National Academies Press: OpenBook

Use of Fiber-Reinforced Polymers in Highway Infrastructure (2017)

Chapter: Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure

« Previous: Chapter Four - Codes, Standards, and Design Guidelines
Page 19
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 19
Page 20
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 20
Page 21
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 21
Page 22
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 22
Page 23
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 23
Page 24
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 24
Page 25
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 25
Page 26
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 26
Page 27
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 27
Page 28
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 28
Page 29
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 29
Page 30
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 30
Page 31
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 31
Page 32
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 32
Page 33
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 33
Page 34
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 34
Page 35
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 35
Page 36
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 36
Page 37
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 37
Page 38
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 38
Page 39
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 39
Page 40
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 40
Page 41
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 41
Page 42
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 42
Page 43
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 43
Page 44
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 44
Page 45
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 45
Page 46
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 46
Page 47
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 47
Page 48
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 48
Page 49
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 49
Page 50
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 50
Page 51
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 51
Page 52
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 52
Page 53
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 53
Page 54
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 54
Page 55
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 55
Page 56
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 56
Page 57
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 57
Page 58
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 58
Page 59
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 59
Page 60
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 60
Page 61
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 61
Page 62
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 62
Page 63
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 63
Page 64
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 64
Page 65
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 65
Page 66
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 66
Page 67
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 67
Page 68
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 68
Page 69
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 69
Page 70
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 70
Page 71
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 71
Page 72
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 72
Page 73
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 73
Page 74
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 74
Page 75
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 75
Page 76
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 76
Page 77
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 77
Page 78
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 78
Page 79
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 79
Page 80
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 80
Page 81
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 81
Page 82
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 82
Page 83
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 83
Page 84
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 84
Page 85
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 85
Page 86
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 86
Page 87
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 87
Page 88
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 88
Page 89
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 89
Page 90
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 90
Page 91
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 91
Page 92
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 92
Page 93
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 93
Page 94
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 94
Page 95
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 95
Page 96
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 96
Page 97
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 97
Page 98
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 98
Page 99
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 99
Page 100
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 100
Page 101
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 101
Page 102
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 102
Page 103
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 103
Page 104
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 104
Page 105
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 105
Page 106
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 106
Page 107
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 107
Page 108
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 108
Page 109
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 109
Page 110
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 110
Page 111
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 111
Page 112
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 112
Page 113
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 113
Page 114
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 114
Page 115
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 115
Page 116
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 116
Page 117
Suggested Citation:"Chapter Five - State of the Art of Fiber-Reinforced Polymer Composites in Highway Infrastructure." National Academies of Sciences, Engineering, and Medicine. 2017. Use of Fiber-Reinforced Polymers in Highway Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/24888.
×
Page 117

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.

19 chapter five State of the art of fiber-reinforced Polymer comPoSiteS in highway infraStructure fiber-reinforced Polymer-reinforced concrete memberS This section provides an overview of FRP-reinforced concrete members and their behavior under flexure, shear, and axial loadings. Miscellaneous subjects such as bond and development are dis- cussed as well. frP reinforcing bars FRP bars, manufactured by pultrusion and braiding techniques (Harris et al. 1998; Portnov et al. 2013), are an anisotropic material. When reinforcing concrete members, the unidirectional mechanical proper- ties of FRPs in the longitudinal direction are considered to be important. The density of FRP bars typi- cally varies from 78 lb/ft3 (1.25 g/cm3) to 175 lb/ft3 (2.8 g/cm3) (Nkurunziza et al. 2005; Nanni et al. 2014), which is about 15% to 35% of the steel density of 500 lb/ft3 (8 g/cm3). FRP bars remain linear elastic until failure, owing to the tensile rupture of the fibers (Figure 3). The function of FRP bars in a concrete structure is the same as that of steel reinforcement, including surface treatment to enhance bond against concrete (e.g., deformed ribs, sand coating, and periodic helical fiber wraps), as shown in Figure 4. FRP bars have been used as flexural reinforcement in most cases, although research revealed their potential as shear stirrups (Shehata et al. 2000; El-Sayed et al. 2007; Lee et al. 2014; Oller et al. 2015) and axial reinforcement (Choo et al. 2006; Zadeh and Nanni 2013; Afifi et al. 2014; Prachasaree et al. 2015). The most notable advantage of GFRP bars can be found in noncorrosive characteristics, which will save on the maintenance and repair expenses of constructed concrete structures, beneficial for bridge decks that are directly exposed to deicing chemicals. The literature reports that the life-cycle costs of FRP-reinforced concrete members are substantially lower than the costs of their steel counterparts (Pearson et al. 2011). FRP bars cannot be bent on site, because they are intrinsically thermoset- based products, whose deformation and properties are irreversible once the resin has cured (see Chapter 2, “Resins”). Manufacturers can produce FRP bars with curved or bent geometries in- plant, employing a special mould. Pre-ordering is necessary if specific dimensions and geometries are required for a construction project. Unlike steel reinforcing bars with standardized nominal diameters and cross-sectional areas, FRP reinforcement has variable geometric properties dependent on manufacturers. For instance, No. 5 GFRP bars produced by Manufacturers A and B have cross-sectional areas of 0.34 in.2 and 0.31 in.2 (218 mm2 and 198 mm2, respectively). The actual area of FRP reinforcement needs to be considered in structural design. Table 5 lists a designation for FRP reinforcing bars taken from ACI 440.6-08 (ACI 2008b) and AASHTO (2009). Although several fiber types are available, GFRP reinforcing bars are exclusively used because of their relatively low costs. BFRP has potential, although further research and site demonstration are required. As discussed in Chapter 2, “FRP Materials,” the properties of FRP bars are con- trolled by the amount of embedded fibers and their types. Because load-carrying fibers are placed along the longitudinal direction, the strength of FRP bars in the transverse direction is controlled by the polymeric resin. The dowel action considered in the shear design of a concrete beam is, therefore, significantly lower than that of steel-reinforced cases. Typical properties of FRP bars

20 FIGURE 3 Stress-strain relationship of GFRP bar (reproduced based on Hughes Brothers’ data sheet). FIGURE 4 Surface treatment of FRP bars (used by permission from Yail J. Kim). Source: ACI (2008b); AASHTO (2009). Bar Size [U.S. customary (metric)] Nominal Diameter [in. (mm)] Nominal Cross- Sectional Area [in.2 (mm2)] Minimum Guaranteed Tensile Strength of GFRP Bars [ksi (MPa)] 2 (6) 0.250 (6.4) 0.05 (32) 110 (760) 3 (10) 0.375 (9.5) 0.11 (71) 110 (760) 4 (13) 0.500 (12.7) 0.20 (129) 100 (690) 5 (16) 0.625 (15.9) 0.31 (199) 95 (655) 6 (19) 0.750 (19.1) 0.44 (284) 90 (620) 7 (22) 0.875 (22.2) 0.60 (387) 80 (586) 8 (25) 1.000 (25.4) 0.79 (510) 80 (550) 9 (29) 1.128 (28.7) 1.00 (645) 75 (517) 10 (32) 1.270 (32.3) 1.27 (819) 70 (480) TABLE 5 DESIGNATION OF FRP BARS

21 collected from manufacturers in the United States are summarized in Table 6. It is worth noting that (1) the size of FRP bars can influence mechanical properties; for example, GFRP bars pro- duced by Manufacturer A show tensile strengths of 120 ksi and 70 ksi (825 MPa and 480 MPa) for No. 2 and No. 10 bars, respectively; (2) FRP manufacturers provide guaranteed strength (i.e., µ - 3s, where µ and s are the mean and standard deviation, respectively, which may be obtained by ACI 440.3R-12); and (3) the material properties shown in Table 6 may not be achievable in a laboratory test, unless special anchorage is employed to prevent stress concentrations at the gripped areas. A number of external attributes can affect the properties of FRP bars such as moisture, ultra- violet radiation, temperature, and acidic chemicals (Nkurunziza et al. 2005). The gradual time- dependent behavior of FRP bars in sustained load may cause creep-rupture, which leads to an abrupt failure without warning. Unlike GFRP bars, CFRP bars do not show noticeable creep- rupture (Al-Salloum and Almusallam 2007). According to some recent experimental studies (Wang et al. 2014; Banibayat and Patnaik 2015), BFRP bars appear to be insensitive to creep. However, further research is warranted to generalize these findings. A stress limit exists in FRP bars below which creep-rupture does not take place, regardless of loading period. Another time- dependent characteristic to mention is the fatigue of FRP bars. It can be noted that the fatigue behavior of FRP bars is controlled by many parameters (e.g., stress ranges, mean stresses, load- ing frequencies, and test methods). As with all engineering materials, an S-N curve (stress range versus failure cycles) provides a convenient means for evaluating the fatigue performance of FRP bars. Research shows that the fatigue life of CFRP and GFRP bars is marginally reduced by 5% to 8% up to 1 to 2 million cycles, irrespective of fiber type (Mandell and Meier 1983; Roylance 1983; Curtis 1989). The fatigue life of CFRP bars tends to be influenced by temperature [e.g., an increase in temperature from room temperature to 104°F (40°C) caused a decrease in fatigue life of CFRP bars of more than 10 times (Adimi et al. 2000)]. A few recent experimental results revealed that BFRP bars have fatigue resistance comparable with other conventional FRP bars (Li et al. 2012; El Refai 2013). flexural behavior FRP-reinforced concrete can include an over-reinforced section, which is not preferred in steel- reinforced concrete. Despite the absence of a yield plateau in FRP-reinforced concrete beams, their impending failure can be revealed by large deflections and excessive cracking. The reserved flexural strength can complement a lack of ductility. This behavior may be referred to as pseudo-plastic failure (Nanni and Faza 2002). Research has been conducted to alter the brittle failure of FRP-reinforced concrete beams to pseudo-ductile failure. Hybrid FRP bars consisting of braided aramid and carbon fibers were developed to induce the step-wise rupture of the fibers (Harris et al. 1998), and a hybrid reinforcing scheme was proposed with steel and GFRP bars to address ductility and durability require- ments (Pang et al. 2016). In some cases, FRP grids were used to reinforce concrete members (Yost et al. 2001). High-strength concrete may be employed to better accommodate high-strength FRP bars (GangaRao and Vijay 1997). The load-deflection behavior of FRP-reinforced beams shows a bilinear trend, with a transition in tangent before and after the cracking load (Figure 5), contrary to the trilinear trend of steel-reinforced Based on ACI (2015a). Bar Fiber Tensile Strength [ksi (MPa)] Tensile Modulus [ksi (GPa)] Strain at Failure (%) AFRP Aramid 250–367 (1,720–2,540) 6,000–18,2000 (41–125) 1.9–4.4 CFRP Carbon 87–535 (600–3,690) 15,900–84,000 (120–580) 0.5–1.7 GFRP Glass 70–230 (483–690) 5,100–7,400 (35–51) 1.2–3.1 TABLE 6 MECHANICAL PROPERTIES OF FRP REINFORCING BARS

22 beams. The post-cracking flexural rigidity of FRP-reinforced concrete beams increases as the reinforcement ratio rises (Tomlinson and Fam 2015). Modification factors were proposed to adjust the deflection of FRP-reinforced concrete beams (Theriault and Benmokrane 1998; Toutanji and Saafi 2000; Kara and Ashour 2012). These factors are empirically calibrated based on limited test data and, consequently, certain limitations exist in application. Cyclic loading may not be a factor decreasing the flexural stiffness of FRP-reinforced concrete beams (Theriault and Benmokrane 1998). Limited moment redistribution can occur in continuous FRP-reinforced concrete beams, because of cracking and bond failure between the FRP and concrete (Kara and Ashour 2013). This redistribution mechanism is intrinsically different from the mechanism of steel-reinforced concrete caused by the yielding of steel bars. FRP-reinforced concrete beams fail in a manner similar to steel-reinforced beams. Figure 6 depicts the crack patterns of FRP-reinforced members at failure. With an increase in flexural load, the crack spacing of an FRP-reinforced beam is reduced, as shown in Figure 7, and its crack pattern stabilizes. The spacing was not influenced by FRP-reinforcement ratios, but was affected by FRP-bar types owing to bond characteristics (Masmoudi et al. 1996; Rafi et al. 2007). The reinforcement ratio may, however, control the development of crack width (Theriault and Benmokrane 1998). (a) (b) FIGURE 5 Load-deflection behavior of FRP-reinforced concrete beams in flexure: (a) CFRP bars (reproduced based on Rafi et al. 2007); (b) GFRP bars (reproduced based on Mias et al. 2013). (a) (b) (c) (d) FIGURE 6 Failure of steel- and FRP-reinforced concrete beams (reproduced based on Toutanji and Saafi 2000; Rafi et al. 2007; Smith et al. 2014; Tomlinson and Fam 2015): (a) steel; (b) CFRP; (c) GFRP; (d) BFRP.

23 Shear behavior Concrete beams reinforced with longitudinal FRP bars demonstrate shear behavior analogous to steel-reinforced beams (Pantelides et al. 2012a), except that the FRP-reinforced case includes reduced aggregate interlock and dowel action. This section is, therefore, devoted to FRP shear- reinforcement with a comparative discussion to steel stirrups. According to experimental inves- tigations, shear resistance provided by FRP stirrups is sufficient to induce the flexural failure of reinforced concrete beams (El-Sayed et al. 2007; Ahmed et al. 2010; Issa et al. 2016). This indicates that conventional steel stirrups can be replaced by FRP stirrups. Although the super- position of FRP stirrups and concrete components (Vf and Vc, respectively) in shear design is commonly accepted, it is important that attention be paid because of FRP’s limited stress redistribution characteristics (Stratford and Burgoyne 2002). Another concern is the premature failure of FRP stirrups at the bend region caused by stress concentrations (Ahmed et al. 2010; Issa et al. 2016). The majority of existing research is dedicated to GFRP and CFRP shear stirrups. Limited experimental data are currently available on BFRP stirrups (Tomlinson and Fam 2015; Issa et al. 2016). ACI 440.3R-12 (ACI 2012) provides two test methods for FRP shear stirrups: B5 (Test Method for Strength of FRP Bent Bars and Stirrups in Bend Locations) and B12 (Test Method for Determining Effect of Corner Radius on Tensile Strength of FRP Bars). The behavior of FRP stirrups was examined in experimental programs (El-Sayed et al. 2007; Ahmed et al. 2010). It is reported that the B5 test method showed higher capacity in comparison with the B12 method. The effect of bend-radius in FRP stirrups appears to be inconclusive (El-Sayed et al. 2007): some tests showed that an increase in bend-radius did not increase stirrups’ shear capacity; however, others exhibited an opposite result. The strength of FRP stirrups at the corners may be limited to 40%–45% of the longitudinal capacity of the FRP (Ahmed et al. 2010). The legs of FRP stirrups (vertical length) embedded in concrete are related to shear resistance. The shear capac- ity of FRP stirrups proportionally increases with an increase in embedment length from 5db to 20db (where db is the stirrup diameter), beyond which the stirrups may rupture and an abrupt loss of the beam’s load-carrying capacity is accompanied (El-Sayed et al. 2007). Figure 8 shows the response of concrete beams reinforced with FRP stirrups. Existing design expressions are gener- ally conservative against the experimental shear capacity of FRP-reinforced beams (Tomlinson and Fam 2015; Issa et al. 2016). Similar to the case of steel stirrups, the spacing of FRP stirrups is an important factor that controls stirrup stresses and crack spacing. In many cases (JSCE 1997; CSA 2014), the allowable crack width of FRP-reinforced concrete beams is between 0.02 in. and 0.028 in. (0.5 mm and 0.7 mm). Experi- mentally measured crack widths in concrete beams with CFRP stirrups were in a similar range [e.g., (b)(a) FIGURE 7 Average crack spacing under flexural load: (a) GFRP-reinforced beams (based on test data of Theriault and Benmokrane 1998); (b) CFRP-reinforced beams (based on test data of Rafi et al. 2007).

24 0.017 in. to 0.021 in. (0.43 mm to 0.54 mm) (El-Sayed et al. 2007)]. Several factors can influence the behavior of FRP shear-reinforced beams. Issa et al. (2016) reported that concrete beams with BFRP stirrups failed in shear compression and diagonal tension, when shear-span-to-depth (a/d ) ratios were 2.5 and 3.5, respectively. As the a/d ratio of beams with FRP stirrups increases, the extent of damage tends to be augmented with a reduced angle in a shear failure plane (Issa et al. 2016). The shear capacity of FRP-reinforced beams is proportional to flexural reinforcement ratios (Tomlinson and Fam 2015). The type of shear-reinforcement, however, may not affect the performance of these beams within a service load range (Tomlinson and Fam 2015). The diameter of stirrups is related to the so-called shear lag, which influences stress transfer between the fibers and the resin (El-Sayed et al. 2007). In addition to traditional vertical stirrups, nonconventional FRP shear reinforcement may be used. Lee et al. (2010) examined the feasibility of internal fiber-sheet stirrups as transverse reinforcement, which wrapped the longitudinal top and bottom bars of concrete beams. The behavior of the sheet stirrups was compared with that of steel and CFRP stirrups, resulting in similar shear resistance. Unlike the CFRP stirrups that failed at the corners owing to stress concentrations, the sheet stirrups ruptured along the side. According to the proposed perfor- mance index, the efficiency of the former was lower than the latter. Kim et al. (2015) tested concrete beams shear-reinforced with perforated GFRP plates instead of steel stirrups. As the ratio between GFRP-width and -spacing rose, shear resistance increased. The dominant failure mode of the beams along with GFRP-rupture was shear compression, depending upon the size of perforated openings. axial behavior The axial behavior of FRP-reinforced concrete columns is similar to that of steel-reinforced columns (De Luca et al. 2101; Afifi et al. 2014). The unique material properties of FRP bars, however, cause distinct column responses. FRP-reinforced columns generally exhibit lower load-carrying capacity than steel-reinforced ones. For instance, Alsayed et al. (1999) reported a 13% lower column capacity when steel reinforcing bars were replaced by GFRP bars. Longi- tudinal FRP bars contribute to the axial capacity of the columns by 5% to 10% (Tobbi et al. 2012; Afifi et al. 2014), which is lower than steel bars showing more than 10% of the capacity (De Luca et al. 2010). ACI 440.1R-15 (ACI 2015a), accordingly, states that the contribution of FRP bars is ignored when determining the capacity of an FRP-reinforced concrete column. Nonetheless, the strength reduction factor of 0.85 for concrete used in steel-reinforced columns may still apply to FRP-reinforced columns (Tobbi et al. 2012). This factor addresses a differ- ence in compressive strength between concrete cylinders and structural columns. Many factors (b)(a) FIGURE 8 Response of FRP stirrups in concrete beams: (a) effect of CFRP-stirrup spacing (reproduced after El-Sayed et al. 2007); (b) comparison of steel and CFRP stirrups (reproduced after Lee et al. 2010).

25 affect the compressive strength and modulus of FRP reinforcing bars; for example, fiber types, fiber-resin volume fraction ratios, manufacturing methods, and geometric configurations (Tobbi et al. 2012). The compressive strength of FRP bars is 10% to 50% of the tensile strength, whereas the elastic moduli in compression and tension are reasonably similar (Kobayashi and Fujisaki 1995; Deitz et al. 2003). Tobbi et al. (2012) proposed that the compressive strength of GFRP bars be 35% of the tensile strength. GFRP bars are predominantly employed for column appli- cation, rather than other FRP types. Hybrid reinforcing schemes with longitudinal steel bars and GFRP spirals (or GFRP longitudinal bars with steel spirals) can also be usable (Pantelides et al. 2013). Existing research programs concerning FRP-reinforced concrete columns include circular and square cross sections subjected to concentric and eccentric loadings (Kawaguchi 1993; Kobayashi and Fujisaki 1995; Lofty 2010; Tobbi et al. 2012). According to a comparative study, steel-reinforced columns under concentric loading show about 10% higher strength than GFRP-reinforced col- umns (Afifi et al. 2014). De Luca et al. (2010) reported that a stiffness change in GFRP-reinforced columns occurred at 60% of their axial capacity, which was lower than steel-reinforced cases showing a change at 80%. The axial behavior of FRP-reinforced columns is shown in Figure 9. The ductility of the columns increases with an increase in FRP-reinforcement ratio and volumet- ric ratio (Sharma et al. 2005; Lofty 2010). As the longitudinal reinforcement ratio of FRP bars increases, strains of the bars and spirals are reduced without a noticeable change in column capacity (Afifi et al. 2014). Various failure modes of FRP-reinforced columns are shown in Figure 10. FRP-reinforced columns fail the same as steel-reinforced columns in terms of crack development and cover spalling, followed by concrete crushing for short columns and buckling for slender columns (Hales et al. 2016). The premature failure of concrete cover can reduce the load-carrying capac- ity of FRP-reinforced columns (Tobbi et al. 2012). The failure of the columns is governed by either FRP-bar buckling or concrete crushing accompanied by tie/spiral-rupture, depending on the pitch of transverse reinforcement. Cover cracking was noticed once the columns reached peak loads (or close to the peak), along with abrupt concrete spalling (Tobbi et al. 2012; Afifi et al. 2014). By adjusting the configuration of GFRP reinforcement in concrete columns, ductile failure can take place. Afifi et al. (2014) reported that a concrete column [11.8 in. (300 mm) in diameter and 4.9 ft (1,500 mm) in height] reinforced with eight No. 5 GFRP longitudinal bars (reinforcement ratio = 1.1%) and GFRP spirals at a pitch of 3.1 in. (80 mm) failed in a ductile manner, including a ductility index of 2.0 (ductility index = ec85/ec1, where ec85 = axial strain at a load of 85% of the column’s post-peak capacity and ec1 = axial strain corresponding to the elastic limit). (b)(a) FIGURE 9 Response of FRP-reinforced columns in axial compression: (a) development of volumetric strain (reproduced based on De Luca et al. 2010); (b) load-strain (reproduced based on Prachasaree et al. 2015).

26 FRP and steel ties show intrinsically the same performance, as far as their contribution to column capacity is concerned (De Luca et al. 2010). FRP spirals or ties in concrete columns become acti- vated when the core concrete is damaged by axial loading in tandem with radial dilatation. Column strength is affected by the diameter of GFRP spirals [the smaller the diameter, the more rapid the strength decays (Afifi et al. 2014)]. The strength of the concrete can also influence the strain develop- ment of GFRP spirals (Hales et al. 2016). The shape of FRP spirals and ties does not affect the load- carrying capacity of the columns (if their pitches are the same), but influences post-peak responses (a) (b) (c) FIGURE 10 Failure modes of FRP-reinforced columns [Tobbi et al. 2012 (used by permission from American Concrete Institute)]: (a) failed columns; (b) buckling of GFRP bars; (c) rupture of GFRP ties.

27 such as deformability (Prachasaree et al. 2015). With an increase in the pitch of FRP spirals, the axial capacity of the columns decreases (Leung and Burgoyne 2001; Francis and Teng 2010; Tobbi et al. 2012). The reason is that tie or spiral spacing affects local buckling of the longitudinal bars. As FRP- tie spacing decreases, crack propagation is delayed (Tobbi et al. 2012). A smaller spiral diameter with a closer pitch may be used to enhance the ductility of FRP-reinforced concrete columns (Afifi et al. 2014). Because FRP bars, unlike steel bars, do not yield, premature buckling of the FRP bars is a concern for columns subjected to eccentric loading. Longitudinal FRP bars buckle when tie or spiral spacing is wide. Also, the spacing controls the cracking pattern of the columns [e.g., a tie spacing of 3 in. (75 mm) resulted in more stable crack progression relative to the case of a 12 in. (305 mm) spacing (De Luca et al. 2010)]. Stringent requirements for the pitch of ties or spirals are, therefore, crucial for FRP-reinforced columns. bond, Splice, and development length Investigations into the bond between FRP reinforcing bars and concrete have been actively conducted, which is salient to the understanding of the flexural behavior of FRP-reinforced concrete beams. At macro-scale, bond-slip responses are frequently examined using the push-out (or pull-out) test of FRP bars, as specified in ACI 440.3R-12 (ACI 2012). This method is easy to conduct; however, detailed bond failure mechanisms cannot be studied owing to the approximate nature (Guo and Cox 2000). To address this limitation, alternative smaller-scale approaches may be considered with a focus on mechanical interlocking between the concrete and the ribs of FRP bars (Yonezawa et al. 1993; Boothby et al. 1995; Pepe et al. 2013; Ju and Oh 2015). The surface treatment of FRP bars impacts the degree of mechanical interlocking. If smooth FRP bars are employed, bond between the bars and concrete is not reliant on a mechanical interlock (Rafi et al. 2007). Test results reveal that bond of FRP bars in concrete is generally lower than that of steel bars (Komiya et al. 1999; Thamrin and Kaku 2007). Insufficient embedment length of FRP bars outside structural supports may weaken the bond of the bars in the shear span of the beam, accompanying horizontal splitting cracks (Thamrin and Kaku 2007). Because FRP bars are primarily used to carry tensile load, specific research on the development length of FRP bars in compression is not available at this time. durability Various exposure conditions deteriorate FRP-reinforced concrete structures, such as wet-dry, saline, water-immersion, alkaline, chlorides, elevated temperatures, ultraviolet radiation, acid, and freeze– thaw (Karbhari et al. 2003; Ceroni et al. 2006). When evaluating the performance of GFRP-reinforced concrete members subjected to aggressive environments, accelerated durability tests are frequently used (Chen et al. 2007). This approach can save experimental effort, time, and expenses. High tem- peratures may be supplemented to accelerate the degradation process of an FRP–concrete interface. The premise in accelerated experimental investigations is that test protocols shorten the response time of materials and systems without altering their physical degradation mechanisms. As elaborated on in this section, the durability of FRP-reinforced concrete members is affected by (1) material- level degradation and (2) bond between the bars and concrete substrate. Chemical and mechanical bonds exist in the FRP–concrete interface. If bond failure occurs, stress transfer in the interface (necessary to carry structural load) becomes invalid. Concrete types and deterioration levels can influence the behavior of embedded FRP bars (Chen et al. 2007; Belarbi and Wang 2012). The properties of matrix resins, rather than those of high-strength fibers, in most cases dominate the durability of the bars. The durability of CFRP bars is generally supe- rior to that of GFRP bars (Karbhari et al. 2003). Chen et al. (2007) showed that the tensile strength of CFRP bars was degraded by 4% when exposed to sodium hydroxide (NaOH)-based chemicals for 70 days, whereas the strength of GFRP bars was reduced by 50% under the same test condition. Similar findings were reported by others (Davalos et al. 2008b; Belarbi and Wang 2012). Unlike aramid fibers (Gerritse 1993), carbon and glass fibers do not absorb water (Wolff and Miesseler 1993). Fatigue of FRP-reinforced concrete is of interest in highway bridge application. Although many laboratory tests were conducted (McBagonluri et al. 2000; El-Ragaby et al. 2007), technical information representing in situ conditions (e.g., variable fatigue load ranges) is still incomplete. When elevated temperatures are

28 expected, attention needs to be paid because the properties of FRP bars change (i.e., the strength and stiff- ness of FRP decrease if the glass transition temperature of the resin is exceeded). The Arrhenius equation (Robert et al. 2009; Zhou et al. 2011) and the Fick’s diffusion model (Tannous and Saadatmanesh 1998; Sen et al. 2002) often represent the deterioration process of FRP-reinforced concrete members. It is important to note that arbitrary extrapolation based on limited test data may cause erroneous prediction. The failure of conditioned FRP-reinforced concrete is caused by the initiation of damage at a weak- est link in the system and its progression with time. Micro-cracks in the resin of FRP bars can be a stress raiser, which causes a durability concern of FRP-reinforced concrete members. GFRP bars are dam- aged through leaching or etching (Chen et al. 2007); leaching is induced by the diffusion of alkali ions, whereas etching is related to hydroxyl ions. Hydrolysis degrades the performance of FRP-reinforced concrete (Chen et al. 2007). It is also known that chloride solutions decrease the strength and stiffness of concrete by increasing its porosity (Zhou et al. 2011). When steel-reinforced columns with GFRP spirals are subjected to corrosion, the steel bars may buckle, followed by the rupture of the spirals (Pantelides et al. 2013). The longevity of the GFRP–concrete interface is controlled by alkali and moisture diffusion. Alkaline solutions deteriorate GFRP bars with large pores on the surface (Robert et al. 2009). GFRP bars made of AR glass fibers show better resistance to an alkaline environment than those made of E-glass fibers (Benmokrane et al. 2002). It is worth noting that the alkali-induced deterioration of GFRP bars in constructed concrete structures may not be represented by accelerated laboratory tests (Mufti et al. 2005). Moisture diffusion alters various characteristics of FRP-reinforcing bars (e.g., physical, chemical, and mechanical properties). The diffusion of hydroxyl ions and water molecules can damage FRP bars, especially when polyester is used as a resin (Chen et al. 2007). As such, epoxy or vinylester resins (well- controlled cross-link density) are preferred in the manufacturing of FRP bars (Riffle et al. 1998; Benmokrane et al. 2002; Karbhari et al. 2003). Surface coating on FRP bars may be another option to retard the chemical diffusion rate (Benmokrane et al. 2002). Moisture diffusion rates arise as the GFRP- concrete interface is immersed in heated water. This circumstance accelerates strength degradation with- out changing the GFRP’s modulus, in conjunction with post-curing of the concrete (Robert et al. 2009). The bond of FRP bars in concrete may be deteriorated by environmental exposure. Belarbi and Wang (2012) reported three potential bond degradation mechanisms, depending on distress types: (1) micro-voids between the FRP bars and concrete that are vulnerable to freeze–thaw damage; (2) a difference in the coefficients of thermal expansion of the bars and concrete when temperature is asso- ciated, and (3) physical damage of FRP induced by chemical diffusion. Pull-out or push-out bond tests are widely used on account of experimental convenience, instead of beam-based bond tests. It can be noted, however, that the slip of reinforcing bars obtained from a pull-out bond test may be greater than the slip in a concrete beam, because flexural cracks in the beam can affect the load-slip behavior of the bars (Ferguson et al. 1965). Interfacial failure between the bar and the concrete ought to occur in a bond test (premature concrete failure sometimes takes place alongside stress concentra- tions). Small bar sizes appear to be more critical to bond degradation than large ones (Belarbi and Wang 2012). This experimental finding may be limited in practice, because the test specimens used by Belarbi and Wang (2012) had inconsistent cover thicknesses. The FRP–concrete interface is vulnerable to sustained loading, as reported by Benmokrane et al. (2002): interfacial failure was observed when loaded to 30% of the interfacial capacity for 30 days. Accordingly, FRP bars may not be loaded higher than a threshold stress level. Thermal mismatching between the FRP bars and concrete resulting from their different coefficients of thermal expansion can cause cracking or spalling of the concrete (Ceroni et al. 2006). Cold temperature (e.g., subzero) results in the hardening and micro-cracking of matrix resins (Karbhari et al. 2003). Freeze-thaw conditioning often increases the strength of FRP-reinforced concrete beams owing to the additional curing of the concrete (Laoubi et al. 2006). Although ultraviolet radiation is deleterious to FRP bars, it may not be a concern for FRP-reinforced concrete (concrete protects the embedded FRP bars). The load-bearing capacity of steel bars (when embedded in concrete) is better than the capacity of GFRP bars in an acidic environment (Zhou et al. 2011). This is attributed to the difference in surface texture (i.e., ribbed steel bars provided a favorable mechanical interlock). Difficulties arise when in situ durability is examined; the reason being that several parameters (e.g., moisture, chlorides, and alkali) are simultaneously mixed and contribute to the degradation of structural

29 members. Mufti et al. (2005) reported on the in situ performance of five GFRP-reinforced concrete structures in Canada (5–8 years of service): the Hall’s Harbor Wharf, Joffre Bridge, Chatham Bridge, Crowchild Trail Bridge, and Waterloo Creek Bridge. The GFRP bars were made of E-glass fibers and a vinylester resin. The structures were cored and examined by an optical microscope, differential scan- ning calorimetry, scanning electron microscopy, infrared spectroscopy, and energy dispersive x-ray. According to the optical microscopy, the interface between the GFRP and concrete was intact (i.e., there was no evidence of bond degradation, although the structures were exposed to a number of envi- ronmental conditions such as freeze–thaw, wet-dry, alkalinity, and deicing salts). The GFRP bars were not damaged either, based on the observation of the scanning electron microscopy. The infrared spec- troscopy revealed that hydrolysis reactions did not occur in the bars. In accordance with the differential scanning calorimetry, there was no change in the glass transition temperature of the GFRP. The study concluded with the explanation of the differences between actual deterioration on site and accelerated laboratory tests: constructed structures experience variable stress states at reasonable temperatures, unlike laboratory specimens subjected to constant stress levels at high temperatures. design aspects General As is the case for traditional reinforced concrete, FRP-reinforced concrete members are designed by means of the LRFD method. Two limit states (ultimate and serviceability) are checked to ensure the safety and functionality of the members. Conforming to most reliability-based design specifications (e.g., AASHTO LRFD BDS Bridge Design Specifications), a safety index of 3.5 to 4.0 is taken when calibrating the load and resistance factors for FRP-reinforced concrete structures ( fib 2007; ACI 2015a). Existing design guidelines and specifications for FRP-reinforced concrete are dedicated to flexural and shear aspects. Design guidelines for FRP-reinforced concrete columns are not available at this time. Readers are expected to have sufficient knowledge of the design of reinforced concrete or prestressed concrete structures. As such, fundamental design contents associated with those struc- tures are not expounded. The focus of this section is on the use of FRP composites, when reinforcing various structural members, including their unique behavior in comparison with steel reinforcement. The physical and mechanical properties of FRP bars are different from those of steel reinforcing bars (e.g., there is no yield plateau in FRP bars), as discussed earlier in this chapter (“FRP-reinforcing Bars”). Conventional design approaches that were developed based on steel bars, therefore, may not be suitable for FRP reinforcement. The engineers are encouraged to understand the application range of FRP bars and corresponding limitations in order to adequately design an FRP-reinforced structural concrete. Typical assumptions made in designing FRP-reinforced concrete are as follows (AASHTO 2009; ACI 2015a): • Plane sections remain plane, as a member is loaded in flexure. • The interface between FRP-reinforcement and concrete is perfectly bonded. • The stresses of FRP-reinforcement and concrete are attained from corresponding strains based on the stress-strain relationships. • The tensile contribution of concrete is ignored when flexural capacity is calculated. • FRP reinforcement behaves linear elastically until rupture occurs. • The maximum usable strain of concrete is 0.003 in compression, and the concrete cracks when the tensile strain reaches the modulus of rupture. • The nonlinear compressive stress profile of concrete is simplified with the equivalent rectangular stress block. Because FRP bars do not yield and accompany more uncertainty in comparison with steel bars, strength reduction factors for FRP-reinforced concrete members are lower than those of steel- reinforced members, as shown in Table 7. A transition between compression- and tension-controlled sections occurs in FRP-reinforced concrete beams, depending on reinforcement ratios. The upper bound of the transition zone may be 1.4rb, where rb is the reinforcement ratio of a balanced section (AASHTO 2009; ACI 2015a). Recent statistics-based research reveals a similar upper bound of the transition zone of 1.5rb (Xue et al. 2016).

30 The manufacturer-reported strength of FRP reinforcement may be reduced by environmental reduction factors, given that several external attributes can degrade the long-term performance of FRP bars (e.g., creep, fatigue, acid, alkali, and moisture). Table 8 lists environmental reduction factors for FRP bars in concrete application (the factors for BFRP are not yet available). The modulus of FRP bars may not be reduced because of environmental exposure (Tomosawa and Nakatsuji 1996, 1997). Flexural Design Depending on the amount of FRP-reinforcement in concrete, three section types are available in flexure: • Balanced section (balanced failure): FRP-reinforcement ruptures and concrete crushes at the same time. • Tension-controlled section (tension failure): FRP-reinforcement ruptures before concrete crushes. It is also known as an under-reinforced section. • Compression-controlled section (compression failure): FRP-reinforcement ruptures after concrete crushes. This case is referred to as an over-reinforced condition. Unlike steel-reinforced concrete, FRP-reinforced concrete may be designed to have a compression- controlled section: FRP-rupture occurs without warning (no yield plateau exists). If an FRP-reinforced concrete member fails in tension (under-reinforced), the compressive strain of the concrete is below the crushing strain of 0.003. In this case, according to force equilibrium and strain compatibility, the equivalent stress block of the concrete in compression needs to be adjusted. Minimum FRP reinforcement is required to avoid the premature failure of an FRP-reinforced concrete beam, when the concrete cracks. This design check is appropriate for a tension-controlled section, where the amount of FRP-reinforcement is less than the amount of the balanced section. Design guides specify the minimum reinforcement requirement: A f b dff c w fu ( ) ( )= ′max 0.16 ;0.33 in U.S. customary units AASHTO 2009 (1),min A ff b d f b df c fu w fu w ( )= ′ ≥4.9 330 in U.S. customary units ACI 2015a (2),min where Af,min is the minimum area of FRP reinforcement needed to prevent the failure of flexural members upon cracking (in.2); f ′c is the specified compressive strength of the concrete (ksi in AASHTO and psi in ACI); ffu is the design tensile strength of the FRP (ksi in AASHTO and psi in ACI); bw is the width of the web (in.); and d is the distance from extreme compression fiber to the centroid of the tension reinforce- Steel-Reinforced Concrete (AASHTO 2015) FRP-Reinforced Concrete (AASHTO 2009) Flexure (tension-controlled) 0.9 0.55 Flexure (compression-controlled) 0.75 0.65 Shear 0.9 N/A N/A = not applicable TABLE 7 COMPARISON OF STRENGTH REDUCTION FACTORS Exposure Type Reduction Factor ACI 440.1R-15 (ACI 2015a) AASHTO (2009) No exposure to earth and weather CFRP 1.0 N/A GFRP 0.8 0.80 Exposure to earth and weather CFRP 0.9 N/A GFRP 0.7 0.70 N/A = not applicable TABLE 8 ENVIRONMENTAL REDUCTION FACTORS

31 ment (in.). The cracking load or moment of an FRP-reinforced section is calculated using elastic beam theory, under the condition that the maximum tensile stress of the section reaches the modulus of rupture. If multiple FRP-reinforcement layers are designed in an under-reinforced concrete section, the sum- mation of all reinforcement areas to determine the flexural capacity of the section is not valid. Contrary to the constant force of yielded steel bars, FRP bars do not yield, and their tensile forces are a function of the linear strain profile across the beam section. The progressive rupture of the multiple-layered FRP bars is, therefore, taken into consideration. This statement also applies to a hybrid-reinforcing scheme (e.g., a combination of GFRP and CFRP bars). When a doubly reinforced beam is designed, the com- pression FRP bars are generally ignored to achieve a conservative design. The moment redistribution mechanism induced by the yielding of steel reinforcement does not occur in the FRP system. The serviceability of FRP-reinforced concrete members demands particular attention, because of the relatively low modulus of FRP bars compared with that of steel bars. Consequently, instead of strength, serviceability may control the design of the members. Service load levels need to be examined in FRP- reinforced concrete members to preclude potential creep-rupture failure, which is irrelevant to steel- reinforced concrete structures. ACI 440.1R-15 (ACI 2015a) provides the following maximum stress levels: 0.20ffu and 0.55ffu for the GFRP and CFRP reinforcing bars, respectively, where ffu is the design tensile strength of the bar. The same limit of 0.20ffu for GFRP bars is specified in the AASHTO guide specifications (AASHTO 2009). FRP-reinforced concrete beams and slabs show more deflection than their steel-reinforced counterparts. A stringent deflection check is, therefore, required to avoid ser- viceability problems. Various expressions are specified in design guidelines to calculate a member’s deflection, as summarized in Table 9. The crack control of FRP-reinforced concrete is intended to satisfy aesthetic requirements, rather than to alleviate the risk of corrosion damage (FRP is a non- corrosive material). A collation of crack control methods for FRP-reinforced concrete structures is provided in Table 10. For the reason that FRP bars do not yield, the traditional concept of ductility does not apply to FRP-reinforced concrete (the ductility of a member is quantified by a ratio between the responses at ultimate and at yield). Deformability may be used as an alternative to assess the flexural characteristics of FRP-reinforced concrete members. The Canadian highway bridge design code (CSA S6) stipulates the following equation: J M M ult ult c c = ψ ψ (3) where J is the deformability index (which shall be greater than 4.0 and 6.0 for rectangular and T-sections, respectively); Mult and Mc are the ultimate moment capacity and the moment corresponding to a maxi- mum compressive concrete strain of 0.001 in the section, respectively; and yult and yc are the curva- ture at the ultimate (Mult) and service (Mc) moments, respectively. Other design guidelines such as fib Bulletin 40 and ACI 440.1R-15 concisely discussed deformability in FRP-reinforced concrete; how- ever, detailed design equations are not presented. The cover of FRP-reinforced concrete members plays an important role in protecting the FRP from various detrimental attributes such as fire, chemicals, and moisture. Some design guidelines explicitly state cover requirements for FRP bars, as listed in Table 11. Spacing requirements for FRP bars are basically the same as those for steel bars (ACI 2015a). AASHTO (2009) specifies that the spacing of GFRP bars shall not exceed 1.5 times the member thickness or 18 in. (457 mm). Because limited research has been done on the shrinkage and temperature reinforcement in FRP-reinforced concrete members, ACI 440.1R-15 (ACI 2015a) provides an FRP-reinforcement ratio, which is scaled from the shrinkage and temperature reinforcement of conventional steel using the modulus and ultimate strength of FRP bars. The surface treatment techniques of FRP bars depend on manufacturers. For this reason, their bond characteristics to concrete cannot be represented by a single expression. Table 12 summarizes development length equations available in various design guidelines. Shear Design In the shear design of FRP-reinforced concrete beams, either conventional steel or FRP stirrups can be used. If steel stirrups are employed, design articles developed for steel-reinforced concrete are exploited. When FRP shear stirrups are utilized, the strength of bar-corners is lower than that of straight bars. FRP- reinforced concrete decks may not require shear reinforcement, similar to steel-reinforced concrete decks,

32 dohteM ecruoS AASHTO guide specifications (AASHTO 2009) Use the effective moment of inertia (Ie): gcr a cr gd a cr e IIM MI M MI ≤−+= 33 1β 0.1 5 1 ≤= fb f d ρ where Icr = moment of inertia of transformed cracked section Ig = gross moment of inertia Mcr = cracking moment Ma = maximum service load moment in member f = GFRP reinforcement ratio fb = GFRP reinforcement ratio producing balanced strain conditions. ACI 440.1R-15 (ACI 2015a) Use the effective moment of inertia (Ie): g g cr a cr cr e I I I M M II ≤ −− = 11 2 γ where Icr = moment of inertia of transformed cracked section Ig = gross moment of inertia = parameter to account for the variation in stiffness along the length of the member ( ( )acr MM /72.072.1 −= ). Canadian Highway Bridge Design Code (CSA 2014) Use the effective moment of inertia (Ie): ( ) g a cr crgcre IM MIIII ≤−+= 3 Intelligent Sensing for Innovative Structures (ISIS Canada 2007) Use the effective moment of inertia (Ie): ( )crg a cr cr crg e II M MI II I −−+ = 2 5.01 fib Bulletin No. 40 (fib 2007) ( )ξδ −+= 112 m cr M M −= max 1 β (based on Eurocode 2 and Model Code 1990) where = deflection 1 and 2 = deflections calculated assuming constant uncracked and cracked sectional moments of inertia along the member Mcr = applied moment causing the occurrence of the first crack Mmax = maximum bending moment under service loads = 1 and 0.8 as per Eurocode 2 and Model Code 1990, respectively m = 2 and 1 as per Eurocode 2 and Model Code 1990, respectively. β ρ ρ ρ γ γ δ δ ξ ξ δ δ δ β TABLE 9 DEFLECTION OF FRP-REINFORCED CONCRETE MEMBERS

33 dohteM ecruoS AASHTO guide specifications (AASHTO 2009) 4 2 2 2, sdk E f w cb f sf += β where (U.S. customary units) =w maximum crack width sff , = tensile strength in GFRP reinforcement at the service limit state (ksi) fE = modulus of elasticity of FRP = ratio of distance between neutral axis and tension face to distance between neutral axis and centroid of reinforcement bk = coefficient that accounts for the degree of bond between GFRP bars and surrounding concrete cd = thickness of concrete cover measured from extreme tension fiber to center of the flexural reinforcement located closest thereto s = spacing of GFRP flexural reinforcement. ACI 440.1R-15 (ACI 2015a) bfs f c bfs f kf wE c kf wE s 92.05.215.1max ≤−= f bfsc E kfd w 2 ≥ where (U.S. customary units) smax = maximum permissible center-to-center bar spacing for flexural crack control Ef = design or guaranteed modulus of elasticity of FRP defined as mean modulus of sample of test specimens ffs = stress level induced in FRP at service loads kb = bond-dependent coefficient (kb = 1.4 if a bond condition is unknown) cc = clear cover. Canadian Highway Bridge Design Code (CSA 2014) 2 2 1 2 2 2 += sdk h h E f w cb FRP FRP cr where (metric units) wcr = crack width (less than 0.5 mm for members subjected to aggressive environments and 0.7 mm for other members) fFRP = stress in the tension FRP reinforcement h1 = distance from the centroid of tension reinforcement to neutral axis h2 = distance from the centroid of the extreme flexural tension surface to neutral axis kb = bond parameter (kb = 0.8 and 1.0 for sand-coated and for deformed bars, respectively). Intelligent Sensing for Innovative Structures (ISIS Canada 2007) ( ) 3/1 1 22.2 Ad h h E f kw c frp frp b= where (metric units) w = crack width at the tensile face of the beam [less than 0.5 mm (0.02 in.)] kb = bond-dependent coefficient (kb = 1.2 if a bond condition is unknown) ffrp = stress in the tension FRP reinforcement at location of the crack h2 = distance from the extreme tension surface to the neutral axis h1 = distance from the centroid of tension reinforcement to the neutral axis dc = concrete cover measured from the centroid of tension reinforcement to the effective tension surface A = effective tension area of concrete surrounding the flexural tension reinforcement and having the same centroid as that reinforcement, divided by the number of rebars. β β TABLE 10 CRACK CONTROL METHODS COMPILED FROM DESIGN GUIDES (continued on next page)

TABLE 10 (continued) dohteM ecruoS fib Bulletin No. 40 (fib 2007) smrmcr sw ε= (based on Eurocode 2) r rm dkks ρ21 25.050 += ( )[ ] sssrssm E//1 221σ −= where (metric units) wcr = crack width = 1.3 k1 = 0.8 and 1.6 for high-bond and plain bars, respectively k2 = 0.5 and 1.0 for bending and pure tension, respectively d = bar diameter r = reinforcement ratio within the effective tension area 1 = 1.0 and 0.5 for high-bond and plain bars, respectively srm = average final crack spacing s = stress in the tension reinforcement in a cracked section sr = stress in tension reinforcement under first crack load Es = modulus of elasticity. Japan Society of Civil Engineers (JSCE 1997) ([ )] +−+= 'max 7.04 csd f fe fb E dcckw where (metric units) maxw = maximum crack width kb = FRP-concrete bond quality coefficient (kb = 1.0 if FRP and steel bars have similar bond characteristics) c = concrete cover to the center of the tension reinforcement cf = center-to-center distance between rebars d = bar diameter fe = stress increase in the reinforcement ' csd = compressive strain due to the effects of creep and shrinkage. β β β β ε σ σ ρ β σ σ σ ε ε σ Source Specified Cover AASHTO guide specifications (AASHTO 2009) • Slabs Top and bottom bars (No. 10 and smaller): 0.75 in. (19 mm). • Beams Stirrups: 1.5 in. (38 mm) Principal reinforcement: 2 in. (50 mm). ACI 440.5R-08 (ACI 2008) • Slabs and joints Top and bottom bars (No. 10 or smaller): 0.75 in. (19 mm). • Beams Primary reinforcement: 2 in. (50 mm) Stirrups, spirals, and ties: 1.5 in. (38 mm). • Walls No. 10 or smaller: 0.75 in. (19 mm) Exposed to earth: 2 in. (50 mm). • Footings Contact with earth: 3 in. (75 mm) Over top of piles: 2 in. (50 mm). Canadian Highway Bridge Design Code (CSA 2014) 1.4 ± 0.4 in. (35 ± 10 mm) Intelligent Sensing for Innovative Structures (ISIS Canada 2007) • External exposure Beam: 2.5db or 2 in. (50 mm) Slab: 2.5db or 1.2 in. (30 mm) where db = bar diameter TABLE 11 CLEAR COVER REqUIREMENTS FOR FRP-REINFORCED CONCRETE

35 Source Development Length (ld) AASHTO guide specifications (AASHTO 2009) b b c f d d d C f f l + − = 6.13 340 ' α where (U.S. customary units) = bar location modification factor [ = 1.0 except for bars with more than 12 in. (305 mm) of concrete cast below for which a value of 1.5 shall be adopted] ff = effective strength in reinforcement C = lesser of the cover to the center of the bar or one- half of the center-to-center spacing of the bars being developed bd = bar diameter. ACI 440.1R-15 (ACI 2015a) b b c fr d d d C f f l + − = 6.13 340 ' where (U.S. customary units) = top bar modification factor (default: = 1) frf = required bar stress ' cf = specified compressive strength of concrete C = spacing or cover dimension. Canadian Highway Bridge Design Code (CSA 2014) Af Ff E EKd kkl cr pu s FRP trcs d + = 4145.0 where (metric units) k1 = bar location factor (1.0 or 1.3) k4 = bar surface factor (ratio of the bond strength of the FRP bar to that of a steel deformed bar having the same cross-sectional area as the FRP bar) dcs = the smaller of either the distance from the closest concrete surface to the center of the bar bending developed, or two-thirds the center-to-center spacing of the bar being developed Ktr = transverse reinforcement index EFRP = elastic modulus of FRP Es = elastic modulus of steel bar F = FRP factor (CFRP: 0.9–1.0; GFRP 0.8–1.0) fpu = tensile strength of steel fcr = cracking strength of concrete A = area of that part of the cross section between the flexural tension face and the centroid of the gross section. α α α α α TABLE 12 DEVELOPMENT LENGTH OF FRP BARS EMBEDDED IN CONCRETE (continued on next page)

36 and conventional theory (e.g., modified compression field theory) is applicable (Liu and Pantelides 2012). FRP stirrups may be produced in closed or U-shaped forms. Some commercially available FRP stirrups are shown in Figure 11. It is important to note again that FRP stirrups cannot be bent on site, contrary to steel stirrups. FRP-reinforced concrete beams demand design approaches different from steel-reinforced con- crete beams. The reason is that the former reveals (1) wider shear cracks (owing to the low modulus of FRP bars), (2) insignificant dowel action of the FRP bars, and (3) the absence of a yield plateau in FRP stirrups. Further, the neutral axis depth of an FRP-reinforced concrete beam at the ultimate state is generally lower than that of its steel counterpart, because of the high tensile strength of the longi- tudinal FRP bars. This load-carrying mechanism reduces shear resistance in the uncracked portion of the concrete above the beam’s neutral axis. The shear strength of an FRP-reinforced concrete beam (Vn) consists of the resistance of concrete (Vc) and the resistance of steel and FRP stirrups (Vf): V V Vn c f= + (4) Owing to the wide crack width of the FRP-reinforced beam, the aggregate interlock of concrete may be negligible. The individual shear-carrying components shown in Eq. 4 vary by design guides and specifications, as listed in Table 13. The strain of FRP stirrups is controlled to limit the width of shear cracks and to avoid the creep-rupture of the stirrups. The maximum usable stirrup strains specified in ACI 440.1R-15 (ACI 2015a) and CHBDC (CSA 2014) are 0.004 and 0.002 in., respectively. Miscella- neous details on FRP stirrups (e.g., minimum shear reinforcement and maximum stirrup spacing) may be the same as those of steel stirrups (ACI 2015a). The punching shear resistance of FRP-reinforced concrete slabs may be determined in a way similar to the resistance of steel-reinforced slabs, using a modified prefactor. AASHTO (2009) and ACI 440.1R-15 (ACI 2015a) provide V f b kdc c ( )= ′10 0 and V f b cc c= ′0.32 0 in U.S. customary units, respectively, where Vc is the punching shear capacity of the slab; b0 is the perimeter of the slab’s critical section; k is the ratio of the neutral axis depth to the reinforce- ment depth; d is the effective depth of the slab; and c is the distance from the extreme compression fiber to the neutral axis. Specific punching shear expressions are not available in other design documents (e.g., Design Manual No. 3 of Intelligent Sensing for Innovative Structures, ISIS Canada 2007). Site application FRP-reinforced concrete bridges have been constructed all around the world. Nanni (2001) sum- marizes field applications of FRP-based bridges in the United States. The first Canadian bridge deck Source Development Length (ld) Intelligent Sensing for Innovative Structures (ISIS Canada 2007) frp cr frp s frp trcs d Af f E E Kd kkl + = 4145.0 where (metric units) k1 = bar location factor (1.0 or 1.3) k4 = bar surface factor (0.8 if test data are not available) dcs = the smaller of the distance from the closest concrete surface to the center of the bar bending developed, or two-thirds of the center-to-center spacing of the bar being developed Ktr = transverse reinforcement index Efrp = elastic modulus of FRP Es = elastic modulus of steel bar ffrp = stress in FRP reinforcement fcr = flexural cracking strength of concrete Afrp = area of FRP bar. TABLE 12 (continued)

(a) (b) (c) FIGURE 11 FRP stirrups (used by permission from American Concrete Institute): (a) C-shape (Tobbi et al. 2014); (b) load test (Nanni and Faza 2002); (c) corner rupture (Nanni and Faza 2002).

38 Source Component AASHTO guide specifications (AASHTO 2009) cbfcbfV cwcc 0 '' 32.016.0 ≤= s dfA V fvfvf = where (U.S. customary units) ' cf = compressive strength of concrete wb = width of web c = distance from extreme compression fiber to neutral axis 0b = perimeter of critical section computed at d/2 away from the concentrated load fvA = amount of FRP shear reinforcement within spacing s fvf = design tensile strength for shear = fbf fE ≤004.0 (ksi) fbf = strength of the bent portion of a GFRP bar. ACI 440.1R-15 (ACI 2015a) ( )kdbfV wcc '5= ( )θ cossin += s dfA V fvfvf where (U.S. customary units) wb = width of beam k = ratio of depth of neutral axis to reinforcement depth d = distance from extreme compression fiber to centroid of tension reinforcement fvf = tensile strength of FRP taken as smallest of design tensile strength, strength of bent portion of FRP stirrups, or stress corresponding to 0.004Efrp s = stirrup spacing θ = angle of inclination of stirrups. Canadian Highway Bridge Design Code (CSA 2014) vvcrccf dbfV βφ5.2= ( ) s dAV vvvFRPsf ασ cotcot + = vvFRPv E ε= or 3.1/3.005.0 pu s v fd r += 002.0210001.0 ' 2/1 ' ≤+= c N vFRPvFRP fus cv fEp Epf θ φ θ σ σ σ ε where (metric units) cfV = factored shear resistance provided by tensile stresses in the concrete β = factor for the shear resistance of cracked concrete c = resistance factor for concrete (0.75) crf = cracking strength of concrete vb = effective web width within depth dv vd = effective shear depth v = strain in an FRP stirrup sfV = factored shear resistance provided by shear reinforcement FRPφ = resistance factor for FRP vA = area of transverse shear reinforcement perpendicular to the axis of a member = angle of inclination of stirrups φ ε θ TABLE 13 SHEAR DESIGN COMPONENTS OF FRP-REINFORCED CONCRETE

39 fully reinforced with GFRP bars, the Val-Alain Bridge, was constructed in 2004 (Benmokrane et al. 2007) and designed in conformance with the Canadian Highway Bridge Design Code (CSA S6). The slab-on-girder bridge (skew angle of 20°) has a single span of 164 ft (49.89 m), with a width of 41 ft (12.57 m). High-performance concrete with a compressive strength of 7,300 psi (50 MPa) was used with two sizes of sand-coated GFRP bars [0.75 in. (19.1 mm) and 0.63 in. (15.9 mm) in diameter]. The reinforcing bars, made of E-glass fibers and a vinylester resin at a fiber volume of 73%, had a guaranteed tensile strength of 94 ksi (650 MPa) and an elastic modulus of 6,100 ksi (42 GPa). The lightweight GFRP bars facilitated the bridge construction. The splice of the bars was not necessary in the transverse direction of the deck, whereas three splices were required at 40 times the bar diameter [31 in. (800 mm)] in the longitudinal direction. Upon completion of the construction, a proof-load test was conducted using two four-axle trucks [73.7 kips (328 kN) and 72.8 kips (324 kN)] to monitor the bridge’s static and dynamic responses. Good agreement was made between the in situ responses and analytically predicted data. Source Component = angle of inclination of shear reinforcement to the longitudinal axis of a member s = stirrup spacing vFRPE = elastic modulus of FRP stirrups r = radius of curvature of the bend of an FRP stirrup sd = diameter of an FRP stirrup puf = specified tensile strength of an FRP bar sp = ratio of the cross-sectional area of the longitudinal FRP reinforcement to the effective cross-sectional area of the beam fuE = elastic modulus of longitudinal FRP bars vFRPp = ratio of the total cross-sectional area of the legs of an FRP stirrup to the product of the width of the beam and the spacing of stirrups N = stress in concrete due to axial loads. Intelligent Sensing for Innovative Structures (ISIS Canada 2007) s frp wcccf E E dbfV '2.0 λφ= s dA V vvfrpvfrpfrp cot = 5.1/3.005.0 frpv s b v fd r += vfrpvv E= or where (metric units) cfV = factored shear resistance attributed to concrete = modification factor for density of concrete c = resistance factor for concrete ' cf = specified concrete strength in compression wb = minimum effective web width within depth d d = distance from the extreme compression fiber to the centroid of the reinforcement frpE = elastic modulus of flexural FRP reinforcement sE = elastic modulus of steel frpvE = elastic modulus of FRP stirrup v = limiting strain in FRP stirrup br = radius of bend sd = diameter of FRP stirrup frpvf = limiting value of tensile strength of stirrup. α σ σ σ σ θφ ε φ λ ε TABLE 13 (continued)

40 Benmokrane et al. (2004) reported case studies based on four bridge decks reinforced with FRP bars: the Joffre, Wotton, Magog, and Morristown bridges. Various combinations of GFRP, CFRP, and steel bars were employed for the top and bottom reinforcement of the decks [thicknesses varied from 7.9 in. (200 mm) to 10.2 in. (260 mm)]. The average tensile strengths of the GFRP and CFRP bars were 92 ksi (634 MPa) and 223 ksi (1,536 MPa), respectively. The bridges were designed according to the AASHTO Standard Specifications and ACI guidelines (ACI 440.1R). Figure 12 shows the GFRP reinforcing schemes of the Magog and Morristown bridges. Field testing was conducted to measure the strains of the GFRP bars. FRP-reinforced concrete panels [each 2.8 ft (860 mm) wide and 1 ft (300 mm) deep] were used for the Walters Street Bridge (Nystorm et al. 2002) that was designed pursuant to ACI 440.1R and the AASHTO Standard Specifications. HS15-44 truck loading was the live load considered on the 24 ft (7.3 m) bridge. The concrete panels were reinforced with CFRP and GFRP bars, as shown in Figure 13. The CFRP bars were made of carbon fibers and an epoxy resin, whereas the GFRP bars comprised E-glass fibers and a vinylester resin. Each panel was connected by steel angles welded together (side-by-side) and was anchored onto abutments. Guardrails were then installed using high-strength bolts along the exterior panels. The total construction time spent for the bridge installa- tion was 10 days. Such an FRP-reinforced panel system can be produced with lightweight concrete (Pantelides et al. 2012b). A corrosion-damaged box culvert bridge was replaced by a GFRP-reinforced (a) (b) FIGURE 12 GFRP reinforcement for bridge decks [Benmokrane et al. 2004 (used by permission from American Concrete Institute)]: (a) Magog Bridge; (b) Morristown Bridge.

41 (a) (b) FIGURE 13 FRP-reinforced concrete panels for the Walters Street Bridge [Nystorm et al. 2002 (used by permission from Antonio Nanni)]: (a) reinforcement cage; (b) installed panels. culvert (Alkhrdaji and Nanni 2001), as shown in Figure 14(a). To ensure the performance of the GFRP- based culvert, two segments were loaded to failure in a laboratory [Figure 14(b)]. The behavior of the cul- vert was controlled by stiffness degradation at midspan, which led to noticeable cracks. Field construction included 18 culvert boxes in two rows [Figure 14(c)]. Matta et al. (2007) showed FRP-reinforced concrete railings for Bridge 14802301 in Greene County, Missouri. The original bridge was 70 years old, and was replaced by a new one (Figure 15). The railing was designed to carry 10 kips (44.5 kN) with a failure mode of diagonal tension at the corner. Deitz et al. (2004) tested the impact resistance of GFRP-reinforced rail- ings and found that the members’ failure load was 2.9 times its design service load. Static and dynamic tests are frequently conducted to examine the in situ response of bridge decks reinforced with GFRP bars. Shekar et al. (2003) reported that the deflection of a GFRP-reinforced concrete bridge was lower than the AASHTO deflection limit [e.g., a recorded deflection of 0.36 in. (9.9 mm) versus the limit of 2.2 in. (54.8 mm)]. Benmokrane et al. (2006) installed fiber optic sensors to measure the strain of the Morristown Bridge subjected to known-weight trucks [65 kips (287 kN) and 72 kips (317 kN)] at a sampling rate of 10 readings/s. The measured strains were converted to load distribution factors. The AASHTO LRFD BDS distribution equations were generally conserva- tive in comparison with the in situ distribution factors. Holden et al. (2014) carried out long-term response monitoring for bridge deck panels reinforced with GFRP bars including deflection, strain, and temperature variations. Finite element models were developed to predict the deck behavior. The performance of the GFRP-reinforced deck (3 years in service) was satisfactory in meeting design requirements. A follow-up monitoring program was concerned with the bridge deck’s dynamic load allowance (Holden et al. 2015). The measured maximum dynamic load allowance of 0.16 was lower than that of the AASHTO LRFD BDS (AASHTO 2012b). The effect of temperature on the behavior of FRP-reinforced concrete bridges appears to be site-specific. Shekar et al. (2003) stated that a tem- perature gradient caused the Mckinnleyville Bridge in West Virginia to crack, whereas Frosch et al. (2006) claimed that temperature did not influence the response of the Thayer Road Bridge in Indiana. fiber-reinforced Polymer-PreStreSSed concrete memberS Instead of conventional seven-wire steel strands, FRP tendons may be used for prestressed concrete members. An outline of FRP-prestressed concrete is discussed in this section, including materials, structural behavior, design aspects, and site demonstration projects along with various members such as pretensioned and post-tensioned girders and decks.

(a) (c) (b) FIGURE 14 GFRP-reinforced box culvert bridge [Alkhrdaji and Nanni 2001 (used by permission from Antonio Nanni)]: (a) reinforcing scheme; (b) load test; (c) constructed culvert.

43 (a) (b) (c) FIGURE 15 GFRP-reinforced railing [Matta et al. 2007 (used by permission from Antonio Nanni)]: (a) 70-year-old bridge in Missouri; (b) GFRP cage for railing; (c) new bridge constricted.

44 frP tendons Various FRP tendons are available in terms of material composition and geometric properties. Carbon fibers are most commonly used, followed by aramid fibers. Because glass fibers are known to be sus- ceptible to creep-rupture, GFRP tendons are not preferred in prestressed concrete application. BFRP may have potential for prestressed concrete. A recent experimental study (Wang et al. 2014) revealed that the creep rate of BFRP tendons was not significant (i.e., a stress loss of less than 3.7% under a sustained load level of 60% of the BFRP’s ultimate strength). Table 14 lists the mechanical proper- ties of typical FRP tendons. Two types of CFRP tendons are available: polyacrylonitrile (PAN) and pitch-based CFRPs. PAN CFRP tendons have a shape similar to seven-wire steel strands, because they are made of twisted carbon fiber yarns. Pitch-based CFRP produced by pultrusion has plain or deformed rod shapes (indented surface profiles are associated to improve the bond against concrete), as shown in Figure 16. The mechanical properties of CFRP tendons are generally superior to those of AFRP tendons, except toughness and impact resistance (ACI 2004). Several shapes of AFRP tendons are manufactured, such as deformed or spirally wounded rods, rectangular or circular rods, and braided cables. flexural behavior The flexural behavior of prestressed concrete beams with bonded FRP tendons is similar to the behav- ior of beams with steel strands, except for the absence of a yield plateau. The capacity of concrete beams prestressed with either FRP or steel is not influenced by prestressing levels (Saiedi et al. 2011). On account of FRP’s lower elastic modulus, FRP-prestressed concrete beams reveal insignificant long-term prestress losses compared with steel-prestressed beams (Youakim and Karbhari 2007). Although FRP-prestressed concrete beams do not exhibit a yield plateau under flexural loading, their sufficiently developed curvature along with extensive cracking can provide a warning of impending failure. FRP-prestressed concrete girders show full composite action with the deck slab (Fam et al. 1997). Figure 17 depicts the load-deflection behavior of FRP-prestressed concrete beams. Prior to cracking of the beams, the effect of tendon types (both steel and FRP) is negligible [Figure 17(a)], because concrete is the primary load-carrying element (Pisani 1998; Zou 2003a). The response of Type Density lb/in.3 (g/cm3) Tensile Strength ksi (MPa) Tensile Modulus ksi (GPa) Elongation at Break (%) Aramid 0.045–0.052 (1.25–1.44) 174–305 (1,200–2,100) 7,832–18,855 (54–130) 1.5–3.8 Carbon 0.054–0.055 (1.5–1.53) 261–370 (1,800–2,550) 1,494 (10.3) 1.3–1.57 Based on ACI (2004). TABLE 14 PROPERTIES OF FRP TENDONS FOR PRESTRESSED CONCRETE APPLICATION FIGURE 16 CFRP tendons for prestressed concrete application (used by permission from Yail J. Kim).

45 prestressed concrete beams with unbonded FRP tendons is different from the response of those with bonded tendons [Figure 17(b)]. This is because concrete sections with unbonded FRP tendons allow more rotation at pseudo-hinges where flexural cracks form, leading to increased beam deflections. Unbonded FRP tendons experience less strain development compared with bonded tendons (Pisani 1998), and can also be used for external strengthening applications (Burningham et al. 2014). Concrete beams with bonded FRP tendons show higher flexural capacity relative to the beams with unbonded FRP tendons. Saafi and Toutanji (1998) reported that prestressed concrete beams with unbonded AFRP tendons demonstrated a 10%–20% lower flexural strength than the beams with bonded tendons. Skew angles in CFRP-prestressed concrete bridges may not alter their dynamic responses [e.g., similar frequencies were obtained from bridges with skew angles of 15° and 30° (Grace and Adbel-Sayed 2000]. Anchorage plays an important role in preserving prestress levels at transfer. MacDougall et al. (2011) showed a prestress loss of up to 60% in an anchor system for CFRP tendons. When jacking FRP tendons, stress concentrations can be alleviated using adequate anchorage. Once an FRP-prestressed concrete beam cracks, the occurrence of subsequent cracks is affected by tendon strains because FRP tendons transmit stresses to the concrete (Lees and Burgoyne 1996). Flexural cracks may develop at the location of shear stirrups, and FRP-tendon types do not contribute to crack formation (Fam et al. 1997). These cracks act as local hinges, allowing the rotation of uncracked concrete in both unbonded and partially bonded FRP-prestressed concrete beams (Lees and Burgoyne 1996). The degree of bond between the FRP tendons and concrete controls the cracking pattern of a prestressed member. A few localized cracks develop in unbonded FRP-prestresed beams, which resulted in the failure of the beams (Lees and Burgoyne 1996; Maissen and de Smet 1998). By contrast, beams with bonded FRP tendons reveal relatively uniform cracks along the span (Toutanji and Saafi 1999). FRP reinforcement ratios do not contribute to the deflection of prestressed concrete beams in both pre- and early post-cracking stages (Grace and Singh 2003). The deformability of FRP-prestressed concrete beams is often defined as a ratio between the responses at ultimate and at cracking (Au and Du 2008), although there is no consensus about the definition in the structural engineering community. Over-reinforced FRP-prestressed concrete beams show higher deformability than under-reinforced beams, which is the opposite case of steel-prestressed members (Saafi and Toutanji 1998; Grace and Singh 2003). This supports the concept that over-reinforced sections are a preferred design option for FRP-prestressed members. A combination of bonded and unbonded FRP tendons can enhance the deformability of FRP-prestressed concrete beams (Toutanji and Saafi 1999; Grace and Singh 2003). Non-stressed FRP tendons may also be added to a prestressed concrete beam to improve deformability (Toutanji and Saafi 1999). CFRP-prestressed concrete girders exhibit good fatigue resistance in a service load range (Grace and Adbel-Sayed 2000). When a CFRP-prestressed concrete beam cracks by fatigue, the beam’s loading and unloading curves are not linear, although CFRP is a linear elastic material (a) (b) FIGURE 17 Load-deflection behavior of prestressed concrete beams: (a) comparison between steel strands versus CFRP tendons (reproduced based on Fam et al. 1997); (b) comparison between bonded and unbonded CFRP tendons (reproduced based on Maissen and de Smet 1998).

46 (Abdelrahman and Rizkalla 1999). This is because the cracked concrete section controls the behavior of the beam. The elastic recovery of steel-prestresed beams upon unloading is not comparable with that of CFRP-prestressed beams: the former shows more residual deflections than the latter (Zou 2003a). Vertically distributed FRP tendons in a prestressed concrete beam can cause progressive fail- ure, because the tendons close to the beam-bottom first rupture when loaded (Dolan and Swanson 2002). Au and Du (2008) state that if a ratio between the neutral axis depth and effective depth of an FRP-prestressed concrete section is greater than 0.3, the section fails in compression. Providing more FRP tendons to negative moment regions can improve the behavior of continuous prestressed concrete beams by retarding flexural failure (Maissen and de Smet 1998). Grace et al. (2003) tested a full-scale double-tee girder in flexure, as shown in Figure 18. The girder was post-tensioned using CFRP tendons and reinforced with steel stirrups. The girder failed by concrete crushing, followed by rupture of the CFRP tendons. Shear behavior Figure 19 shows the construction of full-scale CFRP-prestressed concrete beams with CFRP shear stirrups. When these beams were loaded to failure, significant diagonal tension cracks (or shear cracks) developed before flexural failure occurred, and the crack width was not influenced by web reinforce- ment ratios (Fam et al. 1997). The large crack opening in shear appears to be related to the CFRP’s modulus that is relatively lower than its steel counterpart; in other words, the shear crack width might have been reduced if steel stirrups were used. If GFRP stirrups are used for CFRP-prestressed con- crete beams, the GFRP’s low modulus compared with steel stirrups needs to be considered (Stoll et al. 2000). Noel and Soudki (2014) proposed double-headed GFRP bars to enhance the shear resistance of concrete slabs post-tensioned with CFRP tendons. Marche and Green (1999) examined the punching shear behavior of laboratory-scale bridge decks transversely prestressed with CFRP tendons. The pro- posed benefit of this deck system is that the deck thickness could be reduced by increasing membrane action associated with prestressed CFRP tendons. As the deck was loaded, cracking near the anchor- age was noticed owing to the confining effect provided by the prestressed CFRP. Brittle punching shear failure was observed with a typical punching cone. A comparative study with a steel-prestressed deck revealed that the CFRP-based deck had a lower stiffness, although both decks demonstrated similar cracking patterns. The CFRP-prestressed deck failed at a much higher load (up to 150%) than a steel-prestressed control deck. bond, coupling, and transfer length Research shows that bond between FRP tendons and anchorage is controlled by several parameters, including the geometric configuration of anchors, types of grouting materials, and surface properties of FRP (Zhang et al. 2000). The surface texture of FRP tendons (e.g., sand-coated, spiral-indented, and braided) influences the amount of slip inside the anchorage. The length of commercially avail- able CFRP tendons may be limited on site. In this case, couplers can be used with shrink rubber to protect the steel coupler from corrosion, as shown in Figure 20. To secure the position of FRP ten- dons in the coupler, expansive cement may be used (i.e., Type K shrinkage compensating cement). Lees et al. (1995) reported that the pressure of expansive cement in a coupler for FRP tendons developed to 2,610 psi (18 MPa), 3,630 psi (25 MPa), and 4,350 psi (30 MPa) in 24 h, 48 h, and 72 h, respectively. According to experimental data (Soudki et al. 1997), the instantaneous transfer length of CFRP- prestressed concrete beams was 80 to 90 times the tendon diameter (longer than the transfer length of steel strands), and this trend was maintained for up to 200 days. AFRP tendons appear to have a shorter transfer length than CFRP tendons. Zou (2003b) reported that the transfer length of AFRP tendons was 20 to 30 times their diameter. Concrete strength influences FRP’s transfer length [e.g., an increase in concrete strength from 3,770 psi (26 MPa) to 9,140 psi (63 MPa) reduces the transfer length of CFRP tendons up to 60% (Zou 2003b)]. Sand-coated AFRP tendons show a transfer length comparable with conventional steel strands (Zou 2003b).

47 (a) (b) (c) FIGURE 18 Full-scale, double-tee, CFRP-prestressed concrete girder [Grace et al. 2003 (used by permission from American Concrete Institute)]: (a) fabricated cage; (b) anchors for CFRP tendons; (c) failed girder.

48 (a) (b) (c) FIGURE 19 CFRP stirrups for CFRP-prestressed concrete beams (used by permission from Amir Fam): (a) schematic; (b) fabricated cage; (c) completed beam.

49 durability FRP-prestressed concrete beams show a stiffness decrease under fatigue loading (Grace 2000; Mertol et al. 2006). This is because of a decrease in tension stiffening and cracking, as well as a change in the elastic modulus of concrete (Saiedi et al. 2011). When loaded in fatigue, the camber of FRP-prestressed beams is reduced at an early loading stage and stabilizes after 1 million cycles (Saiedi et al. 2011). CFRP-prestressed concrete beams have a longer fatigue life than steel-prestressed beams, when simul- taneously subjected to low temperature [-18°F (-28°C), Saiedi et al. 2011]. Fatigue and low tempera- ture exposure deteriorate bond between the CFRP tendons and concrete, as shown in Figure 21. The temperature-induced bond damage is attributed to a difference in the coefficient of thermal expan- sion between the tendons and the concrete. A tidal service environment also worsens bond between FRP tendons and concrete (Sen et al. 1993). Anchor systems for FRP tendons in fatigue generally perform well and satisfy the requirements of the Post-tensioning Institute [no failure occurs over 2 million cycles (El Refai et al. 2006)]. Nonetheless, the properties of FRP tendons are still degrad- able by fatigue [e.g., a 4% decrease in the elastic modulus of CFRP tendons (El Refai et al. 2006)]. GFRP tendons progressively deteriorate in aggressive environments. The durability performance of GFRP-prestressed concrete beams was inferior to the performance of the beams with steel strands in (a) (b) FIGURE 20 Coupler for CFRP tendon (used by permission from Amir Fam): (a) close-up view; (b) coupler covered with shrink rubber.

50 (a) (b) (c) FIGURE 21 Bond damage and slip of CFRP tendons (used by permission from Amir Fam): (a) initial; (b) slip in progress; (c) final slip.

51 a marine environment (i.e., load-carrying capacity, ductility, and failure mode), primarily as a result of the degradation of GFRP tendons (Sen et al. 1993). design aspects General The design approach for FRP-prestressed concrete is basically the same as that for steel-prestressed con- crete, except for technical aspects related to the unique characteristics of FRP tendons (e.g., compression- controlled section, creep-rupture, and deformability). The assumptions made in earlier in this chapter (“General”) for FRP-reinforced concrete are valid for FRP-prestressed concrete members. It can be noted that NCHRP Topic 12-97 (Guide Specification for the Design of Concrete Bridge Beams Pre- stressed with CFRP Systems (Belarbi 2016)] is currently underway to develop design specifications for CFRP-prestressed concrete members. Flexural Design Similar to the case of FRP-reinforced concrete (“Flexural Drawings”), three failure modes are consid- ered, dependent on reinforcement ratios (r): balanced (rb), tension (r < rb), and compression (rb < r). The nominal bending resistance of an FRP-prestressed concrete section (Mn) is calculated by strain com- patibility (bonded FRP tendons) and force equilibrium. When a tension-controlled beam fails in flexure, the equivalent concrete stress block does not fully develop. A refined method is required to determine the resultant compressive force of the section, or an alternative stress block proposed by Intelligent Sensing for Innovative Structures may be employed (ISIS Canada 2008b). If a section fails in compres- sion, tendon stress is lower than the strength of the FRP and is not known. To attain the stress at failure, iteration is necessary until force equilibrium is achieved. Tension forces in vertically distributed multiple FRP tendons vary with strain. This is particularly important for a tension-controlled section; the bottom- most tendon ruptures, whereas other tendons above it still carry tensile forces. Although unbonded FRP tendons can be used for prestressed concrete structures, specific design methods are limitedly reported. Table 15 lists allowable stresses in FRP tendons compiled from published design guidelines. The stress limits of FRP are lower than the limits of steel tendons specified in the AASHTO LRFD BDS (e.g., 0.75 fpu at immediately prior to transfer, where fpu is the ultimate strength of the tendon). One of the reasons is that creep-rupture needs to be considered for FRP tendons. ACI 440.4R-04 (ACI 2004) schematically explains the creep-rupture of FRP tendons (Figure 22). By controlling a prestressing limit, premature creep-rupture failure is avoided (i.e, the tertiary stage in Figure 22 cannot be reached). Serviceability often controls the design of FRP-prestressed concrete, rather than strength, which is the same as the case of steel-prestressed concrete. As in the case of conventional prestressed concrete, FRP-prestressed concrete experiences prestress losses. FRP tendons relax as a result of the following components (ACI 2004): relaxation of the resin and fibers as well as fiber straightening. Several param- eters affect the relaxation of FRP tendons: the modular ratio between the fiber and resin, volumetric frac- tion of the fibers, and the quality of a pultrusion process. Literature shows that CFRP tendons do not relax, whereas AFRP tendons relax by 6% to 18% in 100 years (Dolan 1989). Limited research has been done on prestress losses induced by friction. The exponential friction loss equation of steel strands may be used for FRP, until refined design proposals are available. Prestress losses caused by the elastic shortening, creep, and shrinkage of concrete with steel strands are valid for FRP-prestressed members. The anchor-set loss of FRP tendons is a function of the elastic modulus. Source Type Jacking Transfer ACI 440.4R-04 (ACI 2004) CFRP 0.65fpu 0.60fpu AFRP 0.50fpu 0.40fpu CHBDC (CSA 2014) CFRP 0.70fpu 0.60fpu AFRP 0.40fpu 0.38fpu TABLE 15 ALLOWABLE STRESS LIMITS FOR FRP TENDONS

52 Because FRP tendons do not yield, the concept of ductility (the ratio between ultimate and yield responses) is not applicable. An alternative method called deformability is, therefore, used to ensure the sufficient flexural deformation of FRP-prestressed concrete members until failure. Existing deformabil- ity checks for design are shown in Table 16. The 440.4R-04 expression does not have a limit, which may restrict its use. To address this limitation, Kim (2010) proposed a deformability index limit of DI = 2.0 associated with the 440.4R-04 equation. Short-term deflection for FRP-prestressed concrete members is calculated by structural analysis. Their long-term deflections may be obtained using the empirical multipliers listed in Table 17. Although strict crack control does not appear to be necessary (because FRP tendons are noncorrosive), it is still required for aesthetic reasons. Fatigue damage in FRP-prestressed concrete members is minimal, because they are uncracked under service loading. Accordingly, specific design information on fatigue is not available. (a) (b) St ra in Time FIGURE 22 Creep-rupture of FRP tendons (reproduced based on ACI 440.4R-04): (a) three- stage creep behavior; (b) comparison of creep failure between CFRP and AFRP tendons. Source Deformability ACI 440.4R-04 (ACI 2004) ( ) ps pu d kDI     − − = 1 1 1 with '85.0 c pu f dfρ = where DI = deformability index k = ratio of neutral axis depth to FRP tendon depth d = distance from extreme compression fiber to centroid of FRP tendon 1 = stress block factor for concrete pu = ultimate compressive strain in concrete ps = strain in FRP tendon at service = reinforcement ratio puf = design ultimate tensile strength of FRP tendon ' cf = specified compressive strength of concrete. CHBDC (CSA 2014) cc ultult M MJ ψ= where J = overall performance factor (J > 4 for rectangular sections and J > 6 for T-sections) ultM = ultimate moment capacity of the section cM = moment corresponding to a maximum compressive concrete strain in the section of 0.001 ult = curvature at ultM c = curvature at cM . α ε ε β α β ε ε ρ ψ ψ ψ TABLE 16 DEFORMABILITY EXPRESSIONS FOR FRP-PRESTRESSED CONCRETE MEMBERS

53 Shear Design The shear reinforcement of FRP-prestressed concrete beams can be either steel or FRP stirrups. If steel stirrups are used, design approaches are identical to those for conventional prestressed concrete beams. If FRP stirrups are employed, the design approach discussed earlier (“Shear Design”) may be referenced. The expression of Vc is not unified yet, and varies by design guidelines (Table 18). Site application Figure 23 exhibits the use of prestressed CFRP strands for both pretensioned [Figure 23(a)] and post-tensioned bridges [Figure 23(b)]. A bridge in Michigan was constructed with CFRP-prestressed concrete, as shown in Figure 24. The two-lane bridge comprises 12 double-tee girders internally pretensioned with CFRP and externally post-tensioned CFRP tendons. The AASHTO LRFD Specifi- cations were referenced to design the bridge under the Michigan MS-23 standard truck loading (similar to AASHTO HS 25). Upon completion of the bridge construction, a load test was conducted with trucks with a weight of 58 kips (258 kN). The in situ load distribution was less than that calculated by the Loading Stage Deflection CFRP AFRP At erection Deflection due to self-weight 1.85 1.85 Camber due to prestress 1.80 2.00 At final Deflection due to self-weight 2.70 2.70 Camber due to prestress 1.00 1.00 Deflection due to applied loads 4.10 4.00 Source: ACI (2004). TABLE 17 MULTIPLIERS FOR FRP-PRESTRESSED CONCRETE MEMBERS WITHOUT TOPPING Source Nominal shear resistance (Vc) ACI 440.4R-04 (ACI 2004) dbfV wcc '17.0= where (metric units) ' cf = specified compressive strength of concrete wb = width of web d = distance from extreme compression fiber to centroid of FRP tendon. Intelligent Sensing for Innovative Structures (ISIS Canada 2008b) s long vcrc E E dbfV long 5.2= with ++ = zex s000,1 300,1 15001 4.0 where (metric units) = factor to account for shear resistance of cracked concrete =crf cracking strength of concrete =vb effective web width = long d effective shear depth for longitudinal tensile reinforcement longE = elastic modulus of FRP in the longitudinal direction sE = elastic modulus of steel x = longitudinal strain at mid-depth of cross section zes = equivalent crack spacing for influence of aggregate size. β β ε β ε TABLE 18 SHEAR RESISTANCE PROVIDED BY CONCRETE

54 (a) (b) FIGURE 23 Prestressed CFRP tendons for bridge construction (Frankhauser et al. 2016): (a) pretensioning; (b) post-tensioning. AASHTO LRFD BDS equations. Post-tensioned CFRP tendons were utilized for a voided-slab bridge in Maine (Thompson and Parlin 2013). The bridge was a single span with a width of 36 ft (11 m) and was transversely post-tensioned with 1.6-in. (40-mm) diameter tendons. A cost comparison showed that the CFRP tendons and conventional steel strands, respectively, required $525.21 and $469.54 per ft2 ($5,653 and $5,054 per m2) of the deck area. After sheathed CFRP tendons were inserted into the slab [Figure 25(a)], the tendons were tensioned with a hydraulic jack [Figure 25(b)] and end anchor- age was placed to hold the post-tensioning forces [Figure 25(c)]. A completed site view is provided in Figure 25(d). The need for long-term performance monitoring was stated to justify the increased costs owing to the use of the CFRP. fiber-reinforced Polymer-Strengthened concrete memberS Constructed infrastructure deteriorates with time. A number of factors are responsible for degrading the performance of structural members, including environmental and physical. FRP strengthening for deterio- rated members is proven technology with numerous benefits (i.e., noncorrosiveness, favorable strength- to-weight ratio, fatigue resistance, and reduced maintenance costs). This section reviews the fundamentals of FRP strengthening with an emphasis on materials, design approaches, and the behavior of strengthened members. Because FRP strengthening for steel and timber structures remains in development, of interest is the use of FRP sheets or laminates to repair and/or retrofit concrete members (e.g., flexural and shear- strengthening of girders and decks, confinement of columns, and other applications such as for bent caps).

55 (a) (b) (c) FIGURE 24 Bridge Street Bridge [Grace et al. 2004 (used by permission from American Concrete Institute)]: (a) overview; (b) CFRP tendons; (c) load testing.

56 (a) (b) FIGURE 25 Little Pond Bridge [Thompson and Parlin 2013 (used by permission from Maine DOT)]: (a) placing CFRP tendons; (b) post-tensioning. (continued) frP Sheets and laminates or Strips FRP sheets or laminates are bonded to the soffit of a structural member in order to increase the load- carrying capacity. The most widely used bonding agent is epoxy. FRP laminates (frequently called plates) are produced by pultrusion, as discussed earlier in Chapter Three, “Manufacturing Techniques.” FRP sheets are fiber fabrics to be impregnated in a resin matrix on site (this process is called “wet lay-up”). As with other FRP applications in highway infrastructure, two types of fibers are typically employed: CFRP and GFRP. Other fiber types (e.g., AFRP and steel fiber-reinforced polymer) may also be available (Zatar and Mutsuyoshi 2004; Lopez et al. 2007). Table 19 summarizes the engineering properties of FRP sheets and laminates. FRP strips and rods are usable for strengthening concrete structures as near-surface-mounted (NSM) reinforcement. The NSM method uses FRP materials inserted into narrow grooves that are precut along the substrate of a member, followed by permanent bonding with an adhesive. The properties of these reinforcing materials are similar to those of FRP bars and tendons discussed earlier in this chapter under “FRP Reinforcing Bars” and “FRP Tendons,” respectively. Figure 26(a) shows some examples of commercially available FRP products for structural strengthen- ing. High-strength fibers in an FRP composite are unidirectionally aligned in most cases, although bidirectional fiber arrangement can be manufactured. FRP sheets exhibit virtually linear elastic stress- strain behavior until rupture, as shown in Figures 26(b) and (c). A critical concern for FRP sheets and

57 (c) (d) FIGURE 25 (Continued) Little Pond Bridge [Thompson and Parlin 2013 (used by permission from Maine DOT)]: (c) end anchorage; (d) completed work. Type Weight lb/ft2 (g/m2) Design Thickness in. (mm) Tensile Strength Ksi (MPa) Tensile Modulus Ksi (GPa) Sheeta Aramid 0.12 (600) 0.011 (0.28) 290 (2,000) 17,400 (120) Carbon 0.062–0.124 (300–600) 0.007–0.013 (0.17–0.33) 550 (3,800) 33,000 (227) Glass 0.19 (900) 0.015 (0.37) 220 (1,517) 10,500 (72) Laminateb Carbon 0.13–0.39 (618–1,900) 0.04–0.177 (1.0–4.5) 104–406 (717–2,800) 9,450–43,500 (65–300) Glass 0.19 (900) 0.04–0.052 (1.0–1.3) 77–83 (531–575) 3,430–3,785 (24–26) Source: Based on ACI (2007). abased on fiber properties. bbased on composite properties. TABLE 19 PROPERTIES OF FRP FABRIC SHEETS AND PREFABRICATED LAMINATES

58 (a) (b) (c) CFRP laminate (prefabricated) GFRP sheet (wet-lay-up) CFRP sheet (wet-lay-up) FIGURE 26 FRP sheets for structural strengthening (used by permission from Yail J. Kim): (a) various types; (b) stress-strain of CFRP sheet; (c) stress-strain of SRP sheet. laminates is that its performance is temperature-dependent. When FRP sheets are subjected to a tem- perature exceeding their glass transition temperature, the resin matrix is thermally degraded and, as such, limited stress transfer is achieved between the fibers (this process eventually causes the failure of the FRP composite). The glass transition temperature of most epoxy resins is 140°F to 176°F (60°C to 80°C). Further discussions on the effect of elevated temperatures are provided in a subsequent section. flexural behavior The behavior of FRP-strengthened beams in flexure is principally identical to that of steel-plated beams (the first generation of external strengthening). Externally bonded (EB) and NSM methods are widely used FRP strengthening techniques (Figure 27). Because NSM FRP reinforcement is located inside a strengthened beam, the bond performance of the NSM method is superior to that of the EB method (De Lorenzis and Teng 2007). Groove configurations may influence bond between the NSM FRP and concrete (Oudah and El-Hacha 2012). The size of a groove for NSM strengthening may be 1.5 to 3.0 times larger than the CFRP’s diameter or thickness (ACI 2008a). The AASHTO guide specifications for bonded FRP systems do not explicitly provide information on the NSM method (AASHTO 2012a). Figure 28(a) compares the load-displacement behavior of CFRP-strengthened beams with an unstrengthened control beam. When the unstrengthened beam yields, its flexural stiffness is markedly reduced, whereas the strengthened beams reveal a marginal reduction in flexural stiffness. The pre- yield stiffness of all these beams is similar, as shown in Figure 28(b). This can be because the thin FRP sheets or strips do not influence the stiffness of the beams until the steel reinforcement yields. Because

59 FIGURE 27 FRP strengthening methods (used by permission from Yail J. Kim): (a) schematic of EB and NSM FRP; (b) EB beams; (c) NSM beams. FRP sheet FRP strip (a) (b) (c) EB NSM

60 of NSM CFRP’s improved bond performance, the beam strengthened with NSM CFRP shows a higher failure load than the beam with EB CFRP [Figure 28(a)]. Both of these beams fail in a brittle manner owing to CFRP debonding. A number of failure modes are available in FRP-strengthened beams, as shown in Figure 29. FRP-rupture may occur when FRP strains exceed their ultimate capacity [Figure 29(a)]. If end- anchorage is provided, debonding is precluded from the ends, and concrete crushing controls the failure of the beam [Figure 29(b)]. Otherwise, end-peeling can take place [Figure 29(c)]. Local FRP-debonding is frequently observed, which is attributable to intermediate cracks [Figures 29(d) and (e)]. Although intermediate-crack-induced debonding does not imply the immediate failure of a strengthened beam, FRP-debonding eventually propagates as the beam is further loaded and causes the failure of the beam. NSM CFRP strips may accompany stress concentrations at termination points, which can be connected with shear cracks [Figure 29(f)]. Sustained load affects the failure characteristics of NSM CFRP-strengthened beams (Figure 30). With an increase in sustained load levels, the damage of the CFRP–concrete interface is significantly augmented. Post-tensioned CFRP sheets are studied to strengthen existing structural members. El-Hacha et al. (2001) provided a state-of-the-art review of post-tensioned CFRP sheets to upgrade the performance of constructed concrete structures. Various subjects were discussed: post-tensioning methods, flexural and shear-strengthening applications, and field demonstration projects. Wight et al. (2001) examined the efficacy of post-tensioned CFRP sheets on improving the capacity of reinforced concrete beams. Force distributions were predicted when multiple layers of CFRP sheets were post-tensioned. The upgraded beam demonstrated an increase of 41% in flexural capacity compared with a control beam. El-Hacha et al. (2004) showed the effect of cold temperature exposure on the behavior of concrete T-beams retrofitted with post-tensioned CFRP sheets. The flexural strengths of the beams tested at 72°F (22°C) and -18°F (-28°C) were 17% and 32% higher than the strength of an unstrengthened beam, respectively. Kim et al. (2008b) tested prestressed concrete beams strengthened with post-tensioned CFRP sheets. The CFRP’s post-tensioning level was an important factor that controlled the beams’ ductility. The crack progression of the strengthened beams was stable within a service load range, beyond which it became more rapid. The short-term prestress loss of post-tensioned CFRP sheets was examined by Kim et al. (2010), based on experimental and mathematical approaches. Prestress losses of approxi- mately 6% were measured when the post-tensioning force was transferred from a jacking apparatus to the beams, and a 10% loss was suggested for design. NSM CFRP may be post-tensioned to enhance the behavior of flexural members, particularly for serviceability (e.g., crack control and deflection), and to better utilize the tensile strength of the CFRP. Without post-tensioning, 50%–60% of the CFRP’s tensile strength can be used at the failure of the strengthened beams (Hajihashemi et al. 2011). The efficiency of NSM CFRP is augmented with post- tensioning levels; however, the members’ ductility is reduced (Choi et al. 2011). The maximum usable post-tensioning level may, therefore, be 60% of the CFRP capacity (Rezazadeh et al. 2016). Selected post-tensioning techniques for NSM CFRP composites are illustrated in Figure 31. (a) (b) FIGURE 28 Load-displacement behavior of FRP-strengthened beams (used by permission from Yail J. Kim): (a) comparison between EB and NSM CFRP strengthening schemes; (b) effect of shear-span-to-depth ratios.

61 FIGURE 29 Failure modes of FRP-strengthened beams in flexure (used by permission from Yail J. Kim): (a) CFRP rupture; (b) concrete crushing; (c) CFRP debonding. (continued on next page) (a) (c) (b)

62 (e) (d) (f) FIGURE 29 (Continued) Failure modes of FRP-strengthened beams in flexure (used by permission from Yail J. Kim): (d) intermediate- crack-induced CFRP debonding; (e) intermediate-crack-induced NSM CFRP debonding; (f) shear-crack-induced NSM CFRP debonding.

63 (a) (c) (b) FIGURE 30 Effect of sustained load on failure of NSM CFRP (used by permission from Yail J. Kim): (a) 25% of ultimate capacity of beam; (b) 50% of ultimate capacity of beam; (c) 75% of ultimate capacity of beam.

64 (a) (b) (c) FIGURE 31 Post-tensioning NSM CFRP [El-Hacha and Soudki 2013 (used by permission from Canadian Society for Civil Engineering)]: (a) jacking bed; (b) jacking frame; (c) tensioning against beam. When NSM CFRP is post-tensioned, the anchorage experiences eccentric forces and the girder con- crete is subjected to local tension (Kim et al. 2014). Cracking in the vicinity of the anchorage may thus occur, as shown in Figure 32. In spite of complex stress states in the anchorage of the post-tensioned NSM CFRP, the stress magnitude may not be a concern and can be negligible (Kim et al. 2015). With the use of an adequate anchor system, a prestress loss of less than 10% is expected (El-Hacha and Gaafar 2011). The slip of NSM CFRP is proportional to post-tensioning levels (Oudah and El-Hacha 2012). If the NSM CFRP is not properly anchored, premature pull-out failure takes place as the strengthened member is loaded (Taljsten and Nordin 2007; Choi et al. 2011). Post-tensioned NSM CFRP improves the cracking and yield loads of a strengthened beam (Hajihashemi et al. 2011), which is attributable to the load-sharing mechanism between the internal reinforcement and the CFRP. Conventional sectional analysis well predicts the beam’s flex- ural capacity (Badawi and Soudki 2009). The behavior of the strengthened beam is influenced by post-tensioning levels, including ductility and fatigue life (Choi et al. 2011; Wahab et al. 2012). To enhance ductility, post-tensioned NSM CFRP may be unbonded using plastic tubes. Choi et al. (2011) reported that the ductility of a reinforced concrete beam partially bonded with post-tensioned NSM CFRP rods was almost 80% higher than that of its fully bonded counterpart, without sacrificing the load-carrying capacity. The flexural stiffness of the beam, however, decreases as unbonded CFRP length increases (Rezazadeh et al. 2016). Recent research proposes a hybrid strengthening system consisting of nonprestressed (full bond) and post-tensioned (partial bond) NSM CFRP strips to improve both strength and ductility (Rezazadeh et al. 2016).

65 Composite action between the strengthened member and post-tensioned NSM CFRP is crucial to achieving structural integrity. Because of the confining effect provided by the post-tensioned NSM CFRP, the strengthened beams tend to show reduced crack spacing compared with those strengthened with non- prestressed NSM CFRP (Ye et al. 2014; Yao and Wu 2016). The number and width of cracks in the post- tensioned beams are also diminished (Hajihashemi et al. 2011). To maximize strengthening efficiency, the bond length of post-tensioned NSM CFRP may be 60% to 90% of a girder’s span length (Kim et al. 2015). If corrosion damage occurs and propagates in the strengthened girder, the CFRP’s effectiveness increases (Kim et al. 2016). The failure of concrete beams with post-tensioned NSM CFRP is controlled by the interfacial behavior (e.g., concrete-cover splitting and CFRP-debonding), and thereby U-wrap anchors may be bonded to confine the NSM CFRP system (Peng et al. 2014). U-wrap anchorage is particularly beneficial when a load is applied near the support, which enlarges shear effects (Yao and Wu 2016). Fatigue loading degrades the stiffness of a reinforced concrete beam strengthened with post-tensioned NSM CFRP (Oudah and El-Hacha 2012). This concept aligns with the bond-slip of NSM CFRP, indicat- ing the progressive deterioration of the CFRP-concrete or CFRP–epoxy interface (i.e., damage accumula- tion) induced by fatigue loading. Fatigue can also rupture internal steel reinforcement (Wahab et al. 2012), although the stress of the steel is reduced by the presence of the CFRP. Further details on post-tensioned NSM CFRP applications are available in a recently published review paper (El-Hacha and Soudki 2013). Shear behavior FRP sheets are intended to cover diagonal tension cracks in the shear span of a reinforced concrete beam. The applied shear forces are, therefore, shared by the FRP and internal steel stirrups of the beam. FRP strengthening reduces the width of shear cracks (Rahal and Rumaih 2011). NSM FRP rods or strips may be used to shear-strengthen reinforced concrete members (Barros et al. 2007; Islam 2009; Rizzo and De Lorenzis 2009). FRP sheets are also effective to shear-strengthen deep beams (ElMaaddawy and Sherif 2009). Although an inclined bonding scheme with discrete FRP sheets (perpendicular to shear cracks) provides the most effective shear-strengthening efficacy, it may not be preferred because of construction convenience. Many parameters influence the behavior of shear-strengthened concrete members, including FRP reinforcement ratios, shear-span-to-depth ratios, spacing of steel stirrups, FRP properties, and bonding methods (Li et al. 2001; Islam 2009; Panigrahi et al. 2016). Test results reveal that NSM FRP performs better than FRP U-wraps, in terms of increasing the shear capacity of rein- forced concrete beams [e.g., 16% and 44% increases in shear capacity with U-wraps and NSM CFRP rods, respectively (Rizzo and De Lorenzis 2009)]. This is because of the superior bond of NSM FRP (Islam 2009). In addition, NSM FRP enhances the serviceability (deformation) of shear-strengthened members (Barros et al. 2007). Rizzo and De Lorenzis (2009) reported that FRP-debonding accelerated when a distance between FRP-shear reinforcement was reduced, which agrees with the test data of Li et al. (2001) and Belarbi Crack due to high prestressing force Crack caused rotation in fixed bracket FIGURE 32 Cracking in anchorage zone caused by post-tensioning [El-Hacha and Gaafar 2011 (used by permission from Prestressed Concrete Institute)].

66 et al. (2012) showing that FRP strengthening was more efficient for beams with larger internal shear stirrup spacing. This implies that the interaction between the shear reinforcements (either FRP/steel or FRP/FRP) controls the performance of the bonded FRP. Therefore, it can be stated that the simple design equation for shear resistance (Vn = Vc + Vs + Vf in “Shear-Strengthening Design” later in this chapter) is an approximate method. Typical effective strains in NSM CFRP strips for shear vary by 20% to 35% of the CFRP’s ultimate capacity (Triantafillou and Antonopoulos 2000; Adhikari et al. 2004; Islam 2009). The effective strain of FRP U-wraps appears to be independent of a beam size (Belarbi et al. 2012). More testing is, however, necessary to generalize this preliminary conclusion. The thicker an FRP is, the better shear crack growth is controlled in a strengthened beam (Panigrahi et al. 2016). Nonetheless, an optimum FRP-reinforcement ratio exists in shear-strengthening reinforced concrete beams. In other words, an increase in FRP layers does not proportionally augment the shear capacity of a strengthened beam (Khalifa and Nanni 2002). When NSM FRP is used for shear strength- ening, it is essential that attention be paid because an unstrengthened zone between the beam and slab (i.e., outside the NSM FRP bonded region) may fail (Rahal and Rumaih 2011). The failure of concrete beams strengthened with NSM FRP is less brittle than that of the beams with EB FRP (Barros et al. 2007). FRP U-wraps may fail in either debonding or rupture, as shown in Fig- ure 33. FRP-debonding is predominantly observed when a shear-strengthened beam fails [Figs. 33(a) FIGURE 33 Failure mode of CFRP U-wraps (used by permission from Yail J. Kim): (a) complete debonding; (b) partial debonding. (continued) (a) (b) Loading point Bond failure Notch Shear crack

67 (c) (d) FIGURE 33 (Continued) Failure mode of CFRP U-wraps (used by permission from Yail J. Kim): (c) rupture; (d) occurrence of FRP failure (reproduced based on Belarbi et al. 2011). and (b)], unless anchorage is provided [Figure 33(c)] or FRP completely surrounds the beam. Belarbi et al. (2012) tested full-scale reinforced concrete beams strengthened with FRP sheets in shear, and showed that the beams with anchorage resulted in 7%–48% higher shear capacities than those with- out anchorage. Figure 33(d) illustrates the occurrence of individual failure modes collected from published literature. In the case of complete FRP-wrapping, 55% and 45% of the test data show FRP-rupture and debonding, respectively. These trends change when FRP U-wraps and side-bonding are employed: 13% rupture and 64% debonding for U-wraps, and 7% rupture and 81% debonding for side-bonding. axial behavior FRP-wrapping increases the load-carrying capacity and ductility of reinforced concrete columns. FRP-confinement is considered to be passive strengthening, because the contribution of the FRP is negligible until the core concrete is significantly damaged. Figure 34 exhibits CFRP-confined concrete members and their load-displacement response. FRP-wrapping schemes can alter the strain develop- ment of FRP (Pham et al. 2015). As FRP-confined concrete is loaded, concrete stresses are transmitted to the FRP through surface bonding (De Lorenzis and Tepfers 2003). FRP-confinement, however, is not bond-critical (FRP-rupture controls the failure of the strengthened member), unlike flexural and shear-strengthening. Pantelides et al. (2013) proposed a design approach to achieve a certain level of ductility for CFRP-confined concrete columns by precluding the strain-softening behavior. FRP-confinement for circular columns is better than for rectangular columns (rectangular columns cause stress concentrations at the corners). Non-shrink cementitious materials may be used to modify

68 FIGURE 34 CFRP-confinement for axial concrete members (used by permission from Yail J. Kim): (a) axial loading; (b) failed members; (c) load-displacement response (reproduced after Teng et al. 2016). (a) (b) (c)

69 the shape of a rectangular column (e.g., oval shape), so that the efficacy of FRP-confining can be enhanced (Yan and Pantelides 2011). As reported by ElMaaddawy (2008), a strength increase resulting from FRP-wrapping is a func- tion of eccentricity (40% and 8% increases in axial column capacity with e/h ratios of 0.3 and 0.57, respectively, where e and h are the eccentricity and width of the column, respectively). Another thing to note is that partial-wrapping is less effective than complete-wrapping, in terms of increasing the load- carrying capacity and axial stiffness of columns (Yan et al. 2007; Pham et al. 2015). The effectiveness of FRP-wrapping increases when the strength of the core concrete is low (Tao et al. 2008). The extent of pre-damage in concrete before wrapping with FRP sheets affects the performance of FRP-confinement. The pre-damage, however, does not influence the energy absorption characteristics of FRP-wrapped concrete (Dalgic et al. 2016). For rectangular columns, their strength gain by FRP-wrapping is propor- tional to corner radii, and strains at the column corners slowly develop relative to those near the middle of the column section (Tao et al. 2008). The aspect ratios of FRP-confined columns ranging from 2:1 to 5:1 [a column diameter of 6 in. (152 mm)] do not affect the axial capacity and ductility (Mirmiran et al. 1998). The strength of FRP-wrapped concrete is not controlled by monotonic and cyclic loadings (Dalgic et al. 2016). It can be noted that a size effect may exist in FRP-confined columns. Akogbe et al. (2011) reported that CFRP-wrapped concrete tended to show a decrease in axial capacity with an increase in column diameter. When an FRP-strengthened column is loaded laterally, cracks initiate near the connection between the column and base footing, and propagate along the column’s tension side (Jiang et al. 2014). Longitudinally bonded FRP sheets may be added to improve the behavior of slender columns (Fitzwilliam and Bisby 2010), because the sheets carry tensile stresses caused by eccentric loading. FRP- wrapping is effective in relocating the plastic hinge of a column. An experimental program exhibits that a plastic hinge formed at a distance of 2.3 ft (700 mm) above the footing level with FRP-confinement, rather than at the column-footing connection (Rutledge et al. 2013). FRP-confining, however, does not appear to change the length of plastic hinges in rectangular columns (Jiang et al. 2014). FRP-wraps often rupture before reaching the FRP’s ultimate strain obtained from a coupon test. This observation may be explained by the following: (1) a multiaxial stress state develops in the FRP; (2) cracked concrete influ- ences the behavior of the FRP; (3) a curved column geometry affects the FRP deformation; and (4) a size effect exists (Mattys et al. 1999). In situ quality alongside wet lay-up (e.g., fiber alignment) and FRP properties may also affect the failure strain of the FRP (De Lorenzis and Tepfers 2003). Although FRP- confining is a promising method to enhance the behavior of a reinforced concrete column, this technique is not a good option if fractured reinforcing bars exist in a column (Rutledge et al. 2013). durability The durability performance of FRP-strengthened members is influenced by many parameters; for instance, temperature, moisture, chemicals, freeze–thaw, and ultraviolet rays. Unlike FRP materials, adhesives and substrates are susceptible to environmental damage, which can degrade the strengthen- ing system. Chemical effects that may alter the properties of bonding agents include oxidation and hydrolysis along with ultraviolet radiation and humidity (Stewart and Douglas 2012). Of importance is the deterioration of the FRP–concrete interface, because most FRP strengthening applications are bond-critical. The majority of the interface tests was conducted under a pure-shear condition (i.e., pull- out of the bonded FRP, Figure 35), whereas the FRP–concrete interface is subjected to multiple stresses when a strengthened member is loaded. Durability test results summarized in this section are, therefore, intended to provide information on the effect of aggressive service environments, rather than to specify an absolute performance metric. As shown in Figure 36, interfacial failure may happen (1) within the concrete substrate, (2) along the bond line between the adhesive and concrete, and (3) within the adhesive layer. With environmentally induced damage, a failure plane shifts from the concrete substrate to the adhesive layer (Dai et al. 2010; Green et al. 2000). As environmental cycling increases, microscale cracks form in the FRP–concrete interface (Dai et al. 2010). These cracks degrade interfacial stiffness and eventually cause the failure of the interface. FRP–concrete interfacial failure typically occurs at around 0.08 in. (2 mm) within the

70 (c)(b)(a) (d) FIGURE 35 Interface test (used by permission from Yail J. Kim): (a) schematic of test and instrumentation; (b) test protocols; (c) single-lap shear test; (d) freezing inside environmental chamber [-22°F (-30°C)]. (continued) substrate (Colombi et al. 2010), unless noticeable adhesive damage takes place. The failure zone tends to enlarge if the interface is environmentally damaged (Kim et al. 2011; Shrestha et al. 2016), as illus- trated in Figure 37. This observation may be attributable to an increased cohesive zone in the interface, with an increase in environmental conditioning (Subramaniam et al. 2008). Wet-dry cycling may degrade the adhesion capacity of a bonding agent without changing its modulus (Shrestha et al. 2016). Some test programs report that freeze–thaw does not cause a reduction in FRP’s modulus (Subramaniam et al. 2008). Capillary action and diffusion happen when the FRP–concrete interface is exposed to moisture, which irreversibly degrades the micro-pores of the interfacial con- stituents (Tuakta and Buyukozturk 2011; Shrestha et al. 2016). This moisture effect is localized and controls interfacial fracture energy, including adhesive plasticization (Abanilla et al. 2006; Ouyang and Wan 2008). In addition, moisture uptake results in matrix cracking and fiber-matrix debonding with increased void fraction (Karbhari 2002; Abanilla et al. 2006). To preclude environmentally induced deterioration, the FRP–concrete interface may be treated. Wan et al. (2006) used a water-tolerant primer to enhance interfacial bonds subjected to a moisture environment. Interfacial rheology is a function of elevated temperatures, including the softening of adhesives (Di Tommaso et al. 2001). Such softening behavior is also observed in a freeze–thaw test condition (Green 2007). The fracture toughness of the FRP–concrete interface is reduced when exposed

71 (e) (f) FIGURE 35 (Continued ) Interface test (used by permission from Yail J. Kim): (e) wet; (f) drying at room temperature [77°F (25°C)]. to low temperatures [e.g., below -22°F (-30°C)], which leads to the brittle failure of the interface (Di Tommaso et al. 2001). This is also valid for specimens subjected to freeze–thaw (Davalos et al. 2008b), possibly because of the low temperature effect. Low temperature without other environmental attributes may not deteriorate the FRP–concrete interface. For instance, CFRP-strengthened beams exposed to a temperature of -18°F (-28°C) for 48 hours showed increased cracking and ultimate loads (Taljsten and Carolin 2007). Sustained load does not appear to be critical if combined with freeze–thaw, in accordance with test results discussed in Green (2007): there is no influence of these loading conditions on the strength of CFRP-wrapped concrete. Heating and cooling cycles [73°F to 113°F (23°C to 45°C)] degrade the strength of CFRP-wrapped concrete, whereas freeze–thaw conditioning [-0.4°F to 64°F (-18°C to 18°C)] does not noticeably influence the capacity (El-Hacha et al. 2010). Teng et al. (2003) experimentally showed that freeze– thaw damage was not a contributing factor to the behavior of FRP-wrapped concrete. The test results conducted by Teng et al. (2003), however, indicated that freeze–thaw altered FRP strain development. These results are supported by other researchers’ findings: FRP–concrete interface is marginally dete- riorated by freeze–thaw (Green et al. 2000; Colombi et al. 2010). Adhesives may be further cured by heat and moisture that are provided during environmental conditioning (Tuakta and Buyukozturk 2011; Kim et al. 2012). CFRP-wrapping is generally more effective than GFRP-wrapping, in terms

72 (a) (b) (c) FIGURE 36 Interfacial failure of concrete strengthened with NSM CFRP (used by permission from Yail J. Kim): (a) within concrete substrate; (b) along bond line between adhesive and concrete; (c) within adhesive layer.

73 of increasing the strength and ductility of strengthened concrete exposed to aggressive environments (Toutanji 1999; Davalos et al. 2008a). Although FRP-wrapping protects core concrete in aggres- sive environments (Figure 38), environmental exposure may change the stress–strain relationship of FRP-confined concrete (Saenz and Pantelides 2006). If the core of strengthened concrete is envi- ronmentally damaged and dilates, early activation of FRP-confinement is achieved. Saenz and Pantelides (2006) noted that the ultimate radial strain of FRP-confined concrete was not influenced by the extent of environmental exposure. design aspects General FRP strengthening is used to upgrade the flexural, shear, and axial capacities of existing concrete members. Examples are bridge decks, girders, and columns. When seismic deficiency is a concern, (a) (b) 170 166 163 161 159 155 153 151 148 144 140 FIGURE 37 Topological change in CFRP-concrete interface due to environmental damage (used by permission from Yail J. Kim): (a) schematic of damage band; (b) undamaged control. (continued)

74 bridge columns are wrapped with FRP sheets. Design approaches for FRP-strengthened concrete members are similar to the approaches for conventional reinforced concrete (i.e., strain compatibility and force equilibrium), although the contribution of externally bonded FRP needs to be taken into consideration. Existing design guidelines, such as the AASHTO LRFD BDS, can thus be the founda- tion of FRP strengthening. Before a strengthening design is conducted, thorough examinations on structural members to be retrofitted or repaired are necessary. Several conditions need to be con- sidered in selecting FRP types (ACI 440.2R-08): alkalinity or acidity, thermal expansion, electri- cal conductivity, impact tolerance, creep-rupture, and fatigue. The AASHTO guide specifications (AASHTO 2012a) state additional considerations, including moisture equilibrium content, ultra- violet rays, freeze–thaw, and adhesive selection (AASHTO 2012a). NCHRP Report 514: Bonded Repair and Retrofit of Concrete Structures Using FRP Composites (Mirmiran et al. 2004) and NCHRP Report 609: Recommended Construction Specifications and Process Control Manual for Repair and Retrofit of Concrete Structures Using Bonded FRP Composites (Mirmiran et al. 2008) provide construction specifications. When determining the geometric and material properties of FRP sheets or laminates, manufacturers’ data sheets are valuable resources. Some manufacturers provide material properties based on composite thickness, whereas others provide the properties based on equivalent fiber thickness (these two properties cannot be interchangeable). Because FRP materials are susceptible to creep-rupture, FRP stresses in service (e.g., sustained or fatigue loading) may be limited, as shown in Table 20. Environmental reduction factors address potential uncertainties associated with exposure conditions (Table 21). These factors reduce the tensile strength of FRP (ACI 440.2R-08; ACI 2008a): ffu = CE ffu, where CE is the environmental reduction factor and ffu is the specified tensile strength of the FRP as (c) 164 162 160 158 156 153 151 149 147 144 140 136 132 FIGURE 37 (Continued) Topological change in CFRP-concrete interface due to environmental damage (used by permission from Yail J. Kim): (c) 100 cycles of freeze-thaw.

75 (a) (b) FIGURE 38 Comparison of concrete exposed to sulfuric acid (used by permission from Yail J. Kim): (a) plain concrete; (b) CFRP-wrapped concrete. Type CFRP GFRP AFRP Stress limit 0.55ffu 0.20ffu 0.30ffu Source: ACI (2008a). ffu = design ultimate tensile strength of FRP. TABLE 20 STRESS LIMIT FOR FRP UNDER SERVICE LOADING Exposure Conditions Type Environmental Reduction Factor Interior exposure CFRP 0.95 GFRP 0.75 AFRP 0.85 Exterior exposure CFRP 0.85 GFRP 0.65 AFRP 0.75 Aggressive environment CFRP 0.85 GFRP 0.50 AFRP 0.70 Source: ACI (2008a). TABLE 21 ENVIRONMENTAL REDUCTION FACTORS

76 reported by the manufacturer. The AASHTO guide specifications state that the factored resistance of FRP-strengthened members and load combinations are calculated according to the AASHTO LRFD BDS (AASHTO 2012a). FRP-strengthened concrete members may fail by either rupture of the FRP or crushing of the concrete. If anchorage is not provided, FRP-debonding controls the failure of the strengthening system. Insufficient development length from a critical section can also cause pre- mature debonding failure. Flexural-Strengthening Design The concept of flexural strengthening is that externally bonded FRP carries tensile stress, in addition to the internal steel reinforcement, to increase the capacity of the strengthened member. Bond is the most critical requirement in FRP strengthening, because bond failure (either partial or complete) reduces the efficacy of strengthening. A capacity increase by FRP strengthening may be limited up to 40% (ACI 440.2R-08) to avoid problems when the FRP fails unexpectedly. The nominal flexural resistance (Mn) of an under-reinforced rectangular section with FRP sheets is given here: M M M A f d a E A h an s frp s y frp frp frp( ) ( )= + = − + ε −2 2 (5) where Ms and Mfrp are the resistances provided by the steel and FRP reinforcement, respectively; As and Afrp are the cross-sectional areas of the steel and FRP, respectively; d and h are the effective depth and height of the section, respectively; fy is the yield strength of the steel reinforcement; a is the depth of the equivalent stress block; Efrp is the elastic modulus of the FRP; and efrp is the strain of the FRP at failure. When calculating a moment capacity, FRP’s thickness is ignored (FRP is bonded to the tensile soffit of the beam). ACI 440.2R-08 (ACI 2008a) and AASHTO (2012a) use an additional strength reduction factor (yf = 0.85) for the Mfrp component in Eq. 5. For design convenience, perfect bond between the FRP and concrete substrate is assumed until debonding failure occurs. This failure mode is controlled by limiting usable FRP strains (i.e., effec- tive strain). Table 22 lists effective strains (ee) taken from published design guidelines. If NSM FRP is used, bond is improved; accordingly, ACI 440.2R-08 (ACI 2008a) provides ee = 0.7efu, where efu is the Source Method AASHTO guide specifications (AASHTO 2012a) 005.0≤≤ u frpe ηεε where = strain limitation coefficient ( = 0.55, 0.3, and 0.2 for CFRP, AFRP, and GFRP, respectively) u frp = characteristic value of tensile strain of FRP. ACI 440.2R-08 (ACI 2008a) fu ff c e tnE fK 9.0 ' ≤= where K = coefficient (0.083 for US customary units and 0.41 for SI units) ' cf = specified compressive strength of concrete n = number of FRP plies fE = tensile modulus of FRP ft = nominal thickness of one ply of FRP. Intelligent Sensing for Innovative Structures (ISIS Canada 2008a) 006.0≤e CSA S806-02 (CSA 2002) 007.0≤e η η ε ε ε ε ε TABLE 22 EFFECTIVE STRAIN (ee) TO AVOID PREMATURE FRP DEBONDING FAILURE OR CREEP-RUPTURE

77 Source Limit AASHTO guide specifications (AASHTO 2012a) c c c E f '36.0≤ ys 8.0≤ where c = concrete strain s = steel strain ' cf = concrete strength in compression cE = elastic modulus of concrete y = yield strain of steel. ACI 440.2R-08 (ACI 2008a) '45.0 cc ff ≤ ys ff 8.0≤ where cf = concrete stress sf = steel stress. fib Bulletin 14 (fib 2001) '60.0 cc ff ≤ ys ff 8.0≤ ε ε ε ε ε ε TABLE 23 STRESS/STRAIN LIMITS UNDER SERVICE LOADING design-rupture strain of the FRP. Strain computation in concrete, steel, and FRP provides an expected failure mode (e.g., steel-yielding followed by concrete crushing, FRP-rupture before concrete crush- ing, and so on). To determine the failure mode of an FRP-strengthened member, existing strains in the concrete and steel are taken into account. Considering safety, the preferred failure mode is concrete crushing after steel-yielding without FRP-debonding or -rupture (end anchorage may be designed to preclude FRP-debonding). FRP-strengthened members are required to demonstrate sufficient ductil- ity before failure occurs, in conjunction with FRP strains equal to or greater than 2.5 times the strain at the yielding of the steel (AASHTO 2012a). Typical cracked-section analysis (i.e., transformed section) is employed to design FRP-strengthened members in service. It is important that the deflection of these beams be within the limits stipulated in the AASHTO LRFD BDS. The contribution of FRP to beam deflection may be ignored before yielding of the steel reinforcement. However, if post-yield behavior is of interest, the effect of FRP is taken into consideration. Steel and concrete stresses or strains are limited under service loading to prevent fatigue failure, as shown in Table 23. FRP sheets are to be extended over a certain length from a critical sec- tion to avoid premature bond failure. Table 24 summarizes existing development length equations for externally bonded FRP sheets. AASHTO guide specifications (AASHTO 2012a) provide an additional item for end-peeling: f fpeel c≤ ′0.065 , where fpeel is the peel stress at FRP–concrete interface. Shear-Strengthening Design Shear strengthening may be required for reinforced concrete members experiencing deficient shear resistance and plastic hinges caused by reversal loading. To improve shear capacity, FRP sheets are bonded to two or more sides of a beam (the function of FRP is the same as that of conventional shear stirrups). The orientation of fibers is perpendicular to the beam span. Strengthening design for shear is largely empirical, as is the case for reinforced concrete, because of the complexity and insufficient understanding of the shear behavior. Three possible strengthening schemes are as follows: • Complete-wrapping: FRP sheets surround the entire surface of a beam through the slits of the drilled slab monolithically constructed with the beam. • U-wrapping: FRP sheets cover three sides of a beam (i.e., two sides and the bottom). • Both-side-bonding: FRP sheets are bonded to both sides of a beam.

78 When FRP sheets are bonded (e.g., U-wraps), the corners of the beam to be strengthened are chamfered to reduce stress concentrations. The nominal shear resistance (Vn) of an FRP-strengthened member may be obtained by: V V V Vn c s f= + + (6) where Vc, Vs, and Vf are the resistances provided by the concrete, steel stirrups, and FRP, respec- tively. ACI 440.2R-08 (ACI 2008a) employs a reduction factor yf to the Vf term (yf = 0.95 and 0.85 for completely wrapped and U-wrapped/side-bonded FRP, respectively). The AASHTO guide specifications (AASHTO 2012a) provide a similar reduction factor of ff = 0.85 for FRP in shear. An angle of 45° may be assumed for diagonal tension cracks (or shear cracks). Eq. 7 shows the shear resistance of FRP (AASHTO 2012a): V A f d s f f fe f f f f ( ) = α + αsin cos (7) where Af is the FRP area covering two sides of the beam; ffe is the effective stress of the FRP; df is the effective depth of the FRP measured from the top of the FRP to the centroid of the longitudinal reinforcement; af is the angle of the FRP inclination; and sf is the center-to-center spacing of the FRP. Because FRP sheets cover a large area, FRP-bonding with an angle (analogous to bent steel bars in Source Equation AASHTO guide specifications (AASHTO 2012a) frp frp d b T L intτ ≥ where frpT = tensile force in FRP corresponding to its strain of 0.005 int = interface shear transfer length ( 'int 065.0 cf= ) frpb = FRP width. ACI 440.2R-08 (ACI 2008a) ' c ff d f tnE DL = where D = constant ( D = 1.0 for metric units and 0.057 for U.S. customary units) n = number of FRP plies fE = tensile modulus of FRP ft = nominal thickness of one ply of FRP ' cf = specified compressive strength of concrete. fib Bulletin 14 (fib 2001) ctm frpf d fc tE L 2 = where frpt = nominal thickness of FRP 2c = calibration constant ( 2c = 2.0) ctmf = concrete tensile strength. Intelligent Sensing for Innovative Structures (ISIS Canada 2008a) mmtEL frpfd 3005.0 ≥= (12 in.) τ τ TABLE 24 DEVELOPMENT LENGTH (Ld) FOR EXTERNALLY BONDED FRP SHEETS

79 reinforced concrete) is not necessary. The reinforcement ratio of FRP in shear is defined as (AASHTO 2012a): ρ = 2 for discrete strips (8a)n t wb sf f f f v f ρ = 2 for continuous sheets (8b)n tbf f f v where nf is the number of the FRP layers; tf is the FRP thickness; wf is the strip width; bv is the effec- tive web width taken as the minimum web width within the effective depth; and sf is the center-to- center spacing of the FRP. FRP strains are controlled to prevent bond failure, as shown in Table 25. Mechanical anchorage may be installed to preclude FRP-debonding. The maximum spacing of FRP sheets for shear strengthening is either 0.8dv (≤ 24 in.) or 0.4dv (≤ 12 in.) for vu < 0.125f ′c and Source Equation AASHTO guide specifications (AASHTO 2012a) fuffe R= for complete U-wrap or with anchors ( ) 0.14088.0 67.0 ≤≤= −fff ER 004.0≤= fuffe R for side bonding or U-wrap ( ) 0.13066.0 67.0 ≤≤= −fff ER where fu = ultimate strain of FRP f = reinforcement ratio of FRP fE = modulus of FRP. ACI 440.2R-08 (ACI 2008a) fufe 75.0004.0 ≤= for completely wrapped member 004.0≤= fuvfe κ for side bonding or U-wrap 75.021 ≤= fu e v S Lkk 3/2 ' 1     = A fk c fv efv d Ld k − =2 for U-wraps fv efv d Ld 2− = for side bonding ( ) 58.0ffe Ent BL = where ' cf = specified compressive strength of concrete A = constant ( A = 4,000 for U.S. customary units and 27 for SI metric units) S = constant ( S = 468 for U.S. customary units and 11,900 for SI metric units) fvd = effective depth of FRP shear reinforcement B = constant ( B = 2,500 for U.S. customary units and 23,300 for SI metric units) n = number of FRP plies fE = tensile modulus of FRP. ε ε ε ε ε κ ε ρ ρ ρ ε ε ε ε TABLE 25 EFFECTIVE FRP STRAIN (efe) FOR SHEAR-STRENGTHENING (continued on next page)

80 Source Equation Intelligent Sensing for Innovative Structures (ISIS Canada 2008a) fufe 75.0≤ for FRP strength 004.0≤fe for aggregate interlock fuvfe K≤ for bond capacity 75.0 11900 21 ≤= fu e v LkkK 3/2 ' 1 27     = cfk fv efv d Ld k − =2 ( ) 58.0 300,23 ff e Ent L = ε ε ε ε ε ε fib Bulletin 14 (fib 2001) fu ff cm fe E f 30.03/217.0     = for completely wrapped member (CFRP)         ×    = – fu ff cm ff cm fe E f E f 30.03/23 56.03/2 17.0,1065.0min for side bonding or U-wrap (CFRP) where cmf = mean compressive strength of concrete. ε ε ε ε ρ ρ ρ TABLE 25 (continued) vu ≥ 0.125f ′c, respectively, where vu is the shear stress and dv is the effective shear depth calculated in accordance with Art. 5.8.2.9 of the AASHTO LRFD BDS. Axial-Strengthening Design The primary purposes of axial-strengthening (or axial-confining) are to increase the axial and flex- ural capacities of reinforced concrete columns and to improve ductility. This strengthening technique is particularly beneficial for bridge substructures (columns) constructed in seismic regions. FRP sheets are bonded around an existing column (fibers are oriented in the circumferential direction of the member). FRP-wrapping can improve the buckling resistance of a column by confining the longitudinal reinforcing bars. Although an intermittent wrapping scheme with discrete FRP sheets is available, columns are circumferentially wrapped with continuous FRP sheets (similar to steel spirals) to maximize confining effects. FRP-wrapping becomes active as soon as the core concrete is significantly damaged and dilates laterally. For this reason, FRP-confinement is referred to as pas- sive strengthening (if prestress is applied, it then becomes active confinement). Circular columns wrapped with FRP sheets do not exhibit stress concentrations; however, square columns experience uneven stress contours owing to the corners, albeit chamfered. The nominal capacity of an FRP-confined column (Pn) may be calculated by (AASHTO 2012a): 0.85 0.85 for spiral reinforcement (9a){ }( ) ( )= ′ − +P f A A f An cc g st y st 0.80 0.85 for tie reinforcement (9b){ }( ) ( )= ′ − +P f A A f An cc g st y st where f ′cc is the compressive strength of the confined concrete; Ag and Ast are the gross sectional area of the concrete and the total area of the longitudinal steel bars, respectively; and fy is the yield

81 Source Equation AASHTO guide specifications (AASHTO 2012a) += c l ccc f fff 21'' −≤= 75.01 2 2 65.0 ' e cfrp l k f D N f where cf = stress in concrete at strain c frpN = strength per FRP width corresponding to a strain of 0.004 D = external diameter of column ek = constant ( ek = 0.80 and 0.85 for tied and spiral columns, respectively). ACI 440.2R-08 (ACI 2008a) lafccc fff ψ 3.3'' += D ntE f feffl 2 = where f = constant ( f = 0.95) a = efficiency factor ( a = 1.0 for circular column) fE = tensile modulus of FRP n = number of FRP plies ft = nominal thickness of one ply of FRP fe = fu55.0 . fib ulletin 14 (fib 2001) += ' '' 32.0 c l ccc f fff fuffl Ef 2 1 = where fρ = volumetric ratio of FRP. Intelligent Sensing for Innovative Structures (ISIS Canada 2008a) lccc fff 2'' += D ft f fuffl 2 = where f = material resistance factor ( 75.0=f ) fuf = ultimate tensile strength of FRP. ε ε ε ε ε φφ φ ρ κ κ κ ψ ψ TABLE 26 CONFINED CONCRETE STRENGTH ( f ′cc) strength of the steel. Table 26 lists existing expressions about the confined strength of concrete f ′cc. The AASHTO equations shown in this table are used with the following limitations: luD < 8 for circular columns and h/b < 1.5 for rectangular columns, where lu and D are the length and diam- eter of the circular column, respectively; and h and b are the larger and smaller dimensions of the rectangular column, respectively. Service load may be limited to avoid the radial cracking of a strengthened column. ACI 440.2R-08 (ACI 2008a) accordingly requires that the concrete and steel stresses be less than 0.65 f ′c and 0.60 fy, respectively. In the case of columns subjected to axial and flexural loads, the resistance of FRP-confinement is not greater than the resistance corresponding to balanced strain conditions (AASHTO 2012a). The load-moment interaction diagram of FRP-confined columns (frequently referred to as the P-M interaction diagram) can be constructed in a manner similar to reinforced concrete columns.

82 Site application FRP strengthening technologies have been used for bridge structures for more than 20 years. The appli- cation of FRP is virtually unlimited from bridge superstructure to substructure. NCHRP Report 609: Recommended Construction Specifications and Process Control Manual for Repair and Retrofit of Concrete Structures Using Bonded FRP Composites (Mirmiran et al. 2008) is a good resource for site engineers to reference. Figure 39 shows typical strengthening projects to upgrade constructed bridge members using FRP composites. FRP strengthening is required to be conducted by skilled installers FIGURE 39 FRP strengthening of deficient bridge members [Lopez and Nanni 2006; Banthia et al. 2010; Yang et al. 2011 (used by permission from American Concrete Institute)]: (a) flexural strengthening with CFRP laminates; (b) CFRP U-wrapping for shear after flexural strengthening. (continued) (a) (b)

83 FIGURE 39 (Continued) FRP strengthening of deficient bridge members [Lopez and Nanni 2006; Banthia et al. 2010; Yang et al. 2011 (used by permission from American Concrete Institute)]: (c) flexural strengthening with NSM CFRP; (d) fully wrapped bridge piers. (continued on next page) (c) (d) (certified personnel preferred) under the direction of experienced professional engineers. It is particu- larly crucial for wet lay-up application, because dry fibers are impregnated in a resin matrix to form a load-bearing composite. The resin is mixed in accordance with the manufacturer’s guidelines to achieve its full capacity (guaranteed strength). Figure 40 exhibits inappropriately installed FRP strengthening systems, which degrade the performance of the strengthened structures. The Gandy Boulevard Bridge in Tampa Bay, Florida, was strengthened with FRP sheets (Mullins et al. 2006). The bridge was constructed in 1956 and has a length of 2.6 miles (4.2 km), including 275 spans. A demonstration project was conducted to strengthen the bridge’s substructure, as shown in Figure 41. The columns [20 in. by 20 in. (510 mm by 510 mm), Figure 41(a)] reinforced

84 (e) (f) FIGURE 39 (Continued) FRP strengthening of deficient bridge members [Lopez and Nanni 2006; Banthia et al. 2010; Yang et al. 2011 (used by permission from American Concrete Institute)]: (e) partially wrapped bridge pier; (f) shear-strengthened prestressed concrete girder. (continued) with Grade 60 steel bars (8-No. 8) were upgraded using unidirectional and bidirectional CFRP sheets. Because the water depth was 16 ft (4.9 m), a scaffolding system consisting of steel grating and angles was installed to secure a work space [Figure 41(b)]. The concrete surface was cleaned to eliminate debris and marine growth [Figure 41(c)]. After applying a primer, unidirectional CFRP sheets were bonded [Figure 41(d)], followed by additional wrapping with two-layer bidirectional CFRP sheets [Figure 41(e)]. The CFRP-confined columns were wrapped with GFRP [Figure 41(f)] and plastic shrink- age sheets [Figure 41(g)]. Upon curing of the strengthening system, the shrinkage sheet was removed and an ultraviolet-resisting layer was applied to protect the system [Figure 41(h)]. Figure 41(i) shows the

85 (g) (h) (i) FIGURE 39 (Continued) FRP strengthening of deficient bridge members [Lopez and Nanni 2006; Banthia et al. 2010; Yang et al. 2011 (used by permission from American Concrete Institute)]: (g) CFRP splice; (h) strengthened pier cap; (i) discrete CFRP wraps.

86 (a) (b) FIGURE 40 Inappropriately installed FRP-strengthening systems [Arnold and Carr 2010 (used by permission from American Concrete Institute)]: (a) misaligned GFRP patch; (b) unbonded CFRP wrap.

87 FIGURE 41 Gandy Boulevard Bridge repair [Mullins et al. 2006 (used by permission from American Concrete Institute)]: (a) piers to be repaired; (b) installing scaffold; (c) cleaning with pressured water. (continued on next page) (a) (b) (c)

88 (d) (e) FIGURE 41 (Continued) Gandy Boulevard Bridge repair [Mullins et al. 2006 (used by permission from American Concrete Institute)]: (d) bonding unidirectional CFRP sheets; (e) wrapping with bidirectional CFRP sheets. (continued) completed site work. CFRP repair was carried out for a bridge severely damaged by alkali–silica reaction (Williams and Choudhuri 2011). The bridge, located in Houston, Texas, is 25 years old and has 15 spans [Figure 42(a)]. Significant cracking was observed owing to alkali–silica reaction, as shown in Figure 42(b). The damaged girders were wrapped with CFRP sheets to confine existing cracks and to restore the capacity [Figure 42(c)]. The CFRP systems were coated for ultraviolet protection. The Scheifele Bridge in Waterloo, Ontario, Canada, was damaged by corrosion. Significant cracks formed along the girder with concrete spalling, as shown in Figure 43(a). The girder had a compres- sive concrete strength of 4,000 psi (27 MPa), and included a cover of 2 in. (50 mm) to protect the

89 (f) (g) FIGURE 41 (Continued) Gandy Boulevard Bridge repair [Mullins et al. 2006 (used by permission from American Concrete Institute)]: (f) adding GFRP wrap; (g) placing plastic shrinkage wrap. (continued on next page) internal steel reinforcing bars. This bridge was first repaired in 1982 to replace the delaminated deck concrete and the damaged expansion joints. In 2004, the bridge was re-examined and a decision was made to repair the corrosion-damaged girder. Figure 43(b) shows the prepared girder substrate after removing the damaged concrete. To fill the void of the girder, a cementitious grout with a compres- sive strength of 7,250 psi (50 MPa) was injected into the installed wooden form. CFRP U-wraps were bonded to wrap the prepared girder, and longitudinal CFRP laminates were added to preclude the potential debonding of the U-wraps [Figure 43(c)]. The J857 bridge in Phelps County, Missouri, was strengthened with NSM CFRP rods and destructively tested to failure, as shown in Figure 44. The 26 ft (8 m) span bridge (three spans) consisted of solid slabs with a thickness of 18.5 in. (460 mm). Prior to conducting the strengthening

90 (h) (i) FIGURE 41 (Continued) Gandy Boulevard Bridge repair [Mullins et al. 2006 (used by permission from American Concrete Institute)]: (h) installing ultraviolet layer; (i) completed repair. work, the bridge was visually inspected and load rated with HS20 and other trucks. A total of 20 CFRP rods [0.44 in. (11 mm) in diameter] were inserted into precut grooves at a spacing of 15 in. (375 mm) to achieve an increase of 30% in the slab’s capacity. After curing the CFRP strengthening system, reaction steel beams were placed for destructive testing. The bridge failed at a load of 596 kips (2,652 kN), including significant interfacial deterioration between the CFRP rods and the concrete. Post-tensioned CFRP sheets were used to strengthen the Main Street Bridge in Winnipeg, Manitoba, Canada (Kim et al. 2006). The bridge, constructed in 1963, was damaged by the colli- sion of heavy trucks, as shown in Figures 45 (a) and (b). Because of the damage, the load-carrying capacity of the exterior girder was reduced by 18%. The concrete-spalled area was filled with a cementitious material to restore the girder shape [Figure 45(c)], and anchor plates were mounted near the ends of the girder [Figure 45(d)]. After placing three layers of CFRP sheets [Figure 45(e)], post-tensioning was conducted using a hydraulic jack [Figure 45(f)]. The post-tensioned CFRP sheets were anchored to the plates with a temporary support system [Figure 45(g)]. All strengthening tasks were completed in a single day [Figure 45(h)]: the girder’s load-carrying capacity was recovered to the level of the undamaged state. This innovative strengthening work simultaneously enhanced the strength and serviceability of the bridge (Kim et al. 2008a), which is not achievable in conventional FRP-strengthening projects.

91 (a) (b) (c) FIGURE 42 Bridge girders damaged by alkali–silica reaction [Williams and Choudhuri 2011 (used by permission from American Concrete Institute)]: (a) overview; (b) cracked girder; (c) CFRP-repair.

92 (a) (b) (c) FIGURE 43 Corrosion-damaged bridge [Rteil and Soudki 2011 (used by permission from American Concrete Institute)]: (a) corrosion-induced cracking and concrete spalling; (b) prepared substrate and mortar pumping; (c) CFRP-strengthened girder.

93 (a) (b) (c) FIGURE 44 NSM CFRP for strengthening a slab bridge [Alkhrdaji et al. 1999 (used by permission from A. Nanni)]: (a) inserting CFRP rods into grooves; (b) grouting CFRP-inserted grooves; (c) completed strengthening. (continued on next page)

94 Various anchor systems are available for post-tensioning CFRP sheets and laminates, as shown in Figure 46. Adjustable or movable mechanical anchors are frequently used to post-tension CFRP lami- nates [Figures 46(a) and (b)] and, although not common, a specially fabricated headed CFRP laminate can be placed reacting against the steel block [Figure 46(c)]. A pair of jacking and fixed anchors is mounted to the diaphragm and girder side of a bridge [Figures 46(d) and (e)]. Cone-shaped anchors are also usable for strengthening a slab bridge [Figure 46(f)]. The Castlewood Canyon Bridge in Colorado, an arch bridge strengthened with CFRP sheets, is 374 ft (11.4 m) long for two traffic lanes. Because of deterioration, the bridge concrete spalled at many locations (i.e., deck, columns, and arches), as shown in Figure 47(a). The loose concrete was removed using a jack hammer, and the reinforcing steel was sandblasted to eliminate potentially detrimental surface residues. After shotcreting the prepared members to restore the cross-sections [Figure 47(b)], an epoxy adhesive was applied for bonding CFRP sheets [Figure 47(c)]. The arch members were longitudinally and transversely strengthened using multiple CFRP layers [Figure 47(d)]. The bridge was then repainted to protect the installed CFRP from moisture and ultravio- let rays. According to follow-up inspections to monitor the long-term performance of the bonded CFRP, regional debonding was noticed after 4 and 8 years of service [Figure 47(e)]. Pull-off bond tests were then performed to examine the in situ bond of the CFRP system [Figure 47(f)]: average bond stresses varied from 300 psi (2.07 MPa) to 423 psi (2.92 MPa). The poor bond was attributable to incomplete wet-out (the carbon fibers were not completely saturated in the resin) and possibly to environmental damage. A deck overhang was strengthened using NSM CFRP strips in California, as shown in Figure 48, along with the installation of a barrier rail. The purpose of the strengthening was to upgrade the overhang and barrier against an impact load of 54 kips (240 kN). An in situ bond test was conducted to evaluate the performance of the installed NSM CFRP. Figures 49(a–c) exhibit GFRP repair for a timber bridge in West Virginia, which was constructed in the 1900s (Frankhauser et al. 2016). Accord- ing to static and dynamic load tests conducted after the repair, the strains of the piles and the pile cap were reduced by 43% and 46%, respectively. Other repair applications for timber bridges are shown in Figures 49(d–f). fiber-reinforced Polymer deckS and SuPerStructure FRP shapes can be an alternative to reinforced concrete bridge decks such as FRP sandwich panels and pultruded sections. Several core forms are available for modular FRP decks (e.g., honeycomb and (d) Failed NSM CFRP FIGURE 44 (Continued) NSM CFRP for strengthening a slab bridge [Alkhrdaji et al. 1999 (used by permission from A. Nanni)]: (d) destructive testing.

95 (a) (b) (d)(c) FIGURE 45 Post-tensioned CFRP strengthening application [Kim et al. 2006 (used by permission from American Concrete Institute)]: (a) overview of bridge; (b) damaged girder; (c) prepared girder; (d) anchorage mounted. (continued on next page)

96 hexagonal structures). Multiple pieces of FRP elements (either open or closed shapes) are mechanically connected or chemically bonded to form a modular deck, as shown in Figure 50. These types of struc- tural systems are suitable for movable bridges where the replacement of a lightweight deck is desired and for accelerated bridge construction with several advantages (e.g., lightweight, fast erection, minimal disruption to traffic, durability with noncorrosiveness, reasonable labor, and reduced maintenance costs). The following performance criteria of FRP decks are proposed by the Ohio DOT (Alagusundaramoorthy et al. 2006): • For flexure: the maximum strain allowed is less than 20% of the ultimate service strain calcu- lated by dead and live loads with a dynamic load allowance; the maximum dead load strain is less than 10% of the ultimate service strain. • For shear: the shear capacity is equal to or greater than the capacity of a reinforced concrete deck; the maximum shear capacity allowed is less than 45% and 100% of the ultimate shear capacity of nonhybrid and hybrid FRP decks, respectively. • For deflection: the deflection limits vary from L/596 to L/792 and from L/851 to L/1,097 for single and continuous decks, respectively, where L is the span length. (f) (e) (g) (h) FIGURE 45 (Continued) Post-tensioned CFRP strengthening application [Kim et al. 2006 (used by permission from American Concrete Institute)]: (e) CFRP sheets placed; (f) post-tensioning CFRP; (g) temporary support installed; (h) completed strengthening work.

97 FIGURE 46 Various anchor systems for post-tensioned CFRP [Basler et al. 2004 (used by permission from Canadian Society for Civil Engineering)]: (a) adjustable mechanical anchorage; (b) movable anchorage; (c) anchor for headed CFRP laminated. (continued on next page) (a) (b) (c)

98 (d) (e) (f) FIGURE 46 (Continued) Various anchor systems for post-tensioned CFRP [Basler et al. 2004 (used by permission from Canadian Society for Civil Engineering)]: (d) jacking anchor with threaded rods; (e) fixed anchor with an anchor bracket; (f) cone-shape anchors.

(a) (b) (c) FIGURE 47 Strengthening an arch bridge [Atadero et al. 2012 (used by permission from Colorado Department of Transportation)]: (a) deteriorated members; (b) shotcrete patching to restore cross section; (c) CFRP bonding. (continued on next page)

100 (d) (e) (f) FIGURE 47 (Continued) Strengthening an arch bridge [Atadero et al. 2012 (used by permission from Colorado Department of Transportation)]: (d) longitudinal and transverse strengthening scheme; (e) local debonding; (f) pull-off bond test.

(a) (b) (c) FIGURE 48 NSM CFRP strengthening for bridge overhang (used by permission from Jim Gutierrez at California Department of Transportation): (a) overview of Slide Canyon Bridge with original bridge barrier rail; (b) removal of existing rail; (c) saw cutting for grooves. (continued on next page)

102 (d) (e) (f) FIGURE 48 (Continued) NSM CFRP strengthening for bridge overhang (used by permission from Jim Gutierrez at California Department of Transportation): (d) installed CFRP for proof testing; (e) CFRP embedment and reinforcing for railing; (f) completed railing and CFRP-strengthening.

103 FIGURE 49 GFRP-repaired timber structures [Watson 2004 (used by permission from Canadian Society for Civil Engineering)]; Frankhauser et al. 2016): (a) pile cap; (b) piles and stringers; (c) piles; (d) columns. (continued on next page) (a) (b) (c) (d)

104 (e) (f) FIGURE 49 (Continued) GFRP-repaired timber structures [Watson 2004 (used by permission from Canadian Society for Civil Engineering)]; Frankhauser et al. 2016): (e) piles; (f) girder.

(a) (c) (b) FIGURE 50 FRP deck structures (Telang et al. 2006): (a) honeycomb; (b) sold form; (c) pultruded hollow core.

106 Prefabricated FRP decks are positioned on top of bridge girders [Figure 51(a)] and connected using anchor bolts [Figure 51(b)] or clips [Figure 51(c)]. For modular FRP bridge decks, connection plays an important role in maintaining their integrity. Alampalli (2006) reported that an average dynamic load allowance of 0.3 was obtained from an in situ test conducted on an FRP superstructure bridge, which was in agreement with the upper limit of 0.3 of the AASHTO Standard Specifications. Triandafillou and O’Connor (2009) noted that 73 bridge decks and 48 superstructures were constructed with FRP composites in the United States. Various types of FRP-based bridges are shown in Figure 52, includ- ing FRP decks, superstructures, and girders. Technical concerns related to the use of FRP decks are as follows (Reising et al. 2004): • Delamination: local damage on FRP decks can propagate with traffic loading, which then causes delamination of the upper deck panels. Sandwich-type configurations are vulnerable to delamination failure. • Temperature: thermal gradients across the deck may result in misalignment of its components (e.g., regional uplift of installed FRP), thereby influencing the connectivity between elements. Thermal discrepancy between the FRP deck and supporting girders is a design consideration to preclude irregular thermal deformation and to preserve the integrity of the bridge system. Figure 53 reveals the selected failure modes of FRP decks. Because of stress concentrations, con- nection areas or joints are susceptible to damage [Figures 53(a) and (b)]. Blistering or delamination can happen when excessive surface traction applies, as shown in Figure 53(c). Punching-type failure also takes place because FRP consists of discrete fibers [Figure 53(d)]. The long-term performance of in situ FRP decks is not fully addressed owing to a lack of technical data, along with the relatively short application history of FRP decks. Pedestrian bridges can be erected with FRP composites. AASHTO published guide specifications to assist with the design of FRP pedestrian bridges (AASHTO 2008). These specifications comprise three chapters: • Introduction: the scope of the specifications is stated in conjunction with the AASHTO Standard Specifications for Highway Bridges (17th ed.). • Design loads: three load types are specified: (1) pedestrian load for main and secondary bridge members, (2) maintenance vehicle load, and (3) wind load to be horizontally applied to the bridge. • Design details: allowable stresses are stipulated alongside deflection and vibration require- ments. Also included are the minimum thickness of FRP [0.25 in. to 0.375 in. (6.4 mm to 9.6 mm)] and connection methods with galvanized or stainless steel bolts. Recent research shows the concept of a hybrid FRP I-girder (consisting of GFRP and CFRP layers) connected with a high-performance concrete deck, which may be useful to carry the light load associ- ated with pedestrian bridges (Nguyen et al. 2015). Zatar et al. (2014) examined the deflection of the hybrid girder system and stated that it was a good candidate for accelerated bridge construction. The hybrid-girder research was extended to examine connection details (Mutsuyoshi et al. 2016). Paulotto et al. (2014) reported a 708-ft (216-m) long pedestrian bridge constructed with a reinforced concrete deck (prefabricated slabs) carried by CFRP cables, as shown in Figure 54. other aPPlicationS FRP shapes filled with concrete may be employed for bridge girders, piles, and other load-bearing ele- ments. Because the shapes are permanently placed with concrete, additional formwork is not necessary and, thereby, construction costs are saved. Several site projects with FRP stay-in-place forms are shown in Figure 55. The FRP protects core concrete from the environment and carries tensile and compres- sive stresses. Because of material costs, GFRP composites are used in preference to CFRP. Unlike FRP strengthening discussed in “FRP-Strengthened Concrete Members,” multidirectional fiber orientations with multiple stacking are adopted for FRP shapes. The properties of the FRP are thus controlled by fiber angles, stacking sequences, resin thickness, and fiber volumetric ratios. Classical laminate theory is applicable when determining FRP’s properties. Hybrid applications with multiple composite segments

107 (a) (b) (c) FIGURE 51 Installation of FRP decks (Telang et al. 2006): (a) erection; (b) anchor holes; (c) clip connection between girder and deck.

(a) (b) (c) FIGURE 52 Various FRP decks and superstructures constructed in the United States [Triandafillou and O’Connor 2009 (used by permission from Jerome O’Connor)]: (a) FRP deck and girders; (b) FRP shape-panel assembly; (c) FRP superstructure bridge. (continued)

109 (d) (e) (f) FIGURE 52 (Continued) Various FRP decks and superstructures constructed in the United States [Triandafillou and O’Connor 2009 (used by permission from Jerome O’Connor)]: (d) FRP modular bridge; (e) FRP slab bridge; (f) replacing of concrete sidewalk with FRP panels.

110 (a) (b) (c) (d) FIGURE 53 Failure of FRP decks (Telang et al. 2006): (a) cracking at deck-to-deck joint; (b) leaking at joint; (c) delamination; (d) punching failure.

111 may be exploited. The behavior of FRP stay-in-place members is generally similar to that of traditional structural members. Figure 56 shows the responses of FRP stay-in-place members. The low modulus of FRP affects the stress-strain relationship [Figure 56(a)] and load-displacement [Figure 56(b)] of the member compared with its steel counterpart. For bridge deck application, various FRP stay-in-place members are available such as I beams and corrugated shapes [Figure 57(a)]. Upon casting concrete, these members constitute a single load-bearing system. The integrity of the system is achieved by transversely connecting the FRP elements, so that local failure is mitigated (flexural and shear stresses are distributed). In situ monitoring revealed that the transverse movement of FRP stay-in-place forms was insignificant relative to their supporting girders (Reising et al. 2004). The detailing of FRP stay-in-place forms is different from that of ordinary reinforced concrete decks (Nelson et al. 2014). Geometric discontinuities are often required to place FRP forms on top of bridge girders, as illustrated in Figure 57(b). Despite this, existing design approaches can be usable to predict the capacity and behavior of FRP stay-in-place decks (Hanus et al. 2009). When an FRP stay-in-place system fails in flexure (cracking and crushing of the filled concrete), the FRP form is not typically damaged (Hanus et al. 2009). If the system fails by punching shear, a (a) (b) FIGURE 54 Pedestrian bridge with CFRP cables [Paulotto et al. 2014 (used by permission from International Institute for FRP in Construction)]: (a) installed CFRP cables; (b) completed work.

(b) (c) (a) FIGURE 55 Application of FRP stay-in-place members [Ji et al. 2004; Nelson et al. 2014; Frankhauser et al. 2016 (used by permission from American Concrete Institute and Canadian Society for Civil Engineering)]: (a) erected concrete-filled FRP tubes; (b) concrete-filled FRP tubes for arch bridge (underpass); (c) FRP trough form and erection. (continued)

113 (d) FIGURE 55 (Continued) Application of FRP stay-in-place members [Ji et al. 2004; Nelson et al. 2014; Frankhauser et al. 2016 (used by permission from American Concrete Institute and Canadian Society for Civil Engineering)]: (d) installation of corrugated FRP shapes. FIGURE 56 Behavior of FRP stay-in-place members: (a) axial response (reproduced based on Samaan et al. 1998); (b) flexural response (reproduced based on Kim and Fam 2011). (a) (b) gradual load drop occurs (Fam and Nelson 2012), rather than the abrupt drop observed in traditional reinforced concrete decks; the reason being that the transversely integrated FRP forms impede the system’s energy release. Bond between the FRP form and concrete is important to preserve the integrity of the deck (Hall and Mottram 1998). The slip between the concrete and FRP form appears to be negligible (Kim and Fam 2011; Fam and Nelson 2012); nonetheless, shear studs may be installed (Reising et al. 2004). The composite action between the FRP form and concrete, whether intended or unintended, influences the stiffness and therefore the deflection of the bridge deck (Reising et al. 2004). Under service loading, FRP stay-in-place forms may demonstrate substan- tially low strains. For instance, Fam and Nelson (2012) tested a full-scale bridge deck consisting of corrugated GFRP forms filled with concrete, and the forms showed a maximum service strain of less than 2% of the FRP’s ultimate strain without any damage in splice. The dynamic load allow- ance (IM) of bridge decks comprising FRP stay-in-place forms may vary from 0.12 to 0.42 (Reising et al. 2004). Further information on FRP cast-in-place forms is available in a state-of-the-art review article (Nelson et al. 2014).

114 FIGURE 57 FRP stay-in-place members [Nelson et al. 2014 (used by permission from American Concrete Institute)]: (a) various types; (b) detailing methods. (a) (b) Mirmiran and Shahawy (2003) showed the field application of a concrete-filled FRP tube replac- ing traditional prestressed concrete piles. The FRP jacket is expected to extend the service life of the pile when subjected to aggressive marine environments in Florida. This nonconventional pile system provides superior performance to conventional piles with a thick concrete cover or with epoxy-coated steel bars, which will eventually cause maintenance problems. The 26-ft (7.9-m) long FRP pile was subjected to a driving force [Figure 58(a)] and installed down to a depth of 22 ft (6.8 m), as shown in Figure 58(b). According to visual examinations, there was no damage in the FRP and concrete [Figure 58(c)], although the pile was hammered more than 900 times. Fam et al. (2003) reported another project using concrete-filled GFRP tubes for a bridge in Virginia. The precast piles had a diameter of 24.6 in. (625 mm) and were tested in a laboratory [Figure 59(a)] to confirm their structural performance. The composite piles were fabricated and shipped to the site [Figure 59(b)] and installed [Figure 59(c)]. Reinforced cap beams were then cast on top of the piles to transfer traffic load from the superstructure [Figure 59(d)]. The costs of the projects were $95/ft ($311/m) for composite piles and $20/ft ($65/m) for pile driving.

115 (a) (b) (c) Driving Helmet Driving Cap Composite Pile FIGURE 58 Concrete-filled FRP pile [Mirmiran and Shahawy (2003) (used by permission from American Concrete Institute)]: (a) driving device; (b) driving in progress; (c) installed pile.

116 FIGURE 59 GFRP stay-in-place piles [Fam et al. 2003 (used by permission from Prestressed Concrete Institute and Sami Rizkalla)]: (a) laboratory testing; (b) fabrication and handling; (c) driving for installation. (continued) (a) (b) (c)

117 (d) FIGURE 59 (Continued) GFRP stay-in-place piles [Fam et al. 2003 (used by permission from Prestressed Concrete Institute and Sami Rizkalla)]: (d) completed view.

Next: Chapter Six - Life-Cycle Cost Analysis »
Use of Fiber-Reinforced Polymers in Highway Infrastructure Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Synthesis 512: Use of Fiber-Reinforced Polymers in Highway Infrastructure documents the current state of the practice in the use of fiber-reinforced polymers (FRPs) in highway infrastructure. The synthesis identifies FRP applications, current research, barriers to more widespread use, and research needs. The objectives of the study are to synthesize published literature on FRP materials in highway infrastructure and to establish the state of current practice of FRP applications in transportation agencies.

  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!