THE VIABILITY OF PARTIALLY POST-TENSIONED CONCRETE MEMBERS  IN AN AGGRESSIVE ENVIRONMENT, INCLUDING CYCLIC LOADING

ABSTRACT

Partially prestressed concrete offers several advantages over fully prestressed designs including increased ductility, increased energy absorption, decreased cost, decreased camber, and decreased anchor zone congestion.  However, design codes have been slow to adopt provisions for the design of partially prestressed concrete because of concerns over fatigue and corrosion.  The overall objective of this study was to determine the viability of partially post-tensioned concrete members in an aggressive environment, including the impact of cyclic loading.

This objective was investigated through an experimental program, which included both individual strand testing and full-scale, post-tensioned beam specimens.  The strand testing phase consisted of corroding samples to different amounts of section loss, and then testing them under cyclic loading until failure.  The beam testing phase included 12 full-scale, post-tensioned beam specimens combining long-term static exposure testing with fatigue testing.  One of the most important limitations of the study was that the experimental program followed the specific sequence of exposure (corrosion) followed by fatigue to failure.

With regard to the individual strand testing, there was a significant drop-off in fatigue capacity at relatively small pit depths.  On average, the fatigue capacity was reduced 50 percent for pitting that measured 0.010 to 0.015 inches in depth.  Furthermore, an empirical relationship based on an exponential decay function was found to be the most reliable method of predicting the response of corroded prestressing strand for the specific sequence of corrosion followed by fatigue to failure.  In particular, average pit depth offered the best correlation between fatigue capacity and the amount of localized corrosion.

With regard to the full-scale beam testing, there were several conclusions reached as to the behavior of the tendon – duct, grout, and strand – under exposure and fatigue testing.  First, a robust plastic duct is required as it serves as the primary protection method for the tendon and also performs very well under cyclic loading.  Furthermore, the plastic duct requires steel or plastic saddles at tendon deviation points to eliminate the potential for puncturing the duct during the strand stressing operation.  Next, without full encapsulation of the strand by the grout, the tendon will act as a conduit for chloride transport, thus spreading the potential for corrosion from a single breach in the duct.  Furthermore, even within a well grouted tendon, the grout will contain both longitudinal and circumferential cracking, reducing the ability of the grout to protect the strand.  Finally, grouting defects, such as voids and fine cracks, do not adversely affect fatigue performance of the strand, but these defects do adversely affect corrosion protection of the strand, allowing chlorides access to the strand.

A set of best practices are also included in the recommendation section for the durability of partially post-tensioned concrete members in an aggressive environment, including cyclic loading.  These best practices focus on tension and compression stress limits for the concrete, detailing issues for the tendon, and areas of concern with regard to the anchorage zones and long-term durability.  With regard to the study’s overall objective, partially post-tensioned concrete is a viable construction method in an aggressive environment, even with cyclic loading, but it relies on a robust plastic duct system as the

primary protection method for the tendon.

TABLE OF CONTENTS

LIST OF FIGURES ……………………………………………………………………………………….. viii LIST OF TABLES …………………………………………………………………………………………. xiii

ACKNOWLEDGEMENTS …………………………………………………………………………….. xv

Chapter 1  Introduction …………………………………………………………………………………… 1

1.1 Background and Motivation ………………………………………………………………… 1

1.2 Problem Statement ……………………………………………………………………………… 3

1.3 Scope of Study …………………………………………………………………………………… 5

1.4 Objectives and Scope of Work …………………………………………………………….. 7

1.5 Research Plan …………………………………………………………………………………….. 8

1.6 Thesis Organization ……………………………………………………………………………. 9

1.7 Prestressing Terminology ……………………………………………………………………. 9

Chapter 2  Literature Review …………………………………………………………………………… 12

2.1 Fabrication of Prestressing Strand ………………………………………………………… 12

2.2 Corrosion of Prestressing Strand ………………………………………………………….. 15

2.3 Corrosion-Induced Failure of Prestressing Strand …………………………………… 19

2.4 Individual Strand Fatigue Testing ………………………………………………………… 22

2.5 Corrosion Fatigue ………………………………………………………………………………. 25

2.6 Fatigue in Post-Tensioned Members …………………………………………………….. 27

2.7 Chloride-Induced Corrosion in Post-Tensioned Members ……………………….. 28

2.8 Grouts for Bonded Post-Tensioning ……………………………………………………… 30

2.9 Level of Prestress ……………………………………………………………………………….. 31

2.10 Unanticipated Loss of Prestress ………………………………………………………….. 32

2.10.1 Construction Defects ……………………………………………………………….. 33 2.10.2 Material Defects ……………………………………………………………………… 34

2.10.3 Additional Factors …………………………………………………………………… 35

2.11 Code Provisions and Committee Recommendations ……………………………… 36

2.12 Summary …………………………………………………………………………………………. 37

Chapter 3  Individual Strand Testing ………………………………………………………………… 40

3.1 Corroding of Strand Samples ……………………………………………………………….. 41

3.2 Characterization and Measurement of Corrosion ……………………………………. 46

3.3 Fatigue Test Setup, Test Parameters, and Test Results ……………………………. 55

3.4 Corrosion-Fatigue Relationship ……………………………………………………………. 61

3.4.1 Linear Elastic Fracture Mechanics Approach ……………………………….. 64

3.4.1.1 Non-Corroded Strand ……………………………………………………… 67

3.4.1.2 Corroded Strand – Modify Y ……………………………………………. 68

3.4.1.3 Corroded Strand – Modify Stress Range …………………………… 70

3.4.2 Empirical Approach …………………………………………………………………… 71

3.4.3 Discussion ………………………………………………………………………………… 76

3.4.3.1 Seven Wire Strand versus Individual Wire ………………………… 76

3.4.3.2 Fretting Fatigue ……………………………………………………………… 78

3.4.3.3 Defect Location, Pit Location, and Fracture Location …………. 80

3.4.3.4 Hydrogen Embrittlement ………………………………………………… 82

3.4.3.5 Previous Fatigue Testing of Noncorroded Strand ……………….. 83

3.5 Summary …………………………………………………………………………………………… 83

Chapter 4  Beam Specimen Design and Construction …………………………………………. 85

4.1 Beam Specimen Concept …………………………………………………………………….. 85

4.2 Beam Specimen Test Matrix ……………………………………………………………….. 89

4.3 Beam Specimen Design ………………………………………………………………………. 95

4.3.1 Beam Tendon and Mild Steel Design …………………………………………… 95

4.3.2 Local and General Zone Design ………………………………………………….. 100

4.3.3 Deviator Design ………………………………………………………………………… 102

4.3.4 Post-Tensioning Bar and Anchorage Design ………………………………… 104

4.3.5 Corbel Design …………………………………………………………………………… 105

4.3.6 Beam Specimen Design Summary ………………………………………………. 106

4.4 Beam Specimen Construction ………………………………………………………………. 107

4.5 Beam Specimen Instrumentation ………………………………………………………….. 108

Chapter 5  Beam Specimen Testing ………………………………………………………………….. 114

5.1 Determination of Strand Stress …………………………………………………………….. 114

5.1.1 Prior to Fatigue Testing ……………………………………………………………… 115

5.1.2 During Fatigue Testing ………………………………………………………………. 128

5.1.2.1 Moment-Curvature Analysis ……………………………………………. 129

5.1.2.2 Finite Element Analysis ………………………………………………….. 132

5.2 Exposure Testing and Monitoring ………………………………………………………… 136

5.3 Fatigue Testing and Observations ………………………………………………………… 143

5.4 Summary …………………………………………………………………………………………… 148

Chapter 6  Beam Specimen Forensic Investigation …………………………………………….. 152

6.1 General Findings from Forensic Investigation ……………………………………….. 152

6.1.1 Grout Coverage ………………………………………………………………………… 153 6.1.2 Grout Cracking …………………………………………………………………………. 154

6.1.3 Strand Location within Duct ………………………………………………………. 156

6.1.4 Plastic Duct Performance at Deviators …………………………………………. 159

6.2 Specific Findings from Forensic Investigation ……………………………………….. 161

6.2.1 Fatigue Reference Series ……………………………………………………………. 161

6.2.2 Anchor Zone Grout Void Series ………………………………………………….. 162

6.2.3 Anchor Zone Grout Void Corrosion Series …………………………………… 163

6.2.4 Crown Grout Void Corrosion Series ……………………………………………. 164

6.3 Summary …………………………………………………………………………………………… 167

Chapter 7  Findings, Conclusions, and Recommendations …………………………………… 170

7.1 Findings ……………………………………………………………………………………………. 170

7.2 Conclusions ……………………………………………………………………………………….. 174

7.3 Recommendations ………………………………………………………………………………. 177 Bibliography ………………………………………………………………………………………………….

Chapter 1

 

Introduction

1.1 Background and Motivation

Freyssinet’s original design concept for prestressed concrete maintained the concrete in compression under full service load (Billington 2004, D’Arcy et al. 2003).  In fact, Billington (2004) notes that “for Freyssinet, ‘the fields of prestressed concrete and reinforced concrete have no common frontier.’  Either a structure is fully prestressed or it is not to be called prestressed concrete.”  Prestressed concrete design has evolved over time to allow a nominal amount of tension in the concrete under full service load.  Even with this nominal amount of tension, the design is still considered fully prestressed provided the tensile stresses do not exceed the modulus of rupture for the concrete.

As part of the evolution of prestressed concrete, designers, code writers, and researchers have begun to examine the potential for utilizing even lower levels of prestressing that allow the concrete to crack under full service load.  This type of prestressing is often referred to as partial prestressing.  One of the primary reasons for the potential use of partial prestressing is that ultimate strength generally never governs the design of fully prestressed sections.  Instead, the design is controlled by tension stress limits at full service load.  However, that approach may be overly conservative, even with respect to serviceability concerns, if the full live load is infrequently reached during the life of the structure.

The advantages of partial prestressing over fully prestressed designs include increased ductility, increased energy absorption, decreased cost, decreased camber, and decreased anchor zone congestion (Lin and Burns 1981, Nilson 1987).  However, design codes have been slow to adopt provisions for the design of partially prestressed concrete because of concerns over fatigue and corrosion.

The current edition of the ACI Building Code (ACI 318 2008) now recognizes three classes of prestressed flexural members, U, T, and C.  Under full service load, Class

U members are considered uncracked, Class C members are considered cracked, and

Class T members represent a transition between cracked and uncracked behavior (ACI 318 2008).  The dividing lines between the three classes are based on nominal tensile stresses in the precompressed tensile zone of 7.5  and 12  (ACI 318 2008).

Since they will crack under service load, Class C members and some Class T members are, in essence, partially prestressed flexural members.  Consequently, these members are susceptible to fatigue and corrosion.

In order to properly design partially prestressed flexural members in aggressive environments, research is required to understand the role of fatigue and corrosion on these types of members.  Furthermore, repair of existing prestressed concrete members damaged due to corrosion requires an understanding of the remaining capacity of the section prior to designing a fix, such as fiber-reinforced plastic (FRP) composites or external prestressing.

1.2 Problem Statement

This research program proposed to investigate the viability of partially posttensioned concrete members in an aggressive environment, including the impact of cyclic loading.  Corrosion of the nation’s transportation infrastructure is a widespread and costly problem (Koch et al. 2001).  Structures are deteriorating at a faster rate than they can be repaired or replaced.  The most prevalent form of corrosion in highway structures is chloride-induced corrosion of steel.  Limited research has been performed in the area of cyclic loading in post-tensioned members, and the addition of a corrosive environment to this problem has not been studied significantly.

Post-tensioning has many potential benefits, including crack control and rapid construction with minimal traffic interference when combined with precast members.  Although the concept of post-tensioning is not new, post-tensioning as it stands today is a relatively new form of construction, having been used in bridge structures in the United States for less than 50 years.  At this stage in development, construction practices and materials are continuously improving.  And, it is important that durability of the structure be considered during this development process.  In particular, chloride-induced corrosion is a very real concern for all types of bridges.

An additional concern for bridge members is fatigue.  Repeated loadings can cause fatigue damage of the reinforcement as well as increased crack widths and deflections (Balaguru 1981, Naaman 1982).  Traditional fully prestressed members are uncracked at service loads and fatigue would not typically be a problem with these sections.  However, the economic benefits of partially prestressed members are now becoming apparent.  Unfortunately, partially prestressed members may be cracked at service loads and are therefore more susceptible to fatigue than fully prestressed members or cracked nonprestressed members.

Partially post-tensioned members are more susceptible to fatigue due to two reasons.  First, they experience higher stress changes in the steel as compared to fully posttensioned members or cracked non-prestressed member (Naaman 1982).  Second, posttensioned members have the additional concerns of fretting between the strands and fretting between the strands and duct (Rigon and Thurlimann 1985, Wollman et al. 1988).

Fully post-tensioned members may also be susceptible to durability and fatigue problems due to excessive loss of prestress.  This unanticipated prestress loss can occur due to a number of factors, such as construction defects, material defects, overload, corrosion, and retrofitting.  This loss of full prestress does not necessarily condemn the structure.  In truth, the structure typically still possesses sufficient strength to support load over some service life – although likely not the full service life that was originally intended.  However, no design guidelines exist to assist the engineer in determining what that safe service life is.

Larger crack widths from fatigue loading combined with aggressive agents such as chlorides must be investigated in order for partially post-tensioned members to be a safe and viable option in bridges and to provide guidance for fully post-tensioned structures with a loss of prestress beyond traditional losses.  Corrosion protection of the posttensioning system is vital to the integrity of the structure because loss of post-tensioning can result in catastrophic failure.

1.3 Scope of Study

This research program proposed to investigate the viability of partially posttensioned concrete members in an aggressive environment, including the impact of cyclic loading.  This objective was investigated within a limited scope to reduce the number of variables to a manageable level.  One of the most important limitations was that the experimental program followed the specific sequence of exposure (corrosion) followed by fatigue to failure.  The experimental program included both individual strand testing and full-scale, post-tensioned beam specimens.

The individual strand testing examined the effect of corrosion on the fatigue capacity of prestressing strand for the specific sequence of corrosion followed by fatigue to failure.  The methodology developed for this portion of the study involved the following steps: (1) corrode the strand samples to different amounts of section loss; (2) characterize the amount and disposition of the corrosion; (3) test the strand samples under cyclic loading until failure; and (4) determine a relationship between the amount of strand corrosion and the resulting fatigue capacity.  Both analytical and empirical approaches were studied to develop this corrosion-fatigue relationship.  The strand testing used Grade 270, lowrelaxation strand taken from the same spool.

The full-scale beam testing also examined the effect of corrosion on the fatigue capacity of prestressing strand for the specific sequence of exposure (corrosion) followed by fatigue to failure.  After construction, the specimens were loaded under full service load and exposed to chlorides for a period of 6 months.  At that point, the specimens were placed into a load frame and tested in fatigue to 2,000,000 cycles or failure, whichever came first.  Following testing, the specimens were autopsied for signs of corrosion and fatigue damage.

Full-scale beam specimens were necessary in order to evaluate actual posttensioning hardware and the effect of multi-strand tendons.  The full-scale beam test variables included the level of prestress, condition of the post-tensioning system, and exposure to saltwater.  The full prestress design was based on a nominal tensile stress in the precompressed tensile zone of 3  (psi).  The partial prestress design was based on 67 percent of the force required for full prestress, with one specimen using fewer strands and the other using a lower prestressing force.  The specimen with the lower prestressing force was to mimic a fully prestressed design that experienced excessive prestress losses or overload.

The post-tensioning system for the full-scale specimens followed current design practice and construction standards.  The beams used a multi-strand tendon within a plastic duct filled with a prepackaged grout.  The tendon followed a two-point drape profile with steel deviator pipes and Grade 270, low-relaxation strand.  All anchorage hardware was supplied by a post-tensioning hardware manufacturer, and stressing of the tendons was accomplished with a standard, multi-strand tensioning jack.  As part of the study, imperfections in the grout were introduced to determine their effect on durability and fatigue behavior of the tendon.

Finally, findings were presented, and conclusions and recommendations were developed based on the scope of the research program.

1.4 Objectives and Scope of Work

The overall objective of this study was to determine the viability of partially posttensioned concrete members in an aggressive environment, including the impact of cyclic loading.  This goal was accomplished by achieving the following specific objectives:

1. Develop a relationship between the amount of corrosion on the prestressing strand and the resulting fatigue capacity for the specific sequence of corrosion followed by fatigue until failure.
2. Determine the potential of the relationship developed in Item 1 to predict the response of full-scale, partially post-tensioned beam specimens for the same sequence of corrosion followed by fatigue until failure.
3. Characterize the performance of the tendon – duct, grout, and strand – under exposure to chlorides followed by cyclic testing of full-scale, partially posttensioned beam specimens.
4. Quantify the behavior of flexural cracks under exposure to chlorides followed by cyclic testing of full-scale, partially post-tensioned beam specimens.
5. Determine the role of grouting defects on the durability and fatigue performance of prestressing strand for full-scale, partially post-tensioned beam specimens.
6. Develop a set of best practices for the durability of partially post-tensioned concrete members in an aggressive environment, including cyclic loading.

 

The following scope of work was used to pursue these research objectives: (1) review of the applicable literature; (2) development of a research plan; (3) testing of individual strand specimens and development of a predictive model for the response;  (4) design, construction, and testing of full-scale, partially post-tensioned beam specimens; (6) forensic investigation of the full-scale beam specimens; (7) development of findings, conclusions, and recommendations; and (8) preparation of this dissertation documenting the results of the research study.

1.5 Research Plan

The research plan necessary to accomplish the objectives discussed in Section 1.4 involved individual strand testing and full-scale beam testing, and also included the use of several analytical tools during the evaluation of the test results.  It is important to note that to narrow the study to a manageable number of variables, the testing followed the specific sequence of exposure (corrosion) followed by fatigue to failure.  Research has shown that the fatigue capacity of prestressed concrete is a function of either the fatigue capacity of the prestressing strand or the mild steel.  Therefore, it was necessary to examine the strand alone and to evaluate its response to fatigue and the sequence of corrosion followed by fatigue to failure.  The second experimental aspect involved testing of full-scale beam specimens.  Full-scale beam specimens were necessary in order to evaluate actual post-tensioning hardware and the effect of multi-strand tendons.  In support of the experimentation, various analytical tools were applied during evaluation of the experimental test results, including statistical methods, Linear Elastic Fracture Mechanics (LEFM), and Finite Element Analyses (FEA).

1.6 Thesis Organization

In this investigation, experimental and analytical procedures were performed in order to develop a better understanding of the interaction of corrosion and fatigue on partially post-tensioned concrete members.  The different tasks conducted during this study are organized as chapters in this dissertation.  Chapter 1 presents an introduction to the subject, highlighting the problem, objectives, and scope.  Chapter 1 also presents the research plan deemed necessary to achieve these objectives.  Chapter 2 presents a literature review of the current body of knowledge pertaining to partially post-tensioned concrete members.  Chapter 3 presents the testing of individual strand specimens, including evaluation of the results.  Chapter 4 presents details on the design and construction of full scale, post-tensioned beam specimens.  Chapter 5 presents the exposure and fatigue testing of the full scale, post-tensioned beam specimens.  Chapter 6 presents the subsequent autopsy results of the full scale, post-tensioned beam specimens.  The findings, conclusions and recommendations are presented in Chapter 7.

1.7 Prestressing Terminology

Prestressed concrete refers to the construction method by which a compressive stress is placed on the concrete prior to the application of load.  Precompression is extremely beneficial to concrete, which is inherently strong in compression but relatively weak in tension.  There are two categories of prestressed concrete, pretensioned and posttensioned.  The fundamental difference between the two involves the sequence of strand stressing and concrete placement.  In pretensioned concrete, the strand is stressed in tension prior to placement and curing of the concrete.  In post-tensioned concrete, the concrete is placed and cured prior to tensioning of the strand.

Although the two prestressed categories behave in almost identical ways to the application of load, post-tensioned concrete has numerous variations in the way the prestressing force is applied to the concrete through the tendon.  These tendon variations include the following: (1) grouted versus ungrouted; (2) internal versus external; and (3) bonded versus unbonded.

The difference between grouted and ungrouted tendons involves the method of protection of the tendon.  Grouted tendons are placed within either a steel or plastic duct that has been cast into the concrete section.  Once the concrete reaches sufficient strength, the tendons are tensioned, and then a cementitious grout is pumped into the duct to protect the tendon from the elements.  Ungrouted tendons are greased and sheathed in a plastic coating during manufacture.  Ungrouted tendons are placed within the concrete formwork on site.  After the concrete is placed and reaches sufficient strength, the ungrouted tendons are tensioned.  The grease and plastic sheathing provides protection of the tendon from the elements.

The difference between internal and external tendons involves their location relative to the concrete cross-section of the prestressed member.  Internal tendons are located within the confines of the concrete cross-section, while external tendons are located outside of the concrete cross-section.  External tendons include those that are located within the open area inside a concrete box, even though they are not visible “externally.”

The difference between bonded and unbonded tendons significantly affects the strain in the tendon during application of load.  Bonded tendons are continuously connected to the concrete cross-section such that the tendon and concrete experience the same strain at each cross-section of the beam (strain compatibility).  An unbonded tendon is only connected to the concrete cross-section at intermittent locations.  Consequently, the strain in an unbonded tendon as a function of load is member-dependent, not sectiondependent.  When friction loss is excluded, the strain in an unbonded tendon due to load is uniformly distributed along the tendon length.

The final prestressed concrete term discussed in this section involves the difference between fully prestressed and partially prestressed.  For the purposes of this study, the dividing line between fully prestressed and partially prestressed is based on a nominal tensile stress in the precompressed tensile zone of 7.5  .  This value corresponds to the modulus of rupture for concrete (ACI 318 2008).

THE VIABILITY OF PARTIALLY POST-TENSIONED CONCRETE MEMBERS  IN AN AGGRESSIVE ENVIRONMENT, INCLUDING CYCLIC LOADING

Sharing is caring!