimensional positions of retro-reflective targets that were
secured to both the articulating chair and the bony landmarks of the person. All
five cameras contained high power, infra-red light rings mounted around each
lens; the use of the infra-red system allowed testing to occur with ambient light in
the laboratory. The camera speed was variable from 1Hz to a doubled-up
camera system of 120Hz. All force and motion data were collected at 12 Hz.
With pilot testing, 12 Hz was deemed acceptable to capture the static and
dynamic seated movements.

The targets were made of lightweight spherical balls, covered with 3M
high gain, 7610, retro-reflective tape”. The targets were 18 mm in diameter
without the reflective tape and approximately 20 mm once covered. This
reflective tape was used for special effects projection screens and provided at
least 600 times more reflection than a white screen. Together the reflective tape
and the spherical shape of the targets helped to maximize the visibility of the
target. The targets were attached to a flexible material base and it was this base

that was taped to the reference landmarks on the subjects.

Subject Recruitment and UCHRIS Appro val (University Committee on
Research In volving Human Subjects)

Since this research was supported in part by automotive companies, the

work was structured around the automotive environment. Typically the

37

automotive companies designed for accommodation from a small statured
person (5th percentile female) to those of a large stature (95th percentile male),
and the majority of the seat standards were tested with a manikin that
represented a 50th percentile male by weight and stature. For the studies
conducted for this research, 50th percentile men, or mid-sized males were
selected as test subjects, Table 4 (UCHRIS IRB #96-054)[j;he mid-size male
requirements” were derived from the National Health and Nutritional
Examination Survey of 1974 (NHANES ll)41 anthropometric study.

Table 4: Mid-Male Anthropometry

Hight Range 1727-1778 mm (68-70 inches) 2...
Weight Range 72.7-81.0kg (160-180 lbs.) =-

 

 

 

 

 

 

The potential subjects were recruited and brought into the laboratory to
verify their heights and weights prior to testing. At this time, the subjects were
asked several lifestyle questions such as if they had any back injuries, back pain
or back surgery. If the subject answered yes to any of the back pain or injury
questions, he was excluded from the pool of test subjects.

For this investigation, it was desirable to develop a correlation between
posture and force measures. Researchers have stated that people who have
back pain or injury are more likely to produce atypical spinal motion patterns.
Therefore, to decrease the measurement variability, only subjects without back
injury or pain were tested. Before testing, the protocol was reviewed with the
subject and he was asked to sign a consent form to grant permission to be

tested, interviewed and photographed.

38

Two sets of subjects were tested the first set had a sample size of 10
while the second set had a sample size of 13. After the data were analyzed for
the first 10 subjects it was determined that a larger sample size would be
beneficial for statistical analysis. The only change made to the test protocol
between the two subject sets was that the second set of subject had targets
placed on either side of the head, while the first set of subjects did not have head

targets.

Methods for Testing

Application of Targets
For all testing, subject positional data, chair positional data and force data

were collected simultaneously. To collect these data, the subjects were targeted
with light-weight, retro-reflective targets on various bony prominences. These
target sites were surface body points that were easily palpated and could be
located repeatedly.

The list of target locations can be found in Table 5 and the targets sites
can be seen in Figure 21, Figure 22, and Figure 23. The articulating chair was
also targeted to compare the position of the subject to the position of the chair; a

list of chair targets is located in Table 6.

39

Table 5: Targeted Body Landmarks

 

Seventh Cervical Vertebra — C7

 

Sternal Notch

 

Mid-Sternum

 

Left Anterior Superior lliac Spine (ASIS)

 

Right ASlS

 

First Sacral Vertebra (S1)

 

Right and left head at the center of gravity
locaflon

 

Right Acromion Process

 

Right Lateral Epiondyle on the Humerus-Elbow

 

Right Radial Epiconcyle-Wrist

 

Right Mid-Thigh

 

Right Lateral Condyle (Knee)

 

 

Right Lateral Maleolous

 

Table 6: Chair Targets

 

Two targets on the Recline Bar

 

Two targets on the Thoracic Support

 

Two targets on the Pelvis Support

 

 

 

Two targets on the Steering Wheel Base
Two targets on the Base of the Chair

 

 

40

 

Sternal

 

 
    
 
   
 
 

      

Acromion Notch
Process
Mid-
Sternum
Lateral
Epicondyle
Radial Left and Right
Epicondyle ASlS
Figure 21: Anterior target sites.
C7 5‘5
- \eé‘f§ ,-- Right and
23;? Left Head
// ’3‘
y
g:

I Sacrum Target

Figure 22: Posterior target sites.

41

Lateral
Maleolous

 

Figure 23: Leg targets.

The subjects were asked to wear tight fitting shorts and a tight fitting shirt,
or no shirt; once the subject changed into this clothing his height and weight were
measured and he was targeted. Using an alcohol pad, the target sites were
wiped free of any oils; then with double sided medical tape, the targets were
placed onto the landmarks listed in Table 5. In some cases, additional tape was

used on the base of the target to keep it securely attached to the skin.

Test Conditions and Setup
fine BAC was initially set in a mid-seat-height automotive package defined

by an in-depth study performed at General Motors”. The seat height was
located by placing the SAE manikin (not the new ASPECT manikin) in the chair

and measuring the vertical distance from the manikin’s H-point to the top of the

42

foot plate surface. The H-pt, which is the manikin’s representation of a human
hip joint center, was measured in the middle recline angle, 24° and was set at
280 mm relative to the surface of the foot plate, Figure 24. The thigh support
angle was adjusted to be approximately 14° as defined by the mid-thigh line on
the SAE manikin (15° for actual thigh support angle); this angle was
representative of a typical automotive seat pan angle. The angle of the foot rest
was set at 61° based on the 280 mm seat height from the General Motors

automobile interior package study”.

Steering Wheel

  

 

Wheel Height
61 5mm

 

  
  
  

H-Pt.

 

/‘

Pelvis Support

Buttocks Support

 

61k
(

230mm
Foot Plate

Figure 24: Set-up conditions for the BAC

The recline angles of the lab chair were set by placing the SAE manikin in
the test seat while it was configured in the neutral lumbar position (neutral
position: thorax support and pelvis support form a single plane) and measuring

the back angle of the manikin. (See section entitled Test Descriptions for more

43

detail). This technique, where the back angle of the manikin is also the seat
recline angle, is currently used in the automotive industry to obtain a recline
measurement. By using the SAE manikin, the test seat was adjusted to provide
the same recline as an automotive seat. From the GM study”, the average
recline angle for a typical mid-sized automotive package was found to be 24°.
For this study 20°, 24° and 28° recline angles were used, with the most upright
angle being 20° rearward of vertical.

The center of the thoracic support surface was adjusted to be
approximately at the ninth thoracic spinous processof the subjects and the pelvis
support was adjusted so the top of the support was located in the sacrum region
but did not interfere with the lumbar spine motion. Prior to testing, the subjects
were brought in for a height and weight screening to be sure they met the
requirements. At that time, measurements of the height of the ninth thoracic
vertebra (T9) were taken relative to a horizontal seat surface. The thorax support
and pelvis support were adjusted based on an average T9 height of the first ten
subjects, and remained in the same location for the testing of all subjects;

For all tests, the arms were positioned on a steering wheel in front of the
body with a load cell attached to the wheel. The subject was able to adjust the
wheel fore and aft for comfort. The wheel height (center of wheel to foot support)
was fixed at 615 mm. The subject was also allowed to adjust the foot plate fore

and aft for comfort.

44

The entire test time per subject varied between 2.5 hours and 3 hours
from start to finish. This included the informational session, targeting, and

testing.

Test Descriptions

Reference File
first, an initial position of the chair (without the subjects) was taken with

the back upright at 90 degrees relative to horizontal and the buttocks support at 0
degrees relative to horizontal and the thigh plate elevated 15 degrees relative to ,
horizontal. This chair position was used as the reference data file for both the

force plates and chaia
.../

Static Tests

The chair was placed in three different recline angles (20°, 24°, 28°),
based on the angles found in a typical mid-sized automotive interior package.
For each recline angle, four amounts of lumbar curvature were selected. The
order of the three recline angles was randomized, and within each recline, the
four lumbar curvature angles were randomized. The various lumbar positions
were created by adjusting the thorax and pelvis supports of the chair equal and
opposite amounts. These four chair positions induced four different postures:
slouched, neutral, erect and super-erect, Table 7. The neutral chair position,
Figure 25 was determined by aligning the thorax support and pelvis support; in
this condition it connected, the two supports would form a plane. For the

slouched posture, the bottom of the thorax support was rotated away from the

45

subject 5° from neutral and the top of the pelvis support was rotated away from

the subject 5° from neutral, Figure 25.

Table 7: Static Test Postures

 

 

 

 

 

 

 

Static Pastures - 5 second tests

Recline 1 Recline 2 Recline 3
-10°TSA— slumped -10° - slumped -10° - slumped
0° - neutral 0° - neutral 0° - neutral

10° - erect 10° - erect 10° - erect

20° - super erect 20° - super erect 20° - super erect

 

 

 

 

For the erect and super-erect conditions the thorax and pelvic supports
were rotated in the opposite direction from slouched. For the erect condition, the
bottom of the thorax support was rotated toward the subject 5° from neutral and
the top of the pelvis support was rotated toward the subject 5° from neutral,
Figure 25. For the super-erect condition, each of the segments were rotated 10°
in the same direction as the erect condition. For a descriptor of the test, the
thorax and pelvis support movements were added together to describe the Total
Support Angle (TSA): slouched -10°, neutral 0°, erect +10° and super-erect +20°,
Table 7.

When entering the chair, the subject was instructed to make sure the back
of their buttocks were in contact with the pelvis support, but not forcefully pushing
on the support plate. After each chair movement, in the static tests, the subject
was again asked to re-adjust and make sure their buttocks were in contact with
the support. This instruction was given to each subject to provide consistency in

how the pelvic support was loaded.

46

      
      

1 Support
Recline :
Bar ’r'
_~ Thigh
N Support ----- Pelvis
Support

Buttocks
Support

Figure 25: Slumped, neutral and erect positions of the test chair.

Preferred Pastures
Included in the static tests were a set of preferred postures. At three

different times during testing the subject was asked to choose two preferred
postures (Table 8). For trial one of the preferred posture test, the chair was
adjusted to an extreme upright position, then the subject was asked to adjust the
recline and lumbar curvature and find a comfortable position. For trial 2, the
chair was adjusted to an extreme reclined position, and again the subject was
asked to find a comfortable position. Trials 3 and 4 and trials 5 and 6 were
conducted in a similar fashion.

[The method of moving the chair to extreme positions ensured that the
subject would make chair adjustments and not choose the upright or reclined
extremes as comfortable positions. For the preferred posture, the subject was
asked to rest his hands in his lap, adjust the back of the chair until comfortable,
then adjust the foot plate and finally pull the steering wheel to a comfortable

position. The subjects were allowed to make changes after this sequence if

47

necessary. The goal was to have the subject adjust for back comfort without the

——1

influence of the wheel or foot plate.
/

Table 8: Comfort Postures

 

Comfort Posture
Trial duration: 5 seconds

 

Occurred before first set of static postures

 

Trial 1

 

Trial 2

 

 

Occurred after second set of static postures

 

Trial 3

 

Trial 4

 

 

Occurred after dynamic postures

 

Trial 5

 

 

Trial 6

 

 

Table 9: Dynamic Test Conditions

 

Dynamic Postures
Occurred after the static test positions

Trial duration: 10 seconds

 

Recline 1 Recline 2

Recline 3

 

Trial 1 -10° to +20° Trial 1 -10° to +20°

Trial 1 -10° to +20°

 

 

 

Trial 2 -10° to +20° Trial 2 ~10° to +20°

 

 

Trial 2 ~10° to +20°

 

Due to the large volume of data generated for this dissertation, the

preferred data will not be analyzed but will be considered work for the future.

Dynamic Tests

(\

For the dynamic test conditions, the seat was adjusted to move through

the full range of lumbar support positions in one continuous motion. The chair

was operated by electric motors, which allowed the chair to move smoothly while

the subject remained seated. For all of the dynamic tests, a test assistant, out of

view of the cameras, operated the chair movement. Two trials were performed at

48

each recline angle, one trial began in the neutral position and then moved to the
slumped position and then to the super erect condition. The second trial also
began in the neutral position but initially moved to the super erect position and
then to the slumped position.

The subject was placed in the starting position and instructed to adjust the
wheel and foot plate to a comfortable location. The subject was then moved
through the entire chair motion to see if further adjustment was necessary. The
subject was required to keep their hands on the wheel and their feet on the foot
plate for the entire dynamic test.

During the dynamic testing, the subject was instructed to relax and let the
seat move him through changes from lordosis to kyphosis. The dynamic tests

were repeated two times for each recline angle and lasted ten seconds (Table 9).

Hard Seat
After testing in the Biomechanically Articulating Chair (BAC), additional

targets were placed along the spine on the following spinous processes: C7, T6,
T10, T12, L3 and S1, Figure 26. The hard seat data file was collected to
establish a spinal reference position for comparison to the JOHN model. The
JOHN model identified 0° TLC (Total Lumbar Curvature) as a straight lumbar
spine without lordotic or kyphotic curvature. The hard seat was designed to
position the spine in this straight position. The subject’s openness angle was
measured in the hard seat and could be calibrated to JOHN’s 0° TLC position.
(For further information regarding the definition of the JOHN model, see the

Analysis Methods Section.)

49

 

Figure 26: Hard seat reference file.

50

Analysis Methods

This section, Analysis Methods, discusses posture computations, the
calculation of chair orientation, the processing of the force data and the
methodology developed for the estimation of the internal joint forces of the

lumbar region.

Motion Data
Prior to data collection, the motion system was calibrated using the

procedure and calibration stand as described by the Oualisys” motion
measurement system manual. Once calibrated, data collection could begin. The
system was calibrated at the start of each day and usually two subjects were
tested in one day.

After data collection, the motion data were tracked with a computer to
obtain x, y and z coordinates for each target. Using the Qualisys38 system to
track the motion data, the targets were viewed on the computer screen, and each
target was identified as either a specific chair target or a specific body landmark,
Figure 27. Using the calibration file, the software” performed a direct linear
transformation on each frame of data resulting in a set of three dimensional
coordinates. After the targets were named and tracked, the data were exported

for analysis.

51

Cu:

EI

Custom Data Analysis
The orientations of the chair and body consisted of sagittal plane angles.

The data was collected with the calibration identifying the Y axis as anterior to
the seated subject and the Z axis as superior, but was converted to be similar to
the buttock and thigh force plate orientations where the X axis of each plate was
pointing anteriorly. To accomplish the conversion, the motion data was rotated
90° about the Z axis making the X axis anterior, Y left lateral, and Z superior,

Equation 1.

X ' sin6 0050 O X
Y ' = cos6l —sinl9 0 Y where6 equaled 90°
Z ' O 0 1 Z

Equation 1: Rotation of the laboratory coordinate system about the Z axis.

Basic Vector Calculations

/ . .
After the raw data coordInates were transformed, the calculatIon of the

sagittal plane angles followed. First a vector was created between the two
targets located on a chair segment or a portion of the body. To establish a vector
C from target A (which is the position of point A relative to the laboratory) to
target B (which is the position of point B relative to the laboratory), vector A was
subtracted from vector B (Equation 2, a). To subtract vectors, the components of

each vector were subtracted (Equation 2, b).

52

 

 

' Arm
\/ Steering

Wheel
Recline Base
Thorax Bar
Support of (
Chair
—;—\
Pelvis Leg
Pelvis \
Support s
of Chair ‘-

0 Head

Chair Reference
Targets

+

 

 

Figure 27: Computer image of subject being tracked, side view.

c—O —. —.

C = B — A
where (a)
21’ = Axi + ij+ A212

Cx : Bx —Ax (b)

Equation 2: Computing a vector C from target A to target B.

Computation of angles between vectors requires that the vector be first
made into a unit vector, or a vector with a magnitude of one. To create a unit
vector, first the magnitude of the vector must be computed. The magnitude of

the vector was calculated using Equation 3.

53

 

Magnitude of vector 6 = 151: fixz + Cy2 + C22

Equation 3: Computing the magnitude of a vector.

Then, to create a unit vector, each component of the vector was divided by the
vector magnitude, Equation 3. This resulted in a unit, Equation 4, or directional
vector with a magnitude of one.

C,,C,,,CZ

c: -
ICI

 

Equation 4: Creating a unit vector.

Once a unit vector was developed, sagittal plane angles were found by
computing the dot product between two unit vectors and then obtaining the angle
by either computing the arccosine or the arcsine of the dot product. These
angles were computed with the unit vector and a lab coordinate unit vector, such
as lab 2 (vertical), or two unit vectors calculated from the body or chair targets.”
Posture Definitions - JOHN

The method of defining a subject’s posture was based on a model of the
body developed by Haas”, Boughner44, Bush45 and Hubbard”. In the JOHN
model, the thorax and pelvis were treated as rigid bodies with two joints (upper
and lower lumbar) and a lumbar link connecting them (Figure 28). With this
approach, the orientation of the torso was determined by: the position of the torso

in space, the position of the thorax relative to the pelvis and the recline of the

54

entire body in space. For the JOHN model, the orientation of the thorax relative :
to the pelvis was termed Total Lumbar Curvature (TLC) and the recline of the
body was termed Total Recline Angle (TRA). The TRA was measured by the
angle of the lumbar link relative to a vertical, Figure 28. The TLC was measured
by the amount of rotation of the thorax about the upper lumbar joint plus the
amount of rotation of the pelvis about the lower lumbar joint. Since the thorax
and pelvis were treated as rigid bodies, any two points picked consistently on the
thorax and on the pelvis could be used to identify the rotation of each of the body
segments, Figure 29. The initial reference position of the JOHN model was
chosen to be when the joints in the lumbar spine formed a straight line. Thisflwas
termed 0° TLC, Similar measures were developed for the subjects and are

discussed in the following section.

Thorax

  
  
   

Upper Lumbar Joint

Lumbar Link

Lower Lumbar Joint

Pelvis

Figure 28: JOHN model with TRA measure.

55

Thorax Rotation

< 0° TLC

  
 

Pelvic Rotation

Figure 29: Pelvis and thorax rotation about the upper and lower lumbar
joints.

Posture Measures - Sub/eats
The measurement of posture in the seated position is not a trivial issue.

Currently a standard definition of seated posture does not exist, the methods and
protocols developed for this dissertation were some of the first in the field of
seating mechanics and were implemented with other methods developed for the
ASPECT program (described in the literature survey section). These methods
use exterior body landmarks, and treat the thorax and pelvis as rigid bodies to
define posture.

A detailed discussion of the body measures is presented in the following
sections, but in brief, the human openness angle measure was calculated as an
angle between the thorax vector and the pelvis vector (similar to the JOHN model
TLC). The human recline angle was calculated by establishing a vector from the K

midpoint of the ASIS targets to the sternal notch target. The angle of this vector

with respect to the lab vertical vector was called the human recline angle.

56

However, this calculation may not be the optimum method for measuring the

body recline and will be evaluated and discussed in the Results section.

Measurement of Pelvis Orientation
To compute the angle of the pelvis in the sagittal plane, it was necessary

to calculate a point midway between the two Anterior Superior lliac Spine (ASIS)
targets on the pelvis. The midpoint was computed by first finding the magnitude
of the vector between the left and right ASIS targets, then adding half of the
magnitude to the vector identifying the right ASIS target, Equation 5. In cases
where the right ASIS target was missing, the magnitude between the two ASIS
targets could be determined from a reference file and the midpoint could be
computed solely from the left ASIS or right ASIS target (described below).

One concern with using targets was the introduction of data artifact by the
movement of targets due to skin motion. Since extra adipose tissue rests around
the mid-section of the body, the area of most concern for artifactual target motion
was the pelvis region. For this study, subjects were mid-sized males and of
medium weight, so there was little adipose tissue on any of the subjects.
Subjects also wore tight fitting clothing to reduce skin and clothing motion. When
possible, the targets were attached directly to the skin. Calculating the midpoint
of the two ASIS targets produced an average of the two target positions and was

less sensitive to the skin motion than that of a calculation using one target.

57

‘

Right ASIS Vector = R

Unit Vector from right ASIS to left ASIS = 1:1
Magnitude (between lefi and right ASIS) = ILRI

Midpoint = it + (21 *I—Lf—I)

Equation 5: Calculating the midpoint between two targets.

The orientation of the pelvis in the sagittal plane was determined by using
the sacral target and the midpoint of the two ASIS targets, Figure 30. A vector
from the sacrum to the midpoint of the ASIS targets was computed and the
orientation of this vector relative to lab horizontal determined the pelvis
orientation, Equation 6. Since the motion of the pelvis in the seated position was
small, and did not exceed 180° of movement, singularities in the computation did
not arise. However, since the vector from the sacrum to the mid-ASIS location
could align with the lab X axis, an arcsine function was used in conjunction with
the lab vertical axis rather than an arccosine function. This approach computes
the angle with respect to the lab horizontal, using the laboratory 2 axis (vertical)
to perform the calculation, thus the two will never be parallel. To maintain the
right hand rule convention, a negative sign was also used in the calculation,

Figure 30.

Pelvis Angle = - l * arcsin(13 0 Z) = - l * arcsin(Pz)
P2 = Z component of the unit Pelvis vector from the Sacrum to the mid ASIS

Z = Laboratory Z axis

Equation 6: Calculation of pelvis orientation in the sagittal plane.

58

    
  
 

Sacral

Target X

Right ASIS "" Positive Pelvis Angle
Left ASIS

Figure 30: Vector used to determine pelvis orientation.

Measurement of Thorax Orientation
The mid-sternum and sternal notch targets were used to establish a vector

to monitor thorax movement, Figure 32. The thorax and the pelvis vectors were
used together to measure the amounts of torso flexion or extension, also termed
as torso openness. The thorax orientation was determined by calculating the
angle between the vector from the mid—sternum to the sternal notch and the lab
vertical in the sagittal plane. The computation used an arcsine function to

determine whether the thorax angle was anterior or posterior of lab vertical.

Thorax Angle = arcsin( T o X) = arcsin(Tx)
Tx = X component of the unit vector from the mid - sternum to the sternal notch

X = Laboratory X axis

Equation 7: Calculation of thorax orientation in the sagittal plane.

59

Negative
Thorax
Angle

 

Figure 31: Sign convention of the thorax orientation.

Openness Angle
In the JOHN model, the motion of the thorax relative to the pelvis was

defined by the rotation of each segment relative to the lumbar link, and was
termed Total Lumbar Curvature. For the subjects, a similar measure that defined
the rotation of the thorax relative to the pelvis was termed the Openness Angle.
This angle was calculated by using the pelvis vector from the sacrum target to
the mid-ASIS point and the thorax vector from the mid-sternum to the sternal
notch. The angle between these two vectors was the openness angle, Figure 32,
Equation 8. As lordosis increased, the pelvis vector tipped downward and the
thorax vector rotated rearward causing the openness angle to increase. This
was used as a within subject measure to compare the change of a person’s

posture in response to the chair.

60

When used in conjunction with a set of data collected in the hardseat, a
calibration could be devised and comparisons could be made between the JOHN
model TLC and the subject’s openness angle. The subject was asked to sit with
their pelvis and back flat along the hardseat. Thus, the subject’s posture in the
hardseat was considered to have a flat lumbar spine and was chosen to be

equivalent to a TLC of 0° when represented by the JOHN model.

 

Figure 32: JOHN model used to display the openness angle.

Openness Angle = arccos(T o 13)
13 = unit vector from the sacrum to the mid - ASIS point

'1‘ = unit vector from the mid - sternum to the sternal notch
Equation 8: Calculation of the Openness Angle.

Body Recline
If the body is moving with the BAC from lordosis to kyphosis at a given

recline angle and the body recline measure was chosen so that it was

independent of the openness angle, then ideally, the body recline measure

61

should not vary, only the openness angle. However, it may be that with the
human body these two parameters are linked and cannot be considered
independently.

For this study, the body recline angle (BRA) was defined as a vector from
the mid-ASIS location to the sternal notch target. This was an initial best guess
at how to measure the body recline and proved to work well, with only a small

amount of movement in the BRA when the subject changed the openness angle.

Body Recline
Angle
(Negative)

  
  

Figure 33: Body Recline Angle (BRA).

Recline Angle = arcsin(1i o X) = arcsin(R x)
Rx = X component of the unit vector fi'om
the mid - ASIS location to the sternal notch

X = Laboratory X axis
Equation 9: Calculation of the Body Recline Angle.

62

Head Position
The position of the head was measured by placing two targets at the

approximate Center of Mass (CM), also known as the Center of Gravity, level on
either side of the head. These targets were placed 10 mm (0.4 inch) forward and
20 mm (0.8 inch) above the Tragion47. The midpoint of these two targets was
computed and was used as the head CM.

In many trials only one head target was visible by two cameras, this was
due to hair partially covering the target or a slight turning of the head away from
the cameras. An alternative method for computing the CM of the head was
devised for the case where only one target was visible to the cameras. For the
alternative method, a local coordinate system was developed on the thorax. A
vector (A) was computed from the mid-sternum to the sternal notch and a vector
(B) was computed from the sternal notch to the seventh cervical vertebra; a cross
product was calculated to develop vector (C), which was perpendicular to the
plane formed by (A) and (B). The plane formed by vectors (A) and (B)
represented a sagittal plane through the thorax. Since the driving task was a
sagittally symmetric task, if only one head target was available, the unit vector
(C) was used as the directional vector for the computation of the head CM,
Equation 10. The magnitude between the two head targets was determined from
another file that contained the two targets, either a trial file or the hard seat

reference file.

63

 

Figure 34: Calculation of Head CG when only one head target is visible.

[-4. = vector from the mid - sternum to the sternal notch

B = vector from the sternal notch to C7

(5 = unit vecto r (direction al vector) of (3

M = magnitude between th 6 two head targets (from reference file)

fl = Head Target Visible

, , M
CM from one head target = H i (C * '7')

Equation 10: Calculation of head CM from only one head target.

Chair Position
Theposition of the chair was described by a recline angle, and the

relationship of the thorax support to the pelvis support. A set of two targets were
placed on the chair recline bar, on the thorax support, on the pelvis support and

on the steering wheel base. By establishing a vector through each set of targets,

64

the angle of each support plate could be determined in the sagittal plane, Figure
35 and Figure 36.

The buttocks support and the support behind the pelvis were attached and
formed a ninety degree angle, only the pelvis support was targeted and its angle
computed with respect to vertical. The buttocks support orientation had the
same angle as the pelvis support, however the buttocks support angle was with
respect to the laboratory horizontal axis. The thigh support and the foot support
maintained a constant angle throughout the tests; these two support angles were
measured prior to testing. All angles were deemed positive or negative by using
the conventional right hand rule (X positive anterior and 2 positive superior).

A chair total support angle (TSA) which is similar to the subject openness
angle, was calculated by adding the angle of the thorax support relative to
horizontal and the angle of the buttocks plate relative to horizontal, Figure 37.
This identified a larger angle for an erect chair condition as compared to 3

~*-‘
I

slumped chair condition.

,’——4

 

Figure 35: Measurement of wheel angle.

65

Thorax

  
 
 
  

Angle
Recline
Bar Angle
Pelvis and
Buttock
AngIe Thigh Angle

Figure 36: Angle measurements for the test chair, a larger angle is
associated with a more erect chair position.

 
   
    
 

 
   
  

Thorax Angle Thorax Angle

Negative
Buttocks Angle ”(’2' ‘
Positive

Buttocks Angle

Figure 37: The Total Support Angle (T SA) of the test chair. A larger angle is
associated with a more erect chair position.

66

Force Data
The force plates were mounted so that the X axis of the foot, thigh and

buttocks plate pointed anteriorly relative to the subject, while on the thorax and
pelvis force plates, the X axis pointed inferiorly, Figure 38. The Z axis of the
plate was normal to the surface of the plate and the Y axis pointed left lateral.
On the wheel, the X axis pointed superiorly and away from the subject.

To determine the support forces under the body, an initial reference trial
(see Test Descriptions for details on reference position), with the chair unloaded,
was collected. To zero the force plate data, for the reference test only, the force
plate amplifiers were balanced. Because of the high capacity of the load cells,
the readings were close to zero, but not exactly.

To account for this initial load measured on the force plates without a
subject, the reference file forces were subtracted from each test. Since the chair
changed orientation for each trial (Figure 39), a transformation matrix was
developed to convert the components of the reference force values to the new
plate orientation (Figure 40) and then subtracted from the force values obtained

with the subject in the seat.

67

ii” /’I/
ill—t Z

 

 

 

 

 

 

 

___, x
'—4 Tz
4+x

 

 

 

Figure 38: Force plate orientations.

 

 

 

 

ll" \e
Ir In!

” :25

Figure 39: A comparison of the thorax and pelvis supports in a reference
position and a trial position.

 

 

 

 

 

68

 

 

 

 

 

 

 

7 ..
Z t
____' Z »
Z

T 7 Rotation from the

i unprimed system to the

primed system.
X (A negative rotation, -9)

 

-...

Figure 40: Rotation of the plate to recalculate the initial reference forces
--., prior to subtraction from the trial data.

Figure 40 is an example of the change of the plate orientation from the
reference file to a trial condition, (not including the recline angle of the entire
chair). First the locations of the plates for the initial reference trial were
computed, then the new positions of the plates from the trial data were
computed. From these data, a transformation matrix was developed. This
transformation matrix was then used to convert the force components of the

reference forces (unprimed system) to the new orientation (primed system).

F_,' 0036 0 —sin6 Fx
Fy' = O l 0 F

.V

F:' sin0 0 c036 F:

Equation 11: Rotation of the force plates from the reference position to the
trial position.

Once the force data from the reference file was converted to the new plate

orientation, these forces were either added to or subtracted from the forces

69

measured for the trial data. (The moments were computed with a separate set of
.l

’T"

strain gages in the load cells and the computation was performed about the
center of the plate, thus a transformation was not necessary for the moments.

For all tests, the Fy values, or lateral shear forces were low relative to the
other force components. The dominating forces were the normal forces, (F2) and
for some subjects there were substantial shear (Fx) forces primarily on the thorax
and buttocks. The dominating moment value was the moment about the y axis,
My.
Location of Resultant Forces

Another calculation performed with the force data was to determine the
location of the resultant force in the plane of the support plate. This position was
computed to evaluate how the location of the resultant force on the support
surface varied as the posture of the subject varied. For this analysis, the dynamic
files provided a continuous path of the resultant force, whereas the static files
only provided one point per trial. With the trace path from the dynamic files, the
distance of travel of the resultant force was easily identified. One application of
this analysis would be in the medical industry, specifically wheelchair design. A
test of this nature could determine if the chair movement is able to redistribute
the forces, or provide a large enough movement of the resultant forces to reduce

decubitus ulcers or pressure sores.

70

Buttocks: Location of Resultant
150 _ Dyanmic Trial 23drec2a
Anterior

A AA

-300 -200 -1 00 0 1 00
-50 .

 

400

v v

-150 Right Lateral

Figure 41: Example of resultant force travel on the buttock plate for a
dynamic condition, top view.

To determine the position of the resultant force, the forces and moments
were measured with the force plates, and then these values were used to
compute the X and Y distances relative to the center of the force plate. Since the
forces Fx, Fy, and F2, and the moments, Mx, My and M2 were measured by the
force plate, the only additional information needed to compute the location of the
resultant force was the Z distance from the top of the plate to the center of the
plate, which was provided by the manufacturer. The following equations were
used:

M, =(Fx *0)+(-Fy *2)+(F_, *Y)+T_,
My =(Fx*Z)+(Fy*O)+(—-Fz *X)+Ty

where Tx and Ty were the free surface torques

Equation 12: Equations used to determine the location of the resultant
force.

71

The force plates were mounted behind or under the body for the seated
tests, therefore it was not possible for the subject to generate a torque about the
X or Y axis and T)( and Ty were considered to be zero. With this information, and
rearranging the above equations, X and Y were determined. Once X and Y were
computed, the distance the resultant force traveled in the X or Y direction could
be computed, thus determining how much the movement of the chair shifted the

resultant force in either direction, Figure 41.

 

 

Y: M" +Fy*Z
F2 F2
X=E1‘*Z_My 1
F2 F.

Equation 13: Determining X and Y location of the resultant force on the
surface of the force plate.

Internal Joint Estimation
Lastly, the external support forces were used to estimate the internal

forces at a specific point in the lumbar spine. To estimate the internal forces at a
specific location, the seated subject must be in static equilibrium. To evaluate
the equilibrium state of the subject, the Fx and F2 forces of each force plate were
converted to the laboratory axis system and then the horizontal and vertical
components were summed. If the subject was in equilibrium, both the horizontal
and the vertical forces should sum to zero. As an example, subject 17 in the
neutral condition of recline 1 measured horizontal forces with a residual value of

-3.66 N (-0.82 lb), and the vertical forces summed to —727.46 N (163.47 lb).

72

Once the subject’s weight was subtracted from the vertical forces the remaining
force value was 2.34 N ( 0.5 lb). Note that neither the horizontal or vertical forces
summed to zero (see Results Section for full data listing). However, these values
were well within the measurement accuracy of the instrumentation and error was
introduced during the calculations of the support plate angle and the
transformation of the data into the laboratory axes system. Nevertheless, the
overall equilibrium results were deemed reasonable and further internal joint
analysis proceeded. Results of the force data in terms of an equilibrium analysis
are discussed in the Results section.

Force data collected for this research measured support loads exerted by
the body onto the chair, steering wheel or foot support. Equally as important as
quantifying the external support loads was the ability to estimate internal joint
loading from these external loads. For example, support force data could be
used to determine how the internal reaction moments and loads change at the
L5/S1 joint center as a person’s posture changes. Having reference data of this
nature could verify which postures reduce a specific type of loading on the
intervertebral disc, useful information for those with back injuries. Also, these
data are useful in the development of a kinetic model of a seated individual.

Andersson 48 estimated compressive loading of the discs in various seated
postures based on the measurement of spinal disc pressure, however he did not
measure or compute the shear forces or moments. Chaffin” also provided
internal moment estimates on the lumbar spine based on general lifting tasks. To

develop a perspective on the magnitude of the moments of a seated individual,

73

moments estimated during a lifting task will be used for comparison when the
results are presented.

From the external force data, it has been determined that the summation
of the forces for the subjects produced a condition of equilibrium, or one that was
within the measurement error of the instrumentation. Therefore, it was assumed
the following approach would provide reasonable results.

For this dissertation, the methodology for the estimation of joint loads at
the L5/S1 joint center were developed and tested on an individual subject. The
analysis method for determining the internal joint loading is a tedious process
and because a subject’s specific anthropometry is used in the calculations, the
analysis must be performed on an individual basis. As of yet, the analysis has
not been optimized for automatic data analysis of multiple subjects. For these
reasons, a sample set of data has been completed for this dissertation, but the
analysis of the entire subject pool using these methods is considered a future
step for this work.

Using an example subject throughout, the steps used to determine the
joint loads are outlined in this section. The desired outcome of the internal joint
load calculation was to estimate the forces and moments at the approximate
L5/S1 joint center. The internal force results for Subject 17 in the most upright
position (recline1) for all four torso articulations (slumped, neutral, erect and
super erect) are listed in the Results section of this dissertation. Subject 17 was
chosen because all of the target data were available for all four conditions and

the equilibrium results were reasonable.

74

First, the body was divided into anatomical regions. These regions were
based on common anatomical sections used by Dempster and Webb” and
included the head, neck, thorax, lumbar (also called the abdomen), pelvis, upper
arm, forearm and hand, Figure 42. For this analysis, the body segment masses
of the torso, the center of masses (CM) of the torso, and the location of the L5/S1
joint center were necessary. However, knowing the segment mass and CM
location for the legs was not necessary.

Several estimations were made to determine what type of loads were
occurring in the body. Since the center of mass (also known as the center of
gravity) and the mass of each body segment could not be measured easily on a
living subject, these data were estimated based on cadaver work completed by ”A g
other researchers”. Because these masses and locations of the CM were only
estimates, the joint forces and moments were highly susceptible to errors in

these estimations, particularly the moment calculations.

75

WM
Abdomen . _
Q’s/er”
L5/St Joint ’ /
Forces and ,
Moment /

Figure 42: Sectioning of body to determine the center of masses of each
section.

Calculation of Segment Mass as a Percent of Body Weight (%BW)
The mass for each of the body sections was defined based on data from

Webb Associates 49, Table 10. The section masses were estimated as a
percentage of the total body weight. For comparison, the results from Webb
were also checked against those from Dempster, Clauser and Morris 49, Table

11.

76

Table 10: Percentage distribution of total body weight accordin to
different segmentation plans (from Webb Associates, 1978) 4

 

 

 

 

 

 

 

 

 

 

 

 

 

Grouped Segments, % of Individual Segments, % of Example Subject
Total Body Weight Grouped Segments Weight (Subject 17)
Weight 74.5 kg
(164 lbs.)
Head and Neck = 8.4% Head = 73.8% 4.62 kg
Neck = 26.2% 1.64
Torso = 50.0% Thorax = 43.8% 16.33
Lumbar = 29.4 10.96
Pelvis = 26.8% 9.99
Total Arm = 5.1% Upper Arm = 54.9% 2.09
Forearm 33.3% 1.26
Hand = 11.8% 0.45
Total Leg = 15.7% Tthth = 63.7% 7.46
Shank = 27.4% 3.21
Foot = 8.9% 1.04

 

 

 

 

Table 11: Estimates of mass distributions (kg for various male and female
percentiles (a).

 

 

 

 

 

 

 

Male Female

5% T50%j 95% 5%J 50%195%
Hand 0.4 0.4 0.6 0.3 0.4 0.5
Forearm 0.9 1.2 1.6 0.7 1.0 1.4
Upper Arm 1.6 2.1 2.8 1.3 1.7 2.5
Head, Neck and Trunk 33.0 43.4 56.8 27.2 35.8 52.1
Arms, Head, and torso above 27.2 35.8 46.8 22.4 29.5 42.9
L5/S1 disc (b)
Upper Leg 5.7 7.4 9.7 4.7 6.2 8.9
Lower Leg 2.6 3.4 4.5 2.2 2.8 4.2
Foot 0.7 1.0 1.4 0.7 0.9 1.3
Body Weight 57.1 75.2 98.3 47.1 62.1 90.1

 

(a) Estimates are from Dempster (1955) as corrected for fluid loss by Clauser et al. (1969).

(b) Based on Morris et al. (1961)

(0) Based on National Health Survey, Weight, Height and Selected Body Dimension of Adults,
PHS Pub 1000, Series 11, No. 8 (1965).

Calculation of Location of Center of Mass within Each Segment

The calculation of the location of the center of mass within each body

segment was determined by various methods. For the head, two targets were

77

 

placed on either side of the head at the approximate location of the center of
mass (10mm forward and 20 mm above the Tragion). As described earlier, the
center point of these two targets was computed and used as the head CM.

For the upper arm, forearm and hand, the CM was determined as a
percentage of link lengths, Figure 43 and Figure 44. The length of the vector
between the shoulder (acromion) and elbow targets determined the link length of
the upper arm. The length of a vector between the elbow and wrist targets
determined the link length of the forearm. These link lengths were compared to
the link lengths obtained by calculating the lengths as a percentage of standing
height, Table 12. Since a target was not placed on the hand, this method could
not be used to determine the link length of the hand. The comparisons from the
upper arm and forearm were averaged to obtain a percentage of 104.5% and this
percentage was multiplied by the link length obtained from using the Drillis”
percentage of stature method. By combining the anthropometric measures made
with the targets and the link lengths based on a percentage of stature, the link

length of the hand was determined to be194.9 mm.

Table 12: Determining link lengths of arm”.

 

From Targets % of Stature Comparison
(Stature of
Subject 17 =1727 mm)

 

 

 

 

 

 

 

Shoulder to Elbow 324.6 321.25 108%
Elbow to Wrist 273.9 252.17 101%
Wrist to tip of 186.54
middle finger

 

78

 

 

 

t----<> .7. +v+

1‘ I ' 0.l86lt omen 0.I08H
{D
In
0
O

 

 

 

o BIBH
I

5

“R

l

l

1

i

:4

 

 

 

 

 

 

 

 

 

 

 

I
O
a
o‘ I 053OH
3 l l
s P (t--'
Id
0.28514
' i 053's}?
-ol lo— ’
()055H i52

FOOT BREAOTl-l
FOOT LENGTH

“—— .—

Figure 43: Body segment lengths expresses as a proportion of body
stature by Drillis and Contini (Roebuck, Kroemer, and Thomson, 1975) 49

After the link lengths of the arm were computed, the CM location could be

identified as a percentage of the length of the link using the reference data listed

in Figure 44 and Table 13.

79

 

 

   

Head, -
Neck and {0
Trunk .~ .
60.4%
. i
- u '1? .I'
‘E I" ‘5
43.6% $5 I
Arm ”if: 41' O
56.4% 1:1)!
‘I 1.2.3 a '
..i A,
43.0% g lI
Forearm ' ‘1': . . I
57.0% if I? l \ i,
-/' .
r- L 'l i I 4 '{f ’ 9 .r\5

494%, {SE // at”! 43. 3%) 1‘ .b

I '1
5863/. i m K i ”w”

 

5&5}

43.3% W

I
Leg t ,' f
56.7% I, i
42.9%
FOOt $-- I .44,
57.1% _. 5 1' W

Figure 44: Link boundaries and mass center locatiogns as a percentage of
link lengths (Dempster, 1955)”

 

80

Table 13: Distances to segment center of mass in centimeters”.

 

 

 

Male Female Percent
Link
Length
(a)
5% 150% 195% 5% l50% [95%
Wristto hand CM 6.7 7.0 7.4 6.1 6.4 6.7
ElbowtoforearmCM 11.0 11.7 12.3 9.9 10.4 11.0 43.0 %

Shoulder to upper arm CM 12.5 13.2 14.0 11.6 12.1 12.5 43.6 %
Hip to trunk, neck, head 18.1 19.3 22.5 16.7 17.9 19.1 36.6 %
CM (b)

 

 

 

 

 

Knee to upper leg CM 23.0 24.8 26.1 22.2 23.2 24.2 56.7 %
Ankle to lower leg CM 23.0 23.2 24.9 19.3 20.6 22.1 56.7 %
Heel to foot CM 10.6 11.4 12.3 9.4 10.3 11.1 42.9 %

 

(a) All dimensions are based on the Dempster percentage of link length estimates (1955).
(b) When in erect posture. Measured from hip to top of head

    
 
 

x3, y3, 23 .
....... 7
x2, y2, 22 ..............

43.0 °/o 56.4 °/o

Figure 45: Calculation of CM for the upper arm, forearm and hand.

To locate the center of mass of the upper arm, first a vector was created from
the elbow to the shoulder target, (S). The magnitude and the unit directional
vector were computed for (S). According to the information from Dempster”,
Figure 44, the location of the center of mass (CM) was 56.4% of the link length of
the upper arm from the elbow, Figure 45. A vector that had the direction of (S)
with the magnitude of 0.564 * (S) was added to the vector identifying the elbow

I

target, thus producing the CM location of the upper arm, Equation 14.

81

S = vector from the elbow target to the shoulder target

5‘ = unit vector from the elbow target to the shoulder target, also defines the direction of §
IS] = magnitude of vector 3:

E = vector from laboratory center to elbow target

CM of Upper Arm = E + 0564* |S|*.§

Equation 14: Locating the Center of Mass for the upper arm.

To identify the location of the center of mass for the forearm, a vector from
elbow to wrist (W) was created. The magnitude of (W) and the unit directional
vectors were computed. The CM of the forearm was located 43% of the forearm
link length from the elbow along the vector (W). Using the same methodology
used to locate the CM of the upper arm, the CM of the forearm was determined.

Once the link length from the wrist to the tip of the middle finger was
determined, the CM of the hand could be determined. Since a target was not
placed on any of the fingers, an assumption had to be made that the hand was in
line with the forearm therefore having the same direction. From this information,
a vector that was 49.4% of the hand link length in the same direction as the
forearm link was added to the vector defining the wrist target. This defined the
CM of the hand. It was also assumed that the hand CM was the point of the
steering wheel-body contact, and that the forces transmitted between the subject
and the wheel occurred through this point.

The method of defining the CM location based solely on link lengths could
not be used for the location of the neck, thorax, abdomen and pelvis CM. These

body sections were not as clearly defined by targets as the extremities, and both

82

inferior/superior and anterior/posterior coordinates were necessary to define
each body section. To estimate the location of the CM for these body sections,
the seated anthropometric mid-sized male dummy data50 were used, Table 14.

First, data points from the dummy had to be matched to the target data
from the subject. The target locations of C7 and S1 (from the subject) were
adjusted to account for the 10 mm radius of the target and moved the target point
to the skin. This was accounted for by creating a vector from C7 to the sternal
notch target and moving the location of the C7 target 10 mm inward along this
vector. The adjusted C7 target point was treated as the C7 skin point. The S1
target was also adjusted 10 mm inward along a vector from S1 to the mid-point
between the two ASIS targets. These two points were used in conjunction with
the dummy C7 and L5 points to calculate the translation coordinates.

The coordinates of the CM locations were translated from the mid-male
dummy data to the subject data. These data were translated using the adjusted
target landmarks (now skin points) as the original translation point. Translations
were developed from both the adjusted C7 and S1 targets of subject 17. The
translations developed in Table 15 were applied to the data points in Table 14,
moving the dummy points into the subject coordinate system. No rotation was
necessary since both the subject and the dummy data were positioned in a
sagittal plane, thus only translation was necessary to convert between two. Note
that in the dummy specifications, all of the centerline points had a zero
coordinate for the Y axis. However actual subject data showed that the pelvis

and thorax were not in the sagittal plane, there was a small rotation between the

83

thorax and pelvis. This small rotation is common within human data and was
ignored. Between targeting error, and the general human body composition, it Is
rare to find the thorax and pelvis in perfect alignment.

After the translation, the CM of the neck, thorax, abdomen and pelvis were
plotted, Figure 46. Figure 46 shows that there is an anterior/posterior shift in the
CM locations based on the two different methods (using either C7 or 81) for
calculating the CM targets, and a slight vertical shift for the S1 data points. This
vertical shift was due to the use of the L5 dummy point (S1 dummy point was not
available) to translate the S1 target of the subject.

A translation of the dummy L5/S1 point could have been used to define
the approximate L5/S1 joint center in the subject data. However, by using the
measurement of the subject’s pelvic width (ASIS to ASIS), a more accurate
estimate of the location of the L5/S1 joint center was available. Based on work

from Renyolds, 198151

, the location of the L5/S1 joint center was estimated as
26.4% of the pelvic width (PW) reanlvard of the ASIS, along a line from the ASIS
to S1 (pelvis x axis) and 12.6% of PW superior of the ASIS, along a line

perpendicular to pelvis x axis.

84

Table 14: Locations of body landmarks from mid-male crash test dummy
specifications”.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(mm) X Z
C7 — Skin -266 489
L5 — Skin -174 13
Head CM -179 646
Neck CM -195 . 515
Thorax CM -177 267
Abdomen CM ~85 110
Pelvis CM -74 17
T12/L1 - Joint -177 165
L5/S1 — Joint -89 39
HJC 0 0

 

Table 15: Translation coordinates between the dummy points and the
target points.

 

 

Target Locations X Y 2
(mm) (mm) (mm)

C7 -472.6 1 .8 962.4

S1 treated as L5 -406.4 27.4 497.3

 

 

 

Translation Points: Difference between
Dummy Data and Sub'ect Targets

 

 

 

 

 

 

 

C7 Dummy to -206.6 1.8 473.4
nget
L5 Dummy to -232.4 27.4 484.3
Target

 

Research from Seidl and Marchinda” showed that the accuracy of
defining internal pelvic points was increased by obtaining three anthropometric
measures of the pelvis and using these measures to estimate interior pelvic
points. For more accurate definitions of internal pelvic points, it is recommended
that future studies use the following three measures: pelvic width (left ASIS to
right ASIS), pelvic height (mid- ASIS to superior portion of pubic crest) and pelvic

depth, ASIS to PSIS (posterior superior iliac spine).

85

After the masses of each body segment, the location of the center of mass
and the location of L5/S1 were identified, the next step was the summation of the
vertical and horizontal forces and the summation of the moments in the y
direction about L5/S1, Equation 15 -Equation 17.

Finally, by taking the segment masses, the locations of the center of
masses, and the force data, a free body diagram of the body superior to the
L5/S1 estimated joint center was analyzed. Treating all of the body segments as
rigid, and requiring equilibrium, the internal forces and moments could be

determined.

CM estimated from C7
CM Estimated from S1

CM from head
targets
C7 Target
Sternal Notch and
A ‘_____Mid-Sternum
Estimated Thorax Estimated Thorax CM
CM from S1 from C7

Abdomen CM from C7 and S1

Estimated T12/L1 Joint
from C7 and S1

I, (A: . /

L5/S1 from subject measures

ASIS
Target

 

 

 

Estimated L5/Si Joint
from C7 and S1
81 target /'

 

 

PdWsCM

Figure 46: Subject 17, recline 1, neutral. Centers of Mass computed from
the C7 skin point and from the S1 skin point.

86 I

Zfl=0
Horizontal support through wheel + horizontal support through thorax support =
horizontal (shear) loading at a plane between the abdomen and pelvis

Equation 15: Summation of the horizontal forces, according to the
laboratory coordinate system.

ZQ=O

Head Weight + Neck Weight + Thorax Weight + Abdomen Weight

+ 2 * Upper Arm Weight + 2 * Forearm Weight + vertical force on hands through wheel
+ vertical force on throax through thorax support =

vertical loading at plane between the abdomen and pelvis

Equation 16: Summation of the vertical forces, according to the laboratory
coordinate system.

ZMyabout LS/Sl = 0

(F2 * x>Head + (F2 * ”Neck 4' (F2 * ”Thorax 4' (F2 * ”Abdomen + (2 * 1:z * x>UpperArm
'4' (2 * l:z * x)Forearm + (F2 * ”Wheel + (Fx * Z)Wheel + (F2 * ”Thorax Support

4’ (Fx * Z)Thorax Support + My wheel + My thorax support = ’M y about LS/Sl

Equation 17: Summation of the forces and moments about the L5/S1
estimated joint center.

Once the forces and moments were computed at the level of the
estimated L5/S1 joint center, the final step was to transform these forces from the
laboratory coordinate system to a more meaningful system. Since the subject
was seated in a reclined position, the data at L5/S1 joint were rotated to match
the orientation of the pelvis at each test position (slumped, neutral, erect and
super erect). This rotation was assumed to match the angle of the lower lumbar

spine and the sacrum region more accurately than the vertical and horizontal

87

orientation of the laboratory coordinate system. The angle of the pelvis was
measured as the angle of a vector from S1 to the mid-ASIS location with respect
to lab horizontal.

The results of the internal joint analysis are discussed in the section titled

Results and Analysis of Static Data.

88

Results, Analysis and Discussions of Static Data

At this time it is important to note that data similar to those presented in
this dissertation are unavailable from any other source. A study examining the
forces exerted on the seat back of an office chair”, listed data comparable to
those presented and discussed here. However, the study of the office chair53 did
not investigate forces supporting the buttocks, thighs, feet or hands during
postural changes, and only the position of the spine in unsupported seated
conditions was recorded. Further, an investigation measuring seated body
forces and full body posture concurrently has never been performed; and results
of the search done by this author indicate the data presented in this study are the
first of their kind.

The following section discusses the data in terms of the static tests and
analyzes the forces and body positions. The data in raw format are too
numerous to present; rather the raw data have been reduced and summarized so

meaningful conclusions can be made about the entire subject sample.

Test Chair Repeatability
Prior to comparing the force or motion data it was necessary to determine

if the chair was consistently positioned for the various test conditions. The test
assistants moved the chair to the various conditions (recline 1, 2, or 3 and
slumped, neutral, erect or super erect) based on markings positioned on the
chair. Because of inter-operator error, some variance was expected in the chair

positioning.

89

The objective when positioning the recline of the chair was for each of
these recline positions to be varied by four degrees: recline 1 was 20°, recline 2
was 24° and recline 3 was 28°. The thorax support was varied 5° between torso
articulations (slumped, neutral, erect and super erect) and the range from
slumped to super erect was 15°; the pelvis support also moved 5° between each
condition and had a range of 15° between slumped and super erect. Since the
thorax and pelvis movements were coupled, this resulted in a total chair
movement of 30° from slumped to super erect.

Table 1 and Table 2 present the average chair position for all the test
conditions and all of the subjectsl'TThe recline angle, on the average, reflected
the desired location with a standard deviation ranging from 0.8 to 1.3 degrees
between the three recline angles. The desired angles were 20°, 24°and 28° and
the averaged measured angles were 200°, 242° and 284° with the negative

value (seen in Table 1) indicating that the measurement was reanNard of vertical,
\
or reclined—Ll:

Table 16: Average recline measurement of the chair in degrees for each
condition (n=92). The calibration of the recline angle relative to vertical was
computed by subtracting the offset of the recline bar.

 

 

 

 

 

 

Repeatability of Chair .
Recline Recline Chair Recline
(degrees) Bar Bar Std” subtract
Measure offset“
Recline 1 -41.0 1.3 -20.0
Recline 2 -45.2 1.2 -24.2
Recline 3 -49.4 0.8 -28.4
*Std = average standard deviation across slumped, neutral, erect and super erect

conditions.
“21 degree offset between the recline bar and chair recline with respect to vertical.

90

Table 17: Average angle in degrees measurement of the thorax
support and the pelvis support for each recline angle (negative sign means
angle was rearward of lab vertical relative to a right-handed coordinate
system with x anterior and z superior).

 

 

 

 

 

 

 

 

 

 

 

 

Slumped
Thorax Thorax Pelvis Pelvis
(degrees) Std Std
Recline 1 -16.4 2.2 -23.4 1.2
Recline 2 -19.7 2.2 -28.0 1.0
Recline 3 -24.1 2.4 -32.7 1.4
Neutral
Recline 1 -20.6 0.9 —19.0 2.6
Recline 2 -24.7 0.7 -23.4 0.9
Recline 3 -28.8 0.8 -27.4 0.7
Erect
Recline 1 -26.2 1.2 -13.8 2.6
Recline 2 -30.6 0.9 -17.4 0.7
Recline 3 -35.0 2.8 -22.4 1.5
Super Erect
Recline 1 -31.0 1.2 -8.5 1.4
Recline 2 -35.6 2.2 -12.8 0.6
Recline 3 -39.5 2.2 -17.4 1.0

 

 

 

 

 

The data in Table 2 showed that on the average the thorax and pelvis
supports were each adjusted approximately 5° degrees between torso
articulations. However, the pelvis support angle showed a high standard
deviation (>2.0) for recline 1, neutral and erect conditions. The thorax support
angle also showed a high standard deviation for the slumped condition in all
recline angles, the erect condition in recline 3, and in the super erect condition for
recline 2 and 3. The data were re-examined and the single most extreme outlier

was removed from these data, Table 3. In the majority of cases (5 of 8) the

91

outlier came from subject 1 trials. Removing this one point reduced the standard
deviations in all cases without drastically changing the average. These outlier
measures could be due to partial target obstruction or data interpretation during
sections of missing targets.

When examining the entire body of data, it showed that the chair was
positioned in a repeatable fashion meeting the criteria that was outlined for
testing: recline angles of 20°, 24°and 28° and 5° changes of the thorax and pelvis
supports between the conditions designed to articulate the torso. Therefore the
chair was consistently adjusted between subjects and within trials, allowing

comparisons between the different conditions.

Table 18: Average angle measurement of the thorax support and the pelvis
support for each recline angle after removal of one outlier in the shaded

 

 

 

 

 

 

 

 

 

 

 

boxes.
Slumped
Thorax Thorax Pelvis Pelvis
(degrees) Std Std
Recline 1 -16.0 1.1 -23.4 1.2
Recline 2 -20.1 1.0 -28.0 1.0
Recline 3 -23.7 1.5 -32.7 1.4
Neutral
Recline 1 -20.6 0.9 -18.5 1.2
Recline 2 -24.7 0.7 -23.4 0.9
Recline 3 -28.8 0.8 -27.4 0.7
Erect
Recline 1 -26.2 1.2 -13.3 1.4
Recline 2 -30.6 0.9 -17.4 0.7
Recline 3 -34.4 0.8 -22.4 1.5
Super Erect
Recline 1 -31.0 1.2 -8.5 1.4
Recline 2 -35.2 0.4 -12.8 0.6
Recline 3 -39.1 1.4 -17.4 1.0

 

92

 

 

 

 

Force Data for Static Tests

Recline Angle Comparison for Primary Forces
The forces discussed in this section are called the primary forces and are

the maximum average support forces. These data include the normal forces (Fz)
under the feet, thighs and buttocks and behind the pelvis and thorax, Figure 1.
The shear force (Fx) was the largest force on the steering wheel and was the
most likely to be affected by postural change, therefore it was tabulated with the
normal forces from the other support plates.

Due to the body weight of the subject, all of the normal forces acted into
the force plates and these data are reported with a negative sign. Since the
force on the steering wheel was acting in a downward direction, due to the weight
of the arms, this force is also reported as a negative force. Figure 47 depicts the

positive directions of the primary forces acting on the force plates.

 

 

 

 

 

 

Figure 47: Primary forces represent the normal forces under all of the force
plates except for the steering wheel where the shear force, Fx, was the
maximum force.

93

Table 19: Primary forces for recline 1, 2 and 3 for each support plate.
Wheel is sheer load, Fx, while all others are normal loads, Fz).

 

 

 

\ {I/ (Newtons) Recline 1 Recline 2 Recline 3
Thigh -135.1 -129.5 -117.6
Buttocks -41 8.1 -405.3 -395.4
Pelvis -24.8 -31 .3 -34.1
Thorax -184.3 -204.1 -225.8
Foot -101.9 -97.2 -96.5
Wheel LFX) -32.7 -31.6 -33.1

 

 

 

fl

Force data from all of the subjects were first analyzed in terms of the three
recline angles, Table 4, where the data from the four conditions (slumped,
neutral, erect and super erect) were averaged for each recline angle, the
individual conditions are analyzed in a later section. Figure 48 through Figure 51
are graphical representations for each condition; within each graph, the force
data for the three recline angles are represented.

When evaluating the force data between the various recline angles for
statistically significant differences, the conditions (slumped, neutral, erect and
super erect) were lumped together and compared as one set of data for each
recline angleEOne Way Repeated Measures Analysis of Variance54 statistical
test was performed with these data. Both tests on data normality and variance
equality were performed. These data failed the normality test but passed the
equal variance testing, when this occurred, Friedman’s Repeated Measures
Analysis of Variance on Ranks was run. All statistical testing was performed at a
95% confidence level unless otherwise noted.” I

Table 20 shows that a significant differehce exists between all recline

angle combinations for the buttocks, pelvis and thorax when examining the

primary support forces. The thigh and feet exhibited significant differences for

94

two out of the three recline combinations and the wheel support loads (Fx)

showed a significant difference only between reclines 2 and 3.

Table 20: Testing for a significant statistical difference between the primary
support forces at given recline angles. Normal forces (Fz), except wheel
force, which is (Fx).

 

 

 

Fz Recline Recline Recline
1vs.2 2vs.3 1vs.3
Thigh - Y Y
Buttocks Y Y Y
Pelvis Y Y Y
Thorax Y Y Y
Feet Y - Y
Wheel (Fx) - Y -

 

 

L1: hese data support the conclusion that in the upright recline, the body
weight was distributed between the seat pan (under buttocks and thighs), the
footplate and the seat back (behind the thorax and pelvis). As the chair tipped
rearward, or the recline angle increased, the loading was redistributed, with more
of the load going into the seat back and less of the load into the seat pan and
foot plate. This conclusion was best seen by the fact that the average load under
the buttocks was reduced from -418.1 N in recline 1 to -395.4 N in recline 3
(negative sign indicated force into the plate) and the thorax support load was
increased from -183.4 N in recline 1 to -225.8 N in recline 3. As the recline angle
of the chair increased, the load under the thighs and feet also decreased by
17.5 N and 5.4 N respectively. Like the thorax, the load behind the pelvis also
increased by 9.3 N. The loading supported by the wheel stayed constant as the

recline angle changed and was confirmed by the lack of statistically significant

results]

95

Slumped: Mid-male Normal Forces
(Wheel Shear Force) n=23

   

 

 

:3 a
m _
9 — °
,5 E d‘.’ If 63. i
0 4 :.,\\< 9%
-100 ~ §
I g '200 1 §
§
\
l g '300 § lSlumped Reef
400 3% . Slumped RecZ
_ .0 519W”? “9‘”-
'509-

 

 

Figure 48: Primary support forces for the slumped condition at three
recline angles.

Neutral: Mid-male NormalForces
(Wheel Shear Force) n=23

ThIgh
Buttocks
; Pelvis
Thorax
Feet
Wheel

s

    
 

 

m
I:
§
é: -.. NeTItrel Ref“ ‘
thutrel ibcz
D New Recs .

 

 

 

Figure 49: Primary support forces for the neutral condition at three recline
angles.

96

Erect: Mid-male Normal Forces
(Wheel Shear Force) n=23

     

 

  

 

 

 

 

‘3 s -
U) Q)
.c '— L. e—v a)
.9 2 °
,5 E d‘.’ 15 ti” E
o ‘ ‘93 SL‘\‘.
. . § \ \
-100 a
0’ 1;; i -
» g -200 - ~
I
I Q _ a _,A z
I z 300 I Erect Rec1
400 + _ uErect Rec2 . _
D Erect R963 . .
-500

I
1.

Figure 50: Primary support forces for the erect condition at three recline
angles.

Super Erect: Mid-males Normal Forces
(Wheel Shear Force) n=23

  

 

    

 

 

‘3 a
m _
.C " s. .... 8
.9 o
,5 E 8 i5 I35 é
o~ . .\,
\x‘ s\\
‘ X X“
1 - . X \\
i 3 2007 ” ~ 44“
I ' \\
\\
1 § § z--- a
g -300 § lSuper Elect Rect
_400 , i; v ISuper Erect Recz
cistiper Erect Fbc3 .
'599,“ .... _ . _ 7T '

 

 

Figure 51 : Primary support forces for the super erect condition at three
recline angles.

Thus from these results, when determining the force distribution of a
person on a seat, it is important to take into account the recline angle of the seat.
The data from recline 3 (the most reclined position) would not be applicable to a

subject seated in an upright position such as in an office chair. However,

97

because the change in loading distribution is primarily affected by gravity,
interpolations could be developed to estimate the loading for many recline
angles.

The support data least affected by the recline angle is the steering wheel
loading. Since the subject was allowed to pick both a wheel distance and a foot
distance, this trend may be an artifact of the test protocol. In many automobiles,
the foot to wheel distance is fixed and either the preferred elbow angle or knee
angle must be compromised in the driving position. The newest generation of
automobiles offers an adjustable pedal position, and this configuration more
closely relates to the testing procedure implemented for these data.

The only published data pertaining to the measurement of seated support
forces was that by Faiks53 Faiks measured the forces behind the thorax, Figure
52, necessary to lift a relaxed person from a 20° reclined position to an upright
(0°) position. This test was performed on 22 subjects ranging from 5th percentile
stature to 95 "‘ percentile stature. The data for this dissertation began at a recline
angle of 20° and only involved subjects that were 50th percentile in stature. If we
assume the average data reported in the Faiks study is a fair representation of
the 50th percentile individual, then a direct comparison between the two data
sets can occur at the 20° recline angle.

The Faiks53 data reported an average of 25 pounds behind the thorax at
the 20° recline angle, which converts to 111.3 N and compares to 184.3 N
measured for this study. A difference of 73.0 N (16.4 lbs). Faiks also reported

that the force data exhibited a slope of 0.554 (0.554 lbs of force per degree of

98

recline), where the data for this dissertation had a slope of 5.1 N or 1.12 lbs per
degree of recline. It is clear that the two sets of thorax support data have
differing slopes, and that the data for this dissertation is an average of four
amounts of torso articulation while the Faiks data did not allow variations in torso
articulation. It may be that these factors, along with the possibility that the
anatomical regions supported in each study varied, cause the differences in the
thoracic support forces.

However, if the thorax and lumbar forces were added together for the 209
recline of the Faiks53 data, the total normal force would be 19.01 N, or 43.0 lbs. If
the average thorax and pelvis normal forces from recline 1 (209) of this study
were added together, the total force would be 209.0 N, or 47.0 lbs. Therefore, in
terms of total back support, the Faiks data and the data from this study are

similar.

Thoracic Support

40
E's-I'e

I
max

Force (lbs)

min

 

Degrees

Figure 52: Thoracic forces necessary to lift a person from a reclined
position to an upright position“.

99

 

Average Normal Forces for the Thorax

 

250

200 '/¢/R:cline B :28
Recline 2: 24

150 Reclme 1: 20

y = 5.193x + 80.081
R2 = 0.9993

 

 

 

 

 

 

Figure 53: Subject data for thorax support at three recline angles.
Recline 1: 20°, Recline 2: 24°, Recline 3: 28°.

Recline Angle Comparisons for Secondary Forces
The secondary forces (second largest in magnitude) are the shear forces

(Fx) acting in the superior to inferior direction on the thorax and pelvis and acting
in the posterior to anterior direction under the buttocks, thighs and feet, Figure
54. The secondary force for the wheel is the normal force Fz, this is the second
largest force for the wheel and represents the pushing or pulling of the wheel.
The shear forces summarized in Table 21 and Figure 56 through Figure 59 are
the forces that were measured by the force plates. There are two sets of shear
forces, there is a set that acts upon the chair, and then there is a set that is equal
in magnitude and opposite in direction that acts upon the body, Figure 55. The
shear forces discussed for this dissertation always refer to the forces acting upon
the chair. Thus a positive shear force measured by the buttocks support plate is

forward (anterior) as seen by the chair and rearward on the buttocks.

100

 

 

 

 

 

 

 

 

 

 

Figure 54: Secondary forces measured under the buttocks, thighs and feet
and behind the thorax and pelvis Fx and the wheel Fz (positive is In the
direction of the arrows).

 

 

 

 

 

 

 

 

,_=f

 

 

 

 

a

Figure 55: Shear forces measured were those acting on the support plates.
The forces acting on the body were equal in magnitude, but opposite in
direction.

101

Table 21; Secondary forces in Newtons. All support plates were the shear
/drces (Fx) except for the wheel, which was a normal force (Fz).
\

 

 

 

 

 

 

C Shear Forces in Newtons: Averaged Conditions
L.» Recline 1 Recline 2 Recline 3
Thigh 8.5 9.3 7.9
Buttocks -10.5 -15.9 -21.7
Pelvis 14.3 19.2 20.3
Thorax 5.4 5.3 6.9
Foot 47.3 45.4 45.1
Wheel (F2) 59 -5.0 -3.0
" sump... ml}... 5...... 5...}.
(Wheel Normal Force) n=23
.: g 2 2.5 .. g
a :2 E 8 s
80
40 I. W, -

Newton:

-40*

we} teat

    

”*1?

Figure 56: Secondary forces for the slumped condition comparing the three
reclines.

   

-80

 

I Slumped Rect
Slumped Rec2
, -9 S'umwi'fim ‘

 

420

102

 

Neutral: Mid-male Shear Forces
(Wheel Normal Force) n=23

 

   

 

 

=3 :6
m _
5 ~ E 25 a 8
.9 .c
,5 3 3 fi {’8 3
so
40 . , Y
I)
5
‘ i 0 *%r f?
INeutrel nae: '
ElNeutrel Recs ' .
-80 e , .-.e 2 , ... - swings—J I

 

Figure 57: Secondary forces for the neutral condition comparing the three
reclines.

Erect: Mid-male Shear Forces
(Wheel Normal Force) n=23

 

   

‘3 w it? o
.c ... - .. .-
5 a :7: E 3; é’
120
. IErectFlect
80 J IErectRec2
DErectRec3

gm- - - s - ,,- ,,
  .Rijfijfiilm- i—H

 

 

 

Figure 58: Secondary forces for the erect condition comparing the three
reclines.

103

Super Erect: Mid-males ShearForces
(Wheel Normal Force) n=23

 

 

 

Q
8 92 36 E
‘8, z 2 5 "' c”
g a 8 .5 E E
120 , ,
I Super Erect Reef
80 Super Erect ic2
g a Super Erect Reel!
0
1 E 40 k5 §5fi
I o
2
0 J
.40

 

Figure 59: Secondary forces for the super erect condition comparing the
three reclines.

Table 22: Testing for a significant statistical differences between secondary
support forces at given recline angles. Shear forces (Fx), except wheel,

 

 

(Fz).

Fx Recline Recline Recline

1 vs. 2 2vs. 3 1 vs. 3
Thigh - - -
Buttocks Y Y Y
Pelvis Y - Y
Thorax - - -
Feet - - -
Wheel (Fz) - - -

 

 

 

For the secondary forces, the posterior to anterior shear forces under the
buttocks, thighs and feet showed statistically significant differences for the
buttocks between all recline angles, Table 22, but not for the thighs or feet. The
superior to interior shear forces behind the thorax were not significantly different,
however the shear forces behind the pelvis were significantly different for recline
1 vs. 2 and recline 1 vs. 3 but not recline 2 vs. 3. The normal forces for the

wheel (push or pulling on the wheel) did not show any significant differences.

104

When the secondary force data were examined, Table 21 , the average
thigh forces were reported to be positive (forward on the chair) while the buttocks
average forces were reported as being negative (rearward on the chair).
However, this report may be misleading because the data were averaged across
conditions. Upon examination, the forces, particularly the thigh forces, varied
from positive to negative depending on the subject. Each condition showed high
standard deviations which crossed the zero line for the thighs, Figure 57 and
Figure 58 and for the buttocks and thorax Figure 58 and Figure 59.

Aside from the statistical analysis, trends were seen in the secondary forces
under the buttocks, behind the pelvis and on the wheel. As the chair reclined,
the shear force under the buttocks increased in magnitude and had a negative
direction (rearward on the chair, forward on the buttocks) for the slumped, neutral
and erect conditions, Figure 60. For the super erect condition, the forces on the
buttocks were in the opposite direction and decreased in magnitude with
increasing recline, Figure 61. The shear forces behind the pelvis were in the
positive direction (downward on the chair, upward on the pelvis) and increased in
magnitude as the recline increased for all test conditions. For the slumped,
neutral and erect conditions the wheel force decreased as the recline angle
increased, thus there was less pulling on the wheel as the chair reclined.

Overall, as the body reclined, the shear forces increased on the buttocks
and pelvis region demonstrating that this region was sliding into the seat. A

difference was seen in the super erect condition, most likely because the rotation

105

of the buttocks and pelvis supports overcame the effect of the recline angle and

created a “sliding out” effect.

Figure 60: Shear forces on the buttocks and pelvis region increase in
magnitude (negative direction) as the recline angle increases.

Figure 61: Shear forces on the buttocks and pelvis region for the super
erect condition, decrease in magnitude (positive direction) as recline
increases.

The thighs showed mixed responses to the varying recline angles. The shear
forces under the thighs were in the negative direction (reanlvard on the chair,
forward on the thigh) and decreased in magnitude for the slumped condition as
recline increased. The thigh shear forces in the neutral condition had a positive

direction and increased in magnitude from recline 1 to 3. The thigh shear forces

106

were also in the positive direction for the erect and super erect condition, but
decreased in magnitude from recline 1 to 3.

The shear forces for the thorax and feet did not show a trend between the
recline angles, however it is worthy to note that the shear force behind the thorax
was minimal, or the value closest to zero in the erect condition, for all three

recline angles.

Recline Angle Comparison for Lateral Forces
The side-to-side shear forces (Fy) were small for all test conditions, Table

23. Small lateral forces were expected, since the subjects were performing a

sagittally symmetric task and there was little to no lateral motion.

Table 23: Average lateral forces (Fy) and standard deviations. Each recline
Is the average of all test conditions (slumped, neutral, erect and super
erect).

Recline1 STD Recline2 STD Recline3 STD
(N) (3N; (N) (N) (N) (N)

 

 

 

 

Thigh -o.4 . -o.9 4.2 -o.7 5.0
Buttocks -1.5 6.3 -1.5 6.8 -o.4 7.0
Pelvis -1.1 2.2 -1.1 2.8 -0.8 2.1
Thorax -1.6 5.8 -1.3 6.8 -2.5 5.2
Foot 06 3.2 -0.8 3.2 -0.6 3.3
Wheel 1.7 1.5 1.7 1.5 1.8 1.4

 

 

 

 

For the lateral shear forces, which were small in value, only one significant
difference (Table 24) was found and that was for the thorax forces between the

second and third reclines.

107

Table 24: Testing for a significant statistical difference between the lateral
support forces at given recline angles. Lateral shear forces, Fy.

 

Fy Recline Recline Recline
1vs.2 2vs.3 1vs.3
Thigh - - -
Buttocks
Pelvis
Thorax
Feet - - -
Wheel - - -

 

.<

 

 

 

Overall, significant differences were demonstrated for the primary forces
(Fz) exerted under the subject’s thighs, buttocks, and feet, and behind the pelvis
and thorax for recline angles varying by four degrees or more. The primary force
on the wheel (Fx) demonstrated a significant difference for only one recline
comparison. The buttocks shear force (Fx), a secondary force, exhibited
statistically significant differences between all recline angles while the pelvis
showed significantly different secondary forces for two out of the three sets of
recline comparisons (1 vs 2 and 1 vs 3).

The measurement of shear forces in this type of seat may not be realistic
of the forces in a typical office or automotive seat, primarily because most office
and automotive seats do not have such large ranges of motion. Other factors
that could affect the shear forces are the amount of foam and fabric, and the type
of fabric on the seat and on the subject. A typical seat has several millimeters of
foam and suspension between the person and the seat structure, but for this test
seat only a few millimeters of fabric and foam were between the subject and a

rigid support. Also, frictional properties of fabric vary. Certain types of fabric

108

have higher frictional properties than others, such as cloth seats verses leather
seats. Thus, the type of fabric chosen for the subjects and for the seat would
affect the shear forces. Along with the variation in frictional properties, a large
seat pan angle could also cause the seat to grip a person’s pants, continually
placing an undesirable shear force on the buttocks. The shear forces measured
for this study provide initial measurements and a starting point for further shear

data analysis and gathering.

Chair Articulations: Primary Forces
The primary and secondary force data were also examined in terms of the

various chair articulations, slumped, neutral, erect and super erect, Table 25,
Table 26 and Figure 62. The forces on the buttocks, thighs, pelvis and feet
indicated clear trends; as the subjects moved from slumped to super erect, the
forces under the buttocks and thighs decreased, the forces behind the pelvis
decreased, and the forces under the feet increased, Table 25. A clear trend was
not seen in the support forces of the thorax. The forces into the wheel
maintained a constant level for all conditions across all reclines. These trends
confirmed that as a subject was moved into a more erect posture, there was an
increased loading into the floor and a reduced loading into the seat and no

change in wheel forces.

109

Table 25: Primary forces in Newtons for recline 1, 2 and 3 at all support
plate locations.

 

 

 

 

 

Recline 1
Slumped Neutral Erect Super
Erect
Thigh -136.8 -140.2 -135.9 -127.7
Buttocks -432.4 -422.5 -41 4.1 -403.5
Pelvis -27.9 -24.7 -24.1 -22.5
Thorax -183.0 -187.4 -182.1 -184.6
Foot -93.4 -97.9 -104.2 -112.1
Wheel (Fx) 328 -32.3 -33.3 -32.3

Recline 2
Thigh -133.8 -129.5 -128.2 -126.6
Buttocks -418.4 -41 1 .7 -400.3 -390.7
Pelvis -35.6 -31.7 -28.5 -29.5
Thorax 2048 -209.6 -204.3 -1 97.5
Foot -86.3 -93.1 -102.4 -107.0
Wheel (Fx) -31.9 -31.0 -31.8 -31.8

Recline 3
Thigh -127.5 -122.0 -118.0 -103.2
Buttocks -408.6 -400.2 -390.2 -382.5
Pelvis -39.1 -33.3 -33.4 -30.7
Thorax 2300 -229.0 -222.7 -221 .6
Foot -83.3 -94.1 -99.8 -108.9
Wheel (Fx) -32.6 -33.7 -32.9 -33.1

 

 

When evaluating these data for statistically significant differences, the
conditions (slumped, neutral, erect and super erect) were compared to each
other for a given recline angle, Table 26. A One Way Repeated Measures
Analysis of Variance statistical test was performed with these data. Both a test
on data normality and variance equality was performed. These data passed both
the normality and the equal variance test.

When evaluating primary forces, Table 26, for significant differences between

chair articulations, one interesting conclusion was that the thorax and wheel

110

support plates showed no significant differences between conditions. A
consistent difference was seen between the slumped and sugr erect conditions
for the buttocks, feet and pelvis for all recline angles. A significant difference was
also seen between the neutral and super erect conditions for the buttocks and
feet for all three recline angles. The statistical analysis confirmed the apparent
trends reported. Forces shifted from the buttocks-thigh-pelvis region to the foot
support as the chair moved from a slumped or neutral position to a super erect
position.

In simple terms, a trade off occurred between the buttock - thigh - pelvis
forces and the forces into the footplate. As the chair moved from a slumped
position to a more erect position, more support was needed through the footplate
and less through the buttocks, thighs and pelvis to support the body. A
surprising conclusion was that the thorax support forces did not exhibit a clear
trend during this chair movement. lntuitively one would hypothesize that as a
person moves from a slouched position to an erect position, the support force

behind the thorax would increase; however this trend was not seen in these data.

111

Table 26: Comparison of primary support forces between chair
articulations. Slumped (S), Neutral (N), Erect (E), Super Erect (SE).

 

Normal Support Forces (Fz) except for Wheel (Fx - shear)

 

 

 

 

 

 

Recline 1
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh - - - - - -
Buttocks Y Y - Y - -
Pelvis Y - - Y - -
Thorax - - - - - -
Feet Y Y - Y - -
Wheel (Fx) - - - - - -
Recline 2
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh — - - - - -
Buttocks Y Y - - - -
Pelvis Y - - - - -
Thorax - - - - - -
Feet Y Y — Y Y Y
Wheel (Fx) - - - - - -
Recline 3
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh Y Y - — - -
Buttocks Y Y - Y - -
Pelvis Y - - - - -
Thorax - - - - - -
Feet Y Y - Y - Y

Wheel (Fx)

 

112

 

Recline 1: Mid-males Normal Forces

 

    

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

(Wheel Shear Force) n=23
.C - S o-o
.9 == 3 ° 92’
g 5 8 l5 L? 3
0 ..
I -100 1* W -----
l g 7 i 7 i W
I o ‘200’ * ISlumped
’ E -300 e INeutral
z DErect
40° 7‘ 7 7 “7 1T7 77 7 7 7 7 77 ‘ uSuper Erect
500 J
Recline 2: Mid-male Normal Forces
(Wheel Shear Force) n=23
3
8 «a 36 g
‘8 z 2 h "’
'5 a 6‘: 55 I? E
0 Q
N.
100 § "
' N
.. 8
I §-2004~ E, _-
I z -300 7
400 I, , 7* ¥ I
I. ”39575?“
-5007 , , 7 , , 7 , ,. 7 (fl
Recline 3: Mid-male Normal Forces
(Wheel Shear Force) n=23
rn
x
8 2 g g;
«C >
O
E E E .5 E E
0 . l
l
-100 . ,
, 2 -200 - e -
o .‘A 7, . ..
l 3300‘ f , V, 7* I ISlumped _
. INeutral .
I 400.5... —h ¥ ,- ~ 4, DErecl f ...I
' nSuper Erect ‘
750° " 77 77

Figure 62: Primary seated support forces in Newtons. Comparison of chair
articulations

113

In summary, the primary forces associated with torso articulation showed
that statistical differences occurred if the chair’s Total Support Angle (TSA) was
at least 20°. Significant differences were not found between two contiguous chair
positions, but rather between every other condition, i.e. slumped to erect. So,
larger chair motions led to significant differences in the primary forces.

The support forces measured by the footplate were also found to be
significant. Supporting the feet during tasks was important for the ability to
change posture. Petite people most commonly have difficulty choosing their
desired posture because of the inability to obtain support from the chair, or
maintain contact with the floor. In the study of four office chairs by Bushss, petite
subjects were unable to maintain full contact with the seat back in one of the test
chairs. As a result, to maintain floor contact, the buttocks and pelvis were pulled
away from the seat back sacrificing support along the lower back. The
demonstration of the significance of the support force from the floor, or foot rest
showed that in future studies the forces under the feet should be measured to be
assured important force information is not lost.

From a medical perspective, individuals in wheel chairs are at high risks of
developing decubitis ulcers under the buttocks 56' 57. A goal of a medical
professional working with individuals in wheelchairs is to reduce the chances of
the patient developing these ulcers. These data show that a shift in posture
plays a significant roll in changing the loading under the buttocks. To reduce the
loading, the person should be placed in an erect posture; from a slouched

position to an erect position the loading under the buttocks is reduced by 4% of

114

the subjects body weight. For a 170 lb (77 kg) subject this translates into
approximately 6 lbs. (3 kg). An erect posture also has other physiological
benefits, as discussed in the Literature Review Section, of increased ability to
breathe, leading to increased oxygen content in the blood thus reducing muscle
fatigue. Possibly a preferable solution would be to continuously articulate the
individual so that the force distribution is not concentrated in one location but
continuously shifting. Similar benefits exist for users of other chairs, such as

office or automotive seating.

Chair Articulations: Secondary Forces
The secondary forces seen in Table 27, showed a change in direction for

the forces under the buttocks and thighs (negative to positive, or rearward on
chair to forward on chair) as the chair moved from a slumped condition to a more
erect position, Figure 63. This trend was seen across all three recline angles.
The secondary forces on the thorax exhibited a trend opposite to the buttocks
and thighs; the thorax forces moved from a positive value (downward on seat
back) to a negative value (upward on seat back) as the chair position became
more erect, Figure 63. The pelvis and feet secondary forces were all in the
positive direction (downward on the pelvis support and forward on the footplate)
and increased in magnitude as the chair moved from a slumped to erect position.
The wheel secondary forces all begin in the negative direction and decreased in
magnitude; for reclines 2 and 3 the forces switched direction and became a
positive value as the chair moved toward a more erect position. In general the

secondary forces on the wheel indicated less pulling on the wheel as the chair

115

moved to a more erect condition. This may be caused by the chair motion, since
the subject’s torso moved closer to the wheel as the chair moved from the

slouched condition to the super erect condition.

Table 27: Secondary forces in Newtons (Fx) for the thigh, buttocks, pelvis,
thorax and foot and F2 for the wheel.

 

 

 

 

Recline 1
Newtons Slumped Neutral Erect Super
Erect
Thigh -4.6 3.7 13.0 21.8
Buttocks ~39.9 ~18.0 -0.7 16.5
Pelvis 12.7 13.4 15.0 16.0
Thorax 19.9 9.3 0.2 ~8.0
Foot 38.1 42.5 51.4 57.3
Wheel (F2) -8.4 90 32 -3.0
Recline 2
Thigh ~1.3 6.3 13.4 18.9
Buttocks -40.6 -24.6 57 7.4
Pelvis 17.0 18.5 18.8 22.5
Thorax 22.2 9.3 07 ~94
Foot 32.6 41.9 49.9 57.3
Wheel (Fz) -9.4 84 29 0.5
Recline 3
Thigh 04 9.0 9.8 13.4
Buttocks ~50.6 ~31 .0 -9.2 4.2
Pelvis 18.9 19.2 21.8 21.1
Thorax 20.8 9.3 1.5 -4.1
Foot 31.5 42.3 49.6 57.1
Wheel 96 ~46 01 2.3

 

 

116

 

 

Figure 63: A comparison of the shear force directions between the Super
Erect and Slumped chair positions.

A One Way Repeated Measures Analysis of Variance statistical test was
performed with the secondary forces between chair articulations, Table 28. For
this analysis each recline angle was analyzed separately. Both tests on data
normality and variance equality were performed. All secondary force data
passed the equal variance test and all of the support force data except that on
the pelvis passed the normality test.

Significant differences were not found between any conditions for any
recline angles when examining the pelvis shear force data.

All secondary support data, except for the pelvis, showed a significant
difference for all three recline angles between the slumped and super erect chair

positions. A significant difference was seen in Recline 1 and 2 for all support

117

data (except for the pelvis) between the W conditions.
Recline 3 also showed a significant difference except for the thigh and pelvis
data between the neptral and super erect conditions. Reclines 1 and 3 showed a
significant difference between the slumgd and erect condition for all secondary
support data except for the pelvis while Recline 2 showed a significant difference
for all support data except for the wheel and pelvis. Table 27 lists these findings
as can be seen by the shaded regions.

Again, as with the primary support force data, significant differences in the
secondary force data were not found between two contiguous chair positions, but
rather between every other condition, i.e. slumped to erect, or slumped to super
erect, or neutral to super erect. The trends for the secondary forces as seen in
Figure 63 showed that in the slumped chair position the body was being forced
into the seat while the opposite was true for the super erect position. For the
super erect condition the shear forces lifted the thorax, thus extending the lumbar
region. From these data the one would expect that internal loading in the lumbar
spine would have a higher compression force in the slumped condition. This will
hypothesis will be compared in the during the internal joint analysis from this
dissertation.

The next question to be answered is whether these types of shear forces
are seen in other types of seats. When evaluating automotive seats, most likely
the shear forces measured for this study would be larger than those in a typical
automotive seat, primarily because today’s auto seats do not have large ranges

of motion. However, one current office seat (LEAP55) developed by Steelcase

118

does exhibit ranges of motion comparable to the BAC, but shear force data were
unavailable .

The seat pan on the test seat also contained separate thigh and buttocks
plates that did not move together. Since the shear forces are strongly affected
by the orientation of the support plate in space, and the fact that the buttocks and
thigh plates do not move together most likely causes an increase in the shear
forces. The thigh plate remains stationary while the buttocks plate rotates to

accommodate the various postures.

Table 28: Comparison secondary support forces between chair
articulations. Slumped (S), Neutral (N), Erect (E), Super Erect (SE).

 

 

 

 

 

 

 

Shear Support Forces (Fx) except for Wheel (Fz - Normal)
Recline 1
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh Y Y Y Y Y Y
Buttocks Y Y Y Y Y Y
Pelvis - ~ ~ — - -
Thorax Y Y Y Y Y Y
Feet Y Y - Y ~ ~
Wheel (F2) Y Y - Y Y -
Shear Support Forces (Fx) except for Wheel (Fz - Normal)
Recline 2
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh Y Y - Y - Y
Buttocks Y Y Y Y Y Y
Pelvis - - — - - -
Thorax Y Y - Y Y -
Feet Y Y — Y - Y
Wheel (Fz) Y Y - - - -
Recline 3
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh Y - - Y - Y
Buttocks Y Y Y Y Y Y
Pelvis - - - - - -
Thorax Y Y ~ Y Y Y
Feet Y Y - Y - Y
Wheel (Fz) Y Y - Y - -

 

 

119

Support Forces Calculated as Percent Body Weight
The support forces were converted to percent body weight (%BW) for

each subject and then averaged. Normalizing the data in terms of %BW could
establish a method for estimating support loads for any size subject for any
region of the body. By representing the support forces in terms %BW, and with
further investigations of other sized individuals, a pattern may arise that allows
the loading of a seat to be estimated with reasonable accuracy just by obtaining
the subject’s weight and applying a %BW distribution pattern. For the research
reported in this dissertation, this prediction can only be validated for a mid-sized
male subject.

Table 29 lists the average %BW (average of all subjects for all conditions:
slumped, neutral, erect and super erect) for each recline angle with respect to
each support plate. Table 30 and Figure 64 list in tabular and graphical forms
the %BW for each condition within each recline angle. From a macroscopic
viewpoint, a seated subject with the hands on a support device distributes
approximately 70%BW on the seat pan, 30°/68W on the seat back, 10°/68W
under the feet and 4%BW on the support under the hands. Note that the %BW
distribution listed above does not add to 100% because it is not a list of the
vertical force components but rather a combination of vertical and horizontal
force components.

As seen in the statistics of these data, a significant difference between the
postural changes (slumped to super erect, neutral to super erect and slumped to
erect) exists, so the above macroscopic distribution is not universally applicable.

Thus when considering primary and secondary support loads, posture must also

120

be a factor in predicting the loading scenario specifically for the regions under the

buttocks, behind the pelvis and under the feet.

Table 29: Primary forces represented in percent of total body weight (%BW)
for each condition. All forces are normal forces (Fz), except for the wheel
which is a shear force (Fx).

 

 

Average %BW: Average of all conditions.
Recline 1 Recline 2 Recline 3
Thigh 18 17 16
Buttocks 56 54 53
Pelvis 3 4 5
Thorax 25 27 30
Foot 14 13 13
Wheel 4 4 4

 

 

 

 

Representing the data in terms of O/on also mutes the differences seen in
the numerical data. For example the primary pelvis forces show a statistically
significant difference between slumped to super erect, yet the forces only vary by
1%BW. Similarly, when reported in terms of %BW the primary foot forces only
vary 2 to 4%BW between conditions and the primary wheel support forces show

no variation between conditions.

121

Table 30: Primary forces represented in percent of total body weight (%BW)
for each condition. All forces are normal forces (Fz), except for the wheel,
which is a shear force (Fx).

 

 

 

 

 

Recline 1
Slumped Neutral Erect Super
(%BW) Erect

Thigh 18 19 18 17
Buttocks 58 57 56 54
Pelvis 4 3 3 3
Thorax 25 25 24 25
Foot 13 13 14 15
Wheel (Fx) 4 4 4 4

Recline 2
Thigh 18 17 17 17
Buttocks 56 55 54 52
Pelvis 5 4 4 4
Thorax 27 28 27 27
Foot 12 12 14 14
Wheel (Fx) 4 4 4 4

Recline 3
Thigh 17 16 16 14
Buttocks 55 54 52 51
Pelvis 5 4 5 4
Thorax 31 31 30 30
Foot 11 13 13 15
Wheel (Fx) 4 4 4 4

 

 

122

Recline 1: Mid-males Normal Forces
(Wheel Shear Force) n=23

  

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5 a
m _
.c S a "' 8
.9 .c
,5 E 3 E L? 3
E 0 \'
510‘
I 3 -204 8
§
> ~30 ~ §
.40 -- § 1 INeutral
I "' -50 A s , \\\ ,, 7 *VElErect
‘g‘GW " ‘ 7....- ' ’ ' ’ ’ "J DSuperErect
_-70 _ ~_ 5 '7 ‘
Recline 2: Mid-male Normal Forces
0, (Wheel Sheer Force) n=23
x
.5 g
,5 co
0
‘15 ~10 ..
I .9
o - -.,
I; 2°
I ’1 ~30 15 i a ,.
E '40 ‘r — , , ., _ 1‘ Q
I“ I. -.-E- .--- .-l
g '50 g I DErect
'60 " "" ‘ " 7. B Super Erect,
:70 _ _ _ _ __ _ _. .7757"
Recline 3: Mid-male Normal Forces
(Wheel Shear Force) n=23
co
.2
.c g g a "‘ -
g1 ‘5 - g 8
’— CD 0. L],
I 0 7‘ ‘
:7? 7101 I
I g '20 1
l >~ -30 .5 E
§ -40 ~ ~—
' '501 ' ’7 7' " " " ‘nErect g
I 601"“ ‘#'"*‘ "M" "5’ ”I ‘nSuperErectI (
l " " " j

 

 

Figure 64: Primary support forces in terms of Percent Body Weight (%BW).

123

Static Equilibrium
Applying the principal of equilibrium to segments of the human body is not

as simple as applying this principal to a non-living object, such as a pencil resting
on a table. The body has several forces acting internally including abdominal
pressure, passive tissue tension, and active muscular contraction along with fluid
flow, and volume changes which affect mass distribution.

Examining the body from a micro-mechanics perspective produces an
extremely complex model needing numerous measurements and model inputs,
most of which are not easily measured in a living person nor are they available
via published data. Backing away from the micro-mechanics perspective to a
macro-mechanics viewpoint, the body can be treated as a rigid body, or series of
rigid bodies resting in a seat. The macrosc0pic viewpoint ignores all of the
internal happenings yet allows a reasonable analysis of the support forces. The
support forces can be measured and summed to estimate if they balance in an
equilibrium sense.

Therefore, if a subject was relaxed and was allowing the support plates to
hold him in position, the forces measured by the load cells should produce
results that are close to equilibrium. In other words, by breaking the forces into
horizontal and vertical components relative to the lab space, the horizontal forces
should sum to near zero and the vertical forces should sum to the subject’s body
weight. For the static trials, the force data were separated into components and

summed to obtain the equilibrium data presented in Table 31 through Table 34.

124

As can be seen in Table 31, the summed horizontal forces for many of the
subject trials were close to, but not exactly zero. It should be noted that there
were several factors that affected these data and small errors could have been
introduced into the system from any of the following: the identification of the chair
position measured by the targets, the tracking of the targets, the calibration and
zeroing of the force plates, the accuracy of the force plates and the conversion of
these data into horizontal and vertical forces.

The biggest influence on the force data was the force plate accuracy. In
the isolated case where there would be only vertical loading (Fz) the cross talk is
minimized to i 0.2% of full scale, or for a thousand pound plate the error would
be approximately i 2 pounds or i 9 Newtons. This error would increase if there
was a high shear loading or high moment on one of the other force plate
channels that could increase the cross talk and affect the normal force reading.

Table 32 shows the average horizontal forces for all the subjects in each
condition. On the average, the summed horizontal forces approached zero as
the subject moved to a more erect posture. This finding was consistent across
all three recline angles. The question arises, why would the equilibrium
approach zero as the subject’s posture is moved to a more erect condition,
provided that all other errors are consistent? It may be that the subject felt most
relaxed in the super erect condition; however from the testing experience and the
judgement of this author this is probably not the case. The super erect chair

position seemed to be the least tolerated by the subjects. One theory is that in

125

this position, the body was approaching a position closer to an anatomical

neutral position, causing the body segments to be “in balance” for this condition.

Table 31 : Horizontal forces summed for recline 1, showing data for all

 

 

subjects.
Recline 1: Horizontal Forces (Newtons)

Subject Slumped Neutral Erect Super
Erect

1 27.58 26.74 14.06 8.12
2 24.10 18.67 14.46 7.47
3 30.07 4.74 012 -2.26
4 19.17 15.70 12.58 8.75
5 12.98 3.29 0.54 -0.03
6 12.28 1.85 0.32 -6.79
7 20.16 18.44 13.93 2.21
8 17.77 10.89 2.04 085
9 4.12 4.85 4.53 0.43
10 3.26 19.61 10.78 -3.26
11 2.80 228 ~1.99 911
12 3.87 10.30 1.66 -7.14
13 4.74 0.01 ~12.26 -6.89
14 17.16 8.68 4.54 ~7.93
15 3.95 602 68.58 -12.57
16 -4.98 ~10.13 ~13.13 ~14.92
17 038 366 ~12.21 ~14.80
18 -37.62 45.28 4.55 2.37
19 12.74 6.35 3.84 -0.65
20 2.48 3.61 298 -0.48
21 14.14 0.50 ~2.27 -4.09
22 8.30 7.51 -3.29 -13.24
23 26.55 21.51 9.45 0.45

 

 

 

Table 32: Average horizontal forces summed for recline 1, 2 and 3.

 

 

 

(Newtons)
Slumped Neutral Erect Super
Erect
Recline 1 Ave 9.79 8.97 5.11 -3.27
Std 14.17 12.29 16.03 6.95
Recline 2 Ave 16.30 12.11 4.29 0.98
Std 10.91 12.90 8.23 9.49
Recline 3 Ave 16.28 14.46 13.93 5.09
Std 14.05 12.77 22.50 12.47

 

126

 

Table 33: Vertical forces summed for Recline 1, showing data for all

 

 

 

 

 

 

 

 

 

subjects.
Recline1
(Newtons)
- 3
Z \ \ \ \ \ \ ‘- \ \
‘95 i 5:551:69 if E?! 32:392 51:“ 53
a 5 S 85.855 .83 8‘53 .8385 3 83 8‘5
I?) 3 as 53522 a; 52 or 53528 a; '52
1 716.5 ~699.7 16.7 ~23 ~697.4 19.0 0.0 -700.0 16.4 266975 19.0 0.0
2 743.2 ~756.0 ~12.8 1.3 -757.3 -14.1 0.0 ~764.0-20.9 687662 -23.0 89
3 720.9 -723.4 -2.5 3.2 -726.6 -5.7 0.0 -727.4 -6.5 ~0.8-732.6 -11.7 -5.9
4 774.3 808.3 -34.0 6.7 ~815.1 -40.8 0.0 ~816.7~42.4 ~1.7~821.7 -47.4 ~6.6
5 821.0 806.2 14.8 7.1 813.3 7.7 0.0 -810.510.5 2.8-811.6 9.4 1.7
6 738.7 -737.8 0.9 1.6 ~739.4 -0.7 0.0 -743.3 ~4.6 39744.4 57 -5.0
7 714.2 -739.4 -25.2 1.7 -741.1 26.8 0.0 -744.0 -29.7 2.9-740.7 -26.4 0.4
8 725.4 -725.4 0.0 4.8 -730.2 49 0.0 ~730.8 55 ~06 -733.0 ~7.6 2.8
9 801.0 -819.2 ~18.2 2.1 -821.3 -20.3 0.0 8222212 09-8268 258 -5.5
10 856.6 840.7 15.9 1.7 842.4 14.2 0.0 -841.3 15.3 1.1-845.7 10.9 -3.3
11 743.2 ~804.7 -61.6 8.1 -812.8 -69.7 0.0 ~811.7-68.5 1.1-817.0 -73.9 -4.2
12 774.3 868.2 93.9 2.8 ~871.1 -96.8 0.0 -865.7-91.4 5.4-774.9 06 96.2
13 769.9 -819.3 -49.5 4.1 -823.4 53.6 0.0 ~822.0 -52.1 1.5-826.8 -57.0 -3.4
14 725.4 -743.8 ~18.5 0.0 -743.8 ~18.5 0.0 -743.2-17.9 0.6-746.1 20.8 2.3
15 6987-7367 ~38.1 -1.7 -735.0 36.4 0.0 ~718.6 -20.0 16.4 -735.2 ~36.6 -0.2
16 769.9 ~803.1 -33.3 01 -803.0 -33.2 0.0 -802.1-32.3 0.9-803.6 ~33.8 06
17 729.8 -731.5 -1.7 4.0 -727.5 2.3 0.0 -733.2 34 -5.8~728.8 1.0 -1.4
18 734.3 -763.0 28.7 -70.3 -692.6 41.6 0.0 ~726.1 8.1-33.5-724.0 10.3 -31.3
19 774.3 -770.4 3.9 7.1 -777.4 -3.1 0.0 -774.1 0.2 3.3-773.2 1.1 4.3
20 6764-7608 -84.4 6.3 ~767.1 -90.7 0.0 ~762.0 -85.6 5.1-759.9 -83.5 7.2
21 707.6 -727.9 -20.3 2.9 -730.8 -23.2 0.0 -730.0-22.5 0.7-728.0 -20.4 2.8
22 738.7 ~726.5 12.2 2.4 ~728.8 9.9 0.0 -728.7 10.0 0.1-731.8 6.9 -2.9
23 7120-7201 81 2.9 -723.0 -11.0 0.0 ~726.7~14.7 ~3.7~727.6 ~15.6 -4.5
Table 34: Average horizontal forces summed for Recline 1, 2 and 3.
(Newtons) Slumped Neutral Erect Super Erect
52 55 is 53 if 55 32 $3
.83 87,5, 83 8?, 93’ 8‘5 .89 8‘5
a; '52 a; 52 a; 5% a; 15%
Recline1 Ave -20.27 -0.50 ~19.77 0.00 -20.81 -1.04 -18.74 1.02
Std 30.27 15.54 33.89 0.00 30.21 8.45 27.18 21.93
Recline2 Ave ~12.50 2.35 ~14.84 0.00 20.73 -5.88 -22.43 -7.59
Std 30.22 8.16 26.30 0.00 31.18 19.97 31.23 21.77
Recline3 Ave ~15.07 -0.54 -16.04 0.00 —21.45 598 -19.51 ~3.89
Std 25.85 10.68 26.65 0.00 30.36 21.29 23.84 8.80

 

 

127

 

Table 33 and Table 34 display the vertical forces (vertical relative to lab
coordinates) of the subject trials. If the subject was in equilibrium, then the sum
of the vertical forces should total the subject’s weight. The subject’s weight was
subtracted from the totaled vertical force and listed in Table 33 and Table 34
under the column heading ”Differ with Weight”; these remaining force values
were termed the “residual forces”. Even after the data from the reference file of
the force plates (the unloaded condition of the plates) was subtracted from the
trial data, some residual forces remained and for many subjects, the residual
force was consistent across trials. An example of this residual force can be seen
in the data from Subject 1 of Table 33. It should be noted that Subject 18 shows
consistently high residual forces throughout all the conditions as well as Subject
12 in the super erect condition, but all of the other subjects show reasonable
residual force data

Table 33 lists the difference of the vertical forces from the subject’s
weight, and also normalizes the data to the neutral condition. To accomplish the
normalization, the force remaining after subtracting the vertical forces and the
subject’s weight in the neutral condition was subtracted from the other three test
conditions (“Differ with Neutral”). On average, the data varied from the subjects’
weights between -15 and -23 Newtons, with the negative sign meaning that the
calculations produced more loading than the subject’s weight. Again, in an ideal
situation, the weight of the subject would be equal to the sum of the vertical
forces, however as stated earlier some error was introduced through the

measurement process, the accuracy of the system and the series of calculations.

128

So with this in mind, the difference of 15 to 23 Newtons (3.4 to 5.2 lbs) is within

the typical 3% error seen in experimental data collection.

Estimation of Internal Joint Forces
The force data was broken down to vertical and horizontal components

and was used to estimate the forces at a joint location in the lumbar spine.
However, it should be noted again since the joint locations and the segment
centers of gravity must be estimated, the joint forces computed must be
considered only estimations. For subject 17, in recline 1, slumped, neutral, erect
and super erect conditions, the joint forces and moments at the estimated L5/S1
were computed and are listed in Table 35 and visually displayed in Figure 65.
(Subject 17 was chosen because the data set for this subject was complete and
the residual vertical and horizontal forces for all four articulations were

reasonable numbers.)

Table 35: Estimations of internal joint loads at the approximate L5/Sl joint
center. Subject 17, recline 1. Forces are converted to orientation of pelvis.
Fx is shear loading (positive anterior) and F2 is compressive loading
(positive superior).

 

 

 

 

 

 

 

Condition Fx Fz ML5/31
(N) (N) (N-m)
Slumped 84.7 515.4 20.39
Neutral 60.0 498.9 14.32
Erect 38.3 488.5 15.63
Super Erect 34.9 496.0 7.29

 

129

 

Sacrum

Figure 65: Visual display of the direction of loading on estimated L5/S1
joint center.

Table 36: Conditions tested by Andersson58 while monitoring disc pressure
and muscle activity.

 

9. Standing at ease.

 

10. Relaxed (no back support) sittingwith arms at the sides of the body.

 

11.Relaxed sitting with the arms supported.

 

12. Relaxed sitting with the feet unsupported.

 

13. Straight (erect) sitting.

 

14. Relaxed anterior sitting.

 

15. Straight anterior sitting.

 

 

16. Relaxed posterior sittinL

 

Andersson58 measured disc pressures in various seated positions and
converted these data to force estimates, Figure 66 and Figure 67. Notice that
many of the seated tasks exhibited force values (computed between L3 and L4)
between 400 and 500 Newtons. These data are similar to the estimations made

of the F2 forces from the external support forces gathered for this research. This

130

 

comparison shows that using the external support forces to estimate internal
loads provides reasonable results.

From the data listed in Table 35 it is clear that the compressive loads on
the spine decreased as the subject moved from a slumped to a neutral to an
erect posture. However, in the super erect position the loads increased to a
value similar to that of the neutral condition. Andersson58 presented postures
that showed similar trends in Table 36, condition 8 could be considered similar
to the slumped position, condition 2 could be considered similar to the neutral

position and condition 5 could be considered similar to the erect position.

 

 

 

700
N
e 600 1
W
t ..
‘3 532
s 500 ’1 in ,
i"
400 ,1 C}
1.)
300 | | j j I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 66: From Andersson 5", eight different postures with normalized disc
pressure, refer to Table 36 for list of test conditions.

131

 

 

 

700 ~>
N
e 600 L
w
t
o
n
s
500 ,.
400 .1.
Arms
Hanging

 

 

 

 

 

i

.Writing Type— Wig} t

WritinCl

 

 

 

Depression Relaxed
of Pedal

Figure 67: Disc pressure measured in an office chair during simulated work

activities“.

Table 37: Bush dissertation data compared to Andersson58 data.

 

Bush Compressive

Andersson Compressive

 

 

 

 

 

Forces (N) Forces (N)
(estimated from graph)
Drivingpostures General Seated Posture
Slumped 515.4 500
Andersson # 8 position
Neutral 498.9 480
Andersson #2
Erect 488.5 430
Andersson #5
Super Erect 496.0

 

 

 

132

 

 

"3‘” Andersson
" Bush “
,_ _ 1
_L 500 . {I

 

 

 

 

 

 

.. 300

 

 

 

 

 

Neutral Erect Slumped

Figure 68: Comparison between Andersson58 and Bush (shaded) data for
loading of the lumbar spine.

From Table 37 and Figure 68 the differences between Bush and
Andersson58 data are numerically listed and visually represented. As stated
above, a comparison of these data showed similar magnitudes. The difference
between the two sets of data increased as the posture became more erect. This
trend would be expected since Andersson measured the pressures in the
intervertebral disc which is located between bodies of vertebrae (anterior portion
of the vertebrae), and as a more erect posture is attained, loading from the
bodies of the vertebrae is transferred to the posterior facets. Andersson could
not measure this posterior loading; but the calculations using the external support

forces treated the loading as a whole, not distinguishing between anterior or

133

posterior regions. Therefore, the increasing differences between Bush and

Andersson data were logical.

5:31 ‘——> 18.4 cm
A-
1800N 566M .. 15.3cm
V Body Weight
. above L5/St
' 340 N Center of Gravity
350 N

ti 2700N 350N -——

fl
150N /5°-9 150N
Q"

 

 

 

35.0
cm

 

 

 

Figure 69: Comparison of predicted L5/S1 forces during lifting of a large

objects”.

Data for comparison of the shear loading and the moments on the

spine in seated postures is not available. However, data from Chaffin59 an

analysis of lifting tasks helps to put the magnitude of these forces and moments

into perspective. From Figure 69, the lifting tasks list shear loads range from
340N to 500 N for different lifting techniques. Using the numbers provided in

Figure 69, an estimate of the moments at the L5/Sl joint during lifting can be

computed. These moments range from 117 N-m to 130 N-m. As expected, the

lifting shear forces and moments are much larger in magnitude than those

calculated for the general seated posture. This cannot confirm the accuracy of

the calculations, but the numbers seem reasonable in terms of magnitude.

134

Segment Motions for Static Trials

Thorax and Pelvis Angles
In addition to the measurement of the support forces for the body at

the various chair positions, the postures of the subjects were also
measured. In this section, the total body openness (Figure 70) and chair
Total Support Angle (TSA) (Figure 71) will be listed and compared,
followed by the independent angular measurements of the thorax, pelvis,
elbow and knee.

{The posture of the torso was measured as a thorax angle, a pelvis
angle and‘then the combination of the pelvis and thorax position, called
openness, Figure 7E The position of the chair was also measured with
the angle of the thorax support, the angle of the pelvic support and the
Total Support Angle (TSA) which combined the thorax and pelvic support
angles. In both openness and TSA, an increasingly erect position
produced a larger angle.

Table 39 lists the average openness angle for all three recline
positions and the four chair articulations. Table 41 lists the average chair
positions for each condition across subjects. A 5° change of the thorax

support plus a 5° change of the pelvis support should occur between chair

articulations producing an overall 10° change between conditions.

135

fl>

A Openness

w
2

x._i

Figure 70: Openness as measured on the subjects.

Table 38: Average Openness of body (thorax plus pelvis orientation).

 

Body Openness
Degrees Slumped Neutral Erect Super Erect
Angle Std Angle Std Angle Std Angle Std

 

 

Recline 1 92.2 16.4 94.6 16.1 99.2 15.5 101.8 14.9
Recline 2 92.9 22.9 94.6 18.0 97.9 16.9 100.9 16.7
Recline 3 91.4 15.6 94.1 16.0 98.8 15.3 94.0 21.6

 

 

 

 

 

 

The data in Table 38 indicated that the subject moved to a more erect, or
open posture as the chair was moved to a larger TSA. However, the movement
of the body did not have the same magnitude as the chair. Table 39 indicated
that the chair moved 30° between the slumped and super erect conditions while
the subjects on average only moved 10° in reclines 1 and 2 from slumped to
super erect and 7° from slumped to erect in recline 3. Further examination of the
dynamic data in the next section will confirm the inconsistencies between the

chair and body movement ranges.

136

  
  

Thorax Angle

  
  

Negative

  
 

Positive
Buttocks iuttocks
Angle ”9'3

Figure 71: Total Support Angle, TSA is the thorax support angle plus the
pelvic support angle.

Table 39: Average Total Support Angle (TSA) of test chair in degrees.

Slumped Neutral Erect Super Erect

2 81.6 2.8 91.3 1.4 103.2 1.3 112.8 223
4 .4 1.2 102.6 112.1

 

In terms of statistical analysis, a One Way Repeated Measures Analysis of
Variance was performed on the openness data and the chair data. All data
passed the equal variance testing, but the chair data did not pass the normal
distribution test nor did recline 2 of the openness data. Recline 3 of the
openness data could not be tested because the test matrix could not be inverted.

For all recline angles and all chair articulations, the data at each condition
was significantly different from the data at all other chair positions, Table 40.

These results confirm that the articulation conditions were consistently and

137

sufficiently far enough apart from each other to detect a difference. If a
difference in chair positions were undetectable, then it would not be reasonable
to assume that a significant difference could be detected in the body position.
For the body openness angle, a significant difference was measured
between all conditions in recline 1, and only between the slumped and super
erect, and the neutral and super erect conditions for recline 2. An analysis of the
pelvis angle separate from the thorax angle will explain why all conditions were
found significantly different from one another in recline 1 and not in recline 2.

Recline 3 could not be tested.

Table 40: Significant difference testing for body openness and chair TSA.

 

 

 

 

Recline 1
S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N

Openness Y Y Y Y Y Y

TSA Y Y Y Y Y Y
Recline 2

Openness Y Y - - - -

TSA Y Y Y Y Y Y
Recline 3

Openness Matrix not able to be tested

TSA Y Y Y Y Y Y

 

 

 

Motion of Individual Segments: Thorax
Next, the individual motions of the body segments are presented. The

angular data of the thorax for each subject in Recline 1 is listed in Table 41 and
the average thorax position for each recline angle is listed in Table 42. These
data showed that as the thorax support of the test chair was adjusted the
subject’s thorax also changed position. As the chair support rotated rearward to

a more erect position, the thoracic cage, as measured by the sternal notch and

138

mid-sternum targets, also rotated rearward. This was confirmed by the numerical
data, which increased in magnitude as the thorax support was rotated rearward.
The negative sign in Table 41 and Table 42 indicate that the thorax was rearward

of lab vertical with the x axis pointing in the anterior direction.

Table 41 : Subjects’ thorax angles with respect to vertical (wrtv) for

 

 

 

 

 

 

Recline1.
Thorax (Degrees) (wrtv) negative is
Angle rearward of
vertical
Recline1
Subject Slumped Neutral Erect Super Ave Std
Erect

1 -23.0 -23.3 -25.7 -25.8 -24.47 1.52
2 -29.7 - - -30.3 -30.04 0.42
3 -33.9 -32.1 -33.0 -39.6 -34.65 3.40
4 -34.9 -33.3 -37.4 -38.3 -35.98 2.27
5 -17.1 -17.3 -17.2 -18.5 -17.53 0.67
6 -31.1 -32.2 -37.8 -39.6 -35.17 4.16
7 -18.2 -17.4 -20.6 -21.7 -19.49 1.99
8 -19.9 -20.2 -23.0 -20.3 -20.86 1.45
9 -34.6 -34.0 -33.7 -34.6 -34.24 0.44
10 -28.2 -30.4 - -. -29.32 1.56
11 -31.7 -31.2 -30.8 -36.0l -32.42 2.41
12 -47.7 -49.7 -49.3 -49.7 -49.08 0.96
13 -43.1 -44.4 -46.2 -47.6 -45.33 1.99
14 -40.3 -40.0 -38.8 -41.01 -40.03 0.91
15 -36.0 -38.3 -41.3 -42.3 -39.46 2.89
16 -47.3 -46.4 -46.5 -50.3 -47.63 1.82
17 -41.8 -46.6 -47.2 -49.1 -46.18 3.11
18 -44.8 -37.3 -43.5 -47.2 -43.20i 4.18
19 -39.2 -44.0 -45.5 -44.0 -43.17 2.72
20 -41.9 -43.7 -42.8 -41.4 -42.45 1.00
21 -34.8 -34.9 -39.4 -36.6 -36.41 2.16
22 —30.5 -35.2 -34.8 -36.91 -34.34 2.73
23 -43.2 -47.5 -51.4 -55.3 -49.33 5.22

Ave -34.5 -35.4 -37.4 -38.5

Std 9.0 9.6 9.7 10.1

Range 30.6 32.3 34.2 36.8

 

 

139

 

Table 42: Average thorax angles for all subjects across all three recline
angles with respect to vertical. Negative angle is rearward of vertical.

 

 

Thorax Angle (wrtv)

(Degrees)

Subject Slumped Neutral Erect Super

Erect

Recline 1

Ave 345 -35.4 -37.4 -38.5
Std 9.0 9.6 9.7 10.1
Range 30.6 32.3 34.2 36.8
Recline 2

Ave -36.4 -38.9 -40.0 -41 .1
Std 9.8 10.5 11.1 10.9
Range 33.4 35.0 38.4 39.1
Recline 3

Ave -41 .7 —42.1 -45.6 -45.4
Std 8.9 9.6 9.0 10.2
Range 32.5 33.3 35.2 37.4

 

 

 

Table 43: Average angle change of all subjects from the slumped to the
super erect condition for the thorax and pelvis.

 

 

 

 

 

(Degrees) Thorax Pelvis
Recline1 3.7 -5.6
Recline 2 5.2 -6.3
Recline 3 4.8 -5.7

 

 

 

 

Table 44: Testing for significant differences between the thorax angles of
the subjects.

 

 

 

Thorax Angle

Svs.SEst.SE Evs.SE Svs.E st.E Svs.N
Recline 1 Y Y - Y Y -
Recline 2 Y Y - Y - -
Recline 3 Y Y - Y Y -

 

 

 

 

 

 

 

 

140

For the thorax, pelvis and extremity data, a One Way Repeated Measures
Analysis of Variance was performed to determine significant differences, Table
44. The statistical results for the thorax were consistent with what was found in
the analysis of the force data. That is, significant differences were found
between every other articulation position, i.e. slumped to erect or slumped to
super erept or neptrgl tp super erect. However for both recline 1 and recline 3, a
statistically significant difference was found between the thorax angle in the
neutral and erect conditions. This difference was not seen in recline 2. As with
the force data, these data suggest that the chair thorax support must be moved
10° to consistently achieve statistically different thorax positions.

Notice that in recline 3 the average thorax angle change between the erect
and super erect condition was small and opposite the trend. in this reclined
position, it may be that the subjects have reached a physical limitation of the
combination of the pelvis and thorax rotation; this, in combination with the
requirement to maintain foot contact with the floor restricts the subject from
additional movement. The same trend is seen with the pelvis in recline 3 which
will be discussed shortly (Table 45). Following are some theories for what may
cause the reduced movement in recline 3 for both the thorax and the pelvis
segments. It may be that with combination of the large recline angle and the
erect posture, the hamstring tension is at its maximum length and further rotation
of the pelvis is not allowed. Likewise, the passive tissues along the back may be
stretched to their maximum length and further coupled rotation of the pelvis and

thorax is not possible.

141

Motion of Individual Segments: Pelvis
Table 45 lists the pelvis angles for all subjects at all chair articulations for

recline 1, while Table 46 lists the average pelvis angles for all three recline
angles. A trend was seen within and across the subjects: as the pelvis support
moved from a slumped to an erect position, the pelvis angle became smaller.
The negative sign of the pelvis angle denoted that the vector from the S1 target
to the mid-ASIS location was above lab horizontal, or the front of the pelvis was
tipped upward. As the subject became more erect, the front of the pelvis tipped
downward bringing the pelvis angle closer to the horizontal reference line
producing a smaller angle.

The statistical analysis of the pelvis angles shed some surprising results.
For recline 1, all of the conditions were significantly different from one another.
So, contiguous articulations moved the pelvis enough to demonstrate a
difference from each other. In the more upright condition, recline 1, the chair
appeared to have better command over how the pelvis moved and forced the
pelvis to move with the chair.

Recline 2 did not show any significant differences between conditions and
recline 3 showed differences between slumped and super erect and slumped and
erect. If the theory that a more upright recline provides better control over the
pelvis is true, then a higher number of statistically significant differences would
be expected in recline 2 with the least amount of statistically significant
differences found in recline 3. Thus data from reclines 1, 2 and 3 produce

conflicting results toward this theory.

142

Table 45: Pelvis angles for each subject at each chair articulation in recline
1. The pelvis angle is measured with respect the horizontal lab (wrth) with
the negative sign meaning the vector is above horizontal.

 

 

 

 

 

 

Pelvis Angle (wrth) negative is rearward of vertical
(Degrees)
Recline 1
Subject Slumped Neutral Erect Super Ave Std
Erect
1 -34.4 -31.1 -29.4 -28.5 -30.9 2.6
2 -34.7 -33.6 -32.2 -31.3 -33.0 1.5
3 -41.0 -39.3 -34.9 -32.2 -36.9 4.0
4 -41.0 -38.8 -37.9 -35.5 -38.3 2.3
5 -43.1 -42.4 -38.2 -36.2 -40.0 3.3
6 . -33.5 -30.5 -28.3 -30.8 2.6
7 -35.1 -31.9 -28.6 -27.5 -30.8 3.4
8 -34.1 -32.1 -29.1 -25.9 -30.3 3.6
9 -43.7 -41.8 -39.0 -36.0 -40.1 3.4
10 -49.4 -42.1 -38.9 -34.7 -41.3 6.2
11 -36.1 -35.1 -33.3 -32.7 -34.3 1.6
12 -12.6 -12.6 -11.4 -8.8 -11.4 1.8
13 -26.0 -23.3 -21.9 -19.6 -22.7 2.7
14 . -39.3 -36.6 -34.3 -36.7 2.5
15 -24.5 -25.2 -25.2 -22.8 -24.4 1.1
16 -36.0 -34.4 -33.0 -26.7 -32.6 4.1
17 -28.3 -25.6 -22.4 -20.8 -24.3 3.4
18 -26.9 -27.7 -26.1 -23.9 -26.1 1.6
19 -23.0 -23.5 -22.0 -19.7 -22.1 1.7
20 -22.6 -22.4 -19.8 -19.4 -21.1 1.7
21 -32.8 -30.7 -27.5 -26.4 -29.4 3.0
22 -29.9 -28.9 -27.2 -25.1 -27.8 2.1
23 -17.7 -17.3 -14.4 -21.5 -17.7 2.9l
Ave 321 -31 .0 -28.7 -26.9
Std 9.1 8.0 7.6 6.8
Range 36.7 29.7 27.5 27.4

 

 

143

Table 46: Average pelvis angles for all subjects across all three recline
angles, with respect to lab horizontal.

 

 

 

Pelvis Angle (wrth)

(Degrees)

Subject Slumped Neutral Erect Super
Erect

Recline 1

Ave -32.1 -31 .0 -28.7 -26.9

Std 9.1 8.0 7.6 6.8

Range 36.7 29.7 27.5 27.4

Recline 2

Ave 361 -35.0 —32.1 -30.2

Std 9.5 9.4 8.1 8.0

Range 35.5 33.0 29.5 30.5

Recline 3

Ave -40.3 -38.3 -36.6 -36.1

Std 9.2 8.4 8.2 7.6

Range 34.3 32.0 29.6 28.4

 

 

Table 47: Testing for significant differences between the subjects’ pelvis

 

 

 

angles.
Pelvis Angle
Svs.SE st.SE Evs.SE Svs.E st.E Svs.N
Recline 1 Y Y Y Y Y Y
Recline 2 - - - - - -
Recline 3 Y - Y - -

 

 

 

 

 

 

 

144

 

Limb Angles: Elbow and Knee
Finally, to complete the evaluation of the posture of the body, the positions

of the arms and legs will be discussed.

A unique aspect of the protocol used for this study was that the subjects
were allowed to choose both the wheel position and the foot position. Allowing
the subject to choose these positions helped to ensure that the torso posture and
not the effect of the package (wheel and foot position) was the primary influence
on the loading of the chair.

Table 48 lists the elbow angles for all the subjects in all chair articulations
for recline 1, and Table 49 lists the average elbow angles for all three reclines.
Upon examination of Table 48 it appeared as though each subject chose a
preferred elbow angle and then maintained a similar elbow angle across chair
articulations. Table 50, the statistical analysis of the elbow angle, confirms that

there was not a significant change in the elbow angle between chair positions.

145

Table 48: Elbow Angles for all the subjects at all the chair articulations for
recline 1. Note the small deviation within a subject across conditions.

 

 

 

 

 

Elbow Angle
(Degrees)
Recline 1
Subject Slumped Neutral Erect Super Ave Std
Erect
1 101.4 106.4 107.7 104.0 104.9 2.8
2 91.3 86.1 90.2 97.9 91.4 4.9
3 126.8 123.9 118.9 122.6 123.1 3.3
4 102.3 101.3 102.8 101.0 101.8 0.9
5 102.2 105.0 105.8 102.4 103.9 1.8
6 106.6 100.7 110.1 115.6 108.3 6.3
7 101.7 96.3 102.6 102.9 100.9 3.1
8 88.3 86.7 88.8 96.0 90.0 4.2
9 106.9 104.3 103.1 108.9 105.8 2.6
10 91.8 93.7 96.5 111.1 98.3 8.8
11 133.0 131.1 131.1 134.1 132.3 1.5
12 117.6 119.9 104.4 100.5 110.6 9.6
13 113.9 117.4 119.0 118.7 117.3 2.3
14 123.8 127.5 122.1 117.6 122.7 4.1
15 102.7 114.8 109.8 112.4 109.9 5.2
16 80.1 79.0 87.4 76.8 80.8 4.6
17 112.9 116.6 114.0 107.1 112.7 4.0
18 117.1 111.7 116.6 121.0 116.6 3.8
19 92.0 103.6 104.7 100.5 100.2 5.7
20 118.1 133.3 127.1 114.1 123.1 8.7
21 122.6 119.3 116.8 116.3 118.8 2.9
22 82.0 94.8 86.7 96.6 90.0 6.9
23 118.2 120.2 119.1 129.6 121.8 5.3
ave 106.7 108.4 108.1 109.0
std 14.6 14.9 12.5 12.6
rang 52.9 54.3 44.3 57.4

 

146

 

Table 49: Summary of Elbow Angles for Recline 1, 2 and 3.

Elbow Angle Summary
(Degrees)

 

Recline 1
ave 106.7 108.4 108.1 109.0
std 14.6 14.9 12.5 12.6

Recline2
ave 109.8 110.7 111.3 109.6
std 16.1 16.3 15.0 15.1
Recline3

ave 108.2 110.7 109.7 112.0
std 13.5 12.3 12.1 13.3

 

 

 

Table 50: Testing for significant differences of the elbow angle between the
various chair articulations.

 

Elbow Angles
Svs.SE st.SEEvs.SESvs.E st.E Svs.N
Recline 1 - - - - - -
Recline 2 - - - - - -
Recline 3 - - - - - -

 

 

 

Table 51 : Testing for significant differences of the elbow angle between the
recline positions.

 

Elbow Angles
Reclines 1 vs. 2 2 vs. 3 1 vs. 3
wificance - - -

 

 

 

 

Since there were no significant differences exhibited between the slumped,
neutral, erect and super erect groups, Table 50, these data were lumped
together under each recline condition and an ANOVA used to test overall
differences between reclines, Table 51, none were found.

Table 52, Table 53 and Table 54 show the knee angles of the subjects

across conditions. Again, similar to the elbow angles, each subject picked a

147

preferred knee angle and did not deviate from that position across chair
articulations. Table 54 validates these trends by the lack of statistically
significant results across the majority of chair articulations. Recline 2 showed a
statistically significant difference between the slumped and super erect positions,

but this was not supported by the data from the other two recline angles.

Table 52: Knee angles for all the subjects at all the chair articulations for
recline 1. Note the small deviation within a subject across conditions.

 

 

 

 

 

Knee Angle
(Degrees)
Recline 1
Subject Slumped Neutral Erect Super Ave Std
Erect
1 124.7 128.4 133.4 133.6 130.0 4.3
2 125.1 124.7 126.8 123.0 124.9 1.5
3 116.1 114.9 111.3 120.7 115.7 3.9
4 135.9 134.5 139.3 139.7 137.4 2.5
5 120.6 103.9 106.4 104.8 108.9 7.9
6 113.2 111.2 117.2 118.4 115.0 3.3
7 129.0 132.7 133.2 134.6 132.4 2.4
8 120.5 128.2 128.9 128.4 126.5 4.0
9 132.3 130.4 129.9 130.4 130.7 1.1
10 116.8 141.7 140.3 131.1 132.5 11.5
11 150.2 146.8 149.8 147.2 148.5 1.8
12 139.8 147.3 146.1 144.9 144.5 3.3
13 149.5 150.5 153.2 154.9 152.0 2.5
14 146.1 148.3 139.1 136.5 142.5 5.6
15 142.8 144.8 140.3 144.2 143.0 2.0
16 121.8 122.3 124.4 119.1 121.9 2.2
17 140.2 - - - 140.2 -
18 146.3 143.4 141.2 139.5 142.6 3.0
19 165.1 167.8 170.1 168.8 167.9 2.1
20 133.2 136.6 138.2 134.8 135.7 2.2
21 143.6 140.4 144.8 145.2 143.5 2.2
22 153.2 160.2 160.2 159.3 158.2 3.4
23 152.5 157.9 158.9 - 156.4 3.5
ave 135.6 137.1 137.9 136.1
std 14.2 16.0 15.6 14.9
range 51.8 63.8 63.7 63.9

 

148

 

Table 53: Summary of knee angles for recline 1, 2 and 3.

 

Knee Angle Summary
(Degrees)
Recline1
ave 135.6 137.1 137.9 136.1
std 14.2 16.0 15.6 14.9

Recline 2
ave 134.4 135.8 136.7 138.0
std 14.3 15.1 15.5 15.5

Recline 3
ave 131.0 133.4 134.0 132.6
std 13.8 15.5 16.0 16.7

 

 

 

Table 54: Testing for significant differences of the knee angle between the

various chair articulations.

 

Knee Angle

S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Recline 1 - - - - - -
Recline 2 Y - - - - -
Recline 3 - - - - - -

 

 

 

Table 55: Testing for significant differences of the elbow angle between the

recline positions.

 

Knee Angles
Reclines 1 vs. 2 2 vs. 3 1 vs. 3
@gnificance - - -

 

 

 

 

Overall when examining the knee and elbow angles, the data shows that

subjects chose a preferred knee or elbow angle and if they have the choice, they

do not deviate from these angles regardless of the chair position, or the body

posture. These data will be beneficial to the designers of the automobile

interiors. Instead of fixing the horizontal distance between the steering wheel

and pedals, a more desirable alternative may be to allow the driver to choose the

149

position of each, and reduce the overall track motion of the seat. This approach
would accommodate people with varying leg and arm lengths better than the

current approach.

150

Results, Analysis and Discussion of Dynamic Data

In this section, data from the dynamic tests are analyzed and discussed.
Since the dynamic tests were conducted by taking data continuously over a ten
second time period, the data were in the form of traces rather than single points
as generated from the static test data. When possible, comparisons between the
static and dynamic tests were made. Also, new data that was only available from
the dynamic tests were analyzed, such as the travel path of the resultant force

vector.

Chair Motions
Prior to comparisons between the dynamic and static tests, the chair

movement itself was analyzed.

Recline Angle
First, the recline angle of the chair was examined. Again in the dynamic

condition, the three recline angles of 20, 24 and 28 were fixed while the torso
articulation was adjusted in a continuous fashion through the full ranges of
motion (slumped to super erect). Comparing the data in the shaded cells of
Table 56, the recline angle between the dynamic and static tests were similar.
The average dynamic recline angles as measured by the recline bar on the chair
were -40.7°,- 45.4°, -49.4° for recline angles 1, 2, and 3, respectively, compared
to -41.0°,-45.2° and -49.4° for the static chair positions. From these two sets of
data and the statistical analysis, Table 57, it was clear that the recline angles

could be considered the same between the dynamic and static testing.

151

The recline angle data for each subject in the static and dynamic

conditions was statistically analyzed via a one way ANOVA test. However, this

test showed that the data failed to be normally distributed so a Kruskal Wallis one

way ANOVA on Ranks was performed, Table 57. The analysis found no

statistically significant difference between the static and dynamic recline angles,

however a statistically significant difference was shown within dynamic recline

conditions. This demonstrated that the chair was positioned in a repeatable

fashion between recline conditions, that the recline angles were far enough apart

to detect a difference between recline conditions in the dynamic tests, and that

the recline positions were similar between static and dynamic testing.

Table 56: Total Recline Angle (TRA) for the chair during both dynamic and
static testing. All angles in degrees.

 

 

 

 

Chair TRA for

Dynamic Tests
Degrees Recline Recline Recline
1 2 3
Ave Max -40.0 -45.1 -49.0
Max std 2.6 0.8 1.1
Ave Min -41.3 -45.7 -49.7
Min std 2.6 0.9 1.1

 

 

 

 

 

 

 

TRA for Static
Tests
Recline Recline Recline
1 2 3
Average -41.0 -45.2 -49.4
Std 1.3 1.2 0.8

 

 

152

Table 57: Statistical analysis of Total Recline Angle during static and
dynamic testing.

 

 

 

 

 

 

 

 

 

 

Recline 1 Recline 2 Recline 3
Static vs. Dynamic - - -

Rec1vs2 Recl vs3 Rec2vs3
Dynamic Y Y Y

 

 

 

 

 

Y= a statistically significant difference was found

Torso Articulations

For dynamic tests, the recline angle of the chair was held constant while a

test assistant moved the chair through the full range of total support angles to

articulate the torso. The Total Support Angle (TSA) was initially positioned in the

neutral condition and then moved to slumped, back through neutral and to super

erect. For the entire 10 second period, force and motion data were collected at a

12 Hz sample rate. The trials were alternated, so that one trial started toward the

slumped position while the second trial started toward the super erect position

first. Recall that the static tests were positioned at each of the four torso

articulations, slumped, neutral and erect and super erect and data were collected

at each point, not between articulations.

Table 58: Total Support Angle (T SA) for the chair during both dynamic and
static testing. All angles in degrees.

 

 

 

 

 

 

 

 

TSA for Dynamic
Tests
Degrees Recline Recline Recline
1 2 3

Super 113.6 112.3 113.0
Erect

Slumped 78.7 80.5 81.4
Ragqe 34.9 31.8 31.6

 

 

 

 

 

 

 

TSA for Static
Tests
Recline Recline Recline
1 2 3

Super 112.5 112.8 112.1
Erect
Slumped 83.0 81.6 81.4
Range 29.5 31 .2 30.7

 

 

 

 

 

Again, comparing the shaded cell data from Table 58 the average ranges
of chair movement between the maximum and minimum TSA slumped to super
erect in the dynamic test conditions were 349°, 31 .8° and 31 .6° for reclines 1, 2,
and 3 while the ranges of movement for the static conditions were similar in
reclines 2 and 3 (31 .2°, 30.7°) but 5° less for recline 1, (295°). For the dynamic
tests the chair was moved in a smooth continuous fashion and once the end
articulations were reached, the direction of motion was reversed. Since the chair
did not have stops at each setting, but rather markings on the chair that were
aligned for each TSA, it was not surprising that there was some variation in the
ranges. From observation, Reclines 2 and 3 average movements were similar
between the static and dynamic tests, however when examined statistically,
differences between the static and dynamic tests were determined.

The number of trials between the static and dynamic test data for each
recline had large variations between the number of samples for each condition
(dynamic condition had more samples than the static condition). The statistical
analysis was performed using separate tests for comparisons between each
condition (slumped: static vs dynamic, super erect: static vs. dynamic) rather
than an ANOVA. Again the data failed the normality test and a Mann-Whitney
Rank Sum Test was performed on each set of data, Table 59. For each case,
the minimum TSA’s (slumped condition) of the static test was compared to the
minimum TSA’s of the dynamic tests for each recline angle. The maximum
TSA’s (super erect) were also compared. In each recline angle, significant

differences were determined to exist at either the minimum or maximum TSA’s,

154

thus leading to the conclusion that the full ranges of movement of the chair were
significantly different between the static and dynamic conditions.

When appropriate, comparisons will be made between the dynamic and
static test results with regard to torso articulation. However, it is important to
note that the dynamic conditions on average had a larger range of chair motion
than the static conditions and that statistically significant differences existed for

all recline angles either at the maximum or minimum TSA, Table 59.

Table 59: Statistical analysis of Total Support Angle during static and
dynamic testing.

 

 

 

 

 

Static vs. Dynamic Static vs. Dynamic
Slumped (Min) Super Erect (Max)
Recline 1 Y -
Recline 2 Y -
Recline 3 - Y

 

 

 

 

 

Y= a statistically significant difference was found

Force Data for Dynamic Tests

Primary Forces
Next, the primary forces exerted during the dynamic tests will be

presented and compared to the primary forces seen in the static tests. Again,
the primary forces were the normal forces on all of the support plates, except for
the wheel where the load from the arms produced a downward shear force on
the support plate and was considered a primary force, Figure 72. The normal
forces were all loads into the support plate from the body, therefore they were all

identified as negative support forces.

155

Table 60: Average primary forces for both the Static and Dynamic Tests

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(Newtons).

Newtons Static Tests Dynamic Range

Tests Difference
Recline 1 Min Max Range Min Max Range Dyn-Stat
Thigh -127.7 -140.2 12.5 -95.8 -173.0 77.2 64.7
Buttocks -403.5 -432.4 28.9 -406.9 -450.7 43.8 1 4.9
Pelvis -22.5 -27.9 5.4 -19.6 -40.6 20.9 15.5
Thorax -182.1 -187.4 5.3 -154.3 -193.5 39.2 33.9
Foot -93.4 -112.1 18.7 -72.5 -127.1 54.6 35.9
Wheel (Fx) -32.3 -33.3 1 -29.8 -37.4 7.6 6.6
Recline 2
Thigh -126.6 -133.8 7.2 -95.3 -183.0 87.6 80.4
Buttocks -390.7 -418.4 27.7 -393.7 -439.4 45.7 18.0
Pelvis -28.5 -35.6 7.1 -19.8 -42.9 23.1 16.0
Thorax -197.5 -209.6 12.1 -169.4 -205.2 35.9 23.8
Foot -86.3 -107 20.7 -65.6 -119.1 53.5 32.8
Wheel (Fx) -31 -31.9 0.9 -32.4 -39.9 7.4 6.5
Recline 3
Thigh -103.2 -127.5 24.3 -92.6 -183.1 90.5 66.2
Buttocks -382.5 -408.6 26.1 -383.3 -424.4 41 .1 15.0
Pelvis -30.7 -39.1 8.4 -22.4 -42.7 20.3 1 1 .9
Thorax -221 .6 -230 8.4 -194.0 -224.7 30.7 22.3
Foot -83.3 -108.9 25.6 -65.4 -116.4 51.1 25.5
Wheel (Fx) 326 -33.7 1.1 -33.7 -40.4 6.8 5.7

X
Z
Z
1::
=

 

 

Figure 72: Primary forces represent the normal forces under all of the force
plates except for the steering wheel where the shear force, Fx, was the

maximum force.

156

 

Table 60 lists the average maximum, minimum and ranges for the primary
support forces for both the static and dynamic tests. As can clearly be seen by
the last column of data in Table 60, the primary forces exerted by the subject
onto the chair were of a higher magnitude for all reclines and all support plates in
the dynamic tests. The largest differences were seen in the thigh plate with an
average of 70.4 N and the foot and thorax supports followed with average
differences of 31.4 N and 26.7 N, respectively. Recall, the ranges of chair motion
were larger for the dynamic test trials. However, the magnitude of difference
between the static and dynamic forces was more than could solely be attributed
to the larger range of motion of the chair in the dynamic movements. it is unclear
as to what, aside from the different chair ranges, would cause these differences
in loading, specifically why the dynamic conditions consistently produced higher
loading patterns for all support plates.

In Table 60 the forces are represented as maximum and minimum values,
in the static conditions, the location of the maximum and minimums were easily
defined in terms of chair position, i.e. the maximum and minimums had to occur
at one of the four torso positions. Table 61 identifies the position where the
maximum and minimum forces occurred for the static tests. Identifying the
location of the maximum and minimums for the dynamic tests was more difficult,
but as can be seen in Figure 73 the maximum and minimum trends occurred in
similar positions. However these points were slightly before or after the endpoints
of the chair positions, but were in the same region (slumped or neutral vs erect or

super erect). in order to determine the chair position at the maximum or

157

minimum primary force, the points were selected from the graphs and matched to
the same frame number for the graph of the chair position. This tedious
procedure was completed for all the dynamic trials in the middle recline position
(Recline 2) and a sample of these data are presented in Table 62. Thus, the
chair position was determined for the maximum and minimum primary force
value, Table 63.

As an example of the trend of the primary forces during dynamic
movement, force data plots were developed for all subjects, and samples of
these graphs are presented in Figure 73, Figure 74 and Figure 75. These figures
show the chair movement in terms of openness (a larger angle means a more
erect chair condition) and the forces (a more negative force means the force has
a larger magnitude). Comparing Table 61, Table 63 and Figure 73 and it can be
seen that the force trends of the thighs and feet are similar to the static tests for
the majority of subjects. The minimum force magnitude of the thigh occurs near
the super erect chair position and the maximum near the slumped condition. For
the feet, the minimum force magnitude occurs near the slumped chair position
and the maximum near the super erect position. Again, from a visual inspection,
the primary force on the wheel does not appear to be affected by the torso

articulation.

158

Table 61: TSA where the maximum and minimum average primary forces
were exerted for static tests.

 

 

 

 

 

 

 

 

 

 

 

Minimum Magnitude Maximum Magnitude
Rec 1 Rec 2 Rec 3 Rec 1 Rec 2 Rec 3
Thigh SE SE SE N S S
Buttocks SE SE SE S S S
Pelvis SE E SE S S S
Thorax E SE SE N N S
Foot S S S SE SE SE
Wheel No
Change

 

 

 

 

 

 

 

S= Slumped, N= Neutral, E = Erect, SE = Super Erect

Table 62: Sample of data from which Table 63 data was obtained.

 

Trial

1 1drec2a
1 1drec2b
1 Zdrec2a
1 2d rec2b
1 3d rec2a
1 3d rech

Force

Chair
Position
at Max
Force

Chair Range

Estimated Location of S,N,E, SE
Chair Positions based on Chair

Range

Chair
Position
at Max
Force

 

Max
or Min

Thigh
Fz
(N)

Degrees

Max
(Deg)

Min
(Deg)

8
(Deg)

N
(Deg)

E
(Deg)

SE
(Deg)

 

Max

~134.9

109.0

110.3

79.4 79.9 89.9

99.9

109.9

SE

 

Max

-116.0

107.0

110.0

79.7 79.8 89.8

99.8

109.8

E-SE

 

Max

-103.9

112.0

112.7

78.4 80.5 90.5

100.5

110.5

SE+

 

Max

-113.4

108.0

114.1

79.6 81.8 91.8

101.8

111.8

E-SE

 

Max

-153.6

110.0

112.1

79.2 80.6 90.6

100.6

110.6

SE

 

 

Max

 

 

-153.6

106.0

 

 

113.8

 

 

 

82.6 83.2 93.2

103.2

 

113.2

 

 

E-SE

 

Table 63: The most common TSA position of the chair at the maximum and
minimum primary force values for Recline 2 dynamic tests.

 

 

 

 

Minimum Maximum
Magnitude Magnitude
Rec 2 Rec 2

Thigh E-SE S-N
Buttocks Mixed Mixed
Pelvis S-N E-SE
Thorax Mixed Mixed
Foot S SE
Wheel E-SE S

 

 

 

 

S= Slumped,

N= Neutral, E = Erect, SE = Super Erect, Mixed

159

Table 64: Statistically significant difference in static tests between slumped
and super erect conditions for the primary forces.

 

 

 

 

 

 

 

 

 

Statistically Significant Differences
Rec 1 Rec 2 Rec 3

Thigh - - Y
Buttocks Y Y Y
Pelvis Y Y Y
Thorax - - -
Foot Y Y Y
Wheel - - -

 

 

 

 

 

Table 65: Statistically significant differences of primary forces for the
dynamic tests between the maximum and minimum values.

 

Statistically Significant Differences
Rec 1 Rec 2 Rec 3
1m - - -
Buttocks - - -
Pelvis - - -
Thorax - - -
Foot — — -
Wheel - - -
(FX)

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 64 and Table 65 list the results for the statistical analysis of the
primary force data for both the static and dynamic test trials. This analysis
concluded that there was no statistically significant difference between the
maximum and minimum primary force values for the dynamic test trials. A pair-
wise comparison was completed using a one way repeated measures ANOVA on
ranks at a 95% confidence level. The data was not normally distributed thus the
ranks test was performed. These data contradict the findings of the statistical
analysis of the static tests, Table 64, where the primary forces for the thigh,
buttocks, pelvis and feet were found to have statistically significant differences

between the slumped and super erect conditions.

160

150

501

-150

-250

Force! Angle

-350

-450

-550

150 '

50

Force! Angle

-150

-250

-350

Primary Forces: 16drec2a

 

 

 

 

 

 

 

“*g-.".’.~'o" ’ 4 4 5J-
’ ' * V'i‘wtf-‘3‘ft-rmpm‘4-r‘r ‘

 

Super Erect Slumped

.. - - ”i" ’ .
’ 4.2“" M‘V-fifl-L - - .
ins/“’6" ' '.'N-‘-’V."0'w 44.-we» "“

 

 

’WWWW , Mm fl

 

Time

 

Primary Forces: 1 6drec2a

 

 

Super Erect

 

 

 

 

 

Slumped

 

 

 

‘ .. ”MW

 

WW“*W -& A _ -
1+ Chair TSA

i »-—- Thorax Fz
-_ _ . . -0— Behind Pelvis Fz

W Wheel Fx g 9

 

 

 

Time

 

+ Chair TSA .
~-—- Buttocks Fz‘ ‘
—— Thigh Fz ‘
., '1' —< Feet F2

i

 

Increasing Force
Magnitude

i
l 1

 

 

 

 

Increasing Force
Magnitude

 

 

 

Figure 73: Primary forces (Newtons) from Subject 16, in recline 2, dynamic
movement from neutral to super erect to slumped back toward super erect.
All force data is displayed with respect to the chair TSA (degrees).

161

Primary Forces: 9drec3a
Super Erect Slumped

 

 

 

 

 

 

 
    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

150
50 I
- "7 I "11' ~ ~
-50 .
’ 2
‘g’ -150 - . _ -, f
‘ \/ + Oiair TSA
1% -250 7 f ——I— Buttocks F2
'350 t i -_t, - - 4 _. A 1 Thigth
m- “,4..- Fee, F2 .
-450 a. , f V - f ; -- ., .---
lncreasing Force
-550 Magnitude
Time
Primary Forces: 9drec3a
Super Erect Slumped
150
Increasing Force
Magnitude
450 W
,—O—Chair TSA
350 Fan—Thorax Fz
‘ —D— Behind Pelvis Fz

 

 

 

 

Time -~x- Wheel Fx

Figure 74: Primary forces from Subject 9, in recline 3, dynamic movement
from neutral to super erect to slumped back toward super erect. All force
data is displayed with respect to the chair TSA (degrees).

162

Primary Forces: 23drec2b

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

150 Super Erect __ Slumped
50
'50 "3‘ _ _ T ‘5‘1W'é‘wt‘flfififi:26m
. g";16;:---,-§-—"-..f-.-4’.:,,<,dungrgswrrf‘ ,Chi . :rSA
-‘ —+— air
'3 -150 f \ _
5 / ——!—— Buttocks F2
3 _250 —Thigh Fz
I? --—-‘~’:- Feet Fz
Increasing Force
-450 Magnitude
-550
Time
Primary Forces: 23drec2b
Super Erect Slumped
150
50 increasing Force
Magnitude
E -50
I \
g i
E -150 - , -- ~ M 11
-250 +Chair TSA
+ Thorax Fz
—o—Behind Pelvis F2
-350

 

Time

 

- - - Wheel Fx

Figure 75: Primary forces from Subject 23, in recline 2, dynamic movement
from neutral to super erect to slumped back toward super erect. All force
data is displayed with respect to the chair TSA (degrees).

163

Secondary Forces
The secondary forces were also tabulated for the dynamic tests. As a

reminder, the secondary forces were the anterior/posterior shear forces under
the buttocks, thighs and feet and the superior/inferior forces behind the pelvis
and thorax, Figure 76. The secondary forces on the steering wheel were

perpendicular to the wheel and represented a pulling or pushing motion.

 

 

 

 

 

 

 

 

 

 

Figure 76: Secondary forces measured under the buttocks, thighs and feet
and behind the thorax and pelvis Fx and the wheel Fz (positive is in the
direction of the arrows).

When comparing the ranges of the shear forces in the static trials to those
of the dynamic files, the dynamic files generally had a larger range of force
distribution, Table 66. The buttocks plate was the only exception, where the
range of shear motion for the dynamic trials was similar to the static trials. For all

other support plates, the dynamic condition produced ranges two to four times

164

larger than the static files, except behind the pelvis where the increase in range

was as much as seven times greater than the static files.

Table 66: Secondary forces for both the static and dynamic tests in

 

 

 

 

 

 

 

Newtons.

Newtons Static Tests Dynamic Range

Tests Difference
Recline 1 Max Min Range Max Min Range Dyn-Stat
Thigh 21.8 -4.6 26.4 37.1 -12.0 49.1 22.7
Buttocks 16.5 -39.9 56.4 7.6 -47.4 54.9 _-1 .5
Pelvis 16 12.7 3.3 29.6 5.5 24.1 20.8
Thorax 19.9 -8 27.9 36.4 -14.8 51.2 23.3
Foot 57.3 38.1 19.2 79.2 6.6 72.6 53.4
Wheel (F2) -3 -9 6 7.1 -15.1 22.1 16.1
Recline 2
Thigh 18.9 -1.3 20.2 29.4 -13.2 42.7 22.5
Buttocks 7.4 -40.6 48 0.6 -54.1 54.6 6.6
Pelvis 22.5 17 5.5 29.1 5.9 23.2 17.7
Thorax 22.2 -9.4 31.6 32.7 -12.8 45.4 13.8
Foot 57.3 32.6 24.7 71.0 3.7 67.2 42.5
Wheel (Fz) 0.5 -9.4 9.9 14.3 -7.0 21.3 11.4
Recline 3
Thigh 13.4 -0.4 13.8 27.4 -11.9 39.3 25.5
Buttocks 4.2 -50.6 54.8 -1.4 -47.9 46.5 -8.3
Pelvis 21.8 18.9 2.9 29.5 7.1 22.4 19.5
Thorax 20.8 -4.1 24.9 29.4 -15.0 44.4 19.5
Foot 57.1 31.5 25.6 71.5 3.3 68.2 42.6
Wheel (F2) 2.3 -9.6 11.9 17.0 -6.1 23.0 11.1

 

 

 

Since these trials were dynamic, the subject was not able to readjust his

body as the chair moved. For the static trials, after each chair motion, the

subject was encouraged to lift his body off the seat and reposition it to reduce the

shear loading and to make sure his body was positioned with the buttocks in

contact with the pelvis support. Therefore, the higher shear forces would be

expected in the dynamic conditions, since the subject was not able to adjust his

body to reduce the shear forces as the seat was moving. Although slightly larger

165

 

shear forces were expected, the magnitude of increase was higher than
expected solely due to the inability of the subject to adjust his position. Another
possibility for the higher shear forces, was that the chair did not move exactly
with the body as will be shown later in this section, and this discontinuity added

to the increase in the shear forces between the static and dynamic conditions.

Table 67: Statistically significant differences of secondary forces for the
static tests between the slumped and super erect chair positions.

 

 

 

 

 

 

 

 

 

Statistically Significant Differences
Rec 1 Rec 2 Rec 3

Thigh Y Y Y
Buttocks Y Y Y
Pelvis - - -
Thorax Y Y Y
Foot Y Y Y
Wheel Y Y Y
(F2)

 

 

 

 

 

Table 68: Statistically significant differences of secondary forces for the
dynamic tests between the maximum and minimum values.

 

 

 

 

 

 

 

 

 

Statistically Significant Differences
Rec 1 Rec 2 Rec 3

Thigh Y Y Y
Buttocks Y Y Y
Pelvis Y Y Y
Thorax Y Y Y
Foot Y Y Y
Wheel Y Y Y
(F2)

 

 

 

 

 

Table 67 and Table 68 list the statistically significant results for the
secondary forces in both the static and dynamic test conditions. The statistical
analysis showed a significant difference between the maximum and minimum

values of the secondary forces in the dynamic tests for all support plates, Table

166

68. These findings were consistent with the statistical analysis of the static test
trials, except for the case of the support behind the pelvis, where a statistically
significant difference was not found in the static tests, Table 67.

Figure 77, Figure 78 and Figure 79 are sample graphs (subjects randomly
chosen) depicting the trends seen in the secondary force data for the dynamic
test conditions.

For the majority of the subjects, the visual analysis of the trends in the
secondary data showed that the pelvis, thighs and feet had positive shear forces
(anterior in direction when acting on the chair) while the chair was in a maximum
openness (SE) position. The shear forces reversed direction to a posterior force
on the chair while in the slumped position.

The wheel forces also followed the movement of the chair. A pulling
action of the wheel toward the body occurred in the super erect position and this
pulling force decreased as the chair moved into a slumped position. This effect
would seem reasonable since the shoulders were further away from the wheel in
the super erect position as compared to the slumped position.

The changes of the thorax shear forces occurred with the majority of the
subjects having a negative shear force in the super erect chair position (force on
the chair pointed in the superior direction) and a positive shear force in the
slumped chair position.

The shear force on the buttocks varied and was difficult to identify a

consistent trend in the force data relative to the chair position. However, the data

167

showed a statistically significant difference between the maximum and minimum

shear forces, but the location of those forces with respect to chair position varied.

Secondary Forces: 9drec2b

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

150 Super Erect —- Slumped ‘
I
I
100 «-— — 2 - w— — ~ .
lncreasrng Force
Magnitude
9 .
O
C I
1‘ I
§ I
'2 I
I
l ——-— Buttocls Fx 1
-100 j—T'nigh Fx
2 - "Tb , , .‘ "'~‘f‘F9°'F§
Secondary Forces: 9drec2b
150 Super Erect L Slumped
100
l Increasing Force
1 ‘3. 50 I — — — — 2 - 2_ Magnitude
5
8 I
5? ° ‘
-50 , , . _, ,, . , . __ , . ..- Wmfi
.—O—Chair TSA
——I—-Thorax Fx . ‘ V
-100 —D—Behind Pelvis Fx’ ‘
“Hie . . ‘---—-,-:.-... Wheel F2 1

,2. 1

Figure 77: Secondary forces from Subject 9, in recline 2, dynamic
movement from neutral to super erect to slumped back toward super erect.
All force data is displayed with respect to the chair TSA (degrees).

168

Secondary Forces: 11drec28

 

 

 

 

 

 

 

 

 

 

  
 
   

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

150 Super Erect _— Slumped
increasing Force
Magnitude
.9
D
C
<
it
12
+ Buttocls Fx
"00 ‘——-—Thigh Fx
Time -— Feet Fx
Secondary Forces: 11drec2a
Super Erect Slumped
150
I
100 I g
j Increasing Force
%, 50 . -- -- -... _2-m,‘_w L.__ it - Magnitude
I c [
‘4:
E; I
.2 I
50 - 7 , 7, - , ._ 7 ,, -1--.e 1
I-I— Thorax Fx :
-100 —u— Behind Pelvis Fx
“"21 Whee- F2

 

Figure 78: Secondary forces from Subject 11, in recline 2, dynamic
movement from neutral to super erect to slumped back toward super erect.
All force data is displayed with respect to the chair TSA (degrees).

169

Secondary Forces: 13drec2a +681 TSA ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
   

 

 

 

 

 

 

 

 

 

 

150 Si 5 E t ‘____. ~—E~— Buttocks Fx
I umped uper rec Thigh Fx
.2
a
5
9
I2 0 ‘sor - rt fl ...‘I’I. v “—1 w fl
\9
.50 _--,. __ _ _ _ _ pp Increasing Force
Magnitude
.00 J I
Time I
Secondary Forces: 13drecza
150 - Slumped Super Erect
100 4 1
I
.3 50 , p _ _ W, _p- h , Increasing Force
5 ' Magnitude
8 ' :1 ”1'"-
:8- o nu
-50 - a _ .2.
+Chair TSA ;
+Thorax Fx v
"00 —a— Behind Pelvis Fx

Figure 79: Secondary forces from Subject 13, in recline 2, dynamic
movement from neutral to slumped to super erect back toward slumped. All
force data ls displayed with respect to the chair TSA (degrees).

170

Lateral Forces

The lateral forces were the side to side forces on the support plates.

Since the motion of the chair and the subject occurred in the sagittal plane, these

forces were expected to be small and not have significantly different results

between torso articulations. The average maximum and minimum force data for

both the static and dynamic tests are presented in Table 69.

Table 69: Lateral Forces for both the static and dynamic tests in Newtons.

 

 

 

 

 

 

 

Newtons Static Tests Dynamic Range
Tests Difference
Recline 1 Max Min Range Max Min Range Dyn-Stat
Thigh 0.4 -1.3 1.7 1.9 -3.8 5.7 4.1
Buttocks -0.9 -2.0 1.0 3.6 -4.2 7.9 6.8
Pelvis -1.0 -1.3 0.3 0.6 -2.2 2.9 2.5
Thorax 05 -3.2 2.7 1.5 -5.9 7.4 4.8
Foot 04 -0.8 0.4 1.6 —3.5 5.1 4.6
Wheel 1.9 1.6 0.3 3.8 02 4.0 3.7
Thigh -0.3 -1.9 1.6 3.0 -5.1 8.0 6.5
Buttocks -0.4 -2.4 2.0 2.8 -6.1 8.9 7.0
Pelvis -1.0 -1.3 0.3 0.8 -1.8 2.7 2.3
Thorax -0.2 -1.7 1.5 1.8 -5.4 7.2 5.7
Foot 04 -1.1 0.7 1.4 -3.7 5.1 4.4
Wheel 1.8 1.6 0.2 4.0 -0.1 4.1 3.9
Recline 3
Thigh 0.8 -1.9 2.7 2.6 -5.1 7.6 4.9
Buttocks 0.9 -1.7 2.6 2.6 -6.0 8.6 6.0
Pelvis -0.6 -1.0 0.5 0.8 -1.6 2.4 1.9
Thorax -1.8 -3.0 1.2 1.9 -5.3 7.2 6.0
Foot 03 -1.0 0.7 1.5 -3.5 5.0 4.3
Wheel 1.8 1.7 0.1 3.9 0.0 4.0 3.8

 

 

 

As demonstrated in Table 69, the dynamic tests consistently produced

larger lateral forces than the static tests for all recline angles and all support

171

 

plates. This was also the case when the primary and secondary forces were

analyzed between the static and dynamic test conditions.

Table 70: Statistically significant differences of lateral forces for the static
tests between all articulations. (Y) means statically significant differences
found, (-) means no statistical difference found.

 

Lateral Support Forces (Fy)
Recline 1

 

S vs. SE N vs. SE E vs. SE S vs. E N vs. E S vs. N
Thigh - - - - - -
Buttocks - - - - - -
Pelvis - - - - - -
Thorax Y - - - - -
Feet - - - - - -
Wheel - - - - - -

 

Recline 2
Svs. SE st. SE Evs. SE Svs. E st. E Svs. N

 

Thigh - - -

Buttocks - - - Y - -
Pelvis - - - - - -
Thorax - - - - - -
Feet - - - - - -
Wheel - - - - - -

 

Recline 3
Svs. SE st. SE Evs. SE Svs. E st. E Svs. N

 

 

Thigh Y - - - - -
Buttocks - - - - - -
Pelvis - - - - - -
Thorax - - - - - -
Feet - - - - - -
Wheel - - - - - -

 

 

Table 71 : Statistically significant differences of lateral forces for the
dynamic tests between the maximum and minimum values.

 

 

 

 

 

 

 

 

Statistically Significant Differences
Rec 1 Rec 2 Rec 3

Thigh - Y -
Buttocks - - -
Pelvis Y - -
Thorax - Y -
Foot - - —
Wheel Y Y -

 

 

 

 

 

 

172

Since the maximum and minimum lateral forces were not always found at
the slumped or super erect chair position in the static tests, all chair articulations
were listed and compared. For the static tests, there were no consistencies
across recline angles that determined a statistically significant difference
between lateral forces for any support plate. Only three cases, recline 1 for the
thorax, recline 2 for the buttocks, and recline 3 for the thigh showed differences
between lateral forces. The lack of statistically significant results across
conditions and recline angles clearly demonstrated that the lateral forces did not
change for the static tests.

For the dynamic tests, the only plate that showed a statistically significant
difference for more than one recline angle was the support plate attached to the
steering wheel. Since the subject was moving for the entire test it is reasonable
that the wheel would see a change in lateral forces; the wheel would help
stabilize the subject as the body was moved. For the pelvis, thigh and thorax
support plates only one recline position showed a statistically significant
difference in the lateral forces. Again, it was concluded that overall, the lateral

forces did not vary significantly as the chair moved dynamically.

Overall Comments on Dynamic Force Data
In general the dynamic movement trials induced larger forces for the

primary, secondary and lateral forces. The only exception were the secondary
forces under the buttocks for recline 1 and recline 3 where the forces were larger

in the static conditions by 1.5 N and 8.3 N.

173

The primary forces for the dynamic tests showed no statically significant
differences between the maximum and minimum force values, unlike the static
tests where the primary forces associated with the buttocks, pelvis and feet
clearly showed statistically significant. From an examination of recline 2 data,
the maximum and minimum primary forces occurred near the maximum or
minimum chair articulations for the thighs, pelvis, feet and wheel, but had mixed
locations for the buttocks and thorax plates. For the conditions where the chair
position was mixed, it was difficult to make comparisons between the static and
dynamic data. For the static tests, the maximum and minimum primary forces
were located at the chair end points. Even though the dynamic primary forces
did not show a statistically significant difference in magnitude, as discussed in
the next section, a difference in the location of the resultant force on the support
plates was related to changes in TSA.

The secondary force data produced similar results in terms of statistical
significance between the static and dynamic tests. Statistically significant
differences were shown to occur between the maximum and minimum secondary
forces for all recline angles, and these forces primarily occurred at or near the
extreme chair positions.

For the lateral forces, the majority of support plates in the various recline
positions showed that there was no statistically significant difference between the
maximum and minimum lateral forces. The lateral force maximum and minimum
values occurred in various chair locations and were not consistently identified

with slumped or super erect chair positions.

174

In general the primary forces were much larger than the secondary forces
or the lateral forces. The secondary forces were typically only 9-14% of the
magnitude of the primary forces, except for the pelvis and foot plates where the

secondary forces were 60% of the primary force values.

Resultant Forces
Another point of interest when examining the support force data, aside

from the magnitude of the forces, was path of travel of the resultant force. As
stated earlier in this dissertation, one reason to look at the path of the resultant
force was to examine the shift in the center of force. If a chair could be designed
to shift the center of force, this chair design may be helpful in the prevention of
decubitus ulcers. It is unclear as to how much of a shift in the distribution of the
force is necessary to prevent the ulcers altogether, but it is known that postural
change and force shifts reduce the occurrence and severity of the ulcers.

Figure 80 through Figure 83 show sample plots of the shift of the location
of the resultant force in the XY plane, or the plane of the support plate. The
graphs are labeled in terms of superior or anterior relative to the subject. Notice
that the travel of the resultant force was primarily in the superiorfinferior
directions for the thorax and pelvis supports and in the anterior/posterior
directions for the thigh and buttock supports. This type of movement coincided
with the sagittaly symmetric task and the small lateral forces relative to the

primary and secondary forces.

175

The location of the resultant force in the XY plane was computed using the
moment equations, Equation 18, and solving for the X and Y distances to the
forces.

X=(Fx *Z)/Fz —My /FZ

'Y=(Fy *Z)/Fz +Mx/FZ
Equation 18: Calculation of the resultant force In the XY plane. 2 values
were between 54 and 55 mm which included 33 to 34 mm from surface of

force plate to center of plate, 18 mm for board thickness and 3 mm for
compressed foam thickness.

Thorax: Location of Resultant 11drec3b superior

 

100 -
50 + erect
-300 -200 -100 100 200 300
.50 .
slumped
-100 -

 

 

 

-150 -

Figure 80: Location of resultant force for the thorax plate during dynamic
chair movement from slumped to erect.

176

Behind Pelvis: Location of Resultant Force 11drec3b

 

 

 

 

 

superior A

80 .

60 .

40 .

20' slumped

-300 -200 -100 _20 0 100 200 300
..l
erect
.60 .
-80

Figure 81 : Location of resultant force for the pelvis plate during dynamic
chair movement from slumped to erect.

Thighs: Location of Resultant Force 11drec3b A

 

 

 

anterior
A 100 - A
50 -
erect
. . 0 . E - . ,7
-300 -200 -100 o 100 200 300
-5
1 slumped
-100 -
450

Figure 82: Location of resultant force for the thigh plate during dynamic
chair movement from slumped to erect.

177

Buttocks: Location of Resultant 11drec3b
150 4—

 

anterior
I
50 , slumped
, E . E A G . A , ,. . . ,. ,, ,
-300 -200 -100 0 100 200 300 .
-50 - erect

-1flfl .

-150 -

Figure 83: Location of resultant force for the buttocks plate during dynamic
chair movement from slumped to erect.

Table 72: Location of the resultant force with respect to the center of the
force plate in the X direction. X direction is positive inferior behind the
thorax and pelvis, and positive anterior under the buttocks and thighs.

Location of Resultant in the X direction

 

 

 

(mm)

Thorax Pelvis Buttocks Thigh Foot
Rec 1 Ave range 58.5 18.3 27.4 41.9 59.3
Std 17.5 14.6 8.5 20.2 24.9
Rec 2 Ave range 50.1 17.6 28.2 36.4 57.1
Std 16.9 11.9 8.6 14.0 31.8
Rec 3 Ave range 44.0 16.3 29.6 36.1 58.1
Std 15.0 12.4 9.3 11.1 26.8

Average
51 17 28 38 58

 

 

 

 

 

178

Table 73: Location of the resultant force with respect to the center of the
force plate in the Y direction. The Y direction is positive toward the left
side of the body while seated in the chair.

 

 

 

Location of Resultant in the Y direction
(mm)
Thorax Pelvis Buttocks TIMh Foot
Rec 1 Ave range 6.6 14.3 4.7 12.2 15.4
Std 3.0 9.5 2.3 13.8 9.3
Rec 2 Ave range 7.3 14.9 5.5 13.4 18.8
Std 3.4 10.9 2.8 7.6 16.9
Rec 3 Ave range 6.1 13.4 5.7 12.6 17.0
Std 2.9 11.5 3.1 9.5 12.9

 

 

 

 

Table 72 and Table 73 list the average range of movement of the resultant
force in the X and Y directions for a given recline through the full range of TSA’s.
When looking at the ranges of movement of the resultant force in terms of the Y
direction (lateral), the ranges were consistent within a given support plate across
all recline angles, meaning the distance of travel did not vary with the recline
angle. The same observation was made with respect to the motion in the X
direction (anterior/posterior or superior/inferior) for all support plates except the
thorax. The thorax travel path decreased in length as the recline angle of the
chair increased. This trend may be related to the requirement to maintain
contact with the steering wheel. As the seat was reclined, the thorax could have
been pulled slightly forward, away from the top of the support plate. An attempt
was made to prevent this by allowing the subject to adjust the steering wheel

position fore and aft for every trial.

179

For the location of the resultant force behind the pelvis, the lateral motion
(Y direction) was similar in magnitude to the superior/inferior motion (X direction).
Thus, the path produced for the pelvis looked more like a circle rather than a line
trace. For all other support plates, the Y travel range was much smaller than the
travel range of the force in the X direction. These data support that the
Biomechanically Articulating Chair does change the location of the center of force
under the thighs, buttocks and feet and behind the thorax with the change in
torso articulation. The question now arises, is this beneficial to the body, and
does the chair provide a large enough movement in the travel of the center of
force to make a significant difference for a seated individual?

From earlier discussions of the physiology of the body, the answer to the
first question has been addressed. it is clear that movement of the body has
many beneficial aspects to the physiology of the body. It is unknown as to the
degree of movement that is necessary to prevent decubitus ulcers, but it is
known that any movement reduces the pooling of blood and waste products in

the body which is desirable and beneficial.

Equilibrium
The final analysis of the force data for the dynamic tests looks at the

overall summation of the forces in the vertical and horizontal direction. Prior to
performing the analysis of the internal joint force estimation, the static data were
first examined to determine if the subject was in equilibrium. The dynamic data
were not examined for internal joint estimations, however the initial step to

evaluate the equilibrium conditions based on the measured data was completed.

180

To evaluate if the subject was in equilibrium, all of the support forces were
converted to the laboratory horizontal and vertical components. The summation
of the forces in the horizontal direction should result in a value near zero, while
the summation of the vertical forces should result in a value close to that of the
subject's weight.

Figure 84 and Figure 85 show sample data of the equilibrium plots for two
subjects during dynamic tests. As seen in the sample data, similar trends were
present for the other subjects: the horizontal force data was not equivalent to
zero for the entire test, and the forces deviate from zero in a wave pattern,
relating to the chair motion. When a best fit line analysis was performed, for the
majority of the data, the y intercept of the line was near zero with a slope close to
zero.

If the vertical force data were in equilibrium, then as stated above, these
data should sum to that of the subject’s weight. Subject 11 weighed 166 lbs, or
739 N while subject 16 weighed 173 lbs, or 770 N. Since the slope of the best fit
line was near zero, the y intercept value of the best fit line was used as and
average value of the vertical force summation. Subject 11’s best fit line intercept
data had a difference of 77 N or 17 lbs from the body weight, while subject 16
had a 39 N difference, or 8.8 lbs. As discussed with the static data these data

are within reasonable measurement error of the system.

181

Equilibrum: Lab Horizontal

 

 

 

 

 

 

 

 

11drec3b
50.00
40.00 » I
g 30.00- ‘3‘.
o ,3
I
3 20.00 9
O Q .
.5 9’ s
§ 10.001...
- Q
u°.. 0.00 » g \ a
. O
-10.00 - :9. *7
e
20.00
y .. -0.0793x +12.356
Frame Number R2=8E-06
Equilibrum: Lab Vertical
11drec3b
-780.00 a
y=2 00012-81655
790.00 I R "8506
in I
P -800.00 .1
S I I I #
= -810.00 u I.-
75: I "' . '4' .III- "
18 -820.00 I .‘ r- ' I
I - I.
I ' I I I
83000 — l
I
84000 -

Frame Number

Figure 84: Equilibrium analysis of subject 11, recline 3 trial 2 depicting the
vertical and horizontal summed force data, along with the information from
a best fit line analysis.

182

Equilibrum: Lab Horizontal

 

 

 

 

16drec28
30.00
20.00 v
. 4‘ , '
10.00
5 *' ’. O 3
3 0.00
£ -10.00 t .
g 20.00 0 g
12 '30-001' ‘0‘ y2=-0.0242x—1.10
-50.00
Frame Mmber

Equilibrum: Lab Vertical
16drec2a

 

“

C

2

3

Q

.5

0

§

12

v = -0.0433x -809.25

35000“ I R2=0.0088
-860.00- ——----- — -

Frame Mun ber

Figure 85: Equilibrium analysis of subject 16, recline 2 trial 1 depicting the
vertical and horizontal summed force data, along with the information from
a best fit line analysis.

183

Table 74 lists the average data from the best fit lines for all three recline
angles. The slope of the best fit line through the data (for the horizontal forces)
should be near zero, and on the average as seen in Table 74 the slope for all
three reclines was near zero. The average intercept ranged from 7.6 N to 13.4 N
(1.7 to 3.0 lbs); so on average the horizontal force data depicted subjects that
were at or near equilibrium.

If in equilibrium, the vertical force data should have a slope near zero, with
an intercept near the average weight of the subjects. The average weight of all
subjects was 752 N or 169 lbs. The average vertical force data showed larger
deviations from the desired averages, particularly in recline 2, (Table 74), but
these data were near equilibrium requirements; close enough to make the
assumption that the majority of the subjects were in equilibrium with regard to the
vertical forces. if performing an analysis using the requirement of the forces
being in equilibrium, such as that of the internal joint analysis, better results could

be obtained by evaluating the information on an individual subject basis.

Table 74: Averages and standard deviations of best fit line slope and
intercept for the equilibrium horizontal and vertical force data.

 

 

Horizontal Vertical
slope intercept slope intercept
Recline1 Ave 0.02 7.6 0.04 -749.2
Std 0.20 10.8 0.26 141.5
Recline2 Ave 0.04 9.5 0.36 -780.9
Std 0.31 13.4 1.53 59.6
Recline3 Ave 0.01 12.7 -0.03 -747.3
Std 0.57 17.3 0.27 132.0

 

 

 

 

 

184

Torso and Chair Motion for Dynamic Tests
This section will evaluate the ranges of movement of the body, specifically

the thorax and pelvis during the dynamic trials. First the thorax and pelvis
motions will be analyzed, then the movement of the body will be compared and
related to the movement of the test chair. The motions of the limbs (elbow angle

and knee angle) were not analyzed for the dynamic data.

Thorax
The average maximum and minimum thoracic angles for each recline

angle in the dynamic tests are presented in Table 75, along with the average
maximum and minimum thoracic angles for the static tests. The thoracic angle
was measured by forming a vector from the mid-sternum to the sternal notch and
comparing the angle this vector made with laboratory vertical. Recall that in the
static tests, the maximum angles, or the largest deviation from vertical occurred
predominantly in the super erect postures and the minimum thoracic angles
occurred in the slumped chair position. For the dynamic tests, as seen in the
above section discussing forces, the maximum and minimum thoracic angles
were not identified with an exact chair position, but rather a range of positions,
usually near the end ranges of chair motion.

When examining the thoracic angular data, the angle increased as the
recline angle of the chair increased, this was a function of the calculation and
was expected. Overall, the averages between the static and dynamic tests were
similar in terms of magnitude, with the ranges generated by the dynamic testing

larger than those for the static testing (by 3.5 to 4.5 degrees) Again, the larger

185

ranges were probably due to the fact that the chair movement for the dynamic

testing was slightly larger than that of the static testing.

Table 75: Measure of Thoracic Angle in degrees during dynamic motion
from a slumped position to a super erect position. All values are negative,
meaning the thorax is rearward of laboratory vertical. (Maximum equals the

maximum deviation from vertical.)

 

 

 

 

 

 

 

 

 

 

Thorax Dynamic Thorax Static
(Degrees)

Recline Recline Recline Recline Recline Recline

1 2 3 1 2 3
Min 323 -35.5 -39.6 -34.5 -36.4 -41.7
Min std 8.8 9.8 9.6
Max -39.8 -43.7 -48.0 -38.5 -41.1 -45.6
Max std 8.5 9.7 9.6
Range 7.5 8.2 8.4 4.0 4.7 3.9
Rng std 3.0 3.6 3.0
Pelvis

The average maximum and minimum pelvic angles for each recline angle
in the dynamic tests are presented in Table 76 along with the average maximum
and minimum pelvic angles for the static tests. Again, the pelvis angle was
measured by forming a vector from the sacral target to a virtual point formed mid-
way between the left and right ASIS targets. The angle between this vector and
the laboratory x axis (horizontal pointing anteriorly) was the measure of the pelvic
angle. Using the right hand rule, the angle was a negative angle if located above
the horizontal plane. In the static tests, the maximum angles (largest deviation
from horizontal) occurred in the slumped conditions and the minimum pelvic

angles occurred in the erect and super erect chair positions.

186

Like the thorax angles, the pelvic angles increased as the recline angle of
the chair increased. This was related to the pelvic angle calculation, which was
based on the motion of the pelvis relative to the x axis of the laboratory.

As with the thoracic angular data, the static and dynamic tests produced
average pelvic angles that were similar in terms of magnitude, however, the
ranges generated by the dynamic testing were slightly larger than those for the

static testing (by 3.1 to 3.9 degrees).

Table 76: Measure of Pelvic Angle in degrees during dynamic motion from
a slumped position to a super erect position.

 

 

 

 

 

 

 

Pelvis Dynamic Pelvis Static
(Degrees)

Recline Recline Recline Recline Recline Recline

1 2 3 1 2 3

Min -24.6 -27.5 -33.5 -26.9 -30.2 -36.1
Min std 8.1 11.7 12.0
Max -33.7 -36.8 -40.9 -32.1 -36.1 -40.3
Max std 12.4 10.2 12.7
Range 9.1 9.3 7.3 5.2 5.9 4.2
mgstd 7.0 10.1 3.1

 

 

Chair vs. Body: Openness
One of the issues of interest during testing of the subjects was to evaluate

how well the person moved with the chair. To accomplish this evaluation, the
movement of the body or Openness was cross plotted with respect to the
movement of the chair or Total Support Angle (TSA) during the dynamic tests,
and a linear regression using a least squares method was applied to the data.
The TSA and Openness were measures of articulation, one based on chair
motions and the other based on body motions. If these two moved together, then

the cross plot would be in the form of a straight line. If the body moved with the

187

chair, the slope of the line would be near one. This would mean that as the TSA
of the chair varied by five degrees, the body openness also varied by five
degrees. Sample cross plots from two subjects can be seen in Figure 86 and
Figure 87 while Table 77 lists the average slope and y intercept values across
subjects. The averages and ranges of the chair and body motions are listed in

Table 78.

 

 

 

 

 

TSA vs Openness
11drec3b
110
a
Q 105
e y = 0.5084x + 48.025
g ‘00 R2 = 0.9703
3’. 95
0
C .
:90 a»
8
85
50 70 90 110 130

Total Support Angle (Deg)

 

 

 

Figure 86: Chair TSA vs Body Openness for subject 11 in recline 3.

188

 

TSA vs Openness
20drec2a

120

 

y = 0.3563x + 78.289

115 R9 = 0.9087

59'
“a?
”1

50 70 90 1 10 1 30
Total Support Angle (Deg)

Openness Angle (Deg)

 

 

 

 

 

 

Figure 87: Chair TSA vs Body Openness for subject 20 in recline 2.

Table 77: Average data from best fit line through a cross plot of the chair
Total Support Angle (T SA) vs the body Openness. r2 (Pearson product
moment correlation coefficient) is a measure of linear relationship between
the two.

 

 

 

 

TSA vs Openness for all three

reclines during Dynamic Tests
Recline Recline Recline

1 2 3

r2 0.89 0.88 0.85

std 0.15 0.17 0.18

Slope 0.35 0.43 0.43

std 0.27 0.15 0.13

 

 

 

189

1...“?

 

Table 78: Measure of body Openness and chair articulation (Total Support
Angle, TSA) in degrees during dynamic motion from a slumped position to
a super erect position.

 

 

 

 

 

 

 

Openness Chair
TSA
Recline Recline Recline Recline Recline Recline
1 2 3 1 2 3
Max 104.9 105.5 103.3 113.6 112.3 113.0
Max std 15.1 16.1 22.5 3.0 3.9 4.3
Min 88.6 90.9 88.7 78.7 80.5 81.4
Min std 21.0 18.4 22.3 4.2 5.2 5.8
Range 16.3 14.6 14.7 34.9 31.8 31.6
Rng std 12.5 5.5 5.1 6.1 6.5 6.8

 

 

As can be seen from the sample plots (Figure 86 and Figure 87), the data
were linear, however when analyzing the best fit line, the slopes of these lines as
seen in the sample plots were not one to one. This same trend was supported
by the average of the subject data seen in Table 77, where the squared
correlation coefficient (r2) ranged from 0.85 to 0.89, showing a strong relationship
between the subject’s openness and the chair articulation. The average slope of
the lines ranged from 0.35 to 0.43. These data supported the fact that the chair
was moving the body, however it was not being moved at a ratio of 1:1, but
rather at a ratio closer to 2:5. Meaning the subject moved only 40% of the
amount the chair moved. From the range data listed in Table 78, the average
movement of the subject relative to the chair was a little higher, 46% to 47%, or

roughly a ratio of 1:2.

190

These numbers support the conclusions from the static data, that the chair
motion had to be large, or greater than a 10 degree movement, in order to

produce a statistically significant change in the openness of the subject.

Chair vs. Body: Recline
As presented in the methods section, the measurement of the Body

Recline Angle (BRA) was accomplished by establishing a vector from the mid-
point of the ASIS targets to the sternal notch target. It was thought that this
measure would be stable during articulation of the body from a slumped to an
erect posture. The vertical range on the cross plots in Figure 88 and Figure 89,
the BRA measure was slightly affected by the torso articulation. The BRA had an
average range of movement of 6.3 degrees in reclines 1 and 2, and 5.4 degrees
in recline 3, (Table 79). The average range of the chair movement was much
lower, 1.3 degrees in recline 1 and 0.7 degrees in reclines 2 and 3. The offset
between the average chair recline and the average body recline remained
consistent across the three recline angles, Table 80. The question now arises,
does the method of measuring the body recline from these external landmarks
provide a “good enough” method for establishing the recline angle of the body?
An alternative approach for establishing a body recline angle would be to
estimate internal joint centers at two points in the spine and use that as a
reference system for measuring the body recline. However, this method also has
disadvantages, primarily that it is a more complex measurement, requiring more
external targets. A second disadvantage would be that internal joint points have

to be estimated from averaged cadaveric data, as in the internal joint analysis

191

 

performed for this dissertation. Due to the fact that a body of averaged cadaveric
data would be used on an individual with anthropometry that is similar, but most
likely not identical to the averaged data would lead to errors. For example, a test
subject might have a long torso for his stature, while the data applied to every
subject would be that of a person with an average torso.

it is the feeling of this author that with the large variability of subject
anthropometry and movement patterns, the external target method for defining
the recline angle of the body in the types of movement patterns tested provided
reasonable results, as reasonable as can be expected when dealing with the

human body.

Table 79: Measurement of the Body Recline Angle and Chair Total Recline
Angle in degrees for three recline angles. (Maximum angle equals the
maximum deviation from vertical)

 

 

 

 

 

 

 

BRA Chair
TRA
Recline Recline Recline Recline Recline Recline
1 2 3 1 2 3

Min 12.2 16.4 20.8 -40.0 -45.1 -49.0
Min std 3.4 4.0 4.8 2.6 0.8 1.1
Max 18.4 22.8 26.3 -41.3 -45.7 -49.7
Max std 4.0 4.1 5.5 2.6 0.9 1.1
Range 6.3 6.3 5.4 1.3 0.7 0.7
flg std 2.7 3.5 2.5 2.6 0.5 0.4

 

 

Table 80: Offset between the BRA and TRA average maximum and
minimum measures.

 

 

 

 

 

 

 

 

Lngpes) Recline 1 Recline 2 Recline 3
Min 278 -28.7 282
Max -22.9 -22.9 -23.4 g

 

192

Chair Recline vs Body Recline Angle
13drec2

 

Body Fbcllne (Deg)
83833838

71 -I
70

 

 

 

130 132 134 136 138 140
Chair ibcllne (Deg)

Figure 88: A cross plot of the body recline angle (BRA) vs. the chair recline
angle. The chair recline was held constant while the subject was moved
from a slumped to an erect position. Subject 13, recline 2.

Chair Recline vs Body Recline Angle
20drec2

75 - -—-- ~ m-

74 =

Body Fbcllne (Deg)
\l
O

130 132 134 136 138 140
Chair Fbcllne (Deg)

Figure 89: A cross plot of the body recline angle (BRA) vs. the chair recline
angle. The chair recline was held constant while the subject was moved
from a slumped to an erect position. Subject 20, recline 2.

193

 

Discussion:
From the previous analyses and discussions, it has been shown that the

dynamic data trials have provided new information from the static trials and in
some cases different conclusions. In the case of the primary force data, different
conclusions were arrived upon as compared to those found in the static tests. For
example, the primary force data from the dynamic tests did not show statistically
significant differences between maximum and minimum force values as did the
static tests; however, secondary and lateral force data produced similar results
and conclusions as the static test data.

New information from the path analysis of the resultant forces provided
insight as to how the loads were being distributed over time, and showed that the
chair shifted the location of the resultant forces on the support plates as TSA was
vaned.

An equilibrium analysis of the forces was also performed and showed that
the overall trial produced an average set of data that was near equilibrium. If
only using a portion of the dynamic test, that region should be reevaluated for the
equilibrium requirements.

The posture data was also analyzed in terms of dynamic motions and
similar results were obtained in comparison to the static tests, keeping in mind
that the dynamic test trials had slightly larger ranges of chair movement as
compared to the endpoints of the static test trials.

The ability to compare the movement of the chair to the movement of the

body was accomplished with the continuous data plots over the ten second trial

194

period. These data showed than on average, the body did not move with the
chair at an equal rate, but rather moved a portion of the chair motion, 40 to 47%.
Overall, the dynamic data were more complex than those of the static
data, and only an initial analysis of these data have been performed for this
dissertation. These data provide additional information and could be analyzed in

more detail, as a future topic of study.

195

Conclusions

Throughout the results section of this dissertation, interpretations were
presented regarding the meaning or usefulness of these data. The following
section provides general conclusions pertaining to these data while the remaining
two sections list major conclusions and findings from the static and dynamic

tests, respectively.

I. General Conclusions:
Not only were new methodologies developed for the measurement of

human posture, but also the body of data gathered for this dissertation was more
comprehensive than any other work previously available. The ability to measure
both static and dynamic conditions of body posture and support forces has been
newly developed for this work. The methods to break down seated posture into
primary and secondary motions (recline and torso articulation) as used by the
JOHN model, were first applied to the movement of the human body in this
research. Similarly, research measuring the support force data and developing
correlations between body posture and force distributions has not been
performed to this depth in any previous bodies of work. Until now, these unique
aspects of seating have not been studied.

In terms of seat design, it was interesting that the data showed a
combined movement of the thorax and pelvis supports (on the chair) be at least
20 degrees before statistically significant changes in a subject’s posture
occurred. This result will provide guidance to those interested in designing seats

that allow people to truly change their posture. From this information, itis likely

196

automotive seats that only have an adjustable lumbar support and do not allow I
other regions of the seat to move would not induce a significant change in a
person’s posture; primarily because the motion of the seat is too small.

In medical seating, particularly wheelchair design, these data showed that
an articulating seat could induce changes in posture and shifts in the location of
the resultant forces. A goal of a medical professional working with individuals
who are wheelchair-bound is to reduce the chances of an individual developing
decubitis ulcers under the buttocks. These data showed that a seat design which
articulates the torso may be beneficial in shifting support forces and reducing
these ulcers. The ability to change the posture of an individual who is
wheelchair-bound also affects other physiological aspects of the body such as
the ease of breathing and blood flow. There is no doubt that information of this
nature will be useful for advancement of wheelchair design.

Another important discovery was that the loading into the floor was
significantly changed as the posture was varied. This information is especially
important in environments where the foot-to-floor contact is a necessity such as
in the sewing industry or office industry. The change of loading into the floor also
verifies that measurements occurring in the chair alone, independent of the
environment in terms of support forces, do not provide a complete data set.
Therefore, thoroughness necessitates the measurement of the forces from
various surrounding contact zones.

As presented in the Literature Survey section, most researchers believe

that for a single position, the erect seated posture is the optimal posture for your

197

spine. This is based on many factors including increased ability to breath,

reduced muscle fatigue, reduced intervertebral disc pressure, etc. One problem I
with the seated erect posture is that people find it difficult to maintain this position
for a length of time. The static force data showed that in the erect seated
posture, the shear behind the thorax support was at a minimum. This researcher
theorizes it is this shear force that helps support the position of the torso, and
with a reduced shear loading onto the seat, the individual has to do the majority
of work to maintain the position.

Lastly, this study has developed a methodology for estimating internal joint
loading, demonstrated the method, and obtained reasonable results without
using invasive procedures. These results were then compared to published data
from Andersson58 that was gathered via an invasive procedure, and agreed well
with his results. As discussed earlier, these data may be useful in the physical
therapy of an individual with back injury. Studying a larger array of postures may
show a series of body positions that have lower internal spinal loading than
others.

. The availability of these data, combining both posture and force, will be
useful in seating design, including, but not limited to the office, automotive,
aerospace and medical industries. It is also anticipated that these data will be
incorporated into computer models that assist in the design of seats or
environments that include seated individuals. These data provide a significant

advancement in understanding seating mechanics.

198

ll. Static Data Conclusions:
The following section lists the major conclusions found upon analysis of

the static data.

1. The positioning of the Biomechanically Articulating Chair was repeatable in
terms of the test conditions: three recline angles and four torso articulations.

2. When comparing the primary forces (normal forces except for the steering
wheel which was a shear force) between the three recline angles, the forces
were significantly different, as follows: The buttocks, pelvis and thorax
showed significantly different forces between all three recline angles.
Second, the thighs showed significantly different forces between reclines 2
and 3 and reclines 1 and 3. Third, the feet showed significantly different
forces between reclines 1 and 2, and 1 and 3, and fourth, the steering wheel
showed significantly different results between reclines 2 and 3.

3. When evaluating trends that occurred when moving from an upright posture
to a reclined posture, it was clear that the body weight was transferred off the
feet, buttocks and thighs and onto the back of the pelvis and thorax as the
seat reclined.

4. When comparing secondary forces (shear forces except for the steering
wheel which was a normal force) between the recline angles, significant
differences occurred for forces under the buttocks between all three recline
combinations. Trends were apparent behind the pelvis, but were not

statistically significant.

199

. The shear forces behind the thorax were nearest zero for the erect condition

in all three recline angles.

. When examining the primary forces between various torso articulations
(slumped, neutral, erect and super erect) the thorax and wheel support plates
showed no significant differences between conditions. A consistent
significant difference was seen in recline 1 for the buttocks, pelvis and feet
between the slumped and super erect condition. The buttocks and feet
showed a difference between neutral and super erect for all recline
conditions, and between the slumped and erect conditions for reclines 1 and
3 only.

. With torso articulations, forces were shifted from the buttocks, thighs and
pelvis onto the foot plate, when the subjects moved from a slumped to a more
erect position.

. When evaluating secondary forces in terms of torso articulations, there were
no significant differences for the pelvis between any of the test conditions.
The secondary forces showed significant differences for test trials that were
not contiguous, however these differences were not always shown in all three
recline angles. The following statistical differences were found for the
secondary forces for all support plates except the pelvis: a) Slumped verses
super erect all reclines, b) neutral verses super erect for reclines 1 and 2 c)
slumped verses erect for reclines 1 and 3.

. The support force data was analyzed in terms of percent body weight

distribution (%BW). From a macroscopic viewpoint, lumping all recline angles

200

and torso articulations together resulted in the following distribution of forces:
a) 70%BW on the seat pan (buttocks and thighs), b) 30%BW on the seat
back (thorax and pelvis), c) 10%BW into the foot support and d) 4%BW into
the steering wheel. (These forces included both vertical and horizontal
components of the support forces, thus the %BW totaled more than 100%)

10.The movement of the chair in terms of Total Support Angle (TSA) was 30°
from the slumped to the super erect position. However, on average, the
subjects only exhibited a 10° change in torso openness from the slumped to
the super erect position. In an articulating chair such as the BAC, the
subjects exhibited a ratio of chair movement to body movement of

approximately 3 to 1.

III. Dynamic Data Conclusions
The following section lists the major conclusions found upon analysis of

the dynamic data.

11.The dynamic movement trials induced larger magnitudes than the static tests
for the primary, secondary and lateral forces. The only exceptions were the
secondary forces under the buttocks for recline 1 and recline 3 where the
dynamic data had a larger magnitude than the static data by 1.5 N and 8.3 N.
These values were approximately 2.7% and 16% of the total range of the
secondary forces respectively.

12.The primary forces for the dynamic tests showed no statistically significant

differences between maximum and minimum force values. These maximum

201

and minimum force values corresponded with the and ranges of chair
movement, slumped verses super erect.

13.The secondary force data produced similar statistical results between the
static and dynamic tests. These results occurred between the maximum and
minimum secondary forces, at or near the extreme chair positions.

14. For the lateral forces, the majority of support plates in the various recline
positions showed no statistically significant difference between maximum and
minimum lateral forces (or no relationship to chair position). This agreed with
the static data analysis.

15.The primary forces were larger than the secondary forces or the lateral
forces. The secondary forces were typically only 9-14% of the magnitude of
the primary forces, except for the pelvis and foot plate where the secondary
forces were 60% of the primary force values.

16.The subjects moved with the chair, and had correlation coefficients (r2)
ranging from 0.8510 0.89, showing a strong relationship between subjects’
openness and the chair articulation.

17. In the dynamic tests, the chair did not move with the body at a ratio of 1 :1,
but from examination from the slope of a cross plot between chair motion and
body motion, the ratio was closer to 2:5 (which differs from the 1:3 ratio seen
in the static tests). This meant the subjects moved only 40% of the range the
chair moved. Using the ranges of motion, rather than the slope of the line,
the movement was shown to be 46 to 47% of chair motion, or an approximate

ratio of 1:2.

202

 

18. Location of the resultant forces showed that the chair movement shifted the
location of the resultant force on the majority of the support plates under or
behind the body. Average travel of the resultant force in millimeters for each
plate across reclines was: a) thorax: 51, b) pelvis: 17, c) buttocks: 28 d)

thighs: 38 and e) foot support: 58.

Future Work
The study of seating mechanics has many aspects not yet investigated.

Additional data were collected during this project but were not analyzed for this
dissertation and will serve to be future work for this author. Subjective surveys
were taken during the various test conditions and will be analyzed for correlations
with the objective measures discussed here. Preferred seat and posture data
were collected. For the preferred position, the force and positional data were

measured and provide another body of data for future work.

203

References

 

‘ Grieco A. Sitting Posture: an Old Problem and a New One. Ergonomics. Vol.
29, No. 3, 1986. pp. 345-362

2 Grandjean E, and Hunting W. Ergonomics of Posture — Review of Various
Problems of Standing and Sitting Posture. Applied Ergonomics. Vol. 8, No. 3,
1977. pp. 135-140

3 Hobson D. Comparative Effects of Posture on Pressure at the Body Seat
Interface. Journal of Rehabilitation Research and Development. Vol. 29, No. 4,
1992. pp. 21-31

4 Magora A. Investigation of the Relation between Low Back Pain and
Occupation. Industrial Medicine. Vol. 39, No. 11, November 1970. pp. 31 -37

5 Magora A. Investigation of the Relation between Low Back Pain and
Occupation: 3 Physical Requirements: Sitting, Standing and Weight Lifting. .
Industrial Medicine. Vol. 41, No. 12, December 1972. pp. 5-9

6 Pope M, Andersson GB, Frymoyer F, Chaffin D. Occupational Low Back Pain:
Assessment Treatment and Prevention. Mosby Year Book, St. Louis ,MO. 1991

7 Zacharkow D. The Problems with Lumbar Support. Physical Therapy Forum.
September 10, 1990. pp 1-5 .

8 Troup J. Driver’s Back Pain and its Prevention: A review of the postural,
vibratory and muscular factors, together with the problem of transmitted road-
shock. Applied Ergonomics. Vol. 9, No. 4, 1978. pp. 207-214

9 Bush TR, Ekern DF, Hubbard RP. Methodology for Posture Measurement in
Automobile Seats: Experimental Methods and Computer Simulations: Published
through ISATA (International Symposium on Automotive Technology and
Automation). Paper No. 98SAF036. June 1998

\f/Faiks F, Reinecke S. Understanding the Kinematics of the Back and its
Relationship to Back Support. Pending Publication. 1999. pp. 1-13

1‘ Bush TR, Mills F, Thakurta K, Hubbard RP, Vorro J. The Use of
Electromyography for Seat Assessment and Comfort Evaluation. Society of
Automotive Engineers (SAE). Paper No. 950143, 1995 p. 27-32

‘2 Hubbard RP, and Gedraitis C. Biomechanically Articulated Chair Concept and
Prototypes. Society of Automotive Engineers. Paper No. 97051. 1997 p.117-124

204

 

‘3 Bush TR, Gregg S, Hubbard RP. Measuring and Modeling Support Forces of
People to Assist in the Development of the ASPECT Manikin Weighting. Society
of Automotive Engineers. Paper No. 1999-01-0961. 1999

‘4 Society of Automotive Engineers. Automotive Engineering Handbook.
Warrendale, PA. Society of Automotive Engineers, Inc. 1998

‘5 Bush N. Two-Dimensional Drafting Template and Three-Dimensional
Computer Model Representing the Average Adult Male in Automotive Seated
Postures. Master of Science Thesis. Michigan State University. 1992.

‘6 Reed M, Roe R, Schneider L. Design and Development of the ASPECT
Manikin. Society of Automotive Engineers. Paper No. 1999-010963. 1999

‘7 Setyabudhy R, Liu Z, Hubbard RP. Measurement and Analysis of Human
Thigh and Buttock Contours for ASPECT Manikin Development. Society of
Automotive Engineers (SAE). Paper No. 199901-0964. 1999

‘8 Netter F. Atlas of Human Anatomy. Pharmaceuticals Division, ClBA-GEIGY
Corporation, West Caldwell, New Jersey. 1993.

‘9 Moore K. Clinically Oriented Anatomy. Williams and Wilkins, Baltimore, MD.
1985.

2° Bao O, Yuan H. The Artificial Disc. Science and Medicine. January/February
1998. Pp. 615

Keegan J. The Medical Problem of the Lumbar Spine Flattening in Automobile
Seats. Society of Automotive Engineers. Paper No. 640788. 1964 p.57-66

22 Seymour M. The Ergonomics of Seating- Posture and Chair Adjustment
Nursing Times. Vol. 91, Is. 9, 1995. pp. 35-37

tf/Hedman TP, Fernie G. Mechanical Response of the Lumbar Spine to Seated
Postural Loads. Spine. Vol. 22, Is. 7, 1997. pp. 734-743

2‘ Bonney R. Some Effects on the Spine from Driving. Clinical Biomechanics.
Vol. 3, Is. 4, 1988. pp. 236-240

far”.
f/Andersson GB, Murphy R, Ortengren R, Nachemson A. The Influence of

Backrest Inclination and Lumbar Support on Lumbar Lordosis. Spine. Vol. 4, No.
1, 1979. pp. 52-58

205

 

f6 Andersson GB, Ortengren O, Nachemson A, Elfstrom G. Lumbar Disc
Pressure and Myoelectric Back Muscle Activity during Sitting: Studies on and
Experimental Chair. Scandinavian Journal of Rehabilitation Medicine. Vol. 6,
1974. pp. 104-114

, 7 Andersson G, Ortengren O, Nachemson A, Elfstrom G. Lumbar Disc Pressure
‘ and Myoelectric Back Muscle Activity during Sitting: Studies on a Car Drivers
Seat. Scandinavian Journal of Rehabilitation Medicine. Vol. 6, 1974. pp. 128-133

28 Goonetilleke R. Designing to Minimize Discomfort. Ergonomics in Design.
July 1998. pp. 12—19

29 Philippe G. One Methodology to Evaluate Automotive Seat Comfort.
International Symposium on Automotive Technology and Automation. Paper
number : 95A1029 p. 231-241

30 Motavalli S, Ahmad F. Measurement of Seating Comfort. Computers and
Industrial Engineering. Vol 25, No 1-4. 1993. pp 419-422

“Thakurta K, Koester D, Bush N, Bachle S. Evaluating Short and Long Term
Seating Comfort. Society of Automotive Engineers. Paper No. 950144. 1995. pp.
33-37

32 Judic J, Cooper J, Truchot P, VanEffenterre P, Duchamp R. More Objective
Tools for the Integration of Postural Comfort in Automotive Seat Design. Society
of Automotive Engineers. Paper No. 930113, 1993. pp. 67-76

33 Andersson GB, Ortengren O, Nachemson A, Elfstrom G. Lumbar Disc
Pressure and Myoelectric Back Muscle Activity during Sitting: Studies on an
Office Chair. Scandinavian Journal of Rehabilitation Medicine. Vol. 6, 1974. pp.
115-121

3“ Nachemson A, Elfstrom E. lntravital Dynamic Pressure Measurements in
Lumbar Discs. Scandinavian Journal of Rehabilitation Medicine. Supplement 1,
1974. Almqvist and Wiksell, Stockholm.

35 Rebiffe R. An Ergonomic Study of the Arrangement of the Driving Position in
Motor Cars. Center Technique Renault, Rueil-Malmaison, 92, France. 1969 pp.
26-33

36Hubbard RP, Gedraitis C. Biomechanically Articulated Chair Concept and
Prototypes. Society of Automotive Engineers Technical Paper. No. 97051 . 1997
p.117-124

97 Zacharkow D. The Problems with Lumbar Support. Physical Therapy Forum.
September 10, 1990. pp 1

206

 

38 Oualisys Mac Reflex Software and Instruction Manual, Qualisys, 148 Eastern
Blvd. Ste. 110, Glastonbury CT 06033

39 3M Product Literature on 7610 High Gain Tape for Motion Systems. Screen
Products, Industrial Optics Project/ 3M St. Paul, MN 55144. Ph. 612-733-4403
and Industrial Tape Division, 3M PO. Box 358 Southfield, MI 48037, Ph. 1800-
241 -0184

4° ASPECT Task One Report. Not Published for the Public. 1996-1997

[A

f‘yAbraham S,.Johnson C, and Najjar F. Weight and Height of Adults 18-74
Years: United States, 1971 -74 Vital and Health Statistics, Series 11 , Number
208. DHEW Publications Nubmer 79-1656. Hyattsville, MD: US Department of
Health Education and Welfare. [National Health and Nutritional Examination
Survey of 1974 (NHANES ”)1.

I?) Roe, R. General Motors Packaging Study. Not Published for the Public.

43 Haas W. Geometric Model and Spinal Motions of the Average male in Seated
Postures. Master of Science Thesis. Michigan State University. 1989

‘4 Boughner B. A Model of Average Adult Male Human Skeletal and Leg Muscle
Geometry and Hamstring Length for Automotive Seat Designers. Master of
Science Thesis. Michigan State University. 1991

45Bush NJ. Two-Dimensional Drafting Template and Three-Dimensional
Computer Model Representing the Average Adult Male in Automotive Seated
Postures. Master of Science Thesis. Michigan State University. 1992

‘6 Hubbard RP, Haas W, Boughner B, Canole R, Bush NJ. New Biomechanical
Models for Automobile Seat Design. Society of Automotive Engineers Technical
Paper. No. 930110, 1993 p.35-42

‘7 Hubbard R, McLeod D. Definition and Development of a Crash Dummy Head.
Proceedings of the 18th Stapp Car Crash Conference, Ann Arbor, MI and
published in the Transactions of the Society of Automotive Engineers. Paper No.
741193. 1974

‘8 Andersson GB, Ortengren O, Nachemson A, Elfstrom G. Lumbar Disc
Pressure and Myoelectric Back Muscle Activity during Sitting: Studies on and
Experimental Chair. Scandinavian Journal of Rehabilitation Medicine. Vol. 6,
1974. pp. 104-114

207

 

49 Chaffin D, Andersson GB. Occupational Biomechanics: Second Edition. John
Wiley and Sons, New York. 1991

5° Anthropometric Specifications for Mid-Sized male Dummy. Side View with
Skeleton. Transportation Research Institute. University of Michigan. Nov. 30,
1983. Contract Number DTNH22-80-C-07502.

5‘ Renyolds HM, Snow CC, Young JW. Spatial Geometry of the Human Pelvis
(Memorandum Report AAC-119-81-5). Oklahoma City, OK: Federal Aviation
Administration, Civil Aeromedical Institute. 1981

52 Seidl G, Marchinda D. Hip Joint Center Location from Palpable Bony
Landmarks — a Cadaver Study. Journal of Biomechanics, V0. 28, No. 8, pp. 995-
998. 1995

aiks F, Reinecke S. Understanding the Kinematics of the Back and its
Relationship to Back Support. Pending Publication. 1999. pp. 1- 13

5‘ Sigma Stat for Windows. Version 2.03. Copyright 1992-1997, SPSS Inc.

55 Bush TB, Hubbard RP, and Reinecke S. An Evaluation of Postural Motions,
Chair Motions, and Contact in Four Office Seats. Human Factors and
Ergonomics Society. Environmental Design Session. October 1999

56 Hobson D. Comparative Effects of Posture on Pressure at the Body Seat
Interface. Journal of Rehabilitation Research and Development. Vol. 29, No. 4,
1992. pp. 21-31

57 Goonetilleke R. Designing to Minimize Discomfort. Ergonomics in Design.
July 1998. pp. 12-19

58 Andersson GB, Ortengren O, Nachemson A, Elfstrom G. Lumbar Disc
Pressure and Myoelectric Back Muscle Activity during Sitting: Studies on and
Experimental Chair. Scandinavian Journal of Rehabilitation Medicine. Vol. 6,
1974. pp. 104-114

59 Chaffin D., Andersson G. Occupational Biomechanics: Second Edition. John
Wiley and Sons, New York. 1991

208