. . .. y . "11'" “"1. '1‘. ., ;’
.- ‘1 L1” 1:1 ‘2] III. 11 .1, 41.11712.) ”3 , n; ;..1si::=11:.:r: 1‘ ‘1: {.4 II: 3—443? I)! {‘ii‘ffi 1.1.41; '14:? 2‘?‘-‘I‘
. . w _. . '1-11'.' ~"- 6,; ' I.
‘1 I 2.51”“ 1”" . l1flk‘hl‘l‘w' 1., y ""413: A 1'. 6'3. :H E11' 19- 1&1?! Qg‘gfiné
, V .' '15. ‘v ' . ,
' ‘1». 1' 1 [17‘1“ 13:311.“. 1:57:44 II‘
£41“

. ."dgflf‘f I «4“
“II‘IJJ‘W {08.71.1“3:

III; .11..” f» "“ 14

  
  
    

  

I; gIla-[I111 ”all" . I’il

   
  
    
    
 
  
  
   
 
  

   

 
 
  
  

  

     
   
    
   
 
 
   

    
   
    
 
 
  
      

   

    

  
     
  
  

    

. In; 'I :11 III"F 'l’ M’III‘III‘ EEG?
"‘21'1'1-11" I"? I: 1:435; ‘:.11?3|11‘11}:ll'1;l‘.3 1‘24; [131.1% I:
C;: 111x111 ‘I‘III 35'2“ “a . 15‘)" III" ‘I (“114' II‘I‘ 11“ 2-711 1" 11'133
:IéI '* - ., III‘IW'I}, I: , 5,. "AIM I141;
I 3'? .‘EI ' ' ' ' II III: 1
I . [3 :11: ‘ ' ' .1 11:1; “91;“ i191. Ii 1"”1113 l 5.1%3531'4‘3'13?‘ ‘Q‘Ogiglgirflé J $3251,
I . ‘1‘1‘1: ’Iii"?Im- ‘I‘Hé‘ '1th . 11‘14 1’11‘11‘IIL"? 3: ’
I II “"I‘I " 311'“. “-1: ‘L1 '11 I ‘ ' “I '1? 51'
IIII "I.” ”15:;Zifig'gt'flfi ”“5 ‘1 ‘11}31 I11 111:!)
I 'R II‘II‘1'\:I'I ‘ I . ,
1.1937,; £161.] 41.I1 ‘1‘ ' .1 .1: . {LI-5

'11:} 41-31]? :15 '41

   

I; 4 . l' .
:f?“”‘ ‘1 W,“ {F 119?”?! 21-351
”‘13,:th-

. “33:4 4.11“ 1.115,]? 2:: ‘ ‘ 4%] . ;_‘
1.:1‘ ‘1‘”? '. 1 11;;- 1' 1’1
. I “‘I‘ 'II‘I" "I“ ‘31"? 1414’“ IIIIII‘II'! Ier «a:
' 2}“? .§ IKE“? {‘3 ii"
. ' . ‘1 1-111

     
   
 
   

 
 

1"“? “"1511‘K‘II‘EI‘MIII'IWM 131'
h I .114 - “I"I {‘1‘}! I"" '_ 1
1:: 1.1.; 111:;1Ih'1lx’I'I11‘1‘ 3g} -.

 
   
      
     
      
     
     
     
     
       
     
 

  
 

    

      
   
    

‘: f V
‘In .’. 1

I 'I’q’ .. . .
.‘9‘ . :2‘1‘5“ I”

21.!

a. i . .1115: $3. 1111‘.“ ‘IIJ l r I,- _
‘x .? III .IZ, b-zi' III III 54%:in I 2.11.4, £55 A
II I 1.]‘1 4.? “11"” 4L2. 1'1]! ' .3113?” kpf‘ 1‘ "
I 41.4. 3.51. '1‘ “I12". 1 4, "1‘ -; I11
‘ IIII I in} 513:

I
-

    
 
 

v
Ifltlié’

.,.
{13‘ 41. 5,?"
Q'. I

 

  

 

1..

go. -I 1

1 . I I| 1I is .I31.
4 - ,1

| ‘ :1 I" u "‘IIT’VI

, I7I.-I ”1.:

.'I I ”t: .-

., 1-1 ‘

’_g.q :4.
‘
...
-._

     

‘IZ :I‘.
1' ‘ A‘J 3‘11
I " {II III"
I II.{.‘.;' ‘1'»; :I
1: ”HI
;', 1.“ II‘
I' I
', 1|I'II “ I I.
II M 1.,14
I ‘l I
I I
I 7 ‘ I
I
I IN‘U‘I :1 '
1 w ‘
' ‘ LI‘II"
I“ f.‘ ‘
. :I
1- 1:1 I "H“ H. 5
r} :I '1‘ I-

'—
v . -¢'.'-‘
'. -T’C‘ ’.

 

, I. L
I ‘1 ‘ I v: ‘I;'
ly‘I-‘1lI'I‘I; . 11:11? -.
" I I ‘ I - .72,
I 1 :3; i W}; I; - .5211:
p {Ititg «.1141 ,, I; I
I 1‘11 “H‘s/.11“ 'Iti'. I" II. -1 - 5
I H‘II‘A: ' ‘ :11! z ,'
I M .1 “I: I.I'I_:,"‘I‘ III II I-
r ‘ III-{33:11:42 III I .1: ,4 ,
I I IIIIII‘III.I:-.-i II'I 1;}er
I ; ‘1‘.“I'JL‘ 4:1 I l ' ,t’: h1g1 I
‘ :‘ I; 'I‘ 15.: 11m»:
‘ 11 I 1 .1 It ‘ 1:11;
i ;

.
. .
A-
M"
5 ..

'
~»-

llllllllllHHIll“lull“llllilulhlmllllWIHIHUIHHI

3 1293 01397 9491

This is to certify that the

dissertation entitled

A Microscopic Inspection of the Operational
Aspects of Urban Interchanges

presented by

Paul Bennett Wellington Dorothy

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Civil Engineering

 

 

Major professor

 

MSU is an Affirmariw Action/Equal Opportunity Institution 0-12771

 

LIBRARY

Michigan State
University

 

 

 

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

 

MTE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1M ch‘RC/DabprGS—p.“

 

 

 

 

 

 

A MICROSCOPIC INSPECTION OF THE
OPERATIONAL ASPECTS OF URBAN INTERCHANGES

By

Paul Bennett Wellington Dorothy

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Civil and Environmental Engineering

1997

ABSTRACT
A Microscopic Inspection of the Operational Aspects of Urban Interchanges
By

Paul Bennett Wellington Dorothy

The Michigan Department of Transportation (MDOT) is considering the much
needed rehabilitation and upgrading of many interchanges found in urban environments.
Thus, Michigan State University (MSU) undertook an effort to evaluate the
appropriateness of an urban interchange geometric configuration, the Single Point Urban
Interchange (SPUI), as an alternative design to those presently used by MDOT. In
particular, the Michigan Urban Diamond Interchange (MU DI) and the traditional
diamond were investigated.

A field review was conducted to collect information about the geometric design,

signal operation, pedestrian control and pavement markings of SPUIs, as none

currently exist in Michigan. The field review showed that the design and
operation of SPUIs vary greatly from state to state. Thus, the SPUI and MUDI
designs were computer modeled to facilitate a comparison of their respective

operational characteristics. A traditional diamond was also modeled to generate a

frame of reference.

The results showed that the SPUI operation is adversely affected with the
addition of frontage roads. MUDI operation, in most situations, is superior to that
of either a SPUI and diamond interchange configuration. Also, there was less
migration of delay to downstream intersections with a MUDI configuration than

with either a SPUI or diamond. Finally, MUDI operation, in most situations, is

insensitive to the proximity of the closest downstream node, while the SPUI

operation is sensitive.

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to all those who provided the
inspiration, motivation and encouragement to complete this dissertation. Special thanks to
Dr. Thomas Maleck, my advisor, for having the courage to believe in me and the faith to
let me fight my own battles (even the ones I shouldn’t have fought). Your advice,
guidance and friendship have been invaluable to my professional development.

I am also grateful to Dr. William Taylor for his interest, advice and support of this
research. Thanks are extended to Dr. Francis McKelvey and Dr. Roy Erickson of my
advisory committee for their advice and suggestions. Further, I would like to
acknowledge the contribution of Kristy Miller, who put in many a long night with me
during the simulation modeling of this project.

I am grateful to the Michigan Department of Transportation for sponsoring the
research contained in this document. The advice and cooperation of the MDOT staff
helped me to achieve my goal. I would like to single out Laura Aylsworth-Bonzelet and
John Saller for their tremendous effort during the field review.

I would like to extend my gratitude to all my friends and colleagues at Michigan
State University for their friendship and support. Further, I would like to thank my
fiiends outside the university environment, especially “The Horsemen,” for helping me
keep my sanity during this endeavor.

I would close with a quote from two of the philosophers who have helped to
shape what I have become.

“He who knows when he can fight and when he cannot will be victorious.” —Sun Tzu

“Sweet!” —Cartman

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................... vii
LIST OF FIGURES ................................................................................................. ix
CHAPTER 1 INTRODUCTION ............................................................................ 1
1.1 Introduction .......................................................................................................... 1
1.2 Statement of the Problem ..................................................................................... 7
CHAPTER 2 OPERATION AND DESIGN OF THE MICHIGAN URBAN
DIAMOND INTERCHANGE (MUDI) ...................................................... 9
CHAPTER 3 OPERATION AND DESIGN OF THE SINGLE POINT
URBAN INTERCHANGE (SPUI) ............................................................. 11
CHAPTER 4 STATE OF THE PRACTICE ......................................................... 13
4.1 Literature Review ................................................................................................. 13
4.1.1 Single Point Urban Interchange (SPUI) ................................................ 13
4.1.2 Michigan Urban Diamond Interchange (MUDI) .................................. 18
4.2 AASHTO E-mail Survey ..................................................................................... 19
4.3 Telephone Survey ................................................................................................ 21
4.4 Conclusions from the State of the Practice .......................................................... 22
CHAPTER 5 FIELD REVIEW OF THE SPUI .................................................... 25
5.1 Geometric Design ................................................................................................ 26
5.2 Signal Operation .................................................................................................. 31
5.3 Pedestrian Control ................................................................................................ 33
5.4 Pavement Markings ............................................................................................. 34
5.5Conclusions from the Field Review ...................................................................... 36
CHAPTER 6 METHODOLOGY ........................................................................... 38
6.1 Selection of the Computer Model ........................................................................ 38
6.2 Network Configuration ........................................................................................ 39
6.3 Signal Operation .................................................................................................. 47
6.4 Variables and Measures of Effectiveness ............................................................ 51

CHAPTER 7 SIMULATION RESULTS .............................................................. 58

7.1 Interchange Performance without Frontage Roads .............................................. 59
7.2 Migration of Delay without Frontage Roads ....................................................... 67
7.3 Interchange Performance with Frontage Roads ................................................... 67
7.4 Migration of Delay with Frontage Roads ............................................................ 78
7.5 Sensitivity to Proximity of Closest Downstream Node ....................................... 78
CHAPTER 8 CONCLUSIONS ............................................................................... 90
8.1 Conclusions from the State of the Practice and Field Review ............................. 90
8.2 Conclusions from the Simulation Modeling ........................................................ 92
APPENDICES
APPENDIX A: TABULAR PRESENTATION OF SIMULATION

RESULTS ..................................................................................................... 96
APPENDIX B: GRAPHICAL PRESENTATION OF SIMULATION

RESULTS ..................................................................................................... 116
LIST OF REFERENCES ........................................................................................ 152

vi

Table A. 1:

Table A2:

Table A3:

Table A.4:

Table A5:

Table A6:

Table A.7:

Table A8:

Table A9:

LIST OF TABLES

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), without
Frontage Roads, 1.6 kilometers (1 mile), 5-lane Arterial .................. 96

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), without
Frontage Roads, 1.6 kilometers (1 mile), 7-lane Arterial .................. 97

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), without
Frontage Roads, 1.6 kilometers (1 mile), 5-lane Arterial .................. 98

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), without
Frontage Roads, 1.6 kilometers (1 mile), 7-lane Arterial .................. 99

Simulation Results for Modeling Scenarios Involving the
Traditional Diamond Interchange, without Frontage Roads,
1.6 kilometers (1 mile), 5-lane Arterial ............................................. 100

Simulation Results for Modeling Scenarios Involving the
Traditional Diamond Interchange, without Frontage Roads,
1.6 kilometers (1 mile), 7-lane Arterial ............................................. 101

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), with
Frontage Roads, 1.6 kilometers (1 mile), 5-lane Arterial .................. 102

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), with
Frontage Roads, 1.6 kilometers (1 mile), 7-lane Arterial .................. 103

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), with
Frontage Roads, 1.6 kilometers (1 mile), 5-lane Arterial .................. 104

Table A.10: Simulation Results for Modeling Scenarios Involving the

Single Point Urban Interchange (SPUI), with
Frontage Roads, 1.6 kilometers (1 mile), 7—lane Arterial .................. 105

Table A.11: Simulation Results for Modeling Scenarios Involving the

Traditional Diamond Interchange, with Frontage Roads,
1.6 kilometers (1 mile), 5-lane Arterial ............................................. 106

vii

Table A.12:

Table A. 1 3:

Table A.14:

Table A.15:

Table A. 16:

Table A.17:

Table A. 1 8:

Table A. 1 9:

Table A.20:

Simulation Results for Modeling Scenarios Involving the
Traditional Diamond Interchange, with Frontage Roads,
1.6 kilometers (1 mile), 7-lane Arterial ............................................. 107

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), without
Frontage Roads, 1.2 kilometers (3/4 mile), 5-lane Arterial ............... 108

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MU DI), without
Frontage Roads, 1.2 kilometers (3/4 mile), 7-lane Arterial ............... 109

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), without
Frontage Roads, 1.2 kilometers (3/4 mile), S-Iane Arterial ............... 110

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), without
Frontage Roads, 1.2 kilometers (3/4 mile), 7-lane Arterial ............... 111

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), without
Frontage Roads, 0.8 kilometers (1/2 mile), 5-lane Arterial ............... 112

Simulation Results for Modeling Scenarios Involving the
Michigan Urban Diamond Interchange (MUDI), without
Frontage Roads, 0.8 kilometers (1/2 mile), 7-lane Arterial ............... 113

Simulation Results for Modeling Scenarios Involving the
Single Point Urban Interchange (SPUI), without
Frontage Roads, 0.8 kilometers (1/2 mile), S-Iane Arterial ............... 114

Simulation Results for Modeling Scenarios Involving the

Single Point Urban Interchange (SPUI), without
Frontage Roads, 0.8 kilometers (1/2 mile), 7-lane Arterial ............... 115

viii

Figure 1.1:

Figure 1.2:

Figure 1.3:

Figure 1.4:

Figure 5.1:

Figure 5.2:

Figure 5.3:

Figure 5.4:

Figure 5.5:

Figure 5.6:

Figure 6.1:
Figure 6.2:

Figure 6.3:

Figure 6.4:

Figure 6.5:

Figure 6.6:
Figure 6.7:

Figure 6.8:

LIST OF FIGURES

Typical Single Point Urban Interchange (SPUI) Configuration

without Frontage Roads ..................................................................... 2
Typical Single Point Urban Interchange (SPUI) Configuration

with Frontage Roads .......................................................................... 3
Typical Michigan Urban Diamond Interchange (MUDI) Configuration

with Frontage Roads .......................................................................... 4
Typical Diamond Interchange Configuration with Frontage Roads ....... 5

SPUI with cross-road going over the freeway with all signal heads

located on a single overhead tubular beam. Minnesota .................... 27
Confused driver (car with lights on) stopped in middle of a SPUI

while traffic proceeds on either side .................................................. 28
U-tum lane accommodates large trucks .................................................. 28
Dual left-turning traffic on the off-ramp is backing up,

blocking right-turning traffic ............................................................. 30
Pavement marking overlap creates driver confirsion .............................. 35
“Runway” lighting to help illuminate the turning path.

Note buildup of debris ....................................................................... 35
Typical Diamond Interchange Configuration with Frontage Roads ....... 4O
Link/Node Diagram for Diamond Configuration .................................... 41
Typical Michigan Urban Diamond Interchange (MU DI)

Configuration with Frontage Roads ................................................... 42
Link/N ode Diagram for MUDI Configuration ........................................ 43
Typical Single Point Urban Interchange (SPUI) Configuration

with Frontage Roads .......................................................................... 44
Link/Node Diagram for SPUI Configuration .......................................... 45
Phasing Diagram for MUDI Configuration ............................................. 49
Phasing Diagram for MUDI Cross-overs ................................................ SO

ix

Figure 6.9: Phasing Diagram for SPUI Configuration without Frontage Roads ....... 52
Figure 6.10: Phasing Diagram for SPUI Configuration with Frontage Roads .......... 53
Figure 6.11: Phasing Diagram for Diamond Configuration ...................................... 54

Figure 7.1: Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 60

Figure 7.2: Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 61

Figure 7.3: Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 62

Figure 7.4: Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 64

Figure 7.5: Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 65

Figure 7.6: Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 66

Figure 7.7: Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 68

Figure 7.8: Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 69

Figure 7.9: Interchange Area Total Time for 70% Lefi Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 70

Figure 7.10: Interchange Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 72

Figure 7.11: Interchange Area Total Time for 30% Lefi Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 73

Figure 7.12: Interchange Area Total Time for 70% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 75

Figure 7.13: Interchange Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 76

Figure 7.14: Interchange Area Total Time for 30% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 77

Figure 7.15: Downstream Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-1ane Arterial ................................... 79

Figure 7.16: Downstream Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 80

Figure 7.17: Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 5-1ane Arterial, Varying Spacing Scenarios ........................... 82

Figure 7.18: Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 83

Figure 7.19: Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 84

Figure 7.20: Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 87

Figure 7.21: Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 7-1ane Arterial, Varying Spacing Scenarios ........................... 88

Figure B]: Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 116

Figure 8.2: Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 117

Figure 8.3: Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 118

Figure B.4: Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 119

Figure B.5: Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 120

Figure 3.6: Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 121

Figure 8.7: Downstream Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 122

xi

Figure 3.8: Downstream Area Total Time for 50% Left Turns, without Frontage

Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 123

Figure 3.9: Downstream Area Total Time for 30% Left Turns, without Frontage

Figure 3.10:

Figure 3.11:

Figure 3.12:

Figure B.l3:

Figure B.l4:

Figure B.15:

Figure B.l6:

Figure B.l7:

Figure B.l8:

Figure 8.19:

Figure B.20:

Figure B.21:

Figure B.22:

Roads, 1.6 kilometer (1 rrrile), 5-lane Arterial ................................... 124

Downstream Area Total Time for 70% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 125

Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 126

Downstream Area Total Time for 30% Left Turns, without Frontage
Roads, 1 mile, 1.6 kilometer (1 mile), 7—lane Arterial ....................... 127

Interchange Area Total Time for 70% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 128

Interchange Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 129

Interchange Area Total Time for 30% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 130

Interchange Area Total Time for 70% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 131

Interchange Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 132

Interchange Area Total Time for 30% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 133

Downstream Area Total Time for 70% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 134

Downstream Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 135

Downstream Area Total Time for 30% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 5-lane Arterial ................................... 136

Downstream Area Total Time for 70% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-1ane Arterial ................................... 137

xii

Figure B.23:

Figure B.24:

Figure B.25:

Figure B.26:

Figure B.27:

Figure B.28:

Figure 8.29:

Figure B.30:

Figure B.31:

Figure B.32:

Figure 3.33:

Figure 8.34:

Figure B.35:

Figure B.36:

Downstream Area Total Time for 50% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 138

Downstream Area Total Time for 30% Left Turns, with Frontage
Roads, 1.6 kilometer (1 mile), 7-lane Arterial ................................... 139

Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 140

Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 141

Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 142

Interchange Area Total Time for 70% Left Turns, without Frontage
Roads, 7-lane Arterial, Varying Spacing Scenarios ........................... 143

Interchange Area Total Time for 50% Left Turns, without Frontage
Roads, 7-lane Arterial, Varying Spacing Scenarios ........................... 144

Interchange Area Total Time for 30% Left Turns, without Frontage
Roads, 7-lane Arterial, Varying Spacing Scenarios ........................... 145

Downstream Area Total Time for 70% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 146

Downstream Area Total Time for 5 0% Left Turns, without Frontage
Roads, S-Iane Arterial, Varying Spacing Scenarios ........................... 147

Downstream Area Total Time for 30% Left Turns, without Frontage
Roads, 5-lane Arterial, Varying Spacing Scenarios ........................... 148

Downstream Area Total Time for 70% Left Turns, without Frontage
Roads, 7-lane Arterial, Varying Spacing Scenarios ........................... 149

Downstream Area Total Time for 50% Left Turns, without Frontage
Roads, 7-lane Arterial, Varying Spacing Scenarios ........................... 150

Downstream Area Total Time for 30% Left Turns, without Frontage
Roads, 7-1ane Arterial, Varying Spacing Scenarios ........................... 151

xiii

Chapter 1

INTRODUCTION

1.1 Introduction

As Michigan marks its 100th year of auto manufacturing, it should also be noted that
freeways in the Detroit area have been in service since 1942. The first 11 kilometers (7
miles) were constructed in 1942 to reduce the travel time of workers going from Detroit to
the World War II bomber plant at Willow Run. Many of the early interchanges preceded the
Interstate system and, thus, Interstate design standards. The Michigan Department of
Transportation (MDOT) is considering the much needed rehabilitation and upgrading of
many of these interchanges located in the urban environments. MDOT and Michigan State
University (MSU) undertook a joint effort to evaluate the appropriateness of an urban
interchange geometric configuration, the Single Point Urban Interchange (SPUI) (F igures
1.1 and 1.2), as an alternative design to those presently used by MDOT. In particular, the
Michigan Urban Diamond Interchange (MUDI) (Figure 1.3) and the traditional diamond

(Figure 1.4) were investigated.

Most of the pre-interstate freeway interchanges in the city of Detroit and its environs
are directional, partial Cloverleaf and diamond interchanges. Directional interchanges are
normally used to allow a freeway to interchange with another freeway. Conversely, partial
Cloverleaf interchanges are often used when a freeway interchanges traffic with a major
arterial, such as a state trunkline. The loop ramps of the partial Cloverleaf accommodate the

left-tuming movements, thus reducing conflicts on the major arterial. Finally, the simplest and

l

 

 

       

3.an 8 82g

38%

owflcem SEES coufinwmaou
93mg owsmnonBE 583
“Eon chmm Remnant :4 05me

 

 

3385
3:532

ll

 

 

 

   
 

\ seem s 86
mused owficccm ~23 comamcswccoo

SEE “sausages :35
is. swam some? ”3 seem

   
   

 

 

 

 

 

 

 

$38....“
2:332

  

 

 

 

 

 

 

 

 

 

Ans—mom 2 82V

xmvmom owficem £3» :omumnswccou
GOD—2V owcmnonoafi 28:85
:mfiD cmmfiomz 339? ”m..— 053m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

is

 

l

 

 

 

 

 

A

 

 

\ Exam 2 SE J
38% 09885

HEB cognawwcoo owcmnonoufi
98:35 3393. ”v.“ 8&5

 

 

 

 

 

 

E i
in

 

6
perhaps most common interchange used is the urban diamond. Diamond interchanges are used

to accommodate traffic from major city sheets and for freeways with parallel frontage roads.

The frontage roads usually are one-way sheets and run in the same direction as the fieeway.

The configuration shown in Figure 1.4 is an example of an urban diamond interchange
with a city street, freeway and parallel frontage roads. The at—grade intersections of the
frontage roads with the crossroad usually have stop-and-go trafiic signals. If the freeway is
below grade and the crossroad is at grade, then traflic exiting the freeway is going uphill and
traffic entering the freeway is going downhill which is beneficial for both movements. This
design of the diamond interchange allows traffic entering and exiting the freeway to do so at
relatively high speeds. Moreover, if the freeway is depressed, the at-grade intersections have
no sight restrictions typically created by freeway structures or differences in grades.
Unfortunately, this configuration has relatively low capacity because all of the tuming
movements occur at the intersections and left-turning vehicles have to yield to oncoming

traffic. Thus, there are several areas where traffic spillback may exceed the storage space.

The Michigan Department of Transportation (MDOT), borrowing fiom its indirect
left-turn strategy implemented for most at-grade urban boulevards, modified the traditional
urban diamond in an effort to increase the capacity. This modified diamond interchange
configuration will be referred to as the Michigan Urban Diamond Interchange (MUDI)
(Figure 1.3). This configuration evolved during the design and construction of freeways in the

early and mid 19603.

1.2 Statement of the Problem

There are no SPUIs in Michigan and most of the known SPUIs are located in
southern states. Although this interchange design has been around for over 25 years, it has
only recently become more prominent due to claims of its efficient operation. However, the
benefits of the SPUI have been the subject of some debate. As the popularity of these
interchanges increases in other areas of the country, they have been suggested as a logical
alternative to the MUDI. Thus, the Michigan Department of Transportation (MDOT)
commissioned Michigan State University to study the operational characteristics of the
SPUI for application in Michigan. In addition, since both the traditional diamond and MUDI
are widely used in urban areas of Michigan, the operational characteristics of these
interchange configurations were also of interest.

Since no SPUIs exist in Michigan, operational experience with this interchange
configuration was lacking. NHDOT raised several concerns regarding the operation of urban
interchanges in Michigan. These concems affecting urban interchange design included:
ability to progress the arterial cross-road, compatibility with fiontage roads, sensitivity to the
level (volume) of left-turning traffic, migration of delay to downstream intersections, need
to provide special case signing and pavement markings for positive guidance of drivers,
ability to accommodate pedestrians and operational efficiency at volume levels nearing
capacity. As a result, a field review was conducted to collect information about the
geometric design, signal operation, pedestrian control, and pavement markings of SPUIs.

While some evaluation of the SPUI design has been done in the past, the literature
review determined that nothing has been published with regard to the ability to progress the

arterial cross-road, compatibility with frontage roads, sensitivity to left-turning traffic,

8
migration of delay, or traffic levels nearing capacity. Additionally, while the operational

characteristics of a boulevard intersection have been studied and the results published, the
MUDI design, which is unique to Michigan, has never been formally studied and there is no
literature on the subject. Thus, the SPUI and MUDI designs were computer modeled to
facilitate a comparison of their respective operational characteristics. Furthermore, a
traditional diamond interchange was modeled to generate a frame of reference for the

results.

Chapter 2
OPERATION AND DESIGN OF THE

MICHIGAN URBAN DIAMOND INTERCHANGE (NIUDI)

An example of a MUDI is shown in Figrue 1.3. This configuration is an urban
diamond with left-turning vehicles being routed through separate left-turn structures known as
directional cross-overs. Thus, left-turning movements are prohibited at the intersection. As an
example, a driver traveling fi'om bottom to top along the arterial wanting to access the left
enhance ramp to the freeway would make a direct left-turning maneuver at a standard
diamond interchange. For the MUDI, the driver would turn right at the first frontage road,
travel to the directional cross over, make a U-turn through the cross over, have] from right to
left to the arterial, cross the arterial and access the enhance ramp, thus completing the desired
left turn. Similarly, a driver desiring to access a business adjacent to the service road in the
opposite direction would use the cross-overs to change direction and gain access. Evident in

these maneuvers is the associated increased have] distance.

The distance that the directional cross over sh'ucture is placed from the crossroad is a
function of the cycle length of the traffic signals and the speed of the movement. Properly
designed, if the left-tuming maneuver described above began from the start of green, it should
receive a green indication at both the cross over and the arterial. Thus, it does not have to stop

and the total travel time for this indirect left turn would equal approximately one-half of the

cycle length.

10
In urban areas, access to property abutting the freeway is often of such importance as

to require parallel frontage roads. In addition, Intelligent Transportation System (ITS)
shategies, such as ramp metering, firnction better with continuous frontage roads. However,
the intersections of the frontage roads with the cross-road usually require the use of traffic
signals. These closely spaced traffic signals may have a significant negative impact upon
the operation and capacity of the cross-road. This impact may also be influenced by the

cross-section (divided multilane vs. non—divided multilane) of the cross-road.

The addition of U-turn lanes to the cross over sh'uctures, as shown in Figure 1.3, is
cost-effective when there is a major development or other large athactor of haffic located in
the top left or bottom right quadrants of the interchange. For example, freeway haffic
traveling from left to right destined for a development in the top left quadrant would exit
normally at the ramp to the arterial but immediately use the U-turn sh'ucture to access the top
frontage road and, thus, the abutting property. This haffic never enters the intersection with
the arterial and, consequently, this shategy can significantly increase the capacity of the

intersection.

Chapter 3
OPERATION AND DESIGN OF THE

SINGLE POINT URBAN INTERCHANGE (SPUI)

An example of a SPUI without frontage roads is shown in Figure 1.1. The first SPUI
was completed in Clearwater, Florida on February 25, 1974 and was designed by Greiner
Engineering. Since that time several other states have adopted the design and have SPUI
interchanges in place.

The primary feature of the SPUI is that all through and left-turn maneuvers converge
at one signalized intersection area as opposed to two separate, closely spaced signals as with
the h'aditional diamond. In addition, opposing left-turn movements operate to the left of
each other, contrary to the right-hand rule. This allows for a relatively simple phasing
sequence to be used to control conflicting movements. This phasing sequence typically
consists of three phases accommodating: both crossroad through movements, both off-ramp
left-turn movements, and both crossroad left-turn movements. The right-tum movements are
usually allowed to free-flow. However, if frontage roads are present (Figure 1.2), there is a
need to add a fourth phase, resulting in a reduction in capacity of the other phases. In
addition, because of the physical size of many of the SPUIs, a relatively long clearance
interval is required between the phases.

A limitation in the SPUI design is that the close physical relationship of the bridge
abutrnents, roadway cross-sections, and offset left-turn paths may constrain the ability to

easily upgrade the design in the future. In addition, these limitations make it difficult to
11

12

utilize this design in an area where the crossroad and freeway intersect at a skew.
Furthermore, the horizontal alignment of the left-turn paths can affect the amount of right-

of-way needed.

Chapter 4

STATE OF THE PRACTICE

To determine the state of the practice with respect to Single Point Urban
Interchanges, a literature review, AASHTO e-mail survey and telephone survey were
conducted.

4.1 Literature Review

4.1.1 Single Point Urban Interchange (SPUI)

Much of the published literature on the design and operation of single point
interchanges was either generated from a study by Bonneson and Messer at the Texas
Transportation Institute (TTI) or referenced this effort. The TTI study was conducted over a
three year period and generated six papers (3, 4, 5, 6, 7, and 14) on the subject which were
published from 1989 to 1992. The objective of that study was to evaluate the appropriate
design and control shategy for five diamond interchange configurations. These
configurations were: conventional diamond, single point urban diamond, split diamond,
three-point diamond, and three-point stacked diamond.

The first paper published by Bonneson and Messer (3) was a one year status report
on the study. Of the six SPUIs identified for study, none had continuous, one-way fiontage
roads. Although the authors stated that it was possible to have a continuous frontage road,
they had a concern that with the additional signal phase required to accommodate the
frontage road traffic the capacity of the SPUI would be reduced. This concern is echoed in

most of the papers produced by this study. Furthermore, the authors stated a concem that the

13

14
merge capacity of the off-ramp right-tum movement into the arterial cross-road may be a

potential problem. Finally, in all cases there were no crosswalks provided for pedestrians to
use in crossing the arterial cross-road. The authors stated that the typical SPUI signal
phasing does not provide for a protected pedeshian phase to occur across the cross-road.

The authors also identified a wide variation in the geometry and signal strategies at
the studied SPUIs. The cross-section of the arterial cross-road varied from 5 to 10 lanes,
while the ramp-to-ramp widths varied from 24 to 51 meters (80 to 168 feet). The number of
signal heads present in the interchange area varied from 12 to 26. Furthermore, the total of
all phase change intervals varied from 15 to 40 seconds, while the total of the all-red
intervals varied from 3 to 28 seconds. The average daily traffic (ADT) on the arterial cross-
road varied from 28,000 to 52,000.

The second paper by Bonneson (4) presented the results of a sensitivity analysis
relating bridge size and clearance time to various geomehic features of SPUIs. The results
indicated that skew angle, arterial cross-road median width, and ramp-arterial clearance
distance have a significant impact on the length of the bridge and the duration of the
clearance interval.

The third paper by Bonneson (5) dealt with the headway and lost time at SPUIs. The
author collected headway data on 38,000 vehicles at three SPUIs and two at-grade
intersections (AGIs). The results of the analysis indicate that through movement headways
at SPUIs are larger than those at AGIs. However, the larger radii of the SPUI left-turn paths
were found to result in shorter headways for SPUI left-turning vehicles than for AGI left-

tuming vehicles. Finally, start-up lost time was found to be greater for a SPUI than for an

15
AGI. However, the author cautioned that these results were based on data from a small

number of sites and recommended that further study be done.

The fourth paper by Bonneson (6) dealt with the operational efficiency of a SPUI.
The author concludes the there may not be a significant difference in capacity and delay
between small and large SPUIs without frontage roads. While the large SPUIs require a
longer clearance interval resulting in greater lost time than the small SPUIs, this may be
partly offset by lower minimum headways and greater use of the yellow interval associated
with the large SPUI’s larger turning radii. Furthermore, the author states that although the
AGI has higher capacity per signal phase, the SPUI generally operates with a lower average
delay. However, the author concedes that the SPUI’s greater efficiency with respect to an
AGI can be athibuted to the elimination of the arterial cross-road through phase via a grade
separation.

The fifth paper by Bonneson (7) dealt with change intervals and lost time for SPUIs.
The author concludes that driver use of the yellow interval indicates that 2-3 seconds of the
yellow interval can be assumed available for every signal phase. The amount of yellow
interval use increased with an increase in approach speed. Furthermore, increased volrune
was found to cause left-turning drivers to use more of the yellow interval.

The final report from the TTI study (14) endorsed the SPUI as a safe and efficient
design alternative to a Tight Urban Diamond Interchange (TUDI) in reshicted urban
conditions. This report was used as a basis to develop the guidelines for geomehic design
for use with the AASHTO “Green Book” (1). While the original paper by Bonneson and
Messer (3) stated that five interchange configurations were to be studied, the final report

only compared the SPUI to either an AGI or a TUDI. The capacity analyses performed by

16
the authors show that a SPUI utilizing a 3 phase signal is slightly more efficient than a

TUDI, but the advantage diminishes as the size of the SPUI becomes larger. The addition of
a fourth phase to accommodate frontage roads resulted in a reduced capacity for the SPUI.

Of the 36 SPUIs studied, the dominant traffic signal control observed was isolated
haffic actuated operation Dual left turns were typically used on both approach legs of the
off-ramps and arterial cross-sheet.

The authors noted several impacts of the large open central control area of the SPUI.
These impacts included: a slight hesitancy exhibited by through drivers not promptly
moving into the intersection following the onset of the green phase; driver confusion in
finding the on-ramp left-tum slots; and, driver confusion resulting when drivers cannot
complete the left-turn maneuver and become hung-up in the intersection. Further areas of
concern with the SPUI design were noted as: (1on right-tum operations may not
have time to weave across a wide arterial into the SPUI’s left-turn bay unless the distance to
the downsheam node is sufficient; there is typically no provision for a protected arterial
cross-road pedeshian phase; high-quality illumination is desired when the freeway goes over
the interchange area; and, SPUIs cost more than TUDIs to construct.

A separate comparison of the SPUI and TUDI was conducted by Fowler (10). The
author used TRANSYT-7F to model both a SPUI and TUDI using twelve different volume
scenarios. The author concluded that the relative performance of both interchange
configurations are highly dependent on the characteristics of the intersecting movements,
such as: directional split, unbalanced turns, volume, and WC ratio. However, the author
states that the SPUI’s performance is less susceptible to unbalanced off-ramp turns and

directional split and, in smaller configurations, provides greater capacity.

17
Hawkes and Falini (11) also contended that the SPUI accommodates higher haffic

volumes than conventional diamond interchanges. In addition, the authors claimed that the
SPUI provides greater safety and requires less right-of-way. However, the authors presented
no empirical data to reinforce their conclusions.

Leisch (12) also contends that the SPUI is an effective design-«4f applied
appropriately. The author goes further to state that unfortunately the design has been
considered by some as an interchange type to solve all problems, which it does not. The
author suggests that the conditions for which the SPUI might be appropriate are where the
turning volumes are light to moderate and the right—of-way is reshictive. He also contends
that the efficiencies gained by fully utilizing the 3-phase signal are lost if more than one left-
tuming volume requires dual tuming lanes and if the cross-road has more than two through
lanes in each direction.

In a firrther study, Leisch, et. a1. (13), state that many articles and analyses presented
up to May 1989 appear to be based upon limited data, limited analyses and less than
equitable comparisons. The authors tested the concept of two designs, the SPUI and TUDI,
by modeling five real world locations using TRANSYT—7F. With the exception of one
location, the TUDI was detennined to be between 7 and 36 percent more efficient than the
SPUI. In all cases where the TUDI was more efficient, the volumes consisted of heavy
through traffic and heavy, unbalanced left-tums. The site where the SPUI was more efficient
had light cross-road through haffic and all left-turn movements were heavy.

The authors also state several concerns with the SPUI design. First, they state that
the SPUI design is impractical with the addition of fi'ontage roads due to the need for a

fourth phase. Second, they could find no conclusive observation of safety differences

18
between the SPUI and TUDI. Next, their experience showed that the SPUI generally costs

between $1 and $2 million more than a TUDI. Further, the authors state that the maximum
additional right-of-way for the TUDI would rarely exceed more than half an acre. Finally,
they state that the SPUI has little potential for expansion or modification without
reconstruction, while the TUDI can readily be expanded to accommodate more lanes.

Merritt (15) raised additional concems with the operation of the SPUI. The author
stated that drivers who are unfamiliar with the SPUI design may encounter some initial
difficulty. Thus, the SPUI design needs to rely heavily on guide signing, pavement
markings, and lane use signing for the necessary positive guidance of drivers. The author
further stated that the proximity of nearby intersections is a concern.

Abbey and Thurgood (2), based on field observations, consider the SPUI most
appropriate when: an AGI becomes congested to the extent it can no longer function and
there is too much turning demand to provide just an overpass; an existing TUDI is not
performing well because the turning haffic makes it impossible to get optimum functional
use from the coordinated haffic signals at the two TUDI intersections; right-of-way is too
tight to accommodate a traditional diamond or other type of interchange; and, the skew
angle of the intersecting routes is not excessive.

4.1.2 Michigan Urban Diamond Interchange (MUDI)

No literature exists specifically regarding the MUDI. However, the MUDI
configuration is similar to a conventional boulevard intersection with respect to geomeh'y,
signal operation and pavement markings. In a study by Dorothy, et. a1. (8), the operational
aspects of the Michigan design for divided highways were analyzed. The results of this

study showed that the boulevard arterial design utilizing indirect left-turn shategy and

19

signalized cross-overs was superior to both the five-lane and seven-lane center left-turn lane
arterial designs. In addition, the boulevard indirect left-turn shategy was superior to a
boulevard shategy which allowed direct left-turns at the major intersection. Thus,
accommodating left-turns at a major intersection has a greater negative effect on operational
efficiency than the adverse travel distance which results from the use of an indirect left-turn
shategy.

4.2 AASHTO E-Mail Survey

A survey was submitted by e-mail to each of the other 49 state deparhnents of
hansportation. The survey requested information on the design and operation of Single
Point Urban Interchanges (SPUI). 14 state DOTS responded: Arkansas, California, Indiana,
Iowa, Missouri, New Mexico, New York, North Dakota, Oklahoma, Pennsylvania, Texas,
Vermont, West Virginia and Wyoming. Of these, only Califomia, Indiana, Missouri, and
New Mexico have operating SPUIs. In addition, New York is presently designing their first
SPUI. None of the responding states with existing SPUTs reported having frontage roads as
part of the design.

Generally, the respondents reported that the major advantages of a SPUI
configuration with respect to other geomehic configurations are: that it requires the same or
less Right-of-Way, has less delay and user costs, is adaptable to frontage roads, requires
fewer signals, is easier to coordinate the traffic signals with the surrounding system, costs
less, has fewer conflict points, allows for U-turn movements, and, has superior aesthetics.
The responding states also stated that the major disadvantages of a SPUI configuration with
respect to other interchange designs are: it is not an optimal solution if adequate Right-of-

Way is available, it costs more, it has long or special bridge sh'uctures, signals are difficult

20

to mount, it has long clearance intervals, it has unbalanced haffic flows from the off ramps,
it is tough on pedeshians, it should not be considered where the Right-of-Way allows for the
consh'uction of a Partial-Cloverleaf interchange, it has less capacity than a Partial-
Cloverleaf, the downsheam intersections may conhol the flow, left-turn storage capacity on
the cross-road is critical, and, sight distance may be limited.

The responses received from different states varied widely. With respect to delay,
one state reported that delay decreased and another reported no noticeable increase in delay.
Accident rates were reported to be similar to diamond interchanges or having no noticeable
increase in accidents. One state reported that signing was more difficult and two other states
reported that they used conventional signing. One state reported that they used conventional
pavement markings, another state reported that pavement markings may be a problem, and a
third state reported that there is a need for extensive pavement markings. A SPUI was
reported to cost $2 to 4 million more than a conventional diamond, $8 to 12 million for
converting an existing diamond, and, the same as a conventional diamond. Finally, the
Right-of-Way requirements were reported to be similar to a tight diamond, to depend upon
the use of retaining walls, and, to be less than a conventional diamond.

The limited number of responses to the survey reshicted its usefulness for
comparison to the conditions found in Michigan. While maintenance of a SPUI was not a
problem for one state and was "little" problem for another state, snow plowing was not
considered, as none of the responding states with SPUIs are in a climate where snow
plowing would be anticipated to be a problem. In addition, Michigan hies to progress traffic

on most of its major arterials. However, only one state responded that they had a cross-road

21
with good progression, while the other states responding did not make an effort to progress

the cross-road haffic.
4.3 Telephone Survey

The review of the literature and the response to the e-mail survey, while helpful, had
significant inconsistencies and lacked information in key areas. A telephone survey was
subsequently conducted with some of the e-mail states and with several additional states’
Deparhnents of Transportation. The states called in the telephone survey were: Indiana,
Illinois, Minnesota, Florida, Arizona, Missouri, and, Texas. The objective of the phone
survey, in addition to collecting more information, was to locate the most appropriate sites
for a field review. Specifically, it was desired to observe the operation of SPUIS with
frontage roads, the progression of the cross-road haffic, and, the operation of SPUIS under
winter-time conditions.

The individuals having the greatest knowledge of the operations of the SPUIS were
sought out. Thus, most of the phone conversations were with the dishict traffic engineers.
Of the seven state DOTS telephoned, four gave sh'ong favorable recommendations on the
positive aspects of a SPUI. One state DOT could not recall its operation and had ambivalent
feelings. The remaining two state DOTS had unfavorable opinions.

One engineer responded that the SPUI was his preferred design. Conversely, another
engineer responded that the SPUI did not have a single advantage with respect to the design
and operation of a conventional tight diamond. Also on a negative note, another state traffic
engineer responded that when their first SPUI was open to haffic there was a great deal of

driver confusion which negatively impacted the operation of the interchange.

22

When attempting to narrow the search for appropriate field review Sites, it was
discovered that only two of the states had any experience operating a SPUI with frontage
roads. Additionally, only two engineers reported that they progressed the traffic on the
cross-road arterial. Most of the states reported that they rely solely on traffic actuated
signalization along the arterial.

The comments of the Minnesota DOT were of special interest since they have a
similar climate. The dishict haffic engineer in Duluth believed that a SPUI was easier to
operate than a conventional diamond interchange. In addition, he reported that pedeshians
did not have a problem and he knew of no winter time difficulties.

4.4 Conclusions from the State of the Practice

The literature review, e-mail survey and telephone survey pointed out several
aspects of SPUI design that would need to be addressed in the field review and the
simulation modeling. These aspects can best be presented by grouping them into several
topic areas: geomehic design, signal operation, pedeshian conh'ol, pavement markings, and
Simulation modeling.

Several inconsistencies in the geomehic design of SPUIS were discovered. The
studies by Bonneson and Messer (3) and Leisch, et. al. (13) raised the concern that the
operation of a SPUI may be adversely affected by the addition of continuous frontage roads
due to the need for a fourth Signal phase. However, the responses from the e-mail survey
listed the adaptability to fientage roads as one of the major advantages of the SPUI design.
The study by Messer and Bonneson (14) stated that dual left-turns were typically used on

both approach legs of the off-ramps and arterial cross-street. However, Leisch (12) contends

23
that the efficiencies gained by fully utilizing the 3-phase Signal are lost if more than one left-

turning volume requires dual turning lanes.

The Signal operation of SPUIS varied by location. Messer and Bonneson (14) studied
the operation of 36 SPUIS and observed the dominant traffic Signal control to be isolated
traffic actuated operation. This was reinforced by the results of the telephone survey in
which most of the states reported that they rely solely on haffic actuated signalization along
the arterial. However, most arterials in Michigan are operated in a progressed-coordinated
system. Only one state from the e-mail survey and two states from the telephone survey
stated that their agencies progressed the traffic on the arterial.

The reported ability to accommodate pedeshians varied. Bonneson and Messer (3)
reported that the typical SPUI Signal phasing does not provide for a protected pedeshian
phase to occur across the cross-road. However, the dishict engineer for Duluth reported that
pedeshians did not have a problem.

The reported need for pavement markings also varied. AS part of the e-mail survey,
one state reported that they used conventional pavement markings, another state reported
that pavement markings may be a problem, and a third state reported that there is a need for
extensive pavement markings. Menitt (15) stated that the SPUI design needs to rely heavily
on guide Signing, pavement markings, and lane use signing for the necessary positive
guidance of drivers.

The studies by Fowler (10) and Leisch, et. al. (13) used computer modeling to
compare the operation of a SPUI and a TUDI. However, both studies used TRANSYT-7F

which is a macroscopic model and is best suited to modeling large networks, not individual

24
intersections. In addition, the study by Fowler (10) only modeled 24 scenarios and the study

by Leisch, et. al. (13) only modeled ten scenarios.

Chapter 5
FIELD REVIEW OF THE SPUI

While the MUDI configuration can be compared to a boulevard intersection, the
SPUI configuration has no direct comparison. Based on information gathered through the e-
mail and telephone surveys, Sites were selected in several states for inclusion in the field
review. These Sites were located in Indiana, Illinois, Minnesota, Florida, Missouri and
Arizona

During a typical field review, the engineers and technicians responsible for the
operation of the SPUI interchange being studied were interviewed. These interviews
included a visit to the Site where the operation of the SPUI was discussed. If possible, plan
view drawings, Signing plans, aerial photographs, Signal timings, traffic volumes, in-house
studies, and, economic data pertaining to the SPUI in question were collected. In the field,
extensive photographs and video of the interchange were taken.

Based on the field review conducted between January 1996 and May 1996,
subjective observations can be made about the design and operation of a SPUI. These
observations are based upon the consensus of the team which conducted the field review.
The members were, in addition to the author, Dr. Thomas Maleck (Michigan State
University), Laura Aylsworth-Bonzelet (Michigan Deparhnent of Transportation), and John
Saller (Michigan Deparhnent of Transportation). These observations can best be presented
by grouping them into several topic areas: geomehic design, Si gnal operation, pedeshian
control, and pavement markings.

25

26
5.1 Geometric Design

The geomehic features of the SPUIS varied greatly from state to state. The
difference in designs was much greater than anticipated and this difference may explain
some of the inconsistencies in the responses to the e-mail and phone surveys.

The most Significant difference in design is between a SPUI with the cross-road
going over the freeway and a SPUI with the freeway going over the cross-road. The SPUIS
with the cross-road going over the freeway (Figure 5.1) were found to look and operate
more like a conventional signalized intersection. Because of this, driver confusion is
reduced. Conversely, Significant driver confusion was observed at interchanges utilizing the
cross-road under the freeway design. At times, vehicles became happed in the intersection
due to driver confusion, creating a dangerous Situation (Figure 5.2). In addition, routing the
freeway over the cross-road exposes the freeway and major traffic volume to preferential
icing in cold weather climates.

Another significant difference in design is related to the physical Size of the
interchange. Some of the newer SPUI designs include the provision of a dedicated lane to
permit a U-turn maneuver from the exit ramp back onto the entrance ramp (Figure 5.3).
These dedicated sh'uctures were located under the tailspans requiring the tailspans to be
much longer than normal. While the smaller designs can provide for most U-tums, this
dedicated lane is necessary to accommodate large h'ucks and to increase the speed of the
maneuver. Even at interchanges where this maneuver was prohibited, it was still observed to
occur regularly. The smaller designs were observed to function with reduced clearance
times. In addition, the Right-of-Way requirements are obviously much less with the smaller

design.

27

.839 3:53 3.355 29% a co 8:82
memo: Rama =m 53> 330on 06 Ho>o memo» 80.7308 59> 5mm ”mm oSwE

.... ”Eu... .4

a . .

 

28

 

Figure 5.2: Confused Driver (car with lights on) stopped in middle of a SPUI
while hafiic proceeds on either side.

‘fl .
ma"- - 7;."
'Q- ‘. A

fi.

 

Figure 5.3: U-turn lane accommodates large h'ucks.

29
The design of the sh'uctures varied from state to state. They are generally much

longer than those of conventional diamond interchanges. Some of the spans measured were
found to be greater than 146 meters (480 feet) in length. Often there are three spans of
nearly equal lengths. Some of the shuctures were noisy and the resulting booms could be
heard for several kilometers. This noise was reported to be the source of extensive
residential complaints. Because of the large widths and lengths, the road under the sh'uctures
was dark. Lighting was often provided under the Sh'uctures during the day and visibility at
locations that utilized light color bridge paints (e.g. sand or concrete) were noticeably better
than those with dark color bridge paints. These characteristics were not evident when the
cross-road went over the freeway.

The impact of continuous frontage roads on the overall operation of a SPUI was a
key area of interest It was explicitly desired to observe the operation of a SPUI with parallel
frontage roads whose intersections with the cross-road are signalized and accommodate
Significant through haffic. Two of the states visited were anticipated to have these type of
frontage roads based on responses from the e—mail and telephone surveys. However, these
frontage roads did not satisfy MDOT’S requirements. One of the state's frontage roads are
what would be considered to be ramps with private driveways. The other state had a
fi'ontage road that was a two-way road which did not appear to generate the desired through
traffic. Several of the dishict traffic engineers expressed strong opinions that providing for
continuous frontage roads with a SPUI is a poor design and negates the advantages of a
SPUI.

The geometry of the exit ramps often flared from one lane to three at the ramp

terminus. Of these three lanes, two were for left-turning haffic and one for right-taming

30
traffic. The right- and left-tuming lanes are separated by a large charmelized island. The dual

left-turning h'affic on the off-ramp backs up during peak periods. This blocks right-turning
traffic from exiting and locks up the ramp (Figure 5.4).

The geometry of the on-rarnps normally consisted of three lanes near the
interchange, which reduce down to one lane before entering the freeway. The initial three
lanes of the on-ramp are fed by two left—turn lanes, under Signal control, and a free-flow
right-turn lane. Several engineers commented that this merge results in a Sideswipe crash
problem. However, the crash reporting systems are structured in such a way that these
Sideswipe crashes are not referenced to the interchange. Thus, it is difficult to get a clear

picture of the crash experience of the interchange.

Hooper!

 

Figure 5.4: Dual left-turning traffic is backing up, blocking right-turning traffic.

3 1
Most of the SPUI designs, regardless of state, added several additional lanes to the

cross-road basic laneage at the interchange. A typical design would have a 6 lane cross-road
being widened to nine lanes at the interchange. The additional lanes are typically a right-turn
bay and provision for dual left-turn lanes for the on-ramp. In addition to the auxiliary lanes,
most of the cross-roads had raised, concrete medians ranging in width from 1.2 meters (4
feet) to 3.6 meters (12 feet).
5.2 Signal Operation

The operation and placement of haffic signals were of special interest. The MUDI
configuration operates similar to a conventional boulevard intersection. However, each
state's practice differed Significantly for the SPUI. The cycle lengths varied from 80 seconds
to 180 seconds. The SPUIS reviewed that had longer cycle lengths and usually had fully
actuated signal phases for all movements.

Of special interest was the ability to progress haffic on the cross-road. Two of the
SPUIS reviewed have a cross-road arterial which was part of a pre-timed progressed
shategy. While the interchange was operating well below capacity, it was obvious that
providing progression would not be a problem. These interchanges were the smaller designs
which result in Shorter clearance times and allows for a Shorter signal cycle. However, the
impact of the SPUI on intersections downsheam must be considered. Comments were made
to the effect that the SPUI dumps traffic on the downstream nodes causing a migration of
delay. This was hard to judge in the field as none of the SPUIS reviewed were operating near
their capacities.

Most of the SPUIS reviewed had a 3 phase Signal operation. The 3 phases were

usually: left-turn enhance ramp movements, left-turn exit ramp movements, and, cross-road

32
through movements. One state provided for a right-tum exit ramp green arrow during the

left-turn entrance ramp phase. Usually the exiting right turn was accommodated via a free-
flow, charmelized merge with the cross-road haffic. However, a skewed intersection affects
the operation of the signal phasing. At these locations, there are 4 Signal phases: first exit
ramp movement, opposing exit ramp movement, left-tum entrance ramp movements, and,
cross-road through movements. In addition, the Skew requires longer clearance times.

The placement of the traffic signal heads also varied greatly from state to state and
by geomehic design. In the case of a SPUI where the cross-road goes over the freeway, all
of the signal heads are located on a single overhead tubular beam (Figure 5.1). Thus, the 3
phase operation was analogous to a traditional at-grade intersection with a 3-phase Signal.
This design typically took less Right-Of-Way. This SPUI design was observed to function
very well, although the traffic volumes were not heavy. In the case of a SPUI where the
freeway goes over the cross-road, the Signal heads are mounted on the sh'ucture. However,
some states have post-mounted Signals located on traffic islands. In one interchange alone
there were 24 Signal heads. With this proliferation of Signal heads, it was possible to see
green, amber and red indicators at the same time depending on where one looked. In
addition, the signal heads when post-mounted were vulnerable to damage fi'om motorists
nmning into them.

The physical size of the interchange also affected the Signal operation. If the
intersection area is very large, longer clearance times are required for haffic to clear the
intersection prior to allowing the next phase. Additionally, the green Signal arrow for left-
tuming traffic was often canted to give the motorist a sense of direction in these large

intersection areas. Still, there was driver confusion resulting from the large distances needed

33
to clear the intersection (Figure 5.2). There were three common mistakes observed. The first

results when the lead car does not start on green because the driver is (presumably) confused
on which signal indication is theirs. The second results when a motorist entering the
intersection from the exit ramp on a green light has to drive through a red indication meant
for the cross-road. Vehicles were observed stopping in the middle of the interchange and
waiting for a green indication The third results when a motorist starts into the intersection
and becomes (presumably) confused about which path to follow due to the large Size of the
interchange or overlapping pavement markings.

5.3 Pedestrian Control

The ability to accommodate pedeshian movements varied greatly from site to Site.
Since the MUDI configuration is geomehically similar to a boulevard intersection, it is able
to accommodate pedeshians without problems.

When the SPUI was examined, many of the locations Simply had no pedeshian
movements to accommodate. Where pedeshians were present, it was not difficult for them
to move parallel to the cross-road and cross the ramp movements. However, with all
movements going through the center of the interchange and a Signal operation utilizing fully
traffic actuated phases, there is always traffic moving through the intersection. This makes it
difficult for pedeshians to cross the cross-road. In addition, the width of the cross-road,
often 6 to 8 lanes, makes it difficult for pedeshians to cross the cross-road. Often,
pedeshians would become happed on the concrete channelization of the cross-road when
attempting to cross. Some sites actually prohibited pedeshians from crossing. However, this
prohibition was often violated, as the only other opportunity to cross was at the next

intersection which was typically over 400 meters (quarter of a mile) away.

34
5.4 Pavement Markings

With the potential for snow as in Michigan, the need to rely heavily on with lane
markings is a concern. The MUDI has no lane markings that are unique when compared to a
standard intersection. However, for the most part the larger SPUIS have supplemental lane
markings to assist the motorist with the left-turn movement. The need for these pavement
markings is paramount when the SPUI is large. These pavement markings may overlap
creating driver confusion (Figure 5.5). In a skewed configuration, this overlap is more
pronounced and it can be confusing even to a driver familiar with the interchange. However,
the need for supplemental lane lines for the turning movement was not evident for the
locations where the cross-road went over the freeway or the interchange was small in size.

One location had lights placed in the pavement to help illuminate the turning path.
When left turning haffic was given a green light, these “runway” lights would light up green
along the path to be taken by the motorist (Figure 5.6). These lights were reported to be a
maintenance problem as they may fill with dirt, which obscures the lens. The engineer
responsible for maintaining the operation of this location expressed a concern that the lights
may also raise several tort liability issues. For example, if the nmway lights are not working
at the time of an accident, it may be consh'ued that one of the traffic conh'ol devices (TCDS)
was not working.

Many of the SPUIS reviewed have charmelized islands to help guide drivers as they
negotiate though the single-point intersection. On the center island, typically there was also
directional Signing present. The location of this signing may make it vulnerable to damage

fi'om motorists who shay onto the island. During the field review, it became obvious that

35

 

       

Figure 5.6: "Run ay" ligh ' g to help illuminate the turning path.

Note the buildup of debris.

36
motorists frequently shike these islands. Channelized islands are not as popular in Michigan

because of their interference with snow plowing operations.
5.5 Conclusions from the Field Review

Based on this field review, subjective observations can be made about the design and
Operation of the SPUI. These observations were grouped into the areas of geomehic design,
signal operation, pedeshian control, and pavement markings.

The most significant geomehic design difference of the SPUIS reviewed is between
a SPUI with the cross-road going over the freeway and a SPUI with the fieeway going over
the cross-road. The SPUI with the cross-road going over the freeway was found to look and
operate more like a conventional signalized intersection. Another design difference was
related to the physical Size of the interchange. SPUIS without dedicated U-tum lanes
appeared to accommodate U-tums as well as those with dedicated U-turn lanes. The smaller
designs were observed to function better than the larger designs. In addition, the Right-of-
Way requirements are less with the smaller designs. In some cases, the sh'uctures were
noisy resulting in complaints from nearby residents. Because of the large size of these
Sh'uctures required when the fieeway goes over, the roadway under the sh'ucture is dark.
These undesirable sh'ucture characteristics are not present when the cross-road goes over the
freeway.

Furthermore, in the case where the freeway goes over the cross-road, sight distance
is a concern. Several engineers expressed shong opinions that the use of continuous fi'ontage
roads with a SPUI negates the advantages of the design. Finally, the geomehy of the typical

on-ramps may result in a Sideswipe crash problem.

37
The signal operation shategy employed by each state differed significantly. Cycle

lengths varied from 80 seconds to 180 seconds, with longer cycle lengths usually having
fully actuated Signal phases for all movements. The interchanges reviewed were operating
below capacity and, at this level, progression of the cross-road was not a problem. If the
interchange area was very large, the clearance times became quite long and there was
Significant driver confusion. Finally, the best placement of traffic Signal heads occurred in
designs where the cross-road went over the freeway, allowing the Signal heads to be located
on a single overhead tubular beam.

The ability to accommodate pedeshians varied greatly between designs. Typically, it
was not difficult for pedeshians to move parallel to the cross-road and cross the ramp
movements. However, due to the characteristics of the SPUI, there is always haffic moving
through the intersection. This makes it difficult for pedeshians to cross the cross-road.

The need for pavement marking in large SPUIS is paramount. However, these
pavement markings can overlap and cause driver confusion. This resultant driver confusion
is most pronounced when the cross-road is skewed.

Based on the field review, the conclusion of the study team is that the Single
Point Urban Interchange (SPUI), properly situated, is an effective design. However, the
large size of some designs may be counterproductive due to the need for an increase in

clearance times.

Chapter 6

METHODOLOGY

Sufficient haffic volumes were not present at any of the locations visited during the
field review to allow for a field determination of operation at capacity. Thus, to compare the
relative operational characteristics of the interchange configurations in question, computer

modeling of each geomehic configuration was used.

6.1 Selection of the Computer Model

The concept of traffic control is giving way to the broader philosophy of
Transportation Systems Management (TSM), in which the purpose is not to move
vehicles, but to optimize utilization of transportation resources in order to improve the
movement of people and goods without impairing other community values (1). To better
understand this optimization, computer Simulation techniques have been developed.
These models describe a system’s or network’s operational performance based on various
data inputs. This minimizes the need for an existing facility to be expanded or a proposed
facility to be constructed to evaluate their performance.

The computer simulation approach is considered more practical for evaluation of
network changes or operation than field experiments for the following reasons:

0 It is less costly

0 Results are obtained relatively quickly

38

39

o The data generated by simulation includes many measures of effectiveness

that cannot easily be obtained from field studies

0 Disruption of haffic operations, which often accompany a field experiment, is

avoided (1).

TRAP-NETSIM is a stochastic, microscopic model which describes the
operational performance of a road system based on several measures of effectiveness
(MOES). The internal logic of this model describes the movements of individual vehicles
responding to external stimuli including traffic control devices, the performance of other
vehicles, pedeshian activity, and driver performance characteristics. NETSIM applies
interval-based simulation to describe traffic operations. Each vehicle is a distinct object
which is moved every second, and that every variable control device (traffic signals) and
event are updated every second. Each time a vehicle is moved, its position (both lateral
and longitudinal) on the links and its relationship to other vehicles nearby are
recalculated. Its Speed and acceleration are also recalculated. Vehicles are moved
according to car following logic, response to traffic control devices and response to other
demands (1). For these reasons, the TRAP-NETSIM model was selected for use in this
study.

6.2 Network Configuration

To compare the operation of a diamond interchange (Figures 6.1 and 6.2), a MUDI
(Figures 6.3 and 6.4), and a SPUI (Figures 6.5 and 6.6), several decisions were made about the
network geometry. First, it was decided to model the arterial crossroad as both a five-lane and

seven-lane pavement. The cross-section of the five-lane facility consists of four through lanes

40

 

\ Ans—mom 0336 J
museum ommucccm

53> guacamQEVU omen—L885

98:85 3393. ac gamma

agar
é: ............ ll

 

 

 

 

 

 

41

 

 

salelclo ale 0

e
g

 

4—-— 8001!
Q 8043

@lle m

Diamond!

 

 

 

Figure 6.2: Link/Node Diagram for Diamond

Configuration

(Not to Scale)

 

 

42

 

38w 2 82V
m—UMOM OWN—Cain 5T5 Gomumhswwcoo

GODS: E53285 ecoEmmD
5&3 52:32 33%.“. “m6 oSwE

Ema mg

 

 

 

 

 

 

 

 

 

 

a WE ..

A _

WE

 

 

 

 

 

 

 

 

 

 

 

 

 

(fl ,
(MD
I

9

I
e—e—e

®0M9
em 90

st

0
e
I
e
.'

Figure 6.4: Link/Node Diagram for MUDI

Configuration
(Not to Scale)

GO Gm

ea?»

9—:

 

 

MUDI

 

 

 

 

 

\ 33m 2 SE J
38% owes—8m £3, cosmbwcoU

some 953235 fie:
«Eon 2w£m .333. ”we oSwE

 

—

---------
fl

----—---0
fi

—
—

 

 

 

 

 

 

 

 

 

 

 

-----------
fl
-----------

3;»

 

 

 

 

 

 

 

 

 

43
m3 ....

 

 

 

45

 

 

 

SPUI

 

 

 

 

Figure 6.6: Link/Node Diagram for SPUI

Configuration

(Not to Scale)

 

 

46
(two in each direction) and a continuous center lefi-turn lane (CCLTL), while the seven-lane

facility consists of six through lanes (three in each direction) and a CCLTL.

Next, the size of the network had to be determined. Since a major concern with regard
to interchange operation is the interchange’s effect on the downstream nodes of the arterial, it
was decided to model both the interchange area and one arterial downstream node on either
side of the interchange. These downstream nodes were modeled as the intersection of the
arterial with a five-lane arterial with a CCLTL. Since an arterial is said to have “perfect
geometry” if the intersections are 0.8 kilometers (one-half mile) or 1.6 kilometers (one mile)
apart, these downstream intersections were initially placed at 1.6 kilometers from the
interchange. The perfect geometric spacing of these intersections allows for optimal signal
progression, thus minimizing delay. The impact of minor crossroads and driveways was not

modeled.

Once the spacing of these downstream intersections had been determined, their
geometry had to be defined. For each approach to the downstream intersections, a 168 meter
(550 foot) left and right turning bay was provided. In the interchange area, a 168 meter (550
foot) right turn bay was provided on the arterial approach for both the MUDI and diamond
interchange. Additionally, a 168 meter (550 foot) right turn bay was provided on the frontage
road for traffic wishing to make a right turn from the frontage road to the arterial for both
configurations. In the SPUI interchange area, the length of the right turn bays was shortened to

69 meters (225 feet), as the right turn was operating in a free-flow condition.

47

6.3 Signal Operation

For the purposes of the computer model, a free flow speed of 72 kph (45 mph), or 20
meters per second (66 feet per second), was assumed for the arterial, minor crossroads and
frontage roads. Based on this free flow speed and an intersection separation of 1.6 kilometers
(one mile), the optimal cycle lengths were determined to be a multiple of 40 seconds. Longer
cycle lengths will accommodate more vehicles per hour due to the lower frequency of starting
delays and clearance intervals. Thus, an 80 second cycle was selected for the downstream
nodes for all cases. An 80 second cycle was also selected for the operation of the MUDI.
However, since the modeled arterial was to be operated in a progressed-coordinated system, a
160 second cycle (double cycle) was selected for the interchange signals in both the SPUI and
the diamond interchange due to the need for long phase changes and clearance intervals.
Further, given the freeflow spwd of 72 kph (45 mph), the minimum phase change interval
(yellow and overlapping red) for each phase was determined to be 5 seconds. This phase
change interval ensures that approaching vehicles can either stop or clear the intersection

without conflicts.

The modeled arterial was to be operated in a progressed-coordinated system, so a
definite time relationship exists between the start of green intervals at adjacent intersection
signals. Thus, signal offsets had to be determined. Since both downstream intersections were
placed with perfect geometric spacing from the interchange, the fi'ee flow speed was assumed
to be 72 kph (45 mph), and a cycle length of either 80 or 160 seconds was used, an offset of 0
seconds was selected to best provide for progression of traflic along the arterial. When the
spacing of the closest downstream intersection was changed to 0.8 kilometers (one—half mile),

this offset was changed to one half a cycle or 40 seconds. Furthermore, when the spacing of

48
the closest downstream intersection was changed to 1.2 kilometers (three-fourths mile), this

offset was changed to 20 seconds for the closest node and 60 seconds for the node placed at

2.0 kilometers (one and one-quarter mile).

The number of phases used depends upon the geometry of the intersection (number of
approaches, lanes) and the volumes and directional movements of traffic. The purpose of
phasing is to minimize the potential conflicts at an intersection by separating conflicting traflic
movements. However, as the number of phases increases, the total delay is increased and the
total carrying capacity of the intersection may be reduced. Thus, it is desirable to use the

minimum number of phases that will accommodate the traffic demands.

The signal phasing diagram for the intersection of the minor five-lane CCLT'L and the
arterial was the same for both downstream nodes. It was assumed that the volume ratio
between the arterial and the minor crossroads would be 70/30. Thus, the green split between

the arterial and crossroad would also be 70/30.

The phasing diagram for the MUDI signals was determined (Figures 6.7 and 6.8)
using a green split of 60/40. In addition, an offset had to be determined for the crossover
signals of the MUDI design. At the free flow speed of 72 kph (45 mph), or 20 mps (66
fps), a vehicle requires 8.3 seconds to traverse the 168 meters (550 feet) from the
intersection to the crossover. The desired offset for the crossover signal is one which
reduces the delay to arterial traffic wishing to make an indirect lefi turn while not
adversely affecting the progression of the arterial. If a vehicle left the stop bar of the
crossroad intersection at the free—flow speed and there were no cars at the crossover
signal, this offset would be 8.3 seconds. However, there is typically a queue of vehicles,

mostly comprised of exiting freeway traffic wishing to make an indirect left turn onto the

49

 

 

<I>2=4s

11H

 

“m
m .

 

lll'

d>3=43

E

lll'

fit

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

JJJ

 

 

 

 

WT

 

 

 

Figure 6.7: Phasing Diagram for MUDI Configuration

Ht? 1 HP :4 W

M :_ $331 2: (mm 2352-.

I“ _ “I": m"
:LU :LLl —':::LLL

_: W T; 1 "T' T “Tr

 

50

 

 

 

 

 

 

 

CD1=503 <D2=4s CD3=1s
W H W
<D4=203 <D5=43 CD6=1s
_
...... .____ __
,__
...... .__ .
.__
...... ._ .
/ »—— _
TT T'— W

 

 

 

 

 

Figure 6.8: Phasing Diagram for MUDI Cross-overs

 

 

 

51
arterial, waiting at the crossover signal. For the best progression of the arterial traffic, this

queue must begin to dissipate before indirect left turning traffic from the arterial reaches
the crossover signal. This will result in an offset that is less than the 8.3 seconds. The
study done by Dorothy, et. al. (8) determined the best crossover signal offset to be four

seconds.

A signal phasing diagram was developed for the SPUI for the case where no frontage
roads were present (Figure 6.9) and for the case where frontage roads were present (Figure
6.10). A concern with signalizing the SPUI is the need for a long phase change interval to
allow traffic to clear the intersection. Thus, the minimum phase change interval of 5 seconds

was increased to 9 seconds for all SPUI movements except for the frontage road movements.

Finally, the signal phasing diagram for the diamond (Figure 6.11) was determined. A
concern with signalizing the diamond interchange is the need for a clearance interval to allow
time for traffic which has turned left from the ramp and is stored on the structure to begin
clearing before releasing arterial traffic. Thus, a 12 second clearance interval was provided.
This clearance interval advances the green time for traffic stored in the median of the
diamond, allowing it to clear the median area before giving the remaining arterial traffic a

green indication.
6.4 Variables And Measures Of Effectiveness

There were four major variables of interest addressed in this study: traffic
volumes, turning percentages, frontage roads and distance to the closest downstream
node.

The networks were loaded by considering the percent saturation of the entry links

of the arterial. For the entry links of the arterial, it was assumed that each entry lane had a

52

 

 

 

CDa=5$

All Red

 

 

 

 

CD6=53

All Red

 

 

 

 

 

CD 12=Ss

All Red

 

 

 

 

Figure 6.9: Phasing Diagram for SPUI Configuration
without Frontage Roads

53

 

 

 

CD 3=Ss

All Red

 

 

 

 

 

 

CD6=55

All Red

 

 

 

 

 

 

CD9=15

All Red

 

 

 

(D 12=Ss

All Red

 

 

 

     

Figure 6.10: Phasing Diagram for SPUI Configuration
with Frontage Roads

54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(Dr-:61: J“, 02:43 (1.11 _
HQ” .310"
T w T Ill.
@9613 11 F— @4838: 11 *—
lertT" 11k sni—
: TT 2 U
@5-43 11 h— @0813 @7133“ H ‘-
*: f—
lerTT All Red 11 11
J
a a
TT U
was“ J, l :::\_ (Do-ts (bro-12: l J,
.3: E
j; H H All Red H 11
a; 23
11 IT

 

Figure 6.11: Phasing Diagram for Diamond Configuration

 

55
capacity of 1800 vehicles per hour of green. With this in mind, a simple incremental

volume structure was identified for study based on arterial entry link saturation values of
0.3, 0.5, 0.7, 0.9, and 1.0. The minor downstream crossroad entry links were assumed to
have a per lane hourly volume ratio of 30/70 when compared to the arterial entry links.
Furthermore, the network was modeled with an in-balance in traffic flow for both the
frontage roads and exit ramps. It was assumed that there was a 70/30 imbalance in flow
between traffic approaching from the left and traffic approaching from the right (Figures
6.1, 6.3, and 6.5). The maximum frontage road volume was assumed to be 600 vehicles

per hour.

The second variable addressed was turning percentages. First, turns from the minor
crossroad to the arterial were fixed at 20 percent toward the interchange and 10 percent away
from the interchange. Turns from the arterial to the minor crossroad were fixed at 10 percent
lefi and 10 percent right. Second, for arterial traffic approaching the interchange, it was
assumedthat25 percentwantedtotm'nlefttoaccessthe on-ram ,25 percentwantedtoturn
right to access the other on-ramp, and 50 percent wanted to continue on the arterial. Third,
turning traffic exiting the fi'eeway was varied to test the sensitivity of the designs to the
volume of left turning traffic. Thus, values of 30, 50, and 70 percent left turns fiom the exit
ramps were modeled. Finally, it was assumed that the volume of traffic entering on a

particular frontage road would also exit on that frontage road.

The third variable addressed was the existence of fiontage roads. In Michigan,
depressed freeway segments typically are built with frontage roads to access the adjacent
properties. Thus, the operation of a particular interchange configuration with and without

frontage roads was of interest.

56
The final variable addressed was the distance to the closest downstream node. Early in

the project, a concern was raised about the eflect that an interchange would have on a closely
spaced intersection. In addition, it was desired to determine how an interchange configuration
would function in an arterial that did not have perfect geometry. Thus, the distance to the
closest downstream node was varied. To keep the size of the network constant, as a
downstream node was moved closer to the interchange area, its counterpart on the other side
of the interchange was moved and equal distance away fi'om the interchange. The first value
modeled was a spacing of 1.6 kilometers (one mile) to either side of the interchange area
allowing for perfect progression on the arterial. The second value modeled was a spacing of
0.8 kilometers (one-half mile) on one side and 2.4 kilometers (one and one-half mile) on the
other side. This spacing still allows for perfect progression of the arterial. However, the
proximity of one of the downstream nodes to the interchange may be a factor. Finally, a
spacing of 1.2 kilometers (three-fourths mile) to one side and 2.0 kilometers (one and one-
quarter mile) to the other side of the interchange was modeled. This configuration does not
allow for perfect progression along the arterial, but does keep a larger separation between the

closest intersection and the interchange.

A TRAP-NETSIM simulation run produces an output that summarizes the traffic
movements and various measures of effectiveness (MOEs) for both the network as a whole
and for individual links. The MOEs that were selected for this study were: interchange area
total time and downstream area total time. In TRAF-NETSIM, the MOE “total time” is made

up of move time and delay time.

An effort was made to delineate an interchange area and a downstream area in the

computer model. The physical size of these areas was the same for all models. However,

57
inside the area, the size of the interchange may vary. The nodes numbered 7 and 8 were coded

as dummy nodes (i.e. no change in the traffic stream occurs at them) to allow MOEs to be
gathered for both the interchange area (the area bounded on the top by node 7 and on the
bottom by node 8) and the downstream area (the area above node 7 plus the area below node

8).

A criticism of the indirect left-turn strategy used by the MUDI configuration is that
while conflicts from left turning vehicles have been removed from the intersection, these
drivers are penalized by being forced to travel a greater distance to use the cross over. Thus,
delay cannot be used as a MOE, as it would be unclear if the delay savings at an intersection
were being ofi‘set by the extra travel time imposed on left-turning traffic. Therefore, total time,
which represents the amount of time all vehicles spent in the network as a combination of

travel time and delay time, was selected as a MOE.

Chapter 7

SINIULATION RESULTS

Based on the variables selected for study, an hour of operation for 300 individual
models was simulated. Each model used the same random number seed. Since TRAF-
NETSIM brings the simulated network to equilibrium before starting to collect statistics and
the network will be simulated for one hour of operation, the results should be repeatable and
independent of the random number seed. The network was simulated for a saturation up to
100 percent to aid in determining when simulation results become invalid due to delay
occurring outside the environment of the analysis. However, TRAF-NETSIM may produce
unreliable results when run at levels of saturation approaching 100 percent. Thus, the results of

the 100 percent saturation runs will not be discussed.

The values of start-up lost time and headway were not calibrated or validated based on
field data Since Michigan does not have a SPUI, it was not possible to determine what values
would be applicable for Michigan drivers utilizing a SPUI. However, the default values
irnbedded in the model for start-up lost time (2.0 seconds) and headway (1.8 seconds), which

are based upon national averages, were used for each interchange type.

A representative example of the findings are presented in the text, while all findings

are presented in both tabular and graphical format in the appendices.

58

59
7.1 Interchange Performance without Frontage Roads

Figure 7.1 illustrates the performance of the interchange configurations without the
presence of frontage roads and with a five-lane arterial cross-section. The situation modeled in
this scenario is for the extreme case of 70 percent of the vehicles exiting the freeway and
desiring to turn left onto the arterial. At 30 percent saturation, all three interchange
configurations performed approximately the same. However, at 50 percent saturation, the total
time for the MUDI and SPUI configurations was only 60 percent of that for the traditional
diamond. Additionally, at 70 percent saturation, the total time for the MUDI configuration was
25 percent less than the SPUI and 36 percent less than the traditional diamond. Finally, at 90
percent saturation, the total time for the MUDI configuration was 16 percent less than the

SPUI and 20 percent less than the traditional diamond.

Figure 7.2 illustrates the results when the percent left turns is reduced to 50 percent.
Although the operational advantage of the MUDI is less, it is still meaningful and follows the
same pattern as the 70 percent left-turn case outlined above. At 30 percent saturation, all three
interchange configurations still performed approximately the same. At 50 percent saturation,
the total time for the MUDI and SPUI configurations was 50 percent of that for the traditional
diamond. Moreover, at 70 percent saturation, the total time for the MUDI configuration was
18 percent less than the SPUI and 38 percent less than the traditional diamond. Finally, at 90
percent saturation, the total time for the MUDI configuration was 23 percent less than the

SPUI and 32 percent less than the traditional diamond.

As the percentage of left turns is decreased to 30 percent (Figure 7.3), the operational

characteristics of both the MUDI and the SPUI configuration change at higher levels of

$8

9...... ram 3.2. 6 8:838 2.8....
$2.

$3

 

LO
8..

 

 

\

 

 

 

 

 

\

 

VT

 

 

 

 

 

 

 

 

one; .3255 L11
o:m.-m .de III
957m .53.). IQI

 

 

 

 

 

.335 25.» 3.... s 2388.: 3

.353. 35:2". 323 .25.... :3 $3 .8“. 2...... .50... a2< ones—.28... up... 8.5....

com

85

(unou women) own "no; any caunqmozuu

61

$8

9...... rem 3...: .o 8:233 .52....
$2 $3

$8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

257m .ucoEmE Lal

 

 

 

! - 0cm; .de III
257m .535. IOI

 

 

 

 

 

 

cos

 

coo

 

3.2.... 25..» .22. S 882:2... 0.9
.33: 38:2“. 23.3 .22.; to. $8 .0“. 2...... .53 ¢2< 99.2.23... N.» 2.3.“.

(unou aroma) emu. mar. mv OBUIIIOJOWI

62

9...... >55 .3?! .o 2.0.3.35 2.3.3.
28 $2 $8

$8

 

 

 

-oow

cow

 

\r\\

 

 

oov

 

 

 

 

 

 

 

 

 

 

 

 

95.6 .225“... LT
0.57m .de III
2.5.6 .53.). 191

 

 

 

cos

 

com

 

3.2.2 2.0.6 A22: 5 83050.... a...

6.33. 38:2... So...» .25.... no. 3.... 3“. 2...... .88. no.3 09.2.9.2... an... 2.6.“.

(unoq women) our”, pro; euv catamaran"

63
saturation. At 30 percent saturation, all three interchange configurations are again

approximately equal. In addition, at 50 percent saturation, the total time for the MUDI and
SPUI configuration was again 50 percent that of the traditional diamond. However, at 70
percent saturation, the total time for the MUDI was 28 percent less than both the SPUI and
traditional diamond, which perform approximately the same. Finally, at 90 percent saturation,
the total time for the MUDI was 23 percent less than the SPUI and 10 percent less than the
traditional diamond. Thus, at 90 percent saturation, the traditional diamond is operationally

superior to the SPUI, as measured by this single MOE.

Much of the same pattern is shown when the arterial cross-section is changed fiom a
five-lane cross-section to a seven-lane cross-section (Figures 7.4-7.6). The major differences
are that at 30 percent saturation, the total time for both the MUDI and SPUI was 35 to 40
percent less than the traditional diamond for all tunring percentages. In addition, the MUDI
with a seven-lane arterial begins to operationally outperform the SPUI at 50 percent saturation

as opposed to at 70 percent saturation with a five-lane arterial.

Based on the MOE “interchange area total time”, in all cases, the MUDI configuration
either equals or exceeds the operational performance of the SPUI and traditional diamond
configuration. These operational advantages are most pronounced when the percentage of left-
turning traffic is high and the level of saturation is high. The operational advantages of the
SPUI are greatly reduced as the percentage of left-turning trafiic is reduced, with the
traditional diamond outperforming the SPUI at high levels of saturation and low levels of left-

turning trafiic.

9...... Penn .0?! .0 53235 2.33..
$8 $2 $8

 

 

 

 

 

8..

 

 

 

 

 

 

cow

. con

 

 

 

 

 

 

 

 

 

 

25..» 2.0890 +
mam—-5 .de III

 

 

25$ .03.). IOI

 

 

 

 

 

 

com

.8»

 

cow

 

3.22 2....-. ..o..... a 3325.... m...
.258... 355...... 50.3 .25.... to. $2 .0“. 2...... .50... a2< 352.22... 3.5 2.5....

(unoq women) anal 1, "to; any oBueuoumu

65

$8

2...... .55 .0?! .6 3.3.3.“ 2.3.3.

.32.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2a.-.- .ucoEmE If.-
mcmwn .de III
2.9-5 .022 IOI

wllllnl-Itlllll

 

 

 

 

 

are... 2.2-.. .2... P. 828.2... 3
.33... 33:0... 2.0.... .25.... to. $0... .on. 2...... .80... no... 35:22:. "m... 0.3...

N

.8-

(unoq women) emu pro; euv oBuequeun

O
O
‘0

con

coo

66

$8

9...... DEN .6?! .0 53233 2.3....

$2. $8 $8

 

 

 

 

 

 

 

.l'll"-.l.|. r

 

 

 

 

 

 

 

mam: .3955 +1
‘ iii -t -l ---?.1_,-i,1 mam: .Samlll

 

 

 

0cm.-. .535. IOI

 

 

 

 

 

3.2... 2...... ..o__... s 22.52... 3
6.50m. 38...... 2.2.... .25.... to. «Sn .0. 2...... .50... no... 3.5.3.8... 6... 8.5..

8.

SN

con

foam

och

(smoq women) Gal] 1, Inc; 301v aflucqmazul

8...

67
7.2 Migration of Delay without Frontage Roads

In this research effort, there is concern that greatly enhanced urban interchange
configurations may demonstrate an improved operation at the freeway, but may merely move
the delay to the first signalized intersection upstream or downstream. Thus, their advantages
(if any) may be exaggerated. Therefore, this analysis also evaluated the operation of the

downstream nodes.

As illustrated in Figure 7.7, which is a specific case with 50 percent left turns, five-
lane arterial cross-section and no frontage roads, there was no evidence that either the MUDI
or SPUI configuration resulted in moving delay to the downstream nodes. However, the total
time for the downstream area when fed by traffic from the traditional diamond interchange is
greater for all but the 30 percent saturation level, suggesting a migration of delay. In addition,
when the specific case with 50 percent left turns, seven-lane arterial cross-section and no
frontage roads (Figure 7.8) is examined, this trend continues for the traditional diamond. At 70

percent saturation, the modeling of the SPUI also shows this effect.

7.3 Interchange Performance with Frontage Roads

Many, if not most, of the MUDIs in Michigan are located where frontage roads are
provided. Usually these fiontage roads parallel the urban freeway for a considerable distance
and provide access to abutting property. The need for local access in a major urban area was a
primary consideration in the evolution of the MUDI design since frontage roads would need to

be provided.

Figure 7.9 illustrates the performance of the interchange configurations with the

presence of frontage roads, a left-turning percentage of 70 percent and a five-lane cross-road.

68

9...... .5... .0?! no 53835 «:33.

$8 $0»

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

one... .3955 +

slr 2...... Samuel 1!

 

 

0cm; .032 IOI

 

 

 

3.3.2 2.5...." A33. 5 23050.... o...

.33.. 35...... 523 .25.... to. $3 .0. 2...... .30... an... Eagogoa K.» 23..

(canon croquet.) emu, run; any mnmoumoa

8a..

83

.89

69

\\\

9...... .55 .0?! 3 330.35 300.0.
$8 $2. $8

$2.”

 

 

 

 

 

coo

 

 

 

 

coop

 

\

 

 

03... 0:02.05 Lal
0:0: ..:.w III
0:0... .525. IOI

 

 

co:

 

 

 

 

 

 

 

coo.

 

cow.

 

.0..0..< 0.8.... ..0..... 5 0.0.0.3.... 0..
.0000... 38.3.. 2.0.3 .033 80. $3 3. 0...... .50... a0.< 0.00.3530 "a... 0.3..

8N7

(0mm; women) emu [mo-L euv urnmeumoa

9...... .5... .03.. .0 50......» .52....

$8 $2. *on

 

 

 

 

 

 

 

 

 

 

 

 

 

70

 

 

 

 

 

 

 

0.3.6 3.30.0.0 +
0.3.6 .me III

 

 

 

 

 

 

 

0.3.6 .035. +

 

 

 

 

 

 

 

3.3... 2...... ..0..... s 832.8... 3
.300. 33.3.. 5.3 .05.; #0.. $2. .0. 0...... .50... 02¢ 3.3.3.8... "a... 0.3..

§ .3. § § §
autumnal

("“0" 3PM“) Null. I330. 301‘! 9

§

71
At 30 percent saturation, all three interchange configurations performed approximately the

same, which is consistent with the results from simulations without fiontage roads. However,
at 50 percent saturation, the total time for the MUDI configuration was 21 percent less than
the SPUI and 59 percent less than the traditional diamond. This represents a divergence from
the results of simulations without frontage roads, in which the MUDI and SPUI performed
the same at 50 percent saturation. At 70 percent saturation, the total time for the MUDI
configuration was 18 percent less than the SPUI and 29 percent less than the traditional
diamond. Finally, at 90 percent saturation, the total time for the MUDI configuration was 13

percent less than the SPUI and 33 percent less than the traditional diamond.

Figure 7.10 further illustrates the performance of the interchange configurations with
both fiontage roads and five-lane arterial cross-sections. However, the percentage of left-
turning trafiic has been reduced to 50 percent in this case. At 30 percent saturation, all three
interchange configurations continue to perform approximately the same. At 50 percent
saturation, the total time for the MUDI configuration was 12 percent less than the SPUI and
59 percent less than the traditional diamond. These results are consistent with the scenario
involving 70 percent lefi-tums outlined above. However, the results diverge from the results of
the scenario involving no frontage roads, in which the MUDI and SPUI performed similarly at
this level of saturation. At 70 percent saturation, the total time for the MUDI configuration
was 21 percent less than the SPUI and 38 percent less than the traditional diamond. Finally, at
90 percent saturation, the total time for the MUDI configuration was 23 percent less than the

SPUI and 25 percent less than the traditional diamond.

Figure 7.11 illustrates the performance of the interchange configurations with the

presence of fi'ontage roads, 30 percent left-turning trafiic and a five-lane arterial cross-section.

72

8...... .55 .3?! .0 3:23am «52!.

$8 $2. $8

 

 

 

 

 

 

 

 

 

 

 

 

0:3... .ucoEmE lil
9570 .Smw III

 

 

 

 

 

 

 

 

mes-m .532 +

 

 

 

 

 

3:33 25; ..o..:. 5 282:2... o...
.333. 35:2". 5.3 6:5... #3 $3 an. 2...... .80... .92 09.2.28... "3.» 2.6.“.

8-

com

02.

out”, "so; any eBuaqmonu

73

9...... .55 3...: .o 5:238 2.8....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$8 $2. *8 «won
|L o
\\ 8N
con
oov
. 8m
ll 80
953 6.5830 +
1!! E1}:|l,liil .11? ii. .111 1-x-liililli- -j 95-... Emmi-Iii}! 8..
mam—um ._DDE+
coo

 

 

its... 2.2... 3.... s 2225.... 3
.331 38:2... 5.3 .25: :3 $2" .5". 2...... Each a2< 09.2.23... 2...... 2.6.“.

(unoq clamor» our”, "to; Buy catamaran"

74
At the lowest level of saturation, all three interchanges continue to perform approximately the

same. However, at 50 percent saturation, the total time for the MUDI configuration was 25
percent less than the SPUI and 46 percent less than the traditional diamond. This continues the
trend of the SPUI incurring greater total time than the MUDI (at 50 percent saturation) as was
the case for the scenarios without frontage roads. At 70 percent saturation, the total time for
the MUDI configuration was 21 percent less than the SPUI and 18 percent less than the
traditional diamond. Finally, at 90 percent saturation, the SPUI and traditional diamond
continued to perform approximately the same. The total time for the MUDI configuration was

28 percent less than the SPUI and 27 percent less than the traditional diamond.

Much the same pattern is shown when the arterial cross-section is changed from a
five-lane to a seven-lane cross-section (Figures 7.12-7.14). As with the scenarios having no
fiontage roads, one major difference was that at 30 percent saturation, the total time for both
the MUDI and SPUI was 35 to 40 percent less than that of a traditional diamond for all turning
percentages. Additionally, for all lefi turning percentages, at 90 percent saturation, the
traditional diamond operationally outperforms the SPUI. Moreover, for left-turning
percentages of 50 and 30 percent, the SPUI performed similar to the traditional diamond at
saturation levels of 50 and 70 percent. However, in the scenario where the left-turning
percentage was set at 70 percent, the results of the MUDI simulations are not valid past the 70
percent saturation mark. This is due to a spillback of traffic on one of the model’s entry links,

which resulted in delay occurring outside the environment of the analysis.

As with the scenarios involving the performance of the interchange configurations

without frontage roads, based on the MOE “interchange area total time,” the MUDI

75

9...... be... 3...... .o 5:236 .52....
$8 $2. $3

$8

 

 

X...

 

.\\\

 

\\

 

\\\\

 

 

 

 

one: .ucoEmE +
mam—-5 ._Dn.m Ill

 

coo

 

 

 

can...» .522 I?!

 

 

 

 

 

 

 

och

 

 

.235 :5... 3.... s 222.2... 3
.33... 35:2“. 5.3 .25.... to... $2. .0“. 2...... .50... 3.3 3:228... .N E 2.3.“.

(unou woman) emu rue; 301v ofluaqmorul

76

9...... ram. 3.3. .o 3:938 .52....
$8 $2. $8

$9.”

 

 

 

 

§

 

\

 

§

 

 

new: .3855 Id...

 

 

 

96: .5am III

(unoq woman) our; 1, mo; 901v oBucqmozm

 

 

 

 

9575 .53.). lol

 

 

 

och

 

 

5.2.... 2.2... ..o..... p. 9.852... 3
.83.. 35:2“. 5.... .25: 9.... so... 3.. 2.... a...» so... 8:223... a... 2......

8w

77

$8

9...... 3...... 3.... 3 c8238 333..
$9. $8

$8

 

 

 

 

 

 

 

 

 

 

 

 

Tlllll.lltll. Ill. ll

 

 

 

 

 

 

 

 

 

one... .3955 +
9.0.... .5mm III
9.0.-.. .522 ll

 

 

 

 

 

 

 

och

 

 

3.3... can...” ..¢_.E 3 3225.... o...
.33”. 35:9... 5.2. 6E3... :3 $2" 3.... 2...... .80... a8< 3:222... .3.» 2.6...

coo

(mnoq women) our”, rue; 901v ofiunqmouu

78
configuration with fi'ontage roads either equaled or outperformed both the SPUI and the

traditional diamond, except where the MUDI could not be evaluated.

7.4 Migration of Delay with Frontage Roads

When the operation of the downstream nodes was examined for evidence of the
migration of delay, a trend was evident. For example, in the scenario representing 50 percent
left-tuming traflic, frontage roads and a five-lane arterial cross-section (Figure 7.15), there is
evidence of a migration effect from both the SPUI and traditional diamond interchange
configurations. This trend is also exhibited when the arterial cross-section is widened to seven-
lanes (Figure 7.16). Thus, for all cases involving fi'ontage roads, the MUDI was operationally

superior in having less migration of delay to the downstream intersections.

7.5 Sensitivity to Proximity of Closest Downstream Node

The effect that the proximity of the closest downstream node has on either the MUDI

or SPUI interchange operation was also studied. Three spacing scenarios were considered:

0 1.6 kilometers (one mile) which allows for perfect progression along the arterial while

maintaining adequate separation between the intersection and interchange area;

0 1.2 kilometers (three-quarter mile) which does not allow perfect progression along the
arterial, but still maintains adequate separation between the intersection and interchange
area;

0 0.8 kilometers (one-half mile) which allows for perfect progression along the arterial, but

the proximity of the intersection to the interchange area may affect operation.

79

2...... .55 3.2. .0 5.3.3.6 «.323.
$2. $8

$8

 

Lo

\\\.-

 

 

 

 

 

 

 

 

 

8N.

 

 

 

0.53 .3955 LT

 

 

 

 

2.96 ._Dn.mlll
0.57m .5:2+

 

 

 

 

 

coo.

 

 

3.3.... 23.... 3.... .. 33.3.... 3 .
.233. 38.3.... 5.... .uEah #0.. $3 .0“. 0...; .80... a2< Engogoa .m ...h 2.6....

(unoq amnion) can; “no; any mumsumoa

80

9...... Sew 3...: .o 8.3.38 .52....
$8 $2.

$8

 

 

 

 

\

 

\\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mam: .ucoEmE lcl
0.57m .me III
maul. .535. +

 

 

 

 

 

3.3.... 2...... 3.... c cacao... 3

.331 38.8.... 5.... 6...: to. $3 .0“. 2...... .80... a2< Emphasis "ops 2.3.“.

com

com?

8:

89

(canon mum's)» on" 1, "301, Buy muncumoa

81
All the scenarios involving sensitivity testing of the proximity of the downstream node were

modeled without the presence of frontage roads.

When modeled with a five-lane arterial cross-section, 70 percent lefi-tums and 30 to
50 percent saturation, the MUDI configuration (Figure 7.17) performed approximately the
same for all three spacing scenarios. In addition, the MUDI configurations with the closest
downstream node placed at 1.6 kilometers (one-mile) and 1.2 kilometers (three-quarter mile)
from the interchange continued to perform approximately the same for all levels of saturation.
However, at 70 percent saturation and greater, the MUDI configuration with the closest
downstream node placed at 0.8 kilometers (one-half mile) from the interchange exhibited
greater interchange area total time than the other two MUDI spacing scenarios. At 70 percent
saturation, the MUDI 0.8 kilometer spacing scenario had approximately 40 percent more total
time than the other MUDI spacing scenarios, while at 90 percent saturation, the total time was

35 percent more.

When the percent left-turns was reduced to 50 percent (Figure 7.18), the simulation
results for the MUDI configuration were similar to that of the 70 percent lefi-tuming scenario
described above. However, when the percent lefi-tums was reduced to 30 percent (Figure
7.19), the MUDI configuration performed approximately the same for all three spacing
scenarios and all levels of saturation. In addition, when the arterial cross-section was changed
to seven lanes, the MUDI configuration performed approximately the same for all three

spacing scenarios and all levels of saturation.

Thus, the only conditions where the MUDI configuration was affected by the spacing

of the closest downstream node were the scenarios using 70 percent lefi turning traffic, an

$8

9...: Pam 3.8. 3 3.8.3.0 .52....
$2

$0...

$8

 

 

 

 

 

\\

 

 

 

2
00

 

 

 

 

 

m? .1

 

 

 

 

 

 

 

 

 

0...: N: 6:070 ._Dn.m+
0...: VB .0:0_.m ._me+
0...: .. .0:0_.m ._Dn.ml*l
0...: N; .0:0_.m ._DDE+
0...: SM 6:070. ._032+
0...: F .20....» ..032+

 

 

 

 

och

 

coo

 

02.2.30 0502.0

9.5.0) .33.? 2.0.6 .0003. 003:2“. 2.2:. 65...... a... 8.: .8 2...... .53 02¢ 09.0.8.3... K: 2.5....—

(umou from“) own ulnar. wv tau-Momma

83

9...: 3cm 3.8. .o 5:83.» 2.8....

$8 $3 $00 $00

 

 

 

 

Kilt

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\ 0...: N; 6:070 ._Dn.m+

 

 

0...: v8 6:070 ._Dn.m+
0...:P 6:070 ..Dn.wl*l
0...: N: 6:070 ..DDE+

 

 

 

 

 

0...: 30 6:010. ..032+
0...: w 6:070 .532IOI

 

 

 

 

 

 

 

 

00:0:30 2.00:0
05E; ..0..3.< 2.0.0 .0003. 005:9... 2.2.... .053... to.— $om .8 2:... .80... 02< 35:83:. .3.» 250....

8.

§

§

§

§

(""0" "ON.” “all. "301. M WI

1"" I
00:020an 5522......

.350”. 093:9". «:23 £53.. to... $2” .5. 0.5... .30... umg< 09:322.: 59K 033...

05>...) ._0...0t< 0:076

9...: Sam 8.3. .o 8:238 .58...
$8 $2. $8

$00

 

 

 

 

-8.

 

 

 

 

8
N

 

 

 

 

§

 

 

 

 

0...: N: 6:070 .=..n.m+
ii:l--! 0 $191 ....... 0 w , - .2 -lrfi 0.8 an .850 .Samlxl

0...: F 6:070 ._Dn.le|
0...: N... 6:07.... ..DDE+
1.: ._ .ylilui -.|».: . 31:12.-.. i...¥ ill! .1 0...: 30 .30.... ..o:5+
0...: .. 6:070 ..o:2+

 

 

 

 

 

§

 

(anon «om-A) own "no; ION OMNO-lflul

§

 

 

 

 

 

 

 

 

 

00.5330 053:0
05...; .33.; 32.0 .0003. 008:2... .523 .053... to. $9” .3. 05.... .50... no.3 35:23:. "a... 2:9“.

85
arterial cross-section of five lanes, saturation levels of 70 percent or greater and a proximity of

0.8 kilometers (one-half mile). However, the models were coded with an imbalance in trafic
flow of 70/30 between traffic approaching from the left and traffic approaching fi~orn the right
(Figure 6.3). Since, this increase in total time only appeared with a left-turning percentage of
70 percent, an arterial cross-section of five lanes and when the model was operating at near
capacity, the most likely cause of this increase is a spillback from the limited storage available

between the downstream intersection and the interchange.

When modeled with a five-lane arterial cross-section, 70 percent left turns and 30
percent saturation, the SPUI configuration (Figure 7 .17) performed approximately the same
for all three spacing scenarios. However, at 50 percent saturation, the interchange area total
time for the SPUI configurations with 1.2 kilometer (three-quarter mile) and 0.8 kilometer
(one-half mile) separation was approximately 35 percent greater when compared to the 1.6
kilometer (one mile) spacing scenario. At saturation levels of 70 percent or greater, the SPUI
configuration with 1.6 kilometer (one mile) separation performed approximately the same as
the SPUI configuration with a 1.2 kilometer (three-quarter mile) separation. However, at 70
and 90 percent saturation, the total time for the SPUI configuration with 0.8 kilometer (one-
half mile) separation was approximately 15 percent and 20 percent greater, respectively, when
compared to the other SPUI spacing scenarios. These results are also reflected in the

performance of the SPUI configuration with a seven-lane arterial cross-section.

When the percent left-turns was reduced to 50 percent (Figure 7.18), the simulation
results were similar to that of the 70 percent lefi-tum scenario for saturation levels of 30, SO,
and 90 percent. However, at 70 percent saturation, the SPUI configuration performed

approximately the same for all spacing scenarios. When the percent lefi-tums was reduced to

86
30 percent (Figure 7.19), the simulation results were also similar to the 70 percent left-turn

scenario for all saturation levels. At both 50 and 30 percent left-turning traffic, the scenarios

modeled with a seven-lane arterial cross-section reflected similar results.

Unlike the MUDI configuration, the total time for the SPUI configuration was
adversely affected for all percent left-turning scenarios when the spacing to the closest
downstream node was reduced to 0.8 kilometers (one-half mile). In addition, at 50 percent
saturation, the scenarios modeling a separation of 1.2 kilometers (three-quarter mile) resulted
in greater total time than the comparable models with a separation of 1.6 kilometers (one

mile).

In all cases, the performance of the MUDI configuration with a separation of 1.6
kilometers (one mile) or 1.2 kilometers (three-quarter mile) either equals or exceeds the
operational performance of the SPUI. In addition, for levels of saturation of 50 percent or less,
the MUDI configuration with a separation of 0.8 kilometers (one-half mile) also either equals
or exceeds the operational performance of the SPUI. Furthermore, at higher saturation levels,
the operational performance of the SPUI configuration was adversely affected by a separation
of 0.8 kilometers (one-half mile). Thus, in most cases, the MUDI configuration appears to be
insensitive to the proximity of the closest downstream node, while the SPUI configuration is

sensitive to the proximity of the downstream node.

For both arterial cross-sections and all three spacing scenarios of the downstream
node, the MUDI configuration (Figures 7.20 and 7.21) showed no evidence of migration of
delay. In addition, the SPUI configuration with a five-lane cross-section and 1.6 kilometer

(one mile) spacing also showed no evidence of migration of delay to the downstream nodes.

9...... Fem 3...: .o 80......» 28......
$8 2.2 2.8

$00

 

 

 

 

-OON

 

 

\\

 

 

 

 

87

 

 

 

0...: N: .0:0..m .......m+

2.... an .22.... 500+

2.... . .220 .Samlxl

w 2.... N: .820 .53....I1l
2.... ca .820 ..o:2+

illlli - -i! .61 0...: — 6:0...» ._Q32+

 

 

 

 

 

 

 

8N—

 

 

 

 

 

 

 

80.

 

00.5500 0502.0 u:.>.0>
.33.: 0:20 .0003. 008:2“. 52.... .0Eah 60.. $00 .8 2:... .80... 00.3 E02353: SN... 0.50...

(when atom-A) own nor 0w momma

88

9...... Sam 3...: .o 8.538 2.8.2.
$8 $2. 2.8

$8

 

 

 

 

 

 

 

 

coo

§

 

 

 

 

 

 

,§_

8
S‘.‘

 

 

 

 

 

0...: N2. 6:24. ..D....w+
0...: in den: ..Dn.m+

o=E _. 6.5.4. ._Dn.wl*|.
0...: N2. 6:34. ..032+
0...... in .23.-» ..032+
0..E w 6.5.4. ..032+

 

 

 

 

 

.8.

 

 

 

 

 

comp

 

 

notacoom 9....an 9.3.;
.33.: can: 6.33. 33.3.... «so...» .25.... co.— $am 3.. 2...... .80... 3..< Sacha—.300 3N.» 2.5...

com.

(unou 519m“) «nu, um; cm W

89
However, for levels of saturation of 50 percent or greater, all other SPUI configuration

scenarios resulted in higher total times, suggesting a migration of delay.

Chapter 8

CONCLUSIONS

8.1 Conclusions from the State of the Practice and Field Review

Since no Single Point Urban Interchanges exist in Michigan, it was necessary to
determine the state of the practice for SPUI design from other states. This was
accomplished by conducting a literature review, AASHTO e-mail survey, telephone
survey and field review. Special attention was paid to characteristics of the design which
may affect urban interchange operation and design in Michigan.

Much of the published literature on SPUI design and operation was either
generated from a study by Bonneson and Messer (3, 4, 5, 6, 7, and 14) at the Texas
Transportation Institute (TTI) or referenced this effort. The authors identified a wide
variation in the geometry and signal strategies of existing SPUI designs. The capacity
analysis performed by the authors show that a SPUI utilizing a 3 phase signal is slightly
more effective than a tight urban diamond interchange, but the advantage diminishes as
the size of the SPUI becomes larger. The addition of a fourth phase to accommodate
frontage roads resulted in a reduced capacity for the SPUI. Further, the authors stated that
the typical SPUI signal phasing does not provide for a protected pedeshian phase to occur
across the cross-road arterial.

Leisch, et. al. (13) raised the concern that the SPUI has little potential for
expansion and modification. Merrit (15) stated additional concerns that drivers who are

unfamiliar with the SPUI design may encounter some initial difficulty. Thus, the SPUI
90

91

design needs to rely heavily on guide signing, pavement markings, and lane use signing
for the necessary positive guidance to drivers. Merrit further stated that the proximity of
nearby intersections is a concern.

A survey was submitted by e-mail to each of the other 49 state departments of
transportation requesting information on the design and operation of SPUIS. Although the
survey was as succinct as possible, only 14 state DOTS responded, of which only 4 had
existing SPUIS. The responses received from different states varied widely. This, coupled
with the limited number of responses, limited the usefulness of the survey information.

The review of the literature and the response to the e-mail survey, while helpful,
had significant inconsistencies and lacked information in key areas. Thus, a telephone
survey was conducted to collect more information and to locate the most appropriate sites
for field review.

Based on the telephone survey, sites in Indiana, Illinois, Minnesota, Florida,
Missouri and Arizona were chosen for field review. The observations made in the field
review were grouped into the areas of geomehic design, signal operation, pedestrian
control and pavement markings.

The most significant difference in the geomehic designs of SPUIS was between a
SPUI with the cross-road going over the freeway and a SPUI with the freeway going over
the cross-road. The SPUIs with the cross-road going over the freeway were found to look
and operate more like a conventional signalized intersection. Because of this less driver
confusion was observed. In addition, routing the freeway over the cross-road exposes the
freeway and major traffic volume to preferential icing in cold weather climates. Another

geometric design difference was related to the physical size of the interchange. SPUIS

92
without dedicated U-turn lanes appeared to accommodate U-turns, for all but the largest

of trucks, as well as those with dedicated U-turn lanes. The resulting increase in size to
accommodate U-turn lanes may be counterproductive due to an increase in clearance
times.

The signal operation strategy employed by each state differed significantly. Cycle
lengths varied from 80 seconds to 180 seconds, with longer cycle lengths usually having
fully actuated signal phases for all movements. The placement of traffic signal heads in
designs where the cross-road went over the freeway resulted in the signal heads being
located on a single overhead tubular beam.

The ability to accommodate pedeshians varied between designs. Typically, it was
not difficult for pedestrians to move parallel to the cross-road and cross the ramp
movements. However, it was difficult for pedestrians to cross the cross-road.

The need for pavement markings in large SPUIS is paramount. These pavement
markings can overlap and cause driver confusion. This resulting driver confusion is more
pronounced when the cross-road is skewed.

8.2 Conclusions from the Simulation Modeling

The literature review showed that while some evaluation of the SPUI had been
done in the past, no comparative analysis had been published with regard to the ability to
progress the arterial cross-road, compatibility with frontage roads, sensitivity to lefi-
turning traffic, migration of delay, or traffic levels nearing capacity. Additionally, while
the operational characteristics of a boulevard intersection have been studied and the
results published, the Michigan Urban Diamond Interchange (MUDI) design, which is

unique to Michigan, has never been formally studied. Thus, the SPUI and MUDI designs

93
were computer modeled using TRAP-NETSIM to facilitate a comparison of their

respective operational characteristics. Furthermore, a traditional diamond interchange
was modeled to generate a frame of reference for the results.

An hour of operation for 300 individual modeling scenarios was simulated. The
results of the simulation modeling are based on this finite number of scenarios defined by
the four main variables addressed by this study: traffic volumes, turning percentages,
frontage roads and distance to the closest downstream intersection. Only one size of
interchange was modeled for each interchange configuration. Additionally, only one
cycle length and one fixed, interchange area signal timing were coded for each
interchange configuration with the cross-road arterial being modeled as a progressed-
coordinated system.

Not all modeling scenarios that were simulated returned results that were valid. In
a limited number of scenarios, a spillback of traffic on one of the model’s entry links
resulted in delay occurring outside the environment of the analysis.

The measures of effectiveness (MOEs) selected for this study were interchange
area total time and downstream area total time, where “total time” is made up of both
move time and delay time.

Based on the MOE interchange area total time, MUDI operation, in most
situations, is superior to that of a SPUI and traditional diamond interchange
configurations. This is true of scenarios modeled both with and without the presence of
frontage roads. These operational advantages are most pronounced when the percentage
of left-turning traffic is high and the level of saturation is high. In addition, the

operational advantages of the SPUI are greatly reduced as the percentage of left-turning

94
traffic is reduced, with the traditional diamond outperforming the SPUI at high levels of

saturation and low levels of left-turning traffic.

The concern that greatly enhanced urban interchange configurations may
demonstrate an improved operation at the freeway, but may merely move the delay to the
first signalized intersection upstream or downstream was addressed. Based on the MOE
downstream area total time, there was less migration of delay to downstream
intersections with a MU DI configuration than with either a SPUI or traditional diamond
configuration. For all scenarios without the presence of frontage roads, the traditional
diamond interchange configuration resulted in moving delay to the downstream nodes.
While there was no evidence that the SPUI configuration resulted in moving delay to the
downstream nodes when modeled with a five-lane arterial cross-road, when modeled with
a seven-lane cross-road, the SPUI configuration shows this effect at high levels of
saturation. Both the SPUI and the traditional diamond show this effect when modeled
with the presence of frontage roads.

The affect that the proximity of the closest downstream node has on either the
MUDI or SPUI interchange operation was also studied for scenarios without the presence
of frontage roads. Three spacing scenarios were considered: 1.6 kilometers (one mile)
which allows for perfect progression along the arterial while maintaining adequate
separation between the intersection and interchange area; 1.2 kilometers (three-quarter
mile) which does not allow perfect progression along the arterial, but still maintains
adequate separation between the intersection and interchange area; and, 0.8 kilometers
(one-half mile) which allows for perfect progression along the arterial, but the proximity

of the intersection to the interchange area may affect operation.

95

Based on the MOEs interchange area total time and downstream area total time,
MUDI operation, in most situations, is insensitive to the proximity of the closest
downstream node, while the SPUI operation is sensitive to the proximity of the closest
downstream node. In all cases, the performance of the MUDI configuration with a
separation of 1.6 kilometers (one mile) or 1.2 kilometers (three-quarter mile) either
equals or exceeds the operational performance of the SPUI. Furthermore, at higher
saturation levels, the operational performance of the SPUI configuration was adversely
affected by a separation of 0.8 kilometers (one-half mile). For both arterial cross-sections
(five-lane and seven-lane) and all three spacing scenarios of the downstream node, the
MUDI configuration showed no evidence of migration of delay. In addition, the SPUI
configuration with a five-lane cross-section and 1.6 kilometer (one mile) spacing also
showed no evidence of migration of delay to the downstream nodes. However, for higher
levels of saturation, all other SPUI configuration scenarios resulted in higher total times,

suggesting a migration of delay.

APPENDICES

APPENDIX A

96

$5.3. one: ABE. 3 $8080.... 9. £33. 093:8... 505.3 A532. 03323...
28:35 :35 535.2 2.. 9.3.02.— motaaoom 3:25.). .8 3.33. noun—=85 ...< 03......

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Na .2 5.8.: a... 8.. o. o. 2
N8 3.. 3.5%. an. 82 a... o. o.
B. 2.... 9.282 8.: 8mm 8 2 8
Mom a... 32.8 .R 8... 2. E o.
a: a. .35... 8.. 8.: m... 2 cm
3.. .6. .58.: $2 8.. 3 3 o.
N... R. 8.5%. a: 2.2 a... on o.
8.. RN $282 8... 8mm 2. 8 8
E a... 3.225 E 8.: m... o. om
a: 8 .39.. 8.. as. m... o. 8
c8. . z. :28... a... 83 Q. cm 2
a... a. 3.3%. 32 8% a... 8 E
on. 8. 8.53. 8o. 83 S. cm o.
EN mm. 338. E 8.: m... 8 o.
a: .M. .85... S. on... m... cm 2
9.8.. .9. .38.. .3. 2.32.2
08.... 08:. was... was...
.9... .3... 2.....o> 8......> 2.5.3. am 0.. 2.3. .x. a... 2.
8.... 8.... ass. 2.... be... 2:: ram. 5...: so... 8.32
Emubmgon. omgnobfi. SEE 3?. $55.85

 

 

4

0:5 . u 23.828:— .88.U 3 35.3.:

2:..-m n 53:32... 3.32
mono: nus—3.5 «no...» :52

 

 

97

$5.3. 0:2-.. .32.. C £20823 9. £33. owficofi SSE? 502.6 03.2.9.8:—

vcozaa 53.3 5&3...va 05 mag—oz.— motmcuow mam—352 8.. 333% nous—25m ”N4 2%....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

53 3.. 2.588. 9.2 coon c. on on
v? 3... $08.3. mum _ ovmm ad on om
owe ohm ovfloom. owe. ommm 56 on Om
com mm. 3%.? Km oom— md on om
ow. oo. 3.03% mo... owe. md on on
mmo m 3. 2.552: 9m. 83 o. cm cm
3:. 3m .3383 mam. ovmm ad em om
a8 NVm 3.388 98. ommm 56 cm om
wow. am. 2353 Kb cow. We om om
.w. No. .883 m3. owe. md om om
ova «D. .5588. 9.2 ooom o. on on
of. mow $05.06. mwm. ovmm ad cm on
9% gm ovflocm. one. ommm 56 cm on
mum 3. 23.58 Km oom— md Om on
ow. No. .onvm m9. owe. md cm on
.25.. .53 .25.. 55 033$...
0...:. 25% 2:3. 2:3.
.58. .80.. oE:_o> 0833/ oE=_o> Em .x. Emma .x. .901. .x.
.33.. mod... 9:3— 3.... him mxcfi >95 8.82 32”. 8.32
Ewobmcaom $52.88... .052 85.. 09.288:—

 

 

o=E ~ u 5:3285 «88.0 8 35.35

on“: n 3229.0 3?:

2.8.. $85... :5? :5:

 

 

98

..0..0..< 0.....-m A88. .. 88080.... 0.. .853. 09.80.... 82...? 59.3 08.8.0.8...
83.3 .80.. 0.9.8 0... 8.3.08. 8.8.03 88.0.52 .8 8.8.3. 8080.385 2...... 0...“...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nma «9. .5588... 3m. .83 8.. 85 cm
vmw 9... 888mm. mum. ovmm m... 85 cm
mwm 0mm 8.883. 8.8. ommm 5.8 85 8m
#3 8... £3.83 .55 co... m... 85 cm
3. 05 .383 3.. 8.8. m... 85 on
38 9... .5588... 9m. cow... 8.. cm om
NE. mm... 888%. mum. 8.8... 8.8 cm cm
9% omm 8.883. 8.8. ommm 5d om cm
mmm v... 88888 .55 co... m... cm om
v. . 85 .888 8.. 8.8. m... 8m 8m
mg 83. .5588». 9%. 880m 8.. cm 85
omw .8. 808%. mam. ovmm 8.8 cm 85
85m 8.. 8.888. 8.8. ommm 5.8 8m 85
0mm w... 8388 .55 co... m... cm 85
m . . .w .883 8.. cm... m... cm 85
.88.. 80>. $.59. :03 $898.85
08:. 08... 8.89 8......
.808 .06... 08:.o> 082.5 082.5 .mm .x. Em... .x. 80.. .x.
00.< 00.< Q83. 8...... 38m. 8...... EEM. 8.32 25.... .082
88088.33. 08.8.0.8... .082 «0.1... 855.8...

 

 

0:8 . n 8.3088... 3.03.0 o. 0088...:

0.....-m H 9.9.80.0 8.7.2
0.53. 09.28.... «:83 .95

 

 

99

.822 as.-. .2... 8 m=38... 3 .88.. 08.5.. 52...; .58. 8.3.22...

8...... .80.. 0.888 0... 8830.8. 8......00m 88.0802 .0. 8.300.. 8080.888 3...... 0.80...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8v. . 88m .5588. m3. 888 8.. 85 8
8.... vmm .8888. 8m. 88 8.8 85 8m
8... . M...» 8888. 88. 88mm 5.8 85 8...
85m 8N 88.888 .55 88. 8.8 85 8
8... .8. .888 M8... 88. .8 85 8m
88... 888 .5588. 8m. 888m 8.. 8 8m
88. Nmm 888$. 8m. 88 8.8 8 8
588 53. 8.888. 88. 8Nm~ 5.8 8 8
88m 58 88888 .55 88. m8 on 8
8. .8. .888 new 88. m8 8 8
v8. 88 .5588. 9%. 888m 8.. 8m 85
mm. . 8N8 .8888. 88m . 8.8m 8.8 88 85
888 88.. 8.888. 88. 88 5.8 8 85
85m 8m 8888 .55 88. m8 8 85
8. m8. .888 new 88. m8 8 85
.85.. ...0>. .880. ...0>. $888.85
08. .5 08. H 0...:... 8......
.898 .80... 08:.0> 08:.0> 08:.0> 8% o\° Em... .x. 80.. 8
08¢. 00.< 9...... 8...... .08.". 8...... .88.... .082 30... .082
88088.38. 08.8.0.8... .082 00....~ 0888.88...

 

 

0...... . n 8030080.... 3.0.0.0 0. 005......

0.....-5 H 83.80.... .082
£80.. 08580.... 80...» 5mm

 

 

100

iron/w on“: A2? 3 £20823 c; .mvmoz 0938580515 63288:—

wcofima 3:033; o5 wag—9:: mot—88w wan—352 8m 338M coca—nag ”m.< 2an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o S: omv €552: Sum _ 83 A: on om
in: mam 3983 mm»: ovmm ad on 0m
5w 9mm ovaoN— owe ommm 56 on om
mmv am 33.88 2.5 com: md on cm
2; on KNEE m3 Swo— md on em
2 2 X? IRES: mvfl oocm c; on em
2 2 2? $880— 32 ovmm $0 on em
Sm 53. casem— owS ommm 5o om cm
3m 0am £383 :5 com: md om om
NE 3w ENSVm mow one md om cm
mvmfi Hmm €583: $2 8me o; om cu
m :: at“ 3058— wwg ovmm ad em on
wow 03 ovflocfi SS 33 56 cm on
mam mwm £388 :8 com: md cm on
m: S SHE} mow owe md om on
952 53 $32 .55 5mg?
25H 25% ESE EEC.
38h .88. 0628/ oE=_o> oEEo> Em .x. Ewa .x. dog e\a
33 8.14. 95% 823 9cm 3:5 gum 5.82 >55 5?:
anbmgsom 032285 .552 32 09.228:—

 

 

o=E ~ n accuomaouam «mono—U 8 35:5
nun—-m n 39.320 3.32
maneu— ouaaneh «:33 Q7532:—

 

 

fiwog 23: A22: E 820823 e; .mvaom owficem 32E? .omfifibfi—
28:85 182:? H 05 wag—oz: moficoom wEBon 8m 833M noun—=85 “ed 053.

 

 

 

 

 

 

 

 

 

101

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mam; vow 535$ 32 down d._ on dm
mum fl nmv vaodmg «w? dvmm ad on dm
:2 new dvfldofi owe dmmm ad on om
mew mom odeda :5 8E md d5 dm
N2 hm. SQSm m3 22: md d5 dm
:3 anm :39: $2 doom d._ dm om
mi; mam waddmo— wwfl dvmm ad dm om
own: dmv dvfldcmfi dag dmmm ad om om
”we mum odeda EB do“: md om dm
NM: a2 185% new 9:: md on dm
om: owe €532: mvf doom d; cm on
$2 Sc vacdmf 3m— dvmm ad dm on
E Z mmm 33ch one dmmm 5d cm on
50 mom 0353 .2. ddM: md cm on
mm. m3 285% m3 owe md om on
was; .fié 9:5; .fiz 33$?
25% 08:. 2:3. mEE.
.88. .809 oE=_o> 0623/ oEEo> 8m .x. EwE .x. and .x.
83 83 95M 3:: him 835 Sam 8.82 32“— .8 .82
:aobmgoa 0935885 SEE 83. omSfiBBE

 

 

«=8 — N 5:838:— 2385 3 353m:—
o:a_-b n 6.3.58.5 .5an
3.3% oust—PE «:33 G755:—

 

 

102

.355 28; A23. 3 $8082.5— w; 630% 093:8...“ 5.5» ABE $532qu
98839 53.5 fiwEom2 ofi wag—oi: magnum 3:252 8.“ $33M noun—25m “a.< 03a...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wda mdv _aa\ddM: av: doom d._ adedo da dm

aam mam $982 an: dvmm ad EQde da om

mam mom $982 owe dmmm ad on <de da om

mg 5_ erdda _aa 2:: md amtddm da om

am am Swag mow one md aan— da dm

mmm owm Kidd“: $2 doom o._ adedo om om

NNv mdv vacdmo— w”: dvmm ad ENE: om cm

avM m m m 3&ch one dmmm ad dw <de om om

v3 v2 chdda _aa d9: md aNCddm dm dm

3 am :Hfidvm new 9:: md atom: on dm

8? mmm _aa\ddM: mvfl doom o; adedo om da

mmv w 9 «3de $2 dvmm ad $.ng om da

«mm aam dvflooN— 9:: deN ad 9239. dm da.

mm: NM: 9358 _aa do“: md am zoom om da

Va da 135% new owe md atom: dm da

G50: :9: 9:8: .53 $oQ$da $dQ$da
08: 08C. macs... was...
:38. 730,—. uE=_o> 0833/ oE:_o> “mm .x. uE=_o> Ema .x. as .x.
82 83. 95M 3:: him 3:5 Dam 8.32 03.5 32m SE2

8858509 omcasBSE SE2 835m 83 owfiaohxfi

 

 

 

 

 

9:8 — u 5:328:— 2820 3 358m:—
o=a_-m M 32320 .872
233% aunt—Eh as :uD2

 

 

103

flung 057a ABE. C £30823 e; .mumom owflcem .23 A52): 032285
9.955 :35 Swfioaz 2: mag—oi: motaaoom 3:0on 8m 938M cog—scam ”m4 033.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mam com .aadof $2 doom d. _ adedo 2. dm
mmv adv $383 $2 dvmm ad SQofi da om
vow _am 9&ch— dwe dmmm ad ow :de da om
8N wa_ 2353 fiaa 82 md am :ddm da on
2 _ ad_ 185% m3 owe md aa\dw_ da om
maa amm _aa\ddw_ mvfl doom d._ adedo dm om
Nav :% andmo— www.— dvmm ad 188$. om dm
vow d 3 dvfldofl dwdfi damn ad ow :de on dm
mmm Vdm odeda _aa do». md amtddm om dm
d: a3 1883 m3 22: md aadf on on
com amN _aa\ddw_ 32 doom 9. adedo dm da
_am NVM andNE mm»: dvmm ad 7:.ng om da
won 3m 3582 dad. dmmm ad ow <de om da.
mvm dam odeda _aa do“: md am :ddm om da
add _ _ d #353 m3 dwdfi md atom: dm ca
9.52 .5: 9:5; .53 §M§E §M§E
2:? 25a 25a mEPa
38a Boa oE=_o> oEEo> oE:_o> 8m .x. oEBo> Ema .x. cou— .x.
«we 82 95m 834 him $35 him 8qu ”Sta 32m 8.82
5353309 owfingefi SEE 835m 8E uwfinouofim

 

 

 

 

 

 

 

 

«=5 — n 5389.35 .885 3 3.33.5
can—a u 39.9.80 .8an

 

38¢ ems—Sub as new:

 

104

.352 23: fie. : 22082: 3 «.23 omega as, .929 03:55

535 «ion Bwfim 05 wag—oi: 85:on 3:322 8m $163— sedan—58E ”a.< 29¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mvfi wmm _aa\ddm_ Qum— ddcm d._ adedo da dm
33 gm vaoBmS ”a: dvmm ad KNEE da dm
ama mum vadofl dud" dmmm ad 8:de da om
mam EN 2358 fiaa do“: md am:ddm da dm
3: am 3.03% mew owe md atom: da om
a3: mmm .aa\ddw_ 32 down d._ adedo om om
m2 _ 0mm vacdmo— wwfl dvmm ad EQde om om
mom wam dvfldofi one dmmm . ad 8:de dm dm
wwm 9: odeda _aa d2: md amtddm dm dm
a: ma RNBVm mow owe md atom: dm om
awfl 2m _aa\ddw_ $2 doom d._ adedc om da
aad_ mom vaodmf £2 dvmm ad 3.63% dm da
vaa wmv dvfldom. ow? dmmm ad 2289 dm db.
mwm emm owmdda _aa 83 md amtddm dm da
a3 ca dede mow dwdfi md aadw— dm da
:52 :03 9:5; .5: §Q$E $332
oEF 25a 2.5a 2.5a
.maoa 13o a 0E:_o> 0823/ oE£o> 3m .x. oE=_o> 3me .x. tog .x.
83 meg 95% ea: him 915 Esm— SWE o>tQ 30E 8.32
EwomepoQ 0328an SEE 83.5w «oz omfinouowfi

 

 

 

 

 

 

 

 

 

 

0:5 _ u 3:823:— 3820 3 353m:—

o=«_-m n 39.820 3?:
€53— owauneh \3 5mm

 

 

105

.352 253 A2? a 320222 3 £95 0938; £3 528 093525

:35 “Son Bmfim 2: 9:302: morn—Bow 3:032 Sm 833m coca—=86 d_.< 033.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

aam_ cac Kidd“: Qum— ddcm d. fl adedo da dm
mafi mac $383 32 dvmm ad 285% da om
maa wav 9“?ch— dwe dmmm ad 3:de da om
mam vmm odeda _aa ddM: md amzddm da om
mm: v: SQde new 83 md atom: da dm
a _ 2 dma Kaddm. mvfl doom d; 5358 dm dm
mm: mac vawdmc— a»? dvmm ad SQog on dm
daS mav dvfldofi dud. dmmm ad 3:de dm cm
a; mmm odeda .aa 2:: md am<ddm om om
a: w 2 _mm\de mov owe md aa\dw_ on cm
X? .aw _aa\dd§ $2 doom d._ adedo dm da
aw: woo vaodmc— $2 dvmm ad _mm\d¢m dm da
a :: moo 3?ch 82 dmmm ad ow to? dm da.
a; dam odeda _aa do“: md amtddm dm da
Mn: dN_ SQde mow dag md atom: dm da
952 .55 932 .53 goexoa §Q§a
08F 25a mEFa 25a
30% ESE 0:575 0833/ oEBo> 3m .x. oEEo> Ema .x. and ea
83 «we 95% was: him 3:: Dim 8qu 03.5 32» 8.82
Emobmgoa $52235 8&2 outflow 82 owfifieflfi

 

 

0:6 — n 5339.8:— umomso 3 353m:—
o:a_-a u 6.8.58.5 .8732
333— oust—Eh a.» flaw

 

 

106

.Rtog 057m .63: E 320823 e; “mwaom 09355 55 .owqanohefi

988me Recumvfih 05 mEZoZ: 85:08 3:35: no.“ gamed sous—=56 ”__.< 035

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

vmfl dmm _aa\ddw_ Qum— ddom d._ adedc da dm
SN— :m vacdmc— awe dvmm ad Swag da cm
:a dam 3&ch owe dmmm ad dw to? da dm
mmv mam owmdda _aa 83 md amtddm da dm
w: mm Emdvm mew dwdfl md atom: da dm
aa_ _ awm _aa\ddw_ mvfl doom o; adedw dm om
av: amm vaodmm: mam“ dvmm ad EQOVm dm om
max new 3&ch dad“ damn ad 82de dm dm
amv ddv odeda .aa com: md amtddm om om
m3 .a SNBVm mow one md aaBM: om dm
mm” m _a _aa\ddw_ $2 doom d4 adedc dm da
a§ _ v3 vaodmo— mm? dvmm ad Edam om da
aaw wmm dvfldom— dwdfi dmmm ad ow :de om da.
5v 9% erdda _aa ddM: m d am <ddm dm da
a3 add _mm\dvm mow dad. m d atom: om da
952 .53 $52 53 §Q§E .xaexaa
oEF 25a mEPa mEFa
ESE 38a oE:_o> 05:15 2:23». 8m .x. 0823/ Ewa .x. as ea
m8< m2< Edam 9E3 Edam 3:5 beam 5.82 o>tQ 30E 8.82
Eaubmgofl $530th .552 83.5w 83 $53885

 

 

0:8 _ u soccer—8:— 23.20 8 353m:—
o=«_-m H 3839.0 Ema—Z
£8: guano:— E nzo2<E

 

 

107

.3593. 287a Any—ME 3 330523 c; £93m owflaoi 53> .omcmsououfi

28:35 32:3; 05 mag—oz: magnum wnzouoE 8m 333M noun—25m ”S .< 032;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2&— wmm _aa\ddw_ Qum— ddom d._ aondc da om
mag men 3983 an? dvmm ad Sada da om
am: wmv dvfiocN— dwdd dmmm ad ow Comv da om
a5 mom erdda _aa com: md amtdom da 3
a“: a2 ENBVW mow owe md atom: da om
aafl ado daaddw— mvfl doom o; adeoe om om
avfl mac vaSomc— mm»: dvmm ad :dem cm em
3: 0mm 958$ own: dmmm ad on <de om om
moo Gm odeda _aa d2: md amtdom dm om
mm: wE ENBVm 8v dwdfi md atom: on cm
52 9a _aa\dow_ Qum— ddcm d; aonoo om da
cm: Maw .3933 mm: dvmm ad .883 dm da
v2 _ vam ovfldoN— owe ommw ad on (c? om da.
c E 33 335% _aa d3: md am teem om da
ma_ m: ZQOVm new one md atom: om da
$52. :2; $52 $5 gong? $3.52
25a 25a mEPa 2.5;
38% 38a 08:75 oE=_o> 0833/ Km 8 uEEo> Ewa .X. «3 Rd
83 32 95M 3:: ham 3:: Edam 8.32 3cm BoE 5.32
5358500 “$8228:— 852 838m 8?. omSfiBBE

 

 

9::— u n 5339.8:— 2320 8 353m:—
o:a_-a u 3239.0 3.32
33¢ 0935.5 at 9202.5:

 

 

108

_
d

.355 05: fie. $3 $32.22 2 38m amaze; .853, .2936 omfineas
98:85 53“: «Ewing: 05 wag—9:: mew—macaw wfifivoz Sm fiimom cots—zfimm ”m _ .< 2an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ta «3 559: $2 83 3 2 8
Mac 8m $982 a: 8% mo E on
:e SN cases 82 8mm 3 E on
mom 3; 338 E as: no on em
5 3 :sea 8v 82 2 ca 8
8m 8m ESQ: $2 83 3 On 3
Sn 5.. $982 :3 can 2 cm 8
v? 5 958d 82 8mm 3 8 cm
m2 2. 3358 E as: no on on
v: mm 2%: m3 2:: no 8 on
23 new 552: $2 83 3 em 2
So am $282 $2 8% ad 8 2
N? 3m $382 32 8mm 3 on 2
0mm ca 2523 E as: we on E
m: 2 Egg m3 3:: no on E
was; :2; :52 5: $39.55
25a 25% 2:3. v.53.
3,: 35 25.5 o§_o> o§_o> 3m .x. 23 x «3 .x.
82 82 3am ea: ram ea: ram 8.32 so: 8.32
Emobmgoa omSfiBoE— BEE no.2 own—«:88:—

 

 

o=E Em u sateen—8:— .886 8 3:3»:—
o:a_-m N 38395 3.2.2

:33— ouauacah «:23 ~93:

 

 

109

.Rcofix 0.8—-5 Aozfi $8 322—55 N; .mdmom cuss—8r.— Sofig AHA—DEV owgnaoafi
28:85 5&3 5&322 05 9:23.: mogoom wfifidoz Ed 338% noun—25m ”3:4. 038‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vma w; €559: $2 doom d._ on dm
mt. v fl v vaedmm: mam. dvmm ad on cm
was amm 33ch one ommm ad on om
mom 2: chdda Kn 82 md on cm
m: we. EmBVm mow owe md on em
Ra wov Kidd“: m3. doom d._ on dm
ama aam anomS mam: dvmm ad om om
cam wmm 3&ch— owe dmmm ad om om
wwm am. 2358 Kn do“: md on on
NE E: :dem mow dwd. md om cm
3a 23 Shoe: Qum— doom o._ cm on
v: va vacdmo— mm 2 dvmm ad dm on
8m mam 3&ch d»? deN ad cm on
Rm SN odeda in com: md on on
d: R: SQofi m3 92: md dm on
$50; :05 $50; .nué £6?th
08F 25% £25. £5;
:38. 38% oEBo> 0829/ oEgo> 8m .x. Ema .x. mod o\o
«PE. «91¢ ESQ 3:3 Edam 8E5 him 5qu 30E 8.82
anbmgoa amass—885 852 82 03388:—

 

 

o=E in u 5539.3:— 3320 8 3.335

2.5-5 u 39.3.9.0 3?:

£59: ouaanoum «=83 R—DE

 

 

110

.3592 287m Any—ME $9 958822 m; .mdaom 093:on :55? .359 omgnohodfi
53.5 “50m Bwfim 05 mag—oi: moficoom mam—252 8m $38M mamas—58E ”24 visa

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

39 m3 Kaddfi 3m. doom 9. da om
wS _ 03 vawdmf wwmg ovmm ad da om
wmw mmm $38.2 one dmmm ad da dm
dam aa_ odeda _aa com: md da dm
m3 on 3.03% mow one .md da dm
dam: amv Kidd“: me _ doom o._ om om
ma: a; $383 wag dvmm ad on om
aaa V3 95003 one dmmm ad dm om
mam wmm odeda fiaa d2: md dm dm
9; mm 1353 Se own: md om om
Nvfi d3 Kidd»: mvfl doom d._ dm da
5: new vacdmf wwm ~ dvmm ad dm da
do» va dvmdofl owe dmmm ad on da
mam 0mm on Qdda _aa com: md om da
m3 3“ SQog mow owe md om da
3:5; .55 $50; :03 $dQ$oa
25a 25a meal mEPa
ESP an oE=_o> oE=_o> 0823/ Em .x. Ema x. «3 fix.
m2< 35 95M 8E5 Dam BEE Edam 8.52 32» 8.32
anmesoQ 032235 852 mug 0932885

 

 

 

 

 

 

 

«:8 SM n 5389.8:— 3330 8 353m:—

o:a_-m .I. 3239.0 3?:
253— owflueh «=33 flan—m

 

 

111

.332 23: Acme 32 22285 S .382 omega 3er 52.5 033235
.883 “Eon Bwfim 05 mag—oi: mogcoom $5252 8m 338m cots—25m ”24 03$.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a: a3 2282 $2 82 3 2 2
SE a 2922 32 $2. 2 2 2
E _ 2v 2582 22 22 3 2 2
Sq 2m 3252 E as: no 2 2
n: 2: :23 MS 22 no 2 2
mm: 5 22.2: $2 82. 3 2 2
a: an $922 32 $2 2 2 2
22 as $262 22 2mm 3 2 2
E 3». 0252 E as: we 2 2
m: 2: 2S3. 8v 22 no 2 2
8: Sc 2282 $2 82 A: 2 2
22 MS 3822 :2 $2. 3 2 2
22 a 9582 22 22 Z 2 2
5 Sm @252 E as: no 2 2
t: S d :23 a; 22 2 2 2
$52 .53 $52 .53 2222
25H. 25% men—H. man—H.
3,: 30a 055:3,. 2:23, o§_o> am 2 £5 2 «3 2
82 82 mama 825 Sum 925 $5 8.32 3oz 8.32
$058509 omgaaoucm .552 ME owfldseofli

 

 

 

 

 

 

«:8 Sn u 5389.8:— 3829 8 35329
«as: n 3239.0 3?:
3:3— owfiaoah «:23 5.5

 

 

112

.335 05$ .38 NE 922:2: 3 £32 omega Sofia debs: amazes:—
ESEfiQ :35 52:22 2: mag—oi: 882.com 3:082 .89 $38M :33385 ”24 2an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fl :: mow €539: mvm _ occm c; on em
o: gm $933 mm? 3% ad on om
Ymm cum ovflocfi own: ommm \‘d on om
KN mi 9338 Km 83 md on om
a: E. 18va m3 one md on on
Ra 5mm €559: 3m— ooom o; cm cm
can gm $982 wwfl ovmm md om om
wmm Ev $983 9:: ommm 5.0 om om
Nwm 03 £38ko KN. 82 md om cm
m2 5 _mm\ovm mow owe md cm on
voa com 528$ 32 oowm o; cm on
$0 mam $983 wwfl ovmm ad Om on
me awv ovflocfl owe ommm 5o cm on
cmm N2 8353 En com: md cm on
_m_ mm ZNBE m3 92: md om on
$50: .55 $50: :03 £om\o\oo>
25H 2:; 82:. mesh
38 h 38% 0823/ 2:39» oEEo> Em o\e Ema o\o dou— .x.
n2< 33 95% 3:5 Sam 8E5 bucm 8.82 30E 8.82
Emobmgon— 033885 852 8E emu—«:88:—

 

 

o=E N: u stoop—8:— .885 3 353m:—
u=a_-m u 3389.0 3?:

3:3”— oufifizm :5?» mg:

 

 

113

,
,

.352 on“: “Au—ME Q C 988823 ad 630% 093:8”— Eofi? A525 033825
98:85 :35 59:82 05 9:335 motdaoom mam—282 8m 9.38M sowssgm ”a _ .< 2an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ama N 3 Kiddfi mvfl down d._ on om
www wmm $833 32 dvmm ad on on
$6 dmm vadeNH owe dmmm ad on on
3m 3: oonda Km 83 md on om
om: d3 36de mow owed ad on om
wma mmv Stood" 3m d doom o; om cm
Now Ev vaeBNS 32 dvmm ad om om
go now 9‘3ch— dmd. an ad cm on
man 3: 9338 :5 83 md om dm
:— :: .deVm mew owed md om om
Kw mwv Kidd? $2 doom d._ on on
o; w? vaoSNo— wwfl dvmm ad on on
wvm mmv $389 on? dmmm ad on on
03 ma_ chdda 2k. do“: md dm cm
I." 8— SQOVm m3 22: md cm on
$55: had $50; :03 fadflfcdn
25H 25% mFBH mEP—l
120% 30% 053:5 0823/ 0823/ 8m .x. Ema .x. cox— .x.
82 82 gang 9.5 ram 825 ram 5.32 3oz 8E2
Ewobmgofl owamnohodfl 552 32 owcwnBBE

 

 

«:3 N: N 53823:— 3830 3 3.53.:—

25: u 39.320 3.32

2.2% «was; :5? :52

 

 

114

.EEE 287m .33: Q S 380829 ad £98m owfiaoi 5933 ABLE omfiauhzfi
SBHD gem 035m 2: wEZoZ: mongoom wczouoz Ho.“ £38m noun—25m “ajx 23,—.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nam“ adm _aa\ddw_ mvfl doom d._ da cm
2 2 8m vacdmf mwm _ dvmm ad da om
ddw wmv ©3803 owe dmmm ad da om
vam mm: ow Qdda _aa cow“ md da dm
9: aa dmfiowm 8v 2:: md da dm
.mm_ 9% _aa\ddw~ mvfl doom o; dm om
am: m _ m vacdwf an: dvmm ad cm om
mwo amm dvmdom— owe dmmm ad dm on
_am mmm odeda _aa do“: md om om
d2 5 Rdem m9. 9:: ad om on
32 o _ m Eaddfi av: doom o; om oa
ad: Em vaodmf mm: dvmm ad cm da
caa _av vadoN— one dmmm ad dm da
mmm mmm chdoa _aa 83 md om ca
9; mm ~de¢m mow owe. ‘ md cm da
950: .55 $50; :03 §dm>§da
25a 25% ESE mEPa
50a :38. oE=_o> 0823/ o830> “mm .x. 39M .x. to; .x.
m2< m8< 95m 3:: Edam 3:5 Edam 8.32 32m 8.82
8353300 own—£88:— 552 83 omfifiSBE

 

 

 

 

 

 

 

0::— Q— N 5539—8:— 3320 3 3:8»:—

o:a_-m u 39.230 .5732
mini aunt—9:— 323 5mm

 

 

115

.Rtog 237a .33: Q: £08823 wd .mwmom owflcem 305m? .938 «98:80::

52:3 «Sam Bwfim 05 mgr—oz: mongoom mam—02:2 you 838% noun—:85 dm.< oEfia

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a3 _ mam _aa\ddw_ $2 doom d4 da dm
am: new vacdmf «mm. dvmm ad da dm
3.3 dadv dvfldofi one ommm ad da om
va .am @353 _aa com: md da dm
mm. m3 3.0.53 mow owe md da om
d _ m_ dmo _aa\ddw_ $2 doom d._ on om
KS 36 vacdmo— me dvmm ad on em
80— va dvfldom— owe dmmm ad dm om
cav mam odeda _aa 2:: md om dm
3: a3 RQde new owe ad on on
wmfl 3a _aa\ddw_ $2 doom d._ dm da
aadfi w _ m vanmo. mm? dwmm ad om da
d2: 2% dvfidcm— 2:: dmmm ad on da
a 3 3m wadda _aa ddf m d om da
Q: d. _ ENEVm mow own: md om da
$50: .503 $50: .53 oaonada
25a 25a 8.5a 35a
30,—. 30,—. 0828/ 0833» 0838/ 3m o\.. 3me ca tog .x.
82 8.3. 95% .635 abam 3:5 Cam 8.82 263 8.82
Emobmgoa 03328:— 852 32 03.3285

 

 

0:8 N: N 33838:— «885 3 353m:—
u:a_..a u daemmeu 3?:
38¢ away—5.:— 323 Sam

 

 

APPENDIX B

116

$00

8.5. Sam .22. .0 5:233 2322.

$2

$8

 

 

 

 

\

 

8
m

 

 

 

 

 

8
we

 

 

 

 

§

 

 

 

 

 

mam; 65.55 +
9.61.. .me III.
0cm_.m .535. lbl

§

 

 

 

 

 

 

.2532 0:2.» A22... 5 9.32:0...— 9.9
.33". 38:2“. «:23 .25.... to. $2. .3“. 2...... .80... a2< owns—.83... ".mm 2.5.“.

.

(unoq woman) own [3301, am ofiucqoaouu

117

$8

9...... Eucm, 3?: .0 5:83.20. «.322.
$2.

 

 

 

 

 

 

8
N

 

 

 

 

 

.3.

 

 

 

 

 

§

.

 

 

 

men...» .3350 IT

 

 

 

I ,1 . was-.. 5.6+
8a.... .5221

 

g 8
1’
(unoq woman) emu .3301, saw ofluquouu

 

 

 

 

 

 

 

.23.... 23$ 3.... a 222:2... 3
.253. 38:2“. «:23 .oEa... to. $3 .3". 2...... .50... a2< 99.2.23... "ad 959“.

118

$8

9...... Pam 3...: .o 5.8.26 .52....
$2. $8

 

 

 

 

-\\\

 

 

 

 

 

 

 

 

957m .2955 Lil
957m .SQm III

 

 

 

 

 

2.36 .532 IOI

 

 

 

 

 

 

 

 

3.8... 25$ ..o..E S 33050.... Q.
.33: 35:9... 323 .25.... to. $3 .0“. 2...... .80... a2< 3:228... and 2.6...

§ § § §

(unoq opmon) emu .330; any oBuluoaouu

O

119

9...... gem .0?! .0 5.3.0.0.... 2.3.0..
:8 . . $2. $8

$8

 

 

 

 

8F

 

8
N
Bueqmem.

 

\.\\\

§

 

 

\.V\\\ \

 

 

 

 

 

mam: 0:00.05 LT.
mam—K .me III

 

 

 

 

 

 

 

 

mam—é .525. IOI

 

 

 

 

och

 

 

.53.... 2...... ....=... s 2225.... 0..
.0000". 008:0... .00.... .25.... to. $0» .0“. 2...... .80... 02< 09.2.0.3... in 205...

com

Q

(unoq epmen) emu "301, eer e

120

$8

9...... .55 3.02 .0 0030.35 .0020.

$2. $8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0:0... 0:02.05 Lal
0cm... .me III
mcm_-n .522 +

 

 

 

 

 

 

 

 

 

.23.... 2.0.... ..o_.... s 22050.... 0..
.0003. 055:0... .00.... .25.... to. $3 .0. 05F .30... 00... 09.2.0.3... .md 0.09.

§

§ § 8
. co
(unoq epmen) emu I930. ceav eBuequeam

§

121

‘\
\

$8

9...... .35 .0?! .0 0030.23 2.00.0.
$0.. $00

$8

 

 

cow

 

\ \

8N

 

 

¥.\

 

 

 

 

8v

 

 

 

 

 

000.-.. 000505 If
0:0... ._Dn.w III

 

 

 

 

0:0: .532 +

 

 

 

 

 

 

con

 

com

 

.23.... 2...... ..0__... 5 0.0.00.0... 0..
.0000... 095...... 2.0.2. .25.... c0. $00 .0. 0...... .80h 00..< 09.2.0.2... nod 2.5..

(anon clam-A) emu mo; Imv éfiunumum

122

$8

9...... 0.... 3...... .0 0008300 0.8....

$2. $8

 

 

 

 

 

 

 

I‘ll! Y

 

 

 

 

t—w

 

 

 

 

 

 

 

 

000.-.. 0000.05 +
000.6 ..Dn.mlll
000...... .532IOI

 

 

 

 

.0t0...< 0:0... 40...... S 0.3053... a...

.0003. 000.00.. .00.... .000... :0.— $0.. .0. 0.0.... .50... 00.< 50030030.. "hm 0.50..

oov

com

80.

com.

1. Go:

.82

8a..

(unoq emqu) emu ”no; new memeumoa

123

0...... .55 3.0: .0 00.3.5.3 .0020.
$8 $2. $8

 

 

 

 

§

 

§

§

 

 

,§_

 

 

Q

8
s:

 

 

 

 

 

 

 

 

 

 

 

9.0.6 0:00.05 I0...

 

 

mam—A... ._Dn.m III
mcflum .EDEIOI

 

(unoq apnea) emu 1030,], any muonsumoa

 

 

 

 

 

 

 

 

3.2.... 2....-. ..¢..... 5 £225.... 0..
.0000”. 38:0... .30.... .05.; #0.. $3 .0“. 0...... .80.. no.3 E33530 3d 2.6.“.

9.0... .30. 3.02 .0 02.20.00 .0020.
$2. $8

.$on

L

 

 

k...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000.6 0000.05 Lil

 

. . _--|11II-I--& 000.0 5.5+

000.6 .535. IT

 

 

 

 

 

 

 

 

.002... 00...... .00... s 0.825.... 0..
.0000”. 000.00... .000... .005... #0.. ax...” .0. 0.0.... .30.. 02< 0.02.0032. 6.0 2.5..

com

com

coo.

.ooww

000.

com.

com.

(unoq 0|0|q0Ar0w| 1, “no; 001v unansumoa

9.0.. .30. 3.0: .0 00:20.00 .0020.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$8 $2 $8 $8
0
\+ com
\ \ 8..
\\V 80
5
H \ \ 80
80.
80.
lzlnl 1 l 83
000.- 5 000.005 I0|
000.... ..:.m III 80.
000.-. .535. +
80.

.0..0..< 000.... 40...: 5 0.0.052... c...
.0000”. 000.00.. .023 .0E.... :0... $0.. .0. 0.0:. .80.. 02< 0.02.0038 "Sum 2.6..

(anon apnea) emu, '030; 001v wunsumoa

126

9.0... .55 00.0.2 .0 02.0.5.3 .0020.
$8 $2 $8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000.... .000E05 lil
000.-. .S.w III
000.... .522 IOI

 

 

 

 

 

 

 

 

 

 

 

 

.0022 2.0.... .30.. s 00.052: 0..
.0003. 050.02. .00.! .020... #0.. $.00 .0. 02F .80... 02¢ 202.0032. "Sum 0.50..

(unoq woman) emu I030; 001v mamsumoa

com.

127

8.0... b.0m 3.0! .0 02.0.0.5 .0020.
$8 $2 $8

 

 

 

 

 

 

 

 

 

 

 

 

 

-1

 

 

 

 

 

000: 0000.05 Lal

 

000.-. .53.). +

 

 

 

 

 

 

_
fl _ ii: f,i--:-!i-.§-- is. 000... 5.0:...
_

 

 

.0..0..< 0:0... ..0...0 .. 0.0.00.0... 0...
.0000. 000.02. .00.; .005... #0.. $.00 .0. 0.0.... .0.0... 02< 202.0038 “a ...m 0......

$0M

 

com.

83

08..

com.

(anon apnea) emu [0301 may mumsumoa

128

$8

9...... 0.0m. 3.0: .0 0000.500 0.8....

$0. *8

 

 

 

 

 

 

 

 

 

 

 

 

 

000.6 6000.05 Ial
000.6 .5mm III

 

 

 

 

000.6 .532 IOI

 

 

 

 

 

 

 

.008... 000.... ..0_..0 ... 0.0.00.2... 0...
.0000. 000.02. 0...... .02.... 00... $0. .0. 0.0.... .0.0._. 00.< 0000020.:— unrd 0......

 

§ § § §
(anon O|O|I.|QA) emu "no; 001v 06u0q0101u|

§

§

com

129

8.0... 0.00 3.0! .0 02.20.00 .0020.
$8 $2 $8

 

 

 

 

 

\\ \

 

 

 

 

 

 

00070 000.005 +
00070 ..Dn.mll|
000.6 .522 IOI

 

 

 

 

 

 

 

 

 

.002... 2.0.-.. .20.. 0 0.0.00.2... 3
.0003. 000.02. 0...: .020 h 00.. $00 .0. 0.0.... .0.0._. 02¢ 00000280. "3.0 0.00..

8..

8w

co.

coo

emu, '00; 001v ofiunqmmm

130

9...... 0.00 3.00 .0 50.0.0.0 0.8....
$00 08.. $00

$8

 

 

 

 

oo—

 

 

com

 

 

 

 

 

 

 

 

 

 

 

000.6 000.005 L0!
000.6 ._...n.wlll

 

A'll‘! |I|l1|l||4|lilllllL I El.l [lily II'I.

 

 

 

000.6 .522 IOI

 

 

 

 

 

 

 

80

 

.002... 2.0.... 02.... 0 23003.... 0...
.0000”. 000.02. 0...... .020... 000 $00 .0. 0.0.... .0.0... 00.< 00000202. ”0.0 0.00..

(unoq woman) emu mm; 001v ofiuuqmam

131

000... 00.00 .0?! .0 02.20.00 .0020.

$0.. $06 $00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000.-. 000.005 L0...
000.-. .me III
000.-. .535. I6.-

 

 

 

 

 

 

 

 

 

 

.252 2.0... 02.... 0 20.202... 0..
.0000: 000.02. 0..., .020... 00.. $0.. .0. 0.0.... .500 02< 00000280. “09.0 200..

00»

(anon emnen) emu "no; 001v efiucnzuezul

132

$00

000... 2.00 .202 .0 02.20.00 .0020.
$2 $8

 

 

 

 

 

 

 

 

§

 

.3,

 

 

§

 

 

 

 

 

 

000: 000.005 Ial
000.-. .5n.m Ill

§ §
(anon e|0|neA) emu '010 J, eav efluenzueam

 

 

 

 

 

000.-. .53.). +

 

 

 

 

 

.000..¢ 000.... 00:0. ... 0.0.00.2... 0...
.0003. 000.02. 0...... .020 .. 00.. $00 .0. 02.... .0.0._. 02¢ 00000220. K .d 200..

133

$8

9.0... b.0m .3?! .o 02.0.0.3 .0020.".
$2. .08 $8

 

 

 

 

 

 

 

 

\\ g

 

 

 

 

 

 

 

 

 

 

0094. 60005.0 lil
$09-» ._Dn_wlll

 

 

 

 

 

con

 

 

$05-5 .532 IOI

 

 

 

 

 

o8
3032 ea... .20.. s 2223.... 3
.0031 09308.... 5.3 60.0... to. $0» .8“. 20F .53 32 00000280. ”Sun 200....

 

(anon emu.» emu "no; em eBunqcuozm

134

$00

9.0... 2.00 3.0! .o 02.20.00 .0020...

$00-

$00

$00

L

 

 

|\\\

 

lo
8N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.2-0 00000.5 +
$00.6 ._Dn.m Ill
$00.6 .535. IOI

 

 

 

 

 

. 3.03.5 $0070 A220 5 232.00.... 0...
0003. 000.020 0.00 .020... .00.. so» .80 20¢ .89.. u2< 202.9032. "9.0 200.0

com

coo.

oom—

oov _.

com.

com.

(anon woman) emu '00; eav wcaxsumoa

135

$00

9.0... 2.0m .20: .o 02.20.00 .0020....
$2 .08

$00

 

.\\*.
\. 80

 

 

 

 

 

.80

 

 

 

 

g

‘

 

 

 

 

 

$00.0 0000.05 [1|

§

!
8
::

 

 

1- 1 9.0-0 500101

 

 

 

$0070 2632+

 

 

 

 

 

.0000... 2.0.... .20.. s 0.3052... 0..
.0003. 000.020 0.05 .020 0. .00.. $00 .60 02.... .03... 02¢ 202.0032. ”00.0 200.0

80.

(anon C'OIMOA) can “no; say wamsumoa

00.0... 2.00 00.0: .o 02.20.00 .0020..
$8 $2 $2.. $8

L

 

\\\..

 

 

 

 

\ ..

 

\ 8.

 

136

\\\ 8.

 

\\ comp

 

 

 

L-. . .. . i l - ..-- Il-|- - - co:

 

000.0 6.5205 +
0.03 500+ 82
9.0-... 502+

 

 

 

 

 

 

 

 

 

 

80..

.0032 2.2-... 22.... 0 20.2.2... 0..
.0003. 000.020 0.05 .020 .P .00.. $00 30 02.... .20... 02¢ 202.0032. 30.0 200.0

(anon ounrnon) can 1, "301 am waazsumoa

137

0000

00.0... 2.00 3.0: .0 02.20.00 .0020..

.00»

0000

0000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000: 05205 If
$00.4. .500 III
$0070 .522 IOI

(anon omnoa) emu Iago-L nav wamsumoa

I
§

l
8
z.

 

 

 

000.

 

 

. .0022 2.0.... 22.0. 0. 0.32.2... 0..
0080 0000.20 .0.; .05.... 0.... $2 .00 0...: .80.. 02< 502.0038 .000 2:0...

com.

138

$00

00.0... 2.00 3.0: .o 02.20.00 .00200
$2 $00

$00

 

 

 

com

.80

 

 

 

 

 

coop

 

 

comp

 

 

 

 

 

 

. 000.-.. 600205 L0...
$007K .Smw III

 

 

 

 

$00: .522 IOI

(anon omnon) emu Iago; aav waazsumoa

 

 

 

 

 

coo.

 

 

. .008... 2.0.... 00...... 0. 0.3.0.0.... 0..
0.003. 000.020 0...... .020... .00.. $00 .30 02.... .03.. 02¢ 202.0032. "00.0 200.0

83

139

$00

020... 2.00 8.02 8 02.20.00 .0020...

30h

. $.00

 

 

 

 

 

 

 

 

 

 

 

 

000—

 

 

 

com.

 

 

 

 

000: 0820.0 Lal
$00: .500 +
$00.; .535. IOI

00v.

 

 

 

000.

 

 

.2032 002-0 A220 3 20.022... 0...

.0080 000.020 0...... .020 0. .000 $00 .80 02.... .08... 02¢ 202.0280 "00.0 200.0

(anon emnen) emu '00; say mamsumoa

140

9...... .0000 8.0: .0 00080.00 .08....
$8 $2 $00

$00

 

 

 

.00..

 

 

 

§

 

 

 

 

§

 

 

 

 

 

§

§

 

 

x O.=E N\—. .OCNTD JDQW+

f1- 1 1-! ., Ii- 0...: «.0 .0001... 0000+

0...... . 000.0 ._000lxl
0...: 0.. .0020 ..o0....l0l
II 3 111! - l , ; I 0...: 0.0 6.0.0 800+
0...: . .0000 802+

 

 

 

 

 

§

 

 

 

 

 

 

 

 

 

02.00000 00.0000 00.20>
..0...0.._¢ 002-0 .0080 000.020 .02.... .020 h 00.. $0» .8. 02.... .08... 02¢ 00000202. “00.0 200.0

(”"0" "9M“, “I'll. mo; WV “UN-Imam

O

141

9...... .500 .200 .0 50.0.08 0.8.00
$00 . $2 $00

$0.”

 

 

 

 

00.

 

 

 

 

.3.

 

 

 

 

 

§

 

 

§

 

 

 

Aw\ . . f - :-i 2.2 N: 602-... ._300+

22 30 602-0 ..300+

2.2. 602-0 ..300+
22 02 602-0 ._032+
22 30 602-0 ..032+
22 F 602-0 0022+

 

 

 

 

§

 

8
CD

 

 

 

 

 

 

 

000

 

02.00000 00.0000 00.20>
._0..0..¢ 002-0 .0080 000.020 .022. .020 0. 00... 800 .o. 02.... .08... 02¢ 00000202. "00.0 200.0

(anon own») cum mu m 06mm:

A

142

9...... .55 .28. .o 8:235 2.8....
$8 $2 $8

$8

 

 

 

 

 

 

 

 

 

 

.2:

I
i
I
I
§

g:

 

 

 

 

 

 

§

 

 

0:2 N2. 602-m ._Dn_m+
0:2 3n 602$.- .Smm-III

 

0:2 w 602-m ._Dmml*|
0:2 N: 602-m 2032+
..II E:- til-1 - # ii -Jl all 2.2 3m 602-m 2032+

 

 

 

 

 

 

 

 

0:2 P 602-m 2032+

 

§

 

§

 

 

 

 

 

 

 

com

 

302.com @20an u0_>.n>
22.83. 2.2-.» .33.... «0802.. «:83 62:... ca.— .x...” .2 20:. .80... «92 3020.20. find 202....

(um Glow-A) own I'm. WV can-amount

9...... Fem 3...: .0 5.33..» .52....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$8 $2 $8 $8
. o
- r 8.
8.
8n
3
4
l
8..
1| ’ I i 1.1-II]: l- com
2.... a: .25.-. .SawIOI
9.... «a .25.-. ..:n.ml....l
u - - - - - - - Iii-I :. i IL 2.... . .23.-. .Sawlxl 8o
2.... m: .05.-. ._o.._...+
2.... an .25.-. .EDSIII
9.... . .25... ..o:...lol
8.
3023» @2025 9.3.;

._2.3.< 2.2... .233. 00800.... «:23 £20... to.— $3 .8 00....- _So._. a2< 09.2.9.8... "an-m 2:2".

(“"0“ OWN“) “III-l. Ina-l. M “Bu-Imam

144

9...: .0.... .0.»: .o 8.3.38 .08....

$8 $9. {com $0M

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.2 N... 602-5 ..Dn.w+
2.2 v8 602-» .Sme-XI.
2.0.. F 602-... gnaw-III

 

 

 

 

I t: 0...: N: .002-.. 332+
2.... 3n 602-. .EDEIII
2.2 .- 602-5 2032+

 

 

 

 

 

 

 

 

matacoom 00.8.5 9.00;
22.2... 2.2-.. .33: 00500.... «so...» .203 00.. ox...» .3 2...... .50... 02¢ 09.2.82... "mu-m 2.5....

8.

§

(“"0" Claw“) “all. "301 CW .50"!le

§

§

can

con

145

9...: .000 .0.... .0 500.300 2.8.9.
$8 $2 $8

$8

 

 

 

 

§

 

 

 

.3-

 

 

 

 

§

 

 

 

o
S

 

 

2.2 N: 602-. ..Dn.m|.l.
2.2 in 6024. .Smmlll
2.2 w 602$ .505le

 

 

 

2.2 N... 602-.- ._035+
2.2 VB 602.5 2032+
2.2 w 602-5 2032+

 

§

 

 

 

 

 

 

 

 

02.0008 00.025 00.00)
22.0.... 000.... .0003. 00502.... 0.0.... .02....- 00. $00 .0.. 02¢ .80... 02¢ 000000.30. "on-m 0.00.”.

(""0" “3““) CHILI. "301. “V WWI

146

9...... .0.... 8.3. .0 808300 0.8....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$8 $2 $8 $8
0
- o2
\ 8.
08
8..
82
o..... 0.. .803 .0.-8+
- i I it- 2.... v... 6......0 ._0..0|..T o8.
0...... . 68.0 ._0n.0lxl
0...... 0.. .80.... .50....I0I
- IL- - 0...: «a .08.-.. ..00.2III
il 2.... . .20.... 300+ 8...
08.
00000000 00.0000 00...;

..0..0...< 002-0 .0003. 000.00.“. .00.... .020 h 00.. $0.. .0. 0.0: .80... 00.2 800300300 306 0.00...

(”M DION“) Cull-I. “301. 3“ W0

0

147

9...... 00.0 .0.... .0 02.2300 0.8....

$00 $0.. $00 $00

 

 

 

 

 

 

 

 

 

 

 

 

 

2.2 N2 602-0 2300+
2.2 30 602-0 2300+

 

 

 

 

 

a . 2.2 .. 602-0 2300*
2.2 N: 602-0 2032+
2.2 30 602-0 2022‘:-

 

 

 

 

 

i-i- - - ; x - :ll-Ti -.:|,-. iii ilwiiillft-I 1. , 0:9. r 620% ._032+ lIIIII-

 

 

02.00000 00.0000 00.b0>
._0..0..< 002.0 .0003. 00200.0 .00.... .020 h 00.. $0... .0. 00....- .20... 00..< 0.02.0033 nun-m 0.00.0

.SN

80

SN.

8!.

ooo F

(""0" NON“) “"11 "301 M W0

148

00.2-. gm .3." .0 02.05.00 .08....
$00 $00 $00

$00

 

 

 

\\

\, 08

00v

 

 

 

 

000

 

 

é

 

 

 

 

 

 

 

2.2 N: 602.0 ...-..00+
2.2 30 602.0 2200+

 

2.2 F 602-0 2200*
2.2 N2. 6020 2025+
2.2 30 6020 2022+

§

 

 

 

2.2 — 6020 2032+

 

co:

 

000w

 

02.00000 00.0000 00.E0>

._0..0..< 002-0 .0000... 000.020 .00.... .020 0. .00.. $00 .0. 0.0.... .20... 00.< 0.02.0032. "00.0 0.00.0

(""0“ "ON“, mu. "301. M W

149

$00

9.5.. 5.0 3.0: .8 802300 0.8:...
$2

$00

$00

 

 

 

 

 

00w

 

 

 

 

 

 

\\

 

 

\\

 

 

 

 

  

 

 

 

0\

 

 

 

 

0:5 N2. 609$ .5Q0IOI
9...: 30 600.4. ..3Q0+
0...... p 600.-» .5n.0|*|
3.0. N: 600: .5357?
0...: 30 600..n .EDEIII
0...: p 60075 .EDEIOI

(mum: 0.01000) «nu, I001, um «ammo

 

 

 

 

 

 

00.000000 00.0000 00.b0>

..0...0:< 000..» .0003. 000.00.“. 50.3 .0000... 00.. $05 000 000.... .000... 02< 800000300 3.0.0 0.0.0....

9.5.. 00.0 3.8. .0 502.000 0520..
$8 $2 0.8

$00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

150

 

 

 

 

 

 

 

2.0. N: 600..h ._Dn.w+

 

 

 

2.8 :0 .20: 500+
. 2.5 _ .202 500+
_ .- -. - , . . : 1 . 0...: 0: .20-:032Iil

 

003

 

2.0.. _. 600.4. gnu—35+

 

_ ,- . 2.5 an .22... 0002+

 

 

 

 

 

000p

 

000w

 

00000000 00.0000 00.E0>
._0..00< 000:. .0003. 000.02.... 000.3 .0000... 00.. $00 00. 00.... .000... 00..< 8000000300 "00.0 0.50.“.

(anon Glam-A) Own Inc; M tun-mama

151

9...... .00.... .200 .0 5.00.300 0.8....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.8 .8» .00... .08
o
a .\ 80
. 80
\ \

i \ \\ L- I- ..
ii..i:.-.1, 8.:
l! I 1:11! 8....

, 1 2.... 0: .20.-» 0000+
. w 2.... :0 .20.-. .500le 83

m __ 2.... 0 .20.-» ..0n.0lxl
- , W t l I Villll 2.... N: .20.-. ..000+ .11- 80.

_ 2.... in .20.-.. 0000+

2.... 0 .20.-. 0000+
8.:

 

 

 

00000000 00.0000 00.E0>

._0..00.< 000.-» .0003. 000000... 000.? .00.0... 00.. $00 .00 0.0.... .000... 00.< 0.00.000300 "00.0 0.00.0

(unoq clam“) omu Ina; any ma

LIST OF REFERENCES

10.

LIST OF REFERENCES

AASH-TO. A Policy on Geometric Design of Highways and Streets. American
Association of State Highway and Transportation Officials, Washington D.C., 1994.

Abbey, Lester and Glen S. Thurgood. Design and Operational Features of the Single
Point Urban Interchange. Utah Department of Transportation, Salt Lake City, Utah,
June 1991.

Bonneson, James A., and Carroll J. Messer. A National Survey of Single-Point Urban
Interchanges. Texas Transportation Institute, College Station, Texas, March 1989.

Bonneson, James A Bridge Size and Clearance Time of the Single Point Urban
Interchange. Highway Division of the American Society of Civil Engineers, New
York, NY, November 1991.

Bonneson, James A Study of Headway and Lost Time at Single Point Urban

Interchanges. Transportation Research Record 1365, National Research Council,
Washington D.C., 1992, pp. 30-39.

Bonneson, James A. Operational Efi‘iciency of the Single Point Urban Interchange.
ITE Journal, Vol. 62, Number 6, Washington D.C., June 1992, pp. 23-28.

Bonneson, James A. Change Intervals and Lost Time at Single Point Urban
Interchanges. Journal of Transportation Engineering, Vol. 118, Number 5,
Washington D.C., September/October 1992, pp. 631-650.

Dorothy, Paul W., Thomas Maleck, and Laura Aylsworth-Bonzelet The Operational
Aspects of the Michigan Design for Divided Highways. Transportation Research
Record 1579: Geometric Design and Its Effects on Operations, Transportation
Research Board, Washington D.C., 1997.

FHWA. T RAF User Reference Guide. US. Department of Transportation, F HWA,
April 1993.

Fowler, Brian C. An Operational Comparison of the Single-Point Urban and Tight-

Diamond Interchanges. ITE Journal, Vol. 63, Number 4, Washington D.C., April,
1993, pp. 19-24.

152

11.

12.

13.

14.

15.

153

Hawkes, T. W., 111 and Michael D. Falini. The Urban Interchange. Public Works
Journal, Vol. 118, Number 2, Ridgewood, NJ, February 1987, pp. 45-46.

Leisch, J. P. Innovations of Diamond Interchanges in the Urban Environment. ITE
Journal, Vol. 55, Number 11, Washington D.C., November 1985, pp. 24-26.

Leisch, J. P., T. Urbanik and J. P. Oxley. A Comparison of Two Diamond
Interchange Forms in Urban Areas. ITE Journal, Vol. 59, Number 5, Washington
D.C., May 1989, pp. 21-27.

Messer, Carroll, J ., and James A. Bonneson Single Point Urban Interchange Design
and Operations Analysis. National Cooperative Highway Research Program Report
345, Transportation Research Board, National Research Council, Washington D.C.,
1991.

Merritt, David R. Geometric Design Features of Single Point Urban Interchanges.
Transportation Research Board, Transportation Research Record 1385, Washington
D.C., 1993, pp. 112-120.