USF Libraries
USF Digital Collections

Corrosion of dual coated reinforcing steel with through-polymer breaks in simulated concrete pore solution

MISSING IMAGE

Material Information

Title:
Corrosion of dual coated reinforcing steel with through-polymer breaks in simulated concrete pore solution
Physical Description:
Book
Language:
English
Creator:
Accardi, Adrienne
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Zinc
Impedance
Disbondment
Epoxy
Polarization
Dissertations, Academic -- Mechanical Engineering -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: This investigation is an examination of the behavior of dual coated reinforcing steel (DCR) with defects in the polymer coating exposing the only zinc layer in simulated concrete pore solution with and without chlorides. The intentional defects simulated the condition typically experienced by the rebar in service. Specimens were tested at open circuit potential, +100 mV, -500 mV, and -1000 mV for 30 to 100 days. The results were compared with that from previous DCR investigation with to-steel defects and epoxy-coated rebar (ECR). DCR with to-zinc defects had extensive corrosion damage when under strong anodic polarization and exposed to chlorides and was similar to that seen for DCR with to steel defects. The freely corroding (OCP) to-zinc DCR specimens in solutions both with and with no-chlorides experienced initially very active dissolution which ended after ~1 day. The zinc exposed at the coating breaks was not completely consumed even after 100 days and there was no visible corrosion product accumulation. This may be due to the formation of a calcium hydroxyzincate passive film and shows that the zinc passivates in alkaline solutions without the benefit of a crevice environment. The DCR with to-steel defects and the DCR with to-zinc defects had similar amounts of disbondment for all test conditions. Notable disbondment was seen only in highly anodic polarization regime with chlorides and was due to large amounts of solid corrosion product formation. These results suggest then that the overall process of zinc wastage in DCR in concrete pore water is not likely to be rapid, which would be beneficial to extending the period in which the barrier and galvanic properties of the zinc are maintained.
Thesis:
Thesis (M.S.M.E.)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Adrienne Accardi.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0003457
usfldc handle - e14.3457
System ID:
SFS0027772:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003457
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Accardi, Adrienne.
0 245
Corrosion of dual coated reinforcing steel with through-polymer breaks in simulated concrete pore solution
h [electronic resource] /
by Adrienne Accardi.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Thesis (M.S.M.E.)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: This investigation is an examination of the behavior of dual coated reinforcing steel (DCR) with defects in the polymer coating exposing the only zinc layer in simulated concrete pore solution with and without chlorides. The intentional defects simulated the condition typically experienced by the rebar in service. Specimens were tested at open circuit potential, +100 mV, -500 mV, and -1000 mV for 30 to 100 days. The results were compared with that from previous DCR investigation with to-steel defects and epoxy-coated rebar (ECR). DCR with to-zinc defects had extensive corrosion damage when under strong anodic polarization and exposed to chlorides and was similar to that seen for DCR with to steel defects. The freely corroding (OCP) to-zinc DCR specimens in solutions both with and with no-chlorides experienced initially very active dissolution which ended after ~1 day. The zinc exposed at the coating breaks was not completely consumed even after 100 days and there was no visible corrosion product accumulation. This may be due to the formation of a calcium hydroxyzincate passive film and shows that the zinc passivates in alkaline solutions without the benefit of a crevice environment. The DCR with to-steel defects and the DCR with to-zinc defects had similar amounts of disbondment for all test conditions. Notable disbondment was seen only in highly anodic polarization regime with chlorides and was due to large amounts of solid corrosion product formation. These results suggest then that the overall process of zinc wastage in DCR in concrete pore water is not likely to be rapid, which would be beneficial to extending the period in which the barrier and galvanic properties of the zinc are maintained.
590
Advisor: Alberto A. Sagues, Ph.D.
653
Zinc
Impedance
Disbondment
Epoxy
Polarization
690
Dissertations, Academic
z USF
x Mechanical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3457



PAGE 1

Corrosion of Dual Coated Reinforcing Steel with Through Polymer Breaks in Simulated Concrete Pore Solution by Adrienne Accardi A thesis submitted in partial fulfillment of the requirements for the degree of Masters of Science in Mechanical E ngineering Department of Mechanical Engineering College of Engineering University of South Florida Major Professor: Alberto A. Sags, Ph.D. Delcie Durham, Ph.D. Autar Kaw, Ph.D. Date of Approval: March 30, 2010 Key Words: Zinc, Impedance, Disbond ment, Epoxy Polarization Copyright 2010, Adrienne Accardi

PAGE 2

Acknowledgements The author would like to sincerely thank Dr. Alberto Sags for his continued guidance and encouragement during this project. She would also like to thank the mem bers of the committee, Dr. Delcie Durham and Dr. Autar Kaw, for their valuable suggestions. The author would like to thank her colleagues in the Corrosion Lab, Ezeddin Busba, Margareth Dugarte, Mersedeh Akhoondan, and Andrea Sanchez for providing such a wonderful support system during times of stress. The author would like to thank Kingsley Lau especially for providing valuable insight into the project and guidance since the author first came to work in the lab as an undergraduate. The author would lik e to thank her family and wonderful fianc Alex Gady for being patient while long nights were spent at the lab. They provided understanding and encouragement that allowed for the completion of this project. Finally, the author would like to thank Gerdau Ameristeel for the financial support that made this project possible. The findings and opinions presented in the manuscript are those of the author and not necessarily those of the sponsoring organizations.

PAGE 3

i Table of Contents List of Tables ................................ ................................ ................................ ..................... iii List of Figures ................................ ................................ ................................ .................... iv Abstract ................................ ................................ ................................ ........................... vi ii Chapter 1: Introduction ................................ ................................ ................................ ........ 1 Reinforced Concrete ................................ ................................ ................................ 1 Corrosion of Steel in Concrete ................................ ................................ ................. 2 Corrosion Control ................................ ................................ ................................ .... 3 Epoxy Coated Reinforcing Steel ................................ ................................ ............. 4 Galvanized Reinforcing Steel ................................ ................................ .................. 6 Dual Coated Reinforcing Steel ................................ ................................ ................ 8 Chapter 2: Approach and Ob jectives ................................ ................................ ................. 14 Objectives ................................ ................................ ................................ .............. 14 Experimental Approach ................................ ................................ ......................... 14 Chapter 3: Methodology ................................ ................................ ................................ .... 15 Exposure Testing ................................ ................................ ................................ ... 16 Electroch emical Impedance Spectroscopy ................................ ............................ 17 Metallography ................................ ................................ ................................ ........ 21 Coating Disbondment Testing ................................ ................................ ............... 21 Su pplemental OCP Test Exposures ................................ ................................ ....... 23 Chapter 4: Results and Discussion ................................ ................................ ..................... 32 As Received DCR Condition ................................ ................................ ................. 32 Visual Observations ................................ ................................ ............................... 32 Electrochemical Measurements ................................ ................................ ............. 33 Open Circuit Specimens ................................ ................................ ............ 33 Polarized Specimens ................................ ................................ .................. 36 Coating Disbondment ................................ ................................ ............................ 38 Implications on A nticipated Performance of DCR ................................ ................ 39 Chapter 5: Conclusions ................................ ................................ ................................ ...... 5 3 References ................................ ................................ ................................ .......................... 55

PAGE 4

ii Appendices ................................ ................................ ................................ ......................... 59 A ppendix I: Impedance Diagrams ................................ ................................ ......... 60 Appendix II: Metallographic Pictures ................................ ................................ .. 143

PAGE 5

iii List of Tables Table 1: Hot Dipped Galvanized Rebar Coati ng Layers and Characteristics ................... 13 Table 2: Chemical Preparation and Composition of Te st Solutions ................................ .. 31 Table 3: Time of Solution Exposure for Each Test Sample ................................ .............. 31 Table 4: Description of Qualitative Adhesion Ratings ................................ ...................... 31 Table 5: Summary of Phase 1 a nd This Investigation Results ................................ ........... 49 Table 6 : Impedance Fi t Data for Chloride Specimens ................................ ....................... 61 Table 7 : Impedance Fit Data for No chloride Specimen ................................ ................... 62

PAGE 6

iv List of Figures Figure 1: Annual Cost of Corrosion for the Five Government Sections ........................... 11 Figure 2: Microstructure of the Zi nc Layers of Galvanized Reb ar ................................ .... 11 Figure 3: Dependence of Corrosion Ra te of Zinc on pH ................................ ................... 12 Figure 4: Pourbaix Diagram for Zinc ................................ ................................ ................. 12 Figure 5: Metallographic Cross section of Intentional Defect Showing Removal of Much of the Epoxy Coating ................................ ................................ ................ 24 Figure 6: Diagram of DCR Test Sample ................................ ................................ ............ 24 Figure 7: Conceptual Diagram of a 3 Electrode Ce ll ................................ ......................... 25 Figure 8: Simplified Equivale nt Circuit for Uncoated Metal ................................ ............ 25 Figure 9: EIS Excitation Current Paths for a Coated Metal with a Coating Break ........... 26 Figure 10: Equivalent Circuit for Coated Metal Ideal Capacitan ce ................................ 26 Figure 11: Equivalent Circuit for Coated Metals with Capacitors Replaced by CPEs ...... 27 Figure 12: Specimen Subjected to the Disb ondment Measurement Procedure ................. 27 Figure 13: Procedure for Quantifying Adhesion Loss through Disbondment Radius Measurement an d Qu alitative Disbondment Ratings ............................ 28 Figure 14: Diagram of DCR Sample Used for Pull off Test ................................ ............. 29 Figure 15: Pull off Dol ly Attached to Rebar Specimen ................................ ..................... 29 Figure 16: Rebar Specime n and Pull off Testing Device ................................ .................. 30 Figure 17: Supplemental OCP No chloride Exposur e Test Cell ................................ ....... 30 Figure 18: Combined Polymer and Zinc Coating Thickness Distribution ........................ 41 Figure 19: Solid Corrosion Product that Developed on +100m V Chloride Exposure Specimens ................................ ................................ ........................ 41 Figure 20: Appearance of D efects After Immersion Period ................................ .............. 42 Figure 21: Open Circ uit Potential as Function of Time for Chloride (Solid Black Line), No chloride Exposure OCP (Dashed), and No chloride Auxiliary OCP (Gray) Duplicate ................................ ................................ ..... 42 Figure 22: Polarization Current Evolution with Time for No Chloride Exposure Duplicate ................................ ................................ ................................ .......... 43 Figure 23: Pol arization Current Evolution with Time for +100 mV Chloride Exposure Duplicate Specimens ................................ ................................ ........ 43 Figure 24: Polarization Current Evolution with Time for 500 mV and 1000 mV Chloride Exposure Duplicate ................................ ................................ ............ 44 Figure 25: Cumulative Anodic Charge for +100 mV Chloride and No chloride Exposure Duplicate Specime ns ................................ ................................ ........ 44 Figure 26: Nominal Corrosion Current Density from EIS Measurements of Duplicate OCP Specimens as a Function of Exposure ................................ 45 Figure 27: Nominal Zinc Thickness Loss Estimated from EIS Measurements of OCP Duplica te S pecimens as a Function of Exposure Tim e ........................... 46

PAGE 7

v Figure 28: Diagram Showing Possible Passivation Mechanism for Phase 1 DCR ........... 47 Figure 29: Pourbaix Diagram Showing the Effect of Increased pH and Increas ed Zincate Ion Concentrations ................................ ................................ .............. 47 Figure 30: Metallographic Cross Sections of Chloride (YG6) and No Chloride OCP (YB14) Specimens after Exposure ................................ .......................... 48 Figure 31: Average Coati ng Disbondment Radius Results ................................ ............... 51 Figure 32: Average Qualitative Adhesion Loss Rating Results ................................ ........ 51 Figure 33: Nominal Coat ing Pull off Strength Re sults ................................ ...................... 52 Figure 34: Discolored Area Around Defect Observed for the +100 mV Chloride Exposure Specimens ................................ ................................ ........................ 52 Figure 35 : Bode and Nyquist Diagrams for YG6 at 0.06 Days of Exposure .................... 63 Figure 36 : Bode and Nyquist Diagrams for YG6 at 0.09 Days of Exposure .................... 64 Figu re 37 : Bode and Nyquist Diagrams for YG6 at 0.13 Days of Exposure .................... 65 Figure 38 : Bode and Nyquist Diagrams for YG6 at 0.85 Days of Exposure .................... 66 Figure 39 : Bode and Nyquist Diagrams for YG6 at 1.22 Days of Exposure .................... 67 Figure 40 : Bode and Nyquist Diagrams for YG6 at 1.93 Days of Exposure .................... 68 Figure 41 : Bode and Nyquist Diagrams for YG6 at 3.07 Days of Exposure .................... 69 Figure 42 : Bode and Nyquist Diagrams for YG6 at 5.97 Days of Exposure .................... 70 Figure 43 : Bode and Nyquist Diagrams for YG6 at 9.93 Days of Exposure .................... 71 Figur e 44 : Bode and Nyquist Diagrams for YG6 at 13 Days of Exposure ....................... 72 Figure 45 : Bode and Nyquist Diagrams for YG6 at 42 Days of Exposure ....................... 73 Figure 46 : Bode and Nyquist Diagrams for YG6 at 76.83 Days of Exposure .................. 74 Figure 47 : Bode and Nyquist Diagrams for YG6 at 85.92 Days of Exposure .................. 75 Figure 48 : B ode and Nyquist Diagrams for YG21 at 0.0 5 Days of Exposure ................... 76 Figure 49 : B ode and Nyquist Diagrams for YG21 at 0.08 Days of Exposure ................... 77 Figure 50 : B ode and Nyquist Diagrams for YG21 at 0.12 Days of Exposure ................... 78 Figur e 51 : B ode and Nyquist Diagrams for YG21 at 0.84 Days of Exposure ................... 79 Figure 52 : B ode and Nyquist Diagrams for YG21 at 1.21 Days of Exposure ................... 80 Figure 53 : B ode and Nyquist Diagrams for YG21 at 1.92 Days of Exposure ................... 81 Figure 54 : B ode and Nyquist Diagrams fo r YG21 at 3.06 Days of Exposure ................... 82 Figure 55 : Bode and Nyquist Diagrams for YG21 at 5.95 Days of Exposure ................... 83 Figure 56 : Bode and Nyquist Diagrams for YG21 at 9.92 Days of Exposure ................... 84 Figure 57 : Bode and Nyquist Diagrams for YG21 at 12.99 Days of Exposure ................. 8 5 Figure 58 : Bode and Nyquist Diagrams for YG21 at 41.99 Days of Exposure ................. 86 Figure 59 : Bode and Nyquist Diagrams for YG21 at 76.83 Days of Exposure ................. 87 Figure 60 : Bode and Nyquist Diagrams for YG21 at 85.92 Days of Exposure ................. 88 Figure 61 : B ode and Nyquist D iagrams for YB14 at 0.04 Days of Exposure ................... 89 Figure 62 : Bo de and Nyquist Diagrams for YB14 at 0.10 Days of Exposure ................... 90 Figure 63 : B ode and Nyquist Diagrams for YB14 at 0.17 Days of Exposure ................... 91 Figure 64 : B ode and Nyquist Diagrams for YB14 at 0.22 Days of E xposure ................... 92 Figure 65 : Bode and N yquist Diagrams for YB14 at 0.96 Days of Exposure ................... 93 Figure 66 : Bode and Nyquist Diagrams for YB14 at 1.03 Days of Exposure ................... 94 Figure 67 : Bode and Nyquist Diagrams for YB14 at 1.18 Days of Exposure ................... 95 Figure 68 : Bode and Nyq uist Diagrams for YB14 at 2 Days of Exposure ........................ 96 Figure 69 : Bode and Nyquist Diagrams for YGB14 at 2.95 Days of Exposure ................ 97

PAGE 8

vi Figure 70 : Bode and Nyquist Diagrams for YB14 at 5.94 Days of Exposure ................... 98 Figure 71 : Bode and Nyquist Diagrams for YB14 at 7 Days of Exposure ........................ 99 Figure 72 : Bode and Nyquist Diagrams for YB14 at 7.95 Days of Exposure ................. 100 Figure 73 : Bode and Nyq uist Diagrams for YB14 at 9.93 Days of Exposure ................. 101 Figure 74 : Bode and Nyquist Diagrams for YB14 at 10.93 Days of Exposure ............... 102 Figure 75 : Bode an d N yquist Diagrams for YB14 at 12.94 Days of Exposure ............... 103 Figure 76 : Bode and Nyquist Diagrams for YB14 at 14.01 Days of Exposure ............... 104 Figure 77 : Bode and Nyquist Diagrams for YB14 at 15.99 Days of Exposure ............... 105 Figure 78 : Bode and Nyquist Diagrams for YB14 at 16.95 Days of Exposure ............... 106 Figure 79 : Bode and Nyquist Diagrams for YB14 at 19.94 Days of Exposure ............... 107 Figure 80 : Bode and Nyquist Diagrams for YB14 at 21.01 Days of Exposure ............... 108 Figure 81 : Bode and Nyquist Diagrams for YB14 at 21.94 Days of Exposure ............... 10 9 Figure 82 : Bode and Nyquist Diagrams for YB14 at 23.07 Days of Exposure ............... 110 Figure 83 : Bode and Nyquist Diagrams for YB14 at 24.16 Days of Exposure ............... 111 Figure 84 : Bode and Nyquist Diagrams for YB14 at 27.03 Days of Exposure ............... 112 Figure 85 : Bode and Nyquis t Diagrams for YB14 at 34.02 Days of Exposure ............... 113 Figure 86 : Bode and Nyquist Diagrams for YB14 at 63.08 Days of Exposure ............... 114 Figure 87 : Bode and Nyquist Diagrams for YB14 at 97.91 Days of Exposure ............... 115 Figure 88 : B ode and Nyquist Diagrams for YG7 at 0.05 D ays of Exposure .................. 116 Figure 89 : B ode and Nyquist Diagrams for YG7 at 0.13 Days of Exposure .................. 117 Figure 90 : B ode and Nyquist Diagrams for YG7 at 0.20 Days of Exposure .................. 118 Figure 91 : B ode and Nyquist Diagrams for YG7 at 0.24 Days of Exposure .................. 119 Figure 92 : B od e and Nyquist Diagrams for YG7 at 0.99 Days of Exposure .................. 120 Figure 93 : B ode and Nyquist Diagrams for YG7 at 1.03 Days of Exposure .................. 121 Figure 94 : B ode and Nyquist Diagrams for YG7 at 1.19 Days of Exposure .................. 122 Figure 95 : B ode and Nyquist Diagrams for YG7 at 2.01 Days of Exposure .................. 123 Figure 96 : B ode and Nyquist Diagrams for YG7 at 2.95 Days of Exposure .................. 124 Figu re 97 : Bode and Nyquist Diagrams fo r YG7 at 5.94 Days of Exposure .................. 125 Figure 98 : B ode and Nyquist Diagrams for YG7 at 7.01 Days of Exposure .................. 126 Figure 9 9 : B ode and Nyquist Diagrams for YG7 at 7.96 Days of Exposure .................. 127 Figure 100 : B ode and Nyquist Diagrams for YG7 at 9.94 Days of Exposure ................. 128 Figure 101 : B ode and Nyquist Diagrams for YG7 at 10.94 Days of Exposure ............... 129 Figure 102 : B ode and Nyquist Diagrams f or YG7 at 12.95 Days of Exposure ............... 130 Figure 103 : B ode and Nyquist Diagrams for YG7 at 14.06 Days of Exposure ............... 131 Figure 104 : Bode and Nyquist Diagr ams for YG7 at 16 Days of Exposure .................... 132 Figure 105 : B ode and Nyquist Diagrams for YG7 at 16.96 Days of Exposur e ............... 133 Figure 106 : B ode and Nyquist Diagrams for YG7 at 19.95 Days of Exposure ............... 134 Figure 107 : B ode and Nyquist Diagrams for YG7 at 21.02 Days of Exposure ............... 135 Figure 108 : B ode and Nyquist Diagrams for YG7 at 21.94 Days of Exposure ............... 136 Figure 109 : B ode and N yquist Diagrams for YG7 at 23.09 Days of Exposure ............... 137 Figure 110 : B ode and Nyquist Diagrams for YG7 at 24.16 Days of Exposure ............... 138 Figure 111 : Bode and Nyquist Diagrams for YG7 at 27.04 Days of Exposure ............... 139 Figure 112 : B ode and Nyquist Diagrams for YG7 at 3 4.03 Days of Exposure ............... 140 Figure 113 : B ode and Nyquist Diagrams for YG7 at 63.08 Days of Exposure ............... 141 Figure 114 : B ode and Nyquist Diagrams for YG7 at 97.93 Days of Exposure ............... 142

PAGE 9

vii Figure 115 : Metallographic Picture Location Diagram ................................ ................... 143 Figure 116 : YG7 Defect ................................ ................................ ................................ ... 144 Figure 117 : YG7 Off Defect ................................ ................................ ............................ 144 Figure 118 : YG2 Defect ................................ ................................ ................................ ... 145 Figure 119 : YG2 Off Defect ................................ ................................ ............................ 145 Figure 120 : YG4 Defect ................................ ................................ ................................ ... 146 Figure 121 : YG4 Off Defect ................................ ................................ ............................ 146 Figure 122 : YG18 Defect ................................ ................................ ................................ 147 Figure 123 : YG18 Off Defect ................................ ................................ .......................... 147 Figure 124 : YG6 Defect ................................ ................................ ................................ ... 148 Figure 125 : YG6 O ff Defect ................................ ................................ ............................ 148 Figure 126 : YG3 Defect ................................ ................................ ................................ ... 149 Figure 127 : YG3 Off Defect ................................ ................................ ............................ 149 Figure 128 : YG11 Defect ................................ ................................ ................................ 150 Figure 129 : YG11 Off Defect ................................ ................................ .......................... 150 Figure 130 : YG8 Defect ................................ ................................ ................................ ... 151 Figure 131 : YG8 Off Defect ................................ ................................ ............................ 151

PAGE 10

viii Corrosion of Dual Coated Reinforci ng Steel with Through Polymer Breaks in Simulated Concrete Pore Solution Adrienne Accardi ABSTRACT This investigation is an examination of the behavior of dual coated reinforcing steel (DCR) with defects in the polymer coating exposing the only zinc lay er in simulated concrete pore solution with and without chlorides. The intentional defects simulated the condition typically experienced by the rebar in service. Specimens were tested at open circuit potential, +100 mV, 500 mV, and 1000 mV for 30 to 100 days. The results were compared with that from previous DCR investigation with to steel defects and epoxy coated rebar (ECR). DCR with to zinc defects had extensive corrosion damage when under strong anodic polarization and exposed to chlorides and was si milar to that seen for DCR with to steel defects. The freely corroding (OCP) to zinc DCR specimens in solutions both with and with no chloride s experienced initially very active dissolution which ended after ~1 day. The zinc exposed at the coating breaks w as not completely consumed even after 100 days and there was no visible corrosion product accumulation. This may be due to the formation of a calcium hydroxyzincate passive film and shows that the zinc passivates in alkaline solutions without the benefit o f a crevice environment. The DCR with to steel defects and the DCR with to zinc defects had similar amounts of

PAGE 11

ix disbondment for all test conditions. Notable disbondment was seen only in highly anodic polarization regime with chlorides and was due to large a mounts of solid corrosion product formation. These results suggest then that the overall process of zinc wastage in DCR in concrete pore water is not likely to be rapid, which would be beneficial to extending the period in which the barrier and galvanic pr operties of the zinc are maintained.

PAGE 12

1 Chapter 1 Introduction $137.9 billion in 2002. This cost can be broken down into five industry sectors as is shown in Figure 1. Infr astructure alone costs $22.6 billion which is 16.4% of the total cost. Of that, $8.3 billion (37%) is spent to repair corrosion in highway bridges. The indirect costs due to traffic delays and lost productivity caused by bridge repairs are estimated to be 10 times that of the direct cost of corrosion repairs. Of the estimated 583,000 bridges in the United States, 235,000 (~40%) are steel reinforced concrete bridges and ~15% of these have been determined to be structurally deficient due to corrosion related problems ( Koch, et al. 2002). Developing corrosion resistant reinforcing steel would have a large impact on reducing the amount spent to repair corrosion damage in bridges. Reinforced Concrete Concrete is a commonly used building material worldwide. It is typically composed of mortar and aggregate. Although it can handle large compressive loads, it is not strong in tension. The concrete must be reinforced in order to handle tensile and shear forces. Steel reinforcing bar, or rebar, is embedded in the con crete to handle these loads. Rebar is typically made of plain carbon steel (e.g. 0.5 wt% of carbon). In order to transfer the loads from the concrete to the steel reinforcement, the rebar must be bonded to the concrete. For that reason, most rebar in use i s ribbed to provide a stronger mechanical bond than steel rod to concrete (Parker 1968).

PAGE 13

2 Corrosion of Steel in Concrete Passivity is defined by (Fontana and Greene 1978) as the loss of reactivity of a metal under particular environmental conditions. T his is typically due to the formation of a thin oxide surface film that protects the metal surface from corrosion. This film is usually stable at high pH values (9 < pH < 14.5). Concrete pore water tends to have concentrations of calcium, sodium, and potas sium oxides that combine with water to form hydroxides, creating a high pH (minimum of ~12.5 for Portland cement (Derucher, Ezeldin, and Korfiatis 1994)) environment (Broomfield 1998). As a result, steel is typically passive in concrete. Two environmen tal changes can break down the passive film: carbonation of the concrete and chloride attack. Carbonation of the concrete can decrease the pH to 8 or 9 causing the passive film to become unstable and decompose. Chloride ions do not seriously affect the pH of the pore water, but instead attack the passive layer and accelerate the corrosion process (Derucher, Ezeldin, and Korfiatis 1994) (Broomfield 1998). Chloride concentration above a threshold value, C T typically of more that 0.2% by mass of Portland ce ment may, be enough to destroy the protective film when the pH is greater than 11.5 (Derucher, Ezeldin, and Korfiatis 1994). Reinforcing steel in bridges pilings in marine environments or bridge decks subjected to seasonal deicing salts is susceptible to corrosion due to the penetration of chloride ions into the concrete and ensuing passivity breakdown. Chloride ions usually travel slowly by diffusion through the concrete or more quickly along cracks in the concrete. The corrosion process, after the bre akdown of the passive layer, is comprised of two simultaneous reactions: the oxidation or anodic reaction and the reduction or cathodic reaction. The oxidation reaction produces electrons while the reduction reaction consumes electrons. These two reactions must occur at the same rate and time on the surface of the metal for corrosion to take place. The anodic reaction, which consumes the

PAGE 14

3 iron that makes up ~97% of the reinforcing steel, can be expressed as (Fontana and Greene 1978): (Eq. 1) The equati on for the cathodic reaction that typically takes place in high pH solutions such as concrete pore water is: (Eq. 2) These reactions can be combined as: (Eq. 3) where Fe(OH) 2 is a corrosion product (Fontana and Greene 1978). The Fe(OH) 2 can be oxidized further to Fe(OH) 3 the corrosion product typically referred to as rust. These corrosion products are more voluminous than the steel (up to 600% of the original metal volume (Derucher, Ezeldin, and Korfiatis 1994)). This increa se in volume causes tensile stresses that, in turn, cause the concrete to crack and spall resulting in the loss of structural integrity. It is noted that very small amounts (~3 to 4% reduction of cross sectional area of the rebar (T. Ohta 1991)) of corros ion can cause enough internal stress to crack the concrete. Hence, it is very important to minimize the corrosion of reinforcing steel. Corrosion Control Many systems have been developed to prevent or slow down the corrosion of steel in concrete. Some of these prevention systems concentrate on the concrete condition. An increase in the thickness of the concrete cover forces chloride ions to travel through more concrete before reaching the steel. Also, using a lower water to cement ratio (to as little as 0 .32) minimizes the connectivity of the concrete pore network. These two practices combined with following proper casting and curing specifications can delay

PAGE 15

4 corrosion initiation by slowing down chloride transport and consequently extending the period of ti me required to increase the Cl concentration up to the C T of the rebar surface. Physical barrier, such as painted coatings, are sometimes added to the surface of the concrete to block chloride ions from diffusing through the concrete. Corrosion inhibitors that increase C T can be added to the concrete mix and chemically delay the onset of corrosion. Cathodic protection is another commonly used corrosion control system. This system works by supplying the metal intended to be protected with electrons (catho dic polarization), which hinders the metal dissolution described in Eq. 1. Two types of cathodic protection are used: impressed current and sacrificial anode. Both systems act to drive the potential of the protected metal to a more negative value, creating cathodic polarization. The sacrificial anode system relies on a galvanic couple where the protected metal is electrically connected to a metal that is more susceptible to corrosion. This metal corrodes, sending electrons to the protected metal. Zinc and m agnesium are commonly used materials for anodes. This system operates until the anode is consumed (Fontana and Greene 1978). Finally, many types of corrosion resistant reinforcing steels are being increasingly used. These include but are not limited to stainless steel, stainless steel clad carbon steel, galvanized and epoxy coated rebar. Epoxy Coated Reinforcing Steel Epoxy coated reinforcing steel, or ECR, is a surface abraded carbon steel rebar coated with a layer of fusion bonded epoxy polymer that acts as a physical barrier between the steel and the environment. Control of coating imperfections is essential for adequate performance (Manning 1996) (Yeomans 1994). The product has been in use s 1989), typically in environments where chloride induced corrosion is likely.

PAGE 16

5 ECR has, with relatively few exceptions, been reported to perform well in bridge decks where deicing salts are used. Bridge decks were reported to be in good overall conditio n, with corrosion seen only in cracked concrete locations (Smith and Virmani 1996) (Fanous and Wu 2000). However, it was reported that coating adhesion loss could occur in as little as four years after construction, well before chloride ions reach the re bar ECR vulnerable to corrosion when the chloride threshold is reached. It was shown that this disbondment occurs in good quality concrete with ECR that complies with specific There have been notable corrosion incidents in bridge substructure exposed to a marine environment affecting several major bridges. ECR was especially susceptible in the tidal zone where chloride concentrations tend to be higher (Griffith and Laylor 1999). This is especially evident in the Florida Keys were several bridges experienced severe corrosion damage relatively shortly after construction (~6 years) and continued to deteriorate at ~0.1 spal l per bent (pier) per year for 25 additional years with no indication of slowing down (Sags, Powers and Kessler 2009). It was noted in the literature that in instances where ECR performed well, concrete cover was also deeper and the concrete quality was better, which would increase the time to corrosion initiation greatly regardless of the type of rebar (Manning 1996) (Clear 1992). There is an instance where ECR performed well in a marine substructure, despite high chloride concentration. (Cui, Lawler, and Krauss 2007) reported on a bridge built in 1987 that had no evidence of spalling or large spread corrosion even though the chloride concentration was on average greater than 0.079 % per weight of concrete, the epoxy coating was of substandard thicknes s, and adhesion loss was present (Cui, Lawler, and Krauss 2007). This report also stated that one ECR sample contained corrosion, but the chloride concentration at this site was 0.251 % per weight of concrete, which in ~10 pcy, a relatively high concentra tion (Cui, Lawler, and Krauss 2007).

PAGE 17

6 Typical failure mechanisms of ECR seen in the field are loss of coating adhesion and macrocell formation. The loss of coating adhesion, which is caused by water absorption by the coating (Manning 196), anodic blisteri ng, and cathodic delamination (Nguyen and Martin 1996) (Nguyen and Martin 2004), resulted in coating that blistered and cracked (Clear 1992). This was seen in northern and southern United States bridge decks and southern bridge substructures with bars th at had passed inspection. Macrocell formation most often occurred in bars with coating holidays and larger damage, with some already experiencing undercoating corrosion (Clear 1992). Galvanized Reinforcing Steel Galvanized reinforcing steel is a carbon steel rebar coated with a layer of zinc and zinc is applied in a variety of ways including hot dipping, thermal spraying, electro deposition, and diffusion. Hot dipping, where the zinc becomes metallurgically bonded to the steel (Langill and Dugan 2004), is most commonly used for the manufacture of galvanized rebar. During the hot dipping process, several zinc steel alloy layers form on the surface of the carbon steel rebar. Typically four layers develop: gamma (inner layer), delta, zeta, and eta (outer layer), however, all of the layers are not created every time. The layers are shown in Figure 2. The number of layers as well as the thickness of the layers depends on several factors including the composition of the base carbon steel, the surface texture of the base steel, the temperature of the zinc bath, the amount of time the rebar is immersed in the zinc bath, and the speed at which the rebar is removed from the zin c bath. The layers and their characteristics are listed in Table 1. Note that as the layer gets closer to the steel, the more iron that is present in the alloy. The amount of corrosion protection is typically dependent on the zinc coating thickness, rather than the crystal structure of the alloy layers (Langill and Dugan 2004).

PAGE 18

7 Galvanized steel has been shown to withstand chloride concentration 2.5 times that of bare steel rebar and delay the time to corrosion by 4 5 times (Yeomans 2004). Despite this, t here are notable examples where galvanized rebar has not performed well in the field. (Pianca and Schell 2005) reported that there was significant corrosion related concrete damage in three Ontario bridge decks when the C T for black steel was surpassed. The zinc coating acts as both a sacrificial anode, much like a cathodic protection system, and a physical barrier. The zinc corrodes over time and the rate of corrosion depends on the pH of the environment. Zinc is relatively stable at pH values between 8 and 12.5, as seen in the Pourbaix diagram, Figure 4. Above pH 12.5 the corrosion rate of the zinc increase as the pH increases and below pH 6 the corrosion rate of zinc increases as the pH decreases. This is illustrated in Figure 3, which shows the corros ion rate of zinc versus pH. At a certain range of pH values, zinc corrosion products can act as a physical barrier, creating a passive layer. This typically occurs at pH values between 12.5 and 13.3. At pH values above 13.3, the corrosion products tend to create larger crystals which do not form a cohesive passive film. Therefore, at pH values above 13.3, the zinc corrodes readily, eventually leaving the steel unprotected (Bentur, Diamond, and Berke 1997). Other authors report a transition pH of ~13.1 (And rade and Alonso 2004). When galvanized steel comes into contact with freshly cast concrete, as in during the construction process of reinforced concrete structures, typically less than ~10um of the outermost eta zinc layer corrodes in a short time, and then corrosion tends to stop. Zinc oxide product forms first (Andrade and Alonso 2004): (Eq. 4) and is then further oxidized into zinc hydroxide (Andrade and Alonso 2004): (Eq. 5)

PAGE 19

8 Both zinc oxide and z inc hydroxide are white, powdery corrosion products and do not form a protective oxide layer. However, in a strong alkaline environment where calcium is present, such as concrete pore water, the zinc hydroxide further oxidizes to calcium hydroxyzincate: (Eq. 6) which forms a passive oxide layer on the zinc surface. This passive layer increases the C T of the zinc to about twice that of carbon steel (Andrade and Alonso 2004). When a structure with galvanized rebar is placed in serv ice with exposure to external chloride and the C T of the zinc is eventually reached, breakdown of the zinc passive film takes place. The zinc oxides that form then tend to be less voluminous than the iron oxides formed when plain carbon steel rebar first c orrodes and, therefore, create less of the internal stresses that would to cracks and spalls in the concrete. Thus, at least the early stages of galvanized rebar active corrosion are expected to be less damaging to the concrete structure than plain carbon steel rebar. It is noted that this interpretation has been disputed by (Hime and Machin 1993). Their investigation concluded that in concrete with large chloride concentration, another zinc corrosion product, zinc hydroxychloride II (Zn 5 (OH) 8 Cl 2 H 2 O), f ormed on galvanized bar which was more voluminous than iron oxides, expanding to 3.5 times the volume of original zinc. The authors concluded that this may be the reason for varying reports, noted earlier, of the behavior of galvanized rebar in the field. Dual Coated Reinforcing Steel Dual coated reinforcing steel, or DCR, has been developed relatively recently. DCR is composed of a carbon steel rebar core with a thermally sprayed zinc layer and a polymer epoxy coating over the zinc. ASTM Standard A1055 ( 2008) states that the zinc layer must be >0.035 mm and the total coating thickness (zinc and epoxy polymer) must

PAGE 20

9 be between 0.175 and 0.4 mm. Several corrosion evaluations of DCR have been conducted, including tests in concrete and tests in simulated concr ete pore solution (SPS). Tests in concrete (Clemea 2003), which followed the ASTM G 109 standard, resulted in estimated time to corrosion of ~530 days for specimens with defects extending through the polymer and zinc layers exposing steel, ~640 days fo r specimens with defects exposing only the zinc layer and >~740 days for specimens with no intentional defects (Clemea 2003). The lowest time to corrosion for DCR was ~6 times that for black bar. The estimated C T noted in that investigation was ~4460 ppm for specimens with defects exposing steel and >~5200 ppm for specimens with no intentional defects (Clemea 2003). The lowest C T for DCR was ~9 times that of black bar. Other tests in concrete (Darwin, Browning, Locke, and Nguyen 2007) concluded that th e zinc acted as a sacrificial barrier in both cracked and uncracked concrete. This was true for bars with defects extending through the epoxy and defects extending through both epoxy and zinc. The conclusions for this test, however are preliminary and the authors are awaiting the completion of additional tests to evaluate the long term performance of DCR (Darwin, Browning, Locke, and Nguyen 2007). The experiments conducted in SPS solution (Lau and Sags 2009) evaluated DCR specimens with defects through the epoxy and zinc layers, directly exposing steel and ECR specimens with defects exposing steel. Frequent reference is made to this investigation in the following for comparison purposes and detailed results of that investigation are presented in detail in Table 4. The conclusions from that investigation are as follows (Lau and Sags 2009): The DCR coating adhesion depended on the strength of the zinc layer. For all polarization regimes and solutions, the adhesion loss experienced by DCR was less than or equal to that experienced by ECR. ECR and DCR both experienced extensive corrosion for the +100 mV chloride exposure tests; however the damage to ECR was greater. This

PAGE 21

10 was concluded to be due to greater amounts of corrosion product produced by ECR causi ng more coating disbondment. The ECR open circuit chloride exposure specimens developed a negative potential and corrosion product. The DCR open circuit potential, OCP, specimens in both solutions began at very negative potentials which increased to ~ 400 mV. They developed no visible corrosion products. The lack of steel corrosion was concluded to be due to corrosion prevention from the galvanic coupling of the steel in the defect and the rim of zinc around the edge of the defect. For OCP DCR in both solut ions, the zinc was consumed very actively upon immersion and then slowed consumption to a low rate. DCR with no chloride exposure at medium ( 500 mV) and strong ( 1000 mV) cathodic polarization produced cathodic current less than those for ECR. It was conc luded that DCR under these polarization regimes would not support corrosion macrocells greater than ECR.

PAGE 22

11 Figure 1 : Annual C ost of C orrosi on for the Five Government S ections. (Adapted from ( Koch, et al. 2002).) Figure 2: Microstructure of the Zinc Layers of Galvanized Rebar. This sample was exposed to wet concrete for 2 yea Infrastructure, $22,600,000, 16.4% Government, $20,100,000, 14.6% Production and Manufacturing, $17,6 00,000, 12.8% Transportation, $29,700,000, 21.5% Utilities, $47,900,000, 34.7% Eta layer Gamma layer Steel Rebar Zinc layer

PAGE 23

12 Figure 3: Dependence of Corrosion Rate of Z inc on pH. (Adapted from (Bentur, Diamond, and Berke 1997).) Figure 4: Pou rbaix Diagram for Zinc. (Adapted from (Pourbaix 1974).) 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 2 4 6 8 10 12 14 16 pH Corrosion Rate / um*year um*year 1 Zn ZnO 2 2 Z nO Zn 2+ pH E 8 12

PAGE 24

13 Table 1: Hot D ipped Galvanized Rebar Coating Layers and Characteristics. (Adapted from (Langill and Dugan 2004)) Layer Alloy Composition (%Fe) Melting Point ( o C) Crystal Structure Hardness (DPN) Cha racteristics Eta Zinc 0.03 419 Hexagonal 70 72 Soft, ductile Zeta FeZn 13 5.7 6.3 530 Monoclinic 175 185 Hard, brittle Delta FeZn 7 7 11 530 670 Hexagonal 240 300 Ductile Gamma Fe 3 Zn 10 20 27 670 780 Cubic N/A Thin, hard, brittle Base Carbon Steel Carbon Steel 98 99 ~1530 Cubic 150 175 Ductile

PAGE 25

14 Chapter 2 Objective and Approach Objectives The objectives of this investigation are: To evaluate the corrosion behavior of DCR with defects simulating moderate coating damag e most likely to be seen in the field. To compare the behavior of DCR with moderate coating damage to that previously observed with severely damaged DCR and ECR. Experimental Approach To address the first objective the corrosion behavior of DCR was eval uated through electrochemical and coating adhesion loss testing. Electrochemical tests were conducted over a period of time at various polarization values. Two testing solutions were used, simulated concrete pore solution and simulated concrete pore soluti on with NaCl addition. Adhesion loss testing was conducted after the DCR had been immersed in solution and polarized for the period of time. Intentional defects, which extended through the polymer coating exposing the zinc layer, were made in the DCR. Thes e defects represent typical coating damage seen on DCR where normal field handling practices are followed. Supplemental exposure tests were run to ensure the reproducibility of the OCP no chloride results. For the second objective the results were compar ed with those from a previous investigation conducted to explore the behavior of DCR and ECR with extensive coating damage, simulated by intentional defects that extended through polymer and zinc coating layers or the polymer layer, respectively, to expose steel.

PAGE 26

15 Chapter 3 Methodology Dual coated rebar stock, 1.6 cm in diameter, was obtained from the manufacturer and test sample bars, 23 cm in length, were cut from this stock. The two cut ends of the sample bar were patched with Valspar Yellowbar To uchup Components A and B (part number 920Y966) an epoxy patch compound provided by the DCR manufacturer. The two components were mixed together in equal parts, as per the instructions. The coating quality of these samples was assessed through coating thick ness (polymer and zinc) measurements performed with an Elektro Physik S/N Mikrotest magnetic coating thickness gage (model number 014400) and the presence of holidays and mechanical coating damage was assessed by visual observation and the use of a Tinker and Razor Model M 1 holiday detector. For the holiday detection, a small sponge wetted in water with Kodak Photo Flo 200, as suggested on the Tinker and Razor website, was attached to a detecting wand. The wand with sponge was passed along the bar from top to bottom four times, rotating the bar 90 o each time. If a holiday was detected, the detector would beep. This resulted in a failure and that bar was not used in testing. An exception was made if the holiday was detected along the top edge of the sample w here the end of the bar had been patched with polymer epoxy patch compound. In this case, a second thin coat of epoxy polymer was added to the top of the bar and was then retested for holidays. Bars that passed were used in testing. If it failed after repa tching, the bar was not used for testing. The magnetic gage was used by placing the magnetic probe on the surface of the bar, between ribs. The measuring wheel was turned slowly until the magnet no longer held onto the surface. The thickness measurement de noted by the wheel was then recorded. Coating thickness measurements were made in triplicate. The average of the three measurements was calculated and taken as the coating thickness at that particular site. Intentional coating defects, 1.6 mm in diameter, were introduced by locally melting

PAGE 27

16 the polymer and mechanically removing it with a 40 watt soldering iron. This procedure ensured that the defect extended only through the polymer coating layer to expose the zinc layer, but without melting or removing it. Analysis with an optical microscope and metallographic analysis (Figure 5) verified that this method removed nearly all of the epoxy layer with about 2/3 of the zinc exposed and the rest of the defect area covered only by a thin epoxy residue. Defects were located between the ribs of the reinforcing bar. Each bar sample had eight defects, four defects on two opposing sides of the bar. Each defect had an area of ~2 mm 2 so the total exposed area of metal per bar was ~0.16 cm 2 One end of the bar was cast to a depth of 2 cm into a metallographic epoxy cylindrical base. The purpose of this epoxy base was to protect the end from corrosion and to serve as a way for the bar to stand upright. A 6 32 stainless steel screw was tapped on the other end of the bar to se rve as an electrical connection. Connectivity was tested with a multimeter between the screw and all intentional defects. A diagram of the sample bars is presented in Figure 6. Exposure Testing All potentials in this document are given in the saturated ca lomel electrode (SCE) scale. The DCR sample bars were exposed to two simulated concrete pore solutions (SPS): one contained a 3.5 % by weight addition of NaCl and the other had no NaCl addition. The chemical composition of both solutions is presented in Table 1. The samples were partially immersed with the top ~2 cm, including the stainless steel screw, above the solution, leaving ~19cm, including all eight defects, to be exposed to solution. The test tanks were made two 10 gallon glass fish tanks, each c ontaining 8 samples and ~26 L of solution each. The tanks were topped with a Plexiglas cover sealed with weather stripping in order to prevent carbonation of the solution. The solution was kept at lab ambient temperature, 22 2 C.

PAGE 28

17 Four polarization regi mes were used for each sample group: +100 mV, 500 mV, 1000 mV, and open circuit potential (OCP). These values are comparable to those used in similar tests (Lau and Sags 2009). These experiments were conducted with duplicate specimens. These potential s were maintained with a multi potentiostat, capable of maintaining multiple polarization values, and measured regularly with a SCE that was temporarily inserted into the tank. The potentiostat works by maintaining a specified potential difference between the working and reference electrode with an operational amplifier (Orazem and Tribollet 2008). In a multi potentiostat, several operational amplifiers are used to maintain multiple samples at their set potentials at the same time. The potential difference s were adjusted by potentiometric resistors controlling each operational amplifier. Potentials were adjusted regularly to maintain the desired values within 5 mV. There was one multi potentiostat for each tank. For each multi potentiostat, a common activ ated titanium mesh counter and a common activated titanium rod reference electrode (Castro, Sags, Moreno, and Genesca 1996) were used. Each reference electrode was calibrated regularly with respect to the SCE at the time of potential measurements. The d uration of exposure for each test specimen ranged from 55 to 102 days as listed in Table 2. Also listed in Table 2 are the identifiers for each test specimen which begin with either YG or YB. The current produced by each bar sample was measured and record ed regularly. Cumulative charge was calculated by approximating the area under the current versus time curves using the trapezoidal rule of integral approximation. Visual observations of the defects were noted throughout the test. The approximate pH of the solution was determined at the beginning and end of the testing period with pH paper. Electrochemical Impedance Spectroscopy Electrochemical impedance spectroscopy (EIS) can be used to elucidate the electrochemical behavior of a corroding electrode by m easuring the time dependent current response of the electrode to a small alternating potential applied across the interface. This is achieved by using a three electrode cell, shown in Figure 7. In this

PAGE 29

18 investigation, an activated titanium electrode is use d as reference and an activated defined by (Jones 1996): (Eq. 7) where V(t) is the alternating potential (in complex form i.e. V = V o e ) being applied to the syste m, I(t) (in complex form I = I o e ) the angular frequency given by: f (Eq. 8) where f is frequency in hertz, and is the phase angle. Impedance can be separated into real and imaginary components (J ones 1996): (Eq. 9) Some electrochemical processes that occur at the surface of the corroding electrode take some time to respond to changes in the applie d potential, manifested by a finite value of charging and discharging of the interfacial capacitance (Orazem and Tribollet 2008). The electrical response of the interface can be represen ted by equivalent circuits containing resistors and capacitors (Jones 1996). Impedance measurement results are customarily displayed in two graphs, the Bode diagram and the Nyquist diagram. The Bode diagram is a plot of log |Z| versus the

PAGE 30

19 frequency. Th e Nyquist diagram is a plot of the real impedance component versus the imaginary impedance component. Information about the system is gained when software is used to determine the equivalent circuit that best matches the electrical response of the system The values assigned to the components in the equivalent circuit represent different aspects of the system that are related to the corrosion rate and the interfacial and coating capacitances (Jones 1996). For example, for the case of uncoated metal under simplified conditions, the equivalent circuit is as shown in Figure 8, where R s is the ohmic electrical resistance between the reference electrode and the metal/electrolyte interface (Figure 9). C M is the interfacial capacitance, and R p is the polarizatio n resistance. The latter is a key element in the EIS interpretation as it is related to the corrosion current by the Stearn Geary equation: (Eq. 10) where B is equal to: (Eq. 11) slopes for the cathodic and anodic reactions. These are typically approximated to be 0.118 V which gives an approximate value of 0.026 V for B, found to be representative of many corroding systems (Jones 1996). Thus, an electrochemical measurement of R p can be used to obtain an approximate value of the rate at which the metal is being consumed. The corrosion current is related to the rate at which the metal is consumed (moles / second) by the Faradaic conversion: (Eq. 12)

PAGE 31

20 where d m/dt is the consumption rate, n is the valence (2 for zinc), and F is the Faradaic constant (96,500 Coulombs/equivalent). In terms of cumulative metal thickness lost (T), the corrosion current density, i corr (corrosion current divided by the area affected ) can then be converted into mass of metal lost with (Jones 1996): (Eq. 13) where A w and t is time. In the case of a coated metal, corrosion happens mostly at the coating breaks (Figure 9) and the coating acts as a capacitor. In an impedance test, part of the alternating excitation current flows through the electrolytic resistance of the breaks to the metal interface exposed at the breaks and the other part through the coating capacitance. The equivalent circuit used for this case reflects the current part ition among the two paths and is shown in Figure 10, where the C c is this coating capacitance and R b is the resistance associated with the break. The interfacial and coating capacitances typically display non ideal behavior such that their admittance (inve capacitor, but instead can be represented by the admittance of a constant phase element (CPE) given: (Eq. 14) where Y o and n are the two parameters defining the CPE (Orazem and Tribbolet 20 08). EIS measurements were taken regularly for the OCP specimens in both solutions. Scans were run at a frequency range of 100,000 Hz to 10 mHz with an amplitude of 10 mV. The choice of equivalent circuit used to analyze the EIS results is detailed in C hapter 4.

PAGE 32

21 Metallography At least one defect for each polarization regime and solution was analyzed metallographically. A ~2.5 cm long section of the bar around the defect was first cut with a hack saw. A cut was then made down the middle of the defect wi th a slow speed cutter and diamond blade. One half of the resulting cross section revealing the defect was set in metallographic epoxy. Rough polishing was conducted with water on Struers MD Piano resin bonded diamond grinding discs of 120, 600, and 1200 g rit. Water free fine polishing was conducted to ensure that little or no zinc was corroded during polishing. The first step of fine polishing was done on Struers MD Dac cloth using Struers 3 um DP Paste P polycrystalline diamond paste with Streur alcohol b ase blue lubricant. The final polishing step was conducted with Struers MD Floc cloth with Beuhler 0.05 um Alumina powder in ethanol. After polishing, samples were examined with a Reichart metallographic microscope and pictures of the defect, which are sho wn in Appendix II, were taken at a 15x magnification. Coating Disbondment Testing After being exposed to solution for the indicated amount of time, the specimens were removed and assessed for corrosion and coating disbondment. Upon removal from solution, the pH of any solution maintained on the surface of the metal on the defect surface was measured with pH paper. If anodic blistering (which affected only +100 mV chloride exposure samples) had occurred, the pH of the solution under the disbonded coating w as measured with pH paper as well. The samples were then lightly dried with absorbent paper. Pictures were then taken of each defect. The procedure for measuring the disbondment radius and qualitative disbondment ratings was conducted immediately following the drying and photographing of the sample. The coating disbondment was assessed in three ways: by the measured length of disbondment from the defect, by qualitative disbondment ratings given to coating sections near the defect, and by pull off tests co nducted near the defect sites.

PAGE 33

22 The procedure for evaluating the length of disbondment from the defect was as illustrated in Figures 12 and 13. A 3 mm wide strip of epoxy was scribe cut between the ribs on each side of the defect. Two more 3 mm wide strip s were scribe cut above and below the defect across the ribs at a ~90 angle from the ribs, as shown in Figures 12 and 13, forming a skewed cross. The scribe cuts were made through the epoxy layer to the metal using a sharp, thin blade. After the four stri ps were scribe cut, the knife was used to detach a small corner of one of the epoxy strips from the metal at the defect edge. This corner was grasped with precision tweezers and pulled until the epoxy strip broke. The distance from the defect edge to the l ocation where the coating strip broke was then measured and recorded. The same procedure was followed with the remaining three epoxy strips. Four defects on each sample were analyzed in this way. The measurements for each sample bar were then averaged and recorded. This method is similar to that used in (Lau and Sags 2009). Qualitative ratings describing the adhesion level of the coating, listed in Table 3, were used to denote the ease of coating separation. The sections peeled with the precision tweez ers in order to define the disbondment radius, were given a rating of 1 or 2. The remaining portion of the 1 cm strip of epoxy, beyond the disbondment radius was cut into 3 mm segments and removed with the knife and tweezers. Each section was given a ratin g from Table 3 and recorded. The qualitative adhesion ratings of each sample bar were averaged and recorded. The procedure for determining the disbondment radius and qualitative adhesion loss ratings is illustrated in Figure 13. The pull off strength of the coating was measured with a mechanical device designed for this specific use (Sags, et. al 1994) and is shown in Figure 16. A 6 mm diameter dolly was attached with cyanoacrylate adjacent to a defect. For each polarization regime in both chloride and no chloride solution, one of the bars used had as part of its surface on one side the rolled in bar size and make designation. This offered more space between deformation ribs and markings to attach the pull off dollies. A diagram of these particular samp les is seen in Figure 14. The surface of the dolly had

PAGE 34

23 been contoured to match the curvature of the bar. Before attachment the epoxy coating was lightly sanded and degreased by wiping the surface with ethanol to improve adhesion. The epoxy coating around t he perimeter of the attached dolly was removed (Figure 15) using a rotating dental drill bit. The testing device slowly pulled the dolly separated from the bar. This set up is seen in Figure 16. The pull off force was then divided by the dolly area and recorded as the nominal pull off strength. This procedure was used in several other investigations (Sags, et. al 2009) (Sags and Powers 1996). Supplemental OCP Test E xposures An additional open circuit experiment was conducted to ensure reproducibility of the test method. Two DCR samples were prepared in the way explained previously and shown in Figure 6. They were immersed in SPS solution, with 19 cm of the bar expos ed to the solution. The test chamber was a ~14 cm diameter ~30 cm tall Plexiglas cylinder with square Plexiglas pieces used to seal the top and bottom. The square pieces were attached to the cylinder with silicon gel. The container held ~3L of solution. Th e specimens were sealed inside the test chamber to prevent carbonation of the solution. The top square piece contained holes for wires and electrodes which were plugged with rubber stoppers. The open circuit potentials of the specimens were measured regula rly with a SCE.

PAGE 35

24 Figure 5: Metallographic Cross Section of Intentional Defect Showing Removal of Much of the Epoxy C oating. Figure 6: Diagram of DCR Test S ample. 3.8cm 3.8cm 3.8cm 3.8cm 3.8cm 2cm Solution line 19cm Dia = 1.6mm 2cm 10 um Steel Zinc Epoxy Mounting medium

PAGE 36

25 Figure 7: Conceptual Diagram of a 3 Electrode Cell. V is the applied potential and I is the resulting current. An external instrument (potentiostat) makes the necessary current potential control adjustments so that the necessary amount of current, I, is introduced to obtain the desired val ue of V. The counter electrode serves to provide a return path to the excitation current. Figure 8: Simplified Equivalent Circuit for Uncoated Metal. C M

PAGE 37

26 Figure 9: EIS Excitation Current Paths for a Coated Metal with a Coating Break. Figure 10: Equivalent Circuit for Coated Metal Ideal Capacitance. Metal Coating R s I to polymer coated surface I to exposed metal surface I C M

PAGE 38

27 Figure 11: Equivalent Circuit for Coated Metal with Capacitors Replaced by CPEs. (Orazem and Tribollet 2008) Figure 12: Specimen Subjected to the Disbondment Measurement P rocedure. Specimen was polarized to 500 mV with no chloride exposure. R b CPE C CPE M

PAGE 39

28 Figure 13: Procedure for Quantifying Adhesion Loss through Disbondment Radius M eas ure ment and Qualitative Disbondment R atings. Step 1: Scribe ep oxy strips. Step 2: Peel epoxy strips. Measure disbondment radius and assign qualitative ratings. Disbondment Radius Step 3: Scribe 3mm epoxy sections and peel. Assign qualitative ratings to sections.

PAGE 40

29 Figure 14: Diagram of DCR Sample Used for Pull off T est. This figure shows one side of the bar. The opposite side is the same as that shown in Figure 2. Figure 15: Pull off Doll y Attached to Rebar Specimen. 4.8cm 3.8cm 2.6cm 2.8cm 5cm 2cm Solution line 19cm Dia = 1.6mm 2cm

PAGE 41

30 Figure 16: Rebar Specimen and Pull off Testing Device. Figure 17: Supplemental OCP No chloride Exposure Test Cell. Tensioning nut Force measuring ring Cyanoacrylate bond F Test dolly Rebar (held in place by plastic fitting) Link to test dolly

PAGE 42

31 Table 2: Chemical Preparation and C omposi tion of Test S olutions. Solution Type KOH NaOH Ca(OH) 2 NaCl DI Water pH SPS 0.19M 20.9g 0.09M 7.4g 0.03M 4.2g N/A N/A 2L 13.3 SPS + NaCl 0.19M 20.9g 0.09M 7.4g 0.03M 4.2g 0.6M 70.0g 2L 13 Amount in solution. Table 3: Time of Solution E xposure (day s) for E ach Test Sample Designated by its I dentifier. Potential (mV vs. SCE) SPS SPS + NaCl OCP YB14 101 YG7 101 YG21 86 YG6 86 +100 YG15 67 YG2 65 YG17 55 YG3 62 500 YG14 67 YG4 71 YG22 88 YG11 89 1000 YG9 82 YG18 102 YG20 88 YG8 89 Table 4: Descriptio n of Qualitative Adhesion Ratings Qualitative Adhesion Rating Characteristics of Adhesion Loss 1 Coating is easily removed. 2 Coating is disbonded but some force is needed to remove. 3 Coating is disbanded but a large amount of force is required to remo ve. Some parts of coating remain on metal surface. 4 Coating is not disbanded.

PAGE 43

32 Chapter 4 Results and Discussion Note: In the following, frequent reference will be made to the (Lau and Sagues 2009) and Polymer/Zin c Coated Rebar in Simulated Concrete Pore Phase 1 Phase 1 investigation and this investigati on is shown in Table 5 As Received DCR Condition The combined polymer plus zinc coating thickness was on average ~0.28 mm and the standard deviation was ~0.03 mm which meets the ASTM A1055 specification. The cumulative fraction of coating thickness meas urements is shown in Figure 18. The zinc thickness was on average ~0.028mm which also meets the specification. The bars selected contained no visible coating defects. Those specimens used in the test were concluded to be representative of the average produ ct produced by the manufacturing company. Visual Observations During the period of immersion, only the +100 mV chloride exposure specimens showed corrosion product formation. The oxide was reddish brown in color, indicating an iron oxide, likely oxidized to Fe 3+ which formed tubercles that extended outward from the defects (Figure 19 and 20). Corrosion products developed on one specimen after ~1 day and, on the duplicate specimen, after ~3 days of immersion. This oxide formation had developed in all eigh t defects on each of the two samples by the termination of the test. In contrast, in the Phase 1 investigation, oxide formed on ECR and DCR +100 mV chloride exposure specimens in less than an hour, so this stage seems to be delayed in the

PAGE 44

33 DCR with defects exposing only zinc. It was also noted in the Phase 1 investigation that corrosion products developed on ECR +100 mV and 500 mV chloride exposure specimens in locations other than the intentional defect sites as well. The authors concluded this was caused by anodic blistering and subsequent cracking of the polymer coating. On the contrary, in this study, anodic blistering was noted only in the +100 mV chloride exposure specimens and no solid steel corrosion developed under the coating. Moreover, solid whit e zinc corrosion product was not apparent on samples for any of the polarization regimes in either solution as seen in Figure 20. This suggests the development of a zinc hydroxide which dissolves in an alkaline environment to HZnO 2 and ZnO 2 2 (Pourbaix 1 974). Electrochemical Measurements Open Circuit Specimens The open circuit potentials for specimens in both solutions are shown in Figure 21. All six specimens were at very negative potentials ( 1400 mV) immediately after immersion, but increased after ~ 1 day to ~600 mV, where the potentials remained relatively stable for the remainder of the immersion period. There was no significant difference between the OCP values of samples with and without chloride exposure. These potential values are similar to tho se noted with bulk zinc in similar solutions (Videm 2001) suggesting that the bulk zinc in the defects corroded initially and then approached passive behavior. EIS analysis was conducted with two types of equivalent circuits. The impedance diagram of co ated metals with behavior that can be approximated by the circuit in Figure 11 shows two loops, corresponding to the effective time constants associated with the combination of CPE C R b and CPE M R p respectively (Orazem and Tribollet 2008). If corrosion at the break is very slow, as when the metal exposed by the break is in a near passive condition, Rp becomes very large. In that case the loop associated with CPE C R b becomes less apparent and the combination of CPE M R p tends to dominate the impedance spectr um, at least at low frequencies. In that case, the low frequency impedance behavior

PAGE 45

34 tends to resemble that of the circuit in 8, with a CPE (CPE M with parameters Yo M and nM ) instead of the ideal capacitance C M Thus as a working approximation, the circuit in Figure 11 was used to fit the experimental EIS data for the initial part of the exposure, when the metal exposed at the breaks was in the active condition (potentials more negative than ~ 900 mV), and the impedance diagrams exhibited two clearly define d loops. This condi tion is apparent in Figures 35 to 36, 48 to 50, 61 to 64, and 88 to 91 in Appendix 1. When a near passive regime developed (considered to be manifested by potentials more positive than ~ 900 mV), the circuit shown in Figure 8 with th e capacitance replaced with CPE M was used instead. Again, a working approximation was implemented consisting of using only the fou r lowest frequency data points ( 10 mHz to 100 mHz) for the fitting procedure as that frequency range was deemed to be most rep resentative of the impedance behavior dominated by the polarization resistance of the system. It is noted that the value of R s obtained in that case is only a nominal parameter and is not further considered. After the value of R p was obtained for each test it was used to estimate the apparent corrosion current (Eq. 10) and estimated metal consumption rate (Eq. 12). The corrosion current was then divided by the area of metal exposed at the coating breaks to obtain the corrosion current density. The metal th ickness loss as function of immersion time was calculated using this current density (Eq. 13). The EIS experiments provided additional insight on the behavior of the OCP specimens. It is noted, however, that EIS estimated corrosion rates are subject to c onsiderable uncertainty (Orazem and Tribollet 2008) especially considering the uncertainty associated with the working approximations noted above. Consequently, the EIS analysis results obtained here are presented mainly for comparison purposes. In the f measurement s, denotes only an apparent or nominal corrosion current density and it is understood that future analysis using more sophisticated methods may yield updated result s.

PAGE 46

35 The EIS analysis results are summarized in Figure 26. The corrosion current density (calculated by assuming for simplicity that all the corrosion took place at the surface area directly exposed at the intentional defects) of OCP specimens, both in ch loride and no chloride exposure, started at a high value (~1E 4 A/cm 2 ) which quickly dropped to less than 1e 6 A/cm 2 Those results were summed by trapezoidal integr ation, after application of Eq. 13, to calculate the nominal zinc thickness consumed as f unction of time. The result is displayed in Figure 27. The nominal metal loss was in all cases less than that of the average zinc thickness that was determined by direct metallographic observation (denoted by the red line in Figure 27). Considering the afo rementioned uncertainty inherent to the EIS estimates of metal loss, these results are in good agreement with the observation of significant amounts of zinc remaining at the defects after exposure (Figures 20 and 30), as well as the potential evolution evi dence of early onset of passive or otherwise very slow corrosion of the zinc at the OCP. While all of the OCP DCR specimens in this investigation showed a trend toward passive behavior, their EIS estimated corrosion current density was somewhat higher tha n those reported for similar exposure in the to steel defects of the Phase 1 specimens. Moreover, the OCP potential in this investigation tended to stabilize at about 600 mV compared to ~ 400 mV in Phase 1. These differences may be due to different passiv ation mechanisms. The Phase 1 specimens experienced crevice corrosion (Fontana and Greene 1978) of the zinc between the carbon steel bar and the epoxy polymer coating (Lau and Sag s 2009) ( Accardi 2009). The zinc corrosion equations (Eq. 4 and 5) show t hat during dissolution, OH ions are consumed. Because this occurs in a crevice the occluded geometry promotes accumulation of reaction products, as shown schematically in Figure 28. As a result, the solution immediately adjacent to the corroding zinc ten ds to acidify slightly as indicated by the blue arrow in Figure 29. Also, because of the occluded geometry, the concentration of zincate ions may appreciably increase in the crevice which shifts the boundary line on the Pourbaix diagram separating active a nd passive regions to the right (red arrow in Figure 29). The combination of increasing zincate ion concentration and a dropping pH shifts conditions to the passive region of the Pourbaix

PAGE 47

36 diagram. The authors concluded also that passivation due to calcium hydro xy zincate formation should not be discounted, but it was noted that the calcium ion concentration in the solution was very low. The OCP crevice conditions were less likely to be predominant in this investigation as the DCR specimens here had defects were the zinc was freely exposed to solution instead of being present only at the edges of the defect. As indicated earlier, upon removal from the solution, there was still abundant zinc present on the surface of the defects and between the polymer and the steel around the rim of the defect (Figure 30) so crevice conditions had not developed. Nevertheless, passive behavior of the freely exposed zinc appears to have developed here even in the absence of the in crevice beneficial factors noted in Phase 1. A possible explanation is that because there is calcium in the solution, calcium hydroxyzincate may have formed on the surface (per Eq. 6) and passivated the zinc. This passivity may however by imperfect. As noted before, high pH solutions tend to create lar ger calcium hydroxyzincate crystals which may not form protective enough films (Andrade and Alonso 2004) and a pH 13.1 or 13.3 protection may be substantially compromised. The SPS solution used here is very close to that transition pH range, so it is poss ible that a completely cohesive passive layer is not present. This may explain the apparent terminal dissolution rate here being somewhat high er than that observed in Phase 1 Polarized Specimens For the no chloride exposure tests, all polarization regim es initially exhibited high anodic current (e.g. >1mA), indicating active corrosion of zinc. Later on, the +100 mV samples had small anodic current (~0.5 A) for the majority of the testing time. Both the 500 mV and the 1000 mV regimes resulted in small cathodic currents, ~0.5 A and ~1.5 A, respectively. These results are shown in Figure 22. For the chloride exposure tests, the +100 mV specimens demanded high anodic currents (>1 mA) initially and stabilized over a couple of days to a regime of continu ing

PAGE 48

37 corrosion at 200 and 700 A, which was confirmed by the observation of solid corrosion product on the defects. After initial high anodic currents, the 1000 mV and 500 mV specimens stabilized to a very small cathodic current (~1 A) and values in the range of 0.3 A, respectively. The high initial current and subsequent decrease to very small values suggest an initial period of active zinc corrosion followed by the development of a near passive regime possibly by a mechanism similar to that discusse d earlier. The cumulative charges for the +100 mV chloride and no chloride exposure samples were ~1000 coulombs and ~6 coulombs, respectively. This indicates that the chloride exposure specimens sustained considerably more corrosion damage than the no c hloride exposure specimens. This is consis tent with visual observations. Using Eq. 12 and 13 and assuming corrosion occurred only in the area exposed by the coating breaks, the zinc thickness lost was calculated from the cumulative charge. The estimated th ickness lost for the chloride and no chloride exposed specimens was 3mm and 20 um, respectively. This estimate for the no chloride specimens indicates that not all of the zinc in the defect was consumed. The estimate for the chloride specimens is clearly t oo large, indicating that the assumption made that corrosion only occurred in the defect area is not valid for these specimens as is evidenced by the anodic blistering of the coating. A more appropriate analysis that takes into account corrosion in the zon e around the defect is presented in the Coating Disbondment section. As shown in Table 4, the electrochemical results for the Phase 1 DCR specimens and the DCR specimens in this investigation are, for the most part, very similar. There is, however, one notable difference. The Phase 1 open circuit potentials exposed to both solutions follow a similar pattern to this investigation, starting at a very negative value upon immersion and increasing to a relatively stable higher value. However, the potentials f or the Phase 1 investigation are more positive, starting at 1000 mV and stabilizing at 400 mV. The potential trends for both defect types suggest an approach to passive behavior. The difference in potential values may be due to different passivation mech anisms as discussed earlier.

PAGE 49

38 The Phase 1 ECR chloride exposure results differed extensively from the results of the DCR in this investigation. The +100 mV Phase 1 ECR specimens demanded over a comparable exposure period a cumulative anodic charge of 6000 C, which is 6 times that of DCR in both Phase 1 and in the present investigation, showing a tendency for ECR to corrode to a greater extent in this testing scheme. Also, the ECR 500 mV specimens in Phase 1 corroded actively, demanding ~ 1 mA of anodic cur rent. The comparable DCR specimens demanded a high initial anodic current (>1 mA) but quickly stabilized to ~0.75 A of anodic current, indicating a near passive regime. The ECR open circuit potentials in Phase 1 were initially slightly negative and decreased during the test, indicating a transition from passive to active behavior. The DCR samples behaved in the oppos ite manner, transitioning from active to near passive behavior over the period of immersion. The Phase 1 ECR no chloride exposure specimens behaved under most test conditions similarly to the DCR specimens in this investigation with one notable exception As in the chloride exposure tests, the Phase 1 and present investigation open circuit DCR transitioned from active to passive behavior during the test. In contrast, the open circuit ECR remained at an only slightly negative potential for the entire immer sion period, indicating a continuous passive state. Coating Disbondment Figures 31 33 show the disbondment test results for the chloride and no chloride exposure specimens. The results were similar for most of the polarization regimes and solutions, aver aging at ~5 mm disbondment radius, ~13 MPa pull off strength, and ~1.75 qualitative disbondment rating. The +100 mV chloride exposure tests showed very different results, with ~10mm disbondment radius, ~1 qualitative disbondment rating, and negligible pull off strength. This high amount of coating disbondment surrounding the defects is a direct result of active corrosion and anodic blistering. For the +100 mV chloride exposure specimens, a discolored area surrounding each defect under the coating was obse rved as exemplified in Figure 34 A similar

PAGE 50

39 observation was noted during Phase 1 (Lau and Sags 2009). The average diameter of this discolored area in eight defects examined was ~1.7 cm. It was assumed that the zinc was entirely consumed in that region a nd the corresponding Faradaic charge was calculated using Eq. 12 and 13. The average charge calculated was ~1100C. While this estimate is subject to uncertainty as the entire discolored region was not revealed completely in all cases, the value is in appr oximate agreement with the cumulative charge calculated for these specimens from the current output of the system. This observation supports the assumption of near complete consumption of the zinc in the discolored region. The Phase 1 DCR specimens for a ll polarization regimes in both solutions had comparable disbondment results to those observed in this investigation, although somewhat less disbondment was observed in the present study in the 500 mV and 1000 mV chloride exposures. The conclusions on di sbondment behavior relative to that of ECR consequently remains generally comparable to those noted in the Phase 1 investigation. Implications on Anticipated P erformance of DCR This investigation showed that when defects reached only through the polyme r layer and expose the underlying zinc, it reacted initially at a fast rate with the highly alkaline simulated pore water environment. However, the reaction slowed rapidly approaching a passive regime with very low nominal corrosion rate, even in the pres ence of chloride ions. The length of time that the system could sustain this regime cannot be accurately projected at this time, but specimens exposed for up to 100 days still retained much of the initial zinc layer. This finding is encouraging in demonstr ating an ability of the zinc layer to withstand the nominally highly aggressive alkaline test medium used here, even when the layer was freely exposed. These findings supplement the Phase 1 findings that zinc consumption was slow in the crevice surroundin g a to steel defect, which would simulate the conditions prevalent after all the freely exposed zinc was consumed. These results suggest then that the overall process of zinc wastage in DCR in

PAGE 51

40 concrete pore water is not likely to be rapid, which would be b eneficial to extending the period in which the barrier and galvanic properties of the zinc are maintained. The disbondment and anodic/cathodic behavior of DCR with to zinc defects was similar to that observed for to steel defects in Phase 1, in that it was comparable to or less severe than for ECR. Thus the defect mode examined here, more representative of expected in service surface condition, does not appear to introduce further vulnerability. The accelerated evaluation in this work needs to be su pplemented by longer term exposure testing in concrete and field structures. Investigations to that effect are in progress by Florida DOT and elsewhere (Darwin, et. al 2007).

PAGE 52

41 F igure 18: Combined Polymer and Zinc Coatin g T hickness D istribution. Figure 19: Solid C orrosion Product that Developed on +100 mV Chloride Exposure Specimens. Specimen YG3 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.1 0.15 0.2 0 .25 0.3 0.35 0.4 Cumulative Fraction 1 Total Coating Thickness / mm

PAGE 53

42 Figure 20: Appearance of Defects after Immersion Period. Figure 21: Open Circuit Pote ntial as Func tion of Time for Chloride (Solid Black Line), No chloride Exposure OCP (Dashed), and No chloride Auxiliary OCP (Gray) Duplicate Specimens SPS + NaCl SPS OCP +100 mV 500 mV 1000 mV YG6 YG3 YG22 YG8 YB14 YG15 YG4 YG18 E vs. SCE / mV

PAGE 54

43 Figure 22: Polarization Current Evolution with Time for No Chloride Exposur e Duplicate Specimens. Figure 23: Polarization Current Evolution with Time for +100 mV Chloride Exposure Duplicate Specimens. This figure includes results from the Phase 1 investigation and this investigation. 10 8 6 4 2 0 2 4 0 20 40 60 80 100 120 Time / Days I / uA 500 mV 1000 mV +100 mV 0 200 400 600 800 1000 1200 1400 1600 1800 2000 0 20 40 60 80 100 120 Time / Days |I| / uA DCR to Zinc ECR DCR to Steel

PAGE 55

44 Figure 24: Polarization Current Evolution with Time for 500 mV and 1000 mV Chloride Exposure Duplicate Specimens. Figure 25: Cumulative Anodic Charge for +100 mV Chloride and No chloride Exposure Duplicate Specim ens 500 mV 1000 mV 5 4 3 2 1 0 1 2 0 20 40 60 80 100 120 Time / Days I / uA uA 0.1 1 10 100 1000 10000 1 10 100 1000 10000 Time / Ho urs Cumulative Charge / C C Chloride No chloride

PAGE 56

45 Figure 26: Nominal Corrosion Current Density from EIS Measurements of Duplicate OCP S pecimens as a Function of Exposure T ime. Chloride exposure: filled symbols. No chloride exposure: open symbols. Chloride No chloride

PAGE 57

46 Figure 27: Nominal Zinc Thickness Loss Estim ated from EIS Measurements of OCP Duplicate S pecimens as a Function of Exposure Time. Dashed line indicates the average thickness of the sprayed zinc layer. Chloride No chloride

PAGE 58

47 Figure 28: Diagram Showing Possible Passivation M echanism for Phase 1 DCR (Accardi 2009). Figure 29: Pourbai x Diagram S howing the Effect of Increased pH and Increased Zincate Ion C oncentration (Accardi 2009).

PAGE 59

48 Chloride OCP No chloride OCP Figure 30: Metallographic Cross Sections of C hloride ( YG6) and No C hloride OCP (YB14) Spe cimens after E xposure. The rightmost pictures shows the original ~ 250 m thick polymer coating (just outside the defect); defect region starts at center picture

PAGE 60

49 Table 5: Summary of Phase 1 and This Investigation Results. Phase 1 Inv estigation This Investigation DCR ECR DCR Chloride Exposure +100 mV Current Initial high anodic current, stabilized to ~0.1 mA (anodic) ~1 mA (anodic) Initial high anodic current, stabilized to 0.2 0.6 mA (anodic) Cumulative Charge ~1000 C ~6000 C ~1000 C Average Disbondment Radius ~11mm ~9mm ~10mm Average Disbondment Rating ~1 ~1 ~1 Average Pull off Strength Negligible Negligible Negligible 500 mV Current Initial high anodic current, stabilized to 1 3 uA ~1 mA (anodic) Initial high anod ic current, stabilized to ~0.75 uA (anodic) Cumulative Charge N/A ~1000 C N/A Average Disbondment Radius ~8mm ~7mm ~4mm Average Disbondment Rating ~2 ~1.5 ~1.75 Average Pull off Strength ~10 MPa Negligible ~11 MPa 1000 mV Current Initial hi gh anodic current, stabilized to 1 3 uA (cathodic) 2 3 uA Initial high anodic current, stabilized to ~0.5 uA (cathodic) Average Disbondment Radius ~7mm ~10mm ~4mm Average Disbondment Rating ~2 ~1 ~1.75 Average Pull off Strength ~10 MPa ~13 MPa ~15 MPa OCP Potential Initial ~ 1000 mV, stabilized to ~ 400 mV Initial ~ 300 mV, stabilized to ~ 600 mV Initial ~ 1400 mV, stabilized to ~ 600 mV Average Disbondment Radius ~5mm ~5mm ~4.5mm Average Disbondment Rating ~2 ~2 ~2 Average Pull off Stre ngth ~15 MPa ~18 MPa ~14 MPa

PAGE 61

50 Table 5 : (Continued) Phase 1 Investigation This Investigation DCR ECR DCR No chloride Exposure +100 mV Current Initial high anodic current, stabilized to ~0.2 uA ~0.02 uA Initial high anodic current, stabilized to ~ 0.5 uA Cumulative Charge ~1.5 C ~0.15 C ~8 C Average Disbondment Radius ~6.5mm ~3.5mm ~6mm Average Disbondment Rating ~2 ~1.75 ~1.6 Average Pull off Strength ~15 MPa ~14 MPa ~11 MPa 500 mV Current Initial high anodic current, stabilized to 1 3 uA 2 3 uA Initial high anodic current, stabilized to ~0.5 uA Average Disbondment Radius ~6.5mm ~2mm ~5mm Average Disbondment Rating ~2 ~2 ~1.75 Average Pull off Strength ~10 MPa ~20 MPa ~12 MPa 1000 mV Current Initial high anodic current, stabi lized to 1 3 uA 2 3 uA Initial high anodic current, stabilized to ~1.5 uA Average Disbondment Radius 6mm ~6mm ~4mm Average Disbondment Rating ~2 ~1.75 ~1.75 Average Pull off Strength ~10 MPa ~4 MPa ~12 MPa OCP Potential Initial 1000 mV, stabili zes at ~ 400 mV ~ 200 to 100 mV Initial ~ 1400 mV, stabilizes at ~ 600 mV Average Disbondment Radius ~6mm ~3mm ~5mm Average Disbondment Rating ~2 ~2 ~1.75 Average Pull off Strength ~14 MPa ~20 MPa ~13MPa

PAGE 62

51 Figure 31: Average Coating Disbondment Radius Results. Average of duplicate specimens shows by the horizontal line. Figure 32: Average Qualitative Adhesion Loss Rating Results. Average of duplicate specimens shows by the horizontal line.

PAGE 63

52 Figure 33: Nominal Co ating Pull off Strength Results. The white bars indicate a failure of the cyanoacrylate bond. Average of duplicate specimens shows by the horizontal line. Figure 34: Discolored Area Around Defect Observed for the +100 mV Chloride Exposure Specimens. Spe cimen YG3. 0 5000 10000 15000 20000 25000 YB14 YG15 YG14 YG9 YG21 YG17 YG22 YG20 OCP +100 500 1000 OCP +100 500 1000 SPS SPS + NaCl A s Received Pull off Strength / kPa Negligible

PAGE 64

53 Chapter 5 Conclusions The following conclusions apply to dual coated reinforcing steel with moderate damage simulated with defects extending through the polymer epoxy coating layer exposing the zinc layer: DCR with defects penetr ating through the polymer layer exposing but not when under strong anodic polarization and exposed to chlorides. The extent of corrosion was similar to that seen earlier in DCR with defects penetrating through both the polymer zinc and to steel defects exhibited less corrosion than did ECR under the same conditions. The difference in the extent of co rrosion between ECR and DCR may be due to less voluminous corrosion product build up under the coating and, therefore, less surrounding coating disbondment. The freely corroding (OCP) through polymer DCR specimens in solutions both with and with no chlori de s experienced initially very active dissolution which ended after ~1 day. The zinc exposed at the coating breaks was not completely consumed even after 100 days and there was no visible corrosion product accumulation. These observations are consistent with the development of a passive regime with formation of a calcium hydroxyzincate film. The mechanism for passivation seems to be active even without the possible beneficial occluded environment factors that were noted for to steel defects in the Phase 1 investigation.

PAGE 65

54 The DCR with to zinc defects had comparable disbondment results for most polarization regimes both with and with no chloride s. The notable difference was the +100 mV chloride exposure tests which had more extensive disbondment due to acti ve corrosion during the test and the formation of solid corrosion product under the coating. The DCR with to steel defects and the DCR with to zinc defects had similar amounts of disbondment for all test conditions. The present findings are encouraging in demonstrating an ability of the zinc layer to withstand the nominally highly aggressive alkaline test medium used here, even when the layer is freely exposed. These results suggest then that the overall process of zinc wastage in DCR in concrete pore wat er is not likely to be rapid, which would be beneficial to extending the period in which the barrier and galvanic properties of the zinc are maintained. Evaluation in concrete and sustained field experience is needed to assess overall performance of DCR.

PAGE 66

55 References Crevice Zinc Passivation with High pH External Session submission. Atlanta, GA. 5: Electrochemical Aspects of Galvanized Galvanized Steel Reinforcement in Concrete. Ed. Stephen R. Yeoman. Pgs 111 144. Elsevier. 2004. Reinforci Bentur, A., Diamond, S., and Berke, N.S., Steel Corrosion in Concrete: Fundamentals and Civil Engineering Practice. E & FN Spon. New York. 1997. Broomfield, John P. Corrosion of Steel in Concrete: Understanding, Investig ation, and Repair E & FN Spon. New York. 1998. Corrosion. Vol. 52. No 8. Pages 609 617. NACE International. Houston, TX. August 1996. Canadian Strategic Highway Research Program. March 1992. of Several New Metallic Reinforcing Bars to Chloride Transportation Research Council. Virginia. December 2003. Coated Reinforcing Bars i International. Houston, TX. 2007. Federal Highway Admi nistration. Interim Report. July 2007. Derucher, Kenneth N., Ezeldin, A. Samer, and Korfiatis, George P. Materials for Civil and Highway Engineers, 3rd Edition Prentice Hall. New Jersey. 1994.

PAGE 67

56 Fanous, Fouad S. and Wu, Han ge Decks Reinforced with Epoxy Mid continent Transportation Symposium Proceedings. Center for Transportation Research and Education Iowa State University. Pgs 259 262. 2000. Fontana, Mars G. and Greene, Norbert D. Corrosion Engineering: Secon d Edition. McGraw Hill Book Company. New York. 1978. Report. Oregon Department of Transportation. June 1999. eel in Mortar and Concrete. 1993. New Jersey. 1996. Koch, Gerdardus H., Brongers, Michael P. H., Thompson, Neil G., Virmani, Y. Paul. And RD 01 156. Printed in a supplemental to Materials Performance. NACE International. July 2002. Langill, Thomas J. and Dugan Galvanized Steel Reinforcement in Concrete. Ed. Stephen R. Yeoman. Pgs 87 109. Elsevier. 2004. and Polymer/Zinc Coated Rebar in Simulated C 2009. Coated Reinforcing Steel: North 349 365. 1996. Moreno, Eric I. Final Report. FDOT. May 1996. Coated nd Components 7. Vol. 1. Pgs 491 507. 1996. Fusion Bonded Epoxy Research. Vol. 1 No. 2. Pgs 81 92. April 2004.

PAGE 68

57 Concrete. Pg. 459 478. V. M. Malhorta, Ed. Second International Conference. Monteal, CA. 1991. Orazem, Mark E. and Tribollet, Bernard. Electrochemical Impedance Spectroscopy. John Wiley and Sons, Inc. New Jersey. 2008. Parker, Harry. Simplified Design of Reinforced Concrete, 3rd Edition John Wiley & Sons. New York. 1968. Constructed with Galvanized Reinf Transportation. Ontario, CA. 2005. Association of Corrosion Engineers. Houston, TX. 1974. n M., Mokarem, David W., and Coated Reinforcing Steel in R16. VTRC. February 2000. Report FL/DOT/SMO/89 419. Florida Department of Transportation. December 1989. Coated Reinforcing Steel in Concrete Conference. Paper #325. 1996. Sags Report. WPI no. 0510603, Florida Department of Transportation. May 1994. Coated Rebar in Mari ne Bridges Paper No. 4039, 10, 2008, Las Vegas, Nevada, Published by NACE International, Houston, 2009. Smith, Jeff Coated Rebars in Administration. Public Roads Online. Vol. 60 No. 2. Fall 1996. te Pore Water and in Cement

PAGE 69

58 Coated Reinforcing Steels in Chloride Vol. 50 No. 1. Pgs 72 81. Houston, TX. January 1994.

PAGE 70

59 Appendices

PAGE 71

60 Appendix 1 : Impedance Diagrams Tables 6 and 7 include the EIS fit parameters obtained for each of the four OCP specimens evaluated (designated by i dentifiers per Table 3). The blue highlighted entries are for a fit using the circuit in Figure 11. Entries not highlighted resulted from a fit to the circuit in Figure 8 with the modification indicated in Chapter 4. Yellow highlighted entries denote the few instances in which no adequate fit could be obtained by either approach. Fitting procedures involved working assumptions. Refer to Chapter 4 for limitations on the significance of the values obtained by these procedures. Bode and Nyquist EIS represen tations of the EIS results for each sample tested and for the indicated exposure time are shown in Figures 35 to 114

PAGE 72

61 Appendix 1 : (Continued) Table 6 : Impedance Fit Data for Chloride Specimens

PAGE 73

62 Appendix 1: (Continued) Table 7 : Impedance Fit Data for No chloride Specimen

PAGE 74

63 Appendix 1: (Continued) YG6 (Chloride) Impedance Diagrams Test a (0.06 days) Test a (0.06 days) Figure 35 : Bode and Nyquist Diagrams for YG6 at 0.06 Days of Exposure.

PAGE 75

64 Appendix 1: (Con tinued) Test b (0.09 days) Test b (0.09 days) Figure 36 : Bode and Nyquist Diagrams for YG6 at 0.09 Days of Exposure.

PAGE 76

65 Appendix 1: (Continued) Test c (0.13 days) Test c (0.13 days) Figure 37 : Bode and Nyquist Diagrams for YG6 at 0.13 Days of Exposure.

PAGE 77

66 Appendix 1: (Continued) Test d (0.85 days) Test d (0.85 days) Figure 38 : Bode and Nyquist Diagrams for YG6 at 0.85 Days of Exposure.

PAGE 78

67 Appendix 1: (Continued) Test e (1.22 days) Test e (1.22 days) Figure 39 : Bode and Nyquist Diagrams for YG6 at 1.22 Days of Exposure.

PAGE 79

68 Appendix 1: (Continued) Test f (1.93 days) Test f (1.93 days) Figure 40 : Bode and Nyquist Diagrams for YG6 at 1.93 Days of Exposure.

PAGE 80

69 Appe ndix 1: (Continued) Test g (3.07 days) Test g (3.07 days) Figure 41 : Bode and Nyquist Diagrams for YG6 at 3.07 Days of Exposure.

PAGE 81

70 Appendix 1: (Continued) Test h (5.97 days) Test h (5.97 days) Figure 42 : Bode and Nyquist Diagrams for YG6 at 5.97 Days of Exposure.

PAGE 82

71 Appendix 1: (Continued) Test i (9.93 days) Test i (9.93 days) Figure 43 : Bode and Nyquist Diagrams for YG6 at 9.93 Days of Exposure.

PAGE 83

72 Appendix 1: (Continued) Test j (13 days) Te st j (13 days) Figure 44 : Bode and Nyquist Diagrams for YG6 at 13 Days of Exposure.

PAGE 84

73 Appendix 1: (Continued) Test k (42 days) Test k (42 days) Figure 45 : Bode and Nyquist Diagrams for YG6 at 42 Days of Exposure.

PAGE 85

74 Appe ndix 1: (Continued) Test l (76.83 days) Test l (76.83 days) Figure 46 : Bode and Nyquist Diagrams for YG6 at 76.83 Days of Exposure.

PAGE 86

75 Appendix 1: (Continued) Test m (85.92 days) Test m (85.92 days) Figure 47 : Bode and Nyquist D iagrams for YG6 at 85.92 Days of Exposure.

PAGE 87

76 Appendix 1: (Continued) Impedance Diagrams YG21 (Chloride) Test a (0.05 days) Test a (0.05 days) Figure 48 : Bode and Nyquist Diagrams for YG21 at 0.05 Days of Exposure.

PAGE 88

77 Appen dix 1: (Continued) Test b (0.08 days) Test b (0.08 days) Figure 49 : Bode and Nyquist Diagrams for YG21 at 0.08 Days of Exposure.

PAGE 89

78 Appendix 1: (Continued) Test c (0.12 days) Test c (0.12 days) Figure 50 : Bode and Nyquist Diagrams for YG21 at 0.12 Days of Exposure.

PAGE 90

79 Appendix 1: (Continued) Test d (0.84 days) Test d (0.84 days) Figure 51 : Bode and Nyquist Diagrams for YG21 at 0.84 Days of Exposure.

PAGE 91

80 Appendix 1: (Continued) Test e (1.21 days) Test e (1.21 days) Figure 52 : Bode and Nyquist Diagrams for YG21 at 1.21 Days of Exposure.

PAGE 92

81 Appendix 1: (Continued) Test f (1.92 days) Test f (1.92 days) Figure 53 : Bode and Nyquist Diagrams for YG21 at 1.92 Days of Exposure.

PAGE 93

82 Appendix 1: (Continued) Test g (3.06 days) Test g (3.06 days) Figure 54 : Bode and Nyquist Diagrams for YG21 at 3.06 Days of Exposure.

PAGE 94

83 Appendix 1: (Continued) Test h (5.95 days) Test h (5.95 days) Figure 55 : Bode and Nyquist Diagrams for YG21 at 5.95 Days of Exposure.

PAGE 95

84 Appendix 1: (Continued) Test i (9.92 days) Test i (9.92 days) Figure 56 : Bode and Nyquist Diagrams for YG21 at 9.92 Days of Exposure.

PAGE 96

85 Appendix 1: (Continued) Test j (12.99 days) Test j (12.99 days) Figure 57 : Bode and Nyquist Diagrams for YG21 at 12.99 Days of Exposure.

PAGE 97

86 Appendix 1: (Continued) Test k (41.99 days) Test k (41.99 days) Figure 58 : Bode and Nyquist Diagrams for YG21 at 41 .99 Days of Exposure.

PAGE 98

87 Appendix 1: (Continued) Test l (76.83 days) Test l (76.83 days) Figure 59 : Bode and Nyquist Diagrams for YG21 at 76.83 Days of Exposure.

PAGE 99

88 Appendix 1: (Continued) Test m (85.92 days) Test m ( 85.92 days) Figure 60: Bode and Nyquist Diagrams for YG21 at 85.92 Days of Exposure.

PAGE 100

89 Appendix 1: (Continued) YB14 ( No chloride ) Impedance Diagrams Test a (0.04 days) Test a (0.04 days) Figure 61 : Bode and Nyquist Diagrams for YB14 at 0.04 Days of Exposure.

PAGE 101

90 Appendix 1: (Continued) Test b (0.1 days) Test b (0.1 days) Figure 62 : Bode and Nyquist Diagrams for YB14 at 0.1 Days of Exposure.

PAGE 102

91 Appendix 1: (Continued) Test c (0.17 days) Test c (0.17 days) Figure 63 : Bode and Nyquist Diagrams for YB14 at 0.17 Days of Exposure.

PAGE 103

92 Appendix 1: (Continued) Test d (0.22 days) Test d (0.22 days) Figure 64 : Bode and Nyquist Diagrams for YB14 at 0.22 Days of Exposure.

PAGE 104

93 Appendix 1: (Continued) Test e (0.96 days) Test e (0.96 days) Figure 65 : Bode and Nyquist Diagrams for YB14 at 0.96 Days of Exposure.

PAGE 105

94 Appendix 1: (Continued) Test f (1.03 days) Test f (1.03 days) Figure 66 : Bod e and Nyquist Diagrams for YB14 at 1.03 Days of Exposure.

PAGE 106

95 Appendix 1: (Continued) Test g (1.18 days) Test g (1.18 days) Figure 67 : Bode and Nyquist Diagrams for YB14 at 1.18 Days of Exposure.

PAGE 107

96 Appendix 1: (Conti nued) Test h (2 days) Test h (2 days) Figure 68 : Bode and Nyquist Diagrams for YB14 at 2 Days of Exposure.

PAGE 108

97 Appendix 1: (Continued) Test i (2.95 days) Test i (2.95 days) Figure 69 : Bode and Nyquist Diagrams for YB14 at 2.95 D ays of Exposure.

PAGE 109

98 Appendix 1: (Continued) Test j (5.94 days) Test j (5.94 days) Figure 70 : Bode and Nyquist Diagrams for YB14 at 5.94 Days of Exposure.

PAGE 110

99 Appendix 1: (Continued) Test k (7 days) Test k (7 days ) Figure 71 : Bode and N yquist Diagrams for YB14 at 7 Days of Exposure.

PAGE 111

100 Appendix 1: (Continued) Test l (7.95 days) Test l (7.95 days) Figure 72 : Bode and Nyquist Diagrams for YB14 at 7.95 Days of Exposure.

PAGE 112

101 Appe ndix 1: (Continued) Test m (9.93 days) Test m (9.93 days) Figure 73 : Bode and Nyquist Diagrams for YB14 at 9.93 Days of Exposure.

PAGE 113

102 Appendix 1: (Continued) Test n (10.93 days) Test n (10.93 days) Figure 74 : Bode and Nyquist Di agrams for YB14 at 10.93 Days of Exposure.

PAGE 114

103 Appendix 1: (Continued) Test o (12.94 days) Test o (12.94 days) Figure 75 : Bode and Nyquist Diagrams for YB14 at 12.94 Days of Exposure.

PAGE 115

104 Appendix 1: (Continued) Test p (14.01 days) Test p (14.01 days) Figure 76 : Bode and Nyquist Diagrams for YB14 at 14.01 Days of Exposure.

PAGE 116

105 Appendix 1: (Continued) Test q (15.99 days) Test q (15.99 days) Figure 77 : Bode and Nyquist Diagrams for YB14 at 15.9 9 Days of Exposure.

PAGE 117

106 Appendix 1: (Continued) Test r (16.95 days) Test r (16.95 days) Figure 78 : Bode and Nyquist Diagrams for YB14 at 16.95 Days of Exposure.

PAGE 118

107 Appendix 1: (Continued) Test s (19.94 days) Tes t s (19.94 days) Figure 79 : Bode and Nyquist Diagrams for YB14 at 19.94 Days of Exposure.

PAGE 119

108 Appendix 1: (Continued) Test t (21.01 days) Test t (21.01 days) Figure 80 : Bode and Nyquist Diagrams for YB14 at 21.01 Days of Exposure.

PAGE 120

109 Appendix 1: (Continued) Test u (21.94 days) Test u (21.94 days) Figure 81 : Bode and Nyquist Diagrams for YB14 at 21.94 Days of Exposure.

PAGE 121

110 Appendix 1: (Continued) Test v (23.07 days) Test v (23.07 days) Figur e 82 : Bode and Nyquist Diagrams for YB14 at 23.07 Days of Exposure.

PAGE 122

111 Appendix 1: (Continued) Test w (24.16 days) Test w (24.16 days) Figure 83 : Bode and Nyquist Diagrams for YB14 at 24.16 Days of Exposure.

PAGE 123

112 Appen dix 1: (Continued) Test x (27.03 days) Test x (27.03 days) Figure 84 : Bode and Nyquist Diagrams for YB14 at 27.03 Days of Exposure.

PAGE 124

113 Appendix 1: (Continued) Test y (34.02 days) Test y (34.02 days) Figure 85 : Bode and Nyquist Diagrams for YB14 at 34.02 Days of Exposure.

PAGE 125

114 Appendix 1: (Continued) Test z (63.08 days) Test z (63.08 days) Figure 86 : Bode and Nyquist Diagrams for YB14 at 63.08 Days of Exposure.

PAGE 126

115 Appendix 1: (Continued) T est aa (97.91 days) Test aa (97.91 days) Figure 87 : Bode and Nyquist Diagrams for YB14 at 97.91 Days of Exposure.

PAGE 127

116 Appendix 1: (Continued) YG7 ( No chloride ) Impedance Diagrams Test a (0.05 days) Test a (0.05 days) Figure 88 : Bo de and Nyquist Diagrams for YG7 at 0.05 Days of Exposure.

PAGE 128

117 Appendix 1: (Continued) Test b (0.13 days) Test b (0.13 days) Figure 89 : Bode and Nyquist Diagrams for YG7 at 0.13 Days of Exposure.

PAGE 129

118 Appendix 1: (Continued) Test c (0.20 days) Test c (0.20 days) Figure 90 : Bode and Nyquist Diagrams for YG7 at 0.20 Days of Exposure.

PAGE 130

119 Appendix 1: (Continued) Test d (0.24 days) Test d (0.24 days) Figure 91 : Bode and Nyquist Diagrams for YG7 at 0.24 Days of Exposure.

PAGE 131

120 Appendix 1: (Continued) Test e (0.99 days) Test e (0.99 days) Figure 92 : Bode and Nyquist Diagrams for YG7 at 0.99 Days of Exposure.

PAGE 132

121 Appendix 1: (Continued) Test f (1.03 days) Test f (1.03 days) Fi gure 93 : Bode and Nyquist Diagrams for YG7 at 1.03 Days of Exposure.

PAGE 133

122 Appendix 1: (Continued) Test g (1.19 days) Test g (1.19 days) Figure 94 : Bode and Nyquist Diagrams for YG7 at 1.19 Days of Exposure.

PAGE 134

123 Appendix 1: (Co ntinued) Test h (2.01 days) Test h (2.01 days) Figure 95 : Bode and Nyquist Diagrams for YG7 at 2.01 Days of Exposure.

PAGE 135

124 Appendix 1: (Continued) Test i (2.95 days) Test i (2.95 days) Figure 96 : Bode and Nyquist Diagrams for YG7 at 2.95 Days of Exposure.

PAGE 136

125 Appendix 1: (Continued) Test j (5.94 days) Test j (5.94 days) Figure 97 : Bode and Nyquist Diagrams for YG7 at 5.94 Days of Exposure.

PAGE 137

126 Appendix 1: (Continued) Test k (7.01 days) Test k (7.01 days) Figure 98 : Bode and Nyquist Diagrams for YG7 at 7.01 Days of Exposure.

PAGE 138

127 Appendix 1: (Continued) Test l (7.96 days) Test l (7.96 days) Figure 99 : Bode and Nyquist Diagrams for YG7 at 7.96 Days of Exposure.

PAGE 139

128 Append ix 1: (Continued) Test m (9.94 days) Test m (9.94 days) Figure 100 : Bode and Nyquist Diagrams for YG7 at 9.94 Days of Exposure.

PAGE 140

129 Appendix 1: (Continued) Test n (10.94 days) Test n (10.94 days) Figure 101 : Bode and Nyquist Diagra ms for YG7 at 10.94 Days of Exposure.

PAGE 141

13 0 Appendix 1: (Continued) Test o (12.95 days) Test o (12.95 days) Figure 102 : Bode and Nyquist Diagrams for YG7 at 12.95 Days of Exposure.

PAGE 142

131 Appendix 1: (Continued) Test p (14.06 d ays) Test p (14.06 days) Fig ure 103 : Bode and Nyquist Diagrams for YG7 at 14.06 Days of Exposure.

PAGE 143

132 Appendix 1: (Continued) Test q (16 days) Test q (16 days) Figure 104 : Bode and Nyquist Diagrams for YG7 at 16 Days of Exposure.

PAGE 144

133 Appendix 1: (Continued) Test r (16.96 days) Test r (16.96 days) Figure 105 : Bode and Nyquist Diagrams for YG7 at 16.96 Days of Exposure.

PAGE 145

134 Appendix 1: (Continued) Test s (19.95 days) Test s (19.95 days) Figure 106 : Bode and Nyquist Diagrams for YG7 at 19.95 Days of Exposure.

PAGE 146

135 Appendix 1: (Continued) Test t (21.02 days) Test t (21.02 days) Figure 107 : Bode and Nyquist Diagrams for YG7 at 21.02 Days of Exposure.

PAGE 147

136 Appendix 1: (Continued) Test u (21.94 days) Test u (21.94 days) Figure 108 : Bode and Nyquist Diagrams for YG7 at 21.94 Days of Exposure.

PAGE 148

137 Appendix 1: (Continued) Test v (23.09 days) Test v (23.09 days) Figure 109 : Bode and Nyqui st Diagrams for YG7 at 23.09 Days of Exposure.

PAGE 149

138 Appendix 1: (Continued) Test w (24.16 days) Test w (24.16 days) Figure 110 : Bode and Nyquist Diagrams for YG7 at 24.16 Days of Exposure.

PAGE 150

139 Appendix 1: (Continued) Test x (27.04 days) Test x (27.04 days) Figure 111 : Bode and Nyquist Diagrams for YG7 at 27.04 Days of Exposure.

PAGE 151

140 Appendix 1: (Continued) Test y (34.03 days) Test y (34.03 days) Figure 112 : Bode and Nyquist Diagrams for YG7 at 34.03 Day s of Exposure.

PAGE 152

141 Appendix 1: (Continued) Test z (63.08 days) Test z (63.08 days) Figure 113 : Bode and Nyquist Diagrams for YG7 at 63.08 Days of Exposure.

PAGE 153

142 Appendix 1: (Continued) Test aa (97.93 days) Test aa (97.93 days) Figure 114 : Bode and Nyquist Diagrams for YG7 at 97.93 Days of Exposure.

PAGE 154

143 Appendix 2 : Metallographic Pictures The following contains metallographic pictures of specimens. The magnification is 15x except where noted. The locations pi ctured are on the defect and off the defect as illustrated below. These will allow a comparison of zinc thickness between that on the defect surface and that under the coating. Figure 115 : Metallographic Picture Location Diagram

PAGE 155

144 Appendix 2: (Continued) OCP No chloride Figure 116 : YG7 defect Figure 117 : YG7 off defect

PAGE 156

145 Appendix 2: (Continued) +100 mV No chloride Figure 118 : YG2 defect Figure 119 : YG2 off defect

PAGE 157

146 Appendix 2: (Continued) 500 mV No chloride Figure 120 : YG4 defect Figure 121 : YG4 off defect

PAGE 158

147 Appendix 2: (Continued) 1000 mV No chloride Figure 122 : YG18 defect Figure 123 : YG18 off defect

PAGE 159

148 Appendix 2: (Continued) OCP Chloride Figure 124 : YG6 defect Figure 125 : YG6 off defect

PAGE 160

149 Appen dix 2: (Continued) +100 mV Chloride Figure 126 : YG3 defect Figure 127 : YG3 off defect 1 mm

PAGE 161

150 Appendix 2: (Continued) 500 mV Chloride Figure 128 : YG11 defect Figure 129 : YG11 off defect

PAGE 162

151 Appendix 2: (Continued) 1000 mV Chloride Figure 130 : YG8 defect Figure 131 : YG8 off defect