National Academies Press: OpenBook

Smoothness Specifications for Pavements: Final Report (1997)

Chapter: Chapter 4. Results of Smoothness Data Analysis

« Previous: Chapter 3. Data Base Development
Page 41
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 41
Page 42
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 42
Page 43
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 43
Page 44
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 44
Page 45
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 45
Page 46
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 46
Page 47
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 47
Page 48
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 48
Page 49
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 49
Page 50
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 50
Page 51
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 51
Page 52
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 52
Page 53
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 53
Page 54
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 54
Page 55
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 55
Page 56
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 56
Page 57
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 57
Page 58
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 58
Page 59
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 59
Page 60
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 60
Page 61
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 61
Page 62
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 62
Page 63
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 63
Page 64
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 64
Page 65
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 65
Page 66
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 66
Page 67
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 67
Page 68
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 68
Page 69
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 69
Page 70
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 70
Page 71
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 71
Page 72
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 72
Page 73
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 73
Page 74
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 74
Page 75
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 75
Page 76
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 76
Page 77
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 77
Page 78
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 78
Page 79
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 79
Page 80
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 80
Page 81
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 81
Page 82
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 82
Page 83
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 83
Page 84
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 84
Page 85
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 85
Page 86
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 86
Page 87
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 87
Page 88
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 88
Page 89
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 89
Page 90
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 90
Page 91
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 91
Page 92
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 92
Page 93
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 93
Page 94
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 94
Page 95
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 95
Page 96
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 96
Page 97
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 97
Page 98
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 98
Page 99
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 99
Page 100
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 100
Page 101
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 101
Page 102
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 102
Page 103
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 103
Page 104
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 104
Page 105
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 105
Page 106
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 106
Page 107
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 107
Page 108
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 108
Page 109
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 109
Page 110
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 110
Page 111
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 111
Page 112
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 112
Page 113
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 113
Page 114
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 114
Page 115
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 115
Page 116
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 116
Page 117
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 117
Page 118
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 118
Page 119
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 119
Page 120
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 120
Page 121
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 121
Page 122
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 122
Page 123
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 123
Page 124
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 124
Page 125
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 125
Page 126
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 126
Page 127
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 127
Page 128
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 128
Page 129
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 129
Page 130
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 130
Page 131
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 131
Page 132
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 132
Page 133
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 133
Page 134
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 134
Page 135
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 135
Page 136
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 136
Page 137
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 137
Page 138
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 138
Page 139
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 139
Page 140
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 140
Page 141
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 141
Page 142
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 142
Page 143
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 143
Page 144
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 144
Page 145
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 145
Page 146
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 146
Page 147
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 147
Page 148
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 148
Page 149
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 149
Page 150
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 150
Page 151
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 151
Page 152
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 152
Page 153
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 153
Page 154
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 154
Page 155
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 155
Page 156
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 156
Page 157
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 157
Page 158
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 158
Page 159
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 159
Page 160
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 160
Page 161
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 161
Page 162
Suggested Citation:"Chapter 4. Results of Smoothness Data Analysis." Transportation Research Board. 1997. Smoothness Specifications for Pavements: Final Report. Washington, DC: The National Academies Press. doi: 10.17226/6337.
×
Page 162

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

CHAPTER 4. RESULTS OF SMOOTHNESS DATA ANALYSIS Introduction This chapter presents the results of several analyses conducted on the effect of initial pavement smoothness and on the effect of smoothness specifications. The first part of the chapter summarizes the analyses of the effect of initial pavement smoothness on the future smoothness and life of the pavement. This is followed by a section discussing the effect of smoothness specifications on ~rutial pavement smoothness. The final part of the chapter describes a procedure for evaluating the cost effectiveness of initial smoothness levels and presents the results of an evaluation of the cost effectiveness of smoothness specifications. As discussed In chapter 3, the data for these analyses were assembled from several different sources, including existing data bases, literature surveys, and SHAs. The data from SHAs are believed to be the most representative and form the cornerstone for most of the analyses that were conducted. Analysis of the Effect of Initial Smoothness on Future Pavement Smoothness Introduction Very lithe research has been conducted on the effect of Crucial smoothness on the future smoothness of the pavement. Although He design equations developed from the AASHO Road Test (AASHO 1962) imply that initially smoother pavements maintain higher levels of rifle quality, little information is available outside of that work. However, one study reported by Janoff (1991) analyzed historical roughness data spanning 10 years and representing about 400 sections of roadway from Arizona and Pennsylvania. It was found that nutial smoothness is related to long-term roughness, as illustrated in figure 16 danoff 1991~. The smoothness values shown in figure 16 are in terms of the Mays Meter roughness Index, and it is observed that the range of values are rather small, indicating that most of the pavements used in the evaluation had not appreciably deteriorated. lanoff (1991) also determined that the annual maintenance costs were lower for those sections constructed with a higher initial smoothness, as shown in figure 17. While these findings are indications of He beneficial effects of initial pavement smoothness on the ride quality of He pavement, they are based on Innited data over a Ignited range of roughness values. Under this project, pavement projects from a wide range of data sources were used to further evaluate the effect of initial pavement smoothness on the future smoothness of the pavement. The projects used in the analyses include a myriad of pavement designs, subject to various climatic and traffic factors, and characterized by different roughness indices. A description of the data sources used in the analysis is provided In the following section. 41

40 JO 20 10 o Long-Term Roughness (in/mi) - . - 1 . ' ' _ - - - - - - - - 0 10 20 30 40 Initial Pavement Smoothness (in/mi) Figure 16. Filial pavement smoothness versus long-term pavement roughness (Mays Meter roughness index values) Janoff 1991). 1500 1000 500 lo Average Annual Maintenance Costs ($/lane mile) / - - , .. . . - I . . . I . . , 1 10 20 Initial Pavement Smoothness finimi) 30 40 Figure 17. Initial pavement smoothness versus average annual maintenance costs Janoff 1991~. 42 1

Overview of Data Sources Pavement projects from several of the data sources presented in chapter 3 were used to evaluate the effect of initial pavement smoothness on future smoothness. However, the vast majority of the projects analyzed toward this objective were those contributed by SHAs. Roughness data from the Alabama Pavement Roughness Study were also included with the SHA data for use as the centerpiece of the analysis. Roughness data from the various road tests and pavement performance studies were generally much less revealing than the State-furIiished data. This was largely attributed to the time period in which the studies were conducted (1950s and 1960s for most of the road tests), the inadequacy of the experimental designs (factors other than roughness [e.g., pavement design] were usually the main focus), and the lack of standard, reliable smoothness-measur~ng equipment and procedures used throughout the study. Consequently, outside of the SHA data, only data from the AASHO Road Test and the LTPP program were deemed worthy of detailed analysis and reporting. The evaluation results from these two sources are presented as a supplement to the State data evaluation results. Figure iS depicts the States contributing data used in the analysis of the effect of initial smoothness on future smoothness. In performing the analyses of roughness or ride quality over time, it was considered important that each pavement project consist of two or more adjacent "replicate" sections along a highway, as illustrated in figure 19. In this way, the effects of major factors, such as traffic and climate, are eliminated, and direct comparisons could be macle between sections with significantly different initial smoothness values. Table 9 provides a summary of the data sources used In the analysis, and indicates the number of replicate sections within each project. At He AASHO Road Test, for instance, a specific pavement design, subjected to a standard load (e.~. 6,000-Ib [26.7-kN] single-axIe truck load), was duplicated to form two replicate sections. Longer pavement projects, however, such as those from the States, provided numerous replicate sections, as they were broken down into short, uniform segments (typically 0.1 or ~ mi [0.16 or I.6 km]) defined by the available roughness testing interval. The preliminary analysis consisted of preparing time-series (or, in some instances load-series) roughness trends for each of the pavement sections from each data source. This was done by plotting the series of initial and extended roughness measurements for each replicate section within a project. The number of time-series data points varied by data source, as some sources provided sets of routine roughness measurements (e.g., Arizona, Kentucky), whereas others only provided two sets of roughness measurements (e.g., Iowa, RPPR). 43

Description of Data Bases Agency Survey Data Bases Work on the agency survey data bases was initiated with the establishment of the NCHRP 20-7 task 53 State survey data base. The tables In this data base were structured such that States are listec! In the first column and the survey questions are listed, in order, in the column headings to the right. The State responses were then manually entered into the appropriate cells. Once He NCHRP I-31 State survey form was completed, the data base for this survey was fully structured using the same format as the NCHRP 20-7 survey data base. As State responses became available, the results were again manually entered into the appropriate cells. As part of the agency surveys, selected paving contractors were contacted in order to obtain the contractor's perspective on pavement roughness measuring procedures and practices. Both AC and PCC paving contractors were contacted. The same procedure for the clevelopment of the agency survey data base was followed for the contractor data base. Project Analysis Data Bases Work on the project analysis ciata bases began with the development of a list of anticipated data base elements. This list, shown in table 7, was prepared following the Strategic Highway Research Program (SHRP) LTPP Data Collection Guide (FHWA 1993), and contains the necessary data elements for conducting later analyses to address the following key project objectives: i. Determine the effect of initial smoothness on the ride quality of the pavement over its life and on pavement service life. 2. Determine the effect of existing smoothness specifications on He initial as- constructed smoothness. 3. Determine He cost-effectiveness of smoothness specifications, including incentives and disincentives. Sources of Data for Project Analysis Data Base Past Road Tests AASHO Road Test The AASHO Road Test represents one of the most comprehensive studies of pavement performance that has ever been conducted. Over 800 AC and PCC pavement study sections were constructed and evaluated from 1958 to 1960. 30

Paving Project (Same design, traffic, subgrade, etc.) . . . Sect. Sect. 2 . Sect. 3 . Sect. 4 . , 1 , . . Initial Smoothness S1 S2 s3 s4 Time-Series T T T T Smoothness 1 2 3 4 Figure 19. Illustration of Replicate sections along a highway pavement project. As seen in table 9, some pavement projects were tested crucially win one particular roughness device and then tested at a later date with a different device. For those projects in which strong correlations existed between the two testing devices, conversions were made so that a single roughness index could be used to better reflect Me roughness trends. Specific examples of this occurrence include Me three {owe PCC projects (California profilograph used initially for construction acceptance, South Dakota-type profiler used later for pavement management purposes), and the five Minnesota PCC projects (GM-type profilometer used initially for construction acceptance, South Dakota-type profiler used later for pavement management). To some extent, Me ~rutial smoothness data collected represented tests performed within the first few weeks of construction. With several sources, however, such Initial data were not available, either because the construction records containing the test results had been discarded or because the ~rutial pavement smoothness was not being monitored by the SHA at that time. In those instances, pavement management smoothness measurements made within 2 to 3 years of construction were used as Initial smoothness values. Projects that fit this category include those from Illinois, Michigan, and Washington, as well as Minnesota AC and AC overlay projects. 45

Table 9. Summary of data used to analyze the effect of ~rutial pavement smoothness on future smoothness. Number of | Basic Level of I Data Source AASHO Road Test WASHO Road Test San Diego Test Road _ Minnesota _ Investigation 183 Alabama Roughness Study RPPR LTPP GPS Arizona Georgia Illinois Iowa Kentucky Michigan Minnesota South Dakota Washington Wisconsin 1 ft = 0.305 m a For AASHO Road Test, WASHO Road Test, San Diego Road Test, Minnesota Investigation 183, RPPR, and LTPP GPS, a project is equated to a specific pavement design. Full-depth AC. c Time lapse between construction and initial measurement greater than 1 year but less than 3 years. _ Roughness Measuring Device(s) Pavement Replication a Projects Section No. of Pavement Examined Length, ft Sections Types Included Initial Future 32 flexible 1 1 on to '4n 1 2 36 rigid 10 flexible 120 to 240 300 _ 4 flexible 250 5 flexible 16 flexible 26 rigid 18 rigid 2 flexible 1 composite 3 rigid 5 rigid 3 composite 12 flexible 3 rigid 6 composite 9 flexible 37 rigid 4 composite 2 rigid 11 composite 7 flexible 6 rigid 9 flexible 5 rigid 8 flexible AC, JPC, JRC AC FDACD 700 or 1200 2,640 500 to 1,000 500 6 to 8 2 to 6 3 to 18 2 to 6 2 to 8 5250 5 to 14 5,280 528 - 5 _ _ 5,280 5 to 10 ~_ 6tol3 _ 5250 3 to l0 5230 5 to 24 _ 5,280 4 to 9 AC AC, PCC JPC, JRC, CRC AC, FDAC, JPC, CRC PSI (Chloe profiler) BPR Roughometer PSI (Chloe profiler) PSI (BPR Roughometer) . BPR Roughometer Various KJ Law Pro fit o me terC PSI (Chloe profiler) BPR Roughometer PSI (Chloe profiler) PSI (BPR Roughometer) BPR Roughometer Soup Dakota type Profiler KJ Law Profilometer JPC, AC, AC/JPC, AC/AC JPC, AC, AC/JPC, AC/AC CRC, AC/CRC AC/JPC JPC JPC, AC/JPC JRC, AC AC, _ I I I PCC-doweled 13 rigid 528 3 to 11 PCC PCC, AC, AC/AC JPC, JRC, AC CRC, AC/AC, AC/PCC 6 rigid 12 flexible 25 rigid 16 composite 30 flexible _ 5,280 3 to 9 e S;ttlO (typ) 3 to 23 Mays Meter Mays Meter BPR Roughometer California Profilograph PI (Mays Meter) GM ProfilometerC GM Profilometer (PCC) South Dakota-type Profiler (AC)C California Profilograph . PCA Roadmeter PSI (Mays Meter) Mays Meter Mays Meter South Dakota type Profiler South Dakota type Profiler RI (Mays Meter) GM Profilometer South Dakota type Profiler California Profilograph PCA Roadmeter PSI (Mays Meter) 46

Some data sources, such as the AASHO Road Test, Kentucky, and Wisconsin, furnished serviceability-type data. These data took the form of serviceability or rideability indices, derived mostly from roughness measurements. The resulting time-series trends are a reciprocal of the roughness trends of other data sources, as lower index values represent rougher surfaces. Provided in appendix C are pavement project summary tables and time-series roughness plots of each of the sections evaluated in this study. The tables and plots are arranged according to data source. The summary tables include useful information about each pavement project, including project identification, location, design, construction year, evaluation period, and traffic estimates. The plots represent the roughness trends of the various fixed-interval sections that comprise a given pavement project. It should be noted Hat not all of the projects for which data were collected are included In the summary tables and roughness plots. Several projects were identified as having extended smoothness measurements, but no initial measurements (within 3 years of construction). This was the case for the majority of the Michigan projects, where only 15 of the 47 subject sections could be analyzed because of the lack of Initial measurements. Another reason for the exclusion of some projects was the inability to closely align (by milepost, reference point, or station) ~rutial smoothness measurements with extended smoothness measurements. This problem was encountered with some of the Minnesota and Washington projects. Overall Evaluation of Effect of Initial Smoothness Using SHA Data This section describes the results of the overall evaluation of He SHA data. The data sources described in the previous section are analyzed over He entire period for which smoothness data are available to determine whether He ~rutial smoothness of the pavement influences the future smoothness of He pavement. Evaluation Approach An examination of the t~me-series smoothness plots (contained In appendix C) indicate the following: 1. Pavement sections built smoother generally remain smoother over time (all over Wings being equal). 2. The performance curves of two pavements constructed to different initial smoothness levels but of otherwise similar design roughly "parallel" each other, with the rougher section perhaps deteriorating more rapidly due to greater dynamic loadings and/or greater variability in construction. An example chart illustrating the "parallel curve" concept is shown in figure 20. Sections I, 2, and 3 all roughly parallel one another, with section 3, constructed 47

In a) s o Section ~ ~ - Section 2 Time or Traffic Figure 20. Example illustration of pavement performance curves for sections along a given construction project. Initially smoother, remairung smoother over time (or traffic). Section I, constructed rougher than Me other two sections, shows perhaps a slightly greater rate of deterioration. As discussed previously, pavement projects were solicited from SHAs containing multiple sections constructed to different Crucial smoothness values. Because the pavement design, cross section, subgracle support, traffic loadings, age, and climatic forces are approx~nately We same for each of the sections within Me project, the effect of these variables on the performance of all of Me sections is assumed to be constant and Me effect of nutial pavement smoothness can therefore be isolated. For each project containing multiple sections, a statistical regression analysis was conducted relating the smoothness at any time to Me initial smoothness and to the age, as expressed in the following general form: St = aO + a~Si + a2t where: SO = Pavement smoothness at time t. aO, al, a2 = Regression coefficients (at is Me regression coefficient on Si). Si = Initial pavement smoothness. t = Verne (age), years since construction or overlay to time of smoothness testing So. 48

For the purposes of this evaluation, a multiple linear regression analysis was conducted based on observations of the trends shown in the plots. A linear mode] was selected in order to provide an Indication of the significance of the initial smoothness variable. In many cases, nonlinear regression provided a better "fit" of the data, and these improved moclels were used in the analysis on the effect of initial pavement smoothness on pavement life (described later). The output of the linear regression provides the regression coefficients and information on the statistical significance of the independent variables (in this case, initial smoothness and age) on the dependent variable (smoothness at time t). The regression coefficient al is the slope of the So versus Si regression line. If this constant is approximately 1.0, this indicates that a strong one-to-one relation exists between the initial smoothness and the future smoothness. This means that if one section is, say, 10 in/ml (0.16 m/km) smoother at initial construction Man another, then the future smoothness of that section will remain 10 Mimi (0.16 m/km) smoother than the other section over time. If the regression coefficient al is zero, this indicates that future smoothness is not dependent on initial smoothness. This concept is illustrated In figure 21. - T ._ _, Ct to In s o ~ alma /- ,/ al= 1 al=0 , . ~. _ _ _ ~ 1 , 1 , 1 , Initial Roughness Figure 21. Effect of a, regression coefficient In the relationship between Crucial smoothness and future smoothness. 49

Another way to look at the meaning of Me a, coefficient is to consider its effect In the time-series plots of pavement roughness. Figure 22 illustrates the physical significance of the al coefficient for four different cases, assuming four sections of a pavement project with different ~rutial pavement smoothness values. In the first case, when the al coefficient is zero, the Initially smooth and initially rough sections converge at some point within the data range, and no relationship exists between the initial smoothness and the future smoothness. (It is ~rnportant to note that the slopes of the lines shown in figure 22 are indicative of the relationship between time bagel and future smoothness, and not between initial anc! future smoothness.) In case 2 of figure 22, when the al coefficient is between O and I, the initial smoothness does have an effect on the future smoothness; however, the initially smooth and initially rough sections tend toward convergence, but not within the available data range. In other words, the initially smooth and ~rutially rough sections will exhibit the same level of smoothness at some point In the future, but that future point could be 5 years or it could be 25 years. Case 3 of figure 22 shows the effect when the al coefficient is equal to 1. In this case, the smoothness of Individual sections parallels one another, with initially smooth sections remaining smoother over dine. Finally, case 4 of figure 22 illustrates that when the a, coefficient is greater than I, the initially smooth and initially rough sections tend to diverge over the available data range. This would suggest that dynamic loading effects are causing the roughness of the initially rougher sections to increase at a more rapid rate. Tests for the statistical significance of the regression coefficients of We independent variables (Si and t) are also outputs of the regression analysis. The statistical significance of Me independent variables is evaluated using the p-value, which indicates the probability that the significance of the effect of the independent variable fin this case, initial pavement smoothness) on Me dependent variable (future pavement smoothness) is due to chance alone. Obviously, the smaller the p-value, the stronger the indication that the independent variable (initial smoothness) has a truly significant effect on the dependent variable (future smoothness). For this evaluation, a significance level of 10 percent was selected, meaning that if the p-value of the nutial smoothness regression coefficient balk is less than 10 percent, the results are considered significant. An example plot showing regression results is provided in figure 23 for a PCC construction project in Georgia (~-575 NB, Cherokee and Pickens Counties). This figure shows the pavement roughness at year ~ as a function of the initial roughness. The results of the regression indicate a good correlation (R2 = 0.82) for the linear model, and Me a, value of 0.89 demonstrating a strong relationship between the initial pavement roughness at construction and Me pavement roughness at year 8. For this project, each point on the graph is a I-mi (~.6-km) section along the length of the construction project. 50

~a) o Act ° 3= O G 0 ~ can Cat t~ m O=' U:) ~ 0 0 E cn ~ cad co ~ ~ t ~ ~ ~ ~ ~ ~ ~ ~ ~ . -.,\ T~ 8 = ~ ·~ ~° .4pt, . ~ A\ ., ., ., ., ., .\ A\ Iw/ul 'xapul sseu46noj _ ~ ~ ~ ~ At t; ~6 ~6 lime V.. Ace bile E an lo In E o lo C~ E F U. ' ' ' ' :~o !UlIU! 'xapul ssauq6no~ ~ c ~co ~ au O t 0 O ~ O ' ~., ~ 1 O: a) u, u' cn cn ._ cn 0 _ A ~ ~ _ s ~ o Ct ~ E tt cn ~ ~ C~ C,0 O o - cn ·` \ . ~ ·' \ ~ ; \ he ~ -~N . e~ ;` '~ ~ ·. a~ ~ ·. '5 ~ ·` ~ · \; V '"~e \~\ '., C~ U) o ._. \ \ o C~ E u, a O ·~ o .~ a~ ._ ~ 0 + 03 ' (,0 1 (0 ~ 11~=,~ ~ ~ 0 .. ~ ~ ~:,~\ 0 ~ ~ e tV ~ ,= E 0 cn ~w/u! 'xapul ssau46no~ 'eD '~0 *~\ ., ., ., ., . ,. ,` ;, ~u'IUl 'xapul ssau46no~ . 1 . 51

- ~ - - ·= 70 rat in ~ 60 u, a, ~ 50 s a a: in Georgia PCC Project GA I-575 (NB) | | Cherokee and Pickens Co. | ~ Year 8 Roughness L R2 0 82 40 30 40 - - 0089Si + 18.22 50 Initial ways Roughness, in/ml Figure 23. Regression analysis of a Georgia PCC project. Presentation of Results 60 Table 10 summarizes by State the results of the regression analysis for each individual project. This table provides information on Me location of each project, the type of pavement, the year of construction, the number of "replicate" sections widen the project, the age range over which smoothness data are available, and results of the statistical regression: as (y-intercept), a, (initial smoothness) and a2 (time) regression coefficients, We R2 of the multiple linear regression equation, the significance level (p-value) of Me al term, and whether the al term is significant (at a significance level of 0.10~. For each project, statistical regressions were conducted over Me entire range of ages for which smoothness data were available. In some cases, however, where intermecliate smoothness data are also available, regression analyses were conducted over those ranges to provide insight as to whether the significance of initial smoothness on Me future smoothness of Me pavement changes over tune. Four different age ranges were considered In this evaluation and are as follows: · O to 5 years All projects win roughness data available within the first 5 years of Me life of the pavement. O to 10 years All projects win roughness data available widen the first 10 years of the life of the pavement. 52

Table 10. Effect of in~tial smoothness on future smoothness for var~ous SHA Pro jects. Pro ect Pavement Const. No of g aO alof a2 of , -value a j | I T | Range | I T. I R | P T 1 ID I Type I Year I Sect~onsl vear 'I (~nt) I S~ I T~me I I of al I S~gn~f~cant?l I-2~1(18) EB Jefferson & St. Clair Co. I-2~1(18) WB Tefferson & St. Claire Co. I~1(19) SB Mobile Co. I-85-1(15) NB Macon Co. . I-85-1(15) SB Macon Co. P372(4) EB Tuscaloosa Co. F372(5) EB Tuscaloosa Co. F372(5) WB Tuscaloosa Co. P&U122(5) NB Houston Co. F&U122(5) SB Houston Co. F&U122(6) NB Houston Co. F&U122(6) SB Houston Co. I-20-1(14) EB St. Clair & Talladego Co. I-20-1(19) EB St. Claire Co. I-59-2(22) NB DeKalb Co. I-59-2(22) SB DeKalb Co. I-59-2(23) NB DeKalb Co. I-59-2(23) SB DeKalb Co. I-59-2(24) NB DeKalb Co. I-59-2(24) SB DeKalb Co. I-65-3(1) NB Tefferson Co. I~3(1) SB Tefferson Co. I~3(14) NB Jefferson & Blount Co. I-65-3(14) SB Tefferson & Blount Co. , I~3(15) NB Blount Co. I~3(15) SB Blount Co. _ ~ I~3(16) NB Blount Co. I-65-3(16) SB Blount Co. I-65-3(17) NB Tefferson Co. I~3(17) SB Tefferson Co. ~ I-65-3(20) N8 Cullman Co. _ Alabama-BPR Roughometer, in/nu 19.79 17.02 19.59 32.18 40.06 0.05 1.98 3.87 7.54 3.91 27.01 28.38 70.32 19.13 21.72 20.48 21.39 25.10 _11.55 23.88 10.37 25.29 _19.59 60.88 31.62 25.16 12.73 4.05 7.59 -1.51 0.94 1.82 24.48 33.31 6.66 53 o~ 0-6 0-5 ~5 0-5 0-5 ~7 0-7 0-7 ~7 ~7 0-7 0-7 0-7 0-3.5 0-3.5 0-3.5 0-3.5 0-6.6 0-6.6 0-6 0-6 0~ 0-6 0-7 0-7 0-7 0-7 0-7 0-7 0-7 0-7 0-7 0-7 0-7 0.76 0.81 0.82 0.68 0.54 1.01 0.97 0.95 0.91 0.95 0.62 0.60 0.11 0.74 0.81 0.81 0.80 0.76 0.84 0.71 0.91 0.77 0.82 0.38 0.65 0.73 0.87 0.96 0.92 1.00 0.99 0.98 0.73 0.63 0.94 2.61 2.48 0.16 -1.60 3.77 3.57 3.69 4.42 3.75 3.21 3.05 3.12 1.29 3.30 2.12 2.02 3.34 3.32 2.24 2.38 2.96 1.68 2.24 . . 2.72 8.72 7.11 . 4.82 4.77 5.12 . 5.00 5.43 . 4.31 8.90 . 7.42 4.44 0.87 0.90 0.61 0.38 0.91 0.95 0.97 0.92 0.90 0.94 0.78 0.79 0.38 0.86 0.94 0.91 0.92 0.89 0.80 0.82 0.84 0.82 0.75 0.69 0.92 0.92 0.96 0.94 0.95 0.97 0.94 0.95 0.92 0.96 0.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00 l 0.00 0.00 0.00 0.00 0.00 0.55 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Table 10. Effect of in~tial smoothness on future smoothness for various SHA Projects (continued). Project ~ Pavement ~ Const. ~ No. of ID I Type I Year I Sections I-65-3(20) SB Cullman Co. I-65-3(21) NB Cullman Co. I-65-3t21) SB Cullman Co. I~3(22) NB Cullman Co. I-65-3(22) SB Cullman Co. I-17 SB Coconino Co. (2) I-17 SB Coconino Co. (2) I-17 SB Coconino Co. (2) I-17 SB Coconino Co. (2) I-17 SB Coconino Co. (1) I-17 SB Coconino Co. (1) {PC I-17 SB Coconino Co. (1) I-17 SB Coconino Co. (1) I-40 EB Cocoruno Co. (2) I-40 EB Cocon~no Co. (2) I-40 EB Cocon~no Co. (2) I-17 SB Coconino Co. (3) I-17 SB Coconino Co. (3) I-17 SB Coconino Co. (3) I-17 SB Coconino Co. (3) I-10 WB Mancopa Co. I-10 EB LaPaz Co. I-10 EB LaPaz Co. I-10 EB LaPaz Co. I-19 SB Pima Co. AC AC I-19 SB Pima Co. _ I-19SBPima Co. I-40 WB Moiave Co. AC I-40 WB Mojave Co. AC AC I-10 EB Cochise Co. AC I-10 EB Cochise Co. I-15 NB Moiave Co. PCC AC 1961 13 961_ 13 1961 13 1962 6 _ 1962 6 Anzona-Mays Meter, in/nu TPC 1974 1 5 1 0-5 - 1 5 1 0-10 _ TPC 1974 5 0-15 1974 5 0-19 TP JPC 1976 11 0-4 ~1976 _ 11 0-8 TPC 1976 11 0-12 IPC 1976 11 0-17 1972 12 0-4 TP IPC 1972 12 0-8 IPC 1972 12 0-14 TPC ~ - 1974~ _ 4 0-5 TPC 1974 4 0-10 {PC 1974 4 0-15 TPC 1974 4 0-19 . TPC 1985 4 0-5 _ 1985 4 0-9 _ 1973 14 ~4 AC 1973 14 0-8 _ 1973 14 ~13 _ 1976 5 ~4 _ 1976 5 0-8 . _ 1976 5 0-12 _ 1976 5 0-17 _ 1978 8 ~5 _ 1978 8 0-10 . 1978 0-15 976 12 0-10 1976 12 0-15 . _ 1976 12 0-17 _ 1974 8 0-5 54 g aO alof a20f yeag '~ (mt) ~S1 Time R ~p-value~ Significant? 0-7 10.91 0.90 4.59 0.95 o.oo 1 0-7 12.85 0.84 4.99 0.93 0.00 _ 0-7 ~_ 14.48 0.83 4.02 0.93 0.00 1 Yes 0-6 -2.77 1.06 4.09 0.87 0.00 0-6 -5.35 1.11 3.40 0.62 0.00 18.79 1 24.10 15.89 l ~.38 21.11 41.05 49.84 50.79 -33.47 -66.49 -113.95 21.79 34.02 25.29 4.39 44.44 40.87 6.42 12.61 -0.35 5.46 13.46 16.38 21.61 14.79 14.78 3.66 4.03 4.96 5.38 0.18 0.39 0.33 0.31 0.37 0.78 0.57 0.44 0.32 2.10 2.93 4.29 0.20 0.01 0.09 0.44 0.48 0.49 0.37 0.45 0.87 0.71 1 0.34 0.32 0.27 0.65 0.66 0.72 0.68 0.71 l 1 0.73 L1.12 8.79 6.93 9.16 11.75 1 7.54 1 4.07 4.17 l 1 5.72 1 8.48 12.03 1 13.68 8.60 4.98 1 6.54 8.51 l 1.05 2.05 7.79 3.14 l 4.62_ 1.68 4.20 3.57 1 2.70 1.81 1.82 1.58 1 1 - - 1 2.63 219 l 1 2~01 L201 0.82 0.84 0.83 0.79 0.73 0.60 0.73 0.82 0.80 0.83 0.87 0.83 0.78 0.89 0.91 0.23 - 0.57 0.65 0.37 0.68 0.62 0.66 0.76 0.77 0.24 0.40 0.44 0.83 0.82 0.79 0.46 No N~ No

Table 10. Effect of in~tial smoothness on future smoothness for various SHA Projects (continued). Proj ect ID Pavement Type 1 1 I-15 NB Mo eve Co. AC _ 1 _ I-15 NB Mojave Co. AC _ _ I-15 NB Mojave Co. AC _ I-40 EB Navaio Co. . I-40 EB Navajo Co. AC AC AC I~O EB Concon~no Co. (1) I-40 EB Concon~no Co. (1) AC/PCC _ AC/PCC _ _ I-40 WB Cocon~no Co. (1) AC/PCC I~O WB Cocon~no Co. (1) I~O EB Cocon~no Co. (2) I~O EB Coconir~o Co. (2) AC/PCC AC/PCC AC/PCC I-17 SB Yapavi Co. _ ACIAC I-17 SB Yanavi Co. . _ I~O EB Navajo Co. (2) I~O EB Navajo Co. (2) I~40 EB Navaio Co. (3) I~O EB Navaio Co. (3) AC/AC AC/AC AC/AC AC/AC AC/AC I~O WB Navaio Co. (3) AC/AC S-260 EB Yapavi Co. AC/AC I-17 NB Yapavi Co. I-17 NB Yapavi Co. GA 365 NB Hall Co. GA 316 EB G~nnnett Co. I-575 NB Cherokee & Pickens Co. 1 GA 400 NB Dawson & Lumpkin Co. GA 400 NB Dawson & Lumpkin Co. GA 365 NB Habersham Co. GA 365 NB Habersham Co. GA 365 SB Stephens Co. AC/AC AC/AC Const. Year 1974 1974 1974 1979 1979 1979 1987 1987 1987 1987 1987 1987 1980 1980 1980 1980 1982 1982 1981 1981 1980 1980 1980 1980 1980 Age Range, year 0-10 0-15 0-19 0-4 0-8 0-12 ~4 0~ 0-4 0~ _ O 0~ 0-3 0-10 0-3 0-8 0-3 0-8 0 0-8 0-5 0-10 ~14 0-5 0-10 aO (ins) 1.08 1.88 5.97 25.32 33.62 31.24 23.62 9.45 24.58 15.51 -27.24 ~4.37 -11.44 6.78 .04 11.27 41.63 49.77 33.89 58.77 10.09 20.33 32.27 9.40 9.69 alof si 1.09 0.97 0.93 0.39 0.23 0.27 0.47 0.68 0.53 0.72 .37 .42 1.19 0.86 0.77 0.89 o.Og 0.13 0.39 - 0.05 0.83 0.59 0.31 0.82 0.66 a2 of Time 1.83 2.42 1.88 4.39 3.78 4.12 3.62 7.81 0.80 0.46 5.18 l 7.82 3.73 2.88 l 7.74 s.37 . 2.20 4.59 8.70 l 4.18 1.41 l .40 .43 4.78 3.00 R2 0.64 0.77 0.71 . 0.41 0.48 . 0.71 0.29 . 0.67 0.13 0.08 0.65 - 0.71 0.77 0.75 0.46 0.56 0.88 0.65 0.62 0.47 0.38 0.29 0.30 0.62 0.60 | p-value Lo.oo l 1 o.oo o.Oo, 1 0.17 1 0.39 1 1 1 0.2 L 014 1 0.04 1 1 1 0.29 1 0.21 1 o.oo 1 1 o.oo 1 1 ooo 1 1 o.oo 1 1 1 1 o~oo 1 1 ooo 1 0.76 0.67 0.37 1 I 1 0.88 1 I o.o~o 1 1 1 o.oo 1 o~o4 0.02 0.0 Georgia-Mays Meter, in/ml TPC 11982 1 5 1 0-5 1 17.53 1 0.68 1 3.14 1 0.57 1 0.16 1 rPC I1982 1 5 1 0-9 12l.3410.6512.l510.63T 0.: 1982 1 5 1 0-12 1 3906 1 0.49 1 073 1 007 1 054 {PC 11981 1 5 1 0-10 1 25.30 1 0.61 1 -0.51 1 0.60 1 o.oo ~PC I1981 T s 1 0-13 133.6110.52T1.1710.651 0.00 JPC I1986 1 s 1 0-8 14.2210.89ll.75To.821 °.°° 11 1 1 1 1 1 1 AC 1 1981 1 5 1 0-9 1 5.69 1 0.86 1 0.61 1 0.63 1 0.00 AC T 1981 1 5 1 0-13 1 13.93 1 0.65 1 -~0.34 1 0.29 1 o.ol~ I I t I I I I I l AC 1 1988 1 5 1 °~ 1 -1.76 1 1.04 1 0 45 1 0.51 1 0.01 AC 1 1988 1 5 1 06T2.2210.941~.20To.411 0.00 1 1 1 1 1 1 1 1 1 AC 1 1991 1 5 1 0-3 1 15.28 1 0.38 1 1.31 1 0.28 1 0.17 55 No Yes Yes Yes Yes Yes

Table 10. Effect of initial smoothness on future smoothness for various SHA Projects (continued). Proj ect ID L I I I-75 SB Valdosta Co. AC/PCC _ I-75 SB Valdosta Co. I-75 SO Houston & Peach Co. (1) I-75 SB Houston & Peach Co. (1) IN SB Banks & Franklin Co. I-85 SB Banks & Franklin Co. I-75 NB Houston Co. I-75 NB Houston Co. I-75 NB Houston & Peach Co. (2) . I-75 NB Houston & Peach Co. (2) I-85 NB Hart & Franklin Co. I-85 NB Hart & Franklin Co. _ I-95 SB Chatham & Effingham Co. I-95 SB Chatham & Effingham Co. I-185 SB Trou Co. P I-185 SB Troup Co. I-20 EB Newton Co. I-75 NB Whiffield Co. I-20 EB Cobb & Douglas Co. I-20 EB Cobb & Douglas Co. I-75 NB Cobb & Cherokee Co. Pavement Type AC/PCC AC/PCC AC/PCC AC/PCC AC/PCC AC/PCC AC/PCC __ AC/PCC AC/PCC AC/PCC AC/PCC AC/AC AC/AC AC/AC AC/AC I-24 WB Massac Co. I-57 SB Marion Co. I-72 EB Macon Co. (2) _I-74 WB McLean Co. I-55 SB Montgomery Co. I-55 NB McLean Co. (2) I-55 SB McLean Co. (2) I-55 NB McLean Co. (1) I-55 SO McLean Co. (1) Const. Year 1981 1981 1981 1981 1987 1987 1981 1981 1981 1981 1980 1980 1985 1985 1987 1987 1986 1983 1980 1980 1987 Age Range, year 0-9 ~13 0-9 ~13 ~5 0-7 ~5 0-11 0-5 0-11 ~5 _ 0-11 ~5 ~9 ~5 ~7 0-8 0~ 04 ~7 ~5 an (ins) 21.97 24.14 15.36 5.40 -1.28 -10.17 _ -1.24 32.47 9.05 11.80 13.04 20.27 7.72 13.37 17.28 25.44 18.76 2.77 13.68 16.52 0.33 Illinois-BPR Roughometer, in/ml 1972 lg72 1971 1971 1969 1975 1977 1971 1971 1973 197B 1978 1977 1977 1976 56 7.23 8.73 29.47 25.24 19.27 37.76 51.65 -21.47 -5.52 50.34 38.70 58.49 33.00 5.55 4.29 alof sit 0.33 0.30 0.43 0.78 .02 .30 1.29 -0.0 0.66 0.53 0.29 0.16 0.76 0.56 0.22 -0.08 0.41 1.06 0.27 0.21 0.96 0.83 0.77 0.62 0.62 0.61 0.34 0.20 1.22 _1 O_ o.Os 0.46 0.28 0.11 1.00 0.51 a2 of Time -0.10 -0.61 -0.19 -0.07 .5 1.31 -0.69 -0.54 -0.55 o.O9 3.24 1.36 -0.26 l -0.57 0.92 l 0.21 -0.52 l 1.42 2.08 1.09 0.68 2.51 2.85 2.30 2.47 2.54 4.48 s.32 2.02 2.29 l 3.82 3.53 2.87 4.80 5.12 4.49 R2 . 0.02 0.13 . 0.14 0.23 0.58 0.56 0.09 0.14 0.28 0.21 0.74 0.51 0.45 0.24 0.41 0.03 0.11 0.34 0.27 0.20 0.56 0.93 0.97 __78 . 0.91 0.70 0.73 0.60 0.94 0.93 . 0.77 0.92 0.73 0.69 0.78 o.ss p-value 0.54 0.49 0.18 0.02 0.07 0.c2 0.29 o.ss 0.00 0.00 0.39 0.62 0.01 0.06 0.40 0.78 l 0.39 0.01 l 0.09 0.08 l 0.00 . o.oo 0.00 o.OO . o.oo 0.22 0.41 0.34 0.00 0.00 . 0.84 0.09 0.09 0.87 0.02 0.00 Significant No Yes Yes Yes Yes Yes No No No Yes No

Table 10. Effect of initial smoothness on future smoothness for various SHA Projects (continued). Proj ect ID Pavement Type I-55 SB Livingston Co. (2) I-39 NB l~aSalle Co. (1) I-39 SB LaSalle Co. (1) I-39 SB LaSalle Co. (2) I-57 NB Fayette Co. I-70 WB Effingham Co. I-24 EB Gldwell & Trigg Co._ L I-24 EB Caldwell & Trigg Co. I-24 EB Gldwell & Trigs Co. I-24 WB Gldwell & Trigg Co. I-24 WB Caldwell & Trigg Co. I-24 WB Caldwell & Tries Co. . I-24 EB Lyon Co. . I-24 WB Lyon Co. I-64 WB Shelbv Co. . ~ IN We Shelby Co. I-71 NB Galladn Co. _ I-71 NB Galladn Co. I-71 NB Galladn Co. I-71 SB Galladn Co. . I-71 SB Galladn Co. I-75 NB Grant Co. I-75 NB Grant Co. I-75 SB Grant Co. 1-75 SB Grant Co. Mountain Parkway EB Clark Co. Mountain ParhNay EB Clark Co. Mountain Parkway WB Clark Co. Mountain Parkwa WB Clark Co. Y Western KY Parkway EB Ohio Co. Western KY Parkway EB Ohio Co. Western KY Parkwav WB Ohio Co. . Age Range, year 0-16 0 ~9 ~9 ~9 0-5 0-5 ~8 0-8 an (ins) 1.67 3.17 44.80 37.68 17.33 8.94 12.55 0.49 3.94 alof . S 0.91 0.97 0.54 0.63 0.41 0.76 0.68 0.81 0.64 a2 of Time 5.33 7.90 4.34 3.65 5.29 7.14 7.15 9.83 8.44 Kentucky-Mays Meter, converted to rideabilibr index 57 0-4 ~8 ~12 0-4 0-8 ~2 ~4 0~ 0-4 0-10 _ 0-4 ~8 . 0-11 0- 0-8 0-11 0-4 l 0-8 0-4 l l 0-8 . 0 0-8 l . 0-4 08 . 0 0 . . 0-4 0.38 . 0.36 0.53 . 1.31 1.36 . 1.35 . 1.36 -0.94 . 1.66 0.66 3.62 . 4.24 4.40 3.49 4.02 4.23 -0.62 -0.70 _ 0.32 . -0.15 1.78 2.61 3.22 3.25 -1.45 -8.02 . 1.90 0.85 0.85 0.81 0.63 0.61 0.61 0.64 1.26 0.56 0.80 0.09 -0.06 -0.09 0.12 0.00 -0.04 1.15 1.17 _0.91 1.04 0.53 0.32 0.16 0.13 1.31 2.86 0.52 -0.03 -0.03 -0.03 -0.07 -0.04 -0.04 -0.07 -0.09 -0.05 -0.01 0.01 -0.02 -0.04 0.01 -0.02 -0.04 0.01 -0.02 _-0.01 -0.03 -0.05 -0.03 -0.10 -0.05 -0.12 -0.11 -0.09 R2 0.97 0.95 0.89 0.81 0.93 0.98 0.98 0.98 0.99 0.61 0.66 1 _ 0.67 0.42 l 0.42 0.51 l 0.58 0.92 0.49 0.24 1 0.04 1 0.11 1 .45 0.07 1 0.08 .35 0.56 0.62 0.13 .49 0.36 0.40 1 0.62 0.56 1 1 0.72 Lo.76 l 0.67 | p-value 1 o.oo 1 o.oo 1 o.oo 1 o.oo L ooo 1 o.oo 1 L ooo 1 o.oo l 1 o.oo 1 o.oo 1 o.oo 1 ooo _o.oo l 1 o.oo 1 o.oo 1 o.oo 1 o.oo 1 o.oo 1 1 o.oo 1 0.21 1 0.25 1 .o4 0.08 1 1.00 1 0.23 1 o.oo 1 o.oo 1 o. l 0.01 0.10 0.19 _ _0.64 1 0.63 1 0.11 :00_ l 1 0.07 Yes Yes Yes 1 - 11 | Yes | Yes 1 11 Yes No Yes Yes No I No No | No 1 11 I No ,, | Yes 11 t lI

Table 10. Effect of irutial smoothness on future smoothness for various SHA Projects (continued). Project Tn _ l Western KY Parkwav WB Ohio Co. Bluegrass Parkway EB Anderson Co. Bluegrass Parkway EB Anderson Co. Bluegrass Parkway WB Anderson Co. Bluegrass Parkway WB Anderson Co. IN EB Grter Co. Pavement Type AC/PCC AC/PCC AC/PCC AC/PCC AC/PCC ~- AC/AC I-64 EB Carter Co. AC,fAC I-64 EB Carter Co. IN WB Carter Co. AC/AC l AC/AC I-64 WB Grter Co. AC/AC I-64 WB Grter Co. AC/AC I-65 NB Hart Co. AC/AC I-65 NB Hart Co. AC/AC I-65 SB Hart Co. AC/AC IN SB Hart Co. AC/AC I-65 SB Hart Co. Bluegrass Parkway WB Woodford Co. Bluegrass Parkway WB Woodford Co. AC/AC . _ _ AC/AC I-94 WB Calhoun Co. I-96 EB Ionia Co. TRC l ARC I-96 WB Ionia Co. I-94 EB Van Buren Co. I-94 WB Van Buren Co. US 27 SB Hare Co. US 131 NB Osceola Co. US 131 SB Osceola Co. I-75 NB Roscommon Co. (1) I-75 NB Roscommon Co. (2) I-75 SB Roscommon Co. (1&2) I-96 EB Ottawa Co. I-96 WB Ottawa Co. AC/PCC AC/PCC . Const. No. of Year Sections 1986 7 1986 ~5 1986 5 1986 5 1986 5 1984 4 1984 4 1984 _ 4 1983 4 1983 4 1983 4 984 7 .1984 7 1984 7 1984 7 . 1982 4 1982 4 . Age Range, year 0-8 0-4 0-8 0~ 0-8 0~ 0-8 0-12 0-4 __0-8 . ~11 0-4 0-10 . ~4 0-10 0~ _ 0-10 an (ins) 1.09 3.39 . 4.24 _ -1.30 . -3.23 . -0.22 . -1.70 . -2.90 -2.28 . .26 . 0.62 . 2.06 . 2.08 , .56 2.01 -1.98 . _ , -3.26 Michigan-GM Profilometer, in/ml alof sit 0.72 0.15 .07 1.31 1.79 1.05 1.42 1.72 1.59 1.06 0.83 0.50 0.50 1.14 0.51 1.47 1.80 0.85 . 1.16 1.02 . 0.93 1.34 0.97 1.09 1.11 1.52 . 0.62 0.27 0.97 . 1.55 0.83 . 1.17 I ~.O; 1 l 1 -0.13 1 1 -o.og 1 T cog I . . T -007-1 1~0.05 1 l 1 -0.04 1 L: l 1 ~o5 I ~.04 I ~.03 1 1 1 1 ~01 T ~02 T ~.02 . . -003 _-o.o l 1 -0.07 1 I= :4 . 0.40 1 3.00 1 1 1.78 101 l 0.65 1 0.55 0.34 0.81 2.23 1.99 0.15 1 1 _. 9~ R2 0.71 0.85 0.81 0.74 0.86 0.65 0.71 0.80 0.56 0.62 0.66 0.42 0.49 0.64 0.63 0.63 0.84 0.91 0.98 1.00 0.94 0.89 0.96 0.98 0.98 0.70, l 0.51 0.23 0.89 0.82 0.57 1 1 0.86 1 Minnesota-KJ Law Profilometer and PaveTech/South Dakota profiler, in/nu I T MN 27 EB Douglas Co. I AC I . . - . MN 32 NB Red Lake Co. | AC l I-35 SB Rice Co. | PCC (doweled) l 1989 1 6 990 1 4 1990 1 6 58 l 0-6 1 6.05 1 0 85 1 8 91 1 0 49 0-5 1~62.801 460 1 5.96 1 O.55 ~5 1 71.14 1 -0~11 T 26591 0~93 . 0.05 1 Yes l 0.13 1 No 0.87 | No

Table 10. Effect of irutial smoothness on future smoothness for various SHA Projects (cor~tinued). Project ID MN 60 EB Cotton Wood Co. Pavement Type PCC (doweled) MN 210 EB Crow Wing Co. ~ 23 EB Ka_dwohi Co. PCC (doweled) _ AC/AC Const. Year 1990 . 1991 1989 No. of Sections 8 7 _ 4 Age Range, year = ~ 0-5 0~ 0-5 an (ins) _ ~ 5.31 74.42 8.8 alof sir 0.93 0.35 -0.17 a2 of Time - _ ~ 18.02 . 25.42 1.53 R2 0.96 0.76 0.12 Oval= e _ 0.00 . 0.50 086 . South Dakota-California Profilograph, in/ml US 14 EB Beadle Co. (1) US 14 WB Beadle Co. (1) US 14 EB Beadle Co. (2) _ US 14 WB Beadle Co. (2) . . US 8S NB Butte & Lawrence Co. US 16 EB Pigeon Co. _ US 16 WB Pennin~ton Co. _ US 18 EB Fall River Co. (1) US 18 EB Fall River Co. (2) US 18 WB Fall River Co. (2) SD 115 NB Minnehoha Co. US 81 NB Brookin s &e Kin sbu Co. ~ ~ By. US 81 SB Brookin s & Kin sbu Co. g g ry I-90 EB Grant Co. I-90 WB Grant Co. I-90 EB Adams Co. (2) I-90 WB Adams Co. (2) l - 1 I-90 EB King Co. (2) l I -90 WB King Co. (2) T 1 I-5 NB Snohomish Co. | l I-5 SB Snohomish Co. I-90 EB King Co. (1) I-90 WB King Co. (1) I-90 EB Adams Co. (1) I-90 WB Adams Co. (1) IN EB Yakima Co. | I-82 WB Yakima Co. I US 2 EB Kin Co. l US 2 WB King Co. | US 97 SB Yakima Co. | 59 . 11 Significant? _ ~ ,. _ Yes _ No No _ ., ~8 0-8 0-8 1 on 1 7 1 ~8 l 1 ~8 1 1 ~9 1 0-6 1 ~6 0-5 1 1 on 1 o-3 I -0.40 l 1 8.48 1 1 3.07 0.12 1.22 0.36 1 0.07 1.56 1 3.19 6.84 1 1 1.82 L 7.61 l 1 1.15 l 1.17 57 069 l 1.19 0.78 1.03 1 .08 0.85 1- L~ 0.09 1.20 1.14 1 1.08 l 0.83 1 1.02 L_1.49 l 0.60 0.35 0.31 1 0.41 0.69 1 1.06 0.17 1 .73 -1.03 0.89 0.84 0.43 0.67 0.77 0.70 0.90 0.95 0.74 - 0.80 0.01 0.42 Lit Yes | Yes | No | Washington-Cox (PCA) Road Meter, in/ml l 974 1 5 ]~ 1 -17.05 1 0.99 1 739 1 079 1 000 1974 1 5 1 0-11 1 33.49 T 1.16 T 7.65 1 0.76 1 0.00 ~ [- -- 1 - - - -- 1 -- 1 1 - 1 -t t == =~ 1 1 1 1 1 1 1 1 1 1~T I - 1 - 1 - 1 -- 1 1 -1 1 ~ 1 PCC~ = ~ PCC ~ 1979 ~ 4 ~ 0-6 ~ -68.69 ~ 1.61 ~ 8.38 ~ 0.80 ~ 0.00 I AC/AC T ITS AC/AC 1 1973 1 4 1 0-12 1 58.22 1 0.65 1 0.90 1 0.42 1 0.00 AC/AC l 1974 1 4 1 0-14 1 10.39 1 0.90 1 4.94 1 0.63 1 0.00 I -- - ~--- 1 1 1 1 -- 1 -I 1 AC/AC 1 1974 1 4 1 0-14 1 12.10 1 0.83 1 4.59 1 0.83 1 0.00 AC/AC 1 1975 1 5 1 0-10 1 23.66 1 0.81 1 6.06 1 0.62 1 0.00 AC/AC 1 1975 1 5 1 0-10 1 67.S0 1 0.65 1 7.S7 1 O.SS 1 0.05 AC/AC 1 1979 1 6 1 0-7 1 -21.64 1 122 1 6~10 1 0.90 1 0 00 Yes _ | Yes | Yes | Yes | Yes | Yes ll Yes _ || Yes

Table 10. Effect of in~tial smoothness on future smoothness for various SHA Projects (continued). Project T Pavement ~ Const. T No. of I ' ~ aO ra1ofl a20f 1 2 Ra] I e R -value Si Type ~ Year ~ Sections ~g '~ (ins) ~Si ~ Time ~ P WI 21 EB & WB Adams Co. WI 21 EB & WB Adams Co. WI 21 EB & WB Adams Co. WI 21 EB & WB Waushara Co. (1) WI 21 EB & WB Waushara Co. (1) . WI 21 EB & WB Waushara Co. (1) SB Fond Du Lac Co. (1) . WI 26 NB & SB Fond Du Lac Co. (1) WI 26 NB & SB Fond Du Lac Co. (1) WI 80 NB & SB Tuneau Co. , WI 80 NB & SB Juneau Co. US8EB&WBRusk Co. _US8EB&WBRusk Co. US8EB&WB Rusk Co. _ . US 12 EB & WB St. Croix Co. US12EB&WBSt. Croix Co. US12 EB&WB ackson&Monroe Co. T US 12 EB & WB Jackson & Monroe Co. US12 EB&WB Dane Co.(1) US 12 EB & WB Dane Co. (1) US 12 EB & WB Dane Co. (1) WI 13 NB & SB Ashland Co. (1) WI 13 NB & SB Ashland Co. (1) . US18 EB&WBGrantCo. US41 SB Washington Co. US 41 SB Washington Co. . WI80 NB&SBGrar~t Co. US 151 NB & SB Grant & Lafayette Co. US 151 NB & SB Grant & Lafayette Co. . US 10 EB & WB Outagamie Co. (1) US 10 EB & WB Outagamie Co. (1) WI 13 NB & SB Marathon Co. (2) US 41 NB Dod~e Co. AC AC/PCC AC/PCC AC/PCC AC/PCC US 41 NB Dod,re Co. TRC US 41 NB Dod~e Co. . US 10 EB & WB Outagamie Co. (2) Wisconsin-Mays Meter converted to PSI AC 1978 1 9 10-6 1 -1.84 AC 1978 0-12 -3.06 AC 1978 9 ~16 -2.70 AC 1978 0-6 7.82 AC 1978 4 0-12 10.49 _ AC 1978 0-16 12.12 _ _ . AC 1978 8 0-6 1.46 _ AC 1978 0-12 2.19 _ . _ AC 1978 8 0-16 2.37 _ AC 198: 6 0-1.0 2.03 _ AC 1980 6 0-14 1.83 . AC/PCC 1980 ~7 0-5 _ . AC/PCC 1980 7 ~9 -2.00 AC PCC 1980 0-13 -2.45_ / . AC/PCC 1979 8 0-10 1.23. AC/PCC 1979 8 0-14 1.85 . AC PCC 1979 0-10 -.0.93 / . AC/PCC 1979 0-14 0.16 AC/PCC~ ~1981 14 ~ 0-4 ~-0.27 . 1981 14 0-8 -0.18 AC/PCC~1_81 ~14 ~0-12 -0.55 . AC/PCC1979 5 0-10 2.81 AC/PCC ~ ~1979 _ ~ ~__~- _ 0-14 252 _ . _ AC/PCC 1979 10 0-10 1.05 _ AC PCC 1979 15 0-5 1.17 _ / AC/PCC 1979 15 0-9 2.29 AC/PCC 1981 0-10 4.11 . AC/PCC 1980 S ~5 1.63 AC/PCC 1980 0-9 1.77 _ lRC 1981 3 0-5 -1.95 _ TRC 1981 3 ~ 0-11 ~-2.67 _ _ TRC 1989 8 0-5 0.70 _ TRC 1979 5 0-5 -1.36 _ . 1979 0-10 -3.12 C 1979 0-15 ~.02 TR TRC 1980 5 0-10 2.09 _ 60 1.41 1.69 1.61 -0.70 -1.25 -1.60 0.64 0.46 0.44 0.47 0.52 1.49 1.41 1.47 0.73 0.61 1.22 1.04 1.06 1.03 1.11 0.38 0.46 0.75 0.74 0.51 -0.04 0.51 0.45 1.39 1.50 0.81 1.35 1.76 1.75 0.42 -0.04 _0.08 ~.08 -0.09 -0.14 -0.15 -0.06 -0.07 -0.08 -0.08 -0.09 -0.20 -0.19 -0.14 -0.07 -0.12 -0.17 -0.22 -0.20 -0.17 _-0.16 -0.11 -0.12 -0.12 -0.14 -0.18 -0.13 -0.21 -0.17 -0.23 -0.14 -0.06 0.13 -0.13 -0.14 -0.20 1 0.71 Lo.6l 0.74 3 0.91 1 0.93 0.63 1 0.65 0.78 1 1 0.84 1 o.so 1 0.85 0.92 0.84 1 0.67 0.78 0.77 1 0.86 .89 l 0.86 0.87 0.86 0.89 1 .95 0.74 1 0.82 0.81 1 0.92 0.93 1 1 0.93 1 o.so 1 0.66 _0.60 l 1 0.65 t0.97 1 ooo WO.Oo l 1 o.oo L_ 0.15 .17 1 0.06 1 o.oo 1 o.oo 1 o~oo 1 ooo 1 o.oo 1 1 ooo 1 oooo 1 o.oo 1 1 o.oo 0.01 0.04 1 L oo9 1 o.oo l 1__ ooo 1 o.oo l 0.08 0.04 1 1 ooo 1 0.01 1 1 oso9 0.92 1 L007 1 0.10 1 1 o.oo 1 o.oo 1 1 o.oo L 0.02 l 1 o.oo 1 0.02 1 1 0.02 Yes No No Yes Yes Yes Yes Yes Yes No Yes Yes

Table 10. Effect of in~tial smoothness on future smoothness for various SHA Projects (continued). Proj ect ID WI 13 NS & SB MarathOn CO. (1) _ WI 21 EB & WB WaUShara CO. (2) _ _ US 45 NB WaShin~tOn CO. US 45 NB WaShin~tOn CO. _ _ US 10 EB & WB PierCe CO. _ US 10 EB & WB PierCe CO. _ US 51 NB Mar Uette CO. _ _ _ q _ US 61 NB & SB Grant CO. _ US 61 NB & SB Grant CO. I-94 EB St. GOiX & DUnn CO. -_ US 18 WB IOWa & Dane CO Pavement Type JPC w/o Dwls JPC W/O DWlS TPC W/O DW1S TPC W/O DWlS IPC W/O DWlS ~C w/o Dwls JPC W/O DWlS JPC W/O DW1S JPC W/O DWlS TPC W/O DWlS TPC W/O DWlS . US 18 WB Dane CO. TPC W/O DWlS US 18 WB Dane CO. TPC W/O DW1S US 151 NB & SB DOdRe CO. (1) US 151 NB & SB DOdRe CO. (1) TPC W/O DWlS TPC W/O DW1S US 151 SB Iowa Co. TPC w/o DW1S . US 151 SB IOWa CO. JPC W/O DW1S I-43 NB ShebOYRan & ManitOWOC CO. I~3 SB ShebO Ran & ManitOWOC CO. Y I 43 NB BrOWn CO. I~3 SB BrOWn CO. I-43 SB BrOWn CO. US 53 NB Douglas Co. _ US 53 NB DOURlaS CO. CRC _ CRC _ _ _ _ CRC _ _ CRC _ _ CRC _ _ CRC _ _ CRC _ CRC _ _ US 53 SB DOURlaS CO. CRC __- US 12 WB Dane CO. (2) JPC W/ DWlS US 18 EB IOWa CO. JPC W/ DWlS . WI 26 NB & SB Rock & JeRerSOn CO. lPC W/ DwlS . I-94 EB St. GOiX CO. (1) AC/AC _ _ I-94 WB SL Croix Co. (1) AC/AC WI 26 NB Fond Du Lac Co. (2) AC/AC _ WI 26 NB FOrid DU LaC CO. (2) AC/AC _ US 8 EB & WB FOreSt CO. I US 8 EB & WB FOreSt CO. _ AC/AC _ AC/AC Const. Year 1984 1988 1985 1985 1983 1983 1984 - 1982 1982 1984 1983 1984 1984 1982 1982 1980 1980 1_80 1980 1980 - 1981 1981 1981 1981 1981 1981 1981 1981 1988 1988 1989 1990 1989 1989 1979 1979 1985 1985 . _ Range, aD a 1 o f Tim e ~; ~ Y . . ~10_ 2.51 0.42 ~.18 0.820.34 -_ 2.01 0.20 0.820.05 0-5 1.61 0.67 ~.07 0.41 0.06 0-9 4.28 0.10 -0.16 0.70 ~0.74 0-4 -0.41 1.12 -0.05 0.84 0.00 0 ~-0.28 1.13 ~.17 0.81 0.00 0-10 1.77 0.59 -0.11 0.66 0.30 0-5 -2.03 1.43 -0.14 0.68 0.07 0-11 -2.40 1.49 -0.13 0.84 0.01 0-6 -2.89 1.61 -0.22 0.62 _ 0.01 0 ~3 38 0.12 -0.15 0.76 0.70 0-5 0.83 _ 0.86 -0.19 0.64 0.13 0-9 0.30 O.g6 -0.16 0.76 _0.05 ~5 -1.09 1.26 -0.05 0.71 0.01 0-9 2.10 0.49 -0.12 0.82 0.51 0-5 -0.29 1.07 -0.06 0.97 0.00 0-9 -0.48 1.14 -0.09 0.94 0.00 0-13 -0.41 1.13 -0.10 0.94 0.00 0-10 1.91 0.59 -O.IS 0.65 __0.00 0-10 257 0.40 -0.09 0.51 0.00 0-5 1.99 0.59 -0.14 0.61 0.14 _ 0-9 2.48 0.51 -0.19 0.75 0.28 0-5 3.79 0.17 -0.10 0.45 0.51 0-9 4.41 0.04 -0.10 0.70 0.82 0-4 -1.90 1.49 0.08 0.77 0.01 0 ~-2.41 1.63 -0.12 0.87 0.00 __-12 -1.56 1.39 -0.09 0.86 0.00 0-10 1.39 0.60 -0.08 0.80 __0.30 ~5 0.14 0.97 -0.28 0.70 0.05 ~5 0.52 0.88 0.01 0.93 ~0.00 0-4 -0.63 1.16 -0.15 0.85 0.00 _~0- _ 0.61 0.86 -0.06 0.44 0 00 0-5 2.46 0.42 -0.07 0.20 0.38 ~_- 1.08 o.n -0.03 0.13 0.09 0-5 2.53 0.43 -0.16 0.93 0.01 0-15 2.92 0.32 -0.13 0.86 ~0.10 0-3 0.48 0.88 0.02 0.32 0.00 0-7 1.02 1.23 0.14 0.66 0.00 61 Significant? No Yes Yes No Yes Yes Yes No Yes Yes

Table 10. Effect of irutial smoothness on future smoothness for various SHA Projects (continued). Proj ect . ID _ US 8 EB & WB Marinette Co. _ US 8 EB & WB Mannette Co. _ US 10 EB & WB Trempealeau Co. US 10 EB & WB Trempealeau Co. _ US 12 EB &WB Tefferson Co. _ US 12 EB & WB efferson Co. _ ~_ WI13NB&SBTa lorCo. Y _ WI 13 NB & SB Ashland Co. (2) _ WI 13 NB & SB Ashland Co. (2) ~_ US14EB&WBRockCo. _ US 14 EB &rWB Rock Co. _ US 51 NB & SB Vilas Co. _ _ Pavement Type AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC AC/AC US 51 NB & SB Iron Co. (1) AC/AC , US 51 NB & SB Iron Co. (1) US 51 NB & SB Iron Co. (2) US 51 NB & SB Iron Co. (2) AC/AC = AC/AC = AC/AC US 51 NB & SB Iron Co. (2) AC/AC WI 11 EB & WB Walworth Co. WI 11 EB & WB Walworth Co. WI 11 EB & WB Walworth Co. _US 8 EB & WB Barron Co. US 8 EB & WB Barron Co. US 12 EB & WB Eau Claire Co. US 12 EB & WB Eau Claire Co. AC/AC - AC/AC . AC/AC _ AC/AC - AC/AC AC/AC AC/AC US 12 EB Sauk Co. WI13NB&SB Wood Co. WI 13 NB & SB Wood Co. WI 13 NB & SB Wood Co. US14EB&WBIowa&Dane Co. US14EB&WBIowa&Dane Co. US 41 NB Waukesha & Washington Co. . AC/AC AC/AC AC/AC AC/AC AC/AC AClAC_ AC/AC I 43 NB Ozaukee Co. I-43 NB Ozaukee Co. I 43 SB Ozaukee Co. AC/AC AC/AC AC/AC I 43 SB Ozaukee Co. ~AC/AC _ I-43 SB Ozaukee Co. AC/AC Const Year 1980 1980 1984 1984 1982 1982 1988 1984 1984 1982 1982 1989 1987 1987 1981 1981 1981 l 1981 _ 1981 . 1981 1985 1985 . 1984 1984 1986 1986 1982 1982 1982 . 1985 1985 1989 1982 l 1982 1982 1982 1982 . l 1982 aO (ins) . 1.75 1.70 1.44 1.96 0.96 1.62 -2.68 0.57 2.73 -0.30 0.46 2.23 -0.29 2.02 0.95 1.81 2.34 7.42 13.41 18.03 0.12 1.10 -0.87 -2.11 4.57 10.13 -2.06 -0.26 1.59 0.25 -1.95 3.06 1.52 2.73 3.42 0.35 0.81 2.13 | alof | I si I 1 1 1 055 1 0.56 0.68 1 1 0.57 1 1 1 0.72 .65 . . :.88 1 0.42 l 06 .867 l 0.50 057 rO.781 1 058 L 044 l -0.74 -2.18 l 1 -3.27 1 . . 097 0.72 l 9 45 L 0031 l 1 -1.29 1 EU 04 0.63 l 1 095 1 E~ 1 -2.17 1 FO.651 . . 1 033 1 1 0.17 0.92 1 1 0.77 1 1 1 1 0.48 1 a2 of Time -0. -0. -o.os ~.12 -0.16 -0.15 -o.os -0.17 -0.33 -0.27 -0.21 -0.07 -0.13 -0.14 .11 -0.13 -0.12 -0.23 -0.15 -0.15 -0.10 -0.11 -0.07 -0.03 o.oo -0.13 -0.25 -0.16 _-0.15 -o.os -0.12 -0.17 -0.25 -0.13 -0.14 ~.26 -0.15 -0.15 1 R2 1 l 1 0.87 1 o9o 1 0.89 1 0.75 1 0.94 0.94 1 0.62 1 1 1 0.85 0.8 0.89 0.93 0.38 0.72 1 o.9o 1 0.91 1 1 o94 0.94: l 0.94 0.83 1 0.89 1 1 0.88 0.77 o74 0.6 L o.oo 0.46 0.94 0.87 L 0.92 l 1 0.60 1 1 0.51 1 1 1 0.46 1 oe83 0.65 0.82 1 0.84 0.73 0.88 p-value | _ 1 l o.oo 1 o.oo 1 o.oo 1 l 0.01 1 1 o.oo 1 o.oo 1 o.oo 1 1 oooo 1 0.44 0.02 1 1 0.02 1 ooo 1 oo3 0.13 1 o.oo 1 . o.ool o.oo 1 0.35 1 o.os 1 l o.oo 1 . 0.02 1 l o.oo 1 l o.oo 1 0.97 1 l 0.32 1 0.04 1 1 0.23 _0.39 1 o.oo 1 0.02 1 1 0.10 o.oo 1 0.24 0.45 1 o.oo 1 0.04 1 l o.os 1 Significant? Yes Yes Yes Yes No Yes Yes Yes Yes No No No Yes

Table 10. Effect of in~tial smoothness on future smoothness for various SHA Projects (continued). Project ~ Pavement ~ Const. ~ No. of ~ R g ~ aO ~ alof I a2 of 1 R2 1 Va Ue ID ~Type ~ Year ~ Sections ~g '~ (ins) ~ Si ~ Time ~P ~ Sig~ US 53 NB & SB Douglas & Washburn Co. AC/AC 1988 8 0-5 1.46 0.67 -0.14 0.72 0.01Yes WI 67 NB & SB Walworth & Waukesha Co. | AC/AC | 1986 | 6 | ~ 4 | -0.40 | 1.12 | 0.01 0.29 | 0.03| Yes WI 67 NB & SB Walworth & Waukesha Co. ~AC/AC ~1986 ~6 ~ -8 ~-1.71 ~1.51 ~-0.06 0.53 ~0.01 US 151 NB & SB Dodge Co. (2) AC/AC 1983 3 0-10 -0.78 1.16 -0.02 0.34 0.05Yes US151 NB&SBCalumet& Manitowoc Co. T AC/AC | 1984 | 5 T o lo | 8.07 T -l os | -0.26 0.89 T 0.22 r ~ I-94 WB St. Croix (2) AC/PCC 1988 9 0-3 1.13 0.73 -0.06 0.15 0.07 Yes I-94 WB St. Croix (2) ~ AC/PCC T 1988 1 9 T ' -6 1 2 45 T 0 47 1 -0 17 0.46 | 0.29 I-94 WB Dunn Co. AC/PCC 1986 8 0-4 0.91 0.81 -0.03 0.30 0.01 Yes I-94 WB Dunn Co. 1 AC/PCC 1 1986 1 8 1 ~ -8 1 1.85 1 0.64 1 -0.08 0.38 1 0.12 1 No I-94 WBEauaaire Co. 1 AC/PCC 1 1985 1 3 1 ~ 5 1 0.82 1 0.84 1 4.18 0.90 1 0.10 1 No 1-94 WB Eau Claire Co. 1 AC/PCC 1 1985 1 3 1 ~ -9 1 -0.27 1 1.06 1 -0.14 0.90 1 0.01 1 Yes 1-94 WB Tackson Co. | AC/PCC | 1989 1 3 1 ~ -5 7 1~95 1 0.62 | -0.08 0.21 7 0.48 1 No I-94 WB Dane Co. 1 AC/PCC 1 1989 1 3 1 ~ -5 1 1.82 1 0.63 1 -0.16 0.58 1 0.06 1 Yes I-90 EB Juneau Co. AC/PCC 1988 4 0-3 -2.27 1.48 -0.02 0.75 0.00 Yes 1-90 EB Tuneau Co. 1 AC/PCC 1 1988 1 4 1 (-6 1 -0.13 1 1.06 1 -0.09 0.49 1 0.01 1 Yes 1-90 WB Tuneau Co. 1 AC/PCC 1 1987 1 4 1 ~ -4 ~1~4.os 1~o o2 ~1 ~-0.05 0.33 1 0 95 1 No I-90 WB Tuneau Co. 1 AC/PCC 1 1987 1 4 1 ~ -7 1 5.29 1 -0.25 1 -0.10 0.68 1 0 49 1 No 63

· O to 15 years-All projects with roughness data available within the first 15 years of the life of the pavement. O to greater than 15 years All projects with roughness data available spanning more than the first 15 years of the life of the pavement. Thus, several regression analyses may have been performed for a given project, depending on its age and the availability of intermediate data points. For instance, a project having yearly roughness data for the first 4 years following construction would only qualify for evaluation under the O-to-5 age range. A project having yearly roughness data for the first ~ ~ years following construction would qualify for evaluation under all four age ranges. A project with initial roughness data and yearly roughness data for years 13 through 18 following construction would qualify for evaluation under the O-to-15 and 0-to-greater-than-15 age ranges. The information presented in table 10 represents a significant amount of data. To aid In the interpretation of this data, the results of table 10 have been summarized In table At, which indicates by State and by pavement type the percentage of projects in which initial pavement smoothness had a significant effect on the future smoothness of Me pavement. This table is for all paving projects over all age ranges. Table Il. Summary of sigruficar~ce of initial pavement smoothness, all age ranges. No. Of Significant ProjectstTotal No. Of Projects (Percent)* l State AC T PCC | AC/AC 1 AC/PCC || Total, New ~Total, Pavements | Pavements | Pavements | Pavements || Const. | Overlays AL 13/14 26/26 39/40 (93%) (100%) (98%) 12/20 11/17 9/13 3/6 23/37 (60%) (65%) (69%) (50%) (62%) GA 4/5 4/6 6/9 5/12 8/11 (80%) (67%) (67%) (42%) (73%) IL 15/20 4/4 15/20 (75%) (100%) (75%) KY 2/2 6/6 12/12 13/24 8/8 (100%) (100%) (100%) (54%) (100%) 7/7 6/6 2/2 13/13 (100%) (100%) (100%) (100%) 1/2 1/3 0/1 2/5 (50%) (33%) (0%) (40%) SD 11/13 11/13 (85%) (85%) WA 5/5 5/6 7/7 10/11 _ (100%) (83%) (100%) (91 %) 9/11 28/39 36/49 22/30 37/50 (82%) (72%) (73%) (73%) (74%) TOTAL 53/66 113/142 70/91 49/78 166/208 119/169 . (80%) (80%) (77%) (63%) . (80%) (70%) 12/19 (63%) 11/21 (52%) 4/4 (100%) 25/36 (69%) 2/2 (100%) 0/1 (0%) 717 (100%) 58/79 (73%) For example, for Alabama AC pavements, 13 out of the available 14 projects (or 93%) showed a significant correlation between initial pavement smoothness and smoothness measured at some point in the future. 64

Table I: generally shows that initial smoothness has a sigruficant effect on the future smoothness of the pavement for all States and for all pavement types. There is some variability from State to State and from pavement type to pavement type, but overall the trend is that initial pavement smoothness has a significant effect on the future smoothness of the pavement. Figures 24 and 25 illustrate In graphical format the results shown in table I! (by State and pavement type). Figures 26 and 27 summarize the percentage of projects for which initial pavement smoothness is significant by pavement type and type of construction, respectively. It is observed from both figures that initial smoothness is significant for a greater percentage of projects for new AC or new PCC construction than for AC it, , , e ~. ~· e ~ ~. · ~ ~. ~. ~. ~A overlay construction. initial smoothness was slgmucant in flu percent or notn tne AN and PCC pavement projects, with slightly lower percentages for AC overlay projects (77 percent and 63 percent for AC/AC and AC/PCC paving projects, respectively). This is probably explained by the fact that the performance of AC overlays is strongly Influenced by the condition of the underlying pavement; reflection cracking or the presence of localized failed areas significantly detracts from the performance (and smoothness) of the overlay. In particular, it is interesting to note that the percentage of projects for which initial smoothness is significant is least for AC/PCC paving projects, suggesting that reflection cracking (Ion" a performance nemesis for this pavement type) has a strong effect on the performance of AC/PCC pavements. Figure 27 shows, for all age ranges, the percentage of projects for which initial smoothness is significant for new construction or for overlay construction. For new pavement construction (both AC and PCC pavements), 80 percent of the projects show that initial smoothness has a significant effect on the future smoothness of the pavement. The percentage of projects being significant is less for AC overlay projects (70 percent), again probably clue to the strong influence of the underlying pavement condition on the performance of Me overlay. Tables 12 through 15 break out the percentage of projects showing the significance of initial pavement smoothness by age range. Table 12 provides the results for projects in the O-to-5 age range, table 13 provides the results for projects in the O-to-1O age range, table 14 provides We results for projects in Me 0-to-15 age range, and table 15 provides the results for the projects in Me 0-to-greater-than-15 age range. Figure 28 summarizes the results of tables 12 through 15 by age range and pavement type. This figure shows how the percentage of projects for which initial smoothness is significant varies with age. Considering all types of pavement construction, the percentage of projects for which initial smoothness is significant typically ranges from about 60 to 80, although several of the new construction projects do exceed 80 percent in some cases. For example, new AC pavement projects show that 83, S1, and 86 percent are significant in the O-to-IO, O-to-15 and 0- to-greater-than-15 age range categories, respectively. New PCC pavement projects show that 86 percent are significant In the O-to-lO age range, but this drops off to only 59 percent significant in the 0-to-greater-than-15 age range. 65

All Projects, All Age Ranges ~ 80 ._ ._ .= 60 In _, ._ a o 40 o AL AZ GA l IL KY Ml MN SD WA Wl · AC PCC AC/AC 1:Z AC/PCC Figure 24. Percentage of projects showing significance of initial pavement smoothness (by State and pavement type). All Projects, All Age Ranges 100 - ~ 80 - ~_ .= 60 In - ~, 40 o Q 20 o l AL AZ GA IL KY Ml MN SD WA Wl | · New Constr. 1~1 Overlays Figure 25. Percentage of projects showing significance of initial pavement smoothness (by State and type of construction). 66

All Projects, All Age Ranges 100 80 ._ ._ ~ 60 ._ Oh - ~, 40 ._ o 20 o 80 80 77 63 AC PCC AC/AC AC/PCC Figure 26. Overall percentage of projects showing significance of initial pavement smoothness (by pavement type). All Pro sects, All Age Ranges 100 80 co - ._ cn 60 ._ En An ~ 40 ._ o 20 o 80 New Const. 70 Overlays . . ... . . . . Figure 27. Overall percentage of projects showing significance of initial pavement smoothness (by type of construction). 67

Table 12. Summary of significance of initial pavement smoothness, O-to-5 age range. | No. Of Significant Projects/Total No. Of Projects (Percent) State | AC PCC Pavements Pavements AC/AC AC/PCC Total, New Total, Pavements Pavements Cons t. Overlays 4/4 (100%) AZ GA SD 3/S (60%) 1/2 (50%) WA _ WI TOTAL 10/14 (71%) 4/4 (100%) 4/5 (80%) 0/1 (0%) 1 8/8 (100%) 5/5 (100%) 0/1 (0%) 13/16 (81%) l 26/34 (76%) 19/23 (83%) 32/40 (80%) 1/3 (33%) 2/4 (50%) 2/2 (100%) 6/11 (54%) 8/11 (73%) 19/31 (61%) 5/9 (56%) 6/9 (67%) 2/2 (100%) 4/4 (100%) 1/4 (25%) 2/3 (67%) 11/16 (69%) 0/1 (0%) 13/16 (81%) 1! 36/48 (75%) 51/71 (72%) 27/34 (79%) Table 13. Summary of significance of Initial pavement smoothness, O-to-IO age range. No. Of Significant Projects/lotal No. Of Projects (Percent) State AC ~PCC AC/AC AC/PCC Total, New ~Total, Pavements Pavements Pavements Pavements Const. Overlays , . . i .. , 9/10 (90%) 3/6 (50%) 2/2 (100%) 1/1 (100%) WA WI 22/22 (100%) 415 (80%) 1 4/6 (67%) 1 2/3 (67%) 3/3 (100%) 3/3 (100%) AL 31/32 (97%) 7/11 (69%) 5/5 (100%) 3/3 (100%) 2/2 (100%) 13/13 (100%) 1 1 1/1 (100%) 1 2/2 (100%) 6/6 (100%) 9/10 (90%) 4/4 (100%) 2/4 (50%) 5/5 (100%) 3/3 (100%) 13/19 (68%) I/3 (33%) 2/2 (100%) 6/11 (54%) 9/14 (64%) 4/4 (100%) 1 1 1 14/22 (64%) 6/9 (67%) 3/7 (43%) 2/2 (100%) 11/16 (69%) 2/2 (100%) 3/3 (100%) 22/33 (67%) 11 TOTAL 1 25/30 (83%) 1 63/73 (86%) T 27/37 (73%) 1 22/35 (63%) 11 88/103 (85%) 1 49/72 (68%) 11 68

Table 14. Summary of significance of initial pavement smoothness, O-to-15 age range. No. Of Significant ProjectsITotal No. Of Projects (Percent) . State AC PCC AC/AC AC/PCC Pavements Pavements Pavements Pavements l ~AL ~- ~- ~- ~- it - . - AZ 4/6 (67%) 2/4 (50%) 1/1(100%) GA 1/1(100%) 1/2 (50%) 2/5 (40%) IL 3/3 (100%) _ _ 2/2 (100%) ~2/2 (100%) 1/2 (50%) ~- ~- ~- ~- ~1 - L MN ~| | _ | _ || _ | _ SD ~- 11 - I _ WA 5/5 (100%) 1/2 (50%) 4/4 (100%) _ . WI 1 3/4 (75%) 1 5/5 (100%) 1 4/7 (57%) 1 5/5 (100%) 1 1 1l _ TOTAL I 13/16 (81%) 1 14/18 (78%) 1 11/14 (79%) 1 8/12 (67%) 1 , 1 Total, New Total, ConsL Overlays 6/10 (60%) 2/3 (67%) 3/3 (100%) 2/2 (100%) 6/7 (86%) 8/9 (89%) 1/1 (100%) 2/5 (40%) 3/4 (75%) 4/4 (100%) 9/12 (75%) 27/34 (79%) 1 19/26 (73Yo) 1 Table 15. Sununary of significance of initial pavement smoothness, O-to-greater-than-15 age range. l No. Of Significant Projectstrota1 No. Of Projects (Percent) State | AC |PCC | AC/AC | AC/PCC || Total, New | Total, l Pavements Pavements | Pavements | Pavements || Const. | Overlays . AL i i j j | ~- ~ - AZ 1 2/3 (67%) 1 1/3 (33%) 1 _ 1 11 3/6 (50%) it. , , 1 IL r T:l9/l5(64%): ~ KY ~ - ~ - ~ ~ ~ ~ ~ ~ ~ - ~ - ~ - ~ - Id- ~ _ I MN I - I - T - I - 11 - I j WA | | I I || , . . , , , , .... ,, ~ I WI I 3/3 (100%) 1 - I - I - 11 3/3 (100%) 1 1 . . .. . 1 TOTAL 1 5/6 (83%) 1 10/17 (59%) 1 - 1 - 11 15/23 (65%) 1 - 1 69 1 1- 11

100 ~ 80 cut ._ ._ ~ 60 ._ Oh In `1' 40 ._ o 20 o ~ ~ 11 Oto 15Age Oto> 15Age Oto5Age Oto10Age ~ · AC ~ PCC ~ AC/AC 11:1 AC/PCC | Figure 28. Percentage of projects showing significance of initial pavement smoothness (by age range and pavement type). The AC overlay projects show fairly consistent results over time. For example, the AC/AC overlay projects show that 80, 73, and 79 percent are significant in the 0- to-5, O-to-IO, and 0-to-15 age range categories, respectively. The AC/PCC projects show that 6l, 63, and 67 percent are significant in Me 0-to-5, O-to-IO, and 0-to-15 age range categories, respectively. As mentioned earlier, the AC/PCC projects may be showing less significance than AC/AC pavements due to the adverse effects of reflection cracking. Unfortunately for bow overlay types, no projects were available that were greater Wan 15 years old so that age bracket could not be examined. Figure 29 presents overall summaries by type of construction and by age range. Similar to what was observed previously, this figure shows that initial smoothness is significant for a greater percentage of projects in the new construction category (AC and PCC pavements) than for projects in the AC overlay category (AC/AC and AC/PCC). Considering new construction projects, the percentage of projects for which initial smoothness is significant Increases slightly and then begins to fall off for pavements greater than 15 years old. AC overlay projects, on the other hand, show fairly constant values over the different age ranges, although no projects are available for the O-to-greater-than-15 age range category. 70

All Projects, O to 5 Year Age Range 100 100 m6. 15 ~' 2 1. ~ 20 New Const. Overlays All Projects, O to 10 Year Age Range 85 New Const. 68 Overlays All Projects, O to 15 Year Age Range l | All Projects, Greater Than 15 Years Age Ra age 100 ~1 1 1 100 .. ~ 1 New Const. Ovedays I | New Const. Overlays Figure 29. Percentage of projects showing significance of initial pavement smoothness (by age range and type of construction). 71

Another way of looking at the data in table 10 is to examine the magnitude of the al regression coefficient. It is recalled that this value represents the slope of the regression line between the future pavement smoothness and the initial pavement smoothness at some given time period. A value of I.0 indicates a strong one-to-one relationship between initial smoothness and future smoothness, and also suggests that the roughness curves for different pavement sections roughly parallel each other over time. Values greater than ~ indicate divergence of the roughness curves over time, whereas values less than ~ suggest convergence of the roughness curves (although the convergence is at some point outside of the available data range). Table 16 summarizes the average al values by pavement type and age range. The results of table 16 are depicted in figures 30 and 31. Figure 30 illustrates the average a, value by pavement type for all projects and all age ranges. The average al value is highest for PCC pavements, averaging 0.85, indicating Mat every 1 in/ml (0.016 m/km) increase in initial roughness is accompanied by an increase of 0.85 ~n/mi (0.014 m/km) in the future roughness of We pavement over fume. The AC/AC projects show the lowest average a, value, averaging 0.60, with the AC/PCC projects showing a higher average a, value, 0.71. Figure 31 shows Me average al value by type of construction for all projects and all age ranges. This figure indicates higher a. values for new construction Projects ~ cab - ~ ~ - - ~ ~ (0.82) than for overlay projects (0.65~. Again, Allis suggests that the performance of AC overlay projects is more dependent on other factors (e.g., reflection cracking) than on Mutual smoothness. However, the average al value for the overlay paving projects is still 0.65. Table 16. Average initial smoothness regression coefficient Gaul by pavement type and age range. l l Av Rage Initial Smoothness Regression Coefficient' a I Age l | Range AC PCC AC/AC AC/PCC New | Pavements Pavements | Pavements Pavements ||Constructic nl Overlays || 0-5 1 1.02 0.87 1 0.69 0~7411 0~91 1 0 71 1~ . 0-10 1 0.75 0.86 1 0.58 0.7611 0.83 1 0.67 1; 0-15 1 0.66 1 1.01 1 0 39 1 0.4811 0.84 1 0 43 0->15 1 0.40 1 059 1 - 1 - 11 054 1 - .. . . . , 0.75 0.85 0.60 0.71 0.82 0.65 Ranges l l ll l l 72

All Projects, All Age Ranges 1 0.8 ._ ._ It ~ a` O u. cat 0.4 0.2 o 0.85 ~ 0.75 AC 0.71 0.60 PCC AC/AC AC/PCC Figure 30. Average a1 values by pavement type for all projects and age ranges. All Projects, All Age Ranges ~ 0.8 ._ ._ = g 0.6 c' ~ 0.4 an 0.2 o 0.82 New Constr. 0.65 Overlays Figure 31. Average a, values by type of construction for all projects and age ranges. 73

The magnitude of the average al coefficients shown in figures 30 and 31 suggest that the initial smoothness does affect the future smoothness of the pavement; it also suggests that the performance curves are converging and some point is reached at which initial smoothness does not affect future smoothness. However, it is important to note that these relationships are based on linear regression analysis, which can not account for nonlinear effects at different points in the life of the pavement. Because of this, nonlinear regression techniques are used In the evaluation of the effect of initial smoothness on pavement life (discussed later). S~rnilar to the previous evaluation, the average a, value was also examined by age group; figure 32 indicates the average al values for each pavement type grouped by the four different age ranges (0 to 5 years, 0 to 10 years, 0 to 15 years, and 0 to greater Man 15 years). This figure indicates an average a1 value of i.02 for AC pavements in the O to 5 age range, indicating one-to-one correspondence between initial smoothness and future smoothness for AC pavements in that age range. However, the average al value then begins to drop off for AC pavements outside of the O-to-5 age range, showing levels of 0.75, 0.66, and 0.40 for the O-to-IO, O-to-15, and O-to-greater-than-15 age range categories, respectively. This suggests that initial pavement smoothness does not have as significant effect for the older AC pavements. For PCC pavements, We average al value shows an initial increase over the Increasing age ranges, and then falls off. From a value of 0.87 in the 0-to-5 age range, the average al value remains about the same (0.86) in the O-to-IO age range, then 1 .4 - s: ._ (, 1 <' 0.8 `1, 0~6 `1, 0 4 0.2 o Oto5Age Oto10Age Oto15Age Oto>15Age | · AC ~ PCC ~ AC/AC ED AC/PCC | Figure 32. Average a1 values by pavement type for all projects and age ranges. 74

increases to 1.01 In the 0-to-15 age range before falling off to 0.59 In the O-to-greater- than-15 age range. As with the AC pavements, this suggests that Initial pavement smoothness has a less significant effect on the smoothness of older PCC pavements. Figure 32 shows that the average al values for the AC overlay projects are less than those for new construction. The average al values decrease with increasing age range, suggesting a weaker influence of initial smoothness on the future pavement smoothness. The average al values for the overlay projects are generally in the 0.4 to 0.7 range, although data for pavements greater than 15 years old are not available. Summary The results of the evaluation of the SHA data from inservice highways strongly indicate that initial pavement smoothness has a significant effect on the future smoothness of the pavement for AC, PCC, and AC overlay projects. Figure 33 indicates the overall percentage of projects for which Initial pavement smoothness was found to be significant (broken out by pavement type), using multiple linear regression analysis techniques. Considering all age ranges, initial pavement smoothness was shown to have a significant effect in 80 percent of both new AC construction projects and new PCC pavement construction projects. AC overlay projects show a somewhat lesser percentage (70 percent), suggesting that factors other than initial smoothness (such as Me condition of We underlying pavement and the development of reflection cracking) may substantially influence their performance. Figure 34 summarizes the overall significance results for the 10 States providing data to the study for new construction (both AC anc! PCC pavements). All of the States but one show Mat ~rutial pavement smoothness is significant in more than 60 percent of the projects evaluated. Ibis trend is borne out in each of the four primary cInnatic regions (dry-nonfreeze, dry-freeze, wet-nonfreeze, and wet-freeze) of Me U.S. Overall, 80 percent of the new construction projects showed that initial pavement smoothness is significant. A similar chart is shown in figure 35 for AC overlay construction. For this situation, six of the eight States providing data show Mat Initial smoothness is significant to the future smoothness of the AC overlay in 60 percent of the projects; overall, initial pavement smoothness in significant in 70 percent of the AC overlay projects. The a, regression coefficient between Initial smoothness and future smoothness was examined and found to average 0.82 for new pavement construction (0.75 for AC and 0.85 for PCC) and 0.65 for overlay construction (0.60 for AC/AC and 0.71 for AC/PCC). The greater the value of this regression coefficient, the stronger the relation between ~rutial and future pavement smoothness. As most of the average al coefficients were between 0 and l.0, this suggests that the pavement sections are converging at some point outside of the available data range. 75

AC PROJECTS Significant 80% PCC PROJECTS Significant 80% ~ ;;;;;;;;;;;;;;.;, ................. .............. _............ Not Significant 20% AC/AC PROJECTS Significant 77% \.,,,,....,..,....1 `................. A................ ~ - ~.............. qb, - ~........... A,.......... ........ - `,...._ Not Significant 23% Not Significant 20% AC/PCC PROJECTS Signlflcant 63% Not Significant 37% Figure 33. Breakout of overall significance by pavement type. 76

~ ~ ~ ~ -; ~ ~ ~ it{ x ~- , , _ _ ::::::~::^i;::::: :.~:::_ :' 1 ~s~ . ~.-. ~ ~-- ~ ~_ ,,,,. '''''''''~ : ''''''_ - ~. ~::::: ~ 'a 1 ~ ~ - ~ ~ ·_I · · co v) · - v) us o cq a a, pa - · ·= · - # ~:~:~:~ : ~-t ~ # / ~ -~l 77 1 _ CLl ~ a_ / O a' ~0 O I I ad U ~ ~ ·0 - ~ o O cn o A ct u) ~ - v) o o ·~ u v) o u o u · -

for - - - ~ 1 w ~, ~1 - - -t ~¢ t-t / ~ ~ ~- 78 r ~ Y .0 ~ I ~ ~ 1 - O A) _ ~ ·0 _~0 ~ ·_ _ ~ ~ ·0 ~ U. O ~ _ ~ - _ ·0 5 O O U) ' ~' ~ cn cn a; o o Cal - - ·_1 it_ ._ ._ - - ·_I cry __ CO Cal As · _\ - I _ Sin to a) ._

Although long-term smoothness is related to crucial smoothness on many projects, the fact that initial smoothness was not significant for every project is not unexpected. Many factors, including equipment measurement errors; variability In design, materials, and construction; subgrade settlements/heaves; topography variation; and the presence of bridges, culverts, and other structures along the highway, can obscure or eliminate the positive effects of initial smoothness on future smoothness. To reiterate the significance of these findings, the available data show that for two pavement sections with similar design features, the section built smoother will remain smoother for many years after construction. The performance curves of the two sections generally remain parallel for many years. It then follows that the smoother pavement will result in lower vehicle operating costs and longer service life, all other things being equal. . ~ ~ Thus, there is an economic benefit associated with - v ~ ~ smoother pavements, as well as the "ride quality" comfort benefit to highway users. Evaluation of Other Data Sources Two over data sources, the AASHO Road Test and the LTPP GPS program, contain useful data that were also evaluatec! by He research team. The results of those evaluations are presented below. AASHO Road Test Performance data from the AASHO Road Test are available for both AC and PCC pavements and are provided In units of present serviceability index (PSI) over time. Such data are available for paired replicate sections of the same design and subjected to the same traffic loading. One-tailed, paired l-tests were conducted to compare the overall mean of He paired replicates over He 2-year duration of the Road Test. The objective of He one-tailed t-test is to determine if He Initially smoother section maintains a higher overall mean serviceability than He initially rougher section. The null hypothesis (Ho) of the paired t-test is to assume that the mean smoothness of He initially smoother replicate section (lo,) equals the mean of He initially rougher replicate section (~2)' In other words: Ho p~ = p2 where: p~ = Mean smoothness of the Initially smoother replicate section. p2 = Mean smoothness of the lethally rougher replicate section. The alternative hypothesis for the paired t-test is Hat the mean smoothness of the initially smoother replicate section remains significantly higher than He ~rutially rougher replicate section, that is: Ha ~1 > ~2 79 (2) (3)

A t-statistic value (tca~c) is calculated based on the sample means and standard deviations of the serviceability data for each pair of replicate sections. The calculated t-statistic value is then compared to a tabulated t-statistic (table) for a specified confidence level (90 percent was selected for this analysis). If the t-test tums out to be significant (icalc > ttable), then the null hypothesis is rejected and the alternate hypothesis is accepted, meaning that it can be inferred with 90 percent confidence that the initially smoother sections maintain a higher smoothness level than the initially rougher sections. The t-test results for the AC pavement sections and the PCC pavement sections are presented In table 17 and 18, respectively. The difference in the initial serviceability was computed and the sections were categorized into three classes according to that difference: Class A (initial PSR difference less than 0.3), Class B (initial PSR difference between 0.3 and 0.5), and Class C (initial PSR difference equal to or larger Man 0.51. The results shown In the tables 17 and iS indicate a greater #= ~ If 1 -1 1.1 # 1. I've · .1 ~ ~ #1 · 1 _ # · 1 #1 1 ~ rr 1l~elmood of dliterence in the performance ot the paired sections when the ultterence in the initial serviceability is greater than 0.3 (Class B or C). That is, if the difference tial smoothness is greater than 0.3 serviceability units, the initially smoother section stays smoother over time. Table 19 provides an overall summary of the paired l-tests of the AASHO data. This table shows for We replicates In Class A, me results are quite variable, mostly showing no significant difference In performance over tune. This Indicates mat if Me initial smoothness values of the two replicate sections are about the same (i.e. within 0.3 serviceability units), their overall smoothness serviceability value measurements over time are most likely not significantly different from each other. For Me replicates in class B. a majority of the replicates with higher initial serviceability ma~ntainect a higher level over time. For the replicates In class C, most replicates higher initial serviceability stayed smoother over dine, win all of Me AC sections and two-thirds of the PCC sections exhibiting this trend. Thus, the evaluation of the AASHO data indicates that initially smooth pavements generally stay smoother over time provided that a significant difference exists In the crucial smoothness (say, 0.5 or more serviceability units). SHRP LTPP GPS As mentioned In chapter 3, the major limitation to the use of the LTPP GPS data is the absence of initial roughness values. Since Me vast majority of the GPS sections were constructed prior to 1985, significant gaps (> 3 years) exist between the construction date and He first roughness measurement taken under the LTPP program. However, in an attempt to use the data, He GPS sections with roughness measurements (in this case, ~) within 3 years from the initial construction date were selected for analysis. The time series plots of the available roughness data are provided in appendix C. Due to the limited number of sections identified in each experimental type, no sub-group~ng was done for each pavement type. Thus, some differences exist In the 80

Table 17. Summary of the t-test results for AASHO data, AC pavement sections. . . Section No. Initial PSR PSRo Joop 1 | 2 1 | 2 Difference 732 758 . 133 134 108 149 150 569 570 590 416 453 454 476 317 318 306 264 7 731 709 710 757 . 712 . 107 589 615 616 415 475 305 263 760 776 . 143 . . 144 116 161 162 571 572 598 430 455 456 484 329 330 308 272 737 759 741 742 775 738 115 597 629 630 429 483 307 271 4.10 4.3 1 3.80 1 1 3.80 3.70 3.80 3.90 4.20 4.10 1 4.20 4.30 4.30 4.40 4.10 4.30 4.30 4.30 4.20 4.00 4.00 3.90 3.60 4.00 3.60 4.10 4.40 4.10 4.20 4.40 4.50 4.20 4.20 4.20 4.00 4.00 3.60 3.90 3.80 3.80 3.80 4.10 4.40 4.10 4.40 4.20 4.20 4.20 4.40 4.10 4.20 3.90 @E 3.90 4.10 1 1 3.70 1 1 4.10 1 3.40 3.90 1 4.50 4.oo 4.20 4.50 4.40 r4.4O 0.30 0.50 0.70 0.50 0.00 0.20 0.20 | Paired t-test | Significant? Category I p-value (1-tail) 1 (P-1<0.1?) 0.20 0.00 0.10 0.40 I 0.00 1 0.20 0.20 0.10 0.10 0.10 0.10 0.20 1 o.oo 1 0.10 0.40 1 1 o.oo 1 0.50 B A B A A A A A B A A A _ 1 | 0.10 1 A 1 0.13 1 No 1 1 1 1 1 0.30 1 B 1 0.17 1 No 1 0.20 1 A 1 0.19 1 No 1 1 1 1 1 0.20 1 A I 0.41 1 No , 1 0.18 1 No 1 1 1 0.48 1 T- I n12 0.13 No No I No 1 I No ~3i Tail:, 1 A 1 0.18 0.20 1 A 1 0.14 . . . 1 A 1 0.13 1 I 1 1 I A 1 0.20 1 No I A 1 0.22 1 No I A 1 0.18 1 No A 1 0.15 I A | 0.00 l l | B | 0.00 I Yes I A I C 0.42 0.23 0.00 No No No No No Yes A 0.00 0.00 0.00 0.00 Yes | Conclusion | | Not Significant || Not Significant || | Not Significant || l Not Significant | Not Significant || | Not Significant | Not Significant l | Not Significant l | Not Significant l | Not Significant l | Not Significant l | Not Significant | | Not Significant | | Not Significant | | Not Significant | | Not Significant | | Not Significant | | Not Significant l l Not Significant | | Not Significant 1 l Significant | l Significant | | Significant l 1 1 | Significant | | Significant | | Significant | | Significant | Significant | 1 Significant | | Significant | | Significant | . | Significant | Categories: A: 0-0.3; B: 0.3-0.5; C: 2 0.5 81

Table 18. Summary of the t-test results for AASHO data, PCC pavement sections. Section No. Initial PSR Loop 1 1 2 1 1 1 2 2 797 _ 3 221 3 203 3 204 237 . 238 641 . 642 697 671 672 503 533 534 372 367 368 . 363 364 398 519 ~ 520 798 779 780 222 193 194 669 . l 670 504 511 512 371 337 338 777 245 251 252 247 248 705 706 655 687 688 543 539 540 382 389 390 377 378 656 521 522 778 815 816 246 217 218 707 708 544 541 542 381 345 346 4.20 4.90 4.70 4.70 5.00 5.00 4.90 4.70 4.90 4.90 4.60 4.70 4 70 4.70 4.50 4.70 4.40 4.70 4.80 4.70 4.60 4.70 4.30 4 50 4.60 4.60 4.60 4.60 l 4.90 4.80 4.80 4.70 4.80 4.50 4.60 4.50 _ 4.40 4.70 4.50 3.90 l 4.80 l 4.80 4.80 4.80 . ~ 4.80 4.7 4.70 l 4.30 4.70 4.70 4.60 4.70 4.60 4.60 4.80 4.50 4.60 4.30 4.00 4.10 4.80 4.90 l 4.80 4.70 4.7 . ~] 4.40 4.60 4.60 4.70 4.70 j 0.20 | A 1 1 1 0.20 1 A 1 0.20 1 A 1 1 1 0~80 1 C 1 1 0.20 1 A 1 0.20 1 A 1 1 0.10 1 A I 0.10 1 A l 1 1 0.10 1 A l 1 1 I 0.10 1 A I T o.1o 1 A l 1 1 1 1 0.00 1 A I B A | A B 0.40 0.00 . 0.20 1 ~ 0.10 1 ~ 0.30 1 0.10 0.20 0.14 1 0.10 0.00 0~50 1 ~ ~ 0.30 0.20 . 0.20 1 1 0.10 1 1 0.10 1 1 0.30 1 1 0.20 0.10 . 0.10 0.20 l Categones: A: 0-0.3; B: 0.3-0.5; C: > 0.5 ~ 0.16 ~:x 1 0.24 1 0.24 1 0.27 1 0.39 0.42 . 0.36 . 0-44 . 0.43 0.14 0.21 . 0.45 . 0-34 0.31 . 0.21 1 1 0.24 1 1 1 1 1 A 1 0.14 1 1 1 1 I A I 0.19 l 1 1 1 1 A 1 0.07 1 1 1 1 1 A 1 0.02 1 1 1 A 1 0.04 1 1 1 A 1 0.06 1 c 1 o.oo c 1 o.oo l A 1 0.02 1 B 1 000 ~T~ 1 A I 0.01 l A | 0.10 A 1 0.05 1 B I 001 0.02 . 0.09 n.os 82 | PSRo | I Palred t-test 1 Significant? | 11 Difference | Category | p-value (1-tail) 1 (P-1<0.1?) I Conclusion 1l ; I Not significant | Not sigr~ificant | Not significant Not significant Not significant Not significant Not significant Not significant Not significant Not significant Not significant Not significant Not significant Not significant | Not significant Not significant Not significant Not significant Not significant Not significant | Not significant Not significant I | Significant | | Significant | Significant | Significant Significant Significant Significant Significant Significant Significant Significant A 1 0.02 I Yes I Significant A | 0.09 | Yes | Significant | A | 0.08 | Yes | Significant . - I No 1 I No 1 I No l I No 1 No No | No 1 1 I No l 1 1 I No l 1 1 I No I 1 1 | No 1 No . . No . No . No . No No No No . Yes Yes Yes . . Yes 1 | Yes 1 | Yes 1 | Yes 1 | Yes 1 I | Yes | Yes 1 1 | Yes l 1 | Yes | Yes | Yes 1 | Yes 1 I Yes

Table 19. Summary of the paired t-test results using AASHO data. , Category ~Diff Number Of ~ ~Not AC Pavements A 1 0 0.3 1 22 1 5 (23%) 1 1 (4%) 1 16 (73%) B 0.3 0.5 5 3 (60%) O 2 (40%) ~ C ~>0.5 1 4 1 4~100%) T ° I ° I PCC Pavements A 1 0 0.3 1 29 1 10 (35%) 1 3 (10%) 1 16 (55%) 1; B 1 0.3 0.5 1 4 1 2 (50%) 1 0 1 2 (50%) C ~>0 5 T 3 ~2 (67%) T O ~ (33°/0) ] design, climate, traffic loadings, and subgrade support conditions between sections. However, the structural number (SN) of the AC pavements and the slab thickness of the PCC pavements were introduced as independent variables in the regression analysis to account for differences in the design between pavement sections. Similar to the analyses presented earlier In this chapter, multiple linear regression of the future IR] was conducted for each GPS experimental type. The earliest IR] value (serving as the crucial roughness indicators, age, and either SN for AC pavements or slab thickness for PCC pavements were used as the independent variables. The results of the evaluation are presented In table 20. Every experimental time group showed strong indication of the effect of the first TR] value on the future To_ . · .~ · ''. · . ~ .~ · · . · ~ ~ '. . ~ Am_ . n measurements, Wltn regression coelUclents 01 the 1rutla1 First) I In measurement all around 1.0. Thus, the available data from the LTPP GPS program also indicate a strong influence of Axial smoothness on the future smoothness of the pavements. Again, it must be recognized that most of the GPS data did not have a "true" initial smoothness value (the value used was taken within 3 years of construction) and the available time period over which He analyses was conducted is less Han 6 years. An ongoing research study analyzing roughness data from the LTPP program has produced plots of roughness over time for various GPS projects (Kohn et al. 19961. An example of one such plot for the GPS-5 (CRCP) sections is shown in figure 36. This figure shows the lapses that exist for the various projects between the time they were constructed and the tune that they were first measured for roughness. In addition, this figure indicates the relative stability of roughness over tune, as well as a strong degree of "parallelism" between the different roughness curves. That is, even though each of these pavement sections are of different design and are subjected to different traffic and climatic forces, the roughness curves appear to approximately parallel each other, suggesting that those pavements that are constructed smoother stay smoother over tune. 83

Table 20. Summary of the regression results for LTPP data. , . Experiment Pavement No. Of Age . 2 a. of So R revalue Conclusion No. Type Sections Range GPS-5 1 AC 1 7 1 1-5 1 1.4 1 0.82 1 <0.01 . GPS-2 | AC | 2 | 2 ~| Not e' Hugh datafor regression GPS-3 1 JPCP 1 8 1 1-5.5 1 1 1 0.81 GPS-5 1 CRCP 1 3 1 2-5.8 1 1.1 1 0.96 GPS-7A AC OL 2 0-5.2 1.1 0.98 . GPS-9 JPCP OL 2 1-5.5 0.72 0.9 1 - 1 Significant ~. . <0.0 <0.0 <0.0 <0.0 Significant Significant Significant Significant GPS-5: All Regions, Overlay Sections Removed 250 200 150 - A; ~ North Central North Atlantic Western Southern ~- ~ x art' ~- ~- 100 so ~,. - O- 1 1 1 - 1 1 1 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Age, years Figure 36e Historical roughness plots for LTPP GPS-5 experiment (Kohn et al. 1996) 84

Analysis of Effect of Initial Smoothness on Pavement Life Introduction The results of the previous section clearly indicate that a pavement constructed smoother than an otherwise similar pavement will remain smoother over many years. This suggests that added pavement life can be achieved by building a pavement smoother, provided that pavement roughness is a major factor influencing the ctec~s~on to rehabilitate that pavement. However, other factors often come into play In the decision to resurface, restore, or reconstruct a pavement, including such items as agency policy, available funding, surface friction, overall pavement condition, maintenance costs, and politics. Depending on the agency's management policies, roughness may not be very important or it may be the primary consideration in the decision to rehabilitate a pavement. In a recent report, it was fount! that almost all responding highway agencies (51 of 53 State and Provincial agencies and 42 of 44 State agencies) cited roughness as a primary variable used In their pavement management system for establishing Me worth of candidate projects (Zimmerman 1995~. Thus, many highway agencies believe that pavement smoothness is a primary factor considered by the highway users as representing the quality of the highway. In evaluating the relationship between Crucial smoothness and pavement life, two different approaches were taken. The primary strategy involved the clevelopment of roughness models for each project using bow linear and nonlinear regression analyses of the available tune-series roughness data. Since age was a variable Incorporated Into the models, each project mode! was then used to predict, for various Crucial smoothness levels, the service lives associated win We pavement reaching a terminal, or trigger, roughness (or serviceability) level. The resulting Initial smoo~ness-pavement life trends for similar projects (i.e., projects from a given State and of a particular pavement type) were then grouped together in order to examine the effect of Initial smoothness on pavement life (defined In terms of the pavement reaching a critical level of roughness). A second approach consisted of analyzing time-series roughness data and corresponding actual pavement failure (i.e., placement of an overlay) data for many pavement projects from two States. This evaluation yielded the same result (i.e., initial smoothness-pavement life trends for sunilar projects), but used actual pavement life data (defined In terms of the pavement requiring an overlay) in determining the relationship. Overview of Data Sources For the first analysis, times-series roughness data were used from Alabama, Arizona, Georgia, Illinois, Kentucky, Michigan, Minnesota, South Dakota, Washington, and Wisconsin. As with the analysis of the effect of initial smoothness on future smoothness, only those projects with three or more sets of time-series roughness measurements were analyzed, as Weir roughness trends were better defined. 85

For the second analysis of initial smoothness and pavement life, comprehensive serviceability-type data and pavement rehabilitation data from Kentucky and Wisconsin were used. These data ~nclucled nutial and final (pre-overiay) serviceability index values for various projects, along with initial construction and resurfacing dates of each project. As discussed in chapter 3, complete pavement management records were obtained from Kentucky that included annual RI values-converted from roughness measurements made with accelerometers and Mays Meters for nearly every interstate and parkway pavement project, beginning in the early 1960s. These records also contained construction and rehabilitation information, such as the date an activity was performed, Me type of activity (e.g., PCC construction, AC overlays, and the resulting cross-section thicknesses. Pavement distress data and 1994 ADT estimates were also available In these records. Some parts of the Kentucky data were excluded from analysis because it was learned from an official in Kentucky's pavement management division that I-71 and northern portions of I-75 were overlaid early In Me design life because of aggregate durability problems. Also, since a major equipment change (from an accelerometer to a Mays Meter) occurred In the late 1970s, that significantly impacted the consistency of Me RI, only new construction projects built before 1978 were used and only resurfacing projects completed after 1980 were used. In both cases, only a small percentage of projects were excluded, as most new construction In Kentucky occurred in the 1960s and 1970s, and most resurfacing work occurred In the 19SOs and 199Os. Although the Wisconsin data were not nearly as comprehensive as those from Kentucky (PS] data were only available from 1980 through 1995), they did provide a sampling of the relationship between initial PSI and pavement life. The pertinent data included pavement type, construction and rehabilitation dates, pavement distress data, and initial and final PS] values for projects constructed since 1980. Evaluation Approach Although the multiple linear regression models developed earlier in this chapter were useful in determining whether initial smoothness had a significant effect or future roughness, some of those models were limited in their ability to predict future roughness because of their linear form. When plotted three-dimensionally, the models give planar relationships, as shown in figure 37. As a first step to enhancing the prediction capabilities, a second set of multiple linear regression models were developed for the projects, with these models Including an initial smoothness-time Interaction term. The form of these regression models is given below: St = aO + nisi + a2t +a3Sit 86 (4)

Final MaysRI 1000 800 600 400 200 20 ~- Age (years) 5 i~60 ,~ 4n InitialMaysRI ,~20 o o Figure 37. Example thre - dimensional plot of linear regression model. where: SO = Pavement smoothness at time t. aO' al, as, a3 = Regression coefficients. Si = Initial pavement smoothness. = Tune (age), years since construction or overlay to time of smoothness S'. In general, the goodness-of-fit of the interaction-term models was only slightly unproved over Me previous models. Although the interaction term provided more flexibility by allowing the planar relationship to curve and twist (see figure 38), it was determined bat In many cases a better fit of the data particularly with respect to the tune variabIc could be achieved using a nonlinear approach. Observations of all of the lime-series roughness data indicated that a multiple nonlinear regression mode! of exponential form was appropriate for most sections. The equation below represents this model: St = aO + as + a2tb2 ~a3Sib31 where: SO = Pavement smoothness at time I. aO, al, as, a3 = Regression coefficients. 87 (5)

Final MaysRI 1000 800 600 400 200 Age (years) , 40 Initial MaysRI _~ o o Figure 38. Example three-dimensional plot of linear regression mode! with Initial smoo~ness-time interaction term. be, b2, be, b4 = Exponent coefficients for initial smoothness, time, and initial smoo~ness-time Interaction variables. Si = Initial pavement smoothness. = Time (age), years since construction or overlay to time of smoothness So. These models were developed using The SASS System for Windows Release 6.~1 statistical software. The development of Me nonlinear models using this software was an iterative process bat required sewing boundary conditions and starting values for each of the coefficients. Initial starting values for the coefficients included setting the aO coefficient to be set very close to zero (0.001) and We other coefficients equal to one All. If these starting values failed to give a reasonable model, a second trial generally consisted of sewing aO equal to 0.001 and me other coefficients equal to 0.~. If a reasonable mode} still was not obtained, subsequent trials consisted of using the results of the previous trials to set each individual coefficient to either 0.: or I.0 independently. Each developed mode! was checked against the actual historical data for reasonableness. Figure 39 represents a thre - dimensional plot of one such developed model. 88

Final MaysRI 1000 800 600 400 200 20 Age (years) o o 'my 100 Initial MaysRI Figure 39. Example ~re~dimensional plot of multiple nonlinear regression model. Throughout the model development process, the individual section data for each project was closely scrutinized. This was to determine the general trend representative of the majority of sections within a project. In some cases, obvious outliers from the project's overall general trend were excluded from the regression analysis. The decision to exclude such data points was based on the belief that certain anomalies (localized settlements/heaves, maintenance patches) existed in the sections. A best-fit model was selected for each project from one of the three previously discussed model forms (multiple linear, multiple linear with an interaction term, or nonlinear) for use in predicting future roughness. For the majority of the projects, the nonlinear model provided Me best overall fit. In each case, the selected model was checked for reasonableness by plotting its predictions against the actual data. After the selection of the best-fit model for a particular project, initial smoothness-pavement life relationships were developed for each project using that model. This was accomplished by first selecting a trigger roughness value (i.e., a value at which rehabilitation is required) for use in each roughness model such that the tune to reach the trigger roughness value could be computed for various ~rutial smoothness values. This approach is basec! on the assumption that roughness is a principal criterion for rehabilitating a pavement. 89

Since each model was based on a certain roughness index and a common level of roughness was needed to serve as the trigger in all models, approximate correlations were developed between all indices over a wide range In roughness. Table 21 contains the approximate values of several roughness indices for different levels of roughness. It also lists the correlation equations that were used and Me reference source from which the equations were taken. Although many correlations of roughness inclices had been developed in past studies, the correlation equations used here seemed to give the best across-the-board approximations of roughness, especially in light of the roughness data collected for this study. In all roughness mode] analyses, roughness trigger values corresponding to PS] of 2.75 were selected for use. For example, for the Arizona roughness models, a trigger MRN of 175 in/ml (2.76 m/km) was used, whereas for Illinois, a trigger TRI of 150 in/ml (2.37 m/km) was used Once the trigger roughness value and a given initial smoothness value were entered into a given project roughness moclel, the time variable could then be solved for, as demonstrated below for a typical project. Project: Arizona I-10 ED, LaPaz Co. (AC) Model: 51 ~ 12~ '0.96 + 0 22`ty~ 54 + 0.0 (t) T ife Estimate: 175 = 0.51 + 1.12~50~°96 + 0.22~54 + o.o8~50~059(t)~ 79 t = 15.75 yrs (to reach a MRN~ of 175) The resultant value of time represented Me expected age, or service life, of the pavement when the trigger roughness was reached. These values were then plowed as a function of different initial smoothness values to obtain frugal smoo~ness-life relationships. Figure 40 summarizes the steps in this process. The top portion of figure 40 indicates the development of Me regression models for a specific project. Then, as shown in the middle of figure 40, We best-fit mode! was set equal to a uniform trigger value (consistent among smoothness indices) and the hme (life) variable was solved for, assuming different levels of initial smoothness. The final step in the process (shown in the boNom portion of figure 40) consists of constructing plots of predicted life versus initial smoothness for each pavement type widen each State. These plots were generated over the available initial smoothness data range only, and upper life limits of 50 years for asphalt pavements and 75 years for concrete pavements were selected corresponding to baselines or target, nutial smoothness levels of 5 in/ml (0.08 m/km) and 7 ~n/mi (0.11 m/km), respectively. This was to ensure the reasonableness of Me results, and models that did not meet these criteria were excluded. 90

city ~ ~ - m. ~ s~ >= ¢ ~ - =` ~ ;'~E ~ ~- ·oc o. z~ =~ ; A ~ e A, _ _ ¢,, O,v 5m ,,c Cal Cat Cat _ Cut Us Cot LO di Cot CO Cat Cat . ~ . L U) GO U) L . ~ Cat o o Lo _ Cal _ O. LO AL Cut ~ . o ~ . o o · o o · - o ~ En Lo Cat Us Cat o Do Go o Cat Go Cat Lr Cal o o 91 d1 or C~ oo oo o. LO o d°4 ~o c~ o. c~ . di d~ o c~ o. c~ - . - ~ ~ ~ a ~ ~ 1 1 r1 ~L ~ -1~1` 0,)4 m. -.~.1~. c~ ~ 1~ 1'~ o o. 10. 10. ~ ~ 1- 1- = 1~

Step 1: Determine Best Fit Model ;80 1 ,- f/. 1 60 1 ~, ~ ,' t /, -~ ~ z l°° ~- ~- ° --* r y so T ~ 40 -, °'~ -"' 2 0 ~_ 1 1 1 1 1 1 s 0 15 20 25 A ge (years) Step 2: Use Model to Project Pavement Life 1 8 0 1 6 0 - ._ ~ 140 3 ~, 120 ~D ~ 100 =~ 8( ~o 60 ~n - ,St :.c Section 1 D ata a Section 2 D ata ~Section 3 D ata 0 Section 4 D ata Best-Fit M odel at Initial w 10 in/m i Best-Fit M odel at Initial = 20 in/m i Best-Fit M odel at Initial = 30 in/m i --- Best-Fit M odel at Initial ~ 50 in/m i 30 35 40 _ ~ _ _ _ _ Trigger Value ~ 175 in/m ~ ~ 1~-= v ~ ~ ~ ~ ~c Section 1 D ata 0 Section 2 D ata Section 3 D ata ~ Section 4 Data ----- Best-Fit M odel at Initial 1 0 in / m i Best-Fit M odel at Initial ~ 20 in/ml Best-Fit M odel at Initial ~ 30 in/m i ---Best-Fit M odel at Initial :~ 50 in/ml cazo ;~d lrears 29 Year ~-., 0 ~i~ ~ ~ ~ ~=' ~ 0 5 10 15 20 25 A ge (years) 30 35 40 Step 3: Plot Pavement Lives for Range of Initial Smoothness Values and Detemune Relationship 40 T 35 L 30 - 4, 25 ;^ 20 u ~, 15 0 5 10 15 20 25 30 35 In itia I M ay s R id e N u m b er (in /m i) 40 45 50 Figure 40. Conceptual illustration of roughness model procedure for determining smoothness-life relationships. 92

For the Kentucky and Wisconsin pavement failure analysis, the approach consisted of identifying construction projects at least 2-ml (3.2 km) long, sorting them into pavement groups according to general design (i.e., JPC, AC overlay on AC) information and traffic levels, and ranking the projects within each group by initial RI/PSI. Then, for each pavement group, the projects were divided into smoothness classes according to levels of Initial RI/PSI. For example, "initially rough" pavements might have been classified as those with RI/PSI values between 3.4 and 3.8, "initially moderate" as those with RI/PSI values between 3.9 and 4.2, and "~rutially smooth" as those with RI/PS] values greater than 4.3. When dividing the projects into classes according to initial RI/PS] level, efforts were made to "balance" the class sizes by the number of projects included In the analysis. Once the smoothness classes were defined, the effective age of each project-overlaid or not was determined. For overlaid projects, Me effective age was calculated by subtracting the construction year from the documented resurfacing year. In these cases, the effective age represented the service life. For non-overlaid projects, the effective age was calculated by subtracting the construction year from the last reported year (1994) in the data base. In these cases, We effective age represented the ~n-service age. Each smoothness class was next resorted in ascending order by effective age. Then, for each effective age, the percentage of projects overlaid was determined by dividing We cumulative number of projects overlaid through that age by the total number of projects in the smoothness class. Thus, for each class of initial RI/PSI, a series of failure data were generated and plotted, with "pavement life" on the abscissa (x-axis) and "percent sections overlaid" on the ordinate (y-axis), as shown in figure 41. Although the failure curves of some smoothness classes- essentially those win a low percentage of projects overlaid-were not well defined, most seemed to follow an exponential or power pattern, with the curve primarily flat the first ~ to 12 years and then becoming very steep the next few years. For each smoothness class, the service lives corresponding to 10, 25, and 50 percent projects overlaid were estimated using "best-fi~" regression curves (usually exponential or power form), as seen In figure 41. The mean ~rutial RI/PSI values for each smoothness class were then plotted against the corresponding service life estimates, resulting in a relationship between crucial smoothness and pavement life. Presentation of Results Roughness Mode] Analysis Using We procedure described previously, initial smoothness-life relationships were developed for families of pavements in each of We ten States furnishing data to the study. An obvious linear trend prevailed over the typical initial smoothness range. Therefore, a linear regression was fit through the predicted life versus ~rutial smootlu~ess data for each pavement type within each State in order to quantify the general predicted life versus initial smoothness trend. The use of a linear trenc! is 93

loo - 90 t 80 ._ ~70 - o 60 ~n - u ._ o ~40 co 50 30 20 10 0 1 ~ Initially Rough i - Initially Moderate Estimated j ,' Life = 14.5 years | ~ . . ~ 25% Projects Overlaid Estimated Life = 16.5 years l 1 W ~ A'' - , , _ . , ~ ~ f 6'^` ·,' / AA . , ~: Estimated Life = 20 years 5 10 15 Age, years 20 25 30 Figure 41. Conceptual illustration of pavement life estimation using regression failure curves. · Initially Smooth probably conservative for pavements that are constructed smoother Man those presented in the typical crucial smoothness range (i.e., pavements constructed smoother may have even greater lives Wan predicted from the linear model, due to the lack of dynamic loading effects). However, by me same token, me lives of pavements constructed rougher than the available initial smoothness data range may be overestimated by the linear trend if dynamic loadings were to lead to an increased rate of deterioration. Variations in the predicted life versus initial smoothness curves are due to many factors that vary between projects, such as traffic level, pavement thickness and base design, materials, climate, and quality of construction. It is important, however, to note that in Me vast majority of the projects, a general linear smoo~ness-life trend was observed over a reasonable initial smoothness range. The resulting smoothness-life trends for the various pavement projects included in the study are provided in figures D-l through D-35 of appendix D. The project trends have been grouped according to pavement family (i.e., State and pavement type) so that an overall linear trend representative of each family could be determined. An example plot of one such family is shown in figure 42 and illustrates the relationship developed between initial smoothness and projected pavement life. 94

60.00 , 50.00 40.00 c' 2 30.00 20.00 10.00 1 Regression Equation: y = 10.989x - 18.821 x .~ x x ~,: x x x 5 4.75 4.5 4.25 4 3.75 3.5 3.25 3 2.75 Initial Rldeabillty Index Figure 42. Example plot of pavement life versus initial smoothness using the roughness mode! approach. i> 1-64 W B Carter Co. x I-65 NB Hart Co. x 1~65 SB Hart Co. Table 22 summarizes the smoo~ness-life relationships developed for each pavement family. Many of these relationships were used In the cost-effectiveness evaluation of smoothness specifications presented in We last section of this chapter. Another way In which the results of the predicted life versus ~rutial smoothness plots can be interpreted is to plot the rate of increase (or decrease) in life versus the rate of increase (or decrease) In initial smoothness, based on a target Crucial smoothness level. This type of evaluation is commonly referred to as a sensitivity analysis and is useful in examirung the responsiveness of the dependent variable fin this case life) to the independent variable fin this case Crucial smoothness). ~ Me sensitivity analysis carried out here, the target initial smoothness levels for each pavement family were established using the approximate values given In table 21 corresponding to PI values of 7 in/ml (0.~! m/km) for concrete surfaces and 5 n/mi (0.08 m/km) for asphalt surfaces. All initial smoothness levels were expressed as a percentage increase or decrease in ~rutial smoothness with respect to the established target smoothness value. Similarly, the corresponding predicted lives were expressed as a percentage increase or decrease in life with respect to the predicted life at the target initial smoothness level. The predicted life at the target initial smoothness variable varies from project to project. 95

Table 22. Smoothness-life relationships developed using roughness model approach. Pavement Family Smoothness-Life Prediction Equation Life = -0.187BPRAL + 34.994 Life - -0.247BPRAL + 43.025 Life = -0.175MRNAz + 24.883 - _, Alabama PCC Alabama AC Arizona JPC Arizona AC Arizona AC/JPC Arizona AC/AC Georgia JPC Life = -0.229MRNAz + 44.374 Life = -0.098MRNAz + 14.385 Life = 0.138MRNAz + 26.883 Life = -0.454MRNGA + 30.175 Life = -0.260MR>JGA + 20.423 Georgia AC Georgia AC/JPC Georgia AC/AC Life = -0.413MRNGA + 30.351 Life = -0.557MRNGA + 25.714 Illinois CRC Illinois AC/CRC Illinois AC/JRC Kentucky PCC Interstate Kentucky AC Interstate Kentucky AC/PCC Interstate Kentucky AC/AC Interstate Kentucky AC/PCC Parkway Life = -0.180IRI~L + 32.009 - Life =-0.079IRI~ + 14.943 Life = -0.080IRIIL + 14.196 Life = 12.116RIK,, - 23.429 Life = 6.650RIK~,- 14.336 Life = 11.390RIKy~ 22.304 Life = 10~989RIKy- 18.821 Life = 13.583RIK,'- 37.608 Kentucky AC/AC Parkvvay Michigan JRC Michigan AC Michigan AC/PCC Minnesota PCC Minnesota AC Minnesota AC/AC South Dakota PCC Washington PCC Washington AC Washington AC/AC Wisconsin CRC - Wisconsin JPC (w/o dowels) Wisconsin JRC - Wisconsin AC Wisconsin AC on Rigid Base Wisconsin AC on Flexible Base Life = 4~647RIKy - 3.535 Life =-0.153IRIMI + 30.767 Life = -0.040IRIMI + 17.147 Life = -0.123IRIMI + 20.213 Life =-0.051IRIMN + 7.520 Life = -0.113RIMN + 18.060 Life = -0.198IRIMN + 31.891 Life = -1.849PIsD + 49.372 Life =-0.105PCAWA + 36.638 Life = -0.065PCAWA + 28.006 Life = ~.030PCAWA + 28.254 Life = 9.567PSIW, - 27.410 Life = 6.912PSIWl -19.675 Life = 7.220PSIW, - 23.869 Life = 7.896PSIW' -17.293 Life = 5.474PSIW~ -13.725 Life = 5.042PSIw~ -11.207 96

Sensitivity plots, such as the one illustrated In figure 43, were constructed for each project. The plots of similar projects were then grouped together according to pavement family, and an overall relationship for each family was generated using a linear regression of the available data. The complete set of sensitivity plots are provided in figures D-45 through D-79 of appenclLx D. A summary of the results of the sensitivity analysis is presented in table 23. This table lists, for each pavement family, the target initial smoothness level, the sensitivity ratio (i.e., the rate of percentage change in life divided by the percentage change in roughness), and the percentage increases in life corresponding to smoothness increases (from the target level) of 10, 25, and 50 percent. For example, for the first pavement family (Alabama PCC pavements), a 10 percent increase in smoothness (corresponding to a change in the BPR roughness measurement from the target of 100 in/ml [~.58 m/km] to a value of 90 ~n/mi [~.42 m/km]) leads to an 11 percent Increase in pavement life. Two items worth noting about table 23 are the sign and magnitude of the sensitivity ratio. In the cases where roughness is the operating parameter, such as with data from Alabama and Arizona, the negative sign on Me sensitivity ratio simply means Cat negative change in roughness (decreased roughness) results in a positive change In life (increased life). ~ the cases where serviceability is the operating parameter, such as with data from Kentucky and Wisconsin, the positive s - o -50 - 1 00.00 Rougher Smoother 80.00 · - 60.00 40.00 20.004 am _ , , . ^~ ~t i I .00 - 0.00 -30.00 -20.00 ~0. 0 10.00 20.00 30.00 40.00 50 ~r20.0O of ~40.00 : Regression Eq~adon: -60.00 - y · 1 .7174x - 1 E-14 too of t % Change In Initial Smoothness (bawd on tugot-~.1) + I-64 W B Carter Co. -I-65 NB Hart Co. 00 I-65 SB Hart Co. Figure 43. Example sensitivity plot showing percentage change in life versus percentage change in roughness. 97

Table 23. Summary of results of sensitivity analyses performed on various pavement families. Pavement Family Alabama PCC Alabama AC Arizona JPC Arizona AC Arizona AC/JPC Arizona ACiAC Georgia JPC Georgia AC Georgia AC/JPC Georgia ACiAC Illinois CRC Illinois AC/CRC Illinois AC/JRC Kentucky PCC Interstate Kentucky AC Interstate Kentucky AC/PCC Interstate Kentucky AC/AC Interstate Kentucky AC/PCC Parkway Kentucky AC/AC Parkway Michigan JRC Michigan AC - Michigan AC/PCC Minnesota PCC Minnesota AC Minnesota AC/AC South Dakota PCC Washington PCC Washington AC Washington ACiAC Wisconsin CRC Wisconsin JPC (w/o dowels) Wisconsin JRC Wisconsin AC Wisconsin AC on Rigid Base Wi~ce~n.~in ACT on Flexible Bane 1 in/ml = 0.0158 m/km 1 mi = 1.61 km Target Initial Smoothness Level BPRAL = 100 in/ml BPRAL = 80 in/ml MRNAZ = 60 in/ml MRNAZ = 50 in/ml MRNAZ = 50 in/ml MRNAZ = 50 in/ml MRNGA - 25 injmi MRNGA = 20 in/ml MRNGA = 20 in/ml MRNGA = 20 in/ml 8IlL = 55 in/ml IRIS = 50 in/ml IRIIL = 50 in/ml - RIKy = 4.0 RIKY = 4.1 RIKY = 4~1 RIKY = 4~1 RIKY = 4~1 RIM = 4.1 IRIM, = 55 in/ml IRIS = 50 in/ml IRIM' = 50 in/ml IRIMN = 55 in/ml IRIMN = 5.0 in/ml IRISH = 50 in/ml PIED = 7.0 in/ml PCAWA= 100 cts/mi PCAWA = 60 cts/mi PCAWA = 60 cts/mi PSIW~ = 4.0 PSIWl = 4.0 Sensitivity Ratio (%^Life/%^Roughness). Mean Percent Increase in Life Corresponding to Smoothness Increase of: Sensitivity ratios for pavements measured in terms of a roughness index are negative (-), whereas sensitivity ratios for pavements measured in terms of a serviceability index (inverse of roughness index) are positive (a). Percentage increase in life not given because percentage change in roughness/serviceability results in serviceability levels greater than 5. 98

sign on the sensitivity ratio means that positive change in serviceability Pancreases serviceability/decreasec! roughness) results In a positive change in life Pancreases life). As for the magnitude of the sensitivity ratios, since the roughness scales are immensely larger than the serviceability scales, the roughness ratio values are typically much smaller between O and ~ versus between ~ and 5 for serviceability. Although different roughness parameters are used, the vast majority of the pavement families show at least a 9 percent increase in life corresponding to a 25 percent Increase In smoothness. In terms of the profile index, the 9 percent increase in life would correspond to an approximate smoothness increase from 7 to 5 ~n/mi (0.~l to 0.08 m/km) for concrete and 5 to 3.5 Mimi (0.08 to 0.06 m/km) for asphalt. , , ~ ~ ~ · .q ~ · . ~ .~ . ~ .1 ~ ·1 ~ ~ . ~ . ~ _ . lading tne analysis a step nirtner, most or the lamllles snow at least a lo percent increase in life corresponding to a 50 percent increase in smoothness. Thus, an approximate increase In smoothness from 7 to 3.5 in/ml (0.11 to 0.06 m/km) for concrete and 5 to 2.5 in/ml (0.08 to 0.04 m/km) for asphalt could conceivably yield at least a 15 percent increase in life. As seen in table 23, the percentage increases in life corresponding to Me 10 and 25 percent increases in smoothness are much greater for Kentucky and Wisconsin pavement families than for the other families. This is because smoothness for these families is expressed In serviceability, and the serviceability scale is much tighter than the scales of Me other roughness indexes (e.g., MEN, my. The tightened scale causes the specified smoothness Increases to be magnified, so bat a 10 percent increase in serviceability is a much greater change in smoothness than a 10 percent change in MRN or TRI. If the serviceability values for the Kentucky and Wisconsin pavement families were to be converted to MRN or IR] prior to doing the sensitivity analysis, the resulting increases In life would be reduced to levels more comparable to the other pavement families. Pavement Failure Analysis The results of the Kentucky and Wisconsin pavement failure analysis are illustratec! In figures 44 through 52. These figures show Me failure curves of selected Rho and PUS ranges for seven Kentucky pavement groups and two Wisconsin pavement groups, respectively. Although some overlapping of the failure curves was observed for each pavement group, the general trend of fewer projects overlaid for Initially smoother pavements is apparent. For instance, In figure 45, the roughest class (3.4 < Ems ~ 3.~) showed a lower percentage of overlaid projects after 25 years than the next class (3.9 < l~,d' < 4.0~. However, Me remaining smoother classes showed sequentially lower percentages of overlaid projects after 25 years. With a few exceptions, much of Me overlapping among curves occurred over Me low ranges of pavement age or percent projects overlaid. At higher ranges, the curves become more consistent in showing Cat initially smoother pavements provide additional life. This trend is borne out for all pavement types-PCC, AC, AC/AC, and AC/PCC. ~ . , 99

100 90 80 70 c~ o cr. - c, o - 4J 60 50 40 30 20 10 O ~ 100 90 80 70 60 50 40 30 20 10 a t? r- ~ I ; I ....... ...... r~ x~ I.............. 0 5 10 15 20 25 30 35 Age, years 3.4 <= Rl(init) <= 3.8 (25 Projects, Avg Init RI=3.70) ~3.9 <= RI(init) <= 4.0 (24 Projects, Avg Init RI=3.94) ... 4.1 <= RI(init) <= 4.2 (20 Projects, Avg Init Rl=4.16) - 4.2 <= RI(irut) <= 4.3 (20 Projects, Avg Init RI=4.36) 0104.5 <= RI(lrut) <= 4.8 (4 Projects, Avg Alit RI=4.6) Figure 44. Pavement failure curves for PCC interstate pavements in Kentucky (all traffic levels). o . - ~ ...... ... ^1~/ .... ~ _ ~ID Ale-ADMIT , 5 10 15 20 25 30 35 Age, years · 3.4 <a Rl(init) <a 3.8 (24 Projects, Avg Init RIs3~65) 3.9 <a RI(init) <3 4.0 (46 Project, Avg Init Rl=3.95) 4.1 <- Rl(init) <- 4.2 (27 Projects, Avg Init RI~4.15) ~4.3 <= RI(init) <a 4.4 (17 Projects, Avg Init RI-4.34) ~4.5 <= RI(init) c= 4.8 (11 Projects, Avg Init RI,459) Figure 45. Pavement failure curves for PCC parkway pavements in Kentucky (all traffic levels). 100

100 ~ . 90 80 ns 70 - c, o u, - 4, o - c~ 4J 60 50 40 30 20 n o 100 90 80 70 o o' o - X 60 50 40 30 20 10 O t// : I ~...~.R,....~.! 1 2')) I ..~y I .. I/."2'/ . ....~.... . ~ ~ y ---- ~ T;.2,{ I 1~ .:: :: -::: ~ ~ ::. .::l ~' 10 Age' years 3.5 <= RI(init) <= 3.6 (4 Projects, Avg Init RI=355) ~RI(init) = 3.7 (9 Projects, Avg Init RI=3.7) :i: Rl(init) = 3.8 (10 Projects, Avg Init RI=3.8) - -~ - RI(init) = 3.9 (13 Projects, Avg Init Rl=3.9) -- 31t~ RI(init) = 4.0 (6 Projects, Avg Init Rl=4.0) 15 20 Figure 46. Pavement failure curves for AC interstate pavements in Kentucky (all tTaffic levels). :-i; ~ *.......... _ ~ ~ . ~: . ~ ~. ~1 ~ , ;. . ·: ~ I ........ ...... .. f~'' .;....... .;......... _~. ~1 1 o ·-~ 3.5 <= Rl(init) <= 3.6 (20 Projects, Avg Init R1~57) ~3.7 <= RItinit) <= 3.8 (25 Projects, Avg Init Rl=3.77) .. 3.9 <= RI(init) <= 4.0 (31 Projects, Avg Init RI~3.93) ~4.1 <= RI(init) <= 4.4 (17 Projects, Avg Init RI=4.19) . 1 5 10 15 20 25 30 Age, years Figure 47. Pavement failure curves for AC parkway pavements in Kentucky (all traffic levels). 101

100 90 80 70 60 50 40 30 20 10 o 100 90 80 70 - 0 60 os - o 40 50 30 20 10 O . ~_w ~ _ ~_ I ~ 0 2 4 6 8 10 12 14 16 Age, years 3.7 c= RI(init) <= 3.8 (15 Projects, Avg Init Rl=3.77) RI(iriit) = 3.9 (30 Projects, Avg Init Rl=3.9) : :: RI(init) = 4.0 (17 Projects, Avg Init RI=4.0) - x RI(init) = 4.1 (14 Projects, Avg Init Rl=4.1) RI(init) = 4.2 (8 Projects, Avg Init RI=4.2) Figure 48. Pavement failure curves for AC/AC interstate pavements in Kentucky (all traffic levels). en ~ .·~:. ~ ~ ~ _ _ _ ....... / ~ ................... t+# ~.. ' ~ $-.:. 1 0 5 10 15 20 Age, years 3.2 Cal RI(iNt) C= 35 (12 Projects, Avg Init Rls3.43) ~3.6 <= Rl(iIlit) <5 3.7 (44 Projecm, Avg Init Rls3.66) : 3 8 c= RI(illit) <= 3.9 (33 Projects, Avg Init Rl~.84) ~4.0 Cal RI(init) c= 4.2 (19 Projecm, Avg IIIit Rl=4.04) Figure 49. Pavement failure curves for AC/AC parkway pavements in Kentucky (all traffic levels). 102

100 90 80 70 60 50 40 30 20 10 o 100 90 80 - He - - o ~ 40 4, 70 60 50 30 20 10 o .' · ::. ~ 1 1 1 1 0 2 4 6 8 10 12 14 16 Age, years 3.2 <= RI(init) c= 35 (10 Projects, Avg Init RI~.35) ~3.6 <= RI(init) <= 3.8 (21 Project, Avg Init RI=3.76) :...: 3.9 <= RI(init) <= 4.0 (50 Projects, Avg Init RI=3.95) ~4.1 <= RI(init) <= 4.2 (17 Projects, Avg Init Rl=4.11) Figure 50. Pavement failure curves for AC/PCC interstate pavements in Kentucky (all traffic levels). 0 5 10 15 Age, years 20 3:7 < PSI(init) <= 4.0 (15 Projects, Avg PSI(init)~3.86) ~4.0 < PSI(init) <- 4.2 (14 Projects, Avg PSI(init)=4.11) ..~ 4.2 < PSI(init) <= 4.4 (23 Projects, Avg PSI(init)=431) 4.4 c PSI(init) <- 4.6 (18 Projects, Avg PSI(init)=4.48) 31t 4 6 < PSI(init) c= 5.0 (14 Projects, Avg PSI(init 2) Figure 51. Pavement failure curves for AC on flexible base pavements in Wisconsin (all traffic levels, interstate and primary highways, post-1979 construction). 103

loo 9o - 80 70 o in - ~ 50 o - u 60 40 30 - . 20 - 10 O- . >,.~~ .~. 0 5 10 15 20 Age, years · 3.3 < PSI(init) <= 3.9 (13 Project, Avg PSI(init)~3.71) ~3.9 ~ PSI(init) <= 4.2 (22 Project, Avg PSI(init)=4.06) . 4.2 < PSI(init) <= 45 (53 Project, Avg PSI(iriit)=4.32) 4 5 ~ PSI(init) <= 4.8 (32 Projecm, Avg PSI(init~.63) ~4.8 < PSI(init) c= 5.0 ill Projects, Avg PSI(init)=4.94) Figure 52. Pavement failure curves for AC on rind base pavements in Wisconsin (all traffic levels, interstate and primary highways, post-1979 construction). Mere are a few items worth noting about Me failure curves. First, although Kentucky does have AC overlays of PCC pavements (AC/PCC) on their parkway system, failure curves for this pavement group were not constructed due to the very small percentage (less than 2 percent overall) of these pavements Cat have been overlaid to date. Secondly, the failure curves for Me Kentucky AC interstate group were notedly inconsistent, with the smoothest class (R~ = 4.0) showing the greatest failure rate anct the second roughest class (Rim = 3.7) showing Me lowest failure rate. One possible explanation for this phenomenon is that the number of available projects for this group is much lower than for Me over groups. A total of 43 projects were used in the analysis of this pavement group versus 80 or more for Me other groups. Another possible explanation may be heavy traffic volumes. nor example, as seen In figure 46, all of the failure curves are very steep, with the vast majority of Me projects having been overlaid between 9 and 15 years. V 1 . Lastly, as seen in figures 51 and 52, Me Wisconsin pavement failure curves are not nearly as developed as the Kentucky failure curves. This is because they represent oniv Dost-1979 construction. In fact failure curves for CRC and TPC pavements could J 1 not be constructed due to Me lack of overlaid, post-1979 concrete projects. ~1 For each of Me seven Kentucky pavement groups and two Wisconsin pavement groups, graphs of initial serviceability versus life were constructed by taking the 104

estimated pavement lives corresponding to 10, 25, and 50 percent of projects overlaid for each smoothness class and plotting against the mean initial serviceability of that class. Figure 53 illustrates one such graph, representing Kentucky PCC parkway pavements. The complete set of serviceability-life plots is provided in figures D-36 through D- 44 in appendix D. The data points within each of these graphs and for each failure criteria (10, 25, and 50 percent projects overlaid) were fitted with a linear regression curve. Of the 27 total linear regression equations, only six did not show a good fit (i.e., one in which the R2 value is greater than 0.75~. These six included the three Kentucky AC interstate trends-all of which were close to zero-and three individual Kentucky group trends corresponding to 10 percent projects overlaid. Although the slopes (change in life/change in serviceability of the linear regression curves vary by pavement group, the plots largely confirm a direct positive relationship between pavement life and initial serviceability for the serviceability ranges defined by Me failure curves. A summary of the linear regression equations for each pavement group, corresponding to 25 percent projects overlaid, is provided in table 24. 40 ;~` 35 u ._ c Is ;> o lo a) u, o J C E 30 25 20 15 10 s O y = 6.518x + 4.5917 R2 = 0.7999 y = 7.9282x - 5.2905 R2 = 0.7517 y= 8.0631x- 9.0991 R2 = 0.6174 3 3.S 4 4.5 1iIean RI(init) 50% 25% 10% Figure 53. Graph of pavement life versus mean initial serviceability (RI) for Kentucky PCC parkway pavements. 105

Table 24. Summary of serviceability-life regression equations for Kentucky and Wisconsin pavement groups (for 25 percent projects overlaid). l Serviceability-Life Equation Pavement Group ~(based on 25 percent projects overlaid) ~R Kentucky PCC Life = 8.178RI~, t - 2.463 0.91 Interstates Kentucky PCC Life - 7.928RI,~t - 5.291 0.75 Parkways Kentucky AC Interstates Life = 0.840Rlini~ - 6.616 0.01 . Kentucky AC Parkways Life = 16.270RIir,~ - 49.384 0.89 Kentucky AC/PCC Life = Il.133RI~,i~ - 32.721 0.90 Interstates Kentucky AC/AC Life - 4~862RTini~ - 8.317 0.75 Interstates Kentucky AC/AC Life = 15.26IRlir~,' - 43.615 0.95 Parkways Wisconsin AC on Rigid Life = 3.263PSI~,,' + 0.236 0.92 Base Wisconsin AC on Life = 6.516PST,,,~ -12.744 0.92 Flexible Base . , Figures D-SO through D-~8 in appendix D show the percentage changes in life corresponding to various percentage changes in serviceability for each pavement group. Similar to the roughness model analysis, target initial serviceability levels were selected, In this case an RI of 4.0 for concrete and 4.! for asphalt. Hence, a 10 percent increase In serviceability for concrete equaled an increase in serviceability from 4.0 to 4.4 (0.10 x 4.0 + 4.0 = 4.41. Interestingly, the pavement life corresponding to the target Initial smoothness level and 25 percent projects overlaid varied by pavement group. This gives an indication of how service life is also dependent on other factors, such as pavement type and functional classification (traffic). Table 25 summarizes, for each pavement group, Me sensitivity ratios and the percentage increases In life corresponding to serviceability increases of 10, 20, and 25 percent. A comparison of the sensitivity ratios for seven pavement groups, as determined by bow the roughness model (table 23) and pavement failure (table 25) procedures, shows good or fair agreement for three pavement groups (Kentucky PCC interstate, AC/AC interstate, and Wisconsin AC on flexible base), and poor agreement for the other three groups (Kentucky AC interstate, AC/PCC interstate, 106

Table 25. Summary of results of sensitivity analyses performed on Kentucky and Wisconsin pavement groups. Kentucky PCC Interstate Kentucky PCC Parkway Kentucky AC Interstate Kentucky AC Parkway Kentucky AC/PCC Interstate . Target Initial Pavement Serviceability Sensitivity Ratio G roup Level (%^Lif e/%^S erviceab ility) RIini! = 4~0 Refit = 4~1 RI'nit = 4~1 1.120 0-344 3.754 Reknit = 4.1 T 3.428 _ Mean Percent increase in Life Corresponding to Serviceability Increase of: ) 10% 1 20% 1 25% 34 26 75 . 68 Kentucky AC/AC Interstate Kentucky AC/AC Parkway Wisconsin AC on Rigid Base Wisconsin AC on Flexible Base RIinit = 4~1 RIinit = 4~1 PSIjnjt = 4.1 PSIinit = 4~1 1.744 3.477 0.982 1.950 17 10 2() 33 28 Percentage increase in life not given because percentage change in serviceability results in serviceability levels greater than 5. AC/AC parkway, and Wisconsin AC on rigid base). Moreover, win the exception of the Kentucky AC interstate group, all sensitivity ratios are approximately 1.0 or greater. In terms of the percentage increases in life, all pavement groups except Kentucky AC interstate show at least a 10 percent increase in life corresponding to a 10 percent increase In serviceability. Hence, a serviceability jump from 4.0 to 4.4 for concrete or 4.l to 4.5 for asphalt, often results In at least a 10 percent Increase In pavement life. Summary and Significance of Results The results of the evaluation strongly indicate that crucial pavement smoothness has a significant effect on pavement life. Using both roughness mode! and pavement failure analysis techniques on data from several States, it was clearly shown mat added pavement life can be obtained by achieving higher levels of nutial smoothness. The rate at which additional life is achieved is dependent upon, among other things, pavement type, facility type, and location. Moreover, although fixed rates of added life were used In the preceding discussion for purposes of conservative 107

estimation, some indications of increasing rates of added life likely due to the effects of dynamic loading were noted while conducting the analyses. Nevertheless, it stands to reason that as the initial roughness goes to zero (a perfectly smooth pavement), pavement life reaches a maximum value, for even the smoothest pavement ever built will have a finite life due to the fact that other factors are continually at work in the deterioration of the pavement. Sensitivity analyses, in which the percentage change in life as a function of percentage change in smoothness was determined, showed sizable increases In life for most pavement families, corresponding to nominal increases in smoothness. At least a 9 percent increase in life corresponding to a 25 percent increase In smoothness (from a target profile index of 7 in/ml [0.~! m/km] for concrete and 5 in/ml [0.08 m/km] for asphalt) was observed for He vast majority of the pavement families. How such increases affect the smoothness level to be specified must be determined using life-cycle costing techniques, with consideration given to the tradeoff costs of attaining adclitional smoothness. An evaluation of the most cost-effective smoothness levels for a few pavement families is presented in the last section of this chapter. Analysis of the Effect of Smoothness Specifications on Initial Smoothness Introduction l Highway agencies have increasingly been adopting smoothness specifications for both new pavement and overlay construction. According to the results of the questionnaire survey reported in chapter 2, 76 percent of the responding highway agencies now employ a smoothness specification for AC pavements and SS percent of the responding highway agencies employ a smoothness specification for PCC pavements. However, the specifications used among these agencies vary widely In terms of the smoothness-measur~ng equipment used, the target 1units specified, and the inclusion of incentive and disincentive provisions. Although in their crudest form smoothness specifications have been In existence for many years, lithe information is available regarding how successful these specifications have been in achieving increased crucial pavement smoothness. Based on the questionnaire survey responses, 60 percent of He highway agencies employing a smoothness specification for AC pavements believe that the specification has resulted in increased initial pavement smoothness, while 72 percent of the highway agencies employing a smoothness specification for PCC pavements believe that the specification has resulted in increased ~rutial pavement smoothness. This section investigates the effectiveness of these smoothness specifications in providing a smooth pavement surface. Overview of Data Sources Data for evaluating the effect of smoothness specifications on initial pavement smoothness are available from four SHAs: Iowa, Georgia, Lois, and Wisconsin. In 108

addition, some pavement smoothness information is available from the literature for Kansas. Iowa The Iowa DOT provided comprehensive initial smoothness data for new AC and PCC construction and AC resurfacing projects completed between 1982 and 1993. The data consisted of mean overall smoothness measurements for projects in lowa's primary and secondary highway system. Three different testing devices were used by Iowa during the 12 years of study: a BPR Roughometer, a California profilograph, and a Soup Dakota profiler. The BPR Roughometer was primarily used on secondary highways, with test results available for 1982 through 1991. The California profiIograph was primarily used on primary highways, with test results available for all 12 years. The South Dakota profiler data were only available for 1991 through 1993. Georgia The Georgia data Included mean initial smoothness values for AC, PCC, and AC overlaid pavements constructed in Georgia between 1968 and 1995. The pavements built and tested between 1968 and 1980 were tester} with a PCA roadmeter, In which smoothness was expressed In counts/mi (counts/km). In 1981, Georgia switched from the PCA roadmeter to the Mays Meter for pavement smoothness testing of both pavement types. Although the Mays Meter continues to be used to test asphalt pavements, formal testing of concrete pavements was done win the Mays Meter until 1993, at which time a Rainhart profiIograph specification was instituted. However, several concrete projects built from 1993 on have been supplementally tested by the clepar~nent using the Mays Meter. Only for the 1981 to 1995 time frame were initial smoothness values available on a project-by-project basis; the 1970s PCA roadmeter data consisted solely of the average statewide Initial smoothness each year. Illinois As discussed in chapter 3, the Illinois DOT provided historical files containing roughness measurements taken of newly constructed highway pavements built throughout Illinois between 1959 and 1990. The measurements were taken with the BPR Roughometer and, in nearly every case, were made within ~ year of construction. Since the only specification activity Illinois experienced during this period was the implementation of their concrete smoothness specification (California profiIograph) in 1977, initial smoothness data (BPR Roughometer) on interstate CRC construction for this and surrounding years (1975 Trough 1982) were extracted from the files and used for analysis. Mean overall smoothness values for each project were computed from the individual measurements recorded at the time of testing for each direction 109 . . .

and lane combination. To more fairly assess the impact of the specification, only those projects greater than ~ mi (~.6 km) In length were Included In the analysis. The number of interstate CRC projects completed and tested each year during this S-year period varied substantially. Prior to the implementation of the specification (1975 and 1976), a total of 25 construction projects occurred. In the six years following the specification's enactment (1977 through 1982), only 19 projects took place, most of them In 1977 and 1978. Wisconsin As described In chapter 3r the Wisconsin DOT provided a large sampling of their pavement management data base for evaluation under this study. This sampling included annual and biennial serviceability data (1980 through 1995) on all interstate and U.S. highways in Wisconsin, as well as major State highways. The serviceability data were derived from roughness measurements made win a Mays Meter up through 1989 and with a Soup Dakota profiler starting in 1990. For the most part, new construction and resurfacing Projects were tested the Year they were completed and opened to traffic. -= r--~- - -r Wiscons~n's specification history includes the implementation of a concrete smoothness specification in 1984 and an asphalt smoothness specification in 1993. Both specifications were generally reported as being applicable to projects on major highways. Evaluation Approach The first step in Me analysis is to plot the mean, standard deviation, and coefficient of variation of the ~rutial smoothness over time for each State. For the data sources that include ~rutial smoothness measurements bow before and after the Implementation of the smoothness specification, the mean initial smoothness values before the implementation of the specification are compared with those after the unplementation of the specification. Furthermore, a t-test was conducted to examine the effect of Me specification implementation on the mean value of the crucial smoothness values. Finally, graphical plots of pre- and post-specification smoothness distributions were generated to illustrate the effect of smoothness specifications on the resulting pavement smoothness. The results of the analysis are presented In this section for each State. In all cases, the implementation of the smoothness specification showed a significant positive effect on the resulting Initial pavement smoothness. ~0

Presentation of Results By State Towa Iowa implemented a smoothness specification for their AC and PCC pavements In 1981 based on the California profiIograph. That original specification specified a maximum roughness of 15 in/ml (0.24 m/km) and contained a disincentive clause. Over the next 12 years, Iowa continued refining their specification, making the following changes: · November 1985-Added incentive clause. · December 1987 Maximum roughness reduced to 12 Mimi (0.19 m/km). · November 199~Maximum roughness reduced to 7 in/ml (0.01 m/km). An illustration of the mean roughness values corresponding to the time period that Iowa was developing their specification is shown in figure 54 for different pavement types and functional classifications. Although these plots contain a few peaks, the overall trend for all pavement types is a general reduction in Me average initial roughness value. Table 26 summarizes the reductions for each pavement type. For PCC pavements, table 26 shows that the mean roughness values were reduced from about il.5 in/ml (0.~S m/km) in 1982 to about 7 in/ml (0.01 m/km) in 1993 for primary roadways (a 39 percent reduction), and from about 15 in/ml (0.24 m/km) In 1982 to about 6 ~n/mi (0.10 m/km) in 1993 for secondary roadways (a 60 percent 16.00 14.00 it 12.00 tic 10.00 ~ 8.00 .g 6.00 a .= 3 4.oo 2.00 0.00 |12/81 Original Specification- Corrective | Action and Disincentive C lapse. 11/93 Revised S - ification , Maximum Roughness for 100% Pay Reduced from 12 intmi to 7 in/mi. :~ 12/87 Revlsed Speciflcatlon - Ma~mu~n Roughness for 1009 Pay |11/85 Revised Specification - | Reduced from Added Incentive Clause. l 15 in/ml to 12 h/mi. ~~.~~ AL - _ A ,, ~ `, ~ ', . by.. `0 ~ ~ ~ ~_ . ~ 1982 1984 1986 1988 1990 1992 1994 Year Figure 54. Mean smoothness values over time for Iowa pavements. 111 · New PCC- Primary - ~ - New PCC- Secondary · New AC - Primary -~-New AC- Secondary f "AC Overlay - Primary ~ ~ ~ AC Ovalay - Secondary r

Table 26. Summary of reductions In Iowa initial pavement roughness values. ~| 1982 Mean Project | 1993 Mean Project |PercentChangein|| 1! Pavement Type ! Roughness, in/ml ! Roughness, in/ml | Roughness PCC (Primary) 11.5 7.0 39 PCC (Secondary) 15.0 6.0 60 AC (Primary) 9.5 1.0 (1992) AC(Secondary) T 6.5(1984) T ~ 5(1987) ~-31 11 : AC Overlay (Primary) | 6.5 | 3.0 | 54 11 AC OVCIIAV (bccawlao) 7.0 (1985) 4.0 43 . 1 ~n/m~ = 0.0158 m/km reduction). Even greater reductions are observed for primary AC pavements, as the roughness level was reduced from 9.5 ~n/mi (0.15 m/km) to about ~ in/ml (0.02 m/km), an 89 percent reduction. Unfortunately, few data were available for secondary AC pavements, which over a short 3-year period (and for very few projects) actually showed a slight increase In initial roughness. More data were available for AC overlays, and figure 54 and table 26 show a significant decrease In the ~rutial roughness of AC overlays on both the primary (54 percent reduction) and secondary (43 percent reduction) systems. Figure 55 illustrates the standard deviation of the initial roughness values for each pavement type. The trends in this figure are not too clear as there are several spikes in the data and many of the pavements (particularly PCC pavements) show standard deviation values Mat cycle up and down. An examination of the data available for one of Hose cycles (1991) shows that several projects were constructed with extremely low ~rutial roughness values (on the order of 4 in/ml [0.06 m/km]) while one project less than 0.5-mi (O.~-km) long registered an iriitial value of 21 in/ml (0.33 m/km). The validity of including such a short construction section in this analysis is questionable, but the overall effect of including that value is an increase ~ me standard deviation of Hose measurements. However, the new prunary AC pavements and primary AC overlay pavements do show a definite trend in a reduction of the standard deviation of the initial roughness level. An examination of the testing values for He AC pavements shows a smaller range, with pavements commonly registering with an initial smoothness in He range of ~ to 4 in/ml [0.02 to 0.06 m/km. Using He standard deviation values and He mean roughness values, the coefficient of variation (COY) was computed for each year; this information is plotted in figure 56. Similar to the standard deviation plot, this figure shows that the COV for PCC pavements is quite variable and appears to increase in the later years. The COVs for the AC and AC overlay pavements are also quite variable.

9. a 8.00 o ~ .ij .00 ~ ~ 6.00 0 'a ~ '5 tib 5 00 4} 0 _ o 4.00 ~ 3.00 u, 2.00 1.00 0.00 10.00 ~ 12 j81 Ong~nal Specificabon Correc~ve Achon and Disincen~ve [11J85 Revised Specification- | Added Incentive Cause. I . ' 12/B7 Revised Specification Max~mum Roughness for 100% Pay Reduced from 15 ~n/~ru to 12 ~n,/~ru. ^'s _,' "it 'N'. ~As_ ~N _ / By_ ~ . . 11/93 Revised Spe~ificabon - Ma)amum Roughness for 100% Pay Reduced Tom l2 in/ml to 7 in /ml. ,.W 1982 1988 Year 1990 1992 1994 Figure 55. Standard deviation values of nutial pavement roughness for Towa pavements. 1 90 1.00 P: ~ 33 ~ on O o P4 ~ 0 0~ - 4, 0 ~ , ·5 Hi 0.40 0.20 [11/85 Revlsed Slxdllcadon - | |^dd~ ~" ~9~ l 0.00- 1 - 1- 1 1 1 1982 1984 1986 1988 1990 1992 1994 . 11/85 Revlsed Slxdllcadon - Added Incendve C bUSQ [12/81 Original Specked bon - Co~ve 1 11/93 Revlsed Spectflcadon M&,dmum Roughness for 100% Pay Action and Disincentive aa..cP Reduced from 12 in/ml to 7 in/ml. 12/87 Revised Spec~cadon Maximum Roughness for 100% P9y / ~ Reduced bom / \ Li;~_~: AN/ ~ ,,' ,' "of\ : .' ~ ~ ~\_: ~i' ~; ~ ~ New PCC - Primary l - - ~ - New PCC- Secondary - ~ New AC - Primary SAC Overlay - Primary AC Overlay - Secondary -· New PCC - Primary - ~ New PCC- Secondaly · New AC - Primary SAC Overlay - Primary -~-AC Overlay - Secondary Year Figure 56. Coefficient of variation values of Crucial pavement smoothness for Towa pavements. ~3

In summary, data from Iowa show a significant reduction in the ~rutial pavement roughness since the specification was first implemented in 1980. Georgia Prior to 1972, the Georgia DOT used a rolling straight edge for measuring and controlling nutial pavement roughness (Gulden et al. 1983~. In 1968, the State began using the PCA roadmeter on a limited basis to check the roughness of various interstate paving projects. In 1972, the decision was made to begin using the PCA roadmeter for acceptance (but not rejection) of the smoothness of all new paving projects. By 1979, the PCA ride meter was used exclusively within the State for construction control, with 1units of 300 counts/mi (483 counts/km) for AC and AC overlay interstate pavement construction (dens~graded mixtures) and a limit of 500 counts/mi (805 counts/km) for PCC pavements. Because of the difficulty of maintaining a standardized road-meter vehicle over a period of years, and because Me PCA ride meter could not distinguish localized areas of roughness, the use of other roughness measuring devices were investigated (Gulden et al. 1983~. In 1979, a trailer-mounted Mays Meter was purchased and by January I, 1981, Me Mays Meter system was used exclusively for controlling Axial pavement smoothness. The Mays Meter specification limits implemented in 1981 were: . AC pavements (dense-graded mixtures): -New Interstate construction: 35 ~n/mi (0.55 m/km) Overlays: 45 in/ml (0.71 m/km) AC pavements (open-graded mixtures): New interstate construction: 30 in/ml (0.47 m/km) Overlays: 35 in/ml (0.55 m/km) PCC mainline pavements: 75 in/ml (1.18 m/km) The specification for PCC pavements was lowered from 75 ~n/mi (1.~S m/km) to 65 ~n/mi (1.03 m/km) in 1983. In 1986, Me limits for AC pavements were changed to the following: · AC pavements (dense-graded mixtures): -New interstate construction and overlays: 30 in/ml (0.47 m/km) AC pavements (open-gracled mixtures): -- New interstate construction and overlays: 25 ~n/mi (0.39 m/km) In 1993, Georgia adopted a Rainhart (0.1 in [3 nun] blanking band) profiIograph- based specification for PCC pavements (limit of 7 in/ml [0.1 m/km]), while maintaining the Mays Meter specification for AC pavements. However, Mays Meter testing was still conducted on Me PCC pavements. ~4

The mean initial roughness values (PCA Roadmeter) for pavements constructed In Georgia during the late 1960s and 1970s are plotted in figure 57. For all pavement types, a general reduction in initial pavement roughness is observed over that period. Figure 58 shows the mean roughness trends for pavements constructed during the 198Os and early 199Os under the Mays Meter specification. The mean values for AC and AC overlay pavements remained fairly stable over time, with the most recent values registering less than 30 in/ml (0.47 m/km). The PCC pavements show a significant reduction in mean initial roughness from 1981 to 1995. Table 27 summarizes the reductions in mean initial roughness values by pavement type. Figure 59 provides a summary of the standard deviation of the initial smoothness data by year for the period of 1980 to 1995. The standard deviation for AC pavements has gradually decreased over the years, with the standard deviation for the AC overlay pavements remaining relatively stable. However, the standard deviation of the PCC pavements increased over this tone period, particularly in 1988 and 1994. An examination of the data indicates that five projects were constructed in 198S, with initial roughness values ranging from 33 to 59 in/ml (0.52 to 0.93 m/km). In 1994, six projects were available, with initial roughness values ranging from 22 to 72 ~n/mi (0.35 to i.14 m/krn). Again, Me effects of these are increases In the standard deviation for those particular years. 700 ~ 600 p~ ._ _ ~ Be 500 ~54OO 0 - _ ~ 300 ·~m ~ ~ ~ 200 o ~ 100 0 ~1968 - PCA _ ~ Roadmeter used ~ on limited basis 1979 - PCA Roadmeter used for construction control AC and AC Overlays-300 counts/mi PCC-500 counts/mi `~' ~-~% 1972 - PCA Roadmeter used for acceptance AC~400 counts/mi PCC-500 counts/mi ` - ,, ~ _ 1979 - First Mays Meter acquired 968 1970 1972 1974 1976 1978 1980 Year New AC ~ · - New PCC AC overlay Figure 57. Mean smoothness values over time for Georgia pavements (1968-1980~. ~5

100.0 cc v, ~ ._ - ~ 3 0 ~ 60.0 _ ~ ._ ~ C ;-, 40.0 _ ~ ~ _ 80.0 20.0 ~ 0.0 1981 - Mays Meter used for construction control AC: 35 in/ml AC Overlays: 45 in/ml PCC Mainline 75 in/ml 1983 - PCC Mainline: 65 in/ml t1986 - AC and AC Overlays: 30 iamb | -` rat ,~ _r J)000~,~_ ~ 1993 - PCC Mainline: 7 in/ml Rainhart profilo~raph (0.1-in BB) , I , . ~ 1980 1982 1984 1986 1988 1990 1992 1994 1996 Year New AC ~ ~ - New PCC AC overlay Figure 58. Mean smoothness values over time for Georgia pavements (1981-1995~. Table 27. Summary of reductions in Georgia mean project crucial roughness values over time. l | 1981 Smoothness, | 1995 Smoothness, | Percent Change j Pavement Type I in/ml I in/ml I in Smoothness I PCC 1 65 1 33 1 4 AC Overlay 1 35 1 25 1 29 AC 1 27 ~24 1 ~n/mi = 0.0158 m/km 116

25.0 ._ ._.~0.0 i_ 0 ~ ~.0 ._ ~ ;^ ~ \0.0 ~s v' us ~ 5.0 o 0.0 1981 - Mays Meter used for construction control AC: 35 in/ml AC Overlays: 45 in/ml PCC Mainline: 75 in/ml ~1983 - PCC Mainline: 65 in/ml ,1993 - PCC Mainline: 7 in/ml Rainhart profilograph (0.1-in BB) J ~ |1986 - AC and AC Overlays.30 in/m: ~ ~ 'I ~¢ ~ _ 2 `d , . . . . 'a ~ Ha_ 1980 1982 1984 1986 1988 1990 1992 1994 1996 Year Figure 59. Standard deviation values of Crucial pavement smoothness for Georgia pavements. · New AC - ~ - New PCC AC overlay The COV for the Georgia pavements is plotted in figure 60. An initial observation of these data is that the COV values are quite low (typically between 10 and 30 percent). The AC overlay pavements display a fairly stable COV over the 10-year time period, and the AC pavements show a slightly decreasing COV. The PCC pavements generally show the smallest COV, with Me COV gradually increasing over time, again believed to be due to the construction of a few exceptionally smooth pavements. Georgia started using the PCA roadmeter for construction control in 1979. However, for We ~rutial smoothness measurements of the projects constructed In 1980 in the State's data base, some AC overlay projects are covered by the specification whereas others are not. This provides an opportunity to evaluate the effect of the smoothness specification on the initial smoothness values win other factors such as time of construction, paving technology, and smoothness-measuring equipment being similar. The two sample t-test was conducted to test whether there is a significant difference in the mean values of these two groups of initial smoothness data. Table 28 presents the results of the l-test. The results show a highly significant decrease of the overall mean of the group covered by the specification versus the group not covered by Me specification at a 90 percent confidence level (from 353 counts/mi to 250 counts/mi). The standard deviation of the projects covered by the 117

0~80 - ._ ._ _ 0 ~ ~ 2 lo ~ '- -1 ~ ce 0 ~ :~ 0 o 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 1981 - Mays Meter used for construction control AC: 35 in/ml AC Overlays: 45 in/ml PCC Mainline: 75 intmi 1993 - PCC Mainline: 7 ~nJmi Rainhart profilograph (0.1-in BB) 11983 - PCC Mainline: 65 in/ml - J ~ ~1986 - AC and AC Overlays: 30 in/ml I \\ / an" ~ a` ~ _ J 1980 1982 1984 1986 1988 1990 1992 1994 1996 Year · New AC · - New PCC AC overlay Figure 60. Coefficient of variation values of initial pavement smoothness for Georgia pavements. Table 28. Summary of t-test results for Georgia AC overlay projects. || Measurement | No. Of | vMaeluen | DSeavinadtiaOdn | t-test | Significant? | l Time | | counts/mi | counts/mi | P a I (p ue 0. ) I Not Under 1 25 1 53 1 115 1 1 I Specification l l l | <: 0.01 | Yes Under 1 128 1 50 1 83 1 1 Specification l l l l l l 1 mi = 1.61 hen 118

specification (83 counts/mi) is much less than that for the projects not covered by the specification (~15 counts/mi). A comparison of the two distributions is provided in figure 61. ZZinois The TIl~nois Department of Transportation Implemented a smoothness specification for its concrete pavements in 1977. The specification is based on the California profiIograph and calls for corrective action if Me Crucial pavement roughness exceeds 15 in/ml (0.24 m/km). A new specification was Implemented In 1993 containing incentive provisions for Initial AC pavement smoothness values of less than 0.5 in/ml (0.008 m/km) and for initial PCC pavement smoothness values of less than 4.25 m/ml (0.07 m/km). While using the California profiIograph for the smoothness specification implemented In 1977, Illinois also continued measuring nutial smoothness values with the BPR roughometer, which the Department had used in their roughness measuring program since 1957 (Chastain and Burke, 1962~. The BPR data was used In this evaluation of pavement smoothness over tune because of its availability. For sake of comparison, a PI of 17 ~n/mi (0.27 m/km) converts roughly to a BPR roughness index of 100 in/ml (~.58 m/km) for concrete pavements. JO - ~ O Not covered by specification 9 · Covered by specification 7 6 5 4 2 1 / / Covered bv specification Mean = 250 in/nu Standard Decagon = 83 in/ml Not covered bv specifics bon Mean = 353 in/ml 1 ~ . ~; of ~1~- -1 ~ 1 1 1 1 50 100 150 200 250 300 350 400 450 500 550 600 650 Initial Roughness Measurement, PCA counts/mi Figure 61. Comparison of initial pavement smoothness distributions before and after implementation of smoothness specification (Georgia AC overlay projects). ~9

Figure 62 plots the mean initial roughness and corresponding standard deviation of We continuously reinforced concrete pavements (CRCP) constructed on the Illinois interstate system between 1975 and 1982. This figure shows a general downward trend In the initial pavement roughness, with a reduction of about 23 percent observed over that time period. The first year of the specification (1977) shows a definite reduction in initial roughness, although it increases slightly the next 2 years before decreasing again. In addition, the standard deviation of the initial pavement roughness is observed to slightly decrease over tune. Figure 63 illustrates the COV of the initial pavement roughness for the Illinois CRCP pavements from 1975 to 1982. This value is observed to be relatively stable over time, ranging only from about 14 to 28 percent. The COV is observed to drop off moderately after the implementation of the specification (1977) and then levels off over the next few years. Similar to the analysis of the Georgia data, the mean initial roughness of the projects constructed before the smoothness specification and the mean initial roughness of the projects constructed after the specification are compared using a two sample l-test. The results of the t-test are summarized in table 29. As shown in that table, for about the same number of projects, the mean initial roughness of the 90 -, an ._ 70 ,,, ~ 60 in ~ _ ~ ~ 50 ho ~ g ~ 40 ,' to 30 ~ 0 3 ~ 20 In \. 975 1976 1977 1978 Year 1979 1980 1981 1982 +Mean --it Standard Deviation Figure 62. Mean and standard deviation of ~rutial pavement smoothness for Illinois CRCP. 120

0.3 0.25 ~ . o ~ 0 2 ._ .m ° 0.15 ._ ._ ~ 0.1 o 0.05 o A / l ~1 1 1 1 1 1 ~ 1975 1976 1977 1978 1979 1980 1981 1982 Year 1 . cov Figure 63. Coefficient of variation values of initial pavement smoothness for Ill~nois CRCP. Table 29. Summary of t-test results for Ill~nois CRCP projects. Measurement I No. Of | Mean | Standard | l-test I Significant? Time | Sections I Raughness, I DeV~at~on' | p-value | (p-value<°.l) 1 . , , , , ; Not Under 24 81 21 ISpecificahon I l l l l l 1 .1 1 1 1 ~ ).08 1 Yes Under I 19 1 71 1 12 l l I Speciflcahon , 1 in/m = 0.0158 m/km 121

projects constructed after the implementation of the smoothness specification dropped to 71 in/ml (1.1 m/km) from a value of 81 in/ml (1.3 m/km) for projects constructed before the implementation of the specification. The standard deviation of the roughness index after the specification decreased from 21 ~n/mi (0.3 m/km) before the specification to 12 ~n/mi (0.2 m/km) after the implementation of the specification. A two sample t-test gives a p-value of 0.08, which is significant at the 90 percent level. The distribution of the initial roughness of the Blinois pavements both before and after the implementation of the specification is provided in figure 64. Wisconsin The Wisconsin DOT has kept records of project-level ~rutial PST measurements for all new pavement construction for most years from 1978 to 1994. Smoothness specifications for PCC and AC pavements were unplemented In 1984 and 1993, respectively. Plots of the mean initial PS] measurements for each year along with the corresponding standard deviation are shown in figures 65 through 68 for AC, jointed plain, jointed reinforced, and continuously reinforced pavements, respectively. An increase In the serviceability and a general reduction in the standard deviation of Me serviceability is generally observed from these plots. In, - , 9_ 8 7 u ~ 6 0 5 it 2 4 - 3 1 0 Covered bv specification Mean = 71 in/ml Standard Deviation = 12 in/ml \ Not covered by specification \ Mean = 81 in/ml ~ ~Standard Deviation = 21 in/nu ~1 \ ~ |CINot covered byspecificationl ~ ~ ~ ~ ~ ~ an/ | ~ Covered by specification | ~ I I I L\ 1 1 1~ `~` ~ ~ ~ IN n bin : ~ -~ J 20 30 40 50 60 70 80 90 100 110 120 130 Initial Roughness Measurement, BPR inJmi Figure 64. Comparison of ~rutial pavement smoothness distributions before and after implementation of smoothness specification Lois CRCP projects). 122

~ - ~4 ED ~3 . low . · ACE Mean PSI Mean PSI-std ~ Mean PSI+std 89 90 91 92 93 94 95 Initial smoothness measurement year Figure 65. Mean and standard deviation of initial PST over time for Wisconsin AC pavements. - ._ _, A_ ·_I o g ._ - cn a _. a, a 3 2 4 ~~ _ ~ _ _ ~· Arc ~0 - 0 - M&r ~. JPCP Mean PSI Mean PSI-std -Mean PSI+std 79 81 83 85 87 89 Initial smoothness measurement year Figure 66. Mean and standard deviation of initial PST over time for Wisconsin JPC pavements. 123

a] - - - ~ - - ~ a - - ~ 4 Is i; - 4} - . - - - . - ~ Mean PSI + StDev · J1lCP Mean PSI -Mean PST - StDev 79 81 83 85 87 89 Initial smoothness measurement year Figure 67. Mean and standard deviation of initial PST over time for Wisconsin JRC pavements. a, 55 - ._ - - ~_ o 45 g - ._ i} 3.5 3 25 4 2 _ ~ _ _e _ _ _ demean PSI + StDev · CRO? Mean PSI -Mean PSI - StDev ; ; ; . ~1 79 81 83 85 87 89 Initial smoothness measurement year Figure 68. Mean and standard deviation of Crucial PS] over time for Wisconsin CRC pavements. 124

A comparison of the mean initial PS] of the projects before and after the specification implementation was first conducted using a two-sample l-test. However, because a certain amount of skewness was apparent at the high serviceability end (5.0 PSI), the non-parametric Wilcoxon rank sum test was performed to more accurately qualify the significance of the impact of the specification. In the Wilcoxon rank sum test, the rank scores of individual observations within two samples (in this case, before- and after-specification roughness values) are summed and the mean scores are then compared via the z statistic and a specified confidence level. As seen in table 30, the results of both the t-test and the Wilcoxon rank sum test show a significant p-value at a 90-percent confidence level for all Pavement ~es. T T · · lo- . · · · ~·1 · . - - 1 - - -A r Hence, a slgn~ncant increase In serv~ceao~ty was observed corresponding to the implementation of the specification. Figures 69 through 72 provide the comparison of the distribution of the initial PSIs of the pavements covered by the smoothness specification with those not covered by the specification. These figures present very strongly that the smoothness specification did have a positive effect on the initial smoothness values of new pavement construction in Wisconsin. For each pavement type, the mean nutial PSI for the projects built after the Implementation of Me specification is significantly higher than that of the projects constructed before the ~rnplementation of the specification. The standard deviation of We Crucial PST of the projects under specification decreases significantly also. Table 30. Summary of t-test and Wilcoxon rank sum test results for Wisconsin projects. . Tw+Sample Wilcoxon Rank Pavement ~ Measuremen No. Of ~ Mean ~ Standard t-test ~! um Test Type Time Sections PSI Deviation He Significant? l ~P ~ ~p-va~u <0.1) ~ P . AC ~Before Spec 407 ~4.48 ~0.57 <0.01 ~Yet ~<0.01 After Spec 343 4.73 0.42 JPC Before Spec 102 3.84 0.51 <0.01 Yes <0.01 no dowels) ~After Spec ~ 125 ~4.59 ~0.35 JRC ~Before Spe 8 ~3.94 ~0.65 <0.01 ~Ye T 0 04 After Spec 14 4.63 0.26 . CRC || Before Spec 85 | 4.34 | 0.31 <001 T Ye. | cO.O1 ~I . 11 After Spec 82 1 4.78 1 0.25 1 1 1 1 . ~ Significant? (p-value<O.l) Yes Yes Yes Yes 125

160 140 120 v, u ·O 100 o ID 80 60 40 ~ 20 ~ O | Not covered by specification · Covered by specification Not covered bv specification Mean PSI = 4.48 Standard Deviation of PSI = 0.57 Covered by specification Mean PSI = 4.73 Standard Deviation of PSI = 0.42 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 Initial PSI Figure 69. Comparison of initial pavement smoothness distributions before and after implementation of smoothness specification (Wisconsin AC projects). 25 20 ~ v, - ._ ° 15 o 10 5 o · Covered by specification ONot covered by specification Not covered by specification Mean PSI = 3.84 Standard Deviation of PSI = 0.51 -ill Blml n.l.llJ Covered by specification Mean PSI = 4.59 Standard Deviation of PSI = 0.35 / ,111, 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 Initial PSI 4.0 4.2 4.4 4.6 4.8 5.0 Figure 70. Comparison of initial pavement smoothness distributions before and after implementation of smoothness specification (Wisconsin JPCP projects). i26

~ - - 4 v, - u a; ._ ° 3 o 2 1 O · Covered by specification ~N~iV~ Covered by specification Mean PSI = 4.63 Standard Dev~abon = 0.26 \ covered bv spec lion Mean PSI = 3.93 Standard Deviation = 0.65 \ .,: -'le'lI'lI' ' 1 L1 1 ,1 ~ ~ 1 1 1 1 1 ~ _ ~ 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 Initial PSI Figure 71. Comparison of initial pavement smoothness distributions before and after implementation of smoothness specification (Wisconsin JRCP projects). 30 25 20 15 lol- 5 O · Covered by specifica~don 1: , I , , let [lNot covered by specification Covered by specification Mean PA = 4.78 Standard Deviation of PSI = 0.25 \ Not covered by specification Mean PSI = 4.34 Standard Deviation of PSI = 0.31 \~' ~ I 1 ; - r ~ r " I' [1 ~ ~ ~ n 11 ~ all 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 Initial PSI Figure 72. Comparison of initial pavement smoothness distributions before and after implementation of smoothness specification (Wisconsin CRCP projects). 127

Kansas Kansas began experimenting with California profiIograph-based smoothness specifications in the mid-198Os (ParcelIs 19921. In 1985, the westbound direction of a portion of I-70 was reconstructed, while the eastbound direction was reconstructed the following year. The eastbound direction was constructed under a smoothness specification (requiring a smoothness of less than 12 in/ml tO.19 m/km] for full pay) whereas the westbound direction was not constructed under a smoothness specification. The roughness of each direction over the first 5 years after construction is shown in figure 73 (ParcelIs 1992~. The pavement sections in Me eastbound direction (constructed under a smoothness specification) show an initial roughness of 40 percent less than We pavement sections in the westbound direction (not under a smoothness specification). In addition, it is observed that the eastbound lanes have remained smoother over tune. Summary Four SHAs employing initial smoothness specifications have furnished data allowing for an investigation of the effect of Me smoothness specification on initial pavement smoothness. The data from all four States indicate that smoothness specifications have been effective in obtaining pavements that are significantly smoother Man those constructed prior to the Implementation of the specification. Typically, initial smoothness values have improved between about 19 and 89 percent within a few years after implementation. RIDEABILITY OF I- 7 C, DICKINSON CO. COUNTY MP 9.27 1 TO 1 7.000 100 ,_ 80 70 EASTBOUND LN En n LL T WE8T80U~ LN , _ - - `,, 60 ~so 40 1 986 - - - \ 1987 1988 1989 1990 Figure 73. Average pavement smoothness over time for a project In Kansas (eastbound lanes were constructed under a smoothness specification) (Parcelis 1992~. 128

From the available data, it appears that AC pavements are generally constructed smoother than PCC pavements, although significant reductions in initial pavement roughness were observed for all pavement types. For example, ranges In the reductions of Initial pavement roughness were found to be from 22 to 60 percent for PCC pavements, from 19 to 89 percent for AC pavements, and from O to 54 percent for AC overlay pavements. Finally, it also appears that it takes a few years for contractors to become acquainted with smoothness specifications. After the implementation of a specification, the Initial roughness generally decreases and continues to decrease as the contractor becomes more comfortable with the specification and more cognizant of items that can be done to increase the resulting pavement smoothness. Cost-Effectiveness of Initial Smoothness Levels and Smoothness Specifications As discussed in chapter 2, most highway agencies have adopted smoothness specifications (ride quality and/or bump specifications) that are intended to produce higher levels of Initial smoothness on their highway pavements. A review of several States' specifications has revealed those specifications to be largely effective in bringing about the desired higher levels of Initial smoothness. The specified initial smoothness levels, of course, have varied by highway agency and over lime, with some States being more zealous Han others by enacting tighter controls with the progression of time and the accompanying advancements in technology. To a fair extent, the development of smoothness specifications has been based on engineering judgment (offen by committee) or has been patterned after AASHTO or other agency's specifications (which were developed by committee judgment). According to the State survey responses, just over one-third of SHAs use engineering judgment or other specifications as the basis for Heir specified smoothness limits. A similar percentage indicated that these were the basis for selecting the relationship between incentive/disincentive payment and ~rutial smoothness. Although some States indicatec! performing research and analysis towards the establishment of critical Innits, the degree of objectivity included In those efforts is unknown. Because of the apparent subjectivity in many specifications and the everchang~ng state of the technology, it is quite likely that the most cost-effective levels of initial smoothness are not being specified. Some agencies may be operating at too lenient a smoothness level (resulting in shorter life and higher future rehabilitation costs), whereas others may be overly restrictive In their specifications (resulting In Increased initial construction costs). Another issue of concern is the fairness of ncentive/disincentive pay schedules that some SHAs have as part of their smoothness specifications. Even if the most cost-effective level of smoothness is being specified, it is generally not known if He pay adjustment rates are equitable in terms of the overall benefits/costs of building the pavement smoother/rougher. 129

Like so many other aspects of pavement design and construction (e.g., concrete strength, AC density, layer thicknesses), there is an optimum cost-effectiveness level that is defined by the conceptual relationship between initial smoothness and total life-cycle cost, as shown In figure 74. The total life-cycle costs of initial smoothness levels are largely dependent upon the specific conditions of each Individual paving project. Each project is unique In terms of its design, materials, methods used in the construction, and site conditions (e.g., traffic, subgrade, climates. In reality, determining the optimum level of initial smoothness is no different than determining the optimum level of concrete strength, layer thickness, or asphalt density. The contents of this section are twofold. First, a rational procedure for determining the most cost-effective ~rutial smoothness level for a given project is presented. The procedure uses life-cycle cost analysis (LCCA) techniques, along with initial smoo~ness-pavement life relationships and information on the additional costs of building smoother pavements. Secondly, using this evaluation procedure, a cost-effectiveness analysis of five States' specifications is presented. Point of Optimum Cost-Effectiveness ~1 - en' ce a - S ._ Total Cost Construction Cost M & R Cost 1 , 1 . . 1 . .' l _,~ 10 15 20 Initial Smoothness, in/ml Figure 74. Conceptual relationship between initial pavement smoothness and total life-cycle cost. 130

Cost-Effectiveness Analysis Procedures Life-Cycle Cost Analysis (LCCA) A pavement LCCA considers the stream of all costs anticipated to be incurred over a long, specified investment period (i.e., 30 to 50 years). Major cost items include initial construction costs, future maintenance and rehabilitation costs, anc! salvage value (a benefit), if any, at the end of We analysis period. LCC analyses may also include user-related costs, which consist of travel time costs, vehicle operating costs, accident costs, and user cliscomfort costs, although these are costs that are not actually incurred by the highway agency. Life-cycle costs can be expressed in terms of their "present-worth (PW) cost" or their "equivalent uniform annual cost (EUAC)." The PW method converts all future costs to Weir equivalent present-day cost using a selected discount rate. The converted future costs can be combined with the initial construction cost to give a total PW cost over the specified analysis period. If desired, the PW cost can be converted to an EUAC annual cost over the specified analysis period using engineering economics relationships and a selected discount rate. Figure 75 illustrates a typical Investment schedule and corresponding cashflow sequence for a highway pavement over time. In the top portion of the figure, the serviceability levels over time as a result of construction and various upkeep actions are depicted. In the bottom portion of the figure, the cashflow associated with each action is shown, with the downward arrows representing the costs incurred in the initial construction (CC) and upkeep (MCI and RCi) of the pavement and the upward arrow representing the salvage value (SV) of the pavement at the end of the analysis period. In figure 75, the PW cost of each future cashflow is calculated using the following equation: PW = Cashflow / ~ (1 + I)n] (6) where: i = Discount rate. n = Number of years over which costs are to be discounted. By sunun~ng the individual PW costs and the initial construction cost, the total PW cost (PW'O~) can be determined. If desired, the EUAC is then computed using the following equation: EUAC = PEW x ~ ~ i(1 f i/ + in-~ ~ ~ (7) 131

x ~ 4 ~ ~3 ._ - ._ u ._ ~ 1 cat _ ~ Terminal Serviceability = 2.5\ - _ 2 on ~· ······· ... · .... · .... · ... . 25 30 5 10 15 20 CC Age, years RC1 MCI MC2 MC3 ~ ~ i RC2 MC4 ~ l ~1 1 1 ' 11 0 5 10 15 20 25 34 e SV Figure 75. Typical investment schedule and corresponding cashflow of a highway pavement over many years. Several comments are appropriate on Me various cost elements. First, it is anticipated that smoother pavements have lower PW maintenance and rehabilitation costs as a result of the delay in their necessity. Also, the cost of constructing a smoother pavement has been a topic of debate within the pavement community. Many claim Cat there is a definite cost associated with constructing a smoother pavement, whereas others indicate Cat Mere is [idle, if any, additional cost. Still others believe that, for pavements constructed under an incentive/disincentive provision, the actual construction cost is less for a smooth pavement as Me contractor anticipates earrung the incentive payment and factors this into Me bid price. Another important consideration is if user costs are to be considered In the evaluation. User costs have a tremendous effect on the computed life-cycle cost values and Heir inclusion in the LCCA must be carefully considered. In a 1993 AASHTO LCCA survey, fewer than half the States indicated including user costs the LCCA, with a primary concern being He difficulty in estimating the value of delay (FHWA 1994a). Formulas for computing user costs based on pavement smoothness are available in publications by McFarland (1972), Peterson (1985), and others. 132

Conceptual Approach To assess the cost-effectiveness of initial smoothness levels, reliable cashflow estimates and an appropriate discount rate are needed, along with relationships between initial smoothness and construction cost and pavement life. Though not supported by actual data, the nutial construction cost of a pavement as a function of initial smoothness is conceptualized by the graph in figure 76. This graph shows a minimum construction cost corresponding to a marginally smooth, or baseline, pavement (i.e., one that is built by a contractor having reasonable focus on the level of smoothness provided). As a pavement is built smoother and smoother, greater attention to construction procedures is needed, which causes the initial CC to become increasingly greater. A perfectly smooth pavement (zero roughness), therefore, costs substantially more than a marginally smooth pavement. A ceiling limit on this extra cost might be the cost of completely diamond grinding a PCC pavement or milling and overlaying an AC pavement to try to achieve a perfectly smooth surface. Although the cost of constructing a pavement smoother is greater, the added pavement life associated with increased smoothness tends to offset to a certain point the additional construction cost, such that the total life-cycle cost is lower. The added life of the smoother pavement results in a postponement, and possibly some elimination, of future maintenance and rehabilitation costs. These deferred future costs, when converted to PW, are lower than the converted future costs of the marginally smooth, or baseline, pavement. The savings associated with additional pavement life may offset the higher construction cost required. lo .0 - - o o . - In o - ·0 ·_/ Minimum CC corresponding lo baseline Sonar baseline Son ~, , , , 1 , , , , 0 5 10 15 20 Initial Smoothness, inJmi Figure 76. Conceptual relationship between initial smoothness and construction cost. 133

At very high smoothness levels, a point may be reached where the benefit of added pavement life no longer offsets the cost of constructing the pavement smoother. This point is referred to as the point of optimum cost effectiveness, and it is represented by the absolute minimum shown In figure 74. To the right of the point of optimum cost effectiveness, the shorter pavement life (as a result of decreased nutial smoothness) results In higher maintenance and rehabilitation costs, thereby causing Increased life-cycle costs. To the left of the point of optunum cost effectiveness, the cost of constructing a pavement smoother begins to outweigh the · · . He · ~ ~ ~ ~ ~ · ~ ~ ~ . savings associated warn althea pavement fire, also resuming In increased life-cycle costs. The overall goal of this approach is to compare the estimated total life-cycle costs of pavements built to a nommal smoothness level versus those built to increasingly higher levels, and then identify the smoothness point or zone having the least cost. Although the cost effectiveness of a given crucial smoothness level will vary from project to project and from State to State (a function of pavement type, pavement design, construction cost, design reliability, and so on), similar cost-effectiveness levels for common types of projects (e.g., rural JRC construction, AC overlays on nterstates) could be determined for a given SHA, which could then lead to the development of a broad-based specification. Inclusion of Incentive/Disincentive Provisions Many highway agencies incorporate incentive/clisincentive provisions as part of their smoothness specifications. These provisions are intended to encourage the construction of smooth highway pavements through financial incentives for extremely smooth pavements and Trough financial disincentives for unacceptably rough pavements The magnitude of these incentives or disincentives is typically based on subjective judgment of We highway agency and are often expressed as a percentage of Me contract bid amount. Although smoothness specifications containing incentive/disincentive provisions are relatively new, several highway agencies have expressed concerns over the magnitude of incentives that have been paid out. Part of this concern is based on the appropriateness of current smoothness-measuring equipment and methods to accurately discriminate between initial smoothness levels deserving incentive pay. Another part of this concern is based on whether ~ncentive/disincentive provisions for pavement smoothness are an appropriate and cost-effective proposition. The results of the contractor survey presented in appendix A suggest that the overall construction cost, including the bid price and incentive/disincentive payment, is affected lithe by Me nutial smoothness level for most nominal levels of smoothness. More than tw~thirds of Me contractors indicated that project bid prices are generally adjusted to reflect the expected Incentive or disincentive payment. For instance, if a contractor develops a preliminary bid of $~.5 million on a paving project, but anticipates being able to achieve a smoothness level that will result In a $50,000 bonus, his final bid would be close to $~.45 minion. 134

Incentive/dis~ncentive provisions can be viewed as a means of "channeling" contractors toward the most cost-effective smoothness levels. Ideally, the full-pay smoothness level should coincide with the point of optimum cost-effectiveness. This would be the logical point where the SHA is willing to pay 100 percent of the bid price. However, it is not known exactly where this point lies for any given project. For current specifications, beginning at the full-pay smoothness level and moving toward zero roughness, the pay factor increases steadily (as a step or continuous function) with each increment of additional smoothness. However, these increases generally level off as zero roughness is approached; the idea being that the benefits are surpassed by the cost of building the pavement smoother and smoother. On the disincentive side, increasingly greater pay factor reductions are assessed for each incremental drop in initial smoothness, with a huge disincentive in the form of correction (i.e., grinding, milling, overlay, remove and replace) being ultimately reached at the bottom end of the disincentive range. The idea is to deter the contractor from constructing the pavement too rough by making the penalty more and more severe. The evaluation of the cost effectiveness of ~ncentive/dis~ncentive pay provisions is best done following a cost-effectiveness analysis of initial smoothness levels. In this way, the PW life-cycle costs and estunated bid prices (i.e., construction costs) for a range of initial smoothness levels can be Inputted Into the following pay factor formula to generate He theoretical pay factors over that range of smoothness levels: PFaC = [BPa,d + (Lccad--L`CCac)~/BPad where: PFac = Pay factor for as-constructed smoothness level. BPa ~= Bid price corresponding to as-designed (target) smoothness level, $. LCCad = PW life-cycle cost of maintenance and rehabilitation corresponding to as-designed smoothness level, $. LCCac = PW life-cycle cost of maintenance and rehabilitation corresponding to as-constructed smoothness level, $. The above equation is extracted from a report by Darter, et. al. (1993) that describes the development of a prototype performance-related specification (PRS) for concrete pavement construction. The pay adjustment procedure presented in that report was targeted for four concrete construction quality characteristics: strength, air content, thickness, and smoothness. In equation 8, the as-designed, or target, smoothness level is representative of a nominally smooth pavement a pavement built to a smoothness level that is generally achievable by contractors without them incurring significant additional construction costs. The LCCa~ and LCCac terms represent the PW values of He estimated future maintenance and rehabilitation costs for the target and as- constructed smoothness levels, respectively. The BPa~ term is the contractor's bid price corresponding to construction at the target smoothness level. 135

Equation 8 is perhaps better understood with the BPad term ~ the denominator located on the left side of the equation as a direct multiplier of PFaC. When viewed this way, the equation is interpreted as saying that what a highway agency should be willing to pay for is the cost of construction corresponding to a target initial smoothness level plus any added benefits of additional smoothness (represented by the PW difference of future costs associated with the target smoothness level and the as-constructed smoothness level). This amount must be equal to the product of PFaC and BPad By plotting, as a function of initial smoothness, the theoretical pay factors computed using equation ~ and Me pay factors associated with a current specification, an assessment can be made as to the cost-effectiveness of the current pay schedule. However, it is important that We full-pay smoothness levels (pay factor equal to I.0) of both pay factor curves coincide. If not, the theoretical pay factors can be recalculated using a target smoothness level equal to the full-pay smoothness level specified in the current pay schedule. The recalculated pay factors are again plotted and compared with the current pay factors. An example of such an assessment is shown in figure 77. Is , . - 1.4 - _.. ...._....... 1.3 1.2 U 1 ;` 0.9 0.8 0.7 0.6 0.5 ~ ~ Comical pay factor curve at_. ............. _._ ._ . . ~A I_ ' ~. . . _ . ~ A._ ~_ A. .. _ 1 '. '. . ~~CuITent pay schedule Smoothness Levels ... Ode ....... 1 i i 0 20 40 60 80 100 120 140 Initial Roughness, in/ml Figure 77. Example plot of pay factor versus initial smoothness. 136

Evaluation of the Cost Effectiveness of Smoothness Specifications To provide SHAs with a clearer picture as to whether their current smoothness specifying levels and incentive/disincentive pay provisions are appropriate and cost effective, a preliminary assessment was made of a few selected State specifications. These specifications are anonymously listed below. State ~ AC, JPC, AC/AC, and AC/JPC. State 2 CRC. · State 3 AC, JPC, AC/AC, and AC/PCC. State 4 PCC. · State 5 CRC, JPC, JRC, AC, AC/AC, and AC/PCC. The decision to analyze specifications relating to these pavement families was based primarily on the availability and reliability of the smoothness-life relationships developed earlier. Recall that some projects had insufficient or highly variable roughness data, which resulted in poor roughness models and, consequently, unreasonable smoothness-life trends. For State 3 AC, PCC, and AC overlay pavements and State 5 AC overlay pavements, cost effectiveness was assessed using two separate versions of the smoothness-life relationship; the first version generated from the roughness model analysis approach and the second version developed from He pavement failure analysis approach. ~ the case of He former, the smoothness-life trends of multiple projects of a given pavement family were plotted together and an "average" linear trend, representative of all individual trends, was determined. This trend was then used In the [CCA portion of the analysis. The second type of smoothness-life relationship used for States 3 and 5 was that type generated from the pavement failure analyses. The linear trend equations associated with 25 percent projects overlaid were used directly in the LCCA, resulting in an array of life-cycle costs for different crucial smoothness levels. These costs served as a cross-check of the costs computed using the roughness model approach. For the remaining pavement families, only the roughness model approach was used in the determination of smoothness-life relationships. Several assumptions were made in the conduct of the LCCA for the different pavement families, and these are listed in table 31. In this table, the unit bid price represents the construction cost associated with building the pavement to the target smoothness level. For simplicity sake, only the costs of future overlays were considered (i.e., no other maintenance or rehabilitation activities were considered). A standard life of 10 years was given to these overlays, with Heir timings being dependent upon the lives of the newly constructed pavement or newly placed overlay. 137

Table 31. Assumptions made In LCCA of pavement families. LCCA ~Pavement Type Input ~PCC (IO-in) | AC (3-in) | . Analysis Period, yrs 50 40 . Unit Bid Price of Paving $15/yd' $35/ton Bid Price, $/lane-nua 105,600 Il0,880 . Life of Future 4-in AC 10 10 Overlays, yrs Interest Rate, °/ . Inflation Rate, °/O 4 4 Discount Rate, °/O 4 4 (interest rate minus inflation rate) AC Overlay (6-in) 40 $35/ton 83,160 0 8 4 4 a Assumed 12-ft (3.7-m) wide lane and AC density of 150 lb/ft3 (2,430kg/m3). 1 my = 1.61 km 1 yd2 = 0.83 m2 1 ton = 0.907 metric ton The final consideration In the LCCA was the actditional smoothness cost (ASC), which was conceptualized earlier in figure 76. Input on this matter was obtained from three asphalt and three concrete paving contractors Trough a one-page questionnaire survey. The survey requested a graphical depiction of the estimated additional construction costs (in terms of a percentage of the unit bid price) for a range of crucial smoothness values fin terms of Me California PI with a 0.2-~n [5.1 mm] blanl<ing band), begiruiing with a target PI (5.0 in/ml [0.08 m/km] for AC construction, 7.0 in/ml [0.~! m/km] for PCC construction) representing 100 percent of We unit bid price and extending to a PI of O in/ml (O m/km). AC contractors were instructed to assume new construction of a mainline, multi-mile rural pavement, consisting of 8 in (203 mm) of AC (unit bid price of $35/ton [$35.60/metric ton]) placed on a uniform, stable base. Similarly, PCC contractors were instructed to assume new construction of a mainline, multi-mile rural pavement, consisting of lO-,n (254-mm) JPC (unit bid price of $15/yd2 [$17.90/m2~) placed on a uniform, stable base. Both contractor types were asked to exclude anticipated pay adjustments in their cost estimates. Figures 78 and 79 show the individual estimate curves and the resulting mean estimate curves for AC and PCC paving, respectively. As can be seen, there is a significant variation in We estunates, especially for concrete paving. The estimates are also greater than the percentages surmised from the first contractor survey; Pose being between 4 and 6 percent, corresponding to a shift from We full pay region (P] Of 5 to 7 in/ml [0.03 to 0.~l m/km]) to We maximum incentive region (PI of O to 3 in/ml [O to 0.05 m/km. Although some contractors indicated shortly before the 138

50 45 U o · - ._ _, ·Z o 51 40 35 30 25 ~ 20 - 15 10 O Mean Regression Equation: y = -1.9527x + 13.669 . ~ _g _ 0 1 2 3 4 5 6 California profilograph (0.2-in blanking band) PI, in/ml - Contractor 1 ~ Contractor 2 - - ~ - Contractor 3 %, '~ Mean Figure 78. PCC contractor estimates of the cost of additional smoothness. 50 ~ 45 5 40 35 o ~ 30 ._ ¢4 25 ._ ~ 20d ._ ~ 15 o ~ 10 ~ en, 5 O Mean Regression Equation: y = -3.2667x + 16.333 o California profilograph (0.2-in blanking band) PI, in/ml - Contractor 1 ~ Contractor 2 - - ~ - Contractor 3 )( Mean Figure 79. AC contractor estimates of the cost of additional smoothness. 139

survey-that a large portion of the cost of additional smoothness is attributed to achieving a uniform, stable working patio, each contractor was instructed to assume this proposition and base his estimated costs on over smoothness controlling efforts, such as setting more accurate stringlines, maintaining constant paver speeds, and providing more uniform material quality. The mean ASC estimate curves for AC and PCC construction were used respectively in all analyses. However, since these estimates were given as a function of the PI and only one pavement family State 4 PCC utilized the PI statistic, linear approximations for the other roughness statistics were developed. This was done by first converting the PI endpoints (0 and 5 in/ml [0 and 0.08 m/km] for AC, O and 7 in/ml [0 and 0.~1 m/km] for PCC) to the correlative values of the specified roughness statistic. The appropriate cost percentages were then applied to these values and hence a linear relationship was developed. Presentation of Results on Cost-Effectiveness of Initial Smoothness Levels Figures 80 Trough 82 show, for Tree pavement families, the construction, future (overlays), and total life-cycle costs as a function of ~rutial smoothness. Similar graphs for the remaining 17 pavement families were also developed, but are not included herein. The most cost-effective, or optimum, smoothness levels In these plots are the points where the total life-cycle costs are minimum. In most cases, these 250,000 Optimum level = 0 in/ml 200,000 .E AD a - u 100,000 I P" 0 25 50 75 100 125 Initial Roughness (MRN), inJmi Figure 80. Plot of life-cycle costs versus initial smoothness for the State ~ JPC pavement family. 140 _ Rehabilitation Cost Construction Cost Total Cost

160,000 140,000 ._ ~ 100,000 v) Ad 80,000 ~ ;- y ._ PA 120,000 60,000 ' 40,000 20,000 O 250,000 ·_ 200,000 1 - - a, a - ;^ 1 ._ 150,000 100,000 50,000 Optimum Level = 4.6 r ~ ~ U I l 3 Initial RI Figure S1. Plot of lif~cycle costs versus initial serviceability for the State 3 AC/AC interstate pavement family. Optimum Level = 4.55 . ~ | ~ Rehabilitation Cosi | ~ Construction Cost ~ ~ Total Cost _ it_ I · Rehabilitation Cost Construction Cost Total Cost o l I 3 3.5 4 Initial PSI 4.5 5 Figure 82. Plot of life-cycle costs versus initial serviceability for the State 5 AC pavement family. 141

points are not easily distinguishable because of the flatness of the curve, which is an important item to observe. Moreover, these points can be varied In response to changes in bid price, ASC estimate curves smoothness-life relationships and certain LCCA assumptions. 1 ' One of biggest factors in the variation of the optimum smoothness level is the ASC estimate curve. By varying the mean ASC estimate curve up or down Me amount of one standard deviation, a change In the optimum State ~ MRN of 20 to 25 in/ml (0.32 to 0.39 m/km) was effected and a change in Me optimum State 3 serviceability of 0.35 to 0.5 points was produced. Table 32 summarizes the optimum smoothness level, corresponding to Me following three ASC estimate curves: i. Mean minus standard deviation (p - Is) (3rci column). 2. Mean (~) (4th column). 3. Mean plus standard deviation (p + ask (5th column). Because of the variation in contractors' estimates of the ASC, the smoothness level ranges defined by columns 3 and 5 are considered to be most appropriate. Thus, for broader interpretation of findings, these ranges have been expressed in terms of the California PI (0.2-in [5-mm] blanking band), whereby the approximate PI range for each pavement family is listed In column 6 (located at the far right in table 32~. AS can be seen In table 32, many of the optimum PI ranges are in the range of 0.0 to 5.5 in/ml (0.0 to 0.09 m/km). Keying in on pavement type, seven of nine concrete pavement families show the oplunuan cost-effectiveness range as being between 0 and 5.5 in/ml (0 and 0.09 m/km), win all nine families having 0 in/ml as the boundary for the smooth end of Me range. For asphalt pavement families, four of five show Me opium range as being between 0 and 3.5 ~n/mi (0 and 0.06 m/km) and for asphalt overlay families, I! of 13 show the optimum range as being between Oand2in/mi(0and0.03 m/km). In the case of State 3 AC interstate pavements, recall that the smoothness-life relationship developed using pavement failure analysis techniques had a poor fit of the data and was very flat, causing any added life benefits to be far outweighed by additional smoothness costs. This is why the optimum smoothness range for those pavements is 12 to 15 in/ml (0.19 to 0.24 m/km) based on the failure analysis technique. The smoo~ness-life relationship for this family developed using the roughness models, on the other hand, was more reasonable, resulting in a higher optimum smoothness range (0 to 2.5 in/ml [0 to 0.04 m/km. As figures 3 and 4 in chapter 2 indicated, most of today's profilograph specifications begin incentive payments in the 5- to 7-~n/mi (0.08- to 0.11-m/km) range and disincentive payments in the 7- to 10-in/mi (0.11- to 0.16-m/km) range. Hence, on Me whole, Me full-pay, or target, range corresponds to 5 to 10 in/ml (0.08 to 0.16 m/km). This range is consistent with the target smoothness levels that were used in the LCCA: 5 in/ml (0.08 m/km) for AC pavements and 7 in/ml (0.11 m/km) for PCC pavements. 142

Table 32. Cost-effective smoothness ranges for various smoothness families. ~ ~I' Pavement Family Smoothness Index State 1 JPC MRN | State 1 AC State 1 AC/AC State 1 AC/PCC State 2 CRC State 3 PCC Interstate (1) roughness model analysis (2) pavement failure analysis State 3 PCC Parkway (1) pavement failure analysis State 3 AC Interstate (1) roughness model analysis (2) pavement failure analysis State 3 AC Parkway (1) pavement failure analysis State 3 AC/AC Interstate (1) roughness model analysis (2) pavement failure analysis State 3 AC/AC Parkway (1) roughness model analysis (2) pavement failure analysis State 3 AC/PCC Lnterstate (1) roughness model analysis (2) pavement failure analysis State 3 AC/PCC Parkway (1) roughness model analysis . State 4 PCC PI State 5 CRC PSI State 5 JRC State 5 JPC (w/o dowels) State 5 AC State 5 AC/PCC (1) roughness model analysis (2) pavement failure analysis State 5 AC/AC (1) roughness model analysis (2) pavement failure analysis I I Il I Cost-Effective Smoothness Range State Index Range Optimum Level Optimum Level for for low ASC moderate ASC (ASC 11 - a) (ASC p) Optimum Level for high ASC (ASC ~ ~ a) 30 5 o o 5.00 5.00 20 20 20 4.75 4.50 45 30 30 35 4.35 4.15 Approximate Range for CalifoInia PI, . , . 1n/m1 0.0 - 1.0 1.0 - 3.5 0.0 - 1.5 0.0 - 1.5 0.0 - 3.0 0.0 - 2.0 0.0 - 5.0 1 5.00 4.40 4.10 0.0 - 5.5 4.75 3.75 4.50 3.70 5.00 4.9S 4.95 4.75 5.00 4.80 4.60 4.60 4.40 4.90 4.35 3.65 4.65 4.50 4.40 4.20 4.70 0.0 - 2.5 12.0- 15.0 0.0 - 0.5 0.0 - 1.0 0.0 - 2.0 0.0 - 4.0 0.0 - 0.5 5.00 5.00 5.00 0.0 5.00 5.00 5.00 5.00 5.00 4.60 4.95 4.85 1 in/ml = 0.016 m/km 4.75 5.00 4.90 4.55 4.80 4.70 3.0 4.40 4.25 4.25 4.60 4.75 4.30 4.70 4.70 143 0.0 - 1.0 0.0 - 0.0 0.0 - 0.5 5.0 4.10 4.00 3.95 - 4.40 4.60 4.15 4.45 4.50 0.0 - 5.0 0.0 - 5.5 0.0 - 7.0 0.0 - 8.0 0;0 - 2.0 0.0 - 0.5 0.5 - 5.0 0.0 - 1.5 0.0 - 1.0

Based on the results given In table 32, an increase In the target smoothness range to between ~ and 5 in/ml (0.015 anc] 0.08 m/km) would be economically justified for many projects. Although contractor bid prices would likely be increased in response to the contractor's cost of attaining additional smoothness, the benefits of acIded pavement life would more than offset the increases. The ability of contractors to achieve these smoother levels is fairly well substantiated, as mainline pavements In several States are currently being built to smoothness ferrets of PI less than ~ ~n/mi (0.13 m/km). In California, for instance' AC and PCC pavements must be built to PI values of 5 and 7 ~n/mi (0.08 and 0.~! m/kmN, respectively. In another case, the mean PI values for 1992 AC and 1993 PCC ~ . ~ . . _ _ ~ _ . ~ . ,~ a_ _ ~ ~ ~ ~ ,~ ~ ~ . ~ tprunary) pavements In Iowa were 1 and / 1n/m1 (U.U1~ and U.ll m/Rm), respectively (see figure 54~. In still another case, 97 percent of all AC and PCC pavements built under Texas' 1993 specifications had PI values of 6 ~n/mi (0.24 m/km) or below and ~ ~n/mi (0.13 m/km) (easy working conditions) or below, resnectivelv (Frigate 1996~. a, , , ,. , ~ w~ , The re-establishment of the target smoothness level to PI values close to O in/ml (O m/km) would have some negative effects on contractors and SHAs. Contractors- particularly ones who have difficulty reaching smoothness levels of 5 to 7 in/ml (0.08 to 0.~1 m/km)- would be concerned! about their ability to achieve such Increased levels of smoothness, short of doing massive grinding or milling of the new pavement. SHAs, consequently, would be concerned about Me increased bid prices. A more comfortable target level setting for both parties, given these unplications, would be in the 3- to 5-~n/mi (0.05- to O.O8-m/km) range, with application of appropriate incentives and disincentives to help secure initial pavement smoothness In the optimum smoothness range. Further discussion of this type of action is provided ~ the following section. Presentation of Results on Cost-Effectiveness of Incentives/Disincentives A second aspect ~ the cost-effectiveness analysis of smoothness specifications is an evaluation of the appropriateness of incentive/disincentive payment provisions. Even if a highway agency may be correctly specifying the most cost-effective smoothness level, it is quite possible for them to be overly parsimonious or exceedingly generous in their pay adjustment amounts, based on LCCA and pay factor computation results. Although some SHAs believe that their incentive payments are too generous, an evaluation of the pay adjustments inherent in seven actual State specifications (listed below), as applied to various pavement families, showed otherwise. · State 1 PCC. State 2 PCC. State 3 AC and PCC. State 4 PCC. State 5 AC and PCC. 144

Table 33 lists the key components of these smoothness specifications, including the incentive, full-pay, disincentive, and corrective action smoothness limits, as well as the pay adjustments given for exceptionally smooth and less-than-adequate pavement. (It should be noted that the State ~ PCC and State 2 AC smoothness specifications were not evaluatec! the State ~ specification because of its project- specific nature and the State 2 specification because of the absence of roughness data for new AC construction and that State 4 has no AC smoothness specification.) For ease in comparing the pay adjustments, the limits associated with each pay adjustment were converted to the approximate limits of the smoothness statistics used in the LCCA. Figures 83 Trough 85 show the theoretical and actual pay adjustment curves for the State ~ JPC, State 3 AC/AC interstate, and State 5 AC pavement families (recall that the life-cycle cost versus initial smoothness plots for these three families were illustrated in figures 80 through 82~. Similar pay adjus~anent curves were cleveloped for 14 of the 17 remairung families (curves for State ~ AC, AC/PCC, and AC/AC could not be constructed), but again are not included herem. For each theoretical pay adjus~nent curve developed, the target (full-pay) smoothness level given by Me specification was used as the basis for calculating the theoretical pay adjustments. With one exception (State 3 AC interstate based on pavement failure analysis) out of 24, the theoretical pay incentives greatly exceeded the incentives allowed by the current specifications. And, although the State 5 asphalt specification was found to be more punitive with respect to disincentives, the disincentive provisions of me other six specifications were generally less severe than . , ~ the theoretical disincentives. Such findings are a clear indication Cat either the target smoothness level is too low or Cat fairly conservative estimates of added value were used in establishing me pay adjustments. Either way it is viewed, the most cost- effective levels In these cases are probably not being operated at, because the full-pay smoothness level is set too low and because the Incentives are not large enough to entice contractors to build to the optimum smoothness level. / As mentioned before, a more comfortable target level sewing for SHAs and contractors would be In the 3- to 5-in/mi (0.05- to O.OS-m/km) PI range. To determine what pay adjushnents might accompany such shifts in the target level, equation ~ was revisited for each pavement family. Instead of using the LCCad value corresponding to Me full-pay smoothness level of a given specification, the LCC values corresponding to PI values equal to 3 and 5 ~n/mi (0.05 and 0.08 m/km) for AC and PCC pavements, respectively, were used. This shifted each original theoretical pay factor curve toward He smooth end of the chart and consequently reduced the magnitude of the pay adjustments, as shown in figure 86. 145

v) o · - a · - u In On On a: o co ED cry ~5 - AD o - · at D CO Ct r ~ e: ~ . u R i, - ~ ~ 1~ t_ 5f ._ ~ ~ ~ ~ o o ~ c~ ~ o o ° Vl V A. 1 ~ o o o .G ~ ?= o o o o LO ~ C-i LO Vl o o so Go o V ~ - _` · =e o o - ·- V' ~ ~ ~i ~ =_ ~ a~ o-) .e ~ tm O. o o ~. U~ ~ L~ t O o ~i ^1 · S£~ ~ gmm O , ~ o~ ~ .5 ~ ~ ~ . °. o o~ o C~ ~ -, o o o ~ o~ - ,4~ ~ _ .= ~ ~ ~ o C~ . ~ o ~ o, ~ ~ ~ .5 ~ ~ ~ ~ ~ °. o o o o o U~ t' ~ ~ 4, ° o ~ - , ·g ){ -o ~ ._ , , o ~ ~ o o - , oo a~ . . ~ oo - ,c A ._ 00 A C 0. A ~ . .5 o. A _ 1 o A . ~ ·- S? o ~ o ~ ~ °o oo · - 11 o. C Lt~ 1_ ~4 l ° C _ ~Z ~ O ~n ~ ~ to _ ~o ~o o~ o- ~ o o m0 ~ ~ At ~, ~ o o ~ ^1 ~ O .... In O. g. en _o ~o _o o~ o~ o~ U) o ~ ~ ~ ~ °o U~ O , ~ I O O L~ . . . $~,8. v _ V ~ ~_ t}0 · 5 ~ ~,~.~.~,~,,i~D,.,3..,!~ mE .> ~ g~ _ ~.... . - L~ _ ~ - ·= c :m - 146 o o Vl 8= a' ~ ,= o ..° ~ ~ E 8, ~ c .o ° =O 11 {~S :m ~ .5 ~ ~ ~ ~ ~D Cl)

Dolooo 40,000 30,000 20,000 10,000 A} ~O o -10,000 -20,000 -30,000 Initial Roughness (MRN), in/ml · Theoretical pay amount using target MRN=60 . , . in/ me Current approximate pa, schedule Figure 83. Theoretical and actual pay adjustment curves for State 1 JPC pavements. 1.50 1 . . . 1 1.40 1.20 O 1.10 - 1.00 0.90 0.80 0.70 0.60 0.50 3 3.5 4 , '< ~ ~ ~ , 1 Initial Rl · PF using target RI-3.9 and excluding ASC Current Approximate Pay Schedule Figure 84. Theoretical and actual pay adjustment curves for State 3 AC/AC interstate pavements. 147

~ ~ - 1. ~ ~ ! , 1 , 1.20 , L _ 1.15 1.10 O 1.05 - i~, 1.00 0.95 0.90 0.85 0.80 _ ~ ~ ~ ~ ~ ~ ~ - r - - ~ - - _ _ _ _,_ _ _ _ _ _ _ _ _ _ _ _ .~> ·' ~ I ___,#__~____________________ / ~ ' J _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 3 3.5 Initial PSI · PF using target PSI=4.0 and excluding ASC Current approximate pay schedule Figure 85. Theoretical and actual pay adjus~nent curves for State 5 AC pavements. 1.l, ,, 1.05 u ;^ Pa 0.95 0.9 0.85 0.8 0.75 ~i ~. . . -Pay factor for original target smoo~ness level - - ~ - Pay factor for new target smoothness level #1 - ~-Pay factor for new target smoothness level #2 --- ~ --- Current pay schedule :~:1 ... ............... . .. . ,. ..... .. .. ; .; . ~. ............... -- ~---- ---- 1 ., _ ~; ~--- -~ - 1 0 20 40 60 80 100 120 Initial Roughness, inJmi Figure 86. Illustration of revised pay factor curve corresponcI~ng to higher target smoo~ness level. 148

Tables 34 through 38 summarize the results of this additional analysis, which was conducted for the various pavement types included with States ~ through 5. These tables show that the recalculated maximum incentives, corresponding to increased full-pay smoothness levels, still generally exceed those of the current specification. So, essentially, the results indicate that use of the current incentives at a smoother full-pay level, such as 3 to 5 in/ml (0.05 to 0.08 m/km), would be inadequate. Table 34. Comparison of current and hypothetical pay adjushnents for State ~ JPC. Pay Adjushnent Correspond ing to: l Maximum Incentive ~ ~Maximum Disincentive Pavement Family Full-Pay Impel MRN=0 inlmi MRN-60 irL/mi ~ MRN=80 in/ml (PI=0 in/mi) (PI=7 inlmi) (PI=12 inJmi) Current Level MRN=60 in/ml $9,850 $0 -$7,040 (PI=7 inimi) New Level Option 1 _ _ __ State 1 JPC MRN=50 in/ml $27,160 -$7,420 -$20,010 (PI=5 in/mi) New Level Option 2 MRN=40 in/ml $20,440 -$14,150 -$26,740 (PI=3 in/mi) 1 in/ml = 0.016 m/km Table 35. Comparison of current and hypothetical pay adjustments for State 2 CRCP. Pay Factor Corresponding to: Maximum Incentive ~ ~Maximum Disincentive Pavement Family Full-Pay Level IRI=0 in/mi IRI=50 in/ml it IRI=80 in/mi (PI=0 in/mi) (PI=6 in/mi) (PI=15 in/mi) . Current Level IRI=50 in/ml 1.03 1.00 (PI=6 inimi) New Level Option 1 State 2 CRC IRI=40 in/ml 1.14 0.95 (PI=5 in/mi) New Level-Option 2 IRI=30 in/ml 1.10 0.91 (PI-3 in/mi) . 0.90 0.81 0.76 1 in/ml - 0.016 m/km 149

Table 36. Comparison of currer~t and hypothetical pay adjustments for State 3. Pay Factor Corresponding to:' ~ Maximum Incentive ~ ~Maximum Disincenti~ve Pavement Family Full-Pay Level RI=5.0 (PI=O) RI=3.85 (PI=10) ~ RI=3.45 (PI=20) Current Level 1.02 1.00 _ RI=3.9 (PI=9 in/mi) State 3 New Level-Option 1 1.18 0.93 PCC Interstates RI=4.2 (PI=S in/mi) (1.12) (o.93) New Level-Option 2 1.14 0.89 RI=4.3 (PI=3 in/mi) (1.10) (0.91) RI=3.9 (PI=9 2n/mi) 1.02 1.00 State 3 New Level Option 1 (111) (0.96) PCC Parkvvays RI=4.2 (PI=5 inlmi) . New Leve on 2 RI-4.3 (PI=3 in/mi) 1 (1.10) 1 (0.94) _ . . current Leve' 1 15 1.00 RI=3.85 (PI=10 in/mi) State 3 New Level Option 1 1.16 0.89 AC Interstates RI=4.2 (PI=5 in/mi) (1.04) (1.00) New Level Option 2 1.13 0.86 Rl=4.3 (PI=3 in/mi) (1.04) (1.00) | RI=3.85 (PI-ZO ~n/mi) | ~ i5 ~100 State 3 New Level Option 1 (1.29) (0.77) AC Parkways RI=4.2 (PI=5 in/mi) . New Level Option 2 (1.24) (0.72) ~ RI=3.85 (PI ZO in/mi) ~1.15 ~1.00 State 3 AC/PCC New Level Option 1 1.22 0.84 Interstates RI=4.2 (PI=5 in/mi) (135) (0.79) . New Level Option 2 1.19 0.51 Rl=4.3 (PI=3 in/mi) (1.31) (o.74) 1 RI=3.85 (PI=ZO i ~Z 15 1 Z.OO State 3 New Level Option 1 134 0.79 AC/PCC Parkways Rl=4.2 (PI=5 in/mi) _ _ , . New v bon 2 Z.26 0 72 Rl=4.3 (PI=3 in/mi) ~ I Current Levelh ~ i5 T ~ °° State 3 AC/AC New Level Option 1 1.21 0.89 . Interstates Rl-4.2 (PI=5 in/mi) (1.19) (0.89) . New Level-Ophon 2 1.19 0.85 Rl=4.3 (PI=3 in/mi) (1.14) (0.84) _ t. . . Rl=3.85 (Pr-lo in/mi) 1.15 1.00 State 3 New Level~tion 1 1.15 0.91 AC/AC Park~vays RI=4.2 (PI=S in/mi) (1~1) (0.92) New Level Ophon 2 1.13 0.89 | RI=4.3 (PI=3 in/mi) | (l.19) | (0.9l) 0.92 0.79 (0.85) W7 (0.83) 0.92 (0.86) (0.84) 0.85 0.77 (0.96) 0.74 (0.96) 0.85 (0.52) (0.46) 0.85 0.64 (0.50) 0.62 (0.45) 0.85 0.53 0.46 0.85 0.73 (0.78) 0.71 (0.73) 0.85 0.81 (0.82) =7 (0.80) 1 in/ml = 0.016 m/km Values outside of parentheses are those determined from roughness model analysis. Values in parentheses are those determined from pavement failure analysis. 150

Table 37. Comparison of current and hypothetical pay adjustments for State 5 pavements. . Pay Factor Corresponding to:' . Maximum Incentive ~Maximum Disincen~ve Pavement Fam~ly Full-Pay Level PSI=5.0 PSI=4.00 ~ RI=3.60 (PI=0 inJnu) (PI=7 inJmi) (PI=15 in/mi) Current Zevel 1.05 1.00 0.92 PSI=4.0 (PI=7 in/mi) New Level ion 1 State 5 CRC | PSI=4.2 (PI=5 in/mi) | 1.10 | 0.97 0.89 | New Level{)ption 2 | 1.8 | 0.95 0.87 Current l~ve1 1.05 1.00 0.92 PSI=4.00 (PI=7 in/mi) State 5JPC I New Level Option 1 | 1.10 | 0.98 0.92 1. l | New Level Option 2 | 1.09 | 0.96 0 90 1' 1 1 PSI=4.00 (Pl=7 zn/mV 1 1.05 11.00 ~0.92 New Level ion 1 State 5JRC | PSI=4.2 (PI=5 in/zni) | 1.11 |0.98 | 0.91 Ii . | New Level Option 2 | 1.09 | 0.96 | 0.89 Curre7zz Level 1 1.00 1 1.00 ~0.60 State 5 AC I New Level Option 1 | 1.17 | 0.93 | 0.82 New Level ion 2 . | PSI=4.3 (PI=3 in/mi) | 1.14 | 090 I I PS1-4.00 (Pl-7 tn/mi) 1 1.00 1 1.00 1 0.60 St t 5 AC/PCC New Level~tion 1 1.26 0.94 0.82 a e PSI=4.2 (PI=5 in/mi) (1.13) (0.95) (o.9o) __ .. New Level Option 2 1.21 0.89 0.77 PSI=4.3 (PI=3 in/mi) (1.10) (0.92) (0.87) Currertz Level 1 1.00 1 1.00 1 0.60 St t 5 AC/AC New Level~ption 1 1.22 0.96 0.85 a e I PSI=4.2 (PI=5 in/mi) 1 (1.21) 1 (0.92) (0.82) 11 New Level Option 2 1.20 0.94 0.82 PSI - .3 (PI=3 in/mi) (1.19) (0.91) (0.80) 1 in/ml = 0.016 m/km a Values outside of parentheses are those determined from roughness model analysis. Values in parentheses are those determined from pavement failure analysis. 151

Table 38. Comparison of current and hypothetical pay adjustments for State 4 PCC. . Pay Factor Corresponding to: Maximum Incentive ~ | | Maximum Disincentive PI=0 inlmi PI=7 in/ml ~ PI=20 inlmi = 1.05 Pavement Family Full-Pay Level State 4 PCC Current LeKI PI=7 inimi New Level Option 1 PI=5 in/ml 1.10 New Level Option 2 PI=3 in/ml 1.00 0.90 0.95 j 0.33 0.89 0.28 l ' 1.05 1 in/ml = 0.016 m/km As discussed earlier, the SHA should be willing to pay the cost of building a pavement to a nominal smoothness plus any benefits of building it to an even smoother level. Although significantly greater pay adjustments, like those shown in figures 83 through 85, might seemingly entice many contractors to build to O in/ml (O m/km), contractors must take into consideration their cost of Paining additional smoothness. Hence, the level to which Hey build will be dictated by the difference between the anticipated incentive and the ASC. Contractors will build only to the level where He incentive still exceeds the ASC, for as one contractor pointed out In his response to the first contractor survey, "No contractor will pay more to get less." By subtracting the ASC in He numerator of He pay factor equation (equation 9), as shown below, the contractor's perspective on what smoothness level to build at can be better understood. PFaC = [BPa~ + (Lccad. ~ LCCaC)-ASC]/BPad (9) where: PFac = Pay factor for as-constructed smoothness level. BPad = Bid price corresponding to as-designed, or target, smoothness level, $. LCCad = PW life-cycle cost of maintenance and rehabilitation corresponding to as-designed smoothness level, $. LCCac - PW life-cycle cost of maintenance and rehabilitation corresponding to as-constructed smoothness level, $. ASC = Additional smoothness cost, $. The resulting pay factor curve indicates two things to the contractor: the break- even smoothness level and the most profitable smoothness level. These levels are illustrated in figure 87. At the break-even smoothness level, the cost of additional smoothness equals the incentive payment, and the resulting pay factor is 1.0. At the most profitable smoothness level, the difference between He incentive payment and the ASC is greatest, and the resulting pay factor is represented by the maximum 152

· Con~actor's view of pay factor curve (with | consideration of AS C) · Pay factor curve 1.05 - _.................................... 1.00 0.95 0.90 0.8= , , ... Hi,, _ - . ~ I' _ : Contractor s cost of ............ . . . ...... . additional smoothness ............ . . . ...... . . (ASC) . ·- , ~._ ~ e.~. \' At' 2 ...................... ;,,.. 1 a'_ Frost profitable "Break-even" smoothness levet .. smoothness level , _ . : : 4 4.5 5 Initial RI Figure 87. Illustration of contractor's perspective on pay adjustments (based on analysis of State 3 PCC parkway pavements). point on the contractor pay factor curve. It is exactly at this point that the rate at which additional incentive money is obtained becomes less Man the rate at which additional smoothness costs are incurred. Inclusion of User Costs in [CCA The inclusion of user costs in a LCCA is a controversial discussion topic in the transportation community. There is a general consensus that user costs must be included when conducting a LCCA; however, there is much controversy on what user costs should be included, how they should be calculated, and how they should be used. Although most States feel the LCCA should include user costs, there is a need for guidelines on how to accomplish this {ask. The LCCA performed up to now have only considered initial construction and future rehabilitation (overlay only) costs. However, because many people believe that user costs should be included in a comprehensive LCCA, an attempt was made in this study to include the effects of user costs as a function of initial serviceability. Although any estimate of user costs is a crude approximation, the magnitudes of the estimated costs presented in this report are believed to be representative. A discussion of this user cost investigation is provided below. 153

User costs were estimated based on results reported by McFarland (1972~. Four different types of user costs were considered in that investigation, consisting of lime (delay), vehicle operating, accident, and discomfort costs. All four were evaluated In terms of cents per vehicle mile as a function of serviceability (PSI) and the following six highway types: I. Rural 2 lane. 2. Rural- ~ lane, undivided. 3. Rural ~ or more lanes, divided. 4. Urban 2 lane. 5. Urban ~ lane, undivided. 6. Urban ~ or more lanes, divided. The cost tables provided In McFarIand's 1972 report were updated to 1996 costs using an annual Inflation rate of 5 percent. Because user costs are based on the number of vehicles, traffic data were required for the analysis. Average daily traffic (ADT) data was plotted versus time for a particular project. A polynomial regression equation was fit through the ADT versus time data In order to establish a reasonable relationship to serve as a means of extrapolating future traffic over the analysis period. The developed best-fit mode} (a function of Crucial smoothness and limed was used to predict the future roughness on a year-by-year basis for different Initial smoothness levels. Rehabilitation (for simplicity just the placement of an AC overlay) was applied when Me serviceability reached a level of 2.75 (as determined from the predicted-life-versus-initial-smoo~ness curse developed for each pavement type in each State). Overlays were assumed to last 10 years and deteriorate linearly from the target smoothness level to We trigger smoothness level over time. The appropriate cost per vehicle mile is obtained from Me updated cost tables by knowing the projected pavement smoothness and identified pavement type. Yearly user costs are calculated by multiplying the appropriate cost per vehicle mile tunes the projected number of vehicles traveling on the project for a given year (365 days times the projected ADT). For a given project, the cumulative user costs for different initial smoothness levels can be plotted versus time. For this Investigation, two specific Interstate pavement projects were evaluated. These projects consisted of one AC and one PCC project each located In Me same State and built In the mid 1970s. ~- , - - , Figures 88 and 89 show the cumulative PW user costs versus time for these two projects, respectively. Both graphs include all four types of user costs (i.e., time, vehicle operating, accident, and discomfort). Because of the magnitude of these costs, it is difficult to get a clear idea of the cumulative PW dollars saved over time due to 154

80,OOO,oOo 70,000,000 - - ~60,000,000 c lo, ' 50,000,000 40,000,000 30,000,000 1 O,000,000 60,000,000 a,, 50,000,000 ~40,000,000 o U u, 30,000,000 20,000,000 - - U 10,000,000 o Included User Costs: Time, VOC, Accident, Discomfort. Trigger of 17o in/mi, and a target of 60 in/ml were used. . ~ .-~~ T ,~ l I'> it, A' A' it,' O- , 1 1 0 5 1 0 15 20 25 30 35 40 45 50 Age (years) Figure SS. Cumulative PW user costs versus time for selected PCC pavement project. Included User Costs: Time, VOC, Accident, Discomfort. Trigger of 175 ~n/mi, and a target of 50 in/ml were used. 4,¢7# 0 5 10 15 20 25 30 35 40 45 50 Age (years) Figure 89. Cumulative PW user costs versus hme for selected AC pavement project. 155 O in/ml 60 in/ml (target) ---90 in/ml O in/ml 50 in/ml (target) ---70 inlmi

changes in initial smoothness. However, an estimate of the savings associated with a given initial smoothness level as compared to a target smoothness level can be determined using the following equation: SAV~GSX~ = CPWUCTARGEr~-CPWUCX(Y) (10) where: SAV~GSxc$' = Savings in cumulative PW user cost for a given initial smoothness level (X) at a particular year (Y), $. Cumulative PW user cost calculated for the target ~rutial smoothness level at a particular year (Y), $. CPWUCx = Cumulative PW user cost calculated for the given initial smoothness level (X) at a particular year (Y), $. In the case of the PCC project, table 39 contains values of actual yearly user costs, PW yearly user costs, and cumulative PW user costs for initial smoothness levels of 0, 60 (target), and 90 in/ml (0, 0.95, and 1.42 m/km). These three levels correspond roughly to California PI values of 0, 7 (target), and 15 in/ml (0, 0.11, and 0.24 m/km). Figure 90 illustrates the savings in cumulative PW user costs over time for the three different initial smoothness levels, as compared to the yearly values at the target initial smoothness level. For example, at year 15, the cumulative user costs (taken from table 39) for the three initial smoothness levels are as follows: Level 1: 0 in/ml (0 m/km) = $22,994,ooo Level 2: 60 in/ml (0.95 m/km) = $24,883J000 (target level) Level 3: 90 in/ml (1.42 m/km) = $26J143,OOO Therefore, the respective savings in cumulative PW user costs at year 15 (compared to the estimated value at the target of 60 in/ml (0.95 m/km)) are calculated as follows: Savings at O in/ml (0 m/km) => $24,883,000 - $22,994,000 = $1,889,000. Savings at 60 in/ml (0.95 m/km) => $24,883,000 -$24,883,000 = $0. Savings at 90 in/ml (1.42 m/km) => $24,883,000 - $26,143,000 = -$1,260,000. These three respective savings are plotted at year 15 in figure 90. It should be noted that the irregularities seen in figure 90 are due to the applications of overlays at different years for different initial smoothness levels. These different overlay schedules result in different rates of increasing cumulative PW user costs (see table 31 for further explanation of specific values and their interactions). By incorporating into the LCCA the cumulative PW user costs associated with different initial smoothness values and corresponding to a specified analysis period, a comprehensive graph of "total life-cycle costs plus user costs" versus initial smoothness can be developed. Figures 91 and 92 illustrate such graphs for the two subject projects. As can be seen in each of these figures, three different versions of 156

Age (years) 1 2 -3 4 5 6 8 10 11 2 13 14 ~15 16 17 is 19 20 2 22 23 24 25 26 27-- 28 . 29 30 31 32 33 34 35 36 . 37 38 39 40 41 42 43 . 44 45 =46~_ _47 _ 48 49 50 . Table 39. Actual yearly, PW yearly, and cumulative PW user costs for different Crucial smoothness levels ($ thousands). Yearly User Cost Per Mile 1,554 1,636 1,718 1,800 1 882 1964 2,046 2,128 2,210 2,292 2,374 2456 2,538 2,620 2 702 2,784 2,866 2,948 3,030 3,112 3194 3,276 3,358 3,440 3,672 3,811 3,982 4,150 4,315 4491 4,664 4,832 4,995 5,150 4527 4679 4,868 5,053 5,234 5,427 5,617 5800 , 5,975 6,142 5,381 5,546 5,754 5,956 6,153 6,364 Yearly PW User Cost Per Mile 1,494 1,512 1,527 1,539 1 547 1 552 1,555 1,555 1,553 1,548 1,542 1,534 - 1,524 1,513 1500 1,486 1 471 1,455 1,438 1,420 1401 1,382 1,362 1342 1,377 1,375 1,381 1,384 1,384 1 385 1 383 1,377 1,369 1,357 1 147 1,140 1,141 1,138 1,134 1,130 1,125 1 117 1,106 1,093 921 913 911 907 900 895 ~ Represents year of overlay. Cum. PW User Cost Per Mile 1,494 3,007 4,534 6,072 7619 9171 10,726 12,281 13,833 15,381 16,923 18 457 19,981 21,494 22994 24,481 25 952 27,407 28,845 30f265 - 31 666 33,049 34,411 35,753 37,130 38,505 39,886 41,270 42,654 44038 45,421 46,798 48,167 49J5254 50 672 , 51,812 - 5-2,952 54,091 55,~5 56,355 -57,48 - 0 58,597 59703 , 60,797 61,718 _ -62,63 63 542 . , 64,448 65,349 66,244 ~ Yearly User | Cost Per L Mile 1,619 1,711 1,805 1,900 2,002 2108 . ~ 2~16 2,326 2,438 2,551 2,666 2,788 2,911 2,731 2,858 3 008 . ~ 1 3157 . , 3,304 3,461 3,616 3 768 . ~ 1 3 916 . , 4,060 3,586 3,725 3,893 4,060 4~3 1 4,397 1 4 569 . , 4,735 4,897 5,051 4,441 4,592 4,779 4,963 5,142 5,334 5,522 5,703 5877 , , 6,042 5296 5,459 5,665 5,866 6,061 6,270 6,474 157 . Yearly PW User Cost Per Mile 1, 5 5 7 1,582 1,604 1,624 1,645 1,666 1,684 1 700 1,713 1,723 1,732 1,741 1 748 1,577 1,587 1 606 1621 1 631 1,643 1,650 1,654 - 1,653 1,647 1,399 1,397 1,404 1,408 1,408 1,410 1 409 1,404 1,396 1,384 1,170 1,164 1,165 1,163 1,158 1,155 1,150 1,142 1132 1,119 943 935 933 928 923 918 911 _ Cum. PW User Cost Per Mile 1,557 3,138 4,743 T 6,367 8,012 9,678 T 11362 , ~ 1 13061 , ~ 1 14,774 1 16 498 , , 18,229 9,971 21719 , , 1 23,296 24,883 ~26,489 28 109 , , 1 29740 , , 1 31,383 T 33,033 1 34 687 , , 1 36 339 , , 37,986 39,386 4~0,783 42,187 43,595 1 45,003 r 46413 , , 1 47 822 , , 49,226 50,622 52,0~06 53,177 54,340 55,505 56,668 57,826 58,982 - 60,132 61,274 62,406 63,525 64,467 65,402 66 335 , , 67,263 68,186 69,103 70,014 ~ Yearly User| Yearly PW | Cum. PW | Cost Per | User Cost | User Cost | Mile ~ Per Mile | Per Mile 1~774 1 1,706 1 1,706 _1,878 1 1,737 1_ 3,443 1,985 1 1,765 1 5,207 2,094~ -1,790-r- 6,998 2,205 ~1,812 T 8,810 2,318 11,832 ~10,642 2 432 11 848 ~12 490 , . .. . . 2,547 11,861 L 14,351 2,304 11619 1 15970 , .. . . 1 2424 11638 1 17607 , , I, , , 2,565 11,666 1 19,273 2,705 11,690 1 20,963 r ~2,845 T 1,7081-- 22,671 2,992 1 1,728 1 24,399 1 3,140 1 1,743 1 26,143 L_ 3,284 T 1,754 1 27,896 r ~ 3,426 1 1,759 1 29,655 3,564 1 1,759 1 31,415 1 ~ 3,159 1 1,499 1 32,914 r 3,?91T 1,5-02~ --34,416 rim 3,441~ ~1,514 1~_~35,930 r36081 15231 37453 ,. . . . . 13,764 1 1,527 1 38,980 1 .- . 3~929 1 1,533 1 40,513 4,092 1 1,535 1 42,048 14,2521 1,534 r 435811 4,407 1 1,528 1 45,110 14,555 1 1,519 1 46,6296 I~ 4014 1 1,287 1 479161 I~ 1 1 ~ 14,158 1 1,282 1 49,198 1 . i4,336 1 1~286 1 50,483 --4,511 ~1,286 1 51,769 4,683 ~1,284 1 53,053 ,866 t 1~2821 54,335 15,045 1 1,279 1 55,614 i5,219 1 1,272 1 56,886 5,387 1 1,262 1 58,1480 5,547 1 1,250 1 59,397 4,869 1 1,055 1 60,452 5,026 1 1,047 1 61,499 1 ~ Owl ~:~ 1 5,414 r 1,043 1 63,587 1 . . , 1 5,602 1 1,037 1 64,625 1 1 5,802 r-10331 656-581 5,99~1__~ 1,027 l 66,684: 6 187 T 1 018 1 67 703 , . . . . . 6,367 1 1,008 1 68,711 6,53-8 T 9951 69,706 *5,723 T 838 1 70,543 5,893 1 829 1 71,372

.,, 3,000,000 ._ OJ c~ v 2,000,000 U v' 1 ,000 ,000 ._ - U u, ._ U. -1 ,000,000 -2,000,000 ~. ~ _ ~.. .... ,\.. _ 4;0 ~ ; 2 ~Nf, v I T I I I Included User Costs: Time, VOC, Accident, Discomfort. , . . ~ . ., ~ r ~ ~ ~ ~ r ) 2 5 ~, Age (years) , , ~0 s e Figure 90. DIustration of sav~ngs ~n cumulative PW user costs over fime. so,ooo,ooo - . - . . . . . 1 1 1 1 1 70~000tO00 ~- ~ - ; . .. ................................ 60,OOO,OOO 6~ - U 50J000~000 - U 40,000,000 30,000,000 20,000,000 10,000,000 +O in/ml ~60 in/ml (target) ~90 in/ml 1 _ _ _ ~ 0 20 40 60 80 100 120 140 Initial Roughness (MRN), in/ml Total LCC · Total LCC + VOC ~Total LCC + Time Delay + VOC Total LCC + Time Delay + VOC + Accident + Discomfort Figure 91. Graph of total LCC (including user costs) vs. ~nitial roughness for selected PCC pavement project. 158

4D,000,000 1 40,000,000 ~ '_ -A 3~,000,000 30,000,000 25,000,000 20,000,000 15,000,000 10,000,000 5,000,000 0 20 40 60 80 100 120 140 Initial Roughness (MRN), in/ml | Total LCC | · TotalLCC +VOC Total LCC + Time Delay + VOC . Total LCC + Time Delay + VOC ~ Accident + Discomfort Figure 92. Graph of total LCC (including user costs) vs. iriitial roughness for selected AC pavement project. user costs were depicted; the top curve representing all four user cost components, the curve below that representing two user cost components twine and vehicle operating), and the one below Cat representing only vehicle operating costs. The total life-cycle cost curve, cor~sist~ng of initial construction cost and future overlay costs, is located close to the x-axis. The effect of including user costs In the LCCA is obvious in both of these figures. User costs so overwhelmingly dominate the project life-cycle costs that the least overall cost occurs at the smoothest possible level, which is a MRN of O m/ml (O m/km). Although other projects were not analyzed, variations in savings from project to project would be expected, primarily due to factors such as the smoothness-life relationship and traffic. Nevertheless, it is strongly believed that the magnitude of savings associated with smoother levels for other projects would stay relatively uniform, such Cat Me smoothest level is always identified as Me most cost- effective level. Summary This section outlined a procedure for evaluating the cost-effectiveness of initial smoothness levels and presented a detailed evaluation of several current pavement smoothness specifications using that cost-effectiveness procedure. The evaluation procedure introduced is based on LCCA techniques and is contingent on the 159

inclusion of relationships between initial smoothness and initial construction cost and between initial smoothness and pavement life. The evaluation of current smoothness specifications focused on two aspects: the optimum cost-effective smoothness level and the appropriateness of ~ncentive/dis~ncentive payment amounts. Asphalt and concrete specifications from five different States were evaluated using LCCA techniques, contractor estimates on the cost of obtaining additional smoothness, and the smoothness-life relationships developed earlier for various pavement families (i.e., combinations of State and pavement type). Some of the key findings of the smoothness specification cost-effectiveness evaluation are as follows: . Seven of rune concrete pavement families showed Me optimum cost- effectiveness (Pl) range, excluding consideration of user costs, as being between O and 5.5 in/ml (O and 0.09 m/km). Four of five asphalt pavement families showed the optimum cost-effectiveness (Pl) range as being between O and 3.5 in/ml (0 and 0.06 m/km). Eleven of 13 asphalt overlay families showed Me optimum cost-effectiveness (Pl) range as being between O and 2 in/ml (0 and 0.03 m/km). In comparison with actual current pay adjustment curves, the theoretical pay acljustment curves, on the whole, showed that greater incentive and disincentive amounts are warranted in terms of the benefits/disbenefits obtained from various initial smoothness levels. When shifted to coincide with full-pay (Pl) smoothness levels of 5 and 3 in/ml (0.08 and 0.05 m/km) for PCC and AC pavements, respectively, the recalculated theoretical pay adjustment curves still showed greater magnum Incentive amounts and more punitive disincentive amounts in comparison with the current pay adjushnent curves. The inclusion of user costs in a comprehensive LCCA has a profound effect on the determination of the most cost-effective smoothness level. For two selected pavement projects, a comparison of cumulative PW user costs associated with Tree distinct crucial smoothness levels (MRNs of 0, 60, and 90 in/ml [0, 0.95, and 1.42 m/km]) showed significant cost savings for Me pavement constructed to the smoothest level. For both projects, Me addition of user costs to Me project costs (construction cost plus future overlay costs) clearly dominated the total LCC such that the most cost-effective smoothness level was found to be 0 ~n/mi (0 m/km). Overall Chapter Summary This chapter presented the results of several analyses conducted on the effect of Initial pavement smoothness and of pavement smoothness specifications. The results of the analysis of the effect of initial pavement smoothness on the future smoothness of the pavement clearly show that nutial pavement smoothness has a significant effect 160

on the future smoothness of the pavement. This analysis was conducted for different pavement types and for different age ranges, with the results being that initial pavement smoothness is significant to the future smoothness of the pavement. The effect was stronger for new pavement construction than for overlay pavement construction, suggesting that the performance of overlays is governed more by other factors (e.g., reflection cracking). These results were conducted on many sets of independent pavement performance data from across the US and over different pavement types. Analyses were also conducted to investigate the effect of Crucial pavement smoothness on pavement life. The results of the analysis strongly indicate that initial pavement smoothness has a significant effect on pavement life, using both roughness mode! and pavement failure analysis techniques. The analyses show that added pavement life is obtained by achieving a higher level of initial smoothness over the range of initial smoothness values that were available for analysis. The rate at which additional life is achieved is dependent upon, among other things, pavement type, facility type, and location. Sensitivity analyses, in which the percentage change in life as a function of percentage change in smoothness was determined, showed sizable increases In life for most pavement families, corresponding to nominal increases in smoothness. At least a 9 percent increase In life corresponding to a 25 percent increase In smoothness (from a target profile index of 7 in/ml [0.~l m/km] for concrete and 5 in/ml [0.08 m/km] for asphall) was observed for the vast majority of the pavement families. A 50 percent increase in smoothness from these target levels was found to Increase life by at least 15 percent ~ many cases. Me results of Ads analysis are based on the assumption that roughness is a primary factor influencing the decision to rehabilitate a pavement. An analysis of the effect of pavement smoothness specifications on the resulting levels of initial pavement smoothness was also conducted. This analysis indicated that pavement smoothness specifications do have a beneficial effect on the resulting initial smoothness of the pavement. Comparison of pavement smoothness distributions before and after the implementation of a smoothness specification indicate decreased initial roughness values and decreased variability for pavements constructed under the smoothness specification. A procedure for evaluating the cost-effectiveness of initial smoothness levels was developed and presented. The procedure illustrates how the life-cycle costs of a pavement constructed to different smoothness levels (under various incentive/disincentive scenarios) can be determined and Men compared to see which level is most cost-effective. Such an evaluation was conducted on several pavement families, and the results showed Mat Me most cost-effective smoothness (Pl) levels are considerably higher than what is typically used as the current target (PI between 5 and 10 in/ml (0.08 and 0.16 m/km. Seven of nine concrete pavement families showed the optimum cost-effectiveness range as being between O and 5.5 ~n/mi (0 161

and 0.09 m/km), whereas four of five asphalt pavement families showed the optimum range as being between O and 3.5 Mimi (0 and 0.06 m/km). In comparison with actual current pay adjustment curves, the theoretical pay adjustment curves developed in this study showed, on the whole, much greater incentive amounts and much more punitive disincentive amounts. When shifted to coincide with full-pay (Pl) smoothness levels of 5 and 3 in/ml (0.08 and 0.05 m/km) for PCC and AC pavements, respectively, recalculated theoretical pay adjustment curves stiD showed greater maximum incentive amounts and more punitive disincentive amounts than current pay adjustment curves. The inclusion of user costs in a comprehensive LCCA has a profound effect on the determination of the most cost-effective smoothness level. For two selected State pavement projects, me addition of user costs to total life-cycle costs resulted In O n/mi (O m/km) as being the most cost-effective smoothness level. Taken together, the implications of these findings are very significant. Initial smoothness has been shown to provide value in the performance and service life of pavements. This serves to justify the importance of achieving high initial smoothness from a pavement structural standpoint, which is in addition to the importance of smoothness to the nding comfort of the user. And, this strongly supports the benefits of employing smoothness specifications, which have been shown to be an effective means of achieving higher levels of initial smoothness. 162

Next: Chapter 5. Study of Smoothness-Measuring Equipment and Indices »
Smoothness Specifications for Pavements: Final Report Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!