-_

65:1..‘

'; ‘ *1 . ."' A ~ 1:31,”:2'A‘2'yv4
I r . _.‘k_ . ‘ ‘ ‘ . > . ‘
' 3"”. ""r-:. ' ‘ n ' .r , I:

v ‘ . .- . ‘

 

In

.1

v ‘3 onuwyfi-‘i

 

.. x2"
S." 5;“: 0"‘4 51: Am 23:}.

_ \x ‘ V31; afim 1%“? :8
.Jmé 6"? egg-23:,

‘13.. ”“61“?
’r Y

32‘...
H.

IV I ,
.1833“ .3'5' xv: x
J ' t. . ‘ia’fv" ‘ '
Hui}; 35:4,, .
. \ f
x'v. .055 gig” J
4“,”;- , . . _
31%.“: .\ fi‘wf"
a

;um I

.
{31}, "‘t

War'flf‘r- ,

38 {$336 .1.

4' ,6 git-10%;" 4‘

333“
nfsg'fi-fi'

D
k l

., ‘ ‘h

' 4 ,:|' I‘llmu“ 1" ..
' '19'5' t
. . A

';r.5;3 L;

3,. M57 * r
mun
1&1???

4.519;": . '1’"

j; ‘I #1
"if u 1 my,
-' S?

l
5“ 3C;
2: flat; 2:2»

7 .9
.. 127$ "7'-
35%;” $11!?!” ”:5 {*1 an
'33“; "7:5? ‘
raga
W:
“i. .7...
. . 333533? 1
_. n34; “It?“

.

7,

M
:6

$31- "$3.45” .2. 335’ $rfff~r7.,/'""'
mum-.2. '7"31,;"/‘;§,if',‘j‘£§’36
1L3- :.‘. v'
"I

"1L
w: -
" ‘ ' ’ ‘V ‘X‘:»:\..‘

73.5)."Mf-6- ‘

c.“
"$15?"
, .. (r 1,: _
... 3"
5,...
' , v» . ”Sr-.1. "a! .91»- $2};
4 " ' .6- WM .,..- .g' .»w—.w- .,,,,........—*
hr“ m _ . ‘
'22':5‘"':.§‘.'- - ,‘ C > .5 g» :1. . '6?” ,_
‘ II. a: . ,"“__, ~. .- _ .
~ A . .. w...” .7"; m
firm

A:

4
J5- «ct-"'1" paw-17"

"" «a
1.1. 'L‘3'"' ': Jot-”£39"

' < Irirw:-hfi‘::~l
5.6 :63"

r" 3.22.12!

 

*3. t as: |_ .9.

13‘1“} 01‘th

mtii'iti‘in M H mm in w 1 y

3 1293 0061

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

 

 

LIBRARY "
Michigan State
University

 

Geometric Model and Spinal Motions of the

Average Male in Seated Postures

presented by

William A. Haas

has been accepted towards fulfillment
of the requirements for

Master's degree in Mechanics

 

 

 

Major professor

Datwia‘ 9’1

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

 

 

PLACE N RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

iii-X LD 100’s
SEP 2 7 I99:

JUL 1 t 199:

 

 

 

 

 

 

 

 

15W
‘ Hair; 19$6'

-
v—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution

GEOMBTRIC MODEL AND SPINAL MOTIONS
OF THE AVERAGE MALE IN SEATED POSTURES
BY

William Adolf Haas

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

MASTER OF SCIENCE
Department of Metallurgy, Mechanics, and Material Science

1989

(3.
U3

LJh'r
N“)

-3

ABSTRACT

GEOMETRIC MODEL AND SPINAL MOTIONS
OF THE AVERAGE MALE IN SEATED POSTURES

BY

William Adolf Haas

The position of the pelvis relative to the thorax in
the seated posture and the distribution of mid-sagittal
bending in the lumbar spine are not well understood. In
this study, a computer graphics model and a mid-sagittal
bending motion program were created. Initial seated
postures were formed and, through the use of a motion
program which utilized different distributions of lumbar
motion, several published studies were simulated to evaluate
the relative positions of the pelvis and thorax. The
published studies were best simulated by equally
distributing the rotation in the mid-sagittal plane among
the LS/Sl interspace through the T12/L1 interspace and using

the initial postures.

ACKNOWLEDGEMENTS

There were many special people involved in the
production and writing of this master thesis. I would like
here to extend my deep gratitude to those special people,
they are:

Lori Yuhas, Sharon Booker, and Theresa Staton,
undergraduates who helped me with the burdensome work of
entering the computer data and who made the computer
laboratory sparkle with life.

Johnson Controls Inc. for their generous funding of
this project.

Drs. Robert W. Soutas-Little and Nicholas Altiero for
serving on my committee.

My family and friends for their support in my academic

pursuits.

Lastly, but most important; Dr. Robert P. Hubbard, for

tun: years of friendship and guidance and for creating a

uruique and amusing work environment.

ii

TABLE OF CONTENTS

List of Tables.
List of Figures
1. INTRODUCTION
2. METHODS.

2.1 Introduction to Methods. . .
2.2 Computer System and Software
2.3 Skull Modeling . . . .
2.3.1 3-D Skull
2. 3. 2 2-D Skull
2.4 Pelvis . .
2.4.1 3- D Pelvis.
2. 4. 2 2- -D Pelvis.
Thorax
Creation of The Nominal Posture. . .
2. 6.1 Placement of 2- -D Body Segments.
2.6.2 Spinal Linkage. . . . . .
2.7 Spinal Motion. .

NN
mm

2 7 1 Important Seating Variables
2.7.2 Hip Joint Center Location . .
2.7.3 Spinal Rotation Distribution.
2.7.4

Total Lumbar Curvature (TLC).
2.7.4.1 Rigid Thorax . .
2.7.4.2 Non-Rigid Thorax

2.7.5 Effect of Spinal Rotation
Distributions . . . . . . . .

2. 7. 6 Seatback Angle. . . . . .

2. 7. 7 Comparison of Model with Literature

2. 8 Published Data . .

2.8.1 Robbins and Reynolds[4]

2. 8.2 Nyquist and Patrick[5].

2. 8. 3 Andersson et. al. [3]. . . .

2. 9 Method for Comparison of JOHNl Model
with Published Data. . . .

2. 9.1 Step 1: Method for Selection of
Representative Rotation
Distribution(s) . . .

2. 9. 2 Step 2: Method for Modeling
Robbins and Reynolds Study[4]

2.9.3 Step 3: Method for Modeling

Nyquist and Patrick Study[S].

iii

Page

vii

49
50
50

2.9.4 Step 4:
Andersson et. al.

Method for Modeling
Study[3].

2.10 Method of Comparison of JOHNZ Model
with Published Data . . . . .
2.11 Review of Methods

3. RESULTS AND DISCUSSION

3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9

Evaluation of
for the JOHNl
Simulation of
for the JOHNl
Simulation of
for the JOHNl
Simulation of
for the JOHNl
Evaluation of
for the JOHNZ
Simulation of
for the JOHNZ
Simulation of
for the JOHNZ
Simulation of
for the JOHNZ
Choice of the

the Rotation Distributions
Model. . . .

Robbins and Reynolds[4].
Model. . . . .
Nyquist and Patrick[5]
Model. . . . .
Andersson et.

al. [3]
Model. . . . .
the Rotation Distributions
Model. . . . . . .
Robbins and Reynolds[4]
Model. . . . .
Nyquist and Patrick[5]
Model. . . . . .
Andersson et. al.[3]
Model. . . . . . .

JOHNl Model or

the JOHNZ Model. . . . . .
3.10 Movement into the Seatback.

4. SUMMARY AND CONCLUSIONS.

5. LIMITATIONS AND FUTURE WORK. . . . .

.APPENDICES. . . . .

LBIBLIOGRAPHY. . .

iv

51

51
52

54

54
76
77
79
93
112
113
114

125
126

128
132
134
158

Table

1.
2.
3.
4.
5.
6.

10.

11.

12.

13.

14.

15.

16.

Rotation Distributions.

LIST

OF TABLES

TLC for Robbins and Reynolds[4]

TLC for Nyquist and Patrick[5].

Multiplication Factors.

Data from Andersson et.

Distances (mm)
JOHNl Model 10

Distances (mm)
JOHNl Model 10

Distances (mm)
JOHNl Model 20

Distances (mm)
JOHNl Model 30

Results of the

Between
Degrees

Between
Degrees

Between
Degrees

Between
Degrees

al.[3]. . . . . . .

Interspaces for the
Flexion. . . . . .

Interspaces for the
Extension. . . . .

Interspaces for the
Extension. . . .

Interspaces for the
Extension.

Simulation of Robbins and
Reynolds[4] for the JOHNl Model. . . . .

Results of the Simulation of Nyquist and
Patrick[5] for the JOHNl Model. . . . . .

Andersson et.

for the JOHNl Model .

Distances (mm)
JOHNZ Model 10

Distances (mm)
JOHN2 Model 10

Distances (mm)
JOHNZ Model 20

Distances (mm)
JOHNZ Model 30

Between
Degrees

Between
Degrees

Between
Degrees

Between
Degrees

al.[3] TLC and Pelvic Rotations

Interspaces for the
FleXionO O O 0 O O O

Interspaces for the
Extension. . . . . .

Interspaces for the
Extension. . . . . .

Interspaces for the
Extension. . . . . . .

Page
34

41
44
47

48

57

6O

62

65

76

79

93

95

97

99

101

Table Page

17. Results of the Simulation of Robbins and
Reynolds[4] for the JOHN2 Model . . . . . . . . . 113

18. Results of the Simulation of Nyquist and
Patrick[5] for the JOHN2 Model. . . . . . . . . . 114

vi

LIST OF FIGURES

Figure Page
1. 2-Dimensional drawing template. . . . . . . . . . 2
2. 3-Dimensional H-Point machine . . . . . . . . . . 3
3. 3-Dimensional wireframe skull . . . . . . . . . . ll
4. 2-Dimensional skull . . . . . . . . . . . . . . . 13
5 3-Dimensional wireframe pelvis. . . . . . . . . . 15
6. 2-Dimensional pelvis. . . . . . . . . . . . . . . 17
7. 2-Dimensional thorax. . . . . . . . . . . . . . . 18
8. 3-Dimensional ribcage . . . . . . . . . . . . . . 19

9. 2—D thorax/3-D ribcage overlay. . . . . . . . . . 20
10. JOHNl model in nominal position . . . . . . . . . 25
11. JOHN2 model at initial position . . . . . . . . . 27
12. Robbins and Reynolds[4] seated posture. . . . . . 39

13. Figure of data from subject CJM from
Nyquist and Patrick[5]. . . . . . . . . . . . . . 42

14. Figure of data from subject LMP from
Nyquist and Patrick[5]. . . . . . . . . . . . . . . 42

15. Pelvic angle definitions from
Andersson et. al.[3]. . . . . . . . . . . . . . . . 45

16. Multi-screen model of distribution one for -
the JOHNl model . . . . . . . . . . . . . . . . . . 55

17. Calculation of distances between interspaces. . . . 56
18. 10 degrees flexion for the JOHNl model. . . . . . 58

19. 10 degrees extension for the JOHNl model. . . . . 61

vii

Figure Page
20. 20 degrees extension for the JOHNl model. . . . . 63
21. 30 degrees extension for the JOHNl model. . . . . 66

22. Comparison of distribution one and
distribution two at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 67

23. Comparison of distribution one and
distribution three at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 68

24. Comparison of distribution one and
distribution four at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 69

25. Comparison of distribution one and
distribution five at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 7O

26. Comparison of distribution one and
distribution six at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 71

27. Comparison of distribution one and
distribution seven at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 72

28. Comparison of distribution one and
distribution eight at 30 degrees extension
for the JOHNl model . . . . . . . . . . . . . . . . 73

29. Comparison of distribution one and

distribution nine at 30 degrees extension

for the JOHNl model . . . . . . . . . . . . . . . . 74
30. Comparison of distribution one and

distribution ten at 30 degrees extension

for the JOHNl model . . . . . . . . . . . . . . . . 75
31. Andersson -2 posture for the JOHNl model. . . . . . 82

32. Andersson +2 posture for the JOHNl model
with distribution one . . . . . . . . . . . . . . 85

33. Andersson +4 posture for the JOHNl model
with distribution one . . . . . . . . . . . . . . 86

34. Andersson +2 posture for the JOHNl model
with distribution seven . . . . . . . . . . . . . 87

viii

Figure

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Andersson +4 posture for the JOHNl model
with distribution seven . . . . . . . .

Andersson +2 posture for the JOHNl model
with distribution eight . . . . . . . .

Andersson +4 posture for the JOHNl model
with distribution eight . . . . . . . .

Composite figure of Andersson +2 posture
for the JOHNl model for distributions one,
seven, and eight. . . . . . . . . . .

Composite figure of Andersson +4 posture
for the JOHNl model for distributions one,
seven, and eight. . . . . . . . . . . .

10 degrees flexion for the JOHN2 model.

10 degrees extension for the JOHN2 model.
20 degrees extension for the JOHN2 model.
30 degrees extension for the JOHN2 model.

Comparison of distribution one and
distribution two at 30 degrees extension
for the JOHN2 mOdel O O O O O O O O O O 0

Comparison of distribution one and
distribution three at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Comparison of distribution one and
distribution four at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Comparison of distribution one and
distribution five at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Comparison of distribution one and
distribution six at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Comparison of distribution one and
distribution seven at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Comparison of distribution one and

distribution eight at 30 degrees extension
for the JOHN2 model . . . . . . . . . . . .

ix

Page

88

89

9O

91

92

96

98

100

102

103

104

105

106

107

108

109

Figure

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

Comparison of distribution one and

distribution nine at 30 degrees extension

for the JOHN2 model . . . . . . . . .

Comparison of distribution one and
distribution ten at 30 degrees extension
for the JOHN2 model . . . . . . . . . .

Andersson +2 posture for the JOHN2 model
with distribution one . . . . . . . . .

Andersson +4 posture for the JOHN2 model
with distribution one . . . . . . .

Andersson +2 posture for the JOHN2 model
with distribution seven . . . . . . . .

Andersson +4 posture for the JOHN2 model
with distribution seven . . . . . . . .

Andersson +2 posture for the JOHN2 model
with distribution eight . . . . . . . .

Andersson +4 posture for the JOHN2 model
with distribution eight . . . . . . . .

Andersson +2 posture for the JOHN2 model
with distribution ten . . . . . . . . .

Andersson +4 posture for the JOHN2 model
with distribution ten . . . . . . . . .

Composite figure of Andersson +2 posture for

the JOHN2 model for distributions one,
seven, eight, and ten . . . . . . . . .

Composite figure of Andersson +4 posture for

the JOHN2 model for distributions one,
seven, eight, and ten . . . . . . . . .

Page

110

111

115

116

117

118

119

120

121

122

123

124

1. INTRODUCTION

The need for a representation of human seated posture
is apparent in the current practice of automobile seat
design. The current practice in the automotive seating
industry is to layout the interior of the automobile and the
seats using the SAE 2-D drawing templates, Figure 1, and
other SAE accommodation tools (SAE is an abbreviation of
Society of Automotive Engineers). This seating layout is
followed by the development of a prototype automobile seat
which is tested using the three-dimensional (3-D) H-point
machine, Figure 2. The 2-D drawing template and the 3-D H-
point machine provide a location in the automobile seat of
the H-point and the seatback angle. The H-point is a point
defined by the SAE which represents the center of rotation
of the hip joint relative to the pelvis. The seatback angle
is defined as the angle from vertical to a line on the torso
of the 2-D drawing template and the 3-D H-point machine.
Both the position of the H-point and the seatback angle are
often specified prior to the construction of the automobile
seat. The contours of the 2-D template and the 3-D H-point
machine are flat in the lumbar region and curve forward at
the shoulders. This is the posture which has been typical

in American automobile seats for many years.

Figure 1 - 2-Dimensional drawing template

'.P- s-tn1-utfindw a;
‘5' n." L-""‘:

\'
. . ”'1 .31... 5,, U.“ Ivy-Hugh
I
5

    
    

n
‘ 5n fimdmanwevt
'..’ «I . '
‘ ' *tnr.
. ‘.,!u
a .". .' nw"...:‘
‘_ ' waif lsv_. -"!n-u+o~
1. ' ‘7.._H'5"1T .‘IkL-J"

 

nigh": ~ x:"‘_‘.:.. ,-. -...

Figure 2 - 3-Dimensional H-Point machine

4

There has been a problem since the introduction of
highly contoured seats and lumbar support in the automobile
seats. The contours and the lumbar support cause the 2-D
template and the 3—D H-point machine to rotate forward in
the seat. This results in the automobile seat failing to
meet the required specified H-point location and seatback
angle.

Another problem with the 3-D H-point machine is that it
can only be used on a previously constructed automobile
seat. If an automobile seat fails to meet specifications,
then it must be redesigned and a new prototype constructed
and tested. This often results in an expensive cycle of
design and redesign.

Because of the inability to meet the specified H-point
location and seatback angle with the 2-D drawing template
and 3-D H-point machine, a new seating model is needed. A
computerized seating model could be developed with a
computer aided design (CAD) package. To be useful, the
computerized seating model must realistically simulate the
motion of a human. When considering an automobile seat with
lumbar support using the computerized seating model, the
torso of the model must not rotate forward in the seat.
Instead the seating model must represent the lordosis of the
spine as it would occur naturally.

Limited information is available which quantitatively
describes the motion of the lower spine. Most of the

literature does not describe the motion at the individual

5

joint centers. Some literature is available which does
quantitatively describe the motion of the lower spine and
the position of the pelvis.

The most recent and extensive study of human posture in
automobile seats comes from the University of Michigan
Transportation Institute (UMTRI)[1]. Included among the
data from UMTRI[1] is a set of full scale drawings which
show the skeletal position of the seated occupant along with
a table of coordinate locations of important landmarks. The
positions which UMTRI[1] described are representative of the
5th percentile female and the 50th and 95th percentile male
in the then current automobile seats. The data were
gathered through seating occupants in a hard seat which was
representative of a typical automobile seat. After placing
markers on important external landmarks, photographs were
taken and measured providing data on the position of
external landmarks in a seated posture. Generally, the
skeletal placement was based on estimates from external
landmarks based on the Link System of the Human Torso by
Chaffin et. al.(Link Study)[2]. An apparent problem with
the skeletal drawing and the skeletal data is the position
of the pelvis in the body. There is 46mm of tissue between
the bottom of the pelvis and the surface of the buttocks.
This appears to be too much tissue for a person in the
seated posture.

In another important publication, Andersson et. al.[3]

measured radiographs of seated occupants to describe the

6

positions of their pelves relative to their lumbar spines in
both the standing posture and in a variety of seated
postures.

A report by Reynolds and Robbins, "Position and
Mobility of Skeletal Landmarks of the 50th Percentile Male
in an Automotive Seating Posture"[4], is also significant.
This publication provides the positions of several vertebral
interspaces and the position of the pelvis for a single
seated posture. Fortunately, the posture studied by
Reynolds and Robbins was different from the posture used by
UMTRI[1]. This allowed for a comparison between the two
postures.

A study by Nyquist and Patrick, "Lumbar and Pelvic
Orientations of the Vehicle Seated Volunteer"[5], contained
data concerning the positions of the pelvis and lumbar
region for two seated male volunteers obtained through the
use of radiographs.

Data for the distribution of spinal motion among the
vertebral interspaces in the spine are from two sources.

The first source, "Kinematics of Human Spine" by Panjabi and
White[6], quantitatively describes the proportion of
movement for each spinal interspace. The second source for
lumbar motion distribution is by Allbrook, "Movements of the
Lumbar Spinal Column"[7].

Data for the representation of the skeletal pelvis were
obtained from a study by Reynolds, et. al., "Spatial

Geometry of the Human Pelvis"[8]. They measured many

7
skeletal pelves utilizing a digitizing technique then

divided the pelves into three size categories, including the

50th percentile male.

Data for the creation of the skull for a computer model
were furnished by Hubbard[9]. In this publication, Hubbard
extensively described the dimensions and the creation of a
50th percentile male skull model.

The present research to model and study the spinal
motions of an average adult male consists of several parts:
1. Construction of two-dimensional representations of the

thorax, skull, and pelvis.

2. Selection of a nominal position of the skeletal model.

3. Development of a motion algorithm and computer model for
the lumbar region of the spine.

4. Testing of various motion models of the lumbar spine to
determine their effect on the relative positions of the
thorax and the pelvis.

5. Verification of the model created in step 2 by comparison
to available data.

6. Recommendations for the use of the model in the design of

automobile seating.

2 . METHODS

WM

To represent the 50th percentile male pelvis, skull,
and ribcage, geometric models were formed based on available
literature. The pelvis and skull were first modeled as
three-dimensional objects. Then, two-dimensional
representations of the skull and the pelvis were formed
based upon the three-dimensional models. For this study,
three-dimensional representations were not necessary. Only
2-D mid-sagittal motion is studied so two-dimensional
representations were used. Also, at the beginning of this
study, a three-dimensional representation of the thorax was
not available.

Positions of the head relative to the thorax have also
been predicted by the model. The user can place the head of
the model on any angle from the horizontal. The model will
then automatically distribute evenly the motion needed to
obtain this angle throughout the cervical region. No
further research has been done for the cervical region since
the lumbar region, not the cervical region, is the focus of

this study.

'd

S.
U.

Ix)

‘3.

F:
51

2 2 m r m nd f w re

The computer system used to develop and study the
skeletal motion model was the SUN SYSTEM 3/60 utilizing the

SDRC IDEAS 3.8 GEOMOD[10] Solid Modeling package.

WM

2.3.1 3-D Skull

The 3-D skull was created using data supplied by
Hubbard[9]. This publication contained data on the three-
dimensional locations of various skeletal landmarks
representing the skull of the 50th percentile male. This
skull geometry described by Hubbard[9] is the basis for the
head of the Hybrid III crash test dummy. Portions of the
lower jaw and eye socket region could not be adequately
constructed from the data supplied by Hubbard[9]. Figure 3
is a 3-D wireframe representation of the skull.

The origin of the coordinate system for the skull was
at the Nasion which is at the bridge of the nose. The X-Y
plane was defined as the Frankfort plane, which was the
plane formed by the porions and the orbitales skull
landmarks. The positive X axis was in the forward
direction, and the positive Y axis was to the left of the
body. The Z plane was perpendicular to the Frankfort plane

with the positive Z in the superior direction.

10

The cranial portion of the skull was created through
the merging of two spheres as used in modeling the skull by
Khalil and Hubbard[11]. The first and larger sphere which
described the posterior section of the cranium measured 71mm
in radius. The second and smaller sphere which described
the frontal section of the cranium measured 60mm in radius.
The centers of the spheres were positioned 46mm apart as
specified by Khalil and Hubbard[ll]. A conical segment of
length 43.2mm was placed between the two spheres. Once the
Spheres and the cone were positioned, they were joined
together forming the major segment of the superior portion
of the skull.

Next, using landmark locations specified by Hubbard[9],
the remainder of the skull was formed. Those areas of the
skull, mainly the lower part of the jaw, which needed a more
detailed description than that supplied by Hubbard[9], were
described through measurements of a typical skull.

The sockets for the eyes were created as follows.
First, two spheres of 18mm radius and a third sphere of 17mm
radius were formed. The two initial spheres were cut with
the third sphere resulting in was hollow spheres. Next, the
spheres were cut in half and the centers of the cut surfaces
of the spheres were placed at the skull coordinates
-7.6mm,¢31.0mm,-12.7mm. The final step in the creation of
the skull was the placement of two small spheres in the eye
sockets to represent pupils using data obtained through the

measurement of an average skull.

11

 

 

 

 

 

/

Figure 3 - 3-Dimensional wireframe skull

12
2.3.2. 2-D Skull

The two-dimensional skull was created by cutting a 1mm
wide portion of the 3-D solid model of the skull about the

mid-sagittal plane. This can be seen in Figure 4.

2 I E ]v'

2.4.1 3-D Pelvis

The pelvis for the model was created using the data
contained in a report by Reynolds, et al.[8] The
publication contained data on inside and outside skeletal
landmarks of the pelvis. Three types of pelves were
described, the 50th and 95th percentile male and the 5th
percentile female. These pelvic data were gathered for use
in future crash test dummies.

The data used to create a model of the 50th percentile
male pelvis is given in Appendix 1 and a wireframe picture
of the pelvis can be seen in Figure 5. First the
coordinates of the points to form the pelvis were entered
into the computer utilizing the sequence of commands. As
each point was entered using pelvic coordinates, the
computer program assigned to each set of coordinates a label
which became its new name. The points and their

corresponding labels are listed in Appendix 1. After each

 

 

 

Figure 4 - 2-Dimensional skull

14
point was entered and assigned a label, the points had to be

assembled into facets. The computer program asked for the
points, using the labels, needed to assemble facet 1. After
entering the point labels for facet 1, each subsequent facet
was formed until the left half of the pelvis was completed.
Then, the mirror image of the left half of the pelvis was
created and the two halves were simultaneously fused

together.

2.4.2 2-D Pelvis

The two-dimensional pelvis was simply a facet formed
using all the points which described the outer surface of
the pelvis projected onto the mid-sagittal plane as seen
from the left lateral view. Appendix 2 contains these data.
The two-dimensional pelvis was then stored as a separate

object, named 2DPELVIS, Figure 6.

W

The first of two thorax representations was developed
from measurements of diagrams supplied by the University of
Michigan Transportation Research Institute (UMTRI)[1]. This
initial thorax was two-dimensional and was used in the
process to create and debug the spinal motion computer
program. Appendix 3 contains the data for this thorax also

shown in Figure 7. These data include the skeletal

15

 

 

 

 

 

 

Figure 5 — 3-Dimensional wireframe pelvis

16
positions of the twelve thoracic spinous processes, three

points on the sternum and one point to represent the lowest
margin on the 10th rib.

A three-dimensional thorax was created using
measurements of a typical thorax. The coordinates of the
points measured on the thorax were entered into the computer
to create facets in the same manner as for the 3-D pelvis.

Even though a three-dimensional thorax, Figure 8, was
created, it was not used in this motion study. The two-
dimensional model based upon the UMTRI[1] drawing was used
because the 3-D thorax did not have any spinous processes.
It was thought that the lack of spinous processes would
cause problems when trying to fit the model into a seatback.
Figure 9 shows the differences between the two thorax
models. Furthermore, the three-dimensional thorax was not

needed because this is a planar motion study.

2 E 3 I' E I] H . J E !

2.6.1 Placement of 2-D Body Segments

The placement of the pelvis, rib cage, and skull to
represent a 50th percentile male was derived from the
diagrams and data supplied by UMTRI[1], Appendix 4. This
initial posture was the basis for the JOHNl model. The
model was assembled using the SYSTEM ASSEMBLER option

offered by the IDEAS GEOMOD[10] package. The SYSTEM

l7

 

Figure 6 - 2-Dimensional pelvis

 

18

 

Figure 7 - 2-Dimensional thorax

 

19

 

Figure 8 - 3-Dimensional ribcage

RED;

 

Figure 9 - 2-D thorax/3-D ribcage overlay

21
ASSEMBLER package is arranged so that each component to be

assembled was called onto the computer screen individually
in the position designated by the components coordinate
system. As an example, when the two-dimensional skull was
called onto the screen, the nasion was at the origin of the
SYSTEM ASSEMBLER coordinate system.

To complete the body, each component was called onto
the screen then placed in reference to the SYSTEM ASSEMBLER
coordinate system. The SYSTEM ASSEMBLER coordinate system
and subsequently the whole body coordinate system was
designated so that the X axis was parallel to the floor and
positive forward (toward the left of the viewer in all
diagrams). The positive Y axis was lateral to the body in
the left hand direction of the model (pointing out of the
page for all diagrams) and also parallel to the floor. The
Z axis was perpendicular to the floor, with the upward
.direction to be positive.

First, the 2-D pelvis was called onto the screen. When
it appeared on the screen, the midpoint of the anterior
superior iliac spines (ASIS) was at the origin. To place
the pelvis in the orientation specified by UMTRI[1], the
pelvis was first translated to place the midpoint of the hip
joint centers of the pelvis at the origin of the whole body
coordinate system (Omm,0mm,0mm). The pelvis was then
rotated negative -54.07 degrees about the positive Y axis as

specified by UMTRI[1].

22

Second, the 2-D skull was called onto the screen. As
stated earlier the Nasion was at the origin, with the
Frankfort plane aligned with the SYSTEM ASSEMBLER coordinate
system. To move the skull into the position of whole body
coordinates the mid-point of the porions originally at
-86.36mm, Omm,-33.02mm was translated to -185mm,0mm,614mm.
This corresponds with the position as designated by
UMTRI[1]. The skull in the UMTRI[1] drawing and their
literature does not have a level Frankfort plane. Instead,
the skull in the UMTRI[1] literature, and hence the 2-D
skull model was rotated -3.9 degrees about the positive Y
axis with the mid-point of the porions as the center of
rotation.

The third component called onto the computer screen was
the thorax. Since the whole body coordinates as assigned by
UMTRI[1] were used to create the thorax, the thorax was in
the proper position when it was called onto the computer

screen.

2.6.2 Spinal Linkage

The first step in the process of creating the spinal
linkage was to decide how the spine would be modeled. For
the sake of simplicity, the spine was modeled as a chain of
links which would have fixed length. This meant that the

joint centers of rotation were fixed with respect to the

23
vertebra. White and Panjabi[6] described the regions of the

joint centers for flexion and extension. Their conclusion
was that the joint centers moved rearward, especially at the
end of the facet joint range of motion. Although not
documented in the traditional academic manner, Jones[12]
described lumbar motion as a chain of links with joint
centers at the disc centers for the first 36 degrees of
extension from a straight lumbar spine. After this, the
facet spinous process impacted and the joint centers moved
rearward. In future studies, the effects of motions of the
spinal joint centers should be investigated.

The second step in creating the spinal linkage was the
designation of centers of rotation for the motion model.
Initial data were needed for the positions of the centers of
rotation for the spinal discs. The UMTRI[1] study, which is
based on the study by Chaffin et. al.[2], was used to supply
the initial joint centers of motion. The centers of
rotation for the lumbar spine which UMTRI[1] supplied were
the joint center between the twelfth thoracic vertebra and
the first lumbar vertebra (T12/L1), the joint center between
the second lumbar vertebra and the third lumbar vertebra
(L2/L3), and the joint center between the fifth lumbar
vertebra and the first sacral vertebra (LS/81). To get the
other centers of rotation, a line was drawn along the three
known centers of rotation on the drawing, and the remainder
of the centers of rotation were chosen to be at the midpoint

of the interspaces along this line. For the thoracic centers

24
of rotation, UMTRI[1] supplied the center of rotation

between the seventh cervical vertebra and the first thoracic
vertebra (C7/T1), the T4/T5, and the T8/T9 joint centers.
The remainder of the centers of rotation were estimated to
create a smooth curve. A similar method was used for the
cervical centers of rotation. For the cervical region,
UMTRI[1] supplied the C7/T1 and the Cl/skull centers of
rotation. The remainder of the centers of rotation were
estimated through the use of a smooth curve. The
coordinates for the joint centers are contained in
Appendices 5-7.

At this point an initial posture for the model had been
created, the JOHNl model, Figure 10. Since the choices for
the centers of rotation were not necessarily correct a
method was devised by which the centers of rotation could
easily be changed, if a superior set of data would become
available. The computer motion program accessed separate
external data sets which contained the coordinates of the
centers of rotation. These data sets could then easily be
changed without changing the motion program.

The bottom of the pelvis on the UMTRI[1] drawing was
46mm from the surface of the buttocks. This seemed to be
too large a distance. The external surface of the body was
measured in the UMTRI[1] study and the position of the
skeleton was estimated from the location of external body

landmarks. Furthermore, the lengths of the lumbar vertebrae

25

 

 

Figure 10 - JOHNl model in nominal position

26
in the UMTRI[1] study were shorter than the lengths of the

lumbar vertebrae in the Link Study[2]. Because the Link
Study[2] provided a lumbar spine which was 29.6mm longer,
the data from it might provide a superior position of the
pelvis.

To create the new lumbar region of the spine with the
Link Study[2] data the same angles for the spine and the
body as used by UMTRI[1] were used again. The only aspect
which changed was the length of the lumbar vertebrae. To
position the pelvis with the longer lumbar spine the top of
the head was kept in the UMTRI[1] position. Since the top
of the head was an external landmark this position was known
by UMTRI[1] with relatively good accuracy. Because the
UMTRI[1] data for everything except for the length of the
lumbar spine and the position of the pelvis were maintained,
the position of the T12/L1 interspace was the same for both
models. This model is named the JOHN2 model, Figure 11.
Appendices 8-10 contain the coordinates of the joint centers

used for the JOHN2 model.

WM

2.7.1 Important Seating Variables

Once an initial posture had been created, the motion

program for the model was written. Since the motion program

27

 

 

Figure 11 - JOHN2 model at initial position

28
will eventually be used by people with little knowledge of

biomechanics the program must be user friendly. The input
which is required to operate the program must be kept as
simple as possible.

Several meetings with seating engineers of Johnson
Controls Inc., The Society of Automotive Engineers, and
General Motors Chevrolet-Pontiac Canada were arranged to
decide the important variables for the spinal motion
program. These meetings supported the decision that the
three most important variables for the computer program
would be:

1. Hip joint center location.

2. The curvature of the spine.

3. The seatback angle.

2.7.2 Hip Joint Center Location

The common practice of placing the hip joint center at
the origin of the coordinate system was used in the computer
program. An additional capability of the computer program
is that the operator can move the hip joint center of the

model into any desired position.

2.7.3 Spinal Rotation Distribution

The spinal rotation distribution portion of the program

is the portion where the user decides what proportion of

29
total rotation will occur at each of the vertebral

interspace joints. For example if an even distribution of
rotation is chosen, each joint center would rotate through
the same range of motion. An important aspect of studying
the curvature of the spine is to find if there are
differences due to various spinal rotation distributions.
This will be discussed further in Section 2.7.5.

A decision was made that there would be no sacral
motion. By altering the data sets for the centers of
rotation and the program slightly, sacral motion could be
added.

The relative proportion of L5/Sl vertebral rotation was
defined as 1.0 All the input for the spinal motion
distribution was relative to L5/Sl rotation. If, for
example, the L4/L5 joint center rotated through twice the
range of motion the LS/Sl joint center did, then the

proportion of rotation for L4/L5 was 2.0.

2.7.4 Total Lumbar Curvature (TLC)

After a lengthy literature search, no standard
definition for spinal curvature was found. Therefore, a
definition for spinal curvature was created, Total Lumbar
Curvature (TLC). TLC is defined as:

TLC = (pelvic rotation) - (thoracic rotation)
Pelvic and thoracic rotations are about an axis parallel to

the Y-axis and are positive according to the right hand

30

rule, i.e. counterclockwise is positive when viewed from the

left hand side of the model.

2.7.4.1 Rigid Thorax

If the thorax is assumed to be rigid, the resulting
model has 6 mobile lumbar joint centers. The joint centers
are at the interspaces of L5/Sl, L4/L5, L3/L4, L2/L3, L1/L2,
and T12/L1. For a TLC of 30 degrees, the T12 vertebra and
the entire thorax would be rotated through an angle of 30
degrees relative to the pelvis. For an even distribution of
rotation relative to the L5/Sl interspace, each lumbar joint
center would have the same proportion of total rotation.

The L4/L5 joint center has the same rotation as the L5/Sl
joint center. The L3/L4 joint center has the same rotation
as the L4/L5 joint center and the L5/Sl joint center, etc...
Since the proportion of rotation for L5/Sl is defined as 1
the total of the proportions is 6, 1 for each of the mobile
joint centers. The rotation at each joint center is
calculated by dividing the TLC by the total of the
proportions and multiplying this number by the proportion at
each joint center. In equation form:

Rotation at a joint center= W1
Total of all proportions

As an example in the case of 30 degrees TLC and even
distribution, each joint center would have a rotation of

five degrees. The five degrees rotation was calculated by

31
dividing the TLC (30 degrees) by the total of the

proportions (6). Then, this number (30 degrees/6) is
multiplied by the proportion at each joint center. Since
the distribution is even, each proportion is 1. Therefore,
the L5/Sl joint center would have a rotation of (30
degrees/6)*1=5 degrees. The T12/Ll joint center would also
rotate through 5 degrees. Therefore the total rotation of
the L5 vertebra would be 5 degrees. The L4 vertebra would
have rotated 5 degrees relative to the L5 vertebra (5
degrees rotation at the L4/L5 joint center). Therefore, the
total rotation for the L4 vertebra would be 10 degrees. At
the T12 vertebra, the total rotation would be 30 degrees.
For an uneven distribution of rotation, each proportion
will not be equal to 1. If the L4/L5 joint center has half
the mobility of the L5/Sl joint center, the proportion for
the L4/L5 joint center would be 0.5. An example is a model
for which all joint centers other than L5/Sl have half the
mobility of the L5/Sl joint center. This would result in
one joint center with a proportion of 1 and five joint
centers with proportions of 0.5. The total of the
proportions would be 1+(5x0.5)= 3.5. For a TLC of 35
degrees, the L5/Sl joint center would have a rotation of (35
degrees/3.5)x1= 10 degrees. The L4/L5 joint center would
have a rotation of (35 degrees/3.5)x0.5= 5 degrees rotation.
The remaining mobile joint centers would have the same
rotation as the L4/L5 joint center. The total rotation of

the L5 vertebra would be 10 degrees. The rotation of the L4

32
vertebra would be 5 degrees relative to the L5 vertebra (5

degrees rotation at L4/L5). Therefore, the total rotation
of the L4 vertebra would be 15 degrees. The total rotation

of the T12 vertebra would be 35 degrees.

2.7.4.2 Non-Rigid Thorax

It was decided that thoracic bending in the mid-
sagittal plane might be significant enough to warrant the
inclusion of thoracic motion. For a model for which the
motion attributed to lower back motion stops at the T9
vertebra, the first immobile vertebra is the T9 vertebra.
This results in 6 mobile lumbar joint centers, as is the
case for a model with a rigid thorax, and 3 additional
mobile thoracic joint centers with the remainder of the
thorax modeled as rigid. For an even distribution of
rotation there would be 9 proportions. For a TLC of 30
degrees and even distribution of rotation, each joint center
would rotate through 3.33 degrees (30 degrees/9)xl.
Therefore, the L5 vertebra would rotate through 3.33 degree
and the L4 vertebra would rotate through 6.66 degree. The
T10 vertebra would rotate through 26.77 degrees and the
portion of the thorax which is modeled as rigid, the T9
vertebra and above, would rotate through 30 degrees.

For a model with an uneven distribution of rotation the

same procedure would be used as for the case of the rigid

33
thorax with uneven distribution. The only difference would

be the greater number of joint centers.

2.7.5 Effect of Spinal Rotation Distributions

To investigate the various distributions of motion, a
multiple image screen containing 5 images was used. The
nominal position was with a TLC of zero degrees and,
therefore, a straight lumbar spine. The other positions
were for different values of TLC and the images of these
positions were superimposed by the nominal position. The
motion algorithm moved the JOHNl model to different images
with a TLC of 10 degrees flexion, 10, 20, and 30 degrees
extension. This range of motion was chosen as
representative of possible ranges of motion for a person
seated in an automobile seat. Robbins and Reynolds[4]
studied a TLC of 9 degrees extension and Nyquist and
Patrick[5] studied a TLC of 17 degrees extension. A TLC of
10 degrees flexion and up to 30 degrees extension was used
to ensure that a sufficiently large range of motion was
encompassed.

While studying the rotation distributions each pelvis
remained locked in place and superimposed. By doing this
only the movement of the thorax and the spine due to the TLC
was studied. Then by entering various distributions of
motion, the resultant positions of the lumbar interspaces

and the thorax were compared for each distribution of

34
motion. The distributions which were investigated are

listed in Table 1.

Distribution One is the case with even distribution of
rotation, relative to the L5/Sl interspace, and a rigid
thorax. This case will be the distribution which all the
other distributions will be compared with.

Distribution Two is a slightly modified version of
Distribution One. The only change for this case is the
presence of greater rotation at the top of the lumbar spine.

TABLE 1
Rotation Distributions

L2. L1 mmmnmgm

E
f3
If»

1 1 1 1 1 1 1 ------ RIGID THORAX -----
2 1 1 1 2 2 1 ------ RIGID THORAX -----
3 1 1 1 .5 .5 1 ------ RIGID THORAX -----
4 1 l l .5 .5 .5 ------ RIGID THORAX -----
5 l .85 .75 .7 .6 1 ------ RIGID THORAX -----
6 1 1.12 .77 .47 .35 1 ------ RIGID THORAX -----
7 1 .85 .75 .7 .6 .6 .6 .45 .3 .3 ------
8 1 .85 .75 .7 .6 .6 .6 .45 .3 .3 .3 .25
9 l .85 .75 .7 .6 .6 ------ RIGID THORAX -----
10 1 1.12 .77 .47 .35 .35 ------ RIGID THORAX -----

Distribution Three is the opposite of Distribution Two.
This case is where there is less rotation at the top of the
lumbar spine than at the bottom.

Distribution Four is the same as Distribution Three
except for the modification that not only is there less
rotation present at the top of the lumbar spine but there is
also less rotation for the rigid thorax.

Distribution Five is the distribution of rotation

described by White and Panjabi[6] except for the rotation of

35
the rigid thorax. The rotation at T12/Ll for the rigid

thorax rotation was modeled as equivalent to that of L5/Sl
because no data were available from White and Panjabi[6] for
the modeling of a rigid thorax.

Distribution Six is the distribution described by
Allbrook[7]. Once again, a rigid thorax was modeled and
assigned a rotation proportion value of one at T12/L1, since
Allbrook did not research thoracic rotations.

Distribution Seven is the distribution listed by White
and Panjabi[6]. For this case, a non-rigid thorax was
modeled. This distribution has motion up to the T8/T9
interspace.

Distribution Eight is also data from White and
Panjabi[6]. For this case the thorax was modeled as non-
rigid with motion up to the T6/T7 interspace.

Distribution Nine is the same as Distribution Five
_except for this case the rigid thorax was assigned the
relative rotation of the T12/L1 interspace as described by
White and Panjabi[6].

Distribution Ten was the same as Distribution Six
except that the rigid thorax rotation was set equal to the
relative rotation of the Ll/L2 interspace.

The ten distributions were compared with each other to
investigate how the various distributions effected the

positions of the spinal interspaces, and the thorax.

36
2.7.6 Seatback Angle

In the automobile seating industry, the seatback angle
is defined as the angle from the vertical of a line on the
torso of the 2-D drawing template and the 3-D H-point
machine. The seatback angle is measured with the 3-D H-
point machine after it has settled into an automobile seat.
Since the new seating model is a computer model and cannot
settle into an automobile seat, a problem for this project
is how to define the seatback angle without knowledge of the
force-deflection characteristics of the human body in
automobile seats. To use the computer seating model, a
method must be developed by which the torso of the model can
be rotated into a seatback.

The first method used to rotate the torso into a
seatback was to define a tangent line to the body. The line
was defined as the tangent from the most rearward point of
the pelvis to the most rearward point of the torso. The
seatback angle was defined as the angle from the vertical of
this tangent line. The user would input a value for this
seatback angle and the computer would then rotate the model
about the H-point into this position. This attempt to
define the seatback angle did not work for any seatback
except for a straight hard seatback. Since the model was
eventually to be used to develop automobile seats, this

method to define the seatback had to be discarded.

37
The next method to rotate the torso of the model into a

seatback was to utilize a torso angle. The torso angle is
the angle from vertical between two points defined in the
published literature described below. By modifying the
motion program any angle on the torso could be used by the
computer operator. The user could define the angle from the
vertical of any line on the torso. The computer would then
rotate the model about the H-point so that the defined line
on the torso would be on the defined angle from the

vertical.

2.7.7 Comparison of Model with Literature

For comparison of postures between the model and
published literature, the published results were simulated.
First, the TLC used in the published study was entered into
the computer motion program. If the study did not supply
the TLC, then it was estimated from the published diagrams
or tables. Second, the model was rotated into the torso
angle specified in the literature. Third, the resultant
pelvic angles of the model were compared with the pelvic

angles found in the literature.

38
2 P li h Da a

2.8.1 Robbins and Reynolds[4]

Robbins and Reynolds[4] supplied the positions of
several vertebral interspaces. The angle from the vertical
of a line between the T4/T5 interspace and the T8/T9
interspace was 0 degrees. This angle was estimated from a
diagram supplied by Robbins and Reynolds[4] and was the
torso angle used to rotate the torso of the computer model.
Figure 12 is the diagram from the Robbins and Reynolds[4].

The TLC for the posture studied by Robbins and
Reynolds[4] was not given in their study. To calculate the
TLC for this posture, their diagram was used. Lines were
drawn between the given lumbar interspaces. The angle from
the vertical of the line between the L5/Sl interspace and
the L4/L5 interspace was 23 degrees. The angle from the
vertical of the line between the L4/L5 interspace and the
L3/L4 interspace was 28 degrees, and the angle from the
vertical of the line between the L3/L4 interspace and the
L2/L3 interspace was 26 degrees. The next given interspace
was T12/L1. Therefore, the TLC of the Robbins and
Reynolds[4] posture had to be estimated using the three
measured angles since the position of the L1/L2 interspace
was not known. Even though the angles did not progress in a
smooth manner, the TLC was estimated using the overall

change of 3 degrees (23-26). The three degree change

“~04-

 

! lunlgu («In III' I." in“. .

 

39

mm» un- m- an Id!

  
 

“I"

NI“ :9! III-clot I“.

\ Im- in.
“III

C
II. ”a! ”It. liq-ml

 

i i 5 fi h

I
I'll-u I. "III . “.9

Figure 12 - Robbins and Reynolds[4] seated posture

40
occurred at the L4/L5 and the L3/L4 interspaces. TLC is the

change at all mobile interspaces, not just the L4/L5 and the
L3/L4 interspaces, and the 3 degrees change was only a
portion of the entire TLC. The rotations at the remainder
of the mobile interspaces were still not known. If the
model has a rigid thorax and all the interspaces have a
rotation proportion of 1, then the rotation proportions for
the L4/L5 and L3/L4 interspaces are both 1. This results in
6 mobile joint centers each with a rotation proportion of 1
and a total of the rotation proportions equal to 6. The
rotation at the L4/L5 and the L3/L4 interspaces is only 2/6
of the TLC. The total change of 3 degrees is 2/6 the TLC,
therefore the TLC is (6/2)x(3 degrees)=9 degrees. To
calculate the TLC for the Robbins and Reynolds[4] study,
divide the total of all the rotation proportions by the
total of the rotation proportions for the L4/L5 and the
L3/L4 interspaces and multiply by 3 degrees. In equation
form this is:

TLC= ' r

(L4/L5 proportion+L3/L4 proportion)

Table 2 contains the calculated TLC of the Robbins and
Reynolds[4] study for each of the 10 different rotation
distributions.

The last piece of data given by Robbins and Reynolds[4]
was the pelvic angle. They listed the rotation of the
pelvis as 42 degrees from the standing position. Since the

JOHNl posture has a pelvic rotation of 54 degrees from the

41
standing position, the difference is 12 degrees. Therefore,

the pelvic rotation from the nominal JOHNl computer model

position is 12 degrees.

TABLE 2
TLC For Robbins and Reynolds[4]
DISTRIBUTION TLC
NUMBER (DEGREES)
1 9.0
2 12.0
3 7.5
4 6.8
5 9.2
6 7.5
7 11.5
8 12.6
9 8.4
10 6.5

2.8.2 Nyquist and Patrick[5]

Nyquist and Patrick[5] defined and measured the spine
line angle as the angle from the vertical of the line
between the T12/L1 and the L5/Sl vertebral interspace
centers. To match the position of the computer model with
the position of the Nyquist and Patrick[5] data, the spine
line of the computer model was matched to the spine line
found in their data.

The TLC of the Nyquist and Patrick[5] postures was
estimated by measuring the diagrams provided by Nyquist and
Patrick, Figures 13 and 14, in their publication. Lines
were drawn between the lumbar joint centers depicted on the
diagram. The angular change of the L1 vertebra relative to

the L5 vertebra was 7 degrees for the first subject, CJM.

42

   
   
  
   
 

' In no cm In) I
a 0n Volume"

. Vcdebm Inlonuu Cnnm

  

I ":19an an, Conm

0 Number: In “crayon: Item to
Item mm In mu IV

V000!

Horizon“

" j- . Pflflt
.. \ . Rmnncn

X?) a ‘ Panlm Io

"mom Ans

    

   

 

 

Figure 13 - Figure of data from subject CJM from
Nyquist and Patrick[5]

| lo M II.“ M
m... . On Volunmr ' |
ovmmu Inlnmu Ccnm

I Vodcbul Ody Center

0 Ilunbm In "cums MI" '0
"on: mm In mu IV

   
  

lUfl'bI' SDI". «u

[Inc \‘\~°\ 1’}!

    

, ’ mum Io
/' H “mm Am

 

Figure 14 - Figure of data from subject LMP from
Nyquist and Patrick[5]

43
The angular change of the L1 lumbar vertebra relative to the

L5 vertebra was approximately 5 degrees for the second
subject, LMP. These angular changes only measure the
rotations at the L1/L2 through L4/L5 lumbar joint centers.
Since TLC is the total rotation at all the mobile joint
centers, the rotations measured from the Nyquist and
Patrick[5] diagrams are only a portion of the TLC. If the
model has a rigid thorax and all rotation proportions are 1,
then the rotations at the Ll/L2, L2/L3, L3/L4, and the L4/L5
interspaces measured on the diagrams are only 4/6 of the
TLC. Therefore, the angular change equals (4/6)xTLC. The
TLC equals (6/4)x(angular change).

The TLC for the Nyquist and Patrick postures was
calculated by dividing the total of all the rotation
proportions by the total of the rotation proportions of the
L1/L2, L2/L3, L3/L4, and L4/L5 interspaces then multiplying
by the angular change. In equation form, this is:

TLC= (rotaLW'
(total of proportions for L1/L2, L2/L3, L3/L4, L4/L5)

Table 3 contains the calculated TLC for the two male

subjects and the pelvic rotations from the JOHNl position.

44

TABLE 3
TLC for Nyquist and Patrick[5]

DISTRIBUTION TLC (DEGREES)
NUMBER FOR SUBJECT
CJM LMP
1 10.5 7.5

2 9.3 6.6
3 11.7 8.4
4 10.5 7.5
5 11.8 8.4
6 12.1 8.6

7 14.8 10.6

8 16.2 11.6
9 10.9 7.8
10 10.2 7.3

2.8.3 Andersson et. al.[3]

Andersson et. al.[3] provided data for the position of
the pelvis and the relative position of the lumbar region.
He defined and measured the total lumbar angle, the sacral-
horizontal angle, the sacral-pelvic angle, and the pelvic-
horizontal angle, Figure 15. The total lumbar angle is

defined as

’...the angle between perpendiculars from a line drawn
along the superior surface of L1 and a second line
drawn along the sacral endplate.’[3]

The sacral-horizontal angle is defined as:
'...the angle between the sacral endplate and a
horizontal line intersecting the top of the sacrum at
the posterior corner.’[3]

The sacral-pelvic angle is defined as:
’...the angle between the line along the sacral end—
plate and a second line which intersects the line of
the sacral end-plate at the posterior corner of the
sacrum, and is drawn to the most superior point of the
acetabulum.'[3]

The pelvic-horizontal angle is defined as:
'...the angle between a line drawn from the superior
corner of the sacrum to the uppermost point on the

45

Angles measured lrom lhe radlographs ol subjects In lhe
study: 1 - lolal lumbar angle. 2 = sacral-horizontal angle. 3 -
sacral-pelvic angle. 4 - pelvic-horizontal angle.

Figure 15 - Pelvic angle definitions from
Andersson et. al.[3]

46

acetabulum and the horizontal line passing through the
superior-posterior corner of the sacrum.’[3]

A change in the pelvic horizontal angle is the rotation of
the pelvis.

A change in the total lumbar angle is related to a
change in TLC. The change in total lumbar angle and the
change in TLC are not identical because of a discrepancy in
the number of mobile joint centers. Total lumbar angle as
described by Andersson et. al.[3] measures only the rotation
due to the L5/S1 through Ll/L2 lumbar joint centers. TLC
measures the rotation of all the lumbar joint centers and
the mobile thoracic joint centers. Due to the difference in
the number of mobile joint centers, the change in the total
lumbar angle measured must be multiplied by an appropriate
factor to obtain TLC. For the case of even distribution of
motion and a rigid thorax, the total lumbar angle must be
multiplied by 6/5 to obtain TLC. For the case of non-even
distribution of rotation and a rigid thorax, the total of
all proportions must be added together and divided by the
proportions for the L5/Sl through L1/L2 lumbar joint
centers. For the case of a non-rigid thorax, the
proportions of rotation for each joint center must be added
together then divided by the total of the L5/Sl through
L1/L2 proportions. In equation form, this is:
Multiplication

factor = '
Total of Ll/L2, L2/L3, L3/L4, L5/Sl proportions

47
For this study, TLC is calculated by multiplying the total

lumbar angle by the multiplication factor. Table 4 contains
the multiplication factors.

Andersson[3] measured the pelvic horizontal angle, the
total lumbar angle, and the sacral-horizontal angle in one
standing and three seated postures, Table 5. For each
posture, Andersson had the participants sit in a vertical
seatback chair. The chair was equipped with a variable
position lumbar support which could be varied to either be
in back of the seatback, or in front of the seatback. For
the -2 posture, the lumbar support was positioned 2 cm to
the rear of the straight seatback. For the +2 posture, the
lumbar support was positioned 2 cm forward of the straight
seatback. For the +4 posture, the lumbar support was
positioned 4 cm forward of the straight seatback.

TABLE 4
Multiplication Factors

DISTRIBUTION MULTIPLICATION
NUMBER FACTOR
6/5
8/7
5/4
4.5/4
4.9/3.
4.7/3.
6.2/3.
6.7/3.
4.5/3.
0 4.1/3.

I—‘KDCDQQU‘IAWNH

\lkOkOkOQKD

Andersson[3] also measured a 0 posture. For this
posture the lumbar support was flush to the seatback. This

posture was not used for this research because for this

48
posture there was apparent sacral movement. There was

little difference between the sacral-pelvic angles for the
-2, +2, and the +4 postures. But the sacral-pelvic angle
was different for the 0 posture. This indicated that there
was sacral motion. Since the computer model used in this
research did not incorporate sacral movement, the 0 posture
could not be used.
TABLE 5
Data from Andersson et. al.[3]

POSTURE PELVIC HORIZONTAL TOTAL LUMBAR SACRAL HORIZONTAL

ANGLE ANGLE ANGLE
STANDING 63.8 degrees 59.8 degrees 38.0 degrees
-2 cm 34.3 9.7 27.5
+2 cm 47.5 29.9 28.8
+4 cm 54.7 46.8 28.3

Andersson et. al.[3] supplied no data on the position
of the thorax in his study. Therefore, to simulate the of
the Andersson et. al.[3] study, the pelvic horizontal angle
and the TLC for each posture was recreated. The resulting
posture was examined to find if it simulated a reasonable
posture of a human in a vertical seatback. The criteria
used for examining the resultant postures was to look at the
postures and see if the thorax was rotated unrealistically

rearward or forward to simulate a vertical seatback.

y‘ 901 o o 027' on 0- 0H 009‘ wi 1 f 9_l 1‘! .1 9

The method for comparing the model with the results

found in literature was a four step process.

49
1. Chose one or more representative rotation

distributions.

2. Using the rotation distribution(s) chosen in Step 1,
model the data from Robbins and Reynolds[4].

3. Using the rotation distribution(s) chosen in Step 1,
model the data from Nyquist and Patrick[5].

4. Using the rotation distribution(s) chosen in Step 1,

model the data from Andersson et. al.[3].

2.9.1 Step 1: Method for Selection of Representative

Rotation Distribution(s)

To choose between the rotation distributions, the joint
centers for each distribution were compared to the
corresponding joint centers of the first distribution. If
the distance between the compared joint centers was 12.7
millimeters or less, then the distributions were considered
essentially equivalent. The distance of 12.7 millimeters
(0.5 inch) was chosen because in the automobile seating
environment the position of the seated occupant is not
precisely defined, and a 12.7 millimeter difference is
negligible.

If the difference between the corresponding joint
centers was 12.7 millimeters or greater then the two
distributions were considered to be different. All
distributions which were found to be different from

distribution one were then used for Steps 2, 3, and 4.

50

2.9.2 Step 2: Method for Modeling Robbins and Reynolds

Study[4]

To model the Robbins and Reynolds[4] study, the
distributions chosen in Step 1 were used and the procedure
to obtain TLC outlined in section 2.8.1 was followed. To
rotate the torso into the seat, the angle between the T4/T5
interspace and the T8/T9 interspace was used. This line was
measured to be vertical on the diagram supplied by Robbins
and Reynolds[4]. The computer model was rotated about the
H-point until the line between the T4/T5 interspace and
T8/T9 interspace of the computer model was vertical. The
angle which the torso and the pelvis was rotated through was

called the pelvic rotation.

2.9.3 Step 3: Method for Modeling Nyquist and Patrick

Study[5]

To model the Nyquist and Patrick[5] study, the rotation
distributions chosen in Step 1 were used following the
procedure to obtain TLC found in Section 2.8.2. To rotate
the torso into the seat, the spineline angle was used. The
value for this angle was given in the Nyquist and Patrick[5]
publication for both male subjects. The angle which the
torso and the pelvis was rotated through was called the

pelvic rotation.

51

2.9.4 Step 4: Method for Modeling Andersson et. al. Study[3]

To model the Andersson et. al.[3] study, the
distributions chosen in Step 1 were used following the
procedure to obtain TLC as outlined in Section 2.8.3. To
rotate the torso into the seat, the pelvic angle was used.
Andersson et. al.[3] supplied the pelvic-horizontal angle.
The resultant postures were then viewed to see if they were
reasonable for a person in a vertical seatback.

A decision was then made if any of the distributions
could be eliminated on the basis of the resultant postures
or the pelvic rotations. The chosen distribution was the
distribution for which the pelvic rotations of Steps 2 and 3
were closest to the values given in the published data, and
the resultant postures were reasonable for a person in a
vertical seatback.

0 9‘ 900 o .HPQ ' o; 0, O _ 000‘ 5 o ' 9-i 1‘! o. -

This part of the verification process involved using
the Link Study[2] data for the lumbar region. To verify the
position of the pelvis relative to the thorax for the JOHN2
model, the same evaluation procedure which was used for the

JOHNl model was used again.

52
We

The questions addressed in this research are:

1. The effect of various rotation distributions on the

position of the thorax.

2. The position of the thorax relative to the pelvis,

e.g. the angular position of the pelvis and the
choice of the JOHNl or the JOHN2 model.

3. Placement of a computer human postural model into a

seat.

To find the answer to question 1, ten different
rotation distributions were tested for differences in the
positions of the vertebral joint centers. These differences
were considered significant if they were greater than 12.7mm
between the positions of the interspaces for a distribution
when compared to the positions of the same interspaces of
distribution one. Then using any rotation distributions
which were considered different, the data found in Robbins
and Reynolds[4] and Nyquist and Patrick[5] were simulated
with the JOHNl and JOHN2 models. The TLC calculated for the
published studies in Sections 2.8.1 and 2.8.2 was used as
input. Then the spine of the computer model was placed in
the same position as the spine in the literature. The
position of the pelvis of the computer model was then
compared to the position of the pelvis in the literature.

Data from Andersson[3] was simulated, using as input

the position of the pelvis and the TLC. The resultant

53

posture was then reviewed to examine if the posture could be
the posture of a person in a vertical seatback. This
procedure was first done for the JOHNI computer model, then
for the JOHN2 computer model.

To place the computer model into the posture found in
the literature, three methods were used. The first method
for Robbins and Reynolds[4] was to use as input TLC and a
thoracic body angle. The second method for Nyquist and
Patrick[5] was to use as input TLC and a lumbar spine angle.
The third method for Andersson[3] was to use as input TLC
and a pelvic angle.

The recommended method to place the computer model into
the seatback was chosen from these three methods. This
recommended method was the method which supplied results

which were closest to the results of the literature.

3. RESULTS AND DISCUSSION

The distance criteria used for deciding whether the
difference between two positions was significant was 12.7
millimeters. If the position of the same interspace was
different by a distance of 12.7mm or less for two different
rotation distributions, then the rotation distributions were
considered to be not significantly different.

Figure 16 is a representative multi-screen figure of
the rotation distribution evaluations for the JOHNl model.

Appendices 11-14 contain the differences between the
joint centers of distribution one and all other
distributions. The distances between the joint centers is
calculated as indicated in Figure 17. A negative difference
for the x coordinates indicates that the joint center for
that distribution is posterior to the same joint center for
distribution one. A negative difference for the 2 indicates
that the joint center for that distribution is below the
same joint center for distribution one.

Table 6 and Figure 18 contain the results of the
rotation distribution evaluations for 10 degrees flexion.

This table lists the distances for all interspaces from the

54

55

 

 

 

 

Figure 16 - Multi-screen model of distribution one
for the JOHNl model

56

T8/T 9 interspace
for Distribution 7

  
   
 

Distance

Delta Z T8/T9 interspace

for Distribution 1

 

 

Delta X

 

2 2
Distance=V(Delta X) + (Delta Z)

Figure 17 - Calculation of distances between interspaces

57

TABLE 6
Distances (mm) Between Interspaces

for the JOHNl Model
10 Degrees Flexion

 

DI STRIBUTION

INTER- NUMBER

SPACE 2 3 4 5 6 7 8 9 10
T5/T6 1.39 1.03 2.64 0.36 1.39 5.57 8.54 1.61 3.76
T6/T7 1.30 1.08 2.64 0.50 1.44 5.57 8.54 1.66 3.81
T7/T8 1.30 1.08 2.64 0.50 1.44 5.57 8.35 1.66 3.81
T8/T9 1.30 1.08 2.64 0.50 1.53 5.57 7.97 1.70 3.81
T9/T10 1.30 1.08 2.64 0.50 1.53 5.28 7.27 1.73 3.81
T10/T11 1.30 1.08 2.64 0.50 1.44 4.78 6.39 1.66 3.81
T11/T12 1.44 1.03 2.64 0.36 1.39 3.81 5.02 1.61 3.76
T12/L1 1.44 1.03 2.56 0.36 1.39 2.37 3.22 1.61 3.76
L1/L2 1.58 1.25 2.37 0.67 1.70 1.30 1.94 1.48 3.45
L2/L3 1.30 1.03 1.75 0.58 1.39 0.58 1.03 1.08 0.42
L3/L4 0.72 0.58 0.94 0.45 0.94 0.22 0.36 0.72 1.53
L4/L5 0.28 0.28 0.36 0.22 0.22 0.00 0.14 0.36 0.45
L5/S1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

58

_m
28

26

22

2O

12.7mm

 

Anlcriur ll

 

. 1m
Distance l.25 l 70 1.73

From i167 .
Distrilmlinn 0 I

(nnn) l ‘8 Distiibulion Nut bcr

 

l9

 

 

 

 

 

 

 

 

IJ

 

6

 

. . 7
l’uslcnur R S 5

 

 

 

 

[.1 l2. 7mm

20

Figure 18 - 10 degrees flexion for the JOHNI model

59
T5/T6 interspace down. The higher interspaces were not

listed because the distances are the same as the distance
for the T5/T6 interspace. The differences listed are
comparisons between the first rotation distribution with
equal rotations at each interspace in the lumbar spine and
each of the other nine rotation distributions. The largest
differences occurred between the first distribution and the
eighth distribution. This was expected since the eighth
rotation distribution was the distribution with the greatest
amount of thoracic mobility. Since none of the distances
exceeded 12.7mm the conclusion was that for 10 degrees
flexion, all the spinal rotation distributions were judged
to be equivalent.

Table 7 and Figure 19 contain the results of the spinal
rotation distribution evaluations for 10 degrees extension.
Once again the largest differences occurred for the eighth
rotation distribution. The largest difference occurred at
the T5/T6 interspace with a value of 8.58mm. Since this
value is less than the 12.7mm limit, the conclusion was
that, for 10 degrees extension, no difference existed
between the spinal rotation distributions.

Table 8 and Figure 20 contain the results of the spinal
rotation distribution evaluations for 20 degrees extension.
The only rotation distribution pattern which had distance
values greater than 12.7mm was distribution eight. This
rotation distribution involved thoracic mobility. For all

other rotation distribution patterns, there existed no

60

TAUILE '7
Distances (mm) Between Interspaces
for the JOHNl Model
10 Degrees Extension

 

DISTRIBUTION

INTER- NUMBER

SRAQE 2 3 4 5 6 7 8 9 .10
T5/T6 1.35 1.00 2.62 0.45 1.50 5.55 8.58 1.64 3.76
T6/T7 1.42 1.00 2.62 0.42 1.42 5.63 8.58 1.64 3.76
T7/T8 1.25 1.06 2.62 0.42 1.49 5.63 8.44 1.63 3.75
T8/T9 1.36 1.06 2.69 0.50 1.49 5.58 8.05 1.70 3.83
T9/T10 2.10 1.00 2.62 0.42 1.42 5.33 7.43 1.64 3.76
T10/T11 1.28 1.06 2.62 0.50 1.49 4.74 6.41 1.70 3.83
T11/T12 1.28 1.06 2.62 0.50 1.49 3.78 5.00 1.70 3.83
T12/L1 1.36 1.06 2.62 0.42 1.42 2.34 3.20 1.70 3.83
L1/L2 1.56 1.20 2.34 0.64 1.70 1.35 1.98 1.42 3.33
L2/L3 1.28 1.00 1.70 0.58 1.42 0.64 1.00 1.14 2.48
L3/L4 0.64 0.58 1.00 0.50 0.86 0.14 0.42 0.72 1.50
L4/L5 0.28 0.22 0.36 0.22 1.22 0.40 0.14 0.36 0.50
LS/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

30
28
26
24
22

20

Anterior a

Distance 2
Front
Distribution 0
One

(mm) 2

Posterior 8
10
12
I4
I6
I8

20

61

I 2.7mm

 

8.58

 

5.63

 

2.10

 

 

 

Distribution N mtgr

 

 

 

 

 

 

 

2 3 4—6—16 7 8 9 IO
0.64
I

.70 L70

 

 

 

 

 

2.62

 

3.83

 

12.7mm

Figure 19 - 10 degrees extension for the JOHNl

model

62

TABLE 8
Distances (mm) Between Interspaces
for the JOHNl Model
20 Degrees Extension

 

DISTRIBUTION

INTER- NUMBER

SPACE 2 3 4 5 6 7 8 9 10
T5/T6 2.69 2.13 5.30 0.92 2.90 11.19 17.20 3.32 7.57
T6/T7 2.69 2.05 5.23 0.86 2.90 11.19 17.26 3.32 7.57
T7/T8 2.62 2.12 5.30 0.92 2.97 11.19 16.84 3.39 7.57
T8/T9 2.69 2.12 5.23 0.92 2.90 11.19 16.04 3.32 7.57
T9/T10 2.62 2.12 5.23 0.92 2.97 10.70 14.68 3.32 7.57
T10/T11 2.62 2.12 5.23 0.92 2.97 9.56 12.82 3.32 7.64
T11/T12 2.69 2.19 5.30 0.92 2.97 7.57 10.11 3.39 7.23
T12/L1 2.62 2.12 5.23 0.92 2.97 4.60 6.36 3.32 7.57
L1/L2 3.05 2.48 4.74 1.35 3.39 2.62 3.82 2.97 6.79
L2/L3 2.55 2.13 3.54 1.28 2.97 1.06 1.84 2.27 5.02
L3/L4 1.42 1.06 1.84 0.92 1.77 0.42 0.85 1.49 2.97
L4/L5 0.50 0.42 0.64 0.42 0.57 0.00 0.22 0.64 0.92
LS/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

63

30
28
26
24

22

17.26

12.7mm
'2 -- an
m

 

 

Anterior 8

—_ 3.05

 

Distance 2
From
Distribution 0 Distribution Npmbcir

0m 2 3 4 s 6 7' s 9 w
(""") 2 -- - 135

4 _ 2.48 2.97 3 39

 

 

 

 

 

 

 

 

 

 

5.30
7.57

 

 

Posterior 3

 

IO

l2

 

14 l2.7mm

20

Figure 20 - 20 degrees extension for the JOHNl model

64

significant differences between the rotation distribution
patterns.

Table 9 and Figure 21 contain the results of the spinal
rotation distribution evaluations for 30 degrees extension.
Figures 22-30 are the comparisons at 30 degrees extension
TLC between distribution one and each of distribution two
through ten, respectively. As with 20 degrees extension,
the rotation distribution which included thoracic mobility,
distribution eight, had distances much greater than 12.7mm.
The only other rotation distribution which had distances
that exceeded 12.7mm was rotation distribution seven. This
rotation distribution also had thoracic mobility.
Therefore, all spinal rotation distributions which had a
rigid thorax were considered equivalent.

The evaluation of the spinal rotation distribution
revealed that all spinal motion distributions which did not
involve thoracic mobility can be considered to be
essentially equivalent. Distribution one was chosen as the
representative rotation pattern for rigid thorax models
because of its simplicity, assuming even distribution of
rotation. The only spinal rotation patterns which exceeded
the 12.7 mm criteria were the two rotation distributions
which involved thoracic mobility.

At this point in the research work, it was not
determined whether the preferred motion pattern for the
JOHNl model would be distribution one, distribution seven,

or distribution eight. For lumbar mobility only, there

65

TABLE 9
Distances (mm) Between Interspaces
for the JOHNl Model
30 Degrees Extension

 

DI STRIBUTION

INTER‘ NUMBER

SPACE 2 3 4 5 6 7 8
T5/T6 4.03 3.32 7.90 1.35 5.11 16.80 25.89 4.
T6/T7 3.96 3.18 7.82 1.42 4.39 16.80 25.97 4.
T7/T8 3.96 3.26 7.90 1.42 4.46 16.80 25.33 4.
T8/T9 3.93 3.21 7.86 1.45 4.42 16.77 24.11 5.
T9/T10 3.96 3.26 7.90 1.49 4.46 16.03 22.10 5.
T10/T11 3.96 3.26 7.90 1.42 4.46 14.40 19.28 4.
T11/T12 4.03 3.19 7.90 1.41 4.39 11.45 15.12 5.
T12/L1 3.96 3.19 7.90 1.35 4.46 6.92 9.46 4.
L1/L2 4.60 3.69 7.04 1.91 5.11 3.98 5.81 4.
L2/L3 3.89 3.11 5.17 1.84 4.32 1.70 2.83 3.
L3/L4 2.05 1.70 2.83 1.41 2.76 0.57 1.20 2.
L4/L5 0.67 0.57 0.99 0.71 0.85 0.10 0.36 0.
L5/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.

30
28
26
24
22
20
I 8

I6

Anterior 8

Distance 2
From
Distribution 0
One

(mm) 2

Posterior s

20

66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

391
16w
I2.7mm
4w
Distribution Ninmbir
2 3 4 s 6 7 9 m
Lm
1w
5.1! 5m
1w
iLw
12.7mm

Figure 21 - 30 degrees extension for the

JOHNl'model

67

 

0
O
Q
0
Distribution 2 Q
0
‘0
0
Oi

 

Distribution 1

Figure 22 - Comparison of distribution one and distribution
two at 30 degrees extension for the JOHNl model

68

 

Q
0
O
0
Distribution 1 O
O
O
O
9

 

Distribution 3

Figure 23 - Comparison of distribution one and distribution
three at 30 degrees extension for the JOHNl model

69

 

Distribution 1

Distribution 4

Figure 24 - Comparison of distribution one and distribution
four at 30 degrees extension for the JOHNl model

70

_A

 

Distribution 1

 

Distribution 5

' ' ' 'bution
Fi ure 25 - Comparison of distribution one and distri
g five at 30 degrees extension for the JOHNl model

71

 

Distribution 1

 

Distribution 6

Figure 26 - Comparison of distribution one and distribution
six at 30 degrees extension for the JOHNl model

72

 

Distribution 7

Distribution 1

Figure 27 - Comparison of distribution one and distribution
seven at 30 degrees extension for the JOHNl model

73

 

Distribution 8

Distribution 1

' ' ' ' d distribution
' 28 - Com arison of distribution one an
Flgur:ight at 38 degrees extension for the JOHNI model

74

 

Distribution 1 4?
9

 

Distribution 9

Figure 29 - Comparison of distribution one and distribution
nine at 30 degrees extension for the JOHNl model

75

 

 

 

Distribution 1

Distribution 10

Figure 30 - Comparison of distribution one and distribution
ten at 30 degrees extension for the JOHNl model

76
existed no differences between the rotation distributions.

A non-rigid thorax distribution resulted in significant
differences from the distributions with lumbar mobility
only. It was still unknown if the final rotation
distribution should have a rigid or a non-rigid thorax. To
discover which of the patterns best simulated human motion,
each of the three remaining distributions were used in

attempting to duplicate the results found in the literature.

1 '0; 0 R009?! '11

Model

To model the results found in Robbins and Reynolds[4],
the procedure in Section 2.9.2 was followed. Table 2
contains the calculated values of TLC from Robbins and
Reynolds[4] for each of the rotation distributions. Table
10 contains the results from the testing of the three
different motion distributions chosen in Section 3.1.

TABLE 10
Results of the Simulation of

Robbins and Reynolds[4]
for the JOHNl Model

DISTRIBUTION TLC PELVIC ROTATION
NUMBER (DEGREES) (DEGREES)

1 9.0 11.6

7 11.5 14.1

8 12.6 14.8

As noted in Section 2.8.1, the rotation of the pelvis

from the JOHNl position for the Robbins and Reynolds[4]

77
study is 12 degrees. Therefore, the computer model should

ideally give an indicated pelvic rotation of 12 degrees.
The pelvic rotation which was closest to the desired
rotation of 12 degrees was the 11.6 degree rotation for
distribution one. But the indicated pelvic rotations for
all of the three distributions were close to the desired 12
degree rotation. Furthermore, since the TLC and the
positions of the T4/T5 and the T8/T9 interspaces from the
Robbins and Reynolds[4] study were not precisely known, it
remained undetermined which of the rotation distributions
best modeled this study. However, all of the three motion
distributions resulted in a good agreement with the Robbins
and Reynolds[4] study. This reinforced the belief that the
original position of the pelvis in the UMTRI[1] position was

representative of human posture.

°n- a '0; 0_ t o,’ . o f. f A o - o, Moo-l

The second test was to compare the computer model
results to those of Nyquist and Patrick[5] following the
procedure outlined in Section 2.9.3. Nyquist and Patrick[5]
measured the angle from the vertical of the line between the
hip joint center and the ASIS. For the first subject, CJM,
this angle was -7 degrees. For the second subject, LMP,
this angle was -23.5 degrees. In the standing posture as
defined by Reynolds et al.[8], the line between the hip

joint center and the ASIS are is 36 degrees from the

78
vertical. Therefore the pelvic rotation from the standing

posture to the seated posture was -7-36=-43 degrees for CJM
and -23.5-36=-59.5 degrees for LMP. The pelvic rotation
from standing for JOHNl was -54 degrees. Thus, the pelvis
of subject LMP was rotated 11 degrees from the JOHNl
posture. The pelvis of subject CJM was rotated -5.5 degrees
from the JOHNl posture. This posed a problem for subject
LMP. The pelvis was rotated rearward from the JOHNl
posture. Furthermore, the TLC for LMP was approximately 5
degrees. Even if the TLC was 0 degrees, this posture cannot
be obtained from the JOHNl posture. The decision was made
to ignore the results of subject LMP. If the original
position of the pelvis in the JOHNl posture was found to be
incorrect in subsequent study, then this position of the
pelvis for LMP would be further explored.

Table 11 is the results of the computer evaluation of
subject CJM for each of the three distributions.
Distribution one resulted in the pelvic rotation value
closet to the 11 degrees pelvic rotation found by Nyquist
and Patrick[5]. However, all values of the rotation were
close to the desired rotation of 11 degrees. Since the TLC
was not accurately known and the rotations were all similar,
a choice of the best distribution to describe human motion
was still not possible. These results also reinforced the
belief that the original position of the pelvis in the JOHNl

posture was representative of human posture.

79

TABLE 11
Results of the Simulation of
Nyquist and Patrick[5]
for the JOHNl Model

DISTRIBUTION TLC PELVIC ROTATION
NUMBER (DEGREES) (DEGREES)
1 10.5 13.0
7 14.8 13.9
8 16.2 13.9
‘ ii 1 '0! 0 430‘ on e -. o h: OuN 900‘

The third test to discriminate between the rotation
distributions was to model the Andersson et al.[3] data
which is in Table 5. The first step was to determine if the
position of the pelvis in the JOHNl posture was similar to
the position of the pelvis in any of the three Andersson
postures.

For the UMTRI[1] pelvis in the standing posture as
defined by Reynolds, et al.[8] the sacral horizontal angle
is 46.4 degrees. This is the angle from the horizontal of a
line drawn along the top of the sacrum, Figure 15. In the
standing posture for Andersson et. al.[3], the sacral
horizontal angle measured 38 degrees. This was a
discrepancy of 8.4 degrees. Since the standing posture may
differ between the Andersson et. al.[3] study and the
Reynolds et. al.[8] study, this 8.4 degrees discrepancy was
considered to be a constant which must be carried through
the calculations. When the UMTRI[1] pelvis was rotated into
the UMTRI[1] posture, the sacral horizontal angle became

46.4-54.07= —7.67 degrees. From the UMTRI[1] drawing, the

80
measured angle from the horizontal of the superior surface

of the L1 vertebra was 28 degrees. The drawing was used for
this measurement because no better data were available.

This information was used to calculate the total lumbar
angle. Andersson et. al.[3] defined the total lumbar angle
as the angle between the superior surface of the L1 vertebra
and the superior surface of the sacrum. In the UMTRI[1]
posture, this total lumbar angle was calculated as 28-7.64=
20.33 degrees. Assuming the 8.4 degrees discrepancy due to
sacral position was a constant, the total lumbar angle as
defined by Andersson et. al.[3] was 20.33-8.4= 11.9 degrees.
This value compared extremely well with the value of 9.7 for
the total lumbar angle as found by Andersson et. al.[3] for
the case of the -2 lumbar support. The 2.2 degree
difference can be caused by differences between subjects or
errors in the UMTRI[1] diagram or its measurement.

At this point in the research, the position of the
pelvis relative to the spine is well supported by available
data. A good correlation was found between the position of
the pelvis relative to the lumbar spine in the UMTRI[1]
posture and the position of the pelvis relative to the
lumbar spine in the -2 posture found in Andersson et.
al.[3]. Also, the position of the pelvis for the JOHNl
model correlated well with the data of Robbins and
Reynolds[4] and Nyquist and Patrick[5] by using the computer

motion program.

81
A single representative rotation distribution still had

not been found. Three distributions (1,7, and 8) remained
as possibilities. The last step to decide between the
distributions was to model the -2, +2, and +4 Andersson et.
al.[3] postures following the procedure in Section 2.8.4
Since the position of the pelvis relative to the lumbar
spine in the UMTRI[1] original position and the -2 lumbar
support case of Andersson match well, the next step was to
consider the entire body rigid and rotate about the H-point
to place the torso into the -2 lumbar support posture. For
Andersson[3], the pelvic-horizontal angle is 63.8 degrees in
the standing posture and 34.3 degrees in the -2 lumbar
support posture. Therefore, from the standing posture to
the -2 lumbar support posture the pelvis was rotated 63.8-
34.3= 29.5 degrees. For the original UMTRI[1] position the
pelvis was rotated 54.07 degrees from the standing position
to the seated position. So to achieve the -2 lumbar support
posture, the pelvis, and therefore the body of the computer
model must be rotated 54.07-29.5= 24.07 degrees forward,
which is the difference between the two pelvic rotations
from the standing posture. Since no data was given for the
position of the thorax or the lumbar spine, the only way to
match the Andersson et. al.[3] -2 posture was to rotate the
body through the 24.07 degrees to find if the resulting
position was representative of a 50th percentile man in a
vertical seatback. Figure 31 is the resulting posture.

Upon inspection the posture appears to be a representation

82

 

 

Figure 31 - Andersson -2 posture for the JOHNl model

83
of a 50th percentile man in a vertical seatback posture.

The bottom of the torso of the model appears to be in a
position approximately directly above the posterior of the
pelvis. This is what should happen in a vertical seatback.
This posture will be the same for all rotation distributions
since the torso was rotated about the H-point with no change
in TLC.

The pelvic—horizontal angle is 47.5 degrees in the +2
lumbar support posture. Therefore, from the standing
posture to the +2 lumbar support posture of Andersson et.
al.[3], the pelvis was rotated 63.8-47.5= 16.3 degrees
rearward. So from the nominal JOHNl model seated posture,
with the pelvis rotated 54.07 degrees from the standing
posture, to the +2 posture the computer model must be
rotated 54.07-16.3= 37.77 degrees forward. For the +2
lumbar support posture, the total lumbar angle is 20.2
degrees.

The pelvic-horizontal angle is 54.7 degrees in the +4
lumbar support posture. Therefore, from the standing
posture to the +4 lumbar support posture of Andersson et.
al.[3], the pelvis was rotated 63.8-54.7= 9.1 degrees
rearward. So from the nominal UMTRI[1] seated posture to
the +4 posture the computer model must be rotated 54.07-9.1=
44.52 degrees. For the +4 lumbar support posture, the total
lumbar angle is 46.8-9.7: 37.1 degrees.

TLC is calculated by multiplying the total lumbar angle

by the multiplication factor, calculated in Section 2.8.3.

84
Table 12 shows the values of the calculated TLC for each

posture and rotation distribution. Figures 32-37 show the
resulting postures. Figure 38 is a composite picture of the
three distributions in the +2 posture. For the +2 cm
position, the resultant postures for all three distributions
appear to be representative of a person in a vertical
seatback. Figure 39 is a composite picture of the three
distributions in the +4 posture. By viewing the +4
postures, it can be seen that only distribution one resulted
in a posture representative of a person in a vertical
seatback. For distributions 7 and 8 the top of the thorax
is rotated too far rearward to represent a person in a
vertical seatback.

It was concluded that the preferred rotation
distribution is distribution one, the case of even
distribution of rotation with a rigid thorax. Distribution
one resulted in superior postures for simulating the
Andersson et. al.[3] study and in superior pelvic rotations
for simulating Robbins and Reynolds[4] and Nyquist and

Patrick[5].

85

O
O
O
O
O

 

 

 

Figure 32 - Andersson +2 posture for the JOHNl model
with distribution one

86

 

 

Figure 33 - Andersson +4 posture for the JOHNl model
with distribution one

87

O
O
O
O
O

 

Figure 34 - Andersson +2 posture for the JOHNl model
with distribution seven

88

° 0 oo¢N°°°"°oo

   
 

Figure 35 - Andersson +4 posture for the JOHNl model
with distribution seven

89

 

Figure 36 - Andersson +2 posture for the JOHNl model
with distribution eight

90

 

 

Figure 37 - Andersson +4 posture for the JOHNl model
with distribution eight

91

   

Distribution 7

a
4; Distribution 8

"'-oo::°"\

 

Figure 38 - Composite figure of Andersson +2 posture for the
JOHNl model for distributions one,seven, and eight

92

   

Distribution 8

 

0
Distribution 1 °° °° Distribution 7

 

Figure 39 - Composite figure of Andersson +4 posture for the
JOHNl model for distributions one, seven, and eight

93
TABLE 12

Andersson et. al.[3] TLC
and Pelvic Rotations for

the JOHNl Model

 

DISTRIBUTION TLC FOR PELVIC ROTATION FOR
NUMBER POSTURES POSTURES IN DEGREES
+2 +4 +2 +4
1 24.2 44.5 37.77 44.52
2 23.1 42.4 37.77 44.52
3 25.3 46.4 37.77 44.52
4 22.7 41.7 37.77 44.52
5 25.4 46.6 37.77 44.52
6 25.7 47.1 37.77 44.52
7 31.9 58.5 37.77 44.52
8 34.7 63.7 37.77 44.52
9 23.3 42.8 37.77 44.52
10 22.1 40.6 37.77 44.52

At this point, the angular position of the pelvis

relative to the lumbar spine and the thorax appears to be in

good agreement with published literature. But the JOHNl

model is only one possible model for the skeleton. Another

possible model of the

longer lumbar spine.

the lumbar spine does

the JOHN2 model.

But

skeleton is the JOHN2 model with a
The position of the pelvis relative to
not change between the JOHNl model and

the position of the pelvis relative to

the thorax does change between the two models.

The JOHN2 model was used to investigate if the

different lumbar region would affect the results which had

already been found.

To accomplish this, the entire testing

process as described in Section 2.9 was redone with the

94
JOHN2 model, instead of the JOHNl model. The procedure to

test for significant differences among the ten possible
rotation distributions as outlined in Section 2.8.1 was
redone with the JOHN2 model. Appendices 15-18 contain the
differences between the coordinates for the multi-screen
models of the JOHN2 model. The distances between joint
centers are calculated as indicated in Figure 17.

Table 13 and Figure 40 contain the results for 10
degrees flexion. Since no value for the distances exceeded
12.7mm, the conclusion was that there was no significant
differences between the various distributions for 10 degrees
flexion.

Table 14 and Figure 41 contain the results of the
testing for 10 degrees extension. As in the case for 10
degrees flexion, no value for the distances exceeded 12.7mm.
Therefore, there was no significant differences between the
various distributions for 10 degrees extension.

Table 15 and Figure 42 contain the results of the
testing for 20 degrees extension. The only distributions
which resulted in values in excess of 12.7mm were the two
distributions seven and eight which involved thoracic
motion. No other distributions were significantly different
than the nominal distribution of even rotation.

Table 16 and Figure 43 contain the results of the
testing for 30 degrees extension. Figures 44-52 are the
comparisons at 30 degrees extension TLC between distribution

one and each of distributions two through ten, respectively.

TABLE 1 3
Distances (mm) Between Interspaces

95

for the JOHN2 Model
10 Degrees Flexion

 

DISTRIBUTION

INTER- NUMBER

SPACES 2 3 4 5 6 7 8 9 10
T5/T6 1.61 1.30 3.22 0.50 1.75 6.01 9.20 1.97 4.61
T6/T7 1.66 1.25 3.22 0.58 1.75 6.04 9.23 1.97 4.57
T7/T8 1.66 1.25 3.22 0.50 1.75 6.04 9.04 1.97 4.57
T8/T9 1.58 1.34 3.18 0.58 1.84 6.04 8.57 2.06 4.65
T9/T10 1.66 1.39 3.22 0.58 1.84 5.72 7.97 2.06 4.70
T10/T11 1.66 1.25 3.18 0.50 1.75 5.26 7.09 1.97 4.57
T11/T12 1.66 1.30 3.22 0.50 1.75 4.25 5.73 1.97 4.61
T12/L1 1.66 1.30 3.22 0.58 1.84 2.73 3.89 1.97 4.61
Ll/LZ 1.72 1.48 2.87 0.72 2.06 1.61 2.33 1.75 3.79
L2/L3 1.66 1.25 2.20 0.67 1.75 0.81 1.30 1.39 3.09
L3/L4 0.78 0.67 1.03 0.64 1.03 0.28 0.50 0.89 1.70
L4/L5 0.22 0.22 0.36 0.22 0.28 0.00 0.14 0.36 0.50
L5/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Anterior 8

N

Distance
From
Distribution 0
One

(tnnt) 2

Posterior 8

20

96

12.7mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4.70
3.22
La 1% L”
0.72

[2 3 4 s 6 7 a 9 IO

Distribution Numbt r
Ln

am
9.23

 

I2.7ttnn

Figure 40 — 10 degrees flexion for the JOHN2

model

Distances

97

TABLE 14

(mm) Between Interspaces
for the JOHN2 Model
10 Degrees Extension

 

DISTRIBUTION

INTER- NUMBER

SPACESA, 2 3 4 5 6 7 8 9 10
T5/T6 1.91 1.12 3.00 0.40 1.61 6.36 9.51 1.75 4.42
T6/T7 1.70 1.28 3.26 0.50 1.84 6.10 6.28 1.92 4.61
T7/T8 1.64 1.28 3.19 0.50 1.78 6.05 9.13 1.98 4.61
T8/T9 1.77 1.22 3.14 0.54 1.72 6.16 8.71 1.94 4.55
T9/T10 1.64 1.28 3.26 0.58 1.78 5.82 7.99 2.06 4.69
T10/T11 1.64 1.42 3.26 0.58 1.84 5.30 7.10 2.06 4.61
T11/T12 1.64 1.28 3.26 0.58 1.78 4.34 5.70 2.06 4.61
T12/L1 1.64 1.28 3.19 0.50 1.70 2.84 3.91 1.98 4.53
L1/L2 1.99 1.51 2.85 0.73 2.07 1.62 2.33 1.79 4.13
L2/L3 1.64 1.35 2.20 0.78 1.84 0.72 1.14 1.42 3.12
L3/L4 0.86 0.64 1.00 0.50 1.00 0.28 0.45 0.78 1.64
L4/L5 0.32 0.22 0.28 0.28 0.28 0.10 0.10 0.28 0.50
LS/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

98

30
28
26
24
22

20

I2.7mm

 

'0 _ 9.51

 

Anterior s
6.36

 

Distance 2 L9!

From [_—
Distribution 0 Distribution Nttntbqr
Otte 2 3 4 '.‘fi 6 7
(mm) 2 ins
LSI

4 2.07 2-06
3.26
6 4.69

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Posterior 8

 

14 12.7mm

20

Figure 41 - 10 degrees extension for the JOHN2 model

TABLE 15

99

Distances (mm) Between Interspaces

for the JOHN2 Model
20 Degrees Extension

 

DI STRIBUTION

INTER- NUMBER

SPACES 2 3 4 5 6 7 8 9 10
T5/T6 3.33 2.62 6.37 1.00 3.54 12.25 18.67 3.96 9.20
T6/T7 3.26 2.62 6.44 1.06 3.61 12.18 18.60 4.03 9.20
T7/T8 3.26 2.62 6.44 1.06 3.61 12.18 18.31 4.03 9.20
T8/T9 3.80 2.69 6.51 1.14 3.61 12.18 17.44 4.03 9.27
T9/T10 3.33 2.62 6.44 1.14 3.54 11.68 16.08 3.96 9.27
T10/T11 3.19 2.62 6.51 1.14 3.61 10.54 14.23 4.03 9.27
T11/T12 3.26 2.69 6.51 1.13 3.61 8.56 11.39 4.03 9.27
T12/L1 3.33 2.62 6.37 1.00 3.54 5.59 7.78 3.96 9.20
L1/L2 3.82 2.97 5.73 1.49 4.10 3.18 4.74 3.54 8.20
L2/L3 3.33 2.62 4.31 1.41 3.61 1.49 2.48 2.76 6.22
L3/L4 1.56 1.28 2.13 1.06 2.13 0.42 0.92 1.70 3.40
L4/L5 0.57 0.36 0.72 0.50 0.58 0.10 0.28 0.72 1.00
LS/Sl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

100

28
26
24
22

20
18.67

 

I2.7innt
[2 122

 

 

Anterior 3

4 __ 3.30

 

Distance 2
Front . . .
Distribution 0 Distribution Numbg

One 2345678910
(mm) 2

 

 

 

 

 

 

 

 

li.49

 

 

 

4 __ 2.97
3.61
6 4.03

 

 

 

Posterior 8 6.5!

 

 

 

9.27

 

l4 I2.7mnt

20

Figure 42 - 20 degrees extension for the JOHN2 model

101

TABLE 16 _

Distances (mm) Between Interspaces
for the JOHN2 Model
30 Degrees Extension

 

DISTRIBUTION
INTER- NUMBER
SPACES 2 3 4 5 6 7 8
T5/T6 4.95 3.96 9.68 1.64 5.38 18.23 27.89 5.97
T6/T7 4.95 3.89 9.62 1.63 5.38 18.30 27.96 5.97
T7/T8 5.02 3.89 9.62 1.56 5.31 18.30 27.47 5.97
T8/T9 4.95 3.89 9.62 1.70 5.38 18.30 26.13 5.97
T9/T10 4.88 3.96 9.68 1.63 5.38 17.53 24.16 5.97
T10/T11 4.95 3.89 9.62 1.56 5.31 15.98 21.35 5.97
T11/T12 4.88 3.96 9.68 1.70 5.45 12.96 17.20 5
T12/L1 4.95 3.96 9.60 1.64 5.38 8.44 11.62 5
L1/L2 5.66 4.53 8.67 2.33 6.24 4.69 7.02 5.
L2/L3 4.88 3.89 6.52 2.12 5.38 2.27 3.68 4.
L3/L4 2.41 1.84 3.11 1.56 3.11 0.71 1.49 2
L4/L5 0.71 0.71 1.06 0.78 0.92 0.00 0.28 1
L5/Sl 0:00 0.00 0.00 0.00 0.00 0.00 0.00 0

102

30

28 27.96

 

26
24
22

20
I8 18.3

 

I2.7ttitn

 

Anterior 3

6 ___ 5m

 

Distance 2
Front
Distribution 0 Distribution Niinibt r

One 2 3 4 5 6 7 8 9 l0
(iitiit) 2

4 2.33

 

 

 

 

 

 

 

6 4.53

 

 

 

 

5.97
Posterior s 6.24

 

 

 

9.68

 

 

 

I4 12.7mm

 

13.82

Figure 43 - 30 degrees extension for the JOHN2 model

103

 

Distribution 2

 

Distribution 1

' ' ' ' ' ' 'bution
Fi ure 44 - Comparison of distribution one and distri
9 two at 30 degrees extension for the JOHN2 model

104

 

O
Q
0
9
0
Distribution 1 ?
9 l
9 i
O /
0
0
O
0>
_ 0
'0
‘0
‘0
9 Distribution 3

Figure 45 - Comparison of distribution one and distribution
three at 30 degrees extension for the JOHN2 model

105

 

Distribution 1

03°
9:
9:
s.
9:

 

Distribution 4

Figure 46 - Comparison of distribution one and distribution
four at 30 degrees extension for the JOHN2 model

106

 

o
o
o
0
Distribution 1 g
0
o
o
O

 

Distribution 5

Figure 47 - Comparison of distribution one and distribution
five at 30 degrees extension for the JOHN2 model

107

 

‘%
Distribution 1 Q
‘%

 

Distribution 6

Figure 48 - Comparison of distribution one and distribution
six at 30 degrees extension for the JOHN2 model

108

 

 

. . o 0
Distribution 7 0°

 

Distribution 1

Figure 49 - Comparison of distribution one and distribution
seven at 30 degrees extension for the JOHN2 model

109

 

 

O o
o 0
Distribution 8 g:

 

0 Distribution 1

' ' ' ' e and distribution
’ - Com arison of distribution on
Flgurziggt at 33 degrees extenSion for the JOHN2 model

110

 

‘%
Distribution 1 g
‘%

 

Distribution 9

Figure 51 - Comparison of distribution one and distribution
nine at 30 degrees extension for the JOHN2 model

111

 

 

Distribution

Distribution 10

Figure 52 - Comparison of distribution one and distribution
ten at 30 degrees extension for the JOHN2 model

112

For this case, there were three distributions which had
values much greater than 12.7mm. As with the previous
results in Section 3.1, the two distributions which involved
thoracic mobility had values which exceeded 12.7mm. In
addition distribution ten had differences which exceeded
12.7mm. This distribution had a rigid thorax. Therefore,
for simulating the data in Robbins and Reynolds[4], Nyquist
and Patrick[5], and Andersson[3], four distributions were
used. Distributions one and ten represented the rigid
thorax models and distributions seven and eight represented

the models which incorporated a non-rigid thorax.

. 'u_ e :0! o '099°1 elQ " to q ‘ 0 h‘ 0HN2

The next procedure was to model the Robbins and
Reynolds[4] study as outlined in Section 2.9.2. Once again
the TLC, found in Table 17, was entered into the motion
program and the model was rotated about the hip joint center
to place the line between the T4/T5 and the T8/T9 vertebral
interspaces of the computer model on the same angle from the
vertical as the same line in the Robbins and Reynolds[4]
study. Table 17 gives the rotations of the computer model
for each of the four different distributions. Distribution
one resulted in the rotation value closet to the desired
value of 12 degrees. But as was the case with the JOHNl

model, in Section 3.2, all rotations were close to the

113

desired value. Therefore, no conclusive decision was made

at this point.

TABLE 17
Results of the Simulation of
Robbins and Reynolds[4]
for the JOHN2 Model

DISTRIBUTION TLC PELVIC ROTATION
NUMBER (DEGREES) (DEGREES)
l 9.0 11.6
7 11.5 14.1
8 12.6 14.8
10 6.5 9.1
7 in- e '0 o 0L: e e 'g _' . -0 ‘ Oat. 09:1

Once again, the Nyquist and Patrick[5] study was
modeled as described in Section 2.9.3 using only the first
subject since the reliability of the pelvic angular position
had been supported. The same method was used as before,
estimating the TLC and matching the spine lines. Table 18
contains the results of the simulations for each of the four
distributions. For this set of simulations, the
distribution which gave the best fit to the desired value of
11 degrees was distribution one. The worst value was from
distribution ten. Because all values were still close to

each other, no conclusion could be made.

114

TABLE 18
Results of the Simulation of
Nyquist and Patrick[5]
for the JOHN2 Model

DISTRIBUTION TLC PELVIC ROTATIONS
NUMBER (DEGREES) (DEGREES)

1 10.5 13.2

7 14.8 14.1

8 16.2 14.1

10 10.2 14.5

3 't- 1 '01 0 L10‘_ 0 e . . for h: OHN2 Moo-l

The last set of tests involved modeling the Andersson
et. al.[3] data as performed earlier in Section 3.4. Table
12 contains the values of pelvic rotation and TLC for each
distribution. Figures 53-60 show the resulting postures.
Figure 61 is a composite figure of the +2 postures. Figure
62 is a composite figure of the +4 postures. The two +4
postures for the non-rigid thorax cases (distributions 7 and
8) clearly are not representative postures for a man in a
vertical seat back. The top of the torso of the model in
each of these postures is rotated too far rearward to
represent a person in a vertical seatback. All postures for
distributions one and ten appear to be represent a person in
a vertical seatback. Unlike for the JOHNl model, no
decisive conclusion could be drawn from the simulation of
the Andersson et. al.[3] study. For both distribution 1 and
distribution 10, the resultant postures appear to represent

a person in a vertical seatback.

115

 

 

 

Figure 53 - Andersson +2 posture for the JOHN2 model
with distribution one

116

 

 

Figure 54 - Andersson +4 posture for the JOHN2 model
with distribution one

117

O
O
O
O
O

 

 

Figure 55 - Andersson +2 posture for the JOHN2 model
with distribution seven

118

 

 

Figure 56 - Andersson +4 posture for the JOHN2 model
with distribution seven

119

 

 

Figure 57 - Andersson +2 posture for the JOHN2 model
with distribution eight

120

   
 

Figure 58 - Andersson +4 posture for the JOHN2 model
with distribution eight

121

 

   

 

Figure 59 - Andersson +2 posture for the JOHN2 model
with distribution ten

122

O
O
O
O
O

 

Figure 60 - Andersson +4 posture for the JOHN2 model
with distribution ten

123

 

   

Distribution 8

Distribution 1

Figure 61 - Composite figure of Andersson +2 posture for the
JOHN2 model for distributions one, seven, eight, and ten

124

       

 

 

- O
)lflfgo _
a Distribution 8

on
no Distribution 7

Figure 62 - Composite figure of Andersson +4 posture for the
JOHN2 model for distributions one, seven, eight, and ten

125

By examining the results of the simulations of the
literature, the conclusion was reached that distribution one
would be the preferred distribution. The reasoning for this
decision involved two factors. First, distribution one gave
slightly better results for both the Robbins and Reynolds[4]
study and the Nyquist and Patrick[5] study. Second, by
choosing distribution one, both models would then use the

same rotation distribution.

h ' H M l r HN 1

Judging on a basis of rotational position of pelvis and
rib cage, the JOHNl model and the JOHN2 model could be
considered equivalent since both used the same rotation
distribution. Also, the pelvic angular position was correct
for both the JOHNl model and the JOHN2 model. The choice
still remained between using the JOHNl model or the JOHN2
model.

To decide which skeletal model to use, the JOHNl model
or the JOHN2 model, a full scale drawing of the JOHN2 model
was compared to the original UMTRI[1] drawing. There was
not as large a distance between the surface of the buttocks
and the bottom of the pelvis for the JOHN2 model (19mm) as
there was for the JOHNl model (46mm). Furthermore, there
was more confidence in the exterior of the UMTRI[1] study
than in the skeletal placement in that study. Since the

motion distribution did not change between the models and

126
the pelvis of the JOHN2 model was closer to the surface of

the buttocks, the JOHN2 model was chosen as the skeletal

model.

W

The last problem to solve in this research study was
how the model should be moved into the seatback. Up to now
the automobile industry has simply moved the 2-D drawing
template and the three dimensional 3-D H-point machine into
a designated seatback angle as measured on the template and
the machine. This method is no longer valid for automobile
seats which use lumbar support because the 2-D template and
the 3-D H-point machine rotate forward on the lumbar
support. But since the contour between the seatback and the
human body is not known, the computer model cannot be simply
rotated into the seatback angle. Therefore, a new method
must be found which can move the model into the seatback.

To move the computer model in the simulations of
published studies, several methods were used. For the
Robbins and Reynolds[4] study, the TLC was known with the
angle from the T4/T5 interspace to the T8/T9 interspace.

For the Nyquist and Patrick[5] study, the TLC was known with
the spine line angle, the angle from the vertical of the
line between the L5/Sl interspace and the T12/L1 interspace.
For the Andersson et. al.[3] study, the TLC was known with

the pelvic-horizontal angle. In each different case, the

127

TLC and one body angle was known. The angle could be an
angle defined within a rigid part of the body, such as the
pelvic angle of the Andersson et. al.[3] study, or the
thoracic angle for Robbins and Reynolds[4]. Alternatively,
the angle could be an angle between two points in a non-
rigid section of body, such as the spine line of Nyquist and
Patrick[5]. Each of the three methods gave good results, as
shown in sections 3.2, 3.3, 3.6, and 3.7. The method which
gave results which were closest to the data in the
literature was the method for simulating Robbins and
Reynolds[4]. Therefore, to position the model in a seat
design space, the seat designer should no longer use the
seatback angle, but instead use the desired TLC and the

angle between two defined points on the rigid thorax.

4. SUMMARY AND CONCLUSIONS

A computer modeling package was used to create a two-
dimensional human postural model for automobile seating.
The computer modeling package utilized in this research was
the SDRC IDEAS 3.8 GEOMOD[10] Solid Modeling Package run on
a SUN SYSTEM 3/60.

The data for the modeling of a skull were provided by
Hubbard[9]. The data for the modeling of a skeletal pelvis
were provided by Reynolds, et al.[8]. The data for a two-
dimensional representation of the thorax were provided from
drawings and data supplied by UMTRI[1]. To create the
individual body segments, the Object Modeling option of the
GEOMOD[10] package was used.

The two-dimensional representation of the thorax was
modeled with data and drawings from UMTRI[1]. To form the
two-dimensional models of the skull and the pelvis, three-
dimensional models were first created. Two-dimensional
models were then formed from these three-dimensional
representations.

The individual skeletal segments were then assembled to
form a representation of a human seated posture. Two
postural models were created. The first model was a
recreation of the posture supplied in the UMTRI[1] data.

This study measured the external landmarks of the body and

128

129

the position of the skeleton was positioned on the basis of
these external landmarks. This model was called the JOHNl
model. A second model was formed because the bottom of the
pelvis in the UMTRI[1] study was 46mm from the surface of
the buttocks. In this second model, the original JOHNl
model was modified with a 29.6mm longer lumbar spine created
using data supplied by the Link Study[2]. This model was
called the JOHN2 model.

The initial orientation of the JOHNl model in the seat
design space was based on the data from UMTRI[1]. Since the
JOHN2 model was a replica of the JOHNl model except for the
length of the lumbar spine, all points, from the T12/L1
interspace up for the two models were superimposed. The
pelvis of the JOHN2 model was lower in the body than the
pelvis of the JOHNl model. This resulted in a distance of
only 19mm between the surface of the buttocks and the bottom
of the pelvis for the JOHN2 model.

The spine was modeled as a chain of rigid links for
bending in the mid-sagittal plane. The effect of joint
center movement could not be investigated since a rigid link
model was used. Spinal bending was simulated through the
use of a computer motion program. The necessary input for
the computer motion program was the Total Lumbar Curvature
(TLC) and one body angle. The TLC was defined as TLC=
Pelvic Rotation-Thoracic Rotation.

From modeling several different distributions of spinal

mobility, there was no significant difference found between

130
the positions of the joint centers for models which had only

lumbar motion and a rigid thorax based on a criteria of
112.7mm. At the largest value of spinal curvature (TLC = 30
degrees), there were significant differences of greater than
i12.7mm between the positions of the joint centers when a
model which had a non-rigid thorax was compared with a model
which had a rigid thorax. To decide if a non-rigid thorax
was needed for an automobile postural model, the results of
studies by Robbins and Reynolds[4], Nyquist and Patrick5],
and Andersson[3] were recreated. The closest agreements
with published data for both the JOHNl model and the JOHN2
model were obtained using an even distribution of lumbar
rotation and a rigid thorax.

To move the postural model into a simulated seatback,
the seat designers should use as input the desired TLC and
the angle of one rigid part of the thorax. The motion
program can be modified so that any rigid part of the thorax
can be used by the designer. For seat design, the angle for
one part of the rigid torso must be known. Up to this point
in seat design the known angle has been on the torso of the
2-D template and the 3-D H-point machine. Now, the known
angle must be on the thorax of the computer model. Once a
force-deflection model has been created at some future time
it will no longer be necessary to know one angle on the
rigid thorax because the computer model can then simulate
the settling into a car seat. Until a force-deflection

model has been created, the seat designer must specify an

131
angle on the rigid thorax. For this research the angle from

the vertical between the T8/T9 and the T4/T5 interspaces was

used.

5. LIMITATIONS AND FUTURE WORK

The positions of the pelvis, spine, thorax, and head
are still not exactly known for the seated posture. The
angle of the pelvis relative to the spine and the position
of the thorax appear to be consistent with the available
literature. However, not much literature was available.
Further research should also be conducted on the positions
of the pelvis, ribcage and head in the seated posture.
Additional knowledge is needed to confirm or reject the
preference of the JOHN2 model over the JOHNl model.

Using the main conclusion from this study of a uniform
distribution of lumbar rotation, a rigid thorax, and the
placement of rib cage relative to the pelvis for zero lumbar
curvature, the large male and small female could be
simulated. This might best be done by using the geometric
descriptions of these extreme sizes found in the UMTRI
report [1].

The effects of joint center mobility and sacral motion
were not explored in this project. Research should be done
to investigate the effects of both sacral motion and joint
center mobility. The cervical region also was not
investigated in this study. A study similar to this lumbar
research should be conducted for the cervical region.

Furthermore, a force-deflection model is needed to

132

133
adequately describe the position of a computer model in the

seating environment.
It is hoped that this study will provide a new means to

represent human postures for seating design.

APPENDICES

APPENDIX 1
COORDINATES FOR THE
LEFT HALF OF THE SKELETAL PELVIS

POINT COORDINATES
NUMBER

1 -77,100,74
3 -71,95,73

5 -71,87,-30
7 -85,66,-21
9 -103,63,7
11 ~97,58,-14
13 -134,37,29
15 -131,35,11
17 -126,44,4
19 -123,45,-11
21 -111,52,-17
23 -89,60,-45
25 -90,54,-56
27 -84,81,-58
29 -9l,72,-85
31 -99,48,-111
33 -96,60,-118
35 -73,94,-57
37 -61,73,-88
39 -41,25,-121
41 -45,29,-106
43 -29,34,-79
45 ~13,41,-78
47 ~56,62,-50
49 -12,9,-94
51 -24,68,-71
53 -21,64,-53
55 -25,87,-45
57 -34,98,-44
59 -16,95,-33
61 -60,95,-1
63 -15,104,-12
65 -40,108,1
67 -20,132,30
69 -46,133,42
71 -57,110,27
73 -70,89,5

75 -69,82,17
77 -49,96,10
79 -24,105,5
81 -3,112,9
127 -56,80,-5
129 -55,96,15

134

135
APPENDIX 1 (cont’d)

131 -51,109,37
133 ‘94,0,24
134 -104,0,32
135 -122,0,35
136 -134,0,18
137 -141,0,-2
138 -143,0,-25
139 '139,0,-60
140 '130,31,4
142 -99,26,40
144 -83,56,28
146 '87,44,48
148 '89,26,48
150 '85,17,44
152 '97,17,38
154 -8l,l7,31
156 '75,51,-5
158 '81,67,-114
160 -76,39,-132
162 '5,25,-76
164 0,113,0
166 -48,119,60
168 -29,5,-104
170 ‘9,4,-78
171 -79,0,33
172 '59,0,12
173 “69,26,29
175 ~77,29,9
177 -74,0,9
178 -8910'3

I 179 '113,0,-9
180 '123,0,-16
181 '126,0,-23
182 -132,0,-40
183 -132,0,-63
184 -135,9,-62
186 -132,29,-48
188 -118,44,-18
190 '69,26,29
192 -59,0,12

193 -18’123134

136

APPENDIX 2
PELVIC COORDINATES
FOR THE JOHNl AND JOHN2 MODELS

For the coordinates, indicates a pelvic point. Each

Pp()

pelvic point is represented by three numbers. The first

number is the X coordinate, the second number is the Y

coordinate, and the third number is the Z coordinate.
ALL COORDINATES ARE IN MM
pp(1)=-25.27625

PP(2)=0

PP(3)=77.59583
PP(4)=-24.36138

pp(5)=0

pp(6)=58.40825
PP(7)=-7.943755

pp(8)=0
pp(9)=45.27579
pp(10)=-3.508106
pp(11)=0
pp(12)=30.94662
pp(13)=33.32959
pp(14)=0
pp(15)=28.95062
pp(l6)=43.79724
pp(17)=0
pp(18)=12.72015
Pp(19)=48.64297
pp(20)=0
pp(21)=-26.60566
pp(22)=37.01218
pp(23)=0
pp(24)=-61.40113
pp(25)=13.93997
pp(26)=0
pp(27)=-69.38067
pp(28)=6.511439
pp(29)=0
pp(30)=-67.70229
pp(31)=-11.97153
pp(32)=0
pp(33)=-50.60318
pp(34i=-32.74279
pp(35)=0
pp(36)=-28.14089
pp(37)=-58.14952
pp(38)=0
pp(39)=-3.554337

pp(40)=-70.85922
pp(4l)=0
pp(42)=-9.163578
pp(43)=-79.94301
pp(44)=0
pp(45)=-28.51519
pp(46)=-63.86612
pp(47)=0
pp(48)=-57.45536
pp(49)=-51.72010
pp(50)=0
pp(51)=-66.25731
pp(52)=-58.25688
pp(53)=0
pp(54)=-70.16506
pp(55)=-88.94478
pp(56)=0
pp(57)=-52.86613
pp(58)=-106.3951
pp(59)=0
pp(60)=-37.75034
pp(61)=-118.4822
pp(62)=0
pp(63)=-20.34627
pp(64)=-127.3893
pp(65)=0
pp(66)=-13.89151
pp(67)=-138.6540
pp(68)=0
pp(69)=38.73079
pp(70)=-131.9763
pp(71)=0
pp(72)=46.24l34
pp(73)=-130.3799
Ppt74l=0
pp(75)=58.66920
pp(76)=-126.0494
pp(77)=0
pp(78)=62.94081
pp(79)=-102.0266
pp(80)=0
pp(81)=73.93635
pp(82)=-63.36956
pp(83)=0
pp(84)=82.97168
pp(85)=-34.32425
pp(86)=0
pp(87)=80.44778

137
APPENDIX 2

(cont'd)

138

APPENDIX 3
COORDINATES OF POINTS
FOR THE 2-D THORAX

 

LANDMARK COORDINATESLJMML
SUPRASTERNALE -140,0,431
MESOSTERNALE -101,0,385
SUBSTERNALE -75,0,336
MID-POINT OF

END OF 10TH RIBS -93,0,133
T1 SPINOUS PROCESS -256,0,457
T2 SPINOUS PROCESS -266,0,435
T3 SPINOUS PROCESS -273,0,411
T4 SPINOUS PROCESS -279,0,385
T5 SPINOUS PROCESS -284,0,355
T6 SPINOUS PROCESS -285,0,329
T7 SPINOUS PROCESS -28l,0,291
T8 SPINOUS PROCESS -273,0,254
T9 SPINOUS PROCESS -265,0,230
T10 SPINOUS PROCESS -255,0,199
T11 SPINOUS PROCESS -244,0,171

T12 SPINOUS PROCESS -231,0,139

139

APPENDIX 4
UMTRI[1] DATA

 

LANDMARK COORDINATES (MM)
MID-POINT OF

H-POINTS 0,0,0
SUPRASTERNALE -140,0,431
MESOSTERNALE -101,0,385
SUBSTERNALE -75,0,336
MID-POINT

OF 10TH RIBS “93,0,133
GLABELLA -87,0,651
MID-POINT OF

INFRAORBITALES -99,0,620
MID-POINT OF

TRAGIONS -185,0,614
L5/Sl INTERSPACE -89,0,39
L2/L3 INTERSPACE -142,0,115
T12/L1 INTERSPACE -177,0,165
T8/T9 INTERSPACE -215,0,287
T4/T5 INTERSPACE -220,0,397
C7/T1 INTERSPACE -193,0,469

SKULL/C1 INTERSPACE

140

APPENDIX 5
LUMBAR COORDINATES
FOR THE JOHNl MODEL

For the coordinates, lp() indicates a lumbar point. Each
interspace is represented by three numbers. The first
number is the X coordinate, the second number is the Y
coordinate, and the third number is the Z coordinate.

ALL COORDINATES ARE IN MM

These are the coordinates for the T12/L1 interspace
lp(1)=-177

lp(2)=0

lp(3)=165

The coordinates for the L1/L2 interspace
lp(4)=-162

lp(5)=0

lp(6)=141

The coordinates for the L2/L3 interspace
#lpt7)=-142

1p(8)=0

lp(9)=115

The coordinates for the L3/L4 interspace
lp(10)=-127

lp(11)=0

lp(12)=92

The coordinates for the L4/L5 interspace
lp(13)=-109

lp(l4)=0

lp(15)=67 .

The coordinates for the L5/Sl interspace
lp(16)=-89

lp(17)=0

lp(18)=39

141
APPENDIX 6

THORACIC COORDINATES

FOR THE JOHNl MODEL
For the coordinates, tp() indicates a torso point. Each
interspace is represented by three numbers. The first
number is the X coordinate, the second number is the Y
coordinate, and the third number is the Z coordinate.
ALL COORDINATES ARE IN MM

These are the coordinates for the C7/T1 interspace
tp(1)=-193

tp(2)=0

tp(3)=469

These are the coordinates for the Tl/T2
tp(4)=-201

tp(5)=0

tp(6)=452

These are the coordinates for the T2/T3
tp(7)=-209

tp(3)=0

tp(9)=435

These are the coordinates for the T3/T4
tp(10)=-214

tp(11)=0

tp(12)=416

These are the coordinates for the T4/T5
tp(13)=-220

tp(14)=0

tp(15)=397

These are the coordinates for the T5/T6
tp(16)=-225

tp(17)=0

tp(18)=371

These are the coordinates for the T6/T7
tp(19)=-224

tp(20)=0

tp(21)=343

These are the coordinates for the T7/T8
tp(22)=-221

tp(23)=0

tp(24)=315

These are the coordinates for the T8/T9
tp(25)=-215

tp(26)=0

tp(27)=287

These are the coordinates for the T9/T10
tp(28)=-209

tp(29)=0

tp(30)=257

142
APPENDIX 6 (cont'd)

These are the coordinates for the T10/T11
tp(31)=-200

tp(32)=0

tp(33)=227

These are the coordinates for the T11/T12
tp(34)=-190

tp(35)=0

tp(36)=195

These are the coordinates for the T12/L1 interspace
tp(37)=-177

tp(38)=0

tp(39)=l65

For the spinous process points, stp() represents a spinous
process point. The first number is the X coordinate, the
second number is the Y coordinate, and the third number is

the Z coordinate.

ALL COORDINATES ARE IN MM

These are the coordinates of T1

stp(l)=-270

stp(2)=0

stp(3)=462

These are the coordinates for T2
stp(4)=—277

stp(5)=0

stp(6)=437

These are the coordinates for T3
stp(7)=-285

stp(8)=0

stp(9)=410

These are the coordinates for T4
stp(10)=-293

stp(11)=0

stp(12)=380

These are the coordinates for T5
stp(13)=-297

stp(14)=0

stp(15)=352

These are the coordinates for T6
stp(16)=-300

stp(17)=0

stp(18)=325

These are the coordinates for T7
stp(19)=-295

stp(20)=0

stp(21)=285

143
APPENDIX 6 (cont’d)

These are the coordinates for T8

stp(22)=-284

stp(23)=0

stp(24)=253

These are the coordinates for T9

stp(25)=-277

stp(26)=0

stp(27)=225

These are the coordinates for T10
stp(28)=-265

stp(29)=0

stp(30)=192

These are the coordinates for T11
stp(31)=-255

stp(32)=0

stp(33)=165

These are the coordinates for T12
stp(34)=-246

stp(35)=0

stp(36)=146

 

144

APPENDIX 7
CERVICAL COORDINATES
FOR THE JOHNl MODEL

For the coordinates, cp()

indicates a gervical point. Each

interspace is represented by three numbers. The first

number is the X coordinate, the second number is the Y

coordinate, and the third

ALL COORDINATES ARE IN MM
These are the coordinates
(occipital condylar axis)
cp(1)=-196

Cp(2)=0

cp(3)=588

These are the coordinates
cp(4)=-196

cp(5)=0

cp(6)=578

These are the coordinates
cp(7)=-190

CP(8)=0

cp(9)=562

These are the coordinates
cp(10)=-187

cp(11)=0

cp(12)=547

These are the coordinates
cp(13)=-184

cp(14)=0

cp(15)=527

These are the coordinates
cp(16)=-183

cp(17)=0

cp(18)=508

These are the coordinates
cp(19)=-186

cp(20)=0

cp(21)=487

These are the coordinates
cp(22)=-193

cp(23)=0

cp(24)=469

number is the Z coordinate.

for the Cl/SKULL interspace

for the C1/C2 interspace

for the C2/C3 interspace

for the C3/C4 interspace

for the C4/C5 interspace

for the C5/C6 interspace

for the C6/C7 interspace

for the C7/T1 interspace

145

APPENDIX 8
LUMBAR COORDINATES
FOR THE JOHN2 MODEL

For the coordinates, lp() indicates a lumbar point. Each
interspace is represented by three numbers. The first
number is the X coordinate, the second number is the Y
coordinate, and the third number is the Z coordinate.

ALL COORDINATES ARE IN MM

These are the coordinates for the T12/L1 interspace
lp(1)=-194.1

lp(2)=0

lp(3)=189.2

The coordinates for the Ll/L2 interspace
lp(4)=-173.7

lp(5)=0

lp(6)=160

The coordinates for the L2/L3 interspace
lp(7)=-152.9

lp(8)=0

lp(9)=130.3

The coordinates for the L3/L4 interspace
lp(10)=-130.8

lp(11)=0

lp(12)=98.7

The coordinates for the L4/L5 interspace
lp(13)=-110

lp(14)=0

lp(15)=69

The coordinates for the L5/Sl interspace
lp(16)=-89

lp(17)=0

lp(18)=39

146

APPENDIX 9

THORACIC COORDINATES
FOR THE JOHN2 MODEL

For the coordinates, tp()

indicates a torso point. Each

interspace is represented by three numbers. The first

number is the X coordinate,

coordinate, and the third

ALL COORDINATES ARE IN MM

These are the
tp(1)=-210
tp(2)=0
tp(3)=493
These are the
tp(4)=-218
tp(5)=0
tp(6)=476
These are the
tp(7)=-226
tp(8)=0
tp(9)=459
These are the
tp(10)=-231
tp(11)=0
tp(12)=440
These are the
tp(13)=-237
tp(14)=0
tp(15)=421
These are the
tp(16)=-242
tp(17)=0
tp(18)=395
These are the
tp(19)=-241
tp(20)=0
tp(21)=367
These are the
tp(22)=-238
tp(23)=0
tp(24)=339
These are the
tp(25)=-232
tp(26)=0
tp(27)=311
These are the
tp(28)=-225
tp(29)=0
tp(30)=281

coordinates

coordinates

coordinates

coordinates

coordinates

coordinates

coordinates

coordinates

coordinates

coordinates

the second number is the Y

number is the Z coordinate.

for the

for the

for the

for the

for the

for the

for the

for the

.for the

for the

C7/T1 interspace

T1/T2

T2/T3

T3/T4

T4/T5

T5/T6

T6/T7

T7/T8

T8/T9

T9/T10

147
APPENDIX 9 (cont’d)

These are the coordinates for the T10/T11
tp(31)=-217

tp(32)=0

tp(33)=251

These are the coordinates for the T11/T12
tp(34)=-207

tp(35)=0

tp(36)=219

These are the coordinates for the T12/L1 interspace
tp(37)=-194.1

tp(38)=0

tp(39)=189.2

For the spinous process points, stp() represents a spinous
process point. The first number is the X coordinate, the
second number is the Y coordinate, and the third number is

the Z coordinate.

ALL COORDINATES ARE IN MM

These are the coordinates for T1
stp(1)=—270

stp(2)=0

stp(3)=462

These are the coordinates for T2
stp(4)=-277

stp(5)=0

stp(6)=437

These are the coordinates for T3
stp(7)=-285

stp(8)=0

stp(9)=410

These are the coordinates for T4
stp(10)=—293

stp(11)=0

stp(12)=380

These are the coordinates for T5
stp(13)=-297

stp(14)=0

stp(15)=352

These are the coordinates for T6
stp(16)=-300

stp(17)=0

stp(18)=325

These are the coordinates for T7
stp(19)=-295

Stp(20)=0

stp(21)=285

These are the
stp(22)=-284
stp(23)=0
stp(24)=253
These are the
stp(25)=-277
StP(26)=0
stp(27)=225
These are the
stp(28)=-265
stp(29)=0
stp(30)=192
These are the
stp(31)=-255
stp(32)=0
stp(33)=165
These are the
stp(34)=-246
stp(35)=0
stp(36)=146

148

APPENDIX 9 (cont’d)

coordinates

coordinates

coordinates

coordinates

coordinates

for

for

for

for

for

T8

T9

T10

T11

T12

149

APPENDIX 10
CERVICAL COORDINATES
FOR THE JOHN2 MODEL

For the coordinates, cp() indicates a cervical point. Each
interspace is represented by three numbers. The first
number is the X coordinate, the second number is the Y
coordinate, and the third number is the Z coordinate.

ALL COORDINATES ARE IN MM

These are the coordinates for the Cl/SKULL interspace
(occipital condylar axis)

cp(1)=-213

cp(2)=0

cp(3)=612

These are the coordinates for the C1/C2 interspace
cp(4)=-213

cp(5)=0

cp(6)=602

These are the coordinates for the C2/C3 interspace
cp(7)=-207

CP(8)=0

cp(9)=586

These are the coordinates for the C3/C4 interspace
cp(10)=-204

cp(11)=0

cp(12)=571

These are the coordinates for the C4/C5 interspace
cp(13)=-201

cp(14)=0

cp(15)=551

These are the coordinates for the C5/C6 interspace
cp(l6)=-200

cp(17)=0

cp(18)=532

These are the coordinates for the C6/C7 interspace
cp(19)=-203

cp(20)=0

cp(21)=511

These are the coordinates for the C7/T1 interspace
cp(22)=-210

cp(23)=0

cp(24)=493

2150
APPENDIX 11
DIFFERENCES (MM) BETWEEN
INTERSPACES FOR THE JOHNl MODEL

10 DEGREES FLEXION

DIEEI

10

7

DISTRIBUTION NUMBER
5 6

4

2

INTER-
CE
T5/T6

XZXZXZXZXZXZXZXZXZXZXZXZXZ
aaaaaaaaaaaaaaaaaaaaaaaaaa
ttttttttttttttttttttttttCt
l111211;ltiltiltiltlitltlltiltl1111fltl1
eeeeeeeeeeeeeeeeeeeeeeeeee
DDDDDDDDDDDDDDDDDDDDDDDDDD
1:514923023013013013no1§8nv7.lQZJnY4aLOIU
1:11412311131141591231i1flaalaa1flnv03un30
................._......
4.84.94.958594948483796643200
1201iOilnvlnvlnu1ZUthIlnvlnvonuORUnZUnYO
........................

24240464930260J66195321100

828282726262422111nwonw0o0.000

37373737075655111753210000

271J1717171728283917642200

1010.1010101010101010000000

 

T6/T7
T7/T8
T8/T9
T9/T10
T10/T11
T11/T12
T12/L1
L1/L2
L2/L3
L3/L4
L4/L5
LS/Sl

151
APPENDIX 12
DIFFERENCES (MM) BETWEEN
INTERSPACES FOR THE JOHNl MODEL

10 DEGREES EXTENSION

DIEEI

10

7

6

D I STRI BUT ION NUMBER
5

A

 

xzxzxzxzxzxzxzxzxzxzxzxzxz
aaaaaaaaaaaaaaaaaaaaaaaaaa
tttttttttttttttttttttttttt
11111111111111111111111111
eeeeeeeeeeeeeeeeeeeeeeeeee
DDDDDDDDDDDDDDDDDDDDDDDDDD
94948595949595955296294300

101011111011.111011110001000

42333343334343335453432100

0000oooooooooooooooooooooo

070—1.0708070707078531863200

21212121212121211111000000

86868787868787879886532100

T6/T7
T7/T8
T10/T11
T11/T12

152

APPENDIX 13
DIFFERENCES (MM) BETWEEN
INTERSPACES FOR THE JOHNl MODEL
20 DEGREES EXTENSION

 

INTER- DISTRIBUTION NUMBER
SPACE 2 3 4 5 6 7, fl 9 10 DIEEI
TS/T6 -2.0 1.6 3.7 0.7 2.1 -8.4 -13.3 2.4 5.2 Delta
-1.8 1.4 3.8 0.6 2.0 -7.4 -10.9 2.3 5.5 Delta
T6/T7 -2.0 1.5 3.7 0.7 2.1 -8.4 -13.3 2.4 5.2 Delta
-1.8 1.4 3.7 0.5 2.0 -7.4 -1l.0 2.3 5.5 Delta
T7/T8 -2.0 1.5 3.7 0.7 2.1 -8.4 ~13.0 2.4 5.2 Delta
-1.7 1.5 3.8 0.6 2.1 -7.4 -10.7 2.4 5.5 Delta
T8/T9 -2.0 1.5 3.6 0.7 2.1 -8.4 ~12.3 2.3 5.2 Delta
-1.8 1.5 3.8 0.6 2.0 -7.4 -10.3 2.4 5.5 Delta
T9/T10 -2.0 1.5 3.6 0.7 2.1 -8.0 -11.1 2.3 5.2 Delta
-1.7 1.5 3.8 0.6 2.1 -7.1 -9.6 2.4 5.5 Delta
T10/T11 -2.0 1.5 3.6 0.7 2.1 -7.1 -9.6 2.3 5.2 Delta
-l.7 1.5 3.8 0.6 2.1 -6.4 -8.5 2.4 5.6 Delta
T11/T12 -2.0 1.6 3.7 0.7 2.2 -5.4 -7.3 2.4 4.7 Delta
-1.8 1.5 3.8 0.6 2.0 -5.3 -7.0 2.4 5.5 Delta
T12/L1 -2.0 1.5 3.6 0.7 2.1 -3.2 -4.5 2.3 5.2 Delta
-l.7 1.5 3.8 0.6 2.1 -3.3 -4.5 2.4 5.5 Delta
Ll/LZ -2.3 1.8 3.3 1.0 2.4 -1.8 -2.7 2.1 4.7 Delta
-2.0 1.7 3.4 0.9 2.4 -1.9 -2.7 2.1 4.9 Delta
L2/L3 -1.9 1.6 2.6 1.0 2.2 -0.8 -1.4 1.7 3.6 Delta
-1.7 1.4 2.4 0.8 2.0 -0.7 -1.2 1.5 3.5 Delta
L3/L4 -1.1 0.8 1.4 0.7 1.3 -0.3 -0.6 1.1 2.2 Delta
-0.9 0.7 1.2 0.6 1.2 -0.3 -0.6 1.0 2.0 Delta
L4/L5 -0.4 0.3 0.5 0.3 0.4 0.0 -0.2 0.5 0.7 Delta
-0.3 0.3 0.4 0.3 0.4 0.0 -0.l 0.4 0.6 Delta
LS/Sl 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta

N94NXNXNXNXNXNXNXNXNXNXNXNX

153
APPENDIX 14

DIFFERENCES

(MM) BETWEEN
INTERSPACES FOR THE JOHNl MODEL

30 DEGREE EXTENSION

 

INTER- DISTRIBUTION NUMBER
SPACE 3 4 5 6 7 8 9 0 DIEEI
T5/T6 -2.9 2.3 4.9 1.0 3.9 -11.0 -17.7 3.2 6.9 Delta X
-2.8 2.4 6.2 0.9 3.3 -12.7 -18.9 3.8 9.0 Delta Z
T6/T7 -2.8 2.2 4.9 1.1 3.0 -11.0 -17.7 3.2 7.0 Delta X
-2.8 2.3 6.1 0.9 3.2 -12.7 -19.0 3.8 8.9 Delta 2
T7/T8 -2.8 2.2 4.9 1.1 3.0 -11.0 -17.2 3.2 7.0 Delta X
-2.8 2.4 6.2 0.9 3.3 -12.7 -18.6 3.8 9.0 Delta 2
T8/T9 -2.8 2.2 4.9 1.1 3.0 -11.0 -16.3 3.2 7.0 Delta X
-2.8 2.3 6.1 0.9 3.2 -12.7 -17.8 3.8 8.9 Delta 2
T9/T10 -2.8 2.2 4.9 1.1 3.0 -10.4 -14.7 3.2 7.0 Delta X
-2.8 2.4 6.2 1.0 3.3 -12.2 ~16.5 3.9 9.0 Delta 2
T10/T11 -2.8 2.2 4.9 1.1 3.0 -9.3 -12.7 3.2 7.0 Delta X
-2.8 2.4 6.2 0.9 3.3 ~11.0 -14.5 3.8 9.0 Delta 2
T11/T12 ~2.9 2.1 4.9 1.0 2.9 -7.2 -9.7 3.2 6.9 Delta X
-2.8 2.4 6.2 1.0 3.3 -8.9 -1l.6 3.9 9.0 Delta 2
T12/L1 -2.8 2.1 4.9 1.0 3.0 -4.2 “5.9 3.2 7.0 Delta X
-2.8 2.4 6.2 0.9 3.3 -5.5 -7.4 3.8 9.0 Delta 2
L1/L2 -3.2 2.4 4.4 1.3 3.3 -2.5 '3.8 2.9 6.3 Delta X
-3.3 2.8 5.5 1.4 3.9 '3.1 -4.4 3.4 8.0 Delta Z
L2/L3 -2.8 2.1 3.5 1.3 2.9 -1.2 -2.0 2.3 4.9 Delta X
-2.7 2.3 3.8 1.3 3.2 -1.2 -2.0 2.4 5.6 Delta 2
L3/L4 -1.5 1.2 2.0 1.0 1.9 -0.4 -0.9 1.6 3.1 Delta X
-1.4 1.2 2.0 1.0 2.0 -0.4 -0.8 1.6 3.3 Delta 2
L4/L5 -0.6 0.4 0.7 0.5 0.6 -0.1 -0.3 0.7 1.0 Delta X
-0.3 0.4 0.7 0.5 0.6 0.0 -0.2 0.7 1.0 Delta Z
LS/Sl 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta X
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta 2

154

L
E
m
N
mM
wm
TH
EO
:oB.u
1
)E
XMH
IWT
m R
Qv
E O
PwF
PN
A S
WE
EC
A
FP
FS
IR
DE
T
N
I

10 DEGREES FLEXION

DIEEI

10

6

DISTRIBUTION NUMBER
5

4

:7

 

INTER-
SPACE

xzxzxzxzxzxzxzxzxzxzxzxzxz
aaaaaaaaaaaaaaaaaaaaaaaaaa
tttttttttttttttttttttttttt
11111111111111111111111111
eeeeeeeeeeeeeeeeeeeeeeeeee
DDDDDDDDDDDDDDDDDDDDDDDDDD
03020212130203036275584.300
4.2424242424242423121100000
-.......................
70707080807070?05927843200
11111111111111111010000000
....-.-..-......_..._...

87886818576624490217431100

82828202726252312110000000

73707070499997434874220000

.......

48494930~40~ASAQ~JQH4H04965210.0.

10101010.10.1010.10.1110000.0.0.0.

1 2
0 l 1 1
.b .I .8 9 l T T L 2 1. as .3 1
T 1. mi T T l. I, I: .L L in .L S
.l .l ./ /..l 0 1 2 I. l. I: .l ./
.5 .b 7 8 9 1 1 1 1 9. 1. as .3
T T T T T T T T L L L. vu .L

155
APPENDIX 16

DIFFERENCES
INTERSPACES FOR THE JOHN2 MODEL

(MM) BETWEEN

10 DEGREES EXTENSION

10 DIEEL

9

I,

6

DISTRIBUTION NUMBER
5

4

 

xzxzxzxzxzxzxzxzxzxzxzxzxz
3888388388aaaaaaaaaaaaaala
tttttttttttttttttttttttttt
11;11211;]:LIT11f11II1L11;11;11I11I11II
eeeeeeeeeeeeeeeeeeeeeeeeee
DDDDDDDDDDDDDDDDDDDDDDDDDD
57505059605050402640304300

10.11111111111111111000.0.0.0.0.

1011111111.1111111111000000

4.0434352535353436465432200

00000000000000000000000000

A8514.159515151412874862200

21.222221222222222111000000

osoaoaoofiloe1.908082909541200

T10/T11
T11/T12

156
APPENDIX 17
DIFFERENCES (MM) BETWEEN
INTERSPACES FOR THE JOHN2 MODEL

20 DEGREES EXTENSION

10 DIFEL

9

7

6

DISTRIBUTION NUMBER
5

4

XZXZXZXZXZXZXZXZXZXZXZXZXZ
aaaaaaaaaaaaaaaaaaaaaaaaaa
tttttttttttttttttttttttttt
11111111111111111111111111
eeeeeeeeeeeeeeeeeeeeeeeeee
DDDDDDDDDDDDDDDDDDDDDDDDDD
373737474747383J7°~J4538600

666666666666666655L4220000

a889899888838988550.°n316‘.0.0.

O O O I O O O O O O O C O

2222222222222L222222110000.

86878797979788861000874300

0.0.0000000000000.01111000000

46474757565757460110645400

L44444444L44.4-4444433110000

989898J89898390~81198083200

11111121111111112211100000

 

INTER-
SEASE

52515112524151528652204400
22222232222222222222110000
. . . . . . . _ . . . . . . . . . . . . . . . .
1. 2
0 1 1 1

5 7 8 9 1 T T L 2 3 4 5 1
T T T T T / / / L L L L S
/ / / / / O I 2 / / / / /
5 6 7 8 9 1 1 1 1 2 3 4 5
T T T T T T T T L L L L L

157

APPENDIX 18
DIFFERENCES (MM) BETWEEN
INTERSPACES FOR THE JOHN2 MODEL
30 DEGREES EXTENSION

 

INTER- DISTRIBUTION NUMBER
azaga 2 3 4 5, 5 1 a 9 lo DIEEI
15/16 -3.5 2.7 6.0 1.3 3.7 -11.8 -18.8 3.8 8.5 Delta
-3.5 2.9 7.6 1.0 3.9 -13.9 -20.6 4.6 10.9 Delta
T6/T7 -3.6 2.6 5.9 1.2 3.6 -ll.9 -18.9 3.8 8.4 Delta
-3.4 2.9 7.6 1.1 4.0 -l3.9 -20.6 4.6 10.9 Delta
T7/T8 -3.6 2.6 5.9 1.2 3.6 -11.9 -18.5 3.8 8.4 Delta
-3.5 2.9 7.6 1.0 3.9 -13.9 -20.3 4.6 10.9 Delta
T8/T9 -3.6 2.6 5.9 1.3 3.6 -ll.9 -17.5 3.8 8.4 Delta
-3.4 2.9 7.6 1.1 4.0 -13.9 -19.4 4.6 10.9 Delta
T9/T10 -3.5 2.7 6.0 1.2 3.6 -ll.3 -16.0 3.8 8.4 Delta
-3.4 2.9 7.6 1.1 4.0 -13.4 -18.1 4.6 10.9 Delta
110/111 -3.6 2.6 5.9 1.2 3.6 -10.2 -13.9 3.8 8.4 Delta
-3.4 2.9 7.6 1.0 3.9 -12.3 -l6.2 4.6 10.9 Delta
T11/T12 -3.5 2.7 6.0 1.3 3.7 -8.0 -10.9 3.8 8.4 Delta
-3.4 2.9 7.6 1.1 4.0 -10 2 -13.3 4.6 10.9 Delta
T12/L1 -3.5 2.7 6.0 1.3 3.7 -5.0 -7.1 3.9 8.5 Delta
-3.5 2.9 7.5 1.0 3.9 -6.0 -9.2 4.5 10.9 Delta
Ll/L2 -3.9 3.0 5.5 1.6 4.1 -3.0 -4.6 3.5 7.7 Delta
-4.1 3.4 6.7 1.7 4.7 -3.6 -5.3 4.0 9.6 Delta
L2/L3 -3.4 2.6 4.3 1.5 3.6 -l.5 -2.5 2.8 6.1 Delta
-3.5 2.9 4.9 1.5 4.0 -l.7 -2.7 3.0 7.1 Delta
L3/L4 -l.8 1.3 2.2 1.1 2.2 -0.5 -l.1 1.8 3.5 Delta
-l.6 1.3 2.2 1.1 2.2 -0.5 -l.0 1.8 3.7 Delta
L4/LS -0.5 0.5 0.8 0.6 0.7 0.0 -0.2 0.8 1.2 Delta
-0.5 0.5 0.7 0.5 0.6 0.0 -0.2 0.7 1.0 Delta
L5/Sl 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Delta

NXNXNXNXNXNXNXNXNXNXNXNXNX

BIBLIOGRAPHY

10.

11.

BIBLIOGRAPHY

Robbins, D.H., ’Anthropometric specifications for Mid-
Sized Male Dummy,’ QM131;81;§;;2, U.S. Department of
Transportation, National Highway Traffic Safety
Administration, Washington, D.C., 1983.

Synder, R.G., Chaffin, D.B., Schutz, ’Link System of
the Human Torso,’ AMBL;IR;1;;8§, Aerospace Medical
Research Lab., Aerospace Medical Division, Air Force
Systems command, Wright-Patterson AFB, Ohio, 1970.

Andersson, G.B.J., Murphy, R.W., Ortengren, R.,
Nachemson, A.L., ’The Influence of Backrest Inclination
and Lumbar Support on Lumbar Lordosis,’ Spine, Vol. 4,
No 1, pp. 52-58, 1979.

Robbins, D.B., Reynolds, H.M., ’Position and Mobility
of Skeletal Landmarks of the 50th Percentile Male in an
Automotive Seating Posture,’ QM'ESBI'BI-7A'4: Vehicle
Research Institute, Society of Automotive Engineers,
Penn., 1975.

Nyquist, G.W., Patrick, L.M., ’Lumbar and Pelvic
Orientations of the Vehicle Seated Volunteer,’

, Report Num.
760821, pp. 665-696, 1976.

Panjabi, M.M., White III, A.A. Clinical Bigmgghanigfi Qf
Lne_§pine, J.B. Lippincott, Philadelphia, PA. 1978.

Allbrook, D., ’Movements of the Lumbar Spinal Column,’
' , Vol. 398, No 2,
pp 339-345, 1957.

Reynolds, H.M., Snow, C.C., Young, J.W., ’Spatial
Geometry of the Human Pelvis,’ Federal Aviation
Administration, Memorandum Report No. AAQ;1L&;§1;§,
1981.

Hubbard, R.P., McLeod, D.G., ’A Basis for Crash Dummy
Skull and Head Geometry,’ General Motors Corporation
Research Laboratories, Research Publication No. 9M3;

1281, 1972.

SDRC IDEAS 3.8 GEOMOD Solid Modeling Package,
Structural Dynamics Research Corporation, 1986.

Khalil, T., Hubbard, R., ’Parametric Study of Head

Response be Finite Element Modeling,’The Journal of
Biomechanics, 10(2), pp. 119-132, 1977.

158

159

Jones, A., Pollock, M., Graves, J., Fulton, M., Jones,
W.E., MacMillan, M., Baldwin, D.D., Cirulli,J., ’Safe,
' ' T 'n n h ili 'v Ex r ' f h

Muscles of The Lumbar Spine,’ Sequoia Communications,

CA., 1988.

71111111111181“31111155