‘r.

 

‘24“

5
.
2
(I

 

This is to certify that the
thesis entitled

ESTIMATIONS OF HUMAN HIP JOINT CENTER
LOCATIONS IN AUTOMOTIVE SEATS: A
COMPARISON BETWEEN EXPERIMENTAL DATA AND
A PREDICTION MODEL

presented by
VAIBHAV AEKBOTE

has been accepted towards fulfillment
of the requirements for the

MS. degree in Mechanical Engineering

KWQEEM

Major Professor’s Signature
(1 "Z 3 L

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

a—o—o-o—u-.--.-o—.—o-o-u-o-o---o-u--I-o-o-n-n-n-n_n-n—n-.—, -o‘nn.-- - -.

Q 0-... § . .

9096

00-0.

.'."‘”---OO"

O

 

LIBRARY
Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

v”... h‘Lt-c

 

M242008
809

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuepBS-p. 1 5

ESTIMATIONS OF HUMAN HIP JOINT CENTER LOCATIONS IN
AUTOMOTIVE SEATS: A COMPARISON BETWEEN
EXPERIMENTAL DATA AND A PREDICTION MODEL

By

Vaibhav A. Ekbote

A THESIS

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

MASTER OF SCIENCE
Department of Mechanical Engineering

2004

ABSTRACT

ESTIMATIONS OF HUMAN HIP JOINT CENTER LOCATIONS IN
AUTOMOTIVE SEATS: A COMPARISON BETWEEN
EXPERIMENTAL DATA AND A PREDICTION MODEL

By

Vaibhav A. Ekbote

The present study compares experimental data to a prediction model for human hip
joint center (HJC) locations in automotive seats. In phase I of the three phase study, seat
pan stiffnesses for six different seats were tested using an industry standard manikin and
video based motion measurement techniques. For phase II of the study, three of the six
seats tested in phase I were selected to represent a range of seat pan stiffnesses varying
from soft to stiff. Fifteen male subjects from three anthropometric categories were seated
in all three seats and their positional data were recorded. These data were then used to
compute the location their HJC locations. In phase III, a prediction of the HJC locations
for all test cases in the same three seats was obtained using mathematical modeling
techniques developed in previous studies by Radcliffe [6], and Bush and Macklem [3].
Finally the results between the predicted and measured HJC locations were compared.
For this comparison two different approaches were used for the computation of the H] C
producing slightly different locations. The prediction model successfully predicted the
HJC deflections of two of the three subject groups. For the third group a prediction curve

was not necessary as it was located using the industry standard manikin.

ACKNOWLEDGEMENTS

I would like to thank Dr. Tamara Reid-Bush for her erudite guidance during last two
years. Your positive supervision, mentoring, and support have guided me in the right
direction throughout my master’s education at Michigan State University. It was only
your continuous suggestions on my writing that made it possible for me to complete the

seemingly trying task of wording my thesis.

Thank you to my master’s committee members. Thank you, Dr. Hubbard for your
continuous advice and encouragement and showing the significance of focusing on goals.
Thank you, Dr. Liu for helping me settle down and become comfortable in the initial

months of my education at Michigan State University.

I would also like to thank my family members for encouraging me to take up the graduate

education. Mammi, lPappa, and Bhushan, I thank you for your patience and support.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vi
LIST OF FIGURES ................................................................................. ix
NOMENCLATURE .............................................................................. xiv
1. INTRODUCTION ............................................................................... 1
2. BACKGROUND ............................................................................... 6
3. PHASE I .......................................................................................... 8
3.1 Methods used in Phase I ................................................................... 8
3.2 ASPECT Butt-Thigh (ABT) loading steps ............................................. 11
3.3 Coupled Force and Moment kinetic model ............................................. 19
3.4 Seats tested for Phase I. .................................................................. 24
3.5 Seat selection for Phase II. ............................................................... 31
4. PHASE II. ....................................................................................... 36
4.1 Test subjects. ............................................................................... 36
4.2 Test Buck setup. ........................................................................... 38
4.3 Reference seat. ............................................................................. 41
4.4 Test Protocol. .............................................................................. 42
4.5 Testing Procedure. ........................................................................ 44
5. PHASE IH. ...................................................................................... 49
5.1 Background for Phase III. ................................................................ 49
5.2 Calculation of HJC in test seats. ......................................................... 53

iv

5.3 Method to calculate the HJC deflection. ................................................ 55

5.4 Results of comparison between the HJC deflections computed experimentally and

those predicted using the kinetic model. ................................................ 59

5.4.1 Comparison of HJC deflections for 50H50W male subjects .................. 64

5.4.2 Comparison of HJC deflections for 95H95W male subjects .................. 66

5.4.3 Comparison of HJC deflections for 95H5W male subjects. .................. 68

5.5 Comparison with Bush-Macklem offset curves ....................................... 7O

6. CONCLUSION. ................................................................................ 76
6.1 Future Work. ........................................................................... 78
APPENDIX-A SFS Analysis- Phase I. .......................................................... 79

APPENDIX-B Manual measurements of H-point vertical and horizontal deflection. . ....86
APPENDIX-C Experimental data Phase III .................................................... 91

REFERENCES. ................................................................................... lOl

LIST OF TABLES

Table 1. Loads at the H-point and front of ABT for all load steps. ......................... 18
Table 2. Example of input data for the seat factor solver spreadsheet. ..................... 22
Table 3. List showing available information about six seats tested in phase 1. ........... 24
Table 4. Desired human subjects’ anthropometrics as per NTIS. ........................... 37
Table 5. Actual test subjects’ anthropometric measurements ................................. 37
Table 6. Dimension descriptions for test buck set-up. ........................................ 40

Table 7. Package dimensions with J 826 manikin for the typical car segment-seating

environment. .............................................................................. 40
Table 8. Target locations for seat testing. ....................................................... 46
Table 9. Target locations for hard seat trials .................................................... 47

Table 10. Loading under the HJC for various anthropometrics as studied by Bush. ......50

Table 11. HJC deflection comparison between preferred and instructed position for
SOHSOW male subjects in SLK seat. .................................................. 65

Table 12. HJC deflection comparison between preferred and instructed position for
95H95W male subjects in Tahoe seat. ................................................ 67

Table 13. H] C deflection comparison between preferred and instructed position for
95H5W male subjects in Tahoe seat. .................................................. 69

Table A-1. SFS analysis to obtain H-point force Vs. H-Point deflection for seat A ....... 80
Table A-2. SFS analysis to obtain H—point force Vs. H-Point deflection for seat B ....... 81
Table A-3. SFS analysis to obtain H—point force Vs. H-Point deflection for seat C. ..... 82
Table A- 4. SFS analysis to obtain H-point force Vs. H-Point deflection for seat D. 83
Table A-5. SFS analysis to obtain H—point force Vs. H-Point deflection for seat E ....... 84
Table A-6. SFS analysis to obtain H—point force Vs. H-Point deflection for seat F._ ....... 85
Table B-1. Seat A-Audi, manual measurements of H-point vertical deflection. .......... 87

Table B-2. Seat B-Ranger, manual measurements of H-point vertical deflection. ........ 87

vi

Table B—3. Seat C-SLK, manual measurements of H-point vertical deflection. ........... 88

Table B—4. Seat D-Tahoe (Cloth), manual measurements of H—point vertical

deflection .................................................................................. 88
Table B-5. Seat A-Audi, manual measurements of H-point horizontal deflection ......... 89
Table B-6. Seat C-SLK, manual measurements of H-point horizontal deflection ......... 89

Table B-7. Seat D-Tahoe (Cloth), manual measurements of H-point horizontal
Deflection. ................................................................................ 90

Table B-8. Seat E-Tahoe (Leather), manual measurements of H-point horizontal
Deflection ................................................................................. 90

Table C-l. HJC experimental data for SOHSOW category in BMW seat with HJC
calculated using method used by Gutowski. ......................................... 92

Table C-2. HJC experimental data for 50H50W category in BMW seat with H] C
calculated using Bush- Gutowski method. ........................................... 92

Table C-3. HJC experimental data for SOHSOW category in SLK seat with H] C
calculated using method used by Gutowski. ......................................... 93

Table C-4. HJC experimental data for SOHSOW category in SLK seat with HJC
calculated using Bush- Gutowski method. ........................................... 93

Table C-5. HJC experimental data for SOHSOW category in Tahoe seat with HJC
calculated using method used by Gutowski. ......................................... 94

Table C-6. HJC experimental data for SOHSOW category in Tahoe seat with HJ C
calculated using Bush- Gutowski method. ........................................... 94

Table C-7. HJC experimental data for 95H5W category in BMW seat with HJC
calculated using method used by Gutowski. ......................................... 95

Table C-8. HJC experimental data for 95H5W category in BMW seat with HJC
calculated using Bush- Gutowski method. ........................................... 95

Table C-9. HJC experimental data for 95H5W category in SLK seat with HJC
calculated using method used by Gutowski. ......................................... 96

Table C-lO. HJC experimental data for 95H5W category in SLK seat with HJC
calculated using Bush- Gutowski method. ........................................... 96

Table C-1 1. HJC experimental data for 95H5W category in Tahoe seat with HJC
calculated using method used by Gutowski. ......................................... 97

vii

Table C-12. HJC experimental data for 95H5W category in Tahoe seat with HJC
calculated using Bush- Gutowski method. ........................................... 97

Table C-l3. HJC experimental data for 95H95W category in BMW seat with HJC
calculated using method used by Gutowski. ........................................ 98

Table C-14. HJC experimental data for 95H95W category in BMW seat with HJC
calculated using Bush- Gutowski method. ........................................... 98

Table 0-15. HJC experimental data for 95H95W category in SLK seat with HJC
calculated using method used by Gutowski. ......................................... 99

Table C-l6. HJC experimental data for 95H95W category in SLK seat with HJC
calculated using Bush- Gutowski method. ........................................... 99

Table C-l7. HJC experimental data for 95H95W category in Tahoe seat with HJC
calculated using method used by Gutowski. ....................................... 100

Table C-18. HJC experimental data for 95H95W category in Tahoe seat with HJC
calculated using Bush- Gutowski method. .......................................... 100

viii

LIST OF FIGURES

Figure 1: Human Hip Joint Center ................................................................. 2
Figure 2: Representative picture of subject testing in Phase II. ............................... 4

Figure 3: Example of kinetic model by Radcliffe [6] that simulated the seat pan
stiffness response of one of the test seats (Audi Leather) for incremental
loading of ASPECT butt and thigh region. ........................................... 6

Figure 4: Example of buttock and knee loading of ASPECT manikin. ...................... 7

Figure 5: Comparison Between manually measured data and data collected from
Qualisys motion measurement system for Tahoe (Cloth) seat. .................... 9

Figure 6: SAE J 826, 2-D template positioned in Tahoe Leather seat for the
load step zero with a total weight of 0 N. ........................................... 11

Figure 7: (Load step 1) Aspect Butt Thigh (ABT) segment with no weights
added with a total weight 57 N. ....................................................... 12

Figure 8: (Load step 2) ABT segment with two 6-plate H-point weights on H-point axis
with a total weight 125 N. Front and isometric views. ........................... 12

Figure 9: (Load step 3) Load step 2 plusZ five plate H-point weights on H-point axis
with a total weight 185 N. Front and isometric views. ........................... 13

Figure 10: (Load step 4) Load step 3 plu52 five plate H-point weights on H-point axis
with a total weight 231 N. Front and isometric views. ........................... 13

Figure 11: (Load step 5) Load step 4 plusZ torso weights on H-point axis
with a total weight 278 N. Front and isometric views. ........................... 14

Figure 12: (Load step 6) Load step 5 plusZ torso weights on H-point axis
just inside the shell of ABT with slots down and back rear edge resting against
shell with a total weight 310 N. Front and isometric views. ....................... 14

Figure 13: (Load step 7) Load step 6 plus2 torso weights behind the center structure with
slots forward with a total weight 348 N. Front and isometric views. ........... 15

Figure 14: (Load step 8) Load step 7 plusZ torso weights behind the center structure with
slots forward with a total weight 388 N. Front and isometric views. ........... 15

Figure 15: (Load step 9) Load step 8 plusZ torso weights on center structure with slots up

and back lower edge against nylon bushing with a total weight 416 N. Front and
isometric views. ......................................................................... l6

ix

 

 

 

 

F1

Figure 16: (Load step 10) Load step 9 plusZ thigh weights with a total weight 431 N.
Front and isometric views ............................................................... 16

Figure 17: (Load step 11) Load step 10 plu52 thigh weights with a total weight 446 N.
Front and isometric views ............................................................... 17

Figure 18: (Load step 12) Load step 11 plus2 thigh weights with a total weight 461 N.

Front and isometric views. ............................................................. 17
Figure 19: Coupled Force and Moment Kinetic model. ...................................... 20
Figure 20: Seat A, the Audi front and side view ................................................ 25
Figure 21: Seat B, the Ranger front and side view. ............................................ 26
Figure 22: Seat C, the SLK front and side view. ............................................... 27
Figure 23: Seat D, the Tahoe-cloth front and side view. ...................................... 28
Figure 24: Seat E, the Tahoe-leather front and side view. .................................... 29
Figure 25: Seat F, the BMW - front and side view. ............................................ 30

Figure 26: Chart comparing H-point deflection Vs. H-point load for seats tested in phase
I (shown in bold legends). Also shown are H-point deflections with increasing
H—point load for 22 seats tested by Radcliffe . ....................................... 32

Figure 27: Chart comparing H-point deflection versus H-point load for six seats tested in
phase I. The deflections corresponding to a H-point load of 4lON were
compared. ................................................................................. 33

Figure 28: H-point Force Vs H-point deflection for Tahoe (SUV) seat from

Radcliff’ s 2“d order kinetic model. .................................................... 34
Figure 29: H-point Force Vs H—point deflection for BMW (SUV) seat from

Radcliff’ s 2nd order kinetic model. .................................................... 35
Figure 30: H-point Force Vs H-point deflection for SLK (SUV) seat from

Radcliff’s 2“‘1 order kinetic model. .................................................... 35
Figure 31: Test buck dimensions. ................................................................ 39
Figure 32: Reference hard seat dimensions. .................................................... 41
Figure 33: Subject 10 seated in the instructed position in SLK seat ........................ 43

Figure 34: Subject lO seated in the preferred position in SLK seat.
Subject chose more reclined position than the instructed position and
a preferred position of his arms. ....................................................... 43

Figure 35: Pelvic width and pelvic height ....................................................... 45
Figure 36: Target locations on seat and subject for testing in three production seats. ....46
Figure 37: Target locations for testing in hard seat and targeted subject. .................. 48

Figure 38: Averages of HJC forces Vs. Deflections with error bars for various
anthropometries in Town and Country seat obtained by Bush and Macklem
using the data from Gutowski’s study. Each point represents the
averaged HJC deflection of five subjects. A notable difference in average
deflections compared to those predicted by the kinetic model can be
observed for anthropometrics other than SOHSOW males. ........................ 51

Figure 39: Best fit line for SOHSO W and 95H5W male occupants developed by Bush
and Macklem [3]. Each point represents the averaged HJC deflection of five
subjects. HJC deflection for 95H95W estimated to be on the linearly
extrapolated HJC force deflection curve. ............................................. 52

Figure 40: Best fit parabola for female occupants developed by

Bush and Macklem. ..................................................................... 52
Figure 41: Calculation of 6, the vertical deflection of the buttocks ........................ 55
Figure 42: Computation of Seat Deflection from human data. .............................. 57

Figure 43: Un-deflected seat contour scan obtained from Qualisys system used to
calculate 52 . ............................................................................... 58

Figure 44: HJC deflections averaged for all SOHSOW subjects in BMW seat
were found to be higher than that predicted by the kinetic model and was
below the force deflection curve. ..................................................... 59

Figure 45: HJC deflections averaged for all SOHSOW subjects in SLK seat
were close in comparison to the force deflection curve of kinetic model. ...... 60

Figure 46: HJC deflections averaged for all SOHSOW subjects in Tahoe seat
were close in comparison to the force deflection curve of kinetic model. ...... 60

xi

Figure 47: Pressure distribution on BMW seat pan due to ABT loading
of 461N (refer section 3.2). A considerable amount of pressure is distributed
on the bolsters. The BMW seat with prominent seat pan bolsters is
seen in the right. ......................................................................... 62

Figure 48: Pressure distribution on Tahoe seat pan due to ABT loading
of 461N (refer section 3.2). Amount of pressure distributed on the
bolsters is less compared to that in BMW (Figure 47).
Tahoe seat is seen on the right. ........................................................ 62

Figure 49: Pressure distribution on SLK seat pan due to ABT loading
of 461N (refer section 3.2). Amount of pressure distributed on the
bolsters is less compared to that in BMW (Figure 47).
SLK seat is seen on the right. .......................................................... 63

Figure 50: HJC Force vs. Deflection for SOHSOW male subjects in SLK seat. ............ 64

Figure 51: A magnified view of HJC locations in preferred and instructed positions.
HJC in preferred position had a trend of being anterior and distal
(forward and down) with respect to HJC locations in instructed position. ......66

Figure 52: HJC force Vs. Deflection in preferred and instructed position
for 95H95W male subjects in Tahoe seat ............................................. 66

Figure 53: HJC force Vs. Deflection in preferred and instructed position
for 95H5W male subjects in Tahoe seat. ............................................. 68

Figure 54: A plot showing offset line for SLK seat along with the extrapolated force
deflection curve and averaged HJC deflections for each category with
error bars of i 1 standard deviation. ................................................. 71

Figure 55: A plot showing offset line for Tahoe seat along with the extrapolated force
deflection curve and averaged HJC deflections for each category with
error bars of + 1 standard deviation. ................................................. 72

Figure 56: A plot showing offset line for BMW seat along with the extrapolated force
deflection curve and averaged HJC deflections for each category with
error bars. ................................................................................ 73

Figure 57: Comparison of HJC computations based on Bush-Gutowski method

(legends in hollow) and method used in Gutowski’s study (legends in solid)
for Tahoe seat ............................................................................ 74

xii

Figure 58: Comparison of HJC computations based on Bush-Gutowski method
(legends in hollow) and method used in Gutowski’s study (legends in solid)
for SLK seat ............................................................................... 75

Figure 59: Comparison of HJC computations based on Bush-Gutowski method
(legends in hollow) and method used in Gutowski’s study (legends in solid)
for BMW seat .............................................................................. 75

xiii

NOMENCLATURE

ABT ....................................................... ASPECT manikin, butt-thigh segment
ASPECT .................. Automotive Seat and Package Evaluation and Comparison Tools
HJ C ......................................................................... Human Hip-Joint Center
MSU ..................................................................... Michigan State University
SFS ................................................................. Seat Factor Solver Spreadsheet.

Arabic Symbols and Acronyms

F ................................. generic reaction force under the H-point (could be RH or F")
Fn ...................................................... simulation reaction force under the H-pt.
FH ............................................................... simulation input force at the H-pt.
FK ............................................................... simulation input force at the knee
FT ............................................................... simulation input force at the thigh
n K [5 .................. nth order reaction force-deflection stiffness coefficient for simulation
n K129 .................... nth order reaction force-rotation stiffness coefficient for simulation

n K m a ............ rr‘h order reaction moment-deflection stiffness coefficient for simulation

n Km, ............... nth order reaction moment-rotation stiffness coefficient for simulation
KH ...................................................... spring constant of the seat under the H-pt.

xiv

KK ..................... spring constant of the seat under the knee (physically non-existent)

RH ............................................. experimental reaction force under the H-point
Greek Symbols

6" ................................................... relative deflection of the H-pt. (same as AH)
19" .............................. relative rotation of the thigh (referenced to the 2-D template)
AH ................................. relative deflection of the H-pt. (referenced to 2-D template)
AK ................................. relative deflection of the knee (referenced to 2-D template)
9mm” ....................................................... thigh angle referenced to horizontal

XV

1. INTRODUCTION

One of the most important aspects of designing a vehicle’s interior package is
locating the appropriate placement of the seat and the occupant within the vehicle. One
identification of occupant placement is a point at or near the hip joint center (HJC). The
HJC acts as a reference point in designing the car seat and interior packaging because it is
the point on human body with least motion with respect to the car seat during the time
interval of an occupant in a car. So placing the occupant's HJC at an intended position in
seat is of fundamental importance to meet mandated safety regulations and design and
packaging requirements.

Several tools are used to locate or estimate the HJC including Society of Automotive
Engineers (SAE) manikins. There are two versions of the SAE manikins, the first is
nicknamed OSCAR [l] and the newer one is called ASPECT [2]. Both are
representations of average sized male occupants. The point corresponding to human H] C
on the manikin is called the H-point and is a representation of Hip Joint Center of an
average human occupant (Figure 1).

The focus of this study is to locate the hip joint centers of seated human occupants
using experimental techniques and then to compare the measured locations to predicted
locations based on the ASPECT manikin and previously developed mathematical
techniques [3].

Included as part of the vehicle's Computer Aided Design (CAD) data, the H-point
must fall within an envelope in the space that accounts for design variables like cushion
deflection and the seat's for- aft and vertical range of motion. The seat designer's goal is

to locate the H-point in the most advantageous position within this envelope. A smaller

statured human would have the HJC situated at the envelope's forward end, while the

H] C of a larger statured person would be at the rearward part of the envelope. The H-
point enables the designer to position the human model within the CAD data of the
automobile interior and establishes locations of hand and foot controls, overall packaging

and vision requirements.

Right
HJ C \

 

Figure 1: Human Hip Joint Center [8]

Human modeling software such as RAMSIS [4] and JACK [5] simulate various
anthropometric sizes, shapes and body positions and movements in the vehicle-seating
environment. Once the model is positioned correctly within the CAD environment, these
simulations will estimate interior factors such as headroom, legroom, access to controls

and interference with hand brake application or other operational movements.

One of the important inputs in the development of these software simulations is a
mathematical representation of a wide database of vehicle occupant locations, including

the location of occupant’s HJC in a deflected seat for a range of seat, package and

anthrOpometric variables. The present study compares the experimental method of
locating the HJC in automotive seats with a mathematical prediction algorithm. The

present study was broken into three phases, which are briefly described below.

In phase I of the study, the stiffnesses of the seat pans (seat cushions) of six
automotive driver seats were measured using the ASPECT manikin [2] along with video
based motion measurement techniques. The experimental data for vertical and horizontal
deflection measures of the manikin H-point were obtained and used as input into the
mathematical model developed by Radcliffe [6]. The Radcliffe mathematical model
estimated the seat pan stiffness based on experimental data for a sequence of manikin
loading steps. Based on the Radcliffe model, six seats were categorized according to their

seat pan stiffness in comparison to 30 other seats analyzed by Radcliffe.

Three of the six seats were chosen for inclusion in phase 11. Seats were selected to
represent a wide range of seat pan stiffnesses. During phase II, kinematic data from
fifteen male subjects were collected in the three seats selected from phase I. The data
represented three-dimensional locations of various anatomical reference points and points
on the test seats. In each case the location of the subject’s HJC was computed based on

the data gathered from the motion measurement system.

 

Figure 2: Representative picture of subject testing in Phase II

Lastly in phase III, the data from phase H were used as an input to a previously
developed mathematical model by Bush and Macklem [3] to predict the location of HJC.
In the previous study by Bush and Macklem it was observed that the Radcliffe model,
that was based on industry standard manikins, could only predict the deflection of mid-
sized and large male occupants, but offsets to the Radcliffe curve were developed by
Bush and Macklem [3] for other anthropometries. These offset curves had a liner trend
between male occupants of average height and weight and tall but lightweight male
occupants. Also for female occupants the offset curves had parabolic trend. This
prediction model was based on the subjects’ weights and the seat stiffness curves and had

been developed on data from three seats. Data from phase II of the present study was

used to calculate the HJC locations of 15 male occupants in 3 seats .The HJC locations
were also estimated using the Bush—Macklem prediction model. The calculated and

predicted HJCs were then compared and used to improve the Bush-Macklem model.

2. BACKGROUND

Background for phase I

The primary purpose of phase I was to characterize a set of six automobile seats
according to seat stiffness and select three seats to be used in phase II that spanned range
of seat pan stiffnesses. In a previous study by Radcliffe [6], a mathematical model was
developed to represent experimental data collected for seat pan stiffness (Figure3). The
Radcliffe model was used to quantitatively describe and evaluate the mechanical
properties of automotive seat pans. In the study by Radcliffe, the deflections of both the
ASPECT [2] and SAE J 826 [1] manikins into the seat were measured and modeled for 30

production and prototype seats.

 

H-point force Vs H-point vertical deflection

H-polnt turned!)

0 50100150200250300350400450500550600

J A L

O

 

+ From Kinetic Model
_ _____ + From Experimental Data

 

 

.a
O

 

 

H-polnt Vertical Deflection(mm)
8 8 8 8 B

 

 

5‘

 

 

 

 

Figure 3: Example of kinetic model by Radcliffe that simulated the seat pan stiffness response of one
of the test seats (Audi Leather) for incremental loading of ASPECT butt and thigh region.

Each of the two manikins was loaded incrementally and downward motion of the

manikins was measured and recorded. Next, the modified Taylor series based equation

set was found to best fit these experimental parameters thus simulating the experimental
behavior of the manikin. All the experimental data for developing this model was
collected manually with a scale. The kinetic model developed by Radcliffe was a good
predictive tool that described and simulated the buttocks and thigh region of industry

standard seating manikins and their interaction with the seat pan.

Buttock Loads

. Left HJC target
Right HJC target

Thigh Loads

   

_ Left knee target
Right knee target

Figure 4: Example of buttock and knee loading of ASPECT manikin

meas
stud}
was I
Incas
meas

local

”ICES
0f m.

Steps

3.PHASE I

3.1 Methods used in phase I
Experimental data acquisition using ASPECT manikin

In previous work by Hubbard and Gedraitis [7], an experimental technique for
measuring seat pan stiffness with the SAE J 826 [l] manikin was developed. The present
study used a similar technique however a newer manikin, called the ASPECT manikin,
was used to record these measures. The horizontal position of H-point, which was not
measured in previous studies, was also measured in the present study. These
measurements of horizontal shift of H-point were not included in predicting the HJC
location however might be useful in future studies. The butt-thigh section of ASPECT

[2] manikin was removed and used in testing as briefly described below.

All six seats were mounted on flat wooden bases at a cushion pan angle of 150
measured using OSCAR [1] manikin. OSCAR [1] manikin was used because the method
of measuring a cushion pan angle is standardized using that manikin. Following are the

steps used in testing all six seats with the ASPECT [2] manikin.

l’A Using the butt-thigh segment of ASPECT seating manikin (shown in Figure 4),
L/“in'cremental loads were applied. Targets were attached on left and right H-point axis
locations and also on left and right knee locations of the ASPECT Butt Thigh

(ABT) section (Figure 4). The vertical and horizontal deflections were measured
using Qualisys motion measurement system. The position data were collected for 1

second per load increment with a frequency of 12 Hz.

a?!
0

up:
uOURa-Q

~

v.50

E

«E

80

F'gllre.

Hing!"

B. Vertical and horizontal positions of the H-point and knee axes were also measured
manually using a scale after each applied load. The manual measures were later
compared with those from Qualisys motion measurement systein.

C. H-point axis measurements were made on the left and right side by measuring the
position at the tip of a rod extending out from the H-point axis center (Figure 4).
Left and right H-point and knee axis measurements were made at the same distance
from the vertical plane of symmetry of the manikin. The variation in level on each
side of the manikin was averaged for both horizontal and vertical measurements.
Again, the horizontal deflection of H-point into the seat was measured on both sides
of H-point axis using recliner pivot of each seat as reference.

D. Between each loading increments (Figures 6-18) a waiting period of 5 minutes was

maintained so that the seat pan attains equilibrium with the added load.

E. The load deflection data from Qualisys motion measurement was compared with

H point Load (N)

 

 

 

 

 

 

 

 

0 100 200 300 400 500
o l r l i
10 is W“ ' ‘W , ,, —#7 , ,, + H-point deflection
obtained manually

'g 20 - A s z
5
g 30 .. ”T ’A ‘ "" ’ -I—H point deflection
3 obtainedby
g 40 --— 7—“ ,7. ,, . ,L. W Qualisys
O
D
g 50 ~L— ‘ ..... - -
s
:c 60 +— — a -

70 «e w A- ~ ~ a ~

80

Figure 5: Comparison Between manually measured data and data collected from Qualisys motion
measurement system for Tahoe (Cloth) seat.

 

the

her

me

usi

SlU

Ml

ant

dill

3H

pr:

F.

the manually measured data for four of the six seats tested to verify the consistency
between measures from Qualisys system and manual measures. Manual
measurements of H-point and knee heights were taken with lab floor as reference
using a ruler with a minimum scale of 1mm. Hand measurements taken in earlier
studies [6] provided reasonable data. However since the data in phase II was collected
with the motion measurement system, the seat protocols were established in phase I
and were carried into phase H. Figure 5 shows agreement between the motion data
and hand measurements and consistency of the motion measurement data. The
average of difference between the two methods for all four seats was 1.8 mm. After
comparison of data for 4 seats it was felt that target data were sufficient measures.
Manual measurements of horizontal deflection of left H-point were taken for two
seats to observe how much the H-pont moves horizontally. Those are tabulated in
Appendix B-5.These measurements were not used in any of the calculations in the

present study.

The data obtained form motion measurements for all six seats was input to Radcliffe

model to generate stiffness curves.

10

 

)3
I)
I‘-

I"; ‘

Fig“;

dfifle

3.2 ASPECT butt thigh loading steps

Figures 6-18 show the load steps followed on all of the seats tested.

  
 

Left side target, on
[the H-point axis

Figure 6: SAE J826, 2-D template positioned in Tahoe Leather seat for the load step zero with a
total weight of 0 N.

The 2-D template of J 826 manikin was used as a reference to measure the
deflections. The template was placed on the seat so that the template edges fully touched
the surface of the seat cushion and seatback near the mid plane of the seat but avoiding
any indentations in trim. Targets were attached to left and right ends of a thin solid rod
with circular cross-section passing through the H—point axis. This load step represented

location of H-point axis in unloaded condition of seat or Zero load step.

 

Figure 7: (Load step 1) Aspect Butt Thigh (ABT) segment with no weights added with a total weight
57 N

 

Figure 8: (Load step 2) ABT segment with two 6-plate H-point weights on H-point axes with a
total weight 125 N. Front and isometric views.

_ _ i .wxx “Went K Rh... 0 8311'! .2...m..¥.:.~l.rw..w .rw ibupm. n ...4 wk PW! ~ h x .E .3. ct. KKM .mfafluhpfia‘hvihrlhw

 

Flg’m

 

Figure 9: (Load step3) Load step 2 plus 2 five plate H-point weights on H-point axis with a total
weight 185 N. Front and Isometric views.

 

Figure 10: (Load step 4) Load step 3 plus 2 five plate H-point weights on H-point axis with a total
weight 231 N. Front and Isometric views.

13

|l|li ‘ I
_. a. ....
.115 .c t. a m... .%.

.u F

fi-DVva‘lsfnu

.. ...I . . . . .
ER. Eganraw.

  

N W .
m. L

vi

*3

\rhafltw. . ..
. . .

. ...
..l‘i. .

he. she. . .....

. L...
-§ 7...
_, . . J ahthiltEWE.

....
w; L

 

sld

 

Figure 11: (Load step 5). Load step 4 plus two torso weights on H—pt axis just inside shell of ABT
with a total weight 278 N. Front and Isometric views.

 

Figure 12: (Load step 6) Load step 5 plus two torso weights on H—pt axis just inside shell of ABT with
slots down and back rear edge resting against shell with a total weight 310 N. Front and
Isometric views.

 

F . .. . r harshest- 1-..... . hm...

. . . L. -
. .. . . - .. .. I . .
Emu. PEPwuclsvanlw

 

 

Figure 13: (Load step 7) Load step 6 plus two torso weights behind the center structure with slots
forward with a total weight 348 N. Front and Isometric views.

 

Figure 14: (Load step 8) load step 7 plus two torso weights behind the center structure with slots
forward with a total weight 388 N. Front and Isometric views.

15

 

 

:3.

 

 

 

Figure 15: (Load step 9) Load step 8 plus two torso weights on center structure with slots up and
back lower edge against nylon bushing with a total weight 416 N. Front and Isometric views.

 

Figure 16: (Load step 10) Load 9 plus two thigh weights with a total weight 431 N. Front and
Isometric views.

 

(I

 

'.1‘%‘;-‘.r ’. .,_..:.. .u‘ .-‘.

 

 

 

 

 

 

Figure 17: (Load step 11) Load 10 plus two thigh weights with a total weight 446 N. Front and
Isometric views.

 

Figure 18: (Load step 12) Load 11 plus two thigh weights with a total load 461 N. Front and
Isometric views.

17

Loads at the H-point and front (knee axis) of ASPECT butt thigh segment (ABT) for all
load steps are shown in Table 1.

Table]: Loads at the 11-point and front of ABT for all load steps

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Load Step Load on H-point axis Load on front of

(Newtons) ABT. (Knee load)
(Newtons)

O 0 0

l 57 29

2 125 29

3 185 29

4 231 29

5 278 29

6 310 32

7 348 26

8 388 21

9 416 28

10 431 56

11 446 83

12 461 1 1 1

 

 

 

 

 

l8

IJJ

ln
kinetic
Since I
briefly
Th
forces 1

H-poin'

rig-

K1101
e‘Il’i’lllnel

altomom.

3.3 Coupled Force and Moment Kinetic model. |6|

In previous work by Radcliffe [6] it was found that a coupled force and moment
kinetic model was effective in representing the effects of the seat on manikin position.
Since the Radcliffe model was used for seat selection in the phase I of this study, it is
briefly explained in the following paragraphs.

The coupled force and moment kinetic model [6] (figure 19) represented two input
forces FH and FK at the H-point and knee respectively, and two reactions: force, RH at the

H-point axis, and reaction moment M about the H-point axis.

Slider prohibits horizontal translation of
the H-pt.

    

FK

Figure 19: Coupled Force and Moment Kinetic model [6]

Knowing the force—deflection relations of the H-point and knee axes from the
experimental data, this model can be used to simulate the static response of the

automotive seat to the ASPECT manikin loading.

A static mathematical analysis coupled with the use of Microsoft (MS) Excel solver
optimization tool was used to develop the mathematical equations, which represented the
manikin response to seat loading. This mathematical modeling used following steps.

A) Developing experimental and simulation equations.
Two sets of equations, a static set and a simulation set were developed and the aim of
the procedure was to find the stiffness coefficients that would give minimum Root Mean

Square (RMS) error between the simulation and experimental data.

E uationso Statics ure 19

The set of static equations (equations 1&2) represented experimental reaction forces
and moments based upon experimental, incremental input forces F" (Force at H-point)

and FK (Force at Knee).

R” = F” + FK (1)

M H = F K * LK * (COS(GTHIGH )) (2)

Slider prohibits horizontal translation of

 

Figure 19: Coupled Force and Moment Kinetic model [6]

20

 

sane

cal lei

DEIWI

mom:

.1!

II

3) Ext
The
exPerth

Paramel

lg

T0 ca
RadCllffe .
Tdblfi‘z‘ [h‘

spreacghee

Simulation Equations

The set of simulation equations (equations 3 &4) were based on modified Taylor
series expansion and included the stiffness coefficient terms (n K10. , "K1,, ”16",,me )
called seat factor parameters, which needed to be Optimized, so as the RMS difference
between the simulated force and moment (F,l ,M n ) and the experimental force and

moment was minimum.

F, =Z(,,K,,,*6"+,,Kfl,*z9") (3)
0

M. =Z(.K..*6"+.K... *6") (4)
0

B) Experimental data acquisition.

The experimental data needed for the mathematical analysis was acquired using the
experimental data acquisition procedure discussed earlier in section 3.2. The four input
parameters needed from the experimental data for the simulations were:

1. Static incremental load at H-point axis (Fh)
2. Static incremental load at knee axis (front of ABT) (Fk)
3. Average vertical displacement of the H-point at that given load.
4. Average vertical displacement of the knee axis of ABT.
To calculate this mathematical model, an Excel spreadsheet was designed by
Radcliffe that will be referred to as the seat factor solver (SFS) was used. As shown in
Table2, the four input parameters correspond to the first four columns of the analysis

spreadsheet. The last four columns are the calculated values.

21

(LIE

 

Table 2: Input data acquired from experimental incremental loading of the manikin is shown in
white font within dark cells, along with calculated values based on experimental data shown in dark
fonts within gray cells.

Fh(N) Fk(N) Avg (z,mm) Avg (z,mm)
O 0 253.7
57 29
125 29
185 29
29

278 29
310 32
348 26
388 21
416 28

56
446 83

 

Where,
Fh(N) = Static incremental load at H-point (experimental load data).
Fk (N): Static incremental load at knee axis (experimental load data).
H—pt Avg z(mm) = Average vertical displacement of left and right side targets on
the h-point axis of ABT (experimental data measured from motion measurement
system).
Knee Avg 2 (mm): Average vertical displacement of left and right side targets on
the front of ABT (experimental data measured from motion measurement
system).
Thigh Angle (degrees): Angle of thigh segment of ABT with reference to

horizontal obtained from column 3 and 4 and H—point to knee length.
(Calculated).

C) Calculation of static reaction force and moment.
Experimental data from the six seats tested were input into the Radcliffe SFS. The

static reaction forces and moments, which were based on experimental loadings and

22

 

equal

eXpan
hnean
multir
Expan:

Itsults

measurements, were calculated for each incremental loading step according to equations
(1) and (2). These experimental reaction forces and moments were then used as a basis of
comparison to optimize the stiffness coefficients (K’s) in equations (3) and (4) using the

Seat Factor Solver spreadsheet within MS Excel.

D) Optimization of seat factors within modified Taylor Series based simulation

equations.

The simulation equations (3) and (4) were based on Modified Taylor Series
expansion. These Taylor series expansions were modified in that the 0‘h order,

linearization term was neglected. Also neglected were terms in which the coefficients

multiplied by both variable terms (K*[ 5" * 19" D. With increasing order of Taylor series

expansion equations the differences between the experimental data and the simulation

results were decreased by each addition of error correcting higher order terms.

{fFirst (n = 1), second (n = 2), and third (n = 3) order modified Taylor series
expansions were investigated (refer to equations 3 and 4) within the SFS to reduce the
RMS differences between the experimental data and simulation results. It was observed
that the second order Taylor series expansion equations produced the simulation HJ C
deflections, with RMS error less than 5.4mm, compared to the third order equations that
produced the simulation HJC deflections with RMS error less that 4.1mm for all six seats
tested. (Refer to the Appendix.) The second order Taylor series expansion equations that
reasonably simulated the experimental and data were used to plot the seat pan stiffness of

“\

all six seats.

23

 

JJ
(A

suffr

iXSld

repre.

meet

Table

 

3.4 Seats tested for Phase I

The goal of phase I was to select 3 seats that would cover a range of seat pan
stiffnesses. For this purpose, the following six seats listed in Table 4 were tested with
ASPECT manikin for obtaining H-point vertical and horizontal deflections. The seats
represented wide range of car segments. Table 3 shows the available information about

the car, year of manufacture and name of the manufacturing company for each seat.

Table 3: List showing available information about six seats tested in phase 1.

 

 

 

 

 

 

 

 

Name Car Year Seat
Manufacturer

Seat A Audi 1999 Unavailable

Seat B Ranger (Jeep) 2002 J CI

Seat C SLK (Mercedes) 1999 Unavailable

Seat D Tahoe (Cloth trim-Chevy) 2001 Lear Corporation

Seat E Tahoe (Leather trim-Chevy) 2001 Lear Corporation

Seat F BMW Sporting Unavailable Unavailable

 

 

 

 

 

The following is the description of each seat with pictures.

24

 

Figure 20: Seat A, the Audi front and side view.

Seat A, a leather covered 1999 Audi (Figure 20) had motorized adjustable
mechanical lumbar support, cushion lifter, and back recline operations. The seat back
angle for this seat ranged from 9° to 63° rearward from vertical when measured with J 826
manikin. The seatback bolsters were soft and prominent whereas the seat pan had soft flat

bolsters.

25

by Jol
seat b

lllCZlSl

 

w Ira-11m:

The gray and black cloth covered 2002 Ranger seat B, (Figure 21) was manufactured
by Johnson Controls Inc. This seat had manual recline with no lumbar adjustment. The
seat back angle for this seat ranged from 11° forward to 48° rearward from vertical as

measured with J-826 manikin. The bolsters on seat back and seat pan were firm and flat.

 

 

Figure 21: Seat B, the Ranger front and side view.

26

Seat C, the 1999 SLK as shown in Figure 22, was covered with black leather. This
seat had manual recline and with no lumbar support feature. The seat back angle for this
seat ranged from 15° forward to 78° rearward from vertical as measured with J-826

manikin. The bolsters on seat back and seat pan were firm and flat.

 

Figure22: Seat C, the SLK front and side view.

27

The gray cloth covered 2001 Tahoe seat (seat D), shown in Figure 23, was
manufactured by Lear Corp. This seat had manual recline with no lumbar support feature.
The seat back angle for this seat ranged from 10° to 47° rearward, from vertical as

measured with J-826 manikin. The bolsters on seat back and seat pan were soft and flat.

 

Figure 23: Seat D, the Tahoe-cloth front and side view.

28

Seat E, leather covered 2001 Tahoe (Figure 24) seat was manufactured by Lear
Corporation. This seat had motorized mechanical lumbar support, cushion lifter, and back
recline operations. The seat back angle for this seat ranged from 12° to 45° rearward
from vertical as measured with J-826 manikin. The bolsters on seat back and seat pan

were soft and flat.

 

Figure 24: Seat E, the Tahoe-leather front and side view.

29

Seat F, a cloth covered BMW was manufactured by Lear Corp. This seat had manual
recline with no lumbar support feature. This seat ranged from 12° to 75° rearward from
vertical as measured with J -826 manikin. This seat had prominent, firm bolsters on seat

pan and seat back.

 

Figure 25: Seat F, the BMW- front and side view.

30

3.5 Seat Selection for phase 11

Based on the coupled force and moment kinematic model with 2nd order Taylor
series expansions discussed in section 3.3, the seat pan stiffnesses of all six seats were
plotted. These plots of reaction force under h-point versus h-point deflection were
compared to examine which seats should be tested for phase II.

Figure 26 shows comparison of H-point load vs. H-point deflection for all six seats
tested in phase 1 along with the 22 other seats tested by Radcliffe [6] in his study to

develop the SFS and kinematic models.

31

1"

05:065. «501.... Dov-03: Oa floocoaoact «SEW-acoEOOSA-ED

Selma

 

H-Point Deflection vs. H-Point Load

 

 

 

 

 

Displacement (mm) Referenced to Unleaded H-Polnt Machine

adrift:

 

 

 

 

 

 

 

185 285 385 485 585
Force (N)

 

 

 

8 Prototype (Lear)
— —C Prototype (JCI)
.. - - M Vectra Opel (Lear)
_ - -.1 (Audi) .

- s 605 (Peugeot)

E Sport (BMW)SSeries
l Saturn (JCI)

- ‘4 Heavy Trck (Lear)
K (VW)
— —PCamry(JC')
- - - 0(BMW)7Series
_ - '0 Basic (BMW)SSeries

-- — L 806 (Peugeoil

- “A Neon (JCI)
~ - F NS (Atoma)

*- ‘G AS (Atoma)
0 Ford Explorer (Lear)
_ —R Chevy Truck (Lear)
- - - B LH (JCI)
— - 'N Taurus
_.... --—~- H Split Bench (Lear)

“- 9 Aurora (Delphi)
+ Tahoe_ciothe
+Auld_2001
—-0—Tahoe_leather
+Ranger
+SLK_Ieather

m

 

 

 

 

 

 

 

Figure 26: Chart comparing H-point deflection Vs. H-point load for seats tested in phase 1 (shown in
bold legends). Also shown are H-point deflections with increasing H-point load for 22 seats

tested by Radcliffe [6].

The six seats were categorized according to their seat pan stiffness. This was

accomplished by comparing the H-point deflection corresponding to H-point reaction

force of 410 newtons for all six seats®e H-point reaction force of 410 newtons was

selected for comparison because it is 54.3% the body weight of an average sized male

32

the c

UdeOEDlOu now-occaohom «ES.» «coEeonfia‘Q

 

F '8an

occupant [171 lb body weight, 69” height] that passes through the buttocks according to

the dissertation by Bush [9].

 

H-Point Deflection vs. H-Point Load

 

 

+Tahoe_clothe

 

 

 

*Auid_2001

 

'0-SLK_leather

 

 

 

 

Displacement (mm) Referenced to Unleaded
H-Point Machine Shell

 

 

 

 

 

 

 

+Ranger
.40 ._ L c
*‘l’ahoeJeather
-50 wt , i @—
410 N +BMW(Sporting)
-60 I '

 

185 285 385 485 585
Force (N)

 

 

 

Figure 27 :' Chart comparing H-point deflection versus H-point load for six seats tested in phase I.
The deflections corresponding to a H-point load of 410N were compared.

33

Three seats from phase I study were selected to represent a wide range of seat pan
stiffness. The three seats were,

1. Seat E, the Tahoe (2001 SUV) with leather trim and it had H-point deflection of
50 mm corresponding to 410 N, (figure 28). The seat represented a soft seat pan.

2. Seat F, the BMW (sedan) with H-point deflection of 32 mm corresponding to 410
N, (Figure 29) represented stiff seat pan.

3. Seat C, the SLK (1999 sports) with H-point deflection of 45 mm (figure 30)
corresponding to 410 N, represented medium stiff seat pan.

It can be seen from figures 26 and 27 that the three seats selected from phase I

covered wide range of cushion pan stiffness.

 

HJC Force VS Deflection -Tahoe (SUV)

0 1 00 200 300 400 500 600

 

    

 

L-°— Kinetic Modefi

 

1'0
0
l
l
l
|

 

 

Vertical deflection(mm)
8
l

 

 

 

 

Figure 28: H-point Force Vs H-point deflection for Tahoe (SUV) seat from Radcliff’s 2"" order
kinetic model [6].

34

 

Fig

PIgu

 

HJC Force VS Deflection BMW (Sedan)

 

 

 

 

 

O 1 00 200 300 400 500 600

’e‘

E.

g l—e— Kinetic Model
3

“6

u

'5

2

g -50 «f A

>

-60

 

 

 

Force(N)

 

 

 

Figure 29: H-point Force Vs H-point deflection for BMW (Sedan) seat from Radcliff’s 2"" order
kinetic model [6].

 

HJC Force Vs Deflection (SLK sports)

0 1 00 200 300 400 500 600

 

 

 

 

+ Kinetic Model

 

 

Vertical deflection(mm)

 

 

 

 

Force(N)

 

 

 

Figure 30: H-point Force Vs H-point deflection SLK (Sports) seat from Radcliff’s 2'"I order kinetic
model [6].

35

4. PHASE II

The goal of phase II was to collect data to calculate hip joint center (HJC) locations
of a sample of male occupants seated in the three seats selected from phase I. The three
seats included Seat E-Tahoe, Seat C-SLK and Seat F—BMW and selected because they
encompassed a range of seat pan stiffnesses. These three seats were tested with people of
various heights and weights. The locations of their HJCs in the seats were computed and
compared to the prediction model developed by Bush and Macklem [3] (discussed at the
end of section 1.0). The testing protocol, the procedure for data collection, and
calculation of HJC deflection are discussed in this section.

4.1 Test Subjects

The purpose of this study was to collect additional data to validate the method of
HJC prediction by Bush and Macklem [3]. The scope of this study addressed only the
male data. To account for a range of possible HJC locations in the seats, the sample of
male occupants covered a wide range of heights and weights. The development of the
HJC prediction method was based on a previous study by Gutowski[l2] therefore a
sample similar to Gutowski’s was tested including male occupants of average height and
weight, heavy and tall men, and heavy but light men. Thus, data from the present study
could be compared to the prediction developed by Bush and Macklem [3] using
Gutowski data. The goal was to test five male subjects from each of the anthropometric
groups. These groups were based on N ATIK [18] data (shown in Table 4). Subjects were
recruited on a volunteer basis. They were initially screened to see if they fit in the desired
height and weight categories. The actual test subjects varied slightly from their desired

height and weight (refer to Table 5).

36

Table 4: Desired human subjects’ anthropometrics as per NATIK [18].

 

 

 

 

Anthropometric group for Males Weight Stature
50% height and 50% weight (50H50W) 171 lb 69 in.
95% height and 5% weight (95H5W) 135 lb 73 in.
95% height and 95% weight (95H95W) 216 lb 73 in.

 

 

 

 

 

Table 5: Actual test subjects’ anthropometric measurements.

in tall and

 

*STDEV : Standard Deviation

37

The subjects in average height and weight category had average height matched, but
were a little light. The subjects in tall and heavy and tall but light groups were one to two
inches shorter and little lighter than desired.

Generally speaking, the 50%tile height and 50%tile weight (50H50W) subjects
represented male occupants of average height and average weight. The subjects with
95%tile height and 95%tile weight (95H95W) represented tall and heavy male occupants
and the subjects with 95%tile height and 5%tile weight (95H5W) represented tall and
light male occupants.

Six subjects were tested from tall and heavy (95H95W) group, five from average
(50H50W) group and four subjects from tall and light category (95H5W). The
anthropometric measurements along with averages and standard deviations for all
subjects tested are listed in table 5 including pelvic dimensions (refer to Figure 35) which

were necessary for calculating the hip joint centers of the subjects in the reference seat.

38

4.2 Test Buck setup

All subjects were tested in a reconfigurable test buck. For testing, the H-point to Heel
point vertical distance also termed as H30, (Figure 31) was set as per the seat type
according to Johnson Controls Incorporation’s (JCI) seat testing standards, listed in Table
7 [10]. The J 826 manikin and corresponding procedures were used to measure and obtain

the dimensions listed in Table 6.

Zlab

 

H17

 

 

 

L53

Figure 31: Test buck dimensions. [12]

39

Table 6: Dimension descriptions for test buck set-up.

 

SAE #

Dimension Description

 

H3O

Seat (J 826 manikin H-point) Height

 

 

 

 

 

 

 

 

 

L27 Cushion Angle with respect to Horizontal

L11 Steering Wheel to Toe bar (X)

H17 Steering Wheel to Heel Point (2)

H18 Steering Wheel Angle with respect to Vertical
W9 Steering Wheel Diameter (outer)

L53 H-point to Toe bar

L40 Backrest Angle

 

Table 7: Package dimensions with J 826 manikin [1] for the typical car segment-seating environment

 

 

 

 

 

 

 

 

 

 

 

[10].
Package 1 2 3 4

Typical Segment Sporty Passenger Car SUV Van
Torso Angle (°) 27 24 21.5 20
Hip Angle (°) 98 95 95.5 95.5
Knee Angle (°) 132 124 121 115
Foot Angle (°) 87 87 87 88
H—point to Heel point—Z (mm) 190.73 239.82 325 360.69

 

 

The cushion pan angle of all three seats was fixed at 15 degrees using J826 manikin
[1] and the SAE J1100 [l 1] procedure for measuring the cushion angle. Out of the three
seats only Seat E- the Tahoe (SUV) had a lumbar prominence adjustment. To maintain

consistency in testing protocol, testing was performed with the lumbar support in the off

position for the Tahoe seat.

40

 

 

4.3 Reference Seat

Along with the three production seats, subjects were tested in a reference seat also
termed a ‘hard seat’. The hard seat was a wooden seat without any padding, and therefore
no deflection of the seat pan occurred when loaded. The seat was used to collect
positional data on the pelvis and other bony landmarks. The data were later used to
calculate the HJC location and the deflection of the buttocks in the hard seat. The seat
was set with pan angle of 15° and back angle of 23° (refer to Figure 32). These values
corresponded to a cushion pan angle of 11° and a back angle of 24° when measured with

the SAE J 826 manikin [1].

23°

580 mm

 

 

 

500 mm

Figure 32: Reference hard seat dimensions. [12]

41

4.4 Test protocol

From Gutowski’s work [12] it was observed that when subjects were not given any
instructions about the placement of their buttocks in the seat pan, the positions of their
HJC’s had an anterior shift as compared to when they were asked to place their buttocks
against the seat back. Based on this finding, a test protocol was designed to capture these
differences in HJC positions. Each subject was instructed to place his buttocks against the
seat back termed ‘instructed’ position, and then in a preferred position. No instruction
about placing his buttocks was given to the subject in the preferred position.

For both the instructed and preferred positions, the steering wheel position could be
adjusted vertically and horizontally to achieve the preferred H17 distance (refer to figure
31). Also the toe bar could slide forward and rearward to set subject preferred foot
position and thus preferred L11 distance (refer to figure 31).

In the instructed position (figure 33), subjects were asked to sit with their buttocks
firmly against the seat back to achieve the most posterior position of their HJC. The seat
back recline angle of each seat was set to 24° and cushion pan angle to 15° using .1826
[1] manikin. The subjects were asked to maintain contact with a foot support that
represented the gas pedal location. To achieve this, the subjects were allowed to move the
seat fore and aft. Subjects were free to choose the position of their hands relative to the
wheel and were able to slide the wheel fore and aft to their preferred location.

In the preferred position (figure 34), subjects were asked to sit with their buttocks
placed in any preferred position on the seat pan. They were also free to adjust the recline
angle, while cushion pan angle was fixed to 15°. As in the instructed position, the fore/aft

position of the seat as well as the steering wheel height was adjusted by the subject.

42

Again subjects were asked to maintain contact with the foot support. Subjects were free

to choose the position of their hands relative to the steering wheel.

 

Figure 34: Subject 10 seated in the preferred position in SLK seat. Subject chose more reclined
position than the instructed position and a preferred position of his arrm.

43

4.5 Testing Procedure

All testing was performed in the Biomechanical Design and Research Lab (BDRL)
of Michigan State University and was approved under University Committee on Research
Involving Human Subjects IRB#96-054 [13]. Approved consent forms were fully
explained to each subject prior to testing. The same test procedure as described below
was followed for each of the fifteen subjects.

Subjects were asked to wear tight fitting clothes (Figures 33 and 34) so as to reduce
the movement of the clothing relative to the body and thus the motion of the targets that
were attached to the clothing. If the subjects did not have the necessary clothing, it was
provided. Once in the appropriate attire, their height (without shoes), weight and age
were recorded. Manual measurements of pelvic height (PH: the perpendicular distance
from the line joining the right and left Anterior Superior Illiac Spine, ASIS to the top of
the pubic symphisys), pelvic width (PW: distance between the right and left ASIS) and
pelvic depth (PD: distance between right ASIS to the mid Posterior Superior Illiac Spine,
PSIS) were measured with afintropometer (see Figure 35);?)These measurements were
necessary for the computation of HJC location in the hard seat.

Next, targets were put on key locations on the test seats (Figure 36) and on bony
landmarks of the subjects (refer to Table 8). For all subjects only the right side of the
body was targeted and right HJC was calculated. Since motion measurement system only
had five cameras and was limited to 30 targets it was not possible to target both sides of a
subject. Also the motion system is able to catch a maximum of only 30 target locations,

which does not allow putting targets on both sides of the body of a subject. Wherever

possible targets were affixed directly on the skin while the rest were taped to the clothing

at the target locations.

Pelvic Width

Right ASIS , Left ASIS

E Pelvic
Height

 

__________‘!-

 

Figure 35: Pelvic width and pelvic height [14]

{The target on the right Anterior Superior Illiac Spine (ASIS) and the target on
Lateral Femoral Epicondyle of the right knee were the two targets necessary to
calculate the HJC in the seated position using the method developed by Bush and
Gutowski [15]].(‘2The locations of other targets will be useful in further study of various
anatomical landmark positions responses of the subject to the seat, however were
outside the scope of the present study.

Data files were recorded with instructed and preferred positions as discussed in
section 4.4. Two data files for each of the two positions were recorded with 12Hz
frequency for 3 seconds using the Qualisys System. Between the two trails of the
same position, subjects were asked to move around in the seat and then reposition
themselves. To avoid the targets being knocked off during the transition, subjects
were not allowed to get out of the seat between the trials. The order in which a subject

would sit in each of the three seats was randomized.

45

Table 8: Target locations for seat testing.

 

 

 

 

 

 

 

 

 

 

 

 

 

Production Seat Test Subject
Seat Pan Front and Rear Stemal Notch
Recline Top and Bottom Mid-Stemum
Right Toe Bar Left ASIS and Right ASIS
Buck Front and Rear Mid-Thigh
Buck Top Right Knee (Lateral Femoral Epicondyle)
Recliner Pivot Right Ankle (Lateral Maleolous)
Right Ball of Foot
Right Shoulder (Acromion Process)
Right Elbow (Humeral Lateral Condyle)
Right Wrist (Ulnar Condyle)
Right Head (Temple) & Forehead

 

 

 

Right Head

Forehead

Wrist

 
    
  
 

Sternal Notch

 

Right Mid
Shoulder -

e a

Right and Left

 
  
 
 

 

Recline
Top and .2
Bottom ;f;3 Right Knee
Seat Pan Front " 33:-t
and Rear
/'
Recliner Pivot
Toe Bar
Buck Top
Buck Rear Buck Front Right Thigh Right Ankle

Figure 36: Taget locations on seat and subject for testing in three production seats.

After collecting data files in all three-production seats, the subject was asked to sit
in the hard seat with the targets attached to landmarks noted in Table 9 and targets on the
reference locations on hard seat. (See Figure 37 and Table 9). For the hard seat trials, five

additional targets were placed on the spinous process of C7, T8, T12, L1 and L3.

46

To find the HJC in hard seat,.\both right and left Anterior Superior Illiac Spine

x...

(ASIS) locations were targeted. Two hard seat trials were recorded for 3 seconds at a

M...»-

frequency of 12Hzi‘,The subjects were asked to reposition themselves in the hard seat

between the two test files. Again, they were not allowed to get out of the seat between the

two trials.

Table 9: Target locations for hard seat trials.

 

Reference Seat Test Subject

Seat Pan Front Stemal Notch

Seat Pan Rear Mid-stemum

C7 (Seventh cervical vertebra)

T8 (Eighth thoracic vertebra)

T12 (Twelfth thoracic vertebra)

Ll (First lumbar vertebra)

L3 (Third lumbar vertebra)

Left ASIS and Right ASIS
Mid-PSIS

Right Thigh

Right Knee (Lateral Femoral Epicondyle)
Right Ankle (Lateral Maleolous)
Right Ball of Foot

Right Shoulder (Acromion Process)
Right Head (Temple)

Forehead

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

47

Right of head Forehead
Stemal notch
Mid-Stemum

   
  
 
 
  
  
  
  

Right thigh

  

Left and Right
ASIS Right Knee

Right Ankle

Mid

PS'S R' htB II t
Hard seat Fc'igot a 0
front and

rear

 

Right
Shoulder\

Right
Elbow \

 

 

 

 

Figure 37: Target locations for testing in hard seat and targeted subject.

48

5. PHASE III

5.1 Background fomhase III ‘

Elie Radcliffe kinetic model was; based on experimental deflections of butt thigh
segments of ASPECT and J 826 manikins, which represent male occupants of 50th
percentile height and 50th percentile weight (50H50W). Because the kinetic model was
based on the experimental data obtained by manikins of a single size (50H50W) one task
was to determine if the model could be used to predict the HJC deflection people of sizes
other than SOHSOW along with the SOHSOW category.

Bush and Macklem [3] analyzed the applicability of the kinetic model in a previous
study. They used the data from a study by Gutowski [12] in which human occupants of
various sizes and weights were tested in four different seats and their HJC deflections in
each of the seats were calculated directly from the experimental motion data. Bush and
Macklem’s study [3] began by comparing the HJCs computed from Gutowski’s data to
the force deflection curve from the kinetic model. To achieve this, the manikin loading
data for the seats tested by Gutowski was input to the Radcliffe’s kinetic model and the
load deflection curves were obtained. Then using Bush’s [9] loading estimation (Table
10) the deflection was read corresponding to the loading under the occupants’ HJCs (54.3

% of body weight) directly from the load deflection curves obtained from Radcliffe’s

kinetic modelfmk;

s
l

49

Table 10: Loading under the HJC for various anthropometrics as studied by Bush [9]. Body weight
for each anthropometric as per the NHANES [l6]

 

 

 

 

 

 

 

Occupant Category Load under the HJ C in newtons (54.3% of body
weight)

Small Female 271

(5H5W,F)

Medium Male 408

(50H50W,M)

Large Male 432

(95H95W,M)

 

It was observed in the study by Bush and Macklem [3] that using Radcliffe’s [6]
kinetic model, Bush’s [9] loading estimation and Gutowski’s [12] data, the physical
manikins could only predict the HJC deflection of the mid-sized and large male
occupants. A notable deviation in the HJC deflection was observed for other
anthropometries (Figure 38).

Next, Bush and Macklem [3] developed offset curves for predicting the HJC
locations of other sized occupants. These offset equations were based on the HJC
computations in Gutowski’s study on four seats [12]. It was observed that the deviations
in the HJC deflections from the kinetic model had a linear trend between 95H5W males
and SOHSOW males and parabolic trend for 5H5W females, 5H50W females and 5H95W
females (Figure 39 and 40). The HJC deflection for the 95H95W males was found to be
on the extended load deflection curve obtained using Radcliffe’s kinetic model. These
trends in HJC deflection deviations were accounted for by developing mathematical

equations termed as offset curves.

50

 

HJC Force vs. HJC Deflection-Kinetic model and experimental data for
Town & Country seat

HJC Force (N)
0 100 200 300 400 500 600

    

 

10 « L i +2nd order Kinetic

model
- -I- 5H5W Females

N
O

I SOHSOW, Females

8
l
l

@151", "L 4- 95H5W Males
* LL +50HSOW, Males
. —I— Sl-l95W Females

93-55mm
95I-15W,M L. .__._fi#-_ LL +95H95W Males

8 8

 

HJC Vdertical Deflection(mm)
A
O

\l
C
l
l

 

 

 

m
C)

 

 

 

 

 

Figure 38: Averages of HJC forces Vs. Deflections with error bars for various anthropometrics in
Town and Country seat obtained by Bush and Macklem [3] using the data from Gutowski’s study.
Each point represents the averaged HJC deflection of five subjects. A notable difference in average
deflections compared to those predicted by the kinetic model can be observed for anthropometrics
other than SOHSOW males.

A generalized best-fit linear offset equation was developed between 95H5W and
SOHSOW male categories relative to the HJC load deflection curve using Radcliffe’s
kinetic model (Figure 39). A generalized best-fit parabolic equation was developed to
predict the H] C deflections relative to the seat deflection curve of 5H5W, SOHSOW and
5H95W female categories (Figure 40). Bush and Macklem provided offset curves, based
on data from three seats. To refine these curves, (for male occupants’ data only)
additional data were collected on three additional seats. This portion of the work is

considered phase III.

51

 

HJC Force VS Deflection-2nd Order Kinetic Model and
experimental data for LH Tan seat

HJC Force(N)

0 100 200 300 400 500 600 + 2nd Order Kinetic
,___ __._ Nbdel

~10
OIAJ . 1 — . . a 951-15Wlifales

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E

g 10 \ __

G

g 20 \\ + 50wa Males

g 30

a + 951'195W Males

- 40

§ 95H95W, M

8 5° .

> 95H5W; + General Flt Nhle

o 60 *

.I,

I 70 I. sop-mow," —l— Extrapolated
80 Force_Deflection

Curve

 

 

 

 

 

Figure 39: Best fit line for 50H50 W and 95H5W male occupants developed by Bush and Macklem
[3]. Each point represents the averaged HJ C deflection of five subjects. HJC deflection for 95H95W
estimated to be on the linearly extrapolated HJC force deflection curve.

 

HJC Force VS Deflection-2nd Order Kinetic Model and
experimental data With Beat-tit Parabolla for females( LH Tan
seat)

rue Force (N)
o 100 200 300 400 soo soo

 

 

 

—O—2nd Order Kinetic Nbdel

 

” + Si-BW Females

A 50l-60W, Pennies
__ —e— SHQSW Pennies
—— General Flt Fannie

 

 

HJC Vertical collection (mm)

 

 

 

 

 

 

 

 

Figure 40: Best fit parabola for female occupants developed by Bush and Macklem [3].

52

The method used to calculate the deflection of the seat pan under the HJC, the

calculation of HJC locations and the results of the comparison are discussed next.

5.2 Calculation of HJC in test seats

After collecting the data for all 15 subjects in phase II, the next task was to calculate
their HJC locations in each of the three test seats for both the preferred and the instructed
positions (refer section 4.2) and then compute the deflection of the seat pan under the
HJC location.

r"The first step in locating the HJC of the subjects in the production test seats was to
locate their HJC in the hard seat; HJC location in the hard seat was calculated using the
Seidel [17] method, which used of the manually measured pelvis dimensions (refer
section 4.3, figure 35) and the locations of right and left ASIS targets and the mid-PSIS
target (Figure 37). The location of HJC in hard seat was necessary to calculate the
deflection of subject’s buttocks in seated position.

Using the location of HJC in the hard seat and the motion measurement data of the
subject seated in the production (deformable) seat, the HJC for that subject in that
particular production seat was computed.

Two different methods were used to calculate HJC in the production seats. The first
method, used by Gutowski [12] used the coordinates of right ASIS and right lateral
epicondyle (right knee) targets in the actual seat along with the coordinates of right HJC
in the hard seat. In the method used by Gutowski’s [12] three known lengths were used to

calculate HJC coordinates in the sagittal plane which were: the length between right knee

and right ASIS target in production seat, the length between right HJC and right ASIS in

53

the hard seat and the length between right HJC and right knee in hard seat. Using these
three lengths, and the coordinates of the right ASIS and right knee, the right HJC
coordinates were computed. The method used a two-dimensional vector analysis to
obtain the HJC coordinates in sagittal plane.

The method, by Bush-Gutowski [15] used the coordinates of right ASIS and right
lateral epicondyle (right knee) along with the coordinates of right HJC in the hard seat. In
this method, two known lengths were used: the length between right HJC and right ASIS
in hard seat (pelvis length) and the length between right HJC and right knee in hard seat
(femur length). Tish-Gutowski method-assumed that the pelvis length and femur length
remained constant «irrespective of the subject being in hard seat or a deformable seats;
Using the coordinates of right ASIS and right knee in the production seat along with/the
two known lengths, the HJC coordinates were solved for using ariiintersection of sphere
and circle analysis: Thus, Bush-Gutowski method used a three dimensional approach to
solve for HJC coordinates in sagital plane as compared to a two dimensional approach
used in Gutowski’s method. The Bush method however can only be used in a seated
environment.

In previous study by Gutowski [12] the HJC coordinates were calculated only using
the one particular method whereas in the present study both the method used in
Gutowski’s study and Bush-Gutowski method were used to calculate the HJC
coordinates. The HJC vertical deflections then were computed using the HJC coordinates

obtained from both of the above said methods and those were plotted relative to the

deflection predicted by the kinetic model. Both the Bush-Gutowski computations and the

54

computations used in Gutowski’s study were used so data from this study could also be

compared to that obtained from the Gutowski study.

5.3 Method to calculate the H] C deflection
The method used to find the deflection of the production seat under the HJC in the

study by Bush and Macklem [3] was also used in the present study.

   

Estimated
HJC
Location

Hard Seat \‘ Hard seat
reference
Hard seat front target
reference rear
target-Origin for
HJC in hard seat

Figure 41: Calculation of 6, the vertical deflection of the buttocks [3].

Three measurements were calculated to get the HJC location in production seat. First
the vertical distance from H] C to the hard seat pan was calculated and was termed

as 6, (buttocks’ deformation). Next, the distance between a point corresponding to the

H] C vertically downward on the undeflected seat contour to a reference point was

calculated and was termed as 62. The third measurement 6, was calculated as the vertical

distance between the HJC in the production seat and a reference point on the seat.

55

Finally, the HJC deflection (A) was calculated to be A = 5, + 52 - 53. The method is

explained in detail as follows.

From the hard seat data and the measurements of pelvic dimensions the 3-
dimensional coordinates of the HJC with respect to the rear target (origin) on the
reference seat were calculated using Seidel [17] method. The vertical distance between
the HJC and the plane of the hard seat pan was estimated as the measurement of buttocks

deflection and was identified as 6, (Figure 41).

Deflections (both those in the hard seat and in the production seats) are computed
vertical rather that perpendicular to the seat pan. This is because the final seat deflection
was to be compared to that obtained from the kinetic model, which is based on the
vertical deflection of the H-point axis of the manikin.

The vertical distance of the HJC in the production seat was then measured using the

recliner pivot as reference on the seat and was defined as 53. The recliner pivot target

was considered a reference target that did not move during testing. (Refer figure 36).

56

  

Location of HJC relative
to the un-deflected
/contour of the
production seat.

 
 

JC in production seat

 
 

(deformable seat)

  
 

6, _-
Seat /

Recliner
Pivot

 

Seat pan deformation under the HJC, A: 6, + 62 - 63

Figure 42: Computation of Seat Deflection from human data.

To determine the vertical movement of the HJC in a deformable (production) seat,
the location of the HJC relative to the seat pan was needed. So, a point on the undeflected
seat pan contour corresponding vertically downward to the HJC was obtained, and the
distance between that point and a reference point (recliner pivot) for each trial was
calculated as 62. To measure 62 , the seat contour scan was used (refer Figure 43). The
HJC coordinates in the sagital plane (X-Z plane) were obtained using two different
methods as discussed in section 5.1 and the vertical distance between the point on the seat
scan along the Z direction corresponding to the X- coordinate of HJC and the recliner

pivot was measured as 62.

57

 

SLK leather contour

 

 

 

 

 

 

1000 Point on undeflected seat
900 _. _ contour vertically
HJC m. / downward corresponding
800 r production seat // to the H JC

 

 

 

 

 

a \
\
400 62

 

 

 

 

 

Z-Coerdlnates

 

 

 

 

 

 

 

 

 

 

300 4
200 Recliner pivot /
(reference)
1 00
0 l f f

 

-800 -600 -400 -200 0 200
X-Ceerdinates

 

 

 

Figure 43: Un-deflected seat contour scan obtained from Qualisys system used to calculate 52 .

The vertical deflection of the HJC (A) was considered as the vertical deflection of

the production seat under the HJC, which was equal to ((51 + 62) - 63. Where,

6, = Buttocks vertical deflection in the hard seat.
62 2 Vertical distance between the un-deflected seat contour point corresponding

to the HJC in production seat and the recliner pivot reference.

63 2 Vertical distance between the HJC in production seat and the recliner pivot

reference.

The deflections calculated using the above method were then compared with the

deflections produced with the kinetic model.

58

 

5.4 Results of comparison between the HJC deflections computed experimentally and
those predicted using the kinetic model.

The HJC deflection was calculated for each subject in each of the three seats, for
both the preferred and instructed positions, for two trials in each position. The HJC
deflections were averaged over each category in both instructed and preferred positions to
obtain one number per category in each seat and were plotted on the force deflection
curves.

Out of the three seats tested it was observed that the H] C deflection pattern for most
of the SOHSOW subjects in both instructed and preferred positions in the BMW seat had a
different behavior with respect to the kinetic model as compared to the other two seats.
The HJC deflections were consistently larger than that estimated from the kinetic model
(Figures 44’to 46). I

HJC Force VS Deflection BMW( Sedan)
Average of HJC deflections of all SOHSOW subjects in instructed and preferred positions

0 50 100 1 50 200 250 300 350 400 450 500 550 600

 

 

 

 

*Kinetic Model

—I- Average of SOHSOW
.30 . , #. A . , W i 77777 with error bars

 

 

 

i.
o
l
l
l
l
l
l
l

SOHSOW
.50 . L .. ,7 7 -,_,,

JO] ,, i 7 7 77.77.- 7 ,7,,,

 

 

e———— Increasing vertical deflection(mm)

 

-90
Force(N)

Figure 44: HJC deflections averaged for all SOHSOW subjects in BMW seat were found to be higher
than that predicted by the kinetic model and was below the force deflection curve.

59

Increasing vertical deflection(mm)

50

100

150

HJC Force VS Deflection (SLK seat)
Average of HJC deflections of all 50H50W subjects in instructed and preferred positions

200

250 300 350 400 450

550 600

 

 

40‘

 

50H50W

 

*Klnetic Model

+Average of 50H50W
wrth error bars

 

 

 

 

Force(N)

Figure 45: HJC deflections averaged for all 50H50W subjects in SLK seat were close in comparison
to the force deflection curve of kinetic model.

Increasing vertical deflection(mm)

O

50

100

150

HJC Force Vs. Deflection-Tahoe(SUV)
Average of HJC deflections of all 50H50W subjects in instructed and preferred positions

200

250 300 350 400 450 500

550 600

 

d
O

O

 

d
0

~20 1

~404

.50 .

£01

.70 .

 

50H50W

*Kinetic Model

-O- Avearage oi '
50H50W with error I
bars

 

 

 

 

Force(N)

Figure 46: HJ C deflections averaged for all 50H50W subjects in Tahoe seat were close in comparison
to the force deflection curve of kinetic model.

60

CU

Si)

0111

 

 

The BMW seat had prominent, firm seat pan bolsters as compared to the other two
seats and this was thought to be the reason for the different behavior. It was suspected
that the prominent seat pan bolsters on the BMW seat pan, did not allow full contact of
the manikin’s butt thigh segment. Investigation with pressure mapping was performed to
see how the pressure exerted by the ASPECT Butt Thigh (ABT) segment varied among
the seat pans of the three seats (Figures 47 to 49). It can be seen in the figures 47 to 49
that unlike the other two seats, there is a gap in the pressure contours in the elliptically
marked region (buttocks region) for the BMW seat representing lack of contact between
manikin and seat. It was found that the pressure was evenly distributed on the central and
bolster regions of the SLK and Tahoe seat pans while on the BMW seat pan, the pressure
was uneven on central and bolster regions and a part of pressure was concentrated on the
bolsters. This uneven pressure distribution did not allow the (ABT) segment to come
fully in contact with the central portion of the seat pan thereby restricting the vertical
motion of the manikin; which resulted in a kinetic model that would produce a deflection
curve that may be offset higher (less deflection) than actually would occur with a
50H50W occupant.

Thus the kinetic model for the BMW seat produced a force deflection curve based

on the data from the manikin that did not precisely represent a mid-male loading.

61

Front of the seat

 

 

 

 

 

l Pressure on bolster region J

 

Figure 47 : Pressure distribution on BMW seat pan due to ABT loading of 461N (refer section 3.2). A
considerable amount of pressure is distributed on the bolsters. The BMW seat with prominent seat
pan bolsters is seen in the right.

 

 

. 1’ U“

    

 

 

 

 

 

Pressure on bolster region

 

 

 

 

Figure 48: Pressure distribution on Tahoe seat pan due to ABT loading of 461N (refer section 3.2).
Amount of pressure distributed on the bolsters is less compared to that in BMW (Figure 47). Tahoe
seat is seen on the right.

62

 

 

 

 

 

 

Pressure on bolster region

 

 

Figure 49: Pressure distribution on SLK seat pan due to ABT loading of 461N (refer section 3.2).
Amount of pressure distributed on the bolsters is less compared to that in BMW (Figure 47). SLK

seat is seen on the right.

In the process for comparing the calculated HJC deflections to those from the kinetic
model, the following aspects were considered. The HJC deflections were calculated using
the HJC locations obtained from both the method used by Gutowski [12] and Bush-
Gutowski method. Because the Bush-Gutowski method is more recent and is developed
for seated environment, all the comparisons for HJC deflections were made using data
obtained from Bush-Gutowski’s method of calculating HJC. The HJC deflection data was
compared with the HJC force-deflection plots obtained using the Radcliffe’s kinetic
model [6] for each of the three seats.

First, the deflections for each category (50H50W, 95H5W, 95H95W) of subjects in
both the instructed and preferred positions were plotted for all three seats and the HJC
deflection was compared with that predicted by the kinetic model. The HJC deflection

estimated by the kinetic model was read directly from the load deflection curve. As a

63

representation of all comparison plots, only the comparison plots of each category in one

of the three seats are discussed next. The data for all the plots can be found in Appendix.

5.4.1 Comparison of HJC deflections for 50H50W male Stgects.

Vertical deflection(mm)

HJC Force VS Deflection (SLK Sports)50H50W

Bush-Gutowski Method

 

 

 

250 300 350 400 450 500
F F F 7 ,_ I 7 +klnetic Model
8 a 50wa Instructed
L 7 M ‘ A _ __ . .2, ”L, position
6 6) —O— 50wa Preferred
* r r ,_ _, I "r ‘ m "r Position

 

 

 

 

 

Force(N)

Figure 50: HJC Force vs. Deflection for 50H50W male subjects in SLK seat.

The graph in Figure 50 shows the HJC vertical deflection in preferred and

instructed positions for 50H50W male occupants seated in SLK seat. The X- coordinates

on the graph represent the 54.3% of body weight (Bush [9]), which is the loading under

the buttocks of the occupant. The Y-coordinates are the HJ C deflection calculated from

the experimental data. It can be seen that the HJC deflection varied to some extent from

that predicted by the kinetic model. The HJC deflections calculated were consistent

within two trials of the same position for a particular subject in a particular seat (Table

11). Also it was observed that, in most of the trials the HJC deflection in the preferred

64

position was higher than that in the instructed position. There were no large differences in

the HJC deflections between the two positions.

HJC vertical deflection for 50H50W males in SLK
Instructed Position Preferred Position
Subject HJC vertical HJC vertical HJC vertical Deflection HJC
No(TriaI No) Deflection Deflection (Kinematic model- newtons)(54.3
Directly from the chart) °/oot bodyweight)

11 42

1 35 34
35 34 '

 

 

Table 11: HJC deflection comparison between preferred and instructed position for 50H50W male
subjects in SLK seat.

In the preferred position, most of the subjects slid forward in the seat pan with more
recline of the seat back. This movement shifted the HJC more anterior and distal (forward
and down) with respect to the HJ C in the instructed position (Figure 51) and subsequently
increased the HJC deflection in the preferred position (Table 11). The HJC deflections for
subjects 7 and 8 differed by around 10 tol7 mm in preferred position as those subjects

choose a comparatively forward position in the seat pan.

65

 

 

 

 

 

HJC in instructed position

 

 

HJC in preferred

position
\ .

 

 

 

 

 

 

Figure 51: A magnified view of HJC locations in preferred and instructed positions. HJC in
preferred position had a trend of being anterior and distal (forward and down) with respect to HJC
locations' In instructed position.

5.4.2 Comparison of HJC deflections for 95H95W male subjects.

i-UCForceVSDeflectimTahodSUV)
Blah-WW

O 50 1(1) 150 2C!) 250 300 $0 4'!) 450 500 560 600

d
O

 

 

 

 

o —T 'T
+KineliclVbdel

-10 77 7 7 7 7
A -I-%l-$W|rstrmted
Em«~ " A Position
gen-g . 7 7 0 WW
‘6 ..
.240..- 7 3 ~ g
g . ~u—|m5maw
3601 7 7 7 7 '7 -.~
2 £0 9 o
:
gm4fi 7 7 77° 77

.70.- 7

&47 7

-90

 

 

 

 

 

 

Figure 52: HJC force Vs. Deflection in preferred and instructed position for 95H95W male subjects

in Tahoe seat.

MN)

66

The graph in Figure 52 shows the HJC deflection in preferred and instructed
positions for tall and heavy males (95H95W) seated in the Tahoe seat. It can be seen that
the kinetic model force deflection curve, based on 50th percentile manikin data does not
extend to accommodate the HJC force and deflection for 95H95W category. It was
proposed in the study by Bush and Macklem [3] that the HJC deflection for 95H95W
category could be predicted by extrapolating the force deflection curve to reach loading
values for large men. The force deflection curve for each seat was linearly extrapolated
from of last two points of the curve till 550 N of H] C force to compare the HJC
deflections of 95H95W subjects. The HJC deflection calculations for tall and heavy
subjects in the present study were near the extended force deflection curve and supported

the proposition by Bush and Macklem [3].

HJC vertical deflection for 95H95W males in Tahoe

Preferred Position
Subject HJC vertical HJC vertical HJC vertical HJC
No(T rial No) Deflection Deflection Deflection force(newtons)(54.3
cyoOf

1

 

Table 12: HJC deflection comparison between preferred and instructed position for 95H95W male
subjects in Tahoe seat.

67

The HJC deflections calculated were consistent within two trials of the same
position for a particular subject in a particular seat (Table 12). Similar to the 50H50W
category subjects, it was observed that, in most of the trials the HJC deflection in the
preferred position was higher than that in the instructed position, without any large

deviations in the HJC deflections between the two positions.

5.4.3 Comparison of HJC deflections for 95H5W male subiects.

HJC Force VS Deflection Tahoe(SUV)95H5W
Bush-Gutowski Method

 

50 100 150 200 250 300 350 400 450 500

O

 

a
O

 

 

O
.1
fi
.4

+ Kinetic Model

d
O
|

m
o
|

—I- 95H5W Instructed

 

 

 

 

30 . _ DOSiiiOD
.40 3 o 95H5W Preferred
- Position
.50 1 7
I
.0 . £38

*— lncreasing vertical deflection(mm)
ON I 0
O

53 8

 

 

 

Force(N)

Figure 53: HJ C force Vs. Deflection in preferred and instructed position for 95H5W male subjects in
Tahoe seat.

The graph in Figure 53 shows the HJC deflection for tall and lightweight males
(95H5W) in preferred and instructed positions seated in the Tahoe seat. It was observed
that the difference between the HJC deflection in the preferred and instructed positions
was small in most of the trials for all three seats (refer Appendix C). Also from the data

of HJC deflection in Table 13 and Figure 45 it can be observed that there is large

68

deviation between the calculated HJC deflection and that predicted using the kinetic
model. A large deviation in PUC deflections was observed consistently for the trials of
95H5W category subjects in all three seats. This observation lead to the conclusion that
the force deflection curve obtained from the kinetic model needed a correction to

reasonably predict the HJC deflections for occupants in 95H5W category.

HJC vertical deflection for 95H5W males in Tahoe (SUV)
Instructed Position Preferred Position
Subject HJC vertical HJC vertical HJC vertical Deflection HJC
No(Tria| No) Deflection Deflection (Kinematic model- newtons
Directly from the chart) 54.3 %of

bodyweight)

36 309
36

35
35

45
45

46
45

 

Table 13: HJC deflection comparison between preferred and instructed position for 95H5W male
subjects in Tahoe seat.

69

 

§_5 Comparison with Bush-Macklemflmffset curves

In previous study by Bush and Macklem [3], an equation of a line offset to the force
deflection curve of the Radcliffe kinetic model [6] was developed to reasonably predict
the HJC deflection of male occupants of 95H5W group and of groups between the
95H5W and 50H50W. Bush and Macklem also proposed that the HJC deflections for the
95H95W male occupants would lie on the extended force deflection curve produced from
the kinetic model.

To verify these propositions, the offset line equations that depended on the seat pan
stiffness of each particular seat were developed for all three seats. The generalized offset
equation developed by Bush and Macklem [3] was used to get the offset equation for
seats in the present study. When using this method, the offset equation for a particular
seat depended only on the force deflection data from the kinetic model and was
independent of the calculated HJ C deflections.

The HJC deflections for each category were averaged over all subjects in all trials
and between instructed and preferred positions to get one average HJC deflection
corresponding to each anthropometric category. The HJC forces for all subjects in each
category were averaged and a single H] C force corresponding to each category was
calculated. Deviations from the averaged value were represented by a standard deviation
of -_+_ l and were plotted as the error bands for each category. The force deflection curve
was linearly extrapolated from last two points of the curve till 550 N of HJC force.

The results for each seat are discussed in the following text.

70

 

Figure 54: A plot showing offset line for SLK seat along with the extrapolated force deflection curve

HJC Force VS Deflection (SLK seat)

All categories- Average of Instructed And Preferred Position With Offset Line

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

and averaged HJ C deflections for each category with error bars of i 1 standard deviation.

It can be observed from the graph in Figure 54 that for the SLK seat the HJC

 

o 50 100 150 200 250 300 350 400 450 500 550 600
10
o 7 7 7
-1o 4 *Kinetic Model
E .20 . 7 7 +Average of 50H50W
E
E: -30 j 7 g g 7 7 + Average of 95H5W
‘g 40 50H50W 0 Average of 95H95W
g 95H95W —:— Offset Line
a -50 ~ 7 7 r.—
g ”/7. -i- Extrapolated Force- r:
o .60 7.7 Deflection Curve ‘
70 « . " -
Extensron of I
-80 «» - . kinetic model
Offset Line
-90
Force(N)

deflection error bands for the 95H5W category intersect with the offset line meaning the
offset line predicted the HJC deflections within the error range for the 95H5W category.
Also the offset line is just below the error range of HJC deflections for 50H50W category
meaning the offset line did not predict the HJC deflection for 50H50W category in this
seat. The extrapolated force-deflection line intersects with the error bands of 95H95W
category meaning that the extrapolated line predicted the HJC deflection for 95H95W

category.

71

HJC Force Vs. Deflection-Tahoe(SUV)
All categories- Average of Instructed And Preferred Position With Offset Line

 

0 50 100 1 50 200 250 300 350 400 450 500 550 600

 

 

 

 

 

 

 

 

10
0 r w . r . . T r 4 . 4 -l'- Offset _prediction_Macklem
-10 L 7 7 7 7 7 7 7 7 7 7
+Avearaoe of 50H50W
E .20. ‘ 7 7 fl 7 7 -7 7 g
E 0 Avearage of 95H5W
g -30 1r 7 7 ~ 7 4 7 .7 7 7 7
‘5 50H50W
.5 7 7 7 7 7 7 7 7 7 fl # % -0*Averageot95H95w
.-
O
“O
3 -so -7 ~ # 7 7 — 77 7 #- *— 95H95W" . *Kinetic m
E 95H5W , z"
0 7 7 7 7.2:,» 7 7 777 - "4
, :/‘ -O- Extrapolted Force

-70 fl , , _ , 7 7, 7, 7, ' 7 ,7 7 p 7 7, ‘ Deflection curve

0” W»? 7 77 7 , 7 ,7 7 7 .77 7 7 777 7,

~90

 

Force(N)

Figure 55: A plot showing ofl’set line for Tahoe seat along with the extrapolated force deflection
curve and averaged HJ C deflections for each category with error bars of i 1 standard deviation.

It can be observed from the graph in Figure 55 that for the Tahoe seat the HJC
deflection error bands for the 95H5W and 50H50W categories intersect with the offset
line meaning the offset line predicted the HJC deflections within the error range for both
the groups. The extrapolated force-deflection line also intersects with the error bands of
95H95W category meaning that the extrapolated line predicted the HJC deflection for

95H95W category.

72

HJC Force VS Deflection BMW( Sedan)
All categories- Average of Instructed And Preferred Position With Offset Line

 

o so 100 150 200 250 300 350 400 450 500 550 600
10 -°-Kinetic Model

 

 

 

 

Deflection Curve

 

 

 

 

E

.E,

c , .

,9 4° ‘ --- Offset Line

8

m

“c3 -20 <

E + Average of

8 .3 . 50H50W

'E

0 ~40 ,

a A Average of 95H5W

.% .50 .

8

5 -6o « - 0— Average of

E 95H5W 95H95W
.70 . 7

l -80 g -— Extrapoleted Force-
-90

 

 

F0"$3040

Figure 56: A plot showing offset line for BMW seat along with the extrapolated force deflection curve
and averaged HJC deflections for each category with error bars.

As discussed in section 5.3, the BMW seat pan did not fully contacted the ASPECT
butt thigh segment and therefore the force deflection curve developed from the kinetic
model had a slope less than what it should had been. The offset line, which was based on
the force deflection curve, would be shifted downwards than seen in Figure 56, if proper
contact between the ABT and BMW seat pan had been established. Because of these
facts, the HJC deflections were found much larger than that predicted by the kinetic
model.

All the plots presented in this section were calculated based on Bush-Gutowski HJC
computation method. Another set of graphs was plotted with HJC deflections obtained
based on the method used by Gutowski and the Bush-Gutowski method together to study
the difference between HJC obtained using the two methods. The HJC computations with

these two methods are discussed below.

73

 

 

HJC Force Vs. Deflection-Tahoe(SUV)
All categories- Average of instructed And Preferred Position With Offset Line
HJC conputaions based on Bush-Gutowski method and method used in Gutowski's study

050100150200250300350400450500550600

 

10

 

O
i
i

 

-':- Offset_Line

8
T
l
l
l

+Aveerageofsu-lsaw

AWonSi-lsw
* ”5* " " 4f * ’ +Amasslasw T

 

-a-Averageotsu-60\N(Bwh
Method)

 

,7 777 +Awerageof$i£wwueh
I l
_*___J +Averageof$l~mwmuh
Method)

 

 

4wm6tmkrsstudy +Eadrmolatedforoe
.77.. 7 * - —-—~~- 7 ——* -* 7 deflectioncuve

 

 

<———— Vertical deflection(mm)
s a 3 :8 $ 8 2::
i
l
i

 

 

 

 

 

 

is

Force(N)

Figure 57:Comparison of HJC computations based on Bush-Gutowski method (legends in hollow)
and method used in Gutowski’s study (legends in solid) for Tahoe seat.

It can be seen from Figure 57 that the average HJC deflections computed based on
the method used in Gutowski’s study were larger than those computed based on Bush-
Gutowski’s method [12] for all three categories in Tahoe seat. A similar trend was
observed for all categories in all three seats (Figure 58 and 59). It can also be noticed that
the HJC deflections based on the method by Bush-Gutowski were closer to the force
deflection curve obtained from the kinetic model than those based on the method used in
Gutowski’s study. The HJC computations based on Bush-Gutowaski method better

represented the HJ C location because fewer assumptions were used for computing HJC.

74

Vertical deflection(mm)

39.9%“???

10

0

0
d
O

L

r

O
\l
O

is 3

0

HJC Force VS Deflection (SLK seat)

All categories- Average of Instructed And Preferred Position With Offset Line
HJC computaions based on Bush-Gutowski method and method used in Gutowski's study

50

100

1502m250300350400450500550600

 

 

A
17

 

 

 

 

 

 

 

 

 

‘ -+-Extrapolated Force

*KineticModel
+Averaoeof50H50w
+Averageot95H5W
O Averaged 95H95W
-°:—-OffaetLine
—B—Averegeof50H50W(Bueh-
GmwskiMethod)

+Averege of 95H5W(Bueh-
Gutowski Method)
Deflection Curve

—0— Average of 96H95W(8ush—
Gutowski method)

 

 

 

PM")

Figure 58:Comparison of HJ C computations based on Bush-Gutowski method (legends in hollow)
and method used in Gutowski’s study (legends in solid) for SLK seat.

<——-—— Vertical deflection(mm)

3838:";

2'8

if

50

HJC Force VS Deflection BMW( Sedan)

All categories- Average of Instructed And Preferred Position
HJC computaions based on Bush-Gutowski method and method used in Gutowski's study

100

150200250300350400450500550600

 

 

 

 

 

  

 

 

 

}?;1

 

 

*Kinetchodel

+Averageof50l-150w

A Avaageof95i-i5W

O Avuageof95H95w

 

 

 

 

 

7 _7 7 -D-Averageof50H§0W(Bush
95H5W
‘r " - ' ’ r " +Averageofw'i5wmush
HJC based on “M
i ‘ ’ WWW“ ~O-Averageof95i-l95W(Bush
Method)
PM")

Figure 59:Comparison of HJC computations based on Bush-Gutowski method (legends in hollow)
and method used in Gutowski’s study (legends in solid) for BMW seat.

75

6. CONCLUSIONS

If the H] C locations of various anthropometric categories of people in automotive

seats with varying stiffness could be predicted and put into mathematical form then it

 

would aid in better ergonomic design of automotive interior packages. The goal of the
present study was to verify and refine an existing method of HJC prediction.

In the previous study, Bush and Macklem [3] proposed offset equations dependent r I
upon the kinetic model. This approach was used for reasonably predicting the HJC

deflections of groups of males and females other than male occupants of average height L'

 

and weight. In the present study only the offset equations for males were examined by
experimentally calculating the HJC deflections of 15 male subjects and plotting them
with the offset equations for comparison.

The HJCs were computed using both the method used in Gutowski’s study and Bush-
Gutowski method. The first conclusion of the present study was that the HJC calculations
based on Bush-Gutowski’s[15] method gave a better prediction of H] C location than the
HJC calculations based on the method used in Gutowski’s study. All the following
conclusions made about the HJC location prediction for each of the three male categories
were based on the HJC computation using Bush-Gutowski method [15].

The averaged HJC deflections of tall and heavy males (95H95W) in the SLK and
Tahoe seats were close to the linearly extrapolated force deflection curve. The second
conclusion of this study was that the proposition by Bush and Macklem that the averaged
HJ C deflection for 95H95W male group can be predicted by extrapolating the force

deflection curve obtained from the kinetic model holds good for data in this study.

76

 

 

 

The averaged HJ C deflections for the mid-male (50H50W) subjects were close to
that predicted by the kinetic model, which supported the idea that the kinetic model alone
can reasonably predict the HJC deflections of mid-male occupants. The third conclusion
of this study was that only the force deflection curve from the kinetic model is sufficient
enough to predict the HJC deflections of males in the 50H50W category and that the
offset line is not needed for predicting HJC deflections of male occupants in this
category.

The tall and light male (95H5W) category had significantly larger HJC deflections
than those predicted by the force deflection curve. The H] C deflections for 95H5W males
were close to or nearly intersected the offset line for that particular seat. Thus it was
concluded that the offset equation for males developed by Bush and Macklem is able to
predict the averaged HJC deflection for tall and light males for two seats in the present
study. However the offset line equation is expected to predict the averaged HJC
deflections of male occupants ranging from tall and light (95H5W) to mid-males
(50H50W) and further study is necessary to verify this by testing occupants in that range.

In the preferred positions subjects regularly slid forward with more recline of the
back making the HJC shift anterior and distal (forward and down) with respect to that in
an instructed position where in they seated with their buttocks all the way back in the
seat. The shifts in HJC location from instructed to preferred position were only a few
millimeters causing the HJC deflection in preferred position to be consistently more by 5
to 15 mm than that in an instructed position.

It was observed from the HJC deflection data for the BMW seat that the 50th

percentile manikin does not conform to the seat cushion of stiff seats with prominent seat

77

—..r. auger; .A .'

 

pan bolsters and thus does not precisely represent the loading of 50H50W male occupant
for such seats.

The offset line equation developed by Bush and Macklem [3] was based on the HJC
location data for four seats in Gutowaski’s study. Thus the offset line equation was based
on HJC computations from the method used in Gutowski’s study that is less precise in
comparison with the Bush-Gutowski’s method and this provided a scope for

improvement in the offset line equation.

6.1 Future Work

 

The offset line for male occupants developed by Bush and Macklem [3] was able to
predict the HJC deflections for males in tall and light category but it needs to be verified
if the offset line can predict the HJC deflections for males in between the 95H5W(tall and
light) and 50H50W(average height and weight) categories.

As discussed in section 5.3 the BMW seat did not make sufficient contact with the
butt thigh segments of the ASPECT manikin. The experimental HJC deflection data,
which did not correspond with the kinetic model for the BMW seat, initiated a challenge
to investigate the applicability of ASPECT manikin to represent mid-male loading for
stiff seats with prominent seat pan bolsters.

Bush and Macklem in their study proposed offset curves for male as well as female
occupants, but in the present study only male offset equation was verified. The next steps
would include the verification of female offset equation by experimentalstudies for

female occupants.

78

 

APPENDIX A

SFS analysis —Phase I

79

Genoa-=95 eon—:22

€002.32: 805:2

 

 

 

Hoe: ........

 

520 pal.

 

785 .5 ii
fiuiei

 

 

 

1!.

 

\I/
Ill
’,

 

, /%

 

 

 

4

8m

 

 

 

sz 86".

canooaoo 5.: .o> echou 3.:

(lulu) nonsense

 

80

accrue:

 

32...:

82...:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:oaoocon 3.: .o> 8.6“. 3.:

 

81

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 75.0 F 70 .000 020— fl _ _ F _ _ 7 _ _
00-08 F 8.0 N F 00.008 _ F _ _ _ _ _ _
807050 . 807050 807500
00.0 SN 0 A- 0 2 00-000 F N 0 a F0 00.000 N F0 0_ F0 F 0.00
8&3 0 0 F. 0 2. 3.000, F F v N 00 00-000 F F F. N 00 807.7 ........
00-000 F 0 0 N 00 00.000 F 0.0 0 00 00-000 0 0,0 0 00 . 0.00
8.80 0 0 0 0 00 00.080 0 0 0 2. 00.08 F 070 0 00 30.0 BNII
8.08 F. 0 2 0 00 8.08 F 0 0 0 8 8&5 0 F0 000 .005 0.0 5:... 0.0..
8-0000 0» 0.00 00&F0F 00 ~00 00.0703 00 0 00 08+
8.0000 00 N00 00-800 5 000 00.020 5F 000 0.00
8&3 0 0 F. 0 F0 00.08 F 0 0 N 00 00-000 F 0.0 N 00
0.0.070: F. N N 0 0N 00.000 F. F0. F N 00.0: F 0 0 0. N «N 0.8
8&NF 0 0 0 SN 00.02 0 0N. N 0F 00.000 N N N. F.0F
8&0NN FN. 0NF 00-003 00. 0NF 00.00... 0 0.0. FNF 0.0F
8&8 N 0 0. 0.0 3&8 0 F 0. 070 00-000 0 0.0. 00
N F .08 0 0 0 0 o 8&8 0 0,0 0 0 8.08 N 0.0 0 0 . . . . 0.0
8.25 00< 78 £0 «teem 92 so .78 «1.26 93 F00 08 000 000 000 08 8F 0
0.07
00.0 .08 0(000
0 6.0 8.0 0<x
F2 "7
NF,0FF.. 00 0- N800 F0 0F? 0NF. N80
R00 F00 N<x0F00 000 N<x coaoocoofiiéFooFoufii
F2 “7 .2 ”7
F00FFN. 00b 80 0b 3.00. 03F. 80 8.0000. 00:. 80 F _ F
00 08 Fe 9 x 00 0N0. 00 0 x 00 0N0. F0 0 x g r _
2 “7 72 “7 s7 “7 o F 80030700 2 co 8.08.5802 088.585
€002.33: moo—=32 305:5 €32.82: 900302 30525 $092.39.: coat?! one—5:9 _
00 0F 0N F. 80 EE mmv u: :3 09.2 2 3.691
0970 057m 00:0 0720 00:0 072m :80 05 2E 903:sz 0F 8.0000070 02:8
3&80 8&9 ,N 0908 F 3.802
507050 807050 807050
0900.0 N 8RN NN0 8&00 0 880 000 8&0F 0 0080 000 00000 N8 00 0 0.0N 0. F0 39 0. :N 0.0: F FF F9.
8&00 0 FN00N F F0 8&00 N 0080 000 8&0: 00000 000 0:8 0N0 N00 «.9 0.00 89 «SN 8: 00 03.
6&00 0 F FN0N N00 8&8 0 0000N 000 8&0F F. F.0BN 80 F008 000 00... 0F F 0.00 00.: 88 05: 00 F8
0908 N 08F 000 8&3. N «009 08 5.0.5 0 9K? N3. 8FNF 3.0 00.0 .0 N 000 N2: 078 0.0: 0N 0:.
8&000 0000F 0N0 Fo&F.0 F. N000 8.. 6+0: 0 0000 N9. 0:0 000 F00 N 0 F8 00 t 0 8 #08 FN 000
8&8 m 00:: 000 8&8 F F89 00.0 8&F0F F FN0F K0 000: F8 8N Z: 0 00 00 0F FF0N 0 0NF 0N 000
8&F0F 0009 000 F0&F.00 3.03 000 Fo&0N 803 000 00:; N00 RF 00N :0 F3: 008 0 :2 N0 08
8&8 F F 8F 20 5&2 F 0009 000 8&8 F. 00F0F 000 FNmNF 80 2.0 0 N 0 00 00 0F 0 NVN 0 R: 0N 0R
8&0: F00: 0FN 6&9 0 FvFNF F.0N 5&8 0 80: N0N 0N0NF 00N N00. 0 AN 0 0N N0NF 0.08 NF: 0N F0N
8&03 0:0: 0FN 8&0F.N 000: SN 8&0: 000: 00N 88F EN 00 F. 000 F.0F 00 9 N08 0.0.1 0N 00F
8&00F 0003 00F F0&8F. RFNF 00F 6&000 0N0: 00F 0009 00F F00. 000 FNF 80 000N 08F 0N 0NF
F0&R F 809 00 8&F0 F 0003 N0 5&00 0 083 8 3:9 00 B 0. F. 00 0 0 AN 0 F. ANN 0 BF 0N F0
8&000 o 0 090000 0 0 8&80 0 0 0 0 8,0 00 00 007NF 0 8N 0.00: 0 0
«10:0. 7.3.: 3.3: «$35 :25: :EN: N195 :22: :75: 7.33.. 38:5.— 05 Fun in in .2 .3... .55.: use 58.5 9i an 5.
Foot: 8ch new 6ch Pm. AAAmFobos‘ — 7 85‘ _ Hi 24% -0 Fmom.

 

 

.900808520 0.008 .8 02.8.30 .58.: 0> 83F .58.: 07500 o. 2022.. 0am 0.< 0.0..»

 

 

82

 

 

 

30:: I’ll.
50050 DEN I
30.0 0.0 .II

fin Io:

 

 

 

 

 

 

 

 

 

 

 

 

couscoo 3.: .2. 09.0". 3.:

83

 

 

1.1
h
l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o F .80 .9000... F . . F . . F . .. F
80:00 F0008 F F F F F F F F
N<0F§0 tFN<0F§0 I
00.08 F 0F. 30 8.003 0.0 0.00 .02... l... 000
00.000 0 .0F N00 00-003 00 2.0 ..
8.000 0 0F. F.N0 0.0.00: 0.0 000 .050 uch. / . 0.8
00-00N 0 0 0 0.00 3.08N 0.0 0 00 .050 50 .......
8&NF N 0.0 0.00 00.0000 _00 0.00 08+ 0 00
5.000 N 0 0 0 NF. 00.000 0 N F. 0 00 00.000 N We 0 00 .
00-000 0 F F. 0 00 00.08 F 0 0 0 NF. 3.000 0 F00 0NF. 0 9.
00.000 N 0 0 0 F0 300 F 0N 0 00 5.000 0 0N 0.9. 0.00
00.000 N F. N 0 0N 8.08 N 0 F F00 00.0000 0F 0.00 .
00.000 N N F 0 «N 00.000 F 0.0 0 00 00-000 0 0.0 F. 00 . 0.0N
00-0FNN F.0. 00F 00-05 0 0 F. 0.8 8-0000 0.0. 0.NN
8.000 N N N- 0 0 8-000F F0- 0.0 6.0000 F... 0.0 0.9
9.0000 00 0.0 00-0NF.N 0.0 00 00.00F0 0.0 0.0
N035. 00< .00 50 0.1050. 05.. .00 50 «12.5 05. .00 £0 . . 4 . 0.0
80 8F. 000 SN 8F 0
.0F..NF 00F 880 0.9.
0 F0 0. 60. 9.x
.2 u.
00 0N0. NF F- N800 N 03.. 03.. N800
0m 8 m— 0 N<x NF 8 9 p N<x 60900800 uQAL .0) OObOl unu:
s. n. .2 u.
00.000N. 0N 0. 0:0 00.800. 00 0F. 0:0 F2000. 00.00. 0:0 F F 0 F
00000 00.0 x 0N 00F: FN N x F0000. 00 FN. x _ _ . _
s. u. .2 0 F2 0. op "8:50.200 3 co 8:00.580: mocwotEBm
880550 805...... 8005.05.80.05; 08:5: 825:0 2.50 80:5: 085:5.
000N N0 0 8.0 55 one "59.3 00:! 2 .509:
09.0 0.20 .050 02a 5:0 0.20 .000 05 22.. 0002280 0.. 8.000200 0>FFF8Q
8&8 0 0908 F 8&0: 3202
F9058 F9008... FN<0FEB
0908 0 00N0N 00F. 5&0F F. 880 000 8&3 F 0000.. 000 «0000 N00 0N0 0 0N 0.00 00.: 0F F0 0.0FN FFF F00
5&8 F 0N0F.N 00F. 808 0 0800 N00 F0&N0 0 808 0N0 850 08 00.0 0NN N00 BNF 0 t0 0,0FN 00 03.
8&0F F N0NNN 00F. 8&0N N 080N 000 8.0:. F 080 000 890 F00 F00 0 0F NN0 BNF 0 08 0.0NN 00 F8
8&F.0N 0000F 3. 5&2 0 000: 00F. 8&000 F0FNF 03. 000NF 03. 0N0 N0 0.00 00. 0F 0000 NNNN 0N 0FF.
8&0F F. 8:3 .00 5&3 0 0000. 8.. 5&000 0000 0F... 0000 09. 3.0 0 0 0.00 00.9 0 F00 0.0NN FN 000
0908 0 r.003 00F. F0&F0 F 82 F R0 5.05 N FF.0NF v00 08: F00 0NF. 0 NF 0.00 F.0NF 0.08 F.0NN 0N 000
8&00F N000F N00 8&00N 000NF F00 8&0F 0 0F0NF 0N0 0003 N00 000 00F NNF. 00.: 38 88 N0 90
3.08 0 803 000 8&NN v 0F0NF 000 5&00 N 88F N00 809 000 00N 0 t 0.00 0N. FF 0 NN0 0.00N 0N 00N
8.0%. 0 N003 F00 8&0F.0 0F0NF 00N 8&000 F00NF RN 08F 00N 00N 03 000 009 N FN0 000N 0N 8
8.08 0 0000F F.0N 808 0 0009 00N 5&2: F83 NFN 0NF0F 3N 000 N8 0.0N R0 0 90 FSN 0N 00F
8&3 N 88F 03 808 F. F000F 03 Fo&F.0N 0000F N0F 039 09 F0 F. F.F0 0.0N 0F.» 0000 F,N0N 0N 0NF
8.0.8 0 00F: 0... 8&NF F 0N3; 00 883 0009 00 N0N0F 00 00.0. 0.00 0... 03. 0 000 F.00N 0N 00
8&8 0 0 0 8&8 0 0 0 8.080 0 0 0 0 8.0 0.0 0.0 3. 0 F30 0N0N 0 0
«125 3.1:: .53... «£050. :55! :50... «$550. :55. :50: .080... .0805: 0.3 5n .3. an: .2 .5... .55.: 2.5.55.0. 9i F... ....
000:3 EEO new 00.80 Em AAAmFmbosF 00:! 3.1 ~00. FF.FI00.JI..F Salmllms. .000,
5.5 3500300000350 0.58 .2 038:8 5.8.-.. 0> 8.2 5.8+. 5300 2 0.0.02... 000 0.... 0.0-0

 

 

84

 

03:32

82...:

 

 

 

 

 

AAA0F0000<

 

 

 

 

02.8.30 .0... .2 830 .9:

05—03 5.502 33

 

85

APPENDIX-B

Manual measurements of H-point vertical and horizontal deflection -Phase I.

86

 

Table Bl: Seat A-Audi, manual measurements of H-point vertical deflection.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Distance from
H-point Load lab floodmm Verical Deflection(mm)

0 306 0
57 305 2
125 287 20
185 277 30
231 271 36
278 265 41
310 262 44
348 260 46
388 260 47
41 6 257 49
431 256 51
446 250 56
461 248 58

 

 

Table B2: Seat B-Ranger, manual measurements of H-point vertical deflection.

Load lab mm Verlcal mm
0 0
57 3
1 5
22
30
33

35
38
41
42
42
43
44

 

87

 

Table B3: Seat C-SLK, manual measurements of H-point vertical deflection.

Load lab mm Verical
O
57
1 25
1 85
231

278
310
348
388
416
431
446
461

 

Table B4: Seat D-Taboe (Cloth), manual measurements of H-point vertical deflection.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Distance from
H-point Load lab floor(mm) Verical Deflection(mm)

O 355 O
57 348 7
125 329 26
185 320 35
231 313 42
278 306 49
310 300 55
348 299 56
388 295 60
416 290 65
431 291 64
446 289 66
461 288 67

 

 

88

 

 

Table BS: Seat A-Audi, manual measurements of H-point horizontal deflection.

 

Distance from
Load seat mm Horizontal
0 87
57 70
125 62
1 85
231 66

278 66
310 64
348 61
388 61
416 59 h '
431 59
446 60
461 61

 

 

Table B 6: Seat C-SLK, manual measurements of H-point horizontal deflection.

Distance from
Load seat reference mm Horizontal
O 105
57 93
125
185 90
231
278 89
310 87
348 87
388 85
416 85
431 84
446 83
461 82

 

89

Table B 7: Seat D-Tahoe (Cloth), manual measurements of H-point horizoan deflection.

seat
load mm Horizontal
0 1 32
57 1 03
1 25 98
1 85 96
231 94

 

278 91
310 91
348 91
388 91
416 89 L
431 89 -
446 89
461 89

 

 

Table B 8: Seat E-Tahoe (Leather), manual measurements of H-point horizontal deflection.

Distance from
Load seat mm Horizontal

0 1 02
57 81

1 25 78
1 85 78
231 76
278 74
31 O 71
348 70
388 68
41 6 66
431 66
446 67
461 67

 

9O

APPENDIX -C
Experimental data-Phase III

91

-.‘s -111." w Ti
F.

 

Table C1: HJ C experimental data for SOHSOW category in BMW seat with HJC calculated using
method used by Gutowski.

HJC vertical detlectlon tor 50H50W In BMW sedan

~11.1c Instructed _HJC Prelemd "°' “a" '" 110.1

-265

-267
-255

-248

-292
-295

-243

-311

   

Table C 2: HJ C experimental data for 50H50W category in BMW seat with [U C calculated using
Bush-Gutowski method.

HJC vertical deflection tor 50H50W In BMW sedan

HJCPM W

 

92

Table C3: HJC experimental data for 50H50W category in SLK seat with HJC calculated using
method used by Gutowski.

 

HJC vertical deflection for 50H50W in SLK

~HJC Preferred

   

Table C 4: HJC experimental data for 50H50W category in SLK seat with HJC calculated using
Bush-Gutowski method.

vertical deflection tor 50H50W in

Back angi- m _1uc mama vertical vertical

welt (Kim-tie
(Bush-OM (anal-om
from

 

93

Table C5: HJC experimental data for 50H50W category in Tahoe seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection tor 50H50W in Tahoe

-HJC Vertical vented

(Gutowski (Gutowski

   

Table C 6: HJ C experimental data for 50H50W category in Tahoe seat with HJC calculated using
Bush-Gutowski method.

HJC vertical deflection tor 50H50W In Tahoe

me _HJC Preterred vertical vertical

(Kinematic
(Bush-Gutowski (Bush-Gutowski
1mm

42
40 50

42
7O 42

45
45

45
45

 

94

Table C7: HJC experimental data for 95H5W category in BMW seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection tor 95H5W in BMW sedan

Beck angle Back angle _HJC vertlcal vertical
buck buck
rid No) (Gutowski

 

F1.

 

 

Table C 8: HJC experimental data for 95H5W category in BMW seat with HJ C calculated using
Bush-Gutowski method.

HJC vertical deflection for 95H5W ln BMW

Back angle Back ande ,HJC HJC Preferred vertical vertical vertlcai

buck buck
No) (Bush-Gutowski (Bush-Gutowski (Khan-tic

24

 

95

Table C9: HJC experimental data for 95H5W category in SLK seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection for 95st in SLK

Back an’e Back an” vertical vertical
| I | I _HJC instructed AHJC Preterred
rial No) (Gutmeld (Gutowski

 

 

 

Table C 10: H.) C experimental data for 95H5W category in SLK seat with HJ C calculated using
Bush-Gutowski method.

HJC vertical deflection tor

"HJC _HJC Prelerred vertical vertical

(Kinematic
(anathema (Bush-cm:
tron:

 

96

Table C 11: [U C experimental data for 95H5W category in Tahoe seat with HJC calculated using
method used by Gutowski.

WerflcaldeflecfimeIQSHSWlnTahoe

Beckerfle Backside _l-lJCPrelerred venicd
_HJClnetructed

 

 

 

Table C 12: HJC experimental data for 95H5W category in Tahoe seat with HJC calculated using
Bush-Gutowski method.

HJC vertical deflection for 95st in Tahoe

Decimal-fie Maude “m HJC Prelerred vertical
buck buck
No) (Bush-Gutowski

 

97

Table C 13: HJC experimental data for 95H95W category in BMW seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection tor 95H95W in BMW sedan

-""° _HJC Preferred vertical vertical vertical

(Bush-Gutowski (Bush-Gutowski (Kinematic

68 35
70

37

Deviation

 

Table C 14: HJ C experimental data for 95H95W category in BMW seat with HJC calculated using
Bush-Gutowski method.

HJC vertical deflection tor 95H95W in BMW

 

98

 

Table C 15: HJ C experimental data for 95H95W category in SLK seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection for 95H95W In SLK

Back-Ive Baclranfie vertical
I I I ' JiJClnetr-ucted _HJCPreierred

N0)

 

 

.‘rv_
a! v

Table C 16: HJC experimental data for 95H95W category in SLK seat with HJC calculated using
Bush-Gutowski method.

HJC vertical deflection tor in

mm Beckengle -""° _l-lJCPrelerred
m

 

99

 

Table C 17 : HJC experimental data for 95H95W category in Tahoe seat with HJC calculated using
method used by Gutowski.

HJC vertical deflection tor 95H95W in T

[ml ""’" _HJClnetructed _HJCPMerred "'"a' "mu"

(Gutowski

Table C 18: HJ C experimental data for 95H95W category in Tahoe seat with [U C calculated using
Bush-Gutowski method.

HJC vertical tor 95H95W in Tahoe

vertical vertical

 

100

 

 

10.

ll.

12.

REFERENCES

. “Devices for Use in Defining and Measuring Vehicle Seating Accommodation”,

Society of Automotive Engineers Recommended Practice J 826.

Schneider, L., M. Reed, R. Roe, M. Manary, C. Flannagan, R. Hubbard, and G.
Rupp, “ASPECT: The Next Generation H-point Machine and Related Seat Design
and Measurement Tools”, Soc. of Auto. En gin. (SAE) Paper Number 1999-01-
0962, presented at the 1999 SAE International Congress, March 1999.

Bush, T.R., Macklem, W., “Methodology for the Development of Female and

Male Seat Functions for the Kinetic Model A Supplement to an Excel
Spreadsheets”, April 2002.

RAMSIS: 3D-CAD-ergonomics tool by Human Solutions.
http://www.human-solutions.com/produkte_ramsis_e.php

JACK: 3D interactive software environment for controlling articulated figures.
Center for Human Modeling and Simulation at the University of Pennsylvania,
available from Unigraphics Inc. http://www.cis.upenn.edu/~hms/jack.htmi

Radcliffe, N.J,“Kinetic Modeling And Seat Factors Relating To Manikin-Seat
Interaction In Automobiles” dissertation for the degree of MS, Department of
Mechanical Engineering, Michigan State University, 2000.

Hubbard, R., and C. Gedraitis, “Initial Measurements and Interpretations of Seat
Factors for the ASPECT Program,” Society of Automotive Engineers (SAE)
Paper Number 1999-01-0958, presented at the 1999 SAE International Congress,
March, 1999.

Human Pelvis [Online Image]: Available
www.peteresser.de/skizzen/skiz_02.html

Bush, T. R., “Posture and Force Measures of Mid-Sized Men in Seated
Positions,” Thesis for Doctoral Degree, Michigan State University, 2000.

Package dimensions with J 826 manikin [ref] for the typical car segment-seating
environment. Johnson Controls Inc. December, 2003.

Devices for use in defining and measuring vehicle seating accommodation, SAE
Surface Vehicle Standard, SAE J l 100, revised May 1995.

Gutowski, P. E, “Influence of Automotive Seat Factors on Posture and
Applicability to Design Models”, Thesis for Masters Degree, Michigan State
University, 2000.

101

 

13.

14.

15.

l6.

17.

18.

University Committee on Research Involving Human Subjects (UCRIHS),
Michigan State University, IRB # 96-054, August 2003.

Human Pelvis [Online Image]: Available
http://www.hkin.educ.ubc.ca/36l/anatgifs/pubics.GIF

Bush, T.R., Gotowski, P. E., “An approach for hip joint center calculation for use
in seated postures”, Journal of Biomechanics 36 (2003) 1739-1743, March 2003.

National Health and Nutrition Examination Survey.

Seidel, G. K., Marchinda, D. M., Dijkers, M., Soutas-Little, R. W., “Hip Joint
Center Location from Palpable Bony Landmarks - A Cadaver Study,” Journal of
Biomechanics, Vol. 28, No. 8, pp. 995-998, 1995.

NATIK 1988 Anthropometric survey of US. army personnel: Summary statistics
interim report, NATICK/TR-89/027, March 1989.

102

 
    

ullwill[lllgtllnllm

3047