University of Ghana http://ugspace.ug.edu.gh EFFECT OF CLAY POZZOLANA ON THE CORROSION BEHAVIOUR OF STEEL REINFORCEMENT IN CONCRETE. BERNARD KWAME MUSSEY (10552064) IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF MASTER OF PHILOSOPHY IN MATERIALS SCIENCE AND ENGINEERING JULY 2017 University of Ghana http://ugspace.ug.edu.gh DECLARATION It is hereby declared that this thesis is my own work towards the MPhil. Degree and that, to the best of my knowledge, it contains neither material previously published by another person nor material which has been accepted for the award of any degree of the University, except where due acknowledgement has been made in the text. Bernard Kwame Mussey ………………………….. ……………………… (Candidate) Signature Date Certified by: Dr. Lucas Nana Wiredu Damoah………………………….. ……………………… (Supervisor) Signature Date Certified by: David Sasu Konadu ………………………….. ……………………… (Co - Supervisor) Signature Date Certified by: Dr. Abu Yaya ………………………….. ……………………… (Head of Department) Signature Date ii University of Ghana http://ugspace.ug.edu.gh ABSTRACT This study investigated the corrosion resistance of steel reinforcement in two types of concrete mixtures – ordinary Portland (OPC) used for concrete and clay pozzolana cement (CPC) used for concrete. These two concrete types were exposed to 3% and 5% (W/V) of Sodium Chloride (NaCl) and 3% (W/V) Calcium Hypochlorite (Ca(OCl)2) that acted as corrosion media. Mechanical properties and electrochemical analysis were performed over 1, 7, 15 and 23 days. The results indicated that concrete mixtures that contained clay pozzolana cement recorded an average compressive strength of 15.17 MPa while that which contained ordinary Portland cement was 22.75 MPa. Pull – out forces recorded for steel embedded concrete samples that contained clay pozzolana cement in the corrosion media recorded average values of 71, 65.33, 57.7 and 54.67 MegaNewtons (MN) over 1, 7, 15 and 23 days respectively as compared to 80, 70.1, 61, and 49 MN over the same period for concretes without clay pozzolana. Values of corrosion rates obtained from anodic and cathodic curves from Tafel plots revealed that samples exposed to 5%(W/V) NaCl had higher corrosion rates followed by those exposed to 3%(W/V) NaCl and 3% Ca(OCl)2 in that order. XRD analysis showed high presence of Alite (3CaO.SiO2) in all concrete types, however concrete containing clay pozzolana, had higher amounts of alumina – silicates. The XRD also showed low intensity peak for sodium chloride (NaCl) in pozzolana containing concretes and relatively higher intensity peak for ordinary Portland cement (OPC) concretes. SEM showed shallow to deep corrosion pits on steel rods embedded in OPC containing concrete. No corrosion pits were found on steel rods embedded in CPC containing concretes, even though some surface corrosion activities were visible. iii University of Ghana http://ugspace.ug.edu.gh DEDICATION This work is dedicated to my beautiful wife Mary Beann – Mussey, and my lovely children Ellis Ataaku Beann – Mussey, Glenda Lisa Affoh Mussey and Adriel Amihere –Ackah Mussey. Your love and support is greatly appreciated. iv University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT My first appreciation goes to the Almighty God for His grace and mercy, for giving me the strength and knowledge to undertake this research work. My deep-seated appreciation also goes Mr. Emmanuel Okumi Andoh, Pro vice chancellor of Takoradi Technical University, for his immense support he gave in order to pursue the graduate programme. I am also grateful to my supervisor, Dr. Lucas Nana Wiredu Damoah, and co – supervisor, David Sasu Konadu for their invaluable suggestions and supervision. My heart-felt appreciation goes to Nii Ayitey Akoto, for helping me with most of the laboratory analysis. My gratitude also goes to the University of Ghana Physics and Animal Biology departments for helping me with the XRD and SEM scans respectively. Finally, I am very grateful to my entire family for their support and prayers. God richly bless you all! v University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ii ABSTRACT iii DEDICATION iv ACKNOWLEDGEMENT v TABLE OF CONTENTS vi LIST OF FIGURES ix LIST OF TABLES xiii CHAPTER ONE 1 INTRODUCTION 1 1.1 Background 1 1.2 Motivation 2 1.3 Problem statement 3 1.4 Aims 4 1.5 Objectives 4 CHAPTER TWO 5 LITERATURE REVIEW 5 2.1 Introduction 5 2.2 Portland cement 6 2.3 Pozzolanas 7 2.4 Corrosion mechanism of steel in reinforced concretes 9 2.4.1 The penetration mechanism 12 2.4.2 Film or layer breaking mechanism 13 vi University of Ghana http://ugspace.ug.edu.gh 2.4.3 Adsorption mechanism 13 2.5 Corrosion rate quantification and methods 14 2.5.1 Potentiodynamic polarization curve 16 2.5.2 LRP technique 17 2.6 Steel reinforcement – concrete bond 18 2.6.1 Pull – out test 19 2.6.2 Embedded bar test 20 2.6.3 Beam test 20 2.7 Effects of corrosion on concrete structures 22 2.8 Effects of corrosion on bond strength 23 CHAPTER THREE 25 METHODS AND EXPERIMENTS 25 3.1 Introduction 25 3.2 Materials and specimen preparation 26 3.3 Electrolytic solutions and corrosion environments 28 3.4 Accelerated corrosion set-up and measurements 29 3.5 Strength test measurement 30 3.5.1 Pull - Out test set – up and measurement procedure 30 3.5.2 Compressive strength test set – up and measurement procedure 31 3.6 X – Ray diffraction (XRD) 33 3.7 Surface morphology of reinforcing steel 33 CHAPTER FOUR 35 RESULTS AND DISCUSSIONS 35 4.1 Introduction 35 vii University of Ghana http://ugspace.ug.edu.gh 4.2 Pull – out test 36 4.3 Pull – out responses and the effects of corrosion environments and time 39 4.4 Compressive strength test 40 4.4.1 Effect of cement mix proportions and types on compressive strength 44 4.4.2 Effect of Chloride Concentration on Compressive Strength 44 4.5 Electrochemical test 45 4.5.1 Tafel plots and corrosion current density estimation 45 4.5.2 Corrosion rate calculations 52 4.5.3 Effect of cement mix proportions and types 54 4.5.4 Effect of chloride concentration 55 4.5.5 Effect of acceleration time 55 4.6 X – ray diffractometry 56 4.5 Surface morphology investigation 58 CHAPTER FIVE 65 CONCLUSIONS AND RECOMMENDATION 65 5.1 Conclusions 65 5.1 Recommendations 68 REFERENCES 69 APPENDIX 76 MATLAB Codes for tangents drawn from polarization curves 76 viii University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 2.1. Pourbaix diagram for iron in water of temperature of 25 o C (Angst, 2011). 10 Figure 2.2. Polarization curve for steel (MacDougall & Graham, 2002). 12 Figure 2.3. Schematic representation of chloride induced pitting corrosion (MacDougall & Graham, 2002). 14 Figure 3.1: Schematic representation of corrosion rate measurement set – up. 30 Figure 3.2: Pull – Out test system set – up. 31 Figure 3.3: Compressive strength test set – up. 32 Figure 4.1: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 1 day. 37 Figure 4.2: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 7 days. 37 Figure 4.3: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 15 days. 38 Figure 4.4: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 23 days. 38 Figure 4.5: Pull – out forces for reinforcing rod embedded in Type 2 (OPC) and Type 3 (PPC) concrete mix for 1 – 23 days. 39 Figure 4.6: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for duration of 1 days. 41 Figure 4.7: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 7 days. 41 Figure 4.8: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 15 days. 42 ix University of Ghana http://ugspace.ug.edu.gh Figure 4.9: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 23 days. 42 Figure 4.10: Compressive strength of Type 2 (OPC) and Type 3 (PPC) concrete for duration of 1 – 23 days. 43 Figure 4.11: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for a day. 45 Figure 4.12: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for a days. 46 Figure 4.13: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for a day. 46 Figure 4.14: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 7 days. 47 Figure 4.15: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 7 days. 48 Figure 4.16: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 7 days. 48 Figure 4.17: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 15 days. 49 Figure 4.18: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 15 days. 49 Figure 4.19: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 15 days. 50 Figure 4.20: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 23 days. 50 x University of Ghana http://ugspace.ug.edu.gh Figure 4.21: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 23 days. 515160 Figure 4.22: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 23 days. 515161 Figure 4.23: Corrosion rates of steel reinforcement over duration of 23 days. 545464 Figure 4.24: X – ray diffractometry showing some mineralogy type 3 (OPC) concrete mix after exposure to 5% (W/V) Sodium Chloride (NaCl) corrosion environment. 575767 Figure 4.25: X – ray diffractometry showing some mineralogy type 2 (PPC) concrete mix after exposure to 5%(W/V) NaCl corrosion environment. 575768 Figure 4.26: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 3% NaCl. (a) One-day exposure. Dark areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 595969 Figure 4.27: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 3% NaCl. (a) One-day exposure. Dark areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 606070 Figure 4.28: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 5% NaCl. (a) One-day exposure. Rounded particles show areas of corrosion activity and dark areas and spots show corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 616171 Figure 4.29: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 5% NaCl. (a) One-day exposure. Dark xi University of Ghana http://ugspace.ug.edu.gh areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 626272 Figure 4.30: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 3% Ca(OCl)2. Dark areas show areas of corrosion activity or corrosion pits. (a) One-day exposure. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 636373 Figure 4.31: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 3% Ca(OCl)2. Dark areas show areas of corrosion activity or corrosion pits. (a) one-day exposure. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. 646474 xii University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 2.1: Chemical compositions of some natural Pozzolanas (Taylor, 1997). 889 Table 2.2: Compositions of phases in Portland cement clinker in mass percentages (Taylor, 1997). 778 Table 3.1: Mix proportions of concrete and compressive strengths. 272728 Table 3.2: Mechanical properties of high tensile steel. 272728 Table 3.3: Chemical composition of high tensile steel (Wt.%). 272728 Table 3.4: chemical compositions (Wt.% of oxides in ordinary Portland and clay Pozzolana cements. 282829 xiii University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.1 Background Cement is a mixture of powdered gypsum (CaSO4 ∙ 2H2O) and clinker particles. Clinker is obtained by subjecting a mixture of limestone and clay or shale to a high temperature treatment (Lorenzo, Goni, & Guerrero, 2003). The ordinary Portland cement (OPC) and the clay pozzolana cements (CPC) are of much interest in this work. When mixed with other coarse aggregate in the presence of water, a porous solid called concrete is formed. Pozzolanas include a wide range of predominately glassy materials like fly-ash (FA); micro- silica (MS) or silica fume (SF), waste material from the silicon and ferrosilicon metal industry; and natural pozzolanas, and geologic deposits of clay. These materials may be calcined prior to use in order to increase their activity (Boakye, 2012). Pozzolanic materials also bring in other technical advantages such as low heat of hydration and high ultimate strength. The higher strength of concrete with pozzolanas over time is as a result of the pozzolanic reactions increasing the amount of calcium silicate hydrates (C-S-H) while diminishing Ca(OH)2 (Taylor, 1997). Reinforced concrete is the term used to describe a concrete with steel rods embedded in them. Reinforced concrete (RC) combines the good compressive strength properties of concrete and the excellent mechanical strength properties of steel (Bouzoubaâ, Zhang, Malhotra, & Golden, 1999). Thus, designers, architects and civil engineers employ RC materials to attain high mechanical strength, fire resistance, durability, shape adaptability and low cost requirements (Bouzoubaâ et al., 1999). Corrosion is the deterioration of steel reinforcement in the concrete, as a result of interaction with moisture, chloride ions and sulfate ions. 1 University of Ghana http://ugspace.ug.edu.gh 1.2 Motivation Reinforced concrete, despite the aforementioned positive characteristics, are prone to corrosion in the presence of moisture, chloride ions, sulfate ions, to mention a few (Abosrra, 2010; Yuan & Marosszeky, 1991). The deterioration or corrosion of RCs remains a challenge in the construction industry. Currently, a lot of research is ongoing to present the best techniques that can be implemented to lengthen useful life of RCs. Coating the steel rebar with a layer of zinc by submerging it in a bath at temperatures 450 – 460 o C, the so - called hot – dip galvanization has gained interest. The zinc coating serves as shielding barrier against destructive corrosion activity (Hamad & Mike, 2005; Tan & Hansson, 2008; Yeomans, 1998). Impressing current through the reinforcing rebar to create a cathodic protection against corrosion is also an interesting technique (Morozov, Castela, Dias, & Montemor, 2013; Pedeferri, 1996). Different organic compounds have also been found to act as corrosion inhibitors in concrete (Assaad & Issa, 2012; Fayala, Dhouibi, Nóvoa, & Ben Ouezdou, 2013). Another effective yet simple technique is the use of different cement types and admixture proportions to control corrosion activity. Varying mixtures of ordinary Portland and pozzolana cements have been used to study the rate of corrosion attack of reinforced concrete exposed to different corrosive media. Due to the low cost of this technique, the demand for pozzolana cement (chiefly made from pozzolans) as a constituent for concrete admixture has increased (Bouteiller, Cremona, Baroghel-Bouny, & Maloula, 2012; Reyes-Diaz et al., 2011; Román, Vera, Bagnara, M., & Aperador, 2014). The presence of calcium silica hydrate in the concrete mixtures impedes the influence of the respective corrosive media such as chlorides. The use of clay pozzolana cements has gained popularity in Ghana after a campaign by Council for Scientific and Industrial Research – Building and Road 2 University of Ghana http://ugspace.ug.edu.gh Research Institute (CSIR – BRRI) was launched to raise the awareness on its technical, economic and environmental merits. It has been established that replacing 5 – 30% by mass of ordinary Portland cement with clay pozzolana cement and mixing in concrete, gives important characteristics to the resulting admixture. These characteristics range from mechanical to chemical advantages (Boakye, 2012; Hammond, 1983; Lea, 1940; Malquori, 1960). However, pozzolana cements are noted for their slow strength development resulting in low early strengths and slow nature of setting (Boakye, 2012). 1.3 Problem statement One of the main reasons for the premature failure of reinforced concrete structures is corrosion of the reinforcements (Abosrra, 2010). Industries that employ large quantities of steel reinforced concrete have been facing the unending challenge to reduce or prevent corrosion. Investigations have been made on the use of blended cement made with supplementary cementitious materials (such as: fly ash, silica fume, and blast furnace slag) (Canadian Standard Association, 2000; Philips, 1993) and clay pozzolana (Bouteiller et al., 2012; Reyes-Diaz et al., 2011; Román et al., 2014; Taylor, 1997). These reports have studied the rate of corrosion attack of reinforced concrete exposed to different ion environments. Corrosion of reinforcing steel bar causes early deterioration of most concrete structures and it is a very costly problem, not only in terms of its financial implications but also for its structural safety. Multiple protection strategies have been developed to guarantee a long service life including the use of corrosion inhibitors, protective steel coatings and cathodic protection (Abosrra, 2010). A major cause of steel reinforcement corrosion is the presence of chlorides, from chloride-contaminated aggregates and chloride containing admixtures used during construction. Also, corrosion from the penetration of chloride ions from sea spray is 3 University of Ghana http://ugspace.ug.edu.gh also common along the coastal areas of the country. As chlorides diffuse into concrete, chloride ions accumulate on the surface of the reinforcing bars and thus the passive film on the reinforcement breaks down with consequent corrosion of reinforcing steel (Dehghanian & Locke, 1982). This project seeks to investigate the corrosion behaviour of steel reinforcement in concrete containing clay Pozzolana in the presence of various concentrations of chlorides. 1.4 Aim The main aim of this study is to investigate the effect of clay pozzolana on the corrosion behaviour of steel reinforcement in concrete. 1.5 Objectives To achieve the main aim, the following specific objectives were set: i. Investigation of the effect of clay pozzolana addition on the mechanical strength of reinforced concrete. ii. Investigation of the effect of clay pozzolana addition on the pullout force/strength of reinforcing steel in concrete. iii. Investigation of the effect of clay Pozzolana on the corrosion resistance of concrete steel reinforcement to chloride attack. iv. Measurement of the mechanical and electrochemical properties of reinforced concretes following corrosion attack. 4 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.1 Introduction Corrosion of steel reinforcement in concrete occurs as result of moisture, chloride ions and sulfate ions present in its environment. This interaction between the metal and its environment produces by-products than may be solid, liquid or gaseous in state. The physical and chemical characteristics of such by-products are significant since they recurrently affect the subsequent rate of corrosion (Scully, 1975). The pH of concrete, which is commonly in the range 11.5 – 13.6 provides somewhat protective layer on steel surfaces present in reinforced concrete. This protective layer retards corrosion of the steel reinforcements. Furthermore, reinforced concretes with lower water/cement ratios possess minimal permeability values hence decreasing penetration rates of corrosion agents such as chlorides through the concretes and reaching the steel surfaces (Ahmad, 2003). This chapter reviews works on corrosion phenomena of steel in reinforced concretes. The chapter begins with the description of pozzolanas, pozzolana cements and Portland cements. Corrosion principles and mechanisms in reinforced concrete would follow, and then the aqueous environments that stimulate corrosion. The review would also include the passivity of steel in reinforced concretes. Localized corrosion types would also be reported as well as some common principal electrochemical methods used in quantifying corrosion in reinforced concretes. Also, steel – concrete bond strengths and structure behaviours of reinforced concrete subjected to corrosion activity would be reviewed. Statements outlining corrosion prevention and controls adapted from literatures will also be enumerated. 5 University of Ghana http://ugspace.ug.edu.gh 2.2 Portland cement Portland cement is produced by heating a mixture of clay, limestone and other materials of appropriate chemistry at temperatures ranging from 1400 to 1600oC. One important characteristic of cement is its hydraulic nature. It chemically reacts with water to set and harden (Neville, 1996). Calcium silicates are formed after the mixtures under a chemical reaction thermally. Clinker, which is produced after a partial fusion process, is mixed with gypsum and finely grounded to produce cement. Clinker is chemically made of Al2O3, Ca2SiO4 and Fe2O3. Gypsum is calcium sulphate and controls the rate of set and the rate of strength development. Portland cement is 80% composed of clinker and other minor oxide constituents such as K2O, SO3, Na2O and MgO (Lorenzo et al., 2003; Malquori, 1960). Al2O3 also known as alite, is tricalcium silicate (Ca3SiO5) whose chemical composition and crystal structure have been altered by ionic substitution and it makes 50 to 70% of all cement clinkers. Alite reacts quickly with water, and as such for a 23 days strength development, it is considered the most important of the constituent phases. Another development strength – contributing component of cement clinker is belite (Ca2SiO4). It is 15 to 30% abundant in cement clinkers. Belite reacts wit water slowly and is thought to contribute to the later – age development strength beyond 23 days (Taylor, 1997). Aluminate (Al2O3) with a percentage composition of 5 to 10% in cement clinker react quickly with water and makes cement achieve quick setting but only with calcium sulphate present (Lea, 1940). Ferrite (Ca2AlFeO5) makes up 5 to 15% of clinker. It obtained by varying aluminum – iron ratios with ionic substitutions together with compositions modification of tetracalcium aluminoferrite. Its reaction rate with water varies. Due to its chemical composition or other characteristics, its reaction rate with water is sometimes high at the onset and as time elapses the water 6 University of Ghana http://ugspace.ug.edu.gh reaction decreases sharply (Taylor, 1997). Refer to Table 2.1 for the chemical compositions of cement clinker. Table 2.1 Compositions of phases in Portland cement clinker in mass percentages (Taylor, 1997). Sample Na2O MgO Al2O3 SiO2 P2O5 SO3 K2O CaO TiO2 Alite 0.10 1.10 1.00 25.20 0.10 0.10 0.10 71.6 - Belite 0.10 0.50 2.10 31.50 0.10 0.20 0.90 63.5 0.20 Aluminate (cubic) 1.00 1.40 31.30 3.70 - - 0.70 56.6 0.20 Ferrite 0.10 3.00 21.90 3.60 - - 0.20 47.5 1.60 Aluminate 0.60 1.20 28.90 4.30 - - 4.00 53.9 0.50 (Orthorhombic) Aluminate (low Fe) 0.40 1.00 33.80 4.60 - - 0.50 58.1 0.60 Ferrite (low Al) 0.40 3.70 16.20 5.00 - 0.30 0.20 47.8 0.60 2.3 Pozzolanas Pozzolana is defined as a fine siliceous or alumino – siliceous material and in the presence of moisture, chemically reacts at room temperature with calcium hydroxide, a by – product of hydration of Portland cement, to form compounds possessing cementitious properties (Canadian Standard Association, 2000). Pozzolana is used, as an addition to the cement in the manufacturing process, it can be a supplement for a portion of the cement used in mortar and concrete production. There are two types of pozzolanas that occur naturally or artificially. Ashes and lavas that originate from magmas such as leucitic, alkalitrachytic hauynophric and leucotephritic form natural pozzolana. These ashes are volcanic eruptions that have solidify instantaneously to produce pyroclastic glass (Malquori, 1960). There are other 7 University of Ghana http://ugspace.ug.edu.gh classes of natural pozzolanas widely termed pseudonatural pozzolanas. These are pyroclastic glassy minerals that are present in the molten magma and are formed under auto – metamorphism to produce zeolitization and argillization (Malquori, 1960; Steopoe, 1964). Artificial pozzolanas are those materials with undeveloped property and thus have to undergo some pyro – processing before they exhibit full pozzolanic property (Hammond, 1983). Blast furnace slugs, , siliceous shale, burnt sugar cane stalks, opaline shale, burnt clay, rice husk, Fly ash, spent oil shale and bauxite waste are few examples of artificial pozzolanas (Grane, 1980). It is common practice for scientist and engineers refer to natural as well as industrial co – products with some percentage of glassy silica as pozzolanas. Table 2.2 shows different natural pozzolanas and their chemical compositions. Table 2.2 Chemical compositions of some natural Pozzolanas (Taylor, 1997). Sample Na2O MgO Al2O3 SiO2 K2O SO3 CaO Fe2O3 LOI Sacrofano - - 3.05 89.22 - 2.28 0.77 4.67 - (Italy) Bacoli (Italy) 3.08 1.20 18.20 53.08 0.65 7.61 9.05 4.29 3.05 Segni (Italy) 0.85 4.42 19.59 45.47 6.35 0.16 9.37 9.91 4.03 Santorin Earth 3.80 2.00 13.00 63.80 2.50 - 4.00 5.71 4.80 Rhenish trass 1.48 1.20 18.29 52.12 - 5.06 4.94 5.81 11.10 Rhyolite 4.97 1.23 15.89 65.75 1.92 - 3.35 2.54 3.43 pumice Pozzolana cements can be produced from either natural or artificial pozzolana materials and are normally referred to as clay pozzolana cements. When clay is calcined 8 University of Ghana http://ugspace.ug.edu.gh and dehydrated at temperatures ranging from 700 to 750oC, its crystalline structure is distorted and molecules of water are lost. This makes concretes with calcined clay suck lots of water. The process produces a quasi – amorphous material. In the process, a reaction between silica tetrahedral and lime takes place (Massazza, 1998). The lime is produced from the hydration of C3S and C2S of Portland cement. The percentage composition of calcined clay pozzolanas includes silica (SiO2) in excess of 50%, alumina (Al2O3) and hematite (Fe2O3), which usually exceed 20%. Alumina and hematite together is referred to as total R2O3. There is a widely accepted criterion for selecting a good burnt clay pozzolana. The criterion states that the sum of silica, alumina and hematite contents must exceed 70% (Lea, 1940). 2.4 Corrosion mechanism of steel in reinforced concretes Corrosion of the steel reinforcement takes place in the presence of water and oxygen. This electrochemical process can be described with a redox reaction. The oxidation and reduction of iron and oxygen respectively produces Fe(OH)2 as described in Equation 2.1. Fe(OH)2 is not the only product of the reaction as there can be many products formed. The types of corrosion products formed depends on the surrounding environment such as moisture, pH, chlorides, oxygen, temperature but to mention a few (Marcotte, 2001). 1 Fe + H2O + O2 Fe(OH)2 (2.1) 2 A passive layer of iron oxide is formed on the surface of the steel reinforcement. This layer might not be present if the concrete’s pH is not highly alkaline. This concrete is said to be in a passive state. The concepts of passivity, corrosion and corrosion resistance can well be understood using the Pourbaix equilibrium potential – pH diagram (Pourbaix, 1963). Figure 2.1, shows a Pourbaix diagram of iron and its oxides in 10−6 mol – 1 ion concentration solution. 9 University of Ghana http://ugspace.ug.edu.gh Figure 2.1. Pourbaix diagram for iron in water of temperature of 25 o C (Angst, 2011). The equilibrium potentials below which hydrogen or oxygen reduction occurs is depicted as dashed lines in Figure 2.1. Equation 2.2 describes such a hydrogen reduction process. For the case where water and H+ and OH- ions exist in equilibrium, references can be made in-between/the region between the dashed lines in Figure 2.1. Environments with low pH or oxygen deficient can be described with Equation 2.2. Such phenomena occur in corrosion pits. Iron does not dissolve in solution to form passive oxides. It does not also react with water. This behaviour of iron makes it immune to corrosion at very low potentials. Corrosion may occur when oxygen is very low yet with low potentials and high pH values. This occurrence is depicted in Equation 2.4 (Angst, 2011). 2H+ + 2e− → H2 (2.2) 2H2O + 2e − → H −2 + 2OH (2.3) Fe + 2H2O → HFeO − 2 + 3H + + 2e− (2.4) The passive region under anodic control exhibit rapid decline in current as the 10 University of Ghana http://ugspace.ug.edu.gh passive film develops on the steel reinforcement. Dissolution of iron slowly occurs but the associated corrosion rate is infinitesimal(Bardal, 2004). The chemical complexity of concrete makes its passive oxide layers quite different from those formed in aqueous solutions. Thus, it is expected that the chemical composition and microstructure of the passive oxide differ from steel rods that are embedded in concrete as reinforcement from steel exposed to some aqueous media. Furthermore, composition variations that occur in the concrete over the entire length of the steel reinforcement have significant effect on the nature of the passive oxide layer. Other factors such as variations in oxygen and pH of the pore solution and cement hydration components also affect the composition and microstructure of the passive oxide layer formed (Scully, 1975). Some important parameters such as the corrosion potentials and the pitting potentials are important to understanding the principles of corrosion. From Figure 2.2, the potential that lie between the equilibrium potentials of the cathodic and anodic reactions is known as the corrosion potential, Ecorr and is a function of oxygen and moisture. The pitting potential Epit is that critical potential value above which the corrosion potential must exceed for pitting corrosion to occur. It can be seen from Figure 2.2 that pitting corrosion would only occur in the passive region of the polarization curve. The presence of chloride ions and other aggressive media have an indirect proportionality relation with pitting corrosion (Page & Treadaway, 1982) while a strong direct correlation exist between the pitting corrosion and pH of the pore solution, temperature, composition and microstructure of the reinforced concrete (Bertolini & Redaelli, 2009). However, there is still uncertainty within literature on the true quantifiable relationship between the pitting potential and chloride ion concentration for firm inferences concerning the pitting corrosion. 11 University of Ghana http://ugspace.ug.edu.gh Three main mechanisms are responsible for the breakdown of passivity (Strehblow, 2002). They are 1. Penetration mechanism 2. Film breaking mechanism and 3. Adsorption mechanism Figure 2.2. Polarization curve for steel (MacDougall & Graham, 2002). 2.4.1 The penetration mechanism This mechanism involves the penetration of chloride ions in electrolytic solutions through a passive oxide layer due to high potential differences that exist across the thickness of the passive layer (Silva, 2013). The process however requires high chloride concentration, presence of moisture, oxygen and high electrical conductivity. The ions also require a stimulation time gap for migration and breakdown to take place. Chloride ions can also penetrate through defects into lattices (Arya, Buenfeld, & Newman, 1990; Morozov et al., 2013; Peng, Feng, & Song, 2014). The 12 University of Ghana http://ugspace.ug.edu.gh process of chloride ion penetration through concrete occurs mainly by diffusion, a process that is well documented in literature (Angst, 2011; Arya et al., 1990; Bai, Wild, & Sabir, 2003; Dehghanian & Locke, 1982; Marcotte, 2001; Morozov et al., 2013; Peng et al., 2014). 2.4.2 Film or layer breaking mechanism Sudden changes in corrosion potential often cause cracks on the passive layer. These cracks expose the metal surface to chloride ions and they penetrate through the cracks. The reason for cracks caused by stress in the passive layer arises but not limited to the following; 1. Electrostriction pressure due to high potential difference across the layer 2. Interfacial tension of the layer 3. Local stress caused by impurities 4. Internal stress due to layer – steel volume ratio thereby producing more anions 5. Pressure due to partial hydration or dehydration of layer The type of corrosion layer break mechanism causes could be localized at specific points on the metal surface or corrosion of the entire steel surface. However, localized corrosion is severe since the chloride ions have direct access to the steel (Peng et al., 2014; Román et al., 2014). 2.4.3 Adsorption mechanism This mechanism is caused by ion displacement, which results in thinning of the passive oxide layer. The adsorption of chloride ions to the surface of the passive layer progressively dissolves the steel completely. However, the theory surrounding the adsorption mechanism is not fully understood. The vivid processes for which adsorption breaks passive layer in reinforced concrete are insufficient (MacDougall & Graham, 2002). As such this review would not attempt to discuss the mechanism. 13 University of Ghana http://ugspace.ug.edu.gh Figure 2.3, shows a schematic representation of pitting corrosion caused by chloride ion action. When passivity breaks down, iron dissolves to form a pit as shown in Equation 2.1. Electron transfer from the anode to the cathode then takes place for oxygen reduction to occur. Ionic current (movement) takes place such that the cathode repels and attracts cations. In this case, the anions are OH− and Cl− and the cations are Na+, K+ and Ca2+. Electron neutrality is thus maintained and positive charges that appear at the anode are balanced. Acidification takes place in the pit where hydrolysis of dissolved iron occurs. (Pourbaix, 1974). This acidification process follows Equations 2.5 to 2.7. Figure 2.3. Schematic representation of chloride induced pitting corrosion (MacDougall & Graham, 2002). Fe2+ + 2H2O → Fe(OH) + 2 + 2H (2.5) 2Fe2+ + 3H2O → Fe2O3 + 6H + + 2e− (2.6) 3Fe2+ + 4H O → Fe O + 8H+ + 2e−2 2 3 (2.7) 2.5 Corrosion rate quantification and methods Quantifying corrosion is essential to understanding material failure that is 14 University of Ghana http://ugspace.ug.edu.gh caused by corrosion. The rate at which materials, in this case steel rebar, corrode provides information on confined corrosive conditions. The corrosion rate also helps scientists, engineers and building contractors plan the best counteractive action that would prevent corrosion activity (Hammond, 1983). Measuring the weight loss of materials as a way to determine their corrosion rate still remains a simple yet useful technique. Nonetheless, this measurement method suffers significant setbacks. One of such setback is associated with the longer times one has to endure to produce meaningful results. Electrochemistry and electrochemical methods have proved effective in the quantification of corrosion. Based on a common theory, the electrochemical methods for corrosion rate estimation can be grouped into two main techniques (Abosrra, 2010); 1. Potentiodynamic polarization curve 2. Linear polarization resistance (LRP) measurements. This work would employ the potentiodynamic polarization curve technique. The theoretical basis, which is common to the two methods, seeks to give a good approximation of the corrosion current, Icorr. Some authors prefer to work with a much convenient parameter – corrosion current density, to avoid error accumulations associated with area measurements. The corrosion current, Icorr has an inverse relation with the polarization resistance and is given by Equation 2.8 (Song, 2000). 1 𝐼𝐶𝑜𝑟𝑟 ∝ (2.8) 𝑅𝑝 The polarization resistance is a function of the amplitude of the potential shift. If the potential shift away from the corrosion potential ECorr is denoted by ∆E and the corresponding current shift is ∆I, the polarization resistance can otherwise be defined by Equation 2.9 (Song, 2000). 15 University of Ghana http://ugspace.ug.edu.gh ∆𝐸 𝑅𝑝 = . (2.9) ∆𝐼 The slopes of the tangents obtained from the anodic and cathodic curves of the plot represent the constant of proportionality in Equation 2.8. Denoting this proportionality constant as B, and anodic and cathodic Tafel slopes as βa and βc respectively, an expression for B can be obtained (Song, 2000). (𝛽𝑎 × 𝛽𝑐) 𝐵 = (2.10) 2.3(𝛽𝑎 + 𝛽𝑐) 2.5.1 Potentiodynamic polarization curve In this measurement technique, the potential driving of the anodic and cathodic reactions is varied to obtain the net current change. The Potentiostat measures the applied current for which there is an increase in the force to move charges round. The applied current registers zero at the open circuit potential. This measurement technique provides more information about corrosion activity over wider potential ranges as compared to other techniques (Burstein, 2005). This measurement method accumulates less error thus making its corrosion rate measurements more accurate. Unlike the LRP methods, no assumptions are made for the constant B neither does slopes from the Tafel plots have to be extrapolated (Burstein, 2005). Many authors have applied the potentiodynamic polarization technique. Dehghanian & Locke, 1982 used potentiodynamic polarization technique to determine the corrosion rates of steel reinforcement embedded in concretes prepared from Type I and Type V Portland cement and exposed to chloride ions. Their set up scanned a potential of ±1700 mV for 10 minutes at a rate of 1 mV/s. Their curves did not show clear passive range for the rebar. However, an increase in corrosion rate occurred due to increase in chloride ion concentration. The ion concentration was directly inferred from increases in the corrosion potential and the current density. Similar phenomena 16 University of Ghana http://ugspace.ug.edu.gh occurred when steel rebars placed in concretes prepared with plain and silica fume blended cement (Jarrah, Al-Moudi, Maslehuddin, Ashiru, & Al-Moana, 1995). The time of chloride exposure this time was 2 years. Active corrosion activity occurred in rebars embedded in concretes prepared from plain cement as compared to their silica fume blended counterparts. The differences in electrical resistivity of the samples played important roles in impeding corrosion activity. The potentiodynamic polarization technique has been the preferred technique within literature specifically for measuring corrosion rates of steel reinforcements exposed to chloride environments. The ease with which the corrosion rate and the Tafel slopes are obtained simultaneously with the test results may be one of the numerous reasons for its preference. 2.5.2 LRP technique Due to its ability to stabilize the corrosion system and its simplicity, the linear polarization resistance technique is used widely in field tests and laboratory experiments. Nonetheless, there is a setback to this technique. The assumption of the proportionality constant B introduces error and constituently results in erroneous estimation of the corrosion rate. A B value of 26 mV and 52 mV have been found to be best predictions for active and passive states respectively (González, Molina, Escudero, & Andrade, 1985). Alternatively, Tafel slopes of 120 mV/decade if substituted into Equation 2.10 would give a B value of 26 mV with an error of ±2 in the value of corrosion rate calculated (González, Andrade, Alonso, & Feliu, 1995). On the other hand, if one of the Tafel slopes is infinity and the other maintained at 120 mV/decade, Equation 2.10 would give a B value of 52 mV. Furthermore, B value in the passive state lie between 8 mV and infinity under different conditions (Song, 2000). The LRP technique has produced good results when used to measure the 17 University of Ghana http://ugspace.ug.edu.gh corrosion rate of reinforced concrete prepared with slag cement and exposed to 0 – 5% (W/V) CaCl2 aqueous solution (Gouda, Shater, & Mikhail, 1975). The calculated corrosion rate lied between the values of 0.05 to 0.34 mpy with minimum error. 2.6 Steel reinforcement – concrete bond Many works have enriched the theoretical understanding of the bond that exist between steel reinforcement and the concretes in which they are embedded. However, its experimental and practical application is widely insufficient. The complexity of this bond, which is a force transfer, is one of the reasons for the lack of appreciation of the subject matter. Treatment of the force transfers can be grouped in two categories (Treece & Jirsa, 1989). First is the bearing component on the lugs, which causes splitting of the concrete. Second is the friction component, which consists of true friction and the effect of any secondary chemical and bonding effects. The major influences on bond include (Nawy, 1996); 1. Gripping effect resulting from the drying shrinkage of the surrounding concrete 2. Frictional resistance to sliding and interlock on the reinforcing elements subjected to tensile stress. 3. Effect of concrete quality and strength in tension and compression 4. Adhesion between the concrete and the reinforcing elements. 5. Mechanical anchorage effects of the ends of the bars through the development length, splicing, hooks and crossbars. 6. Diameter, shape and spacing of reinforcement as they affect crack development. Recent literatures tend to give much reference to the effects of bond rather than absolute mechanisms. For instance some authors have developed a model that views bonds as containing both a splitting and non – splitting components (Cairns & Jones, 1995, 1996). The amount of confinement experienced by the rebar varies as the splitting 18 University of Ghana http://ugspace.ug.edu.gh component while on the other hand the non-splitting component is fixed. Concrete confinement or restriction is one important parameter that has significant effect on concrete – steel reinforcement bonds. It has been established that there is a direct proportionality between concrete – steel reinforcement bond strength and concrete confinements around a reinforcing steel bar (Leet, 1996). Stirrups can give reinforced concrete confinement such that the transverse steel reinforcement supplies the confinement. The stress field that exists in the concrete also supplies the necessary confinement in the following way: for a steel beam – column configuration for instance, stress from the column acts perpendicular to the longitudinal steel beam. This configuration supplies stresses that confine the steel beam and increases the concrete – steel reinforcement bond strength. The impossibility of engineers and scientist to ensure the existence of a stress fields makes stress fields less significant in designing structures. However they are well accepted (Rashid, Khatun, Uddin, & Nayeem, 2010). Three test methods have widely been adopted in literature to measure or quantify bond strength and they include (Nawy, 1996); 1. Pull-out tests 2. Embedded bar tests 3. Beam tests. This work would employ the Pull – out test method to determine the bond strength between the steel reinforcement and concrete. All the test types are used to ascertain how well a steel reinforcement transfers load to a surrounding concrete and they have their advantages and disadvantages. These would be outlined in the following section. 2.6.1 Pull – out test A known length of steel is embedded into a concrete sample and pulled out of 19 University of Ghana http://ugspace.ug.edu.gh the restrained concrete. Alternatively, the steel could be pushed out of the concrete. Both methods work well in accordance with Newton’s third law of motion. The steel yields if the pull out or push out action is continued (Nawy, 1996). This method has many advantages. It is simple, and easy to apply for the determination of bond strengths. Simultaneous measurement of slip between the steel reinforcement and concrete is easily achieved. The disadvantage that is associated with the pull – out test method lies in the stress field or force fields that arise. The steel reinforcement is under tensional force while the concrete is under compressional force. As such, concretes cracks at small tensile loads since its tensile strength is very low. The test does not capture these events even though it is a standardized procedure (C234-91a, 1994). It however has the capacity to evaluate different concrete types. For structural design purposes the pull – out test is not preferred. 2.6.2 Embedded bar test This test is almost similar to the pull – out test, just that for this test method, both ends of the steel reinforcement are pulled until the concrete cracks. The bond stresses are then calculated from the widths of the cracks (Nawy, 1996). This method captures stress fields more appropriately and the set up is comparatively simple to organize. However, its shortcoming lie with the difficulties associated with the accurate monitoring of the width of the cracks and the space in-between cracks. Direct stress estimation inferred from data interpretation is also a major challenge. The case of insufficient understanding of what is occurring and how this relates to stress is another shortfall (Reyes-Diaz et al., 2011). 2.6.3 Beam test The test can be performed in many ways while the fundamental aim of modeling 20 University of Ghana http://ugspace.ug.edu.gh a section of a beam with a known length of reinforcing steel embedded inside it. The configuration is forced to bend, which places the steel reinforcement and the concrete under tension. Service conditions are well modeled if the process is performed well which is quite simple to understand and interpret. The extraordinary geometries involved in bending make the beam test measurement technique somehow difficult to implement. Unlike standardized tests employed by all researchers, the literature has supplied various set ups to cater for the unfamiliar geometries (Hamad & Mike, 2005). The service conditions for normal ribbed bars vary, thus the Eurocode supplies a scheme for their use in design works. Such values given by the Eurocode have a safety factor of 1.5 and are a function of concrete strength. The code’s values vary from 1.6, 3.4 to 4.3 MPa for concrete of strengths of 12, 35 to 50 MPa respectively. The steel reinforcement’s diameter, actual stress of the steel reinforcement and casting direction are accounted for by modifying the Eurocode (Abosrra, 2010). Cairns & Jones, (1996) investigated steel – concrete bond strengths for samples with concrete compressive strengths within 30 MPa. Their bond strengths ranged from 3 to 5 MPa for reinforced concrete specimen dimensions of 320 mm × 225 mm or 100 mm × 225 mm. There are many formulas within literature that can be used to estimate reinforcement – concrete bond strength but that of Leet, (1996) is widely used. The formula states 20√f ′c σ = (2.11) ds where σ denote the bond strength in metric units, f ′e denote the compressive strength of the concrete and ds denote the diameter of the steel reinforcement. Equation 2.11 holds for σ ≤ 5.52 MPa and as such, weak concretes with larger reinforcement diameters would have σ values within the range of 1.5 MPa, while strong concretes with smaller 21 University of Ghana http://ugspace.ug.edu.gh reinforcements have σ values around 5.5 MPa. 2.7 Effects of corrosion on concrete structures Deterioration of steel reinforcement in concrete will have effect on the performance of the reinforced structure. The effect of steel corrosion on concrete structures and performance will be discussed in this section. Corrosion effects on steel – concretes bonds are outlined in the next section. Firstly, the loss of the steel reinforcement section is an effect of the corrosion influence on the reinforcement’s properties. From Equation 2.1, the corrosion reactions convert iron to Fe(OH)2 and other product as discussed in section 2.2. The molecules or atoms of these reaction products form a weak brittle material or substance. These substances do not contribute to the entire load distribution of the reinforcement. It has been investigated in many typical analysis, that the effect of the load distribution reduces as the steel reinforcement continues to deteriorate (Ting & Nowak, 1991; Yuan & Marosszeky, 1991). Three years on, this assertion was confirmed by Philips, (1993) in his PhD work. The investigations and confirmation directs obvious conclusion on the corrosion effects on safety. Spalling is the second effect reinforcement corrosion has on structural performance. As the steel reinforcement expand, surrounding concrete is lost. This phenomenon is known as spalling. It would be realized in the next section that spalling leads to loss of bond and loss of the concrete unit as well. It becomes more dangerous when spalling occurs in the compressed area. Spalling occurs at the steel rust regions that control shrinkage and thermal movement and not regions that control primary reinforcement. Unlike the concrete in the tension region, the compression region resists load and that loss of concrete would reduce the capacity of the member in a reinforcement configuration. Due to design safety factors, the foregoing may not be 22 University of Ghana http://ugspace.ug.edu.gh critical. However, intense decrease in reinforced concrete integrity can occur if concrete loss was allowed to continue. Furthermore, the reinforcement configuration will thus be failing in a brittle – type failure (compression), and that is highly detrimental (Bai et al., 2003). 2.8 Effects of corrosion on bond strength Corrosion also affect the steel – concrete bond strength and it has been found that a direct relationship exist between them (Almusallam, Ahmed, Gahtani, & Rauf, 1995; Cabrera, 1996). Their studies show bond strength to increase with corrosion to some critical value corrosion beyond which a decrease in bond strength occurs with rapid corrosion. The early increase in bond strength, they argued, can be linked to the increased roughness of the steel reinforcement surface that occur due to the development of a strong corrosion layer. The loss in bond strength was associated with increased corrosion to severe localized corrosion at the steel ribs, reduction in the load carrying capacity and the lubricating effect of the flaky corrosion substance on the surface of the rebar. As part of their conclusions, they stated that the overall load in flexure is not affected by low level of corrosion but rather reduced at high levels because of reduction that occur in the diameter of the steel reinforcement. Studies have been conducted using pull – out tests on the correlation that exist between rebar corrosion and bond strength by varying concrete cover thickness and compressive strengths (Lee, Noguchi, & Tomosawa, 2002; Rashid et al., 2010). Their studies found the steel – concrete bond strength to decrease for increasing corrosion activity. They also mentioned the influence on the bond strength by concrete cover thickness and compressive strength and that an increase in either may lessen the chances of corrosion activity. As mentioned in earlier sections, confinement has some effects on steel – 23 University of Ghana http://ugspace.ug.edu.gh concrete bond strengths. Unconfined steel reinforcements with no corrosion activity have bond strengths slightly lower than confined ones and for the case of corroded rebars the difference in bond strengths were significant (Fang, Lundgren, Plos, & Gylltoft, 2004, 2006). However, the best means to neutralize steel – concrete bond loss is confinement (Auyeung, Balaguru, & Chung, 2000). Reasonable understandings of the relationships that exist between corrosion of steel reinforcement and the loss of steel – concrete bond have been drawn from the above researches and discussions. Cracks, rust spots or spalling are the external signs that visually confirms the deteriorating effect on structures caused by the corrosion of steel reinforcements. Therefore, the corrosion service life reinforced concrete structures can best be predicted by investigating the corrosion characteristics of steel reinforcements in concrete and the steel – concrete bond strengths from corrosion rates calculations or experiments (Abosrra, 2010). 24 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE METHODS AND EXPERIMENTS 3.1 Introduction This chapter describes the experimental procedure and corrosion test methods designed to study the corrosion behaviour of a steel rod in concrete exposed to corrosion environments used during this work. The materials used for the experimental investigation were high tensile steel Grade Fe-415 with corrugated surface of cross sectional area 74.69 mm2 , 30 % clay pozzolana and ordinary portland cements, concrete aggregates. Steel in concrete was exposed to 3% and 5%(W/V) of NaCl and 3%(W/V) Ca(OCl)2 and subjected to impressed current of 0.4 A at various time periods to study induced corrosion activity on the metal rods. The potentiodynamic polarization technique (Tafel plots) was used to assess the corrosion behaviour of the steel reinforcement. The specimens were also subjected to tension to determine the concrete consolidation as well as the bond strength between the concrete and the reinforcing steel rod using the Controls Compressive Strength and Flexure Testing Machines (CCSFTM) at the Ghana Standard Board Authority. X-ray diffractometry (XRD) was performed on crushed and powered samples of the concrete samples to determine and confirm the mineral components of the concrete mixtures and also to look out for chloride crystals that may have formed due to the corrosion media into the reinforced concrete. Finally a surface morphology analysis was performed on the steel exposed to corrosion using the scanning electron microscope (SEM) to study the extent of corrosion activity. 25 University of Ghana http://ugspace.ug.edu.gh 3.2 Materials and specimen preparation Material constituents used include clay pozzolana cement (CPC), ordinary Portland cement (OPC), sand, 2.36 mm and 4.75 mm size coarse aggregate and deionized water. Table 3.1 shows the mixed proportions of three different concrete types for 100 mm cubes molds used for the test. Table 3.2 shows the mechanical properties of steel used. Three concrete types were prepared. One contained only ordinary Portland cement (named OPC), another contained 30% clay pozzolana as partial replacement of ordinary Portland cement (named PPC) and a another concrete which contained only pozzolana cement. The test specimen which consisted of a 10 mm diameter of high tensile steel was embedded in 100 mm concrete cubes. The rods were washed with acetone before they were embedded into the concrete. The length of steel embedded was 90 mm, which is 9 times the diameter of the steel bar. This was done to ensure that bond slip failure dominates other types of failure such as yielding of the steel reinforcement. The concrete specimens were kept in molds and immersed in deionized water of temperature (21 ± 2)𝑜 C for first 7 days, after which de-molding was done and specimen kept in a curing room of humidity corresponding to (22 ± 2)𝑜 C for the following 21 days before testing. The compressive strengths of the various specimens were measured before exposing them to the corrosion environment. Table 3.4, shows the chemical compositions of oxides present in the ordinary Portland and clay pozzolana cements. By composition the oxides can be grouped into major and minor oxide constituents. Larger amounts of CaO, Al2O3, Fe2O3 and SiO2 were present in the ordinary Portland cement, which have been termed as major oxide constituents of ordinary Portland cement. MgO, Na2O, K2O and SO3, which have lower fractions, are the minor constituents. 26 University of Ghana http://ugspace.ug.edu.gh Table 3.1 Mix proportions of concrete and compressive strengths. Concret De- CPC OPC Sand 4.75 2.36 mm Avera Compressiv e ionised (kg/m3) (kg/m3) (kg/m3 mm Chipping ge e Strength Mix water ) Gravel s Weigh after curing Type (kg/m3) (kg/m3) (kg/m3) t (kg) (MPa) 1 187.00 340.00 - 775.00 605.40 403.60 2.23 7.80 2 187.00 - 340.0 775.00 605.40 403.60 2.27 21.00 3 187.00 102.00 238.0 775.00 605.40 403.60 2.29 20.50 Table 3.2 Mechanical properties of thermo mechanically treated (TMT) high tensile steel obtained from Tema Steel Company. Yield Load Tensile Yield Tensile Young (N) Load (N) Strength Strength Modulus (MPa) (MPa) (GPa) 31133.00 41043.00 416.81 549.48 200.00 Table 3.3 Chemical composition of high tensile steel (Wt.%) obtained from Tema Steel Company. Elements C Si Mn P S Sn Cr Cu Fe Wt.% 0.69 0.32 0.89 0.03 0.03 0.02 0.54 0.25 97.23 27 University of Ghana http://ugspace.ug.edu.gh Table 3.4 chemical compositions (Wt.% of oxides in ordinary Portland and clay pozzolana cements from CSIR - BRII. Ordinary Portland Cement (OPC) Clay Pozzolana Cement (CPC) Oxide Composition Oxide/element Composition (%) (%) CaO 65.00 SiO2+ Al2O3+ Fe2O3 70.00 SiO2 21.50 SiO2, Min 23.05 Al2O3 5.00 Reactive SiO2, Min - Fe2O3 3.50 MgO 2.50 MgO 2.00 S in SiO3 2.50 Na2O 0.50 Na2O 1.90 K2O 0.50 Cl 0.05 SO3 2.00 - - 3.3 Electrolytic solutions and corrosion environments 825 ± 0.001 g of sodium chloride (NaCl) solute was dissolved in 27500 ± 5 mL of deionized water to produced 3%(W/V) of sodium chloride (NaCl) solution. Same procedure was followed with Ca(OCl)2 solute to produce 3% (W/V) of Ca(OCl)2 solution. 1375 ± 0.001 g of sodium chloride (NaCl) solute was dissolved in 27500 ± 5 mL of deionized water to produced 5% (W/V) of sodium chloride (NaCl) solution. These solutions served as electrolytes and different media for corrosion acceleration environments. This accelerated corrosion process is necessary to meet the objectives of the work. 28 University of Ghana http://ugspace.ug.edu.gh Concrete specimens were then placed into the solution and corrosion test measurements were made after 1, 7, 15 and 23 days. Specimens were submerged so as to have the top of the concrete just below the solution surface. Few specimens were placed into large amounts of the solutions to minimize diffusion effects or depletion of the solution concentration as the experiment was set to run for 23 cumulative days. 3.4 Accelerated corrosion set-up and measurements After the first day of exposure to the corrosion environments, the first set of testing and measurements were done on the samples. At start a current was passed at varying potential differences across the reinforced steel rod as they stayed in solution and a Tafel plot was generated from the current – voltage characteristic. The set up was aligned such that the reinforcing steel is the anode (working electrode) while an immersed graphite rod with its associated reference electrode served as the cathode as shown in Figure 3.1. The corrosion cell consists of a graphite rod (counter electrode), a saturated calomel reference electrode (SCE) and a reinforcing steel embedded in concrete specimen (working electrode). The corrosion rate of the reinforcing steel was then obtained from a linear polarization test performed connecting the set – up as shown in Figure 3.1 to the Autolab NOVA – Metrohm Autolab ultramodern machine and its software which generates both the rectified current and records corresponding values. The test was performed using a scanning potential range of −1000 mV ≤ V ≤ 100 mV, with a scanning rate of 10 mV/𝑠. The data were obtained for duration of 2 hours at room temperature. 29 University of Ghana http://ugspace.ug.edu.gh Connecting wire Potentiostat Reference electrode Graphite electrode Reinforcing steel Reinforced concrete Electrolytic solution PVC pipe Wooden block Figure 3.1: Schematic representation of corrosion rate measurement set – up. This test was performed on the specimens for the first day of corrosion environment exposure, then subsequently 7, 15 and 23 days. The linear polarization plot obtained was a pure current – voltage characteristic. The Tafel plot was then drawn from the characteristic plot by simply converting the current axis to logarithm scale. Corrosion current density Icorr (𝜇A/cm 2) was extrapolated from the Tafel plot and the corrosion rates calculated using the following equation (Trethewey and Chamberlain 1995): mils 0.13 𝑊 𝐶orrosion Rate ( ) = 𝐼𝐶𝑜𝑟𝑟 (3.1) yr 𝜌𝐴 where W is the atomic weight of steel (56 for iron), 𝜌 is the density of steel (7.85 gm/cm3) and A is the surface area. 3.5 Strength test measurement 3.5.1 Pull - Out test set – up and measurement procedure Once electrochemical measurements have been made, pull – out test was performed on the specimen. The first test was performed on the specimen after curing and drying before exposure to the corrosion environments. The pull – out test was performed at the Ghana 30 University of Ghana http://ugspace.ug.edu.gh Standards Board Authority and the RILEM Technical Recommendation (RILEM TC., 1994) was followed during the test. The flexure-testing machine was used to carry out the pull – out test. A schematic diagram of the machine is shown in Figure 3.2. The set – up comprised a compressive force which was applied to the top of the reinforcing steel rod, a dial gauge recorded the least slip between the steel rod and the concrete at a constant loading rate of 400 mm/min integrity Compressive Force Steel support Reinforcing steel rod Reinforced concrete PVC pipe Dial gauge Figure 3.2: Pull – Out test system set – up. 3.5.2 Compressive strength test set – up and measurement procedure The reinforced concrete specimens were subjected to compressive force to test for their compressive strength after curing and drying had been done, before exposure to the corrosion environments. This was done to ascertain the compressive strengths of the reinforced concrete specimen after they have undergone curing and drying. This 31 University of Ghana http://ugspace.ug.edu.gh also is very necessary to gain insight into the best mix proportions of aggregate and cement. The test was also performed after the 1, 7, 15 and 23 days of concrete specimen exposure to the corrosion environments. A schematic representation of the compressive strength Test device is shown in Figure 3.3. The concrete specimen, with a small 100mm of the reinforcing steel was simply placed and subjected to a compressive force. The minimum force at which the first crack is detected is then recorded. The device used, from the Ghana Standards Board Authority is such that, the machine itself releases stress (which has a constant loading rate of 100 mm/min) and records once the concrete fails by cracking. The compressive test was carried out using the Controls Compressive Strength machine and Figure 3.3 shows a schematic of the machine. Up and down movement of compression force block Compression force block with dial gauge Reinforced concrete Figure 3.3: Compressive strength test set – up. 32 University of Ghana http://ugspace.ug.edu.gh 3.6 X – Ray diffraction (XRD) In order to ascertain invasion of the chloride aqueous solutions penetration through the porous concrete mixture as it approached the embedded steel rods, an X – ray diffraction was performed on the concrete sample that was around 0.5 centimeters of the rod. This study was necessary to confirm the rate at which the chloride solution reach the reinforcing steel rod and also to understand retarding effect of clay pozzolana constituents in the concrete subjected to corrosive environment. After each electrochemical and strength measurements were completed, the test samples were sun dried to ensure evaporation and crystallization of compounds. The samples in granulated form were further crushed in a crucible to powdered forms. These samples were sent to the Department of Physics, University of Ghana for X – ray diffractometry measurements. The experimental diffractograms were compared with those of models and reference compounds, in order to make preliminary hypothesis about sample structure and composition. Then, a full pattern structural refinement was performed in order to achieve a first quantitative understanding of the crystallographic structure of the samples. The results from the XRD are detailed in Chapter Four. 3.7 Surface morphology of reinforcing steel The scanning electron microscopy (SEM) set – up through its secondary electron imaging technique gives the surface morphology of the steel – reinforced rod. Once the reinforcing steel rod had been taken from the concrete, the steel samples were prepared (polished and cleaned of any concrete materials) and cut into 5.0 mm sizes across their cross-sections. The surface of each 5.0 mm sample was viewed in the SEM to ascertain the extent of corrosion and to study corrosion pits after respective days of corrosion. The 5.0 mm sizes were to allow the samples to fit into the microscope’s 33 University of Ghana http://ugspace.ug.edu.gh sample chamber as well as prevent the objective lens from being crushed by the metallic samples. Secondary electron images of the steel surfaces were obtained. SEM facility could view surfaces with magnification ranging from 20 times to approximately 30,000 times and with spatial resolution of 50 to 100 nm. 34 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR RESULTS AND DISCUSSIONS 4.1 Introduction This chapter reports the results of the corrosion tests conducted on a various steel reinforcement in concretes prepared with clay pozzolana and ordinary portland cement prepared concrete that was exposed to 3% and 5% (W/V) of NaCl and 3%(W/V) Ca(OCl)2 respectively. Curing and hardening times of PPC was longer than OPC. Each of the steel reinforced concrete samples were exposed to the stated environments at different exposure times as described in chapter 3. Various tests were then conducted on the samples. Compressive strength tests and Pull – out tests were performed on the concrete samples to ascertain the concrete integrity in terms of both material strength, and bond strength between the embedded steel reinforcement and concrete. Three different tests were then conducted to each sample to confirm and complement the corrosion behaviour of the steel reinforcement in the different mixes of concrete. At start, an electrochemical test was performed, where direct current is passed through the steel rebars and corresponding Tafel plots (Burstein, 2005) obtained. Corrosion rates were calculated based on Equation 3.1 and corrosion current density data from the Tafel plots. Second, X – ray diffractometry technique was employed to confirm the presence and quantity of chlorides that were responsible for the corrosion activity. Finally the extent of corrosion on the metal surface, the so-called surface morphology, was ascertained using the Scanning Electron Microscopy technique. In this chapter, results from the stated measurements are presented and discussed. It is also important to state, that concrete mixtures that fall under concrete mix type 1 of Table 3.1 couldn’t harden during the curing process. Therefore, subsequent measurements of strength and corrosion characteristics could not be carried out on concrete Type 1. Its purpose as a 35 University of Ghana http://ugspace.ug.edu.gh control was not used. Hence concrete mix Types 2 and 3 only would be mentioned for discussion in subsequent sections in this chapter. 4.2 Pull – out test For the purpose of clarity and uniformity through the discussion, concrete mix Type 2 which is made of ordinary Portland cement and its associated aggregates mix is henceforth labeled OPC and concrete mix Type 3 which is composed of 30% clay pozzolana cement, and 70% ordinary Portland cement as cementitious materials and its associated aggregates is labeled PPC. These concrete mixes have been detailed in Table 3.1. Figure 4.1 – 4.5 shows the pull – out forces recorded for the two concrete mixes that were subjected to different corrosion environments over a period of 1, 7, 15 and 23 days respectively. From Figure 4.1, it was observed that Type 2 showed higher pull-out strength in the NaCl solutions but was lower in the 3% Ca(OCl)2 than the Type 3 for the first day. The same behaviour of pull-out strengths of the rebar was seen in days 7, 15, and 23 (Figures 4.2 – 4.4). The higher pull-out strength of Type 2 might be due to the good setting of ordinary Portland cement compared with pozzolana clay setting. This revealed that ordinary Portland cement had a higher pull-out strength of concrete in NaCl solution but lower in the Ca(OCl)2. It could also be observed that the 3% NaCl had higher pull-out strengths than the 5% NaCl. A compilation of the days (1 – 23) clearly showed that Type 2 in 3% NaCl was highest for all the days (Figure 4.5). 36 University of Ghana http://ugspace.ug.edu.gh 100 90 80 70 60 50 PPC 40 OPC 30 20 10 0 5% NaCl 3% NaCl 3% Ca(Ocl)_2 Corrosion media Figure 4.1: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 1 day. 90 80 70 60 50 40 PPC 30 OPC 20 10 0 5% NaCl 3% NaCl 3% Ca(Ocl)_2 Corrosion media Figure 4.2: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 7 days. 37 Pull - Out force (MN) Pull - Out force (MN) University of Ghana http://ugspace.ug.edu.gh 90 80 70 60 50 40 PPC OPC 30 20 10 0 5% NaCl 3% NaCl 3% Ca(Ocl)_2 Corrosion media Figure 4.3: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 15 days. 80 70 60 50 40 PPC 30 OPC 20 10 0 5% NaCl 3% NaCl 3% Ca(Ocl)_2 Corrosion media Figure 4.4: Pull – out forces for reinforcing rod embedded in type 2 (OPC) and type 3 (PPC) concrete mix for duration of 23 days. 38 Pull - Out force (MN) Pull - Out force (MN) University of Ghana http://ugspace.ug.edu.gh 100 90 80 70 OPC in 3% NaCl 60 PPC in 3% NaCl 50 OPC in 5% NaCl 40 PPC in 5% NaCl 30 OPC in 3% 20 Ca(OCl)_2 10 0 0 5 10 15 20 25 Number of Days Figure 4.5: Pull – out forces for reinforcing rod embedded in Type 2 (OPC) and Type 3 (PPC) concrete mix for 1 – 23 days. The pull – out strength, quantifies bond strength between the reinforcing steel rod and the concrete mixture. From Figures 4.1 – 4.5, it can be seen that there were higher pull – out strengths in the OPC mixes than their PPC counterparts. Pull – outs performed on steel rebars embedded in 3% (W/V) Sodium Chloride (NaCl) tend to show higher bond strengths than those exposed to 5%(W/V) of NCl and 3%(W/V) Ca(OCl)2. Also lower values of pull - out were recorded for samples that were exposed to 3%(W/V) Ca(OCl)2. 4.3 Pull – out responses and the effects of corrosion environments and time In general, the concrete Type 2 (OPC) having only Portland cement exhibits higher pull out strength compared to the PPC counterpart. This was particularly more pronounced in low chloride environment and in the early days of exposure to the corrosive environment. However, in high chloride environments such as 5% NaCl and 39 Pull - out Force (MN) University of Ghana http://ugspace.ug.edu.gh also in the 3% Ca(OCl)2 the PPC showed higher long term pull – out strength than the OPC counterparts. It is yet to be understood why 3% Ca(OCl)2 solution resulted in lower pull – out strengths compared to the other chloride environment since it is expected to produce the least corrosion activity due to its lower chloride concentration. This could be as a result of the hypochlorite slowing down the setting reaction, thereby weakening the bond strength between the steel rod and the concrete. The PPC showed a slower decline in pull – out strengths over time as compared to the OPC. This can be attributed to the fact that the clay pozzolan content in the concrete mix results in a slower development of strength compared to concretes with only ordinary Portland cement. Concrete strength develops by hydration and densification and since hydration is slow in the PPC, chloride diffusion to the steel concrete interface could be high in the early stages. 4.4 Compressive strength test The compressive strength of the concrete describes the integrity of the concrete. Normally inferred from the water-cement ratio, cement strength, quality of concrete material and curing processes, its defining characteristics ascertains whether or not a concrete is prepared well. Figures 4.6 – 4.9 shows the compressive strengths of Type 2 and 3 concrete mixes in 3% (W/V) NaCl, 5% (W/V) NaCl and 3% (W/V) Ca(OCl)2. 40 University of Ghana http://ugspace.ug.edu.gh 30 25 20 15 PPC 10 OPC 5 0 5% NaCl 3% NaCL 3% Ca(Ocl)_2 Corrosion media Figure 4.6: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for duration of 1 days. 30 25 20 15 PPC OPC 10 5 0 5% NaCl 3% NaCL 3% Ca(Ocl)_2 Corrosion media Figure 4.7: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 7 days. 41 Compressive Strenght (MPa) Compressive Strenght (MPa) University of Ghana http://ugspace.ug.edu.gh 25 20 15 PPC 10 OPC 5 0 5% NaCl 3% NaCL 3% Ca(Ocl)_2 Corrosion media Figure 4.8: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 15 days. 25 20 15 PPC 10 OPC 5 0 5% NaCl 3% NaCL 3% Ca(Ocl)_2 Corrosion media Figure 4.9: Compressive strength of type 2 (OPC) and type 3 (PPC) concrete mix for a duration of 23 days. 42 Compressive Strenght (MPa) Compressive Strenght (MPa) University of Ghana http://ugspace.ug.edu.gh 27 25 23 OPC in 3% NaCl 21 PPC in 3% NaCl OPC in 5% NaCl 19 PPC in 5% NaCl 17 OPC in 3% 15 Ca(OCl)_2 PPC in 3% Ca(OCl)_2 13 0 10 20 30 Number of Days Figure 4.10: Compressive strength of Type 2 (OPC) and Type 3 (PPC) concrete for duration of 1 – 23 days. Figure 4.10 shows a collation of all the compressive strength data. The figure shows that, Type 2 (OPC) concrete in 3% (W/V) NaCl recorded higher values of compressive strength. Its compressive strength decreased by just 1.5 MPa over the 23 days of exposure. OPC concrete in 5% (W/V) NaCl initially recorded the highest compressive strength. Its strength reduced by 3 MPa over the 23 days of exposure. OPC concrete in 3% (W/V) Ca(OCl)2 recorded a 5.5 MPa decrease in compressive strength over 23 days of exposure. However Type 3 (PPC) concrete in 3% (W/V) NaCl recorded 3.5 MPa decrease in compressive strength over the 23-day exposure period. A 2.5 MPa decrease in compressive strength occurred for PPC in 3% (W/V) NaCl and 3 MPa decrease in compressive strength in 5% (W/V) NaCl. A 2.5 MPa decrease in compressive strength also occurred for PPC in 3% (W/V) Ca(OCl)2. 43 Compressive Strength (MPa) University of Ghana http://ugspace.ug.edu.gh 4.4.1 Effect of cement mix proportions and types on compressive strength Like Pull – out strength, the cement types and mix have great influence on the compressive strength of the concrete after exposure to the chloride environment. A careful scrutiny of Figures 4.6 to 4.10 shows that, the PPC that contains clay pozzolana has lower values of compressive strength as compared to their OPC counterparts that contain only Portland cement fraction. This could be due to the presence of clay that swells upon taking in water thereby weakening the bond strength of the various constituents in the concrete. However the OPC concrete tend to harden more in the chloride environment by imbibing the water present in the aqueous medium and hardens up over time. The presence of clay pozzolana in concrete makes the concrete more resistant to chloride attack. The pozzolana therefore helps to maintain the compressive strength of the concrete in chloride environment. 4.4.2 Effect of Chloride Concentration on Compressive Strength Figures 4.6 to 4.10 do not show any clear characteristic effect by Chloride levels on the compressive strength of concrete. This phenomenon was also observed by Gurdian, Garcia - Alcocel, Beaza-Brotons, Garces, & Zornoza, (2014). However there is the possibility of the porous concrete allowing the chlorides to fill their pore spaces thereby initially strengthening the concrete and then weakening it later as a result of Friedel’s salt formation (Peng et al., 2014). There is also a general observation that higher chloride environment reduces the compressive strength of the OPC in the long run. For the PPC on the other hand, the chemistry of the chloride environment made very little impact on the compressive strength. 44 University of Ghana http://ugspace.ug.edu.gh 4.5 Electrochemical test 4.5.1 Tafel plots and corrosion current density estimation Figures 4.9 – 4.11 shows Polarization curves (Tafel plots) obtained from electrochemical measurements performed on the reinforcing steel rods, as they stay embedded in the concrete and immersed into the various accelerated corrosion environments. The plots show current values on the ordinate and applied voltages on the abscissa for a day’s corrosion exposures in 3% and 5%(W/V) of NaCl and 3%(W/V) Ca(OCl)2. 0.001 0.0001 0.00001 OPC in 3% NaCl PPC in 3% NaCl 0.000001 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 Applied Voltage (V) Figure 4.11: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for a day. 45 Current (A) University of Ghana http://ugspace.ug.edu.gh 0.001 0.0001 0.00001 OPC in 5% NaCl PPC in 5% NaCl 0.000001 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 Applied Voltage (V) Figure 4.12: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for a days. 0.0001 0.00001 OPC in 3% Ca(OCl)_2 PPC in 3% Ca(OCl)_2 0.000001 -1.2 -0.7 -0.2 Applied Voltage (V) Figure 4.13: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for a day. 46 Current (A) Current (A) University of Ghana http://ugspace.ug.edu.gh The following are polarization curves for reinforcing steel rods embedded in 3% and 5%(W/V) of NCl and 3%(W/V) Ca(OCl)2 for a duration of 7 days. 0.001 0.0001 0.00001 OPC in 3% NaCl PPC in 3% NaCl 0.000001 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 Applied Voltage (V) Figure 4.14: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 7 days. 47 Current (A) University of Ghana http://ugspace.ug.edu.gh 0.001 0.0001 0.00001 OPC in 5% NaCl PPC in 5% NaCl 0.000001 -1.2 -0.7 -0.2 Applied Voltage (V) Figure 4.15: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 7 days. 0.0001 0.00001 OPC in 3% Ca(OCl)_2 PPC in 3% Ca(OCl)_2 0.000001 -1.2 -1 -0.8App-li0e.6d Volt-a0g.4e (V)-0.2 0 0.2 Figure 4.16: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 7 days. 48 Current (A) Current (A) University of Ghana http://ugspace.ug.edu.gh The following plots show the same experiments for 15 days accelerated corrosion exposure time. 0.001 0.0001 0.00001 OPC in 3% NaCl PPC in 3% NaCl 0.000001 -1.2 -0.7 -0.2 Voltage (V) Figure 4.17: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 15 days. 0.001 0.0001 0.00001 OPC in 5% NaCl PPC in 5% NaCl 0.000001 -1 -0.8 -0.6 -0.4 Voltage (V) Figure 4.18: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 15 days. 49 Current(A) Current(A) University of Ghana http://ugspace.ug.edu.gh 0.001 0.0001 0.00001 OPC in 3% Ca(OCl)_2 PPC in 3% Ca(OCl)_2 0.000001 -1.2 -0.7 -0.2 Voltage (V) Figure 4.19: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 15 days. Finally the following plots represent polarization measurements after 23 days accelerated exposure time. 0.001 0.0001 0.00001 OPC in 3% NaCl PPC in 3% NaCl 0.000001 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 Applied Voltage (V) Figure 4.20: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% sodium chloride (NaCl) for 23 days. 50 Curremt (A) Current(A) University of Ghana http://ugspace.ug.edu.gh 0.001 0.0001 0.00001 OPC in 5% NaCl PPC in 5% NaCl 0.000001 -1.2 -0.7 -0.2 Applied Voltage (V) Figure 4.21: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 5% sodium chloride (NaCl) for 23 days. 0.0001 0.00001 OPC in 3% Ca(OCl)_2 PPC in 3% Ca(OCl)_2 0.000001 -1.2 -0.7 -0.2 Applied Voltage (V) Figure 4.22: Polarization curves of reinforcing steel rods type 2 (OPC) and type 3 (PPC) concrete mix in 3% Ca(OCl)2 for 23 days. 51 Curremt (A) Curremt (A) University of Ghana http://ugspace.ug.edu.gh 4.5.2 Corrosion rate calculations Corrosion rate calculations are made after extrapolations are made from Figures 4.6 to 4.14. Corrosion currents 𝐼𝑐𝑜𝑟𝑟 and corrosion potentials 𝐸𝑐𝑜𝑟𝑟 are obtained from intersections of tangents drawn from the anodic and cathodic polarization curves. To ensure accuracy and to limit the extent of human errors a MATLAB program was written (see appendix) to draw the tangents to the polarization curves and the corrosion currents and potentials determined. Tables 4.1 to 4.3 shows the corrosion currents and potentials values obtained from the respective plots. These values have been majorly grouped into the accelerated corrosion environments. Under each environment, type 2 (OPC) and type 3 (PPC) concrete mix have been captioned to display their corresponding accelerated exposure times, days 1 to 28. Corrosion rate calculations have been performed using equation 3.1 on the EXCEL platform. An average surface area of 29.85 cm2 and average cross- sectional area of 0.74 cm2 was used. Atomic weight of 56 and density of 7.88 g/cm3 was also used in the calculation of the corrosion rates. Table 4.1: Corrosion rates of reinforcing steel rods embedded in type 2 (OPC) and type 3 (PPC) concrete mix and in 3% sodium chloride (NaCl) aqueous environment over a period of 23 days. OPC PPC Time I_corr E_corr Corr Rate I_corr E_corr Corr Rate (days) (μA) (mV) (mills/yr) (μA) (mV) (mills/yr) 1 170.00 680.00 7.04 150.00 670.00 6.21 7 170.00 660.00 7.04 70.00 580.00 2.90 15 90.00 580.00 3.73 70.00 600.00 2.90 23 81.00 530.00 3.36 21.00 640.00 0.87 52 University of Ghana http://ugspace.ug.edu.gh Table 4.2: Corrosion rates of reinforcing steel rods embedded in type 2 (OPC) and type 3 (PPC) concrete mix and in 5% sodium chloride (NaCl) aqueous environment over a period of 23 days. OPC PPC Time I_corr E_corr Corr Rate I_corr E_corr Corr Rate (days) (μA) (mV) (mills/yr) (μA) (mV) (mills/yr) 1 200.00 650.00 8.28 65.00 620.00 2.69 7 240.00 510.00 9.94 120.00 500.00 4.97 15 170.00 520.00 7.04 110.00 510.00 4.56 23 95.00 680.00 3.94 90.00 640.00 3.73 Table 4.3: Corrosion rates of reinforcing steel rods embedded in type 2 (OPC) and type 3 (PPC) concrete mix and in 3% Ca(OCl)2 aqueous environment over a period of 23 days. OPC PPC Time I_corr E_corr Corr Rate I_corr E_corr Corr Rate (days) (μA) (mV) (mills/yr) (μA) (mV) (mills/yr) 1 47.00 550.00 1.95 40.00 560.00 1.66 7 100.00 620.00 4.14 40.00 600.00 1.66 15 63.00 470.00 2.61 33.00 690.00 1.37 23 11.00 640.00 0.46 13.00 670.00 0.54 53 University of Ghana http://ugspace.ug.edu.gh 12.00 OPC in 3% NaCl 10.00 PPC in 3% NaCl OPC in 5% NaCl 8.00 PPC in 5% NaCl OPC in 3% Ca(OCl)_2 6.00 PPC in 3% Ca(OCl)_2 4.00 2.00 0.00 0 5 10 15 20 25 Number of Days Figure 4.23: Corrosion rates of steel reinforcement over duration of 23 days. From the above Tables and Figure 4.23 it can be established that the corrosion current densities and rate of corrosion in all the different accelerated corrosion media tend to decrease over time. The rate of decrease is however minimal in steel rods that were embedded in concrete mix PPC. The decrease in the rate of corrosion over time could be attributed to passivation – the formation of a protective layer that acts as a diffusion barrier that slows down the diffusion of oxidizing species reaching the steel surface. In general there is an initial increase in the corrosion rate within the first seven days followed by a decrease in the corrosion rate when a stable coating is formed. This is seen in concretes containing ordinary Portland cement. 4.5.3 Effect of cement mix proportions and types It is quite evident from Tables 4.1 to 4.3 and Figure 4.23 that concrete mix PPC recorded low values of corrosion current density likewise corrosion rates. That for OPC 54 Corrosion rate (mills/yr) University of Ghana http://ugspace.ug.edu.gh had relatively higher values. The presence of clay materials is assumed to be responsible for lower corrosion rates in PPC concrete mix. Its fine particle size is also a contributing factor not only to the chemical resilience to corrosion activity but also lessen the porosity of the entire concrete material, thereby delaying the contact time for which the chlorides in the aqueous media would reach the steel surface. This can also be attributed to packing of pozzolana grains amongst other grains in the concrete (Arya et al., 1990; Guerrero, Hernández, & Goñi, 2000), formation of secondary cementitious materials thereby decreasing permeability (Bai et al., 2003; Luke, 2006) as well as inclusion of the chloride ions binding Al2O3 from pozzolana (Lorenzo et al., 2003). 4.5.4 Effect of chloride concentration Steel rods in concretes that were exposed to 5%(W/V) sodium chloride (NaCl) recorded higher corrosion currents densities and corrosion rates compared to those exposed to 3% (W/V) sodium chloride (NaCl) and 3% Ca(OCl)2. However, samples exposed to 3% Ca(OCl)2 corrosion environment recorded relatively lower corrosion current densities and corrosion rates. The 3% Ca(OCl)2 results in the least amount of chloride ions in solution at 1.49 % (W/V) upon dissolution followed by 3 % NaCl with 1.82 % chloride ions and then 5 % NaCl with chloride ion concentration of 3.03 %. This could explain the corrosion rates results. Additionally, sodium chloride dissociates readily in water and hence breaks its chloride bonds freely and readily as opposed to the calcium hypochlorite (Gaur, et al 1994). It is thus not much surprising to find higher concentrations of sodium chloride (NaCl) render higher corrosion activity as depicted in the corrosion current density readings and corrosion rate calculations. 4.5.5 Effect of acceleration time OPC contains no clay material; hence it is expected to have a higher porosity compared to PPC. The longer the exposure time, the higher the possibility of chloride 55 University of Ghana http://ugspace.ug.edu.gh contact. In all the three chloride solutions, the rate of corrosion inferred from the corrosion current density tend to increase sharply over the first 1 – 7 days. Initially, the chloride actively attacks the rebars until a solid corrosion barrier forms on the metal – concrete surface. Within the periods from 8 – 23 days further chloride attack is less significant as a result of the formation of a surface layer. 4.6 X – ray diffractometry In order to identify the true phase composition of cement and or concrete mix, the samples were crushed and sent for X – ray diffractometry in which the diffraction peak were identified and phase composition determined. Figures 4.24 and 4.25 shows the X – ray diffraction patterns of various minerals as expected to be seen in a Portland cement mix and clay pozzolana concrete mix. Concrete mixes embedded into 5%(W/V) NaCl alone are shown. Samples exposed to 3% (W/V) NaCl and 3% Ca(OCl)2 could not be shown because the clear peaks for chlorides have been masked by other crystalline structures. 3000 3CaO.SiO_2 2500 2000 1500 2CaO.SiO_2 3CaO.Al_2O_3 3CaO.SiO_2 1000 500 NaCl 0 0 20 40 60 80 100 2θ (degrees) 56 Intensity (counts) University of Ghana http://ugspace.ug.edu.gh Figure 4.24: X – ray diffractometry showing some mineralogy type 3 (OPC) concrete mix after exposure to 5% (W/V) Sodium Chloride (NaCl) corrosion environment. In Figure 4.24, Alite (3CaO.SiO2) tends to be abundant followed by Belite (2CaO.SiO2) then Aluminate (3CaO.Al2O3) and the ferrite counterparts (4CaO.Al2O3.Fe2O3). This is expected for a concrete mix with ordinary Portland cement component. The presence of small fractions of NaCl in the sample only indicates its crystallization after the concrete mix was dried after its exposure to the accelerated corrosion environment. 3000 3CaO.SiO_2 2500 2000 2CaO.SiO_2 3CaO.Al_2O_3 3CaO.SiO_2 1500 Al_2O_3 SiO_2 1000 CaO 500 NaCl 0 0 20 40 60 80 100 2θ (degrees) Figure 4.25: X – ray diffractometry showing some mineralogy type 2 (PPC) concrete mix after exposure to 5%(W/V) NaCl corrosion environment. Figure 4.25 shows similar diffraction pattern but with Silicon (SiO2), Aluminium (Al2O3) and Calcium oxides (CaO) present. These oxides are a clear evidence of the presence of clay pozzolana materials in the concrete mixture. This is expected for a concrete mix with ordinary Portland cement component. Comparing 57 Intensiy (counts) University of Ghana http://ugspace.ug.edu.gh fractions in the two plots, NaCl happens to show a relatively lower peak, as the clay pozzolana constituents inhibits the imbibition of Chlorides into the concrete mixture. 4.7 Surface morphology investigation Figures 4.26 – 4.31 show the surface views and Scanning Electron Microscopy (SEM) views at 500-μm views. The images show the extent of corrosion of the reinforcing steel as it is embedded in the concrete samples and exposed to various concentrations of accelerates corrosion environments. The dark spots in the SEM images depict corroded areas or corrosion pits. a b c d a b 500 μm Dark spots c d 58 University of Ghana http://ugspace.ug.edu.gh Figure 4.26: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 3% NaCl. (a) One-day exposure. Dark areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. a b c d a b 500 μm c d 59 University of Ghana http://ugspace.ug.edu.gh Figure 4.27: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 3% NaCl. (a) One-day exposure. Dark areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. right: 23 days exposure. a b c d a b 500 μm c d 60 University of Ghana http://ugspace.ug.edu.gh Figure 4.28: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 5% NaCl. (a) One-day exposure. Rounded particles show areas of corrosion activity and dark areas and spots show corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. a b c d a b 500 μm c d 61 University of Ghana http://ugspace.ug.edu.gh Figure 4.29: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 5% NaCl. (a) One-day exposure. Dark areas show areas of corrosion activity or corrosion pits. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. a b c d a b 500 μm c d 62 University of Ghana http://ugspace.ug.edu.gh Figure 4.30: Surface view and corresponding SEM (500 μm) images of steel embedded in type 2 (OPC) concrete mix immersed in 3% Ca(OCl)2. Dark areas show areas of corrosion activity or corrosion pits. (a) One-day exposure. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. a b c d a b 500 μm c d 63 University of Ghana http://ugspace.ug.edu.gh Figure 4.31: Surface view and corresponding SEM (500 μm) images of steel embedded in type 3 (PPC) concrete mix immersed in 3% Ca(OCl)2. Dark areas show areas of corrosion activity or corrosion pits. (a) one-day exposure. (b) 7 days exposure. (c) 15 days exposure. (d) 23 days exposure. From Figures 4.26 – 4.31 it is quite clear that steel samples that were embedded in concrete type 3 (PPC) mix generally experienced less corrosion activity. This could be attributed to the clay pozzolanic minerals that were present in the concrete mix. Even though it was exposed to accelerate aqueous corrosion media, the pozzolana minerals inhibited chloride activity. Unlike PPC, OPC concrete mix could not provide much corrosion resistance for its embedded reinforcing steel as their corrosion areas and pits deepened with time. Chloride attack was much prevalent in OPC since there was no principal corrosion resistant mineral present. 64 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE CONCLUSIONS AND RECOMMENDATION 5.1 Conclusions This study investigated the effect of clay pozzolana on the corrosion resistance of steel reinforcement in concrete as the concrete samples were exposed to various accelerated corrosion environments namely, 3% and 5%(W/V) NaCl and 3%(W/V) Ca(OCl)2. Three concrete types were prepared. One contained only ordinary Portland cement (named OPC), another contained 30% clay pozzolana as partial replacement of ordinary Portland cement (named PPC) and a another concrete which contained only pozzolana cement (which was discarded). The following conclusions have been drawn based on the experimental results:  Concrete mixtures that contained clay pozzolana cement recorded an average compressive strength of 15 MPa while that which contained only ordinary Portland cement is 22 MPa. While the presence of the clay pozzolana in the concrete reduced its compressive strength, it made the concrete more resistant to chloride attack. The PPC type concrete maintained its compressive strength in the various chloride environments over the same time; the OPC type saw a significant decrease in compressive strength with higher chloride concentration. It is believed that due to their relative smaller particle sizes and shape the clay pozzolana result in higher packing densities within the concrete. This combined with a slow rate of chloride ion conduction resulted in the slow rate of chloride diffusion through the PPC concrete. Also due to the presence of aluminosilicates, which are not easily broken down by chloride ions as the belite and alite phases in the OPC, compressive strengths in PPC remain fairly constant in various chloride environments. 65 University of Ghana http://ugspace.ug.edu.gh  Reaction time in the various chloride environments made significant difference in the measured compressive strength of both concrete types. Longer exposure times resulted in a decrease in the compressive strength of both concretes.  Pull – out forces recorded for steel embedded concrete samples that contain clay Pozzolana cement in accelerated corrosion media recorded average values of 71, 65, 57 and 54 MN over 1, 7, 15 and 23 days respectively. Those embedded in concrete mixtures containing ordinary portland cement as the only cement portion recorded average pull – out forces of 80, 70, 61, and 49 MN over 1, 7, 15 and 23 days respectively. The pull – out strength decreased with the number of days in the corrosion environments for both concrete types. However, the rate of decrease was lower between the embedded steel and the PPC type concrete than that between the steel and the OPC type concrete. This implies that, despite the compressive strengths being lower in clay pozzolana containing concretes, steel rods tend to stay longer without significant corrosion activity, which consequently would have weakened the bond strength between the steel surface and concrete.  Values of corrosion rates obtained from anodic and cathodic polarization curves which constitute the Tafel plots reveal that samples exposed to 5%(W/V) Sodium Chloride (NaCl) had higher corrosion rates followed by those exposed to 3%(W/V) sodium chloride (NaCl) then 3% Ca(OCl)2. The hypochlorite has a lower dissociation constant in water compared to NaCl that is ionic. The hypochlorite has a lower chloride concentration compared to NaCl. Furthermore, the Ca2+ bond is stronger than the Na+ bond to chloride ions. These reasons make excess chlorides available within the NaCl aqueous medium. 66 University of Ghana http://ugspace.ug.edu.gh Therefore, higher chloride ions concentration and longer exposure times increase the rate of corrosion.  Also, PPC recorded lower corrosion rates as compared to concrete OPC. The presence of clay pozzolana materials in concrete can therefore be said to decrease chloride-corrosion attack on the embedded steel.  XRD analysis of PPC and OPC shows higher presence of Alite (3CaO.SiO2). However concrete PPC, contained more aluminosilicates that formed secondary cementitious materials thereby decreasing permeability as observed by Bai et al., (2003) and Luke, (2006). This is good indication of the chemical stability of PPC concrete.  Curing and hardening times of PPC was longer than OPC. This implies the presence of clay pozzolana delays the time for hardening of the concrete since clay materials tend to suck more water into its crystal structure.  XRD also showed low intensity peak for sodium chloride (NaCl) in concrete mix type 3 and relatively higher intensity peak for concrete mix type 2. This confirms the inhibitor role clay pozzolana presence play in restricting the movement of chlorides across its concrete boundaries.  SEM showed shallow to deep corrosion pits on steel rods embedded in OPC concrete exposed to the chloride environment. No such corrosion pits were found on steel rods embedded in PPC. However, some surface corrosion activities like powdered rust were visible. This shows that the absence of clay pozzolana allows chlorides to travel faster through the porous concrete mixture and get to the reinforcing steel for corrosion activity to be centred on specific areas on the steel rod. 67 University of Ghana http://ugspace.ug.edu.gh  From the results obtained from this study, 3% Ca(OCl)2 is the least corrosive environment in the absence of clay pozzolana materials. 5.1 Recommendations  Further studies into the Pozzolana – concrete mix proportions for same corrosion environments should be undertaken to ascertain the strengths and steel reinforcement corrosion behaviour.  Other corrosion environments and concentrations could be explored to confirm the resistance potentials of clay Pozzolana cement in reinforced concrete. 68 University of Ghana http://ugspace.ug.edu.gh REFERENCES Abosrra, L. R. (2010). Corrosion of steel reinforcement in concrete. PhD Thesis, University of Bradford, UK. Ahmad, S. (2003). Reinforcement corrosion in concrete structures, its monitoring and service life prediction. Cement and Concrete Composites, 25, 459–471. Almusallam, A., Ahmed, S., Gahtani, A., & Rauf, A. (1995). Effect of reinforcement corrosion on bond strength. Construction and Building Materials, 10, 123–129. Angst, U. (2011). Chloride induced reinforcement corrosion in concrete - Concept of critical chloride content - methods and mechanisms. PhD Thesis NTNU, Trondheim, Norway. Arya, C., Buenfeld, N. R., & Newman, J. B. (1990). Factors influencing chloride - binding in concrete. Cement and Concrete Research, 20, 291–300. Association, C. S. (2000). Cementitious materials compendium. Canadian Standards Association, Canada. Auyeung, Y., Balaguru, P., & Chung, L. (2000). Bond behavior of corroded reinforcement bars. ACI Materials Journal, 97, 214–220. Bai, J., Wild, S., & Sabir, B. B. (2003). Chloride ingress and strength loss in concrete with pc-pfa-mk binder compositions exposed to synthetic seawater. Cement and Concrete Research, 33, 353–362. Bardal, E. (2004). Corrosion and protection: Engineering Materials and Processes. (1st ed.). London, UK: Springer. Bertolini, L., & Redaelli, E. (2009). Depassivation of steel reinforcement in case of pitting corrosion: detection techniques for laboratory studies. Materials and Corrosion, 60(8), 608–616. Boakye, K. A. (2012). Improvement of setting time and early strength development of 69 University of Ghana http://ugspace.ug.edu.gh pozzolana cement through chemical activation. (Unpublished Master’s Thesis). Kwame Nkrumah University of Science and Technology, Ghana. Bouteiller, V., Cremona, C., Baroghel-Bouny, V., & Maloula, A. (2012). Corrosion initiation of reinforced concretes based on Portland or GGBS cements: Chloride contents and electrochemical characterizations versus time. Cement and Concrete Research, 42, 1456–1467. https://doi.org/10.1016/j.cemconres.2012.07.004 Bouzoubaâ, N., Zhang, M. H., Malhotra, V. M., & Golden, D. M. (1999). Blended fly ash cement – A review. ACI Materials Journal, 96(6), 641–650. Burstein, G. T. (2005). A century of Tafel’s equation: 1905 – 2005 a commemorative issue of corrosion science. Corrosion Science, 47(12), 2858–2870. https://doi.org/10.1016/j.corsci.2005.07.002. C234-91a, A. (1994). Standard test method for comparing concretes on the basis of bond developed with reinforcing steel. Annual Book of ASTM Standards (Volume 04.). Philadelphia: ASTM. Cabrera, J. G. (1996). Deterioration of concrete due to reinforcement steel corrosion. Cement and Concrete Composites, 18, 47–59. Cairns, J., & Jones, K. (1995). The splitting forces generated by bond. Magazine of Concrete Research, 47(171), 153–165. Cairns, J., & Jones, K. (1996). An evaluation of the bond – splitting action of ribbed bars. AC1 Materials Journal, 93(1), 10–19. Dehghanian, C., & Locke, C. E. (1982). Electrochemical behaviour of steel in concrete as a result of chloride diffusion into concrete: Part 2. Corrosion, 38, 494–499. Fang, C., Lundgren, K., Plos, M., & Gylltoft, K. (2004). Bond behaviour of corroded reinforcing steel bars in concrete. Cement and Concrete Research, 34, 2159 – 2167. 70 University of Ghana http://ugspace.ug.edu.gh Fang, C., Lundgren, K., Plos, M., & Gylltoft, K. (2006). No Title. Cement and Concrete Research, 36, 1931–1938. González, J. A., Andrade, C., Alonso, C., & Feliu, S. (1995). Comparison of rates of general corrosion and maximum pitting penetration on concrete embedded steel reinforcement. Cement and Concrete Research, 25, 257–264. González, J. A., Molina, A., Escudero, M. L., & Andrade, C. (1985). Errors in the electrochemical evaluation of very small corrosion rates. Part I. Corrosion Science, 25(10), 917–930. Gouda, V. K., Shater, M. A., & Mikhail, R. S. (1975). Effect of slag cement on corrosion behaviour of steel in concrete. Cement and Concrete Research, 5, 1–13. Grane, P. O. (1980). Noncement - based hydraulic binder. Monographs on Appropriate Industrial Technology, 12, 48–57. Guerrero, A., Hernández, M. S., & Goñi, S. (2000). Effect of simulated radioactive liquid waste on the microstructure of cementious materials: portlandite orientation and saturation factors in the pore solution. Journal of American Ceramic Society, 83, 2803–2808. Gurdian, H., Garcia - Alcocel, E., Beaza-Brotons, F., Garces, P., & Zornoza, E. (2014). Corrosion behaviour of steel reinforcement in concrete with recycled aggregates, fly ash and spent cracking catalyst. Materials, 7(4), 3176–3197. https://doi.org/doi: 10.3390/ma7043176 Hamad, B. S., & Mike, J. A. (2005). Bond strength of hot-dip galvanized reinforcement in normal strength concrete structures. Construction and Building Materials, 19, 275–283. Hammond, A. A. (1983). Pozzolana cements for low cost housing. In Proceedings Of A Symposium On Appropriate Building Materials For Low Cost Housing (pp. 73– 71 University of Ghana http://ugspace.ug.edu.gh 83). Nairobi. Jarrah, N. R., Al-Moudi, O. S., Maslehuddin, M., Ashiru, O. A., & Al-Moana, A. I. (1995). Electrochemical behaviour of steel in plain and blended cement concrete in sulphate and/or chloride environments. Construction and Building Materials, 9, 97–103. Lea, F. M. (1940). Investigations on pozzolanas – pozzolanas and pozzolana – lime mixes. Garston, Herts. Lee, H. S., Noguchi, T., & Tomosawa, F. (2002). Evaluation of the bond properties between concrete and reinforcement as a function of the degree of reinforcement corrosion. Cement and Concrete Research, 32, 1313–1318. Leet, K. (1996). Reinforced concrete design (Second). Toronto: McGraw-Hill Inc. Lorenzo, M., Goni, S., & Guerrero, A. (2003). Activation of pozzolanic reaction of hydrated portland cement fly ash pastes in sulfate solution. Institute of Construction Science. Eduardo Torroja. Luke, K. (2006). Microstructure of early age hydration of natural zeolite with pure cement pastes. In 28th Conference on Cement Microscopy (pp. 10–21). Denver, Colorado. MacDougall, B., & Graham, M. J. (2002). Growth and stability of passive films in corrosion mechanisms in theory and practice. (Second). New York, USA: Marcel Dekker Inc. Malquori, G. (1960). Portland pozzolana cement, chemistry of cement. In Fourth International Symposium (pp. 983–1005). Washington. Marcotte, T. D. (2001). Characterization of chloride-induced corrosion products that form in steel-reinforced cementitious materials. PhD Thesis, University of Waterloo, Waterloo, Canada. 72 University of Ghana http://ugspace.ug.edu.gh Massazza, F. (1998). Pozzolana and Pozzolana Cements. In P. C. Hewlett (Ed.), Lea‘s Chemistry of Cement and Concrete (pp. 471–602). London: Elsevier. Morozov, Y., Castela, A. S., Dias, A. P. S., & Montemor, M. F. (2013). Chloride – induced corrosion behaviour of reinforcing steel in spent fluid cracking catalyst modified mortars. Cement Concrete Research, 47(4), 1–7. Nawy, E. G. (1996). Reinforced concrete: a fundamental approach. (3rd ed.). Toronto: Prentice – Hall Canada Inc. Neville, A. M. (1996). Properties of concrete. (4th ed.). London: ELBS. Page, C. L., & Treadaway, K. W. J. (1982). Aspects of the electrochemistry of steel in concrete. Nature, 297, 109–115. Pedeferri, P. (1996). Cathodic protection and cathodic prevention. Construction and Building Materials, 10(5), 391–402. https://doi.org/10.1016/0950- 0618(95)00017-8 Peng, G., Feng, N., & Song, Q. (2014). Influence of chloride – ion adsorption agent on chloride ions in concrete and mortar. Journal of Materials, 7(5), 3415–3426. https://doi.org/10.3390/ma7053415 Philips, J. (1993). The effect of corrosion on the structural performance of new and repaired one - way slabs. PhD Thesis, University of Toronto. Pourbaix, M. (1974). Applications of electrochemistry in corrosion science and practice. Corrosion Science, 14(1), 25–82. Pourbaix, M. J. N. (1963). Atlas d’equilibres electrochimiques. Paris: Gauthier-Villars. Rashid, M. H., Khatun, S., Uddin, S. M. K., & Nayeem, M. A. (2010). Effect of strength and covering on concrete corrosion. European Journal of Scientific Research, 40, 492–499. Reyes-Diaz, E. P., Maldonado, B. E., Almeray, C. F., Bastidas, D. M., M., B. Z., J., C. 73 University of Ghana http://ugspace.ug.edu.gh N., … Tiburcio, G. C. T. (2011). Corrosion behavior of steel embedded in ternary concrete mixtures. International Journal of Electrochemical Science, 6, 1892– 1905. Román, J., Vera, R., Bagnara, M., M., C. A., & Aperador, W. (2014). Effect of chloride ions on the corrosion of galvanized steel embedded in concrete prepared with cements of different composition. International Journal of Electrochemical Science, 9, 580–592. Scully, J. C. (1975). The fundamentals of corrosion. (2nd ed.). U.K.: International Series on Material Science and Technology. Silva, N. (2013). Chloride induced corrosion of reinforced steel in concrete. Chalmers PhD Thesis, University of Technology, Gothenburg, Sweden. Song, G. (2000). Theoretical analysis of the measurement of polarization resistance in reinforced concrete. Cement and Concrete Composites, 22, 407–415. Steopoe, A. (1964). The Romanian trasses, their characteristics and technical uses. In Proceedings of symposium on Pozzolanas, their survey, manufacture and utilization (pp. 95–101). New Delhi: CRRI Publication. Strehblow, H. H. (2002). Mechanisms of pitting corrosion and corrosion mechanisms in theory and practice. (M. P., Ed.) (2nd ed.). New York: Marcel Dekker Inc. Tan, Z. Q., & Hansson, C. M. (2008). Effect of surface condition on the initial corrosion of galvanized reinforcing steel embedded in concrete. Corrosion Science, 59(9), 2512–2522. Taylor, W. (1997). Cement chemistry (2nd ed.). London: Thomas Telford Publishing. TC., R. (1994). RILEM recommendation for the testing and use of construction materials. Paris: E & FN SPON. Ting, S., & Nowak, A. (1991). Effect of reinforcing steel area loss on flexural behaviour 74 University of Ghana http://ugspace.ug.edu.gh of reinforced concrete beams. ACI Structural Journal, 88(3), 309–314. Treece, R. A., & Jirsa, J. O. (1989). Bond strength of epoxy coated reinforcing bars. ACI Materials Journal, 86(2), 167–174. Yeomans, S. (1998). Corrosion of the zinc alloy coating in galvanized reinforced concrete. Corrosion, 98, 653. Yuan, Y., & Marosszeky, M. (1991). Analysis of corroded reinforced concrete sections for repair. Journal of Structural Engineering, 117(7), 2018–2034. 75 University of Ghana http://ugspace.ug.edu.gh APPENDIX MATLAB Codes for tangents drawn from polarization curves function [x y tan] = GetPointListForDisplay(m_Length, r1, r2, count) global nn ca ce length; GetLength = m_Length; length = GetLength; ca = GetCurvatureAtDeltaLength(0.0); ce = GetCurvatureAtDeltaLength(length); radius = 1.0 ./ max(abs(ca), abs(ce)); nn = 3 + (180.0 * length/(2*pi*radius)); % Using modified formula of arc here lengthStep = length/nn; currLen = -lengthStep; while (1) currLen = currLen + lengthStep; if (currLen > m_Length) currLen = m_Length; end 76