corrosion protection of grinding mills in...

CORROSION PROTECTION OF GRINDING MILLS IN THE PHOSPHATE INDUSTRY USING IMPRESSED CURRENT TECHNOLOGY

FINAL REPORT

D. Tao Department of Mining Engineering UNIVERSITY OF KENTUCKY

234E Mining and Mineral Resources Building Lexington, KY 40506 USA

B.K. Parekh Center for Applied Energy Research

UNIVERSITY OF KENTUCKY 2540 Research Park Dr.

Lexington, KY 40514 USA

Prepared for

FLORIDA INSTITUTE OF PHOSPHATE RESEARCH 1855 West Main Street

Bartow, FL 33830 USA

Project Manager: G. Michael Lloyd, Jr. FIPR Project Number: 00-01-170

June 2004

DISCLAIMER

The contents of this report are reproduced herein as received from the contractor. The report may have been edited as to format in conformance with the FIPR Style Manual. The opinions, findings and conclusions expressed herein are not necessarily those of the Florida Institute of Phosphate Research, nor does mention of company names or products constitute endorsement by the Florida Institute of Phosphate Research. © 2004, Florida Institute of Phosphate Research.

iii

PERSPECTIVE

Ball mill grinding of phosphate rock to prepare the phosphate rock for use in the manufacture of phosphoric acid by reacting the phosphate rock with sulfuric acid is essential when central Florida phosphate rock is used as the raw material for phosphoric acid production. In general, grinding costs have also increased with the use of the lower-grade, higher silica-containing rock mined south of the original Bone Valley deposits. The practice of using acidic process water for make-up water for wet grinding has also contributed to the more rapid wear of the grinding media. With the recent pressures on profits, lower-cost grinding of phosphate rock has become even more attractive than it has been in the past. This study looks at a technique that has been employed for corrosion control in widely varying industrial situations and adapts it to address the phosphate industry’s grinding costs. G. Michael Lloyd, Jr. Research Director, Chemical Processing

v

ABSTRACT

Grinding mills are commonly used in the Florida phosphate industry to reduce particle size. The corrosion of metallic grinding media and mill liner is a very serious problem, particularly in acidic conditions as encountered in the Florida phosphate fertilizer industry. Approximately 50% of the total wear of grinding mills can be attributed to metal corrosion. It is known that the corrosive wear of grinding mill leads to an increase in operating cost and plant downtime, a loss of process efficiency, and product contamination. An effective protection of grinding mills from corrosion will improve the process performance and economics, enhance product quality, and increase the lifetime of mills and grinding media.

The proposed project is aimed at developing a practical and effective technique

for minimizing corrosive wear of the ball mill and its grinding media. Metal corrosion reactions are electrochemical in nature and their reaction rates are controlled by the electrochemical potential at the surface. A reduction in potential with excess electrons will depress the anodic dissolution reaction of metal (e.g., M = M2+ + 2e-). This can be accomplished by supplying an impressed current to the object to create a negative potential change called cathodic polarization which reduces the rate of metal reaction. The impressed current was supplied by an EG&G PARC, Model 273 potentiostat for a specially designed ball mill. The results from this study more closely resemble those to be expected from industrial ball mills.

Cathodic protection using impressed current was employed to reduce mill wall

wear during ball milling. Wear rates were determined from weight-loss measurements made on three 3/8-inch-diameter mild carbon steel and high chromium alloy coupons that were flush with the interior surface of the mill wall. Experimental results indicate that the corrosive wear rate was reduced by 92.9% to 94.5% and total wear was reduced by 47.8% to 49.6% for 1018 carbon steel when a potential of -1.0 V was applied. Similar results were obtained with the high chromium alloy, except that the required polarization potential was -0.7 V.

The polarization diagram gives a fundamental quantitative assessment of the

decrease in corrosion rate caused by cathodic polarization. The effects of solution pH, electrode material, gaseous environment, and solution composition on polarization curves have been performed. Polarization diagrams indicate that the current density is higher in nitrogenated solution than in oxygenated solution at the same pH value. The polarization diagrams also suggest that the current density of 1018 carbon steel was higher than that of high-chromium alloy under the same condition. Experimental results also indicate that the current density is higher in buffer solution than that in pond water solution under the same operating conditions.

A statistical Box-Behnken Design (BBD) of experiments was performed to

evaluate effects of individual variables and their interactions on wear rate of grinding ball mill used in phosphate industry. The variables examined in this study included grinding

vi

time, solution pH, rotation speed, mill crop load, and solids percentage. The most significant variables and optimum conditions were identified from statistical analysis of the experimental results using response surface methodology. Solution pH had the most significant effect on the wear rate for both 1018 carbon steel and high chromium alloy. The optimum process parameters for minimum wear rate were solution pH at 7.36, rotation speed at 70.31 RPM, solid percentage at 75.50, and crop load at 71.94% for 1018 carbon steel; solution pH at 8.69, rotation speed at 61.13 RPM, solid percentage at 64.86, and crop load at 57.63% for high-chromium alloy.

To understand the cathodic protection process, scanning electron microscopy

(SEM), energy dispersive spectrometer (EDS), and X-ray diffraction (XRD) methods were used to investigate the corrosion products, surface morphology, and composition when grinding phosphate rock with and without cathodic protection in different solutions. The main corrosion type and corrosion products for 1018 carbon steel and high chromium alloy under different conditions have been determined.

vii

ACKNOWLEDGEMENTS

The financial support for the program is provided by the Florida Institute of Phosphate Research (FIPR) Grant Number 00-01-170. The project manager, Mr. Michael Lloyd, provided valuable advice and support, which is greatly appreciated. Special thanks are given to IMC Phosphate, CF Industries, Inc., and Magotteaux, Inc. for supplying phosphate and grinding media specimens.

ix

TABLE OF CONTENTS PERSPECTIVE................................................................................................................... iii ABSTRACT .......................................................................................................................v ACKNOWLEDGEMENTS.................................................................................................vi EXECUTIVE SUMMARY ..................................................................................................1 INTRODUCTION ................................................................................................................5 LITERATURE REVIEW .....................................................................................................7 Importance of Corrosion Studies ..............................................................................7 Type of Corrosion Damage.......................................................................................8 Factors Affecting Corrosion ...................................................................................10 Techniques Used to Protect Against Corrosion ......................................................12 SCIENTIFIC DISCUSSION ..............................................................................................17 Fundamentals of Electrochemical Corrosion..........................................................17 Cathodic Protection Principles................................................................................21 METHODOLOGY .............................................................................................................23

Construction of a Specially Designed Ball Mill .....................................................23 Specially Designed Ball Mill .........................................................................23 Coupons and Dimensionally Stable Anode ...................................................25 Phosphate Acquisition and Characterization ..........................................................27 Determination of Polarization Diagram for Fundamental Assessment ..................30 Polarization Diagram .....................................................................................30 Polarization Cell and Electrodes....................................................................31 Polarization Study..........................................................................................34 Characterization of Corrosion Rate under Various Operating Conditions .............61 Parametric Study of Corrosion Rate ..............................................................63 Effect of Grinding Time..........................................................................63

x

TABLE OF CONTENTS (CONT.)

Effect of Solution pH..............................................................................63 Effect of Rotation Speed.........................................................................64 Effect of Mill Crop Load ........................................................................65 Effect of Feed Solids Percentage ............................................................65 Statistical Box-Behnken Design of Experiments...........................................70

Interactions of Individual Variables........................................................................71

Determination of Optimal Conditions ...........................................................71 Effect of Polarization Potential on Corrosion Rate.................................................90 1018 Carbon Steel Corrosion Rate ................................................................90 High-Chromium Alloy Corrosion Rate .........................................................91 Determination of Corrosion Current Density .........................................................98 1018 Carbon Steel..........................................................................................98 High-Chromium Alloy...................................................................................99 Determination of Required Current Density for Effective Protection ..................106 1018 Carbon Steel........................................................................................106 High-Chromium Alloy.................................................................................107 Effect of Slurry Conductivity on Wear Rate.........................................................115 Effect of Sodium Sulfate on Wear Rate without Cathodic Protection.........115 Effect of Sodium Sulfate on Wear Rate with Cathodic Protection..............115 Slurry Conductivity Measurement...............................................................125 Eh-pH Diagram.....................................................................................................127 Corrosion Mechanism Investigation .....................................................................131 Scanning Electron Microscopy (SEM) ........................................................132 SEM Analysis for 1018 Carbon Steel...................................................132 SEM Analysis for High-Chromium Alloy............................................141 Energy Dispersive X-Ray Spectrometer (EDS)...........................................148

xi

TABLE OF CONTENTS (CONT.)

EDS Analysis for 1018 Carbon Steel....................................................149 EDS Analysis for High-Chromium Alloy.............................................157 X-Ray Diffraction (XRD) ............................................................................164 XRD Analysis for 1018 Carbon Steel...................................................165 XRD Analysis for High-Chromium Alloy............................................173

ECONOMIC EVALUATION ..........................................................................................181

CONCLUSIONS...............................................................................................................187

REFERENCES .................................................................................................................189

xiii

LIST OF FIGURES

Figure Page

1. Forms of Corrosion: (a) Crevice Corrosion, (b) Stress Corrosion, and (c) Pitting Corrosion ................................................................................... 9

2. Schematic of Simultaneous Metal Dissolution and Hydrogen Evolution ............... 18 3. Anodic and Cathodic Half-cell Reactions for Iron in Acid Solution ...................... 19 4. Cathodic Polarization Protection of Steel by an Impressed Current in an

Acid Solution................................................................................................... 22 5. Schematic of the Specially Designed Mill for Grinding Tests ................................ 24 6. Photograph of the Specially Designed Mill for Grinding Tests .............................. 24 7. Photograph of the Mill on Rollers ........................................................................... 25 8. Photograph of the EG&G PARC, Model 273 Potentiostat ..................................... 25 9. Two Different Designs for Coupon Assembly for (a) Mild Steel and

(b) High Chromium Alloy............................................................................... 26 10. Photograph of the Specially Designed Mill with Coupons and Electrodes........... 27 11. Cumulative Weight Percent vs. Particle Size for Phosphate Sample .................... 29 12. Photograph (a) and Illustration (b) of Polarization Cell ........................................ 32 13. Photograph of the Working Electrodes ................................................................. 33 14. Photograph of the Electrochemical System........................................................... 33 15. Polarization Curves for 1080 Steel in 1 N H2SO4 ................................................. 36 16. Effect of Solution pH on Cathodic Polarization of Iron........................................ 37 17. Potentiostatic Anodic Polarization Curve of Iron, 18Cr SS, and 18Cr-8Ni SS

in 1 N H2SO4 ................................................................................................... 38 18. Potentiodynamic Anodic Polarization Curve of 1018 Steel in

Nitrogenated Buffer Solutions......................................................................... 41

xiv

LIST OF FIGURES (CONT.)

Figure Page

19. Potentiodynamic Anodic Polarization Curve of 1018 Steel in Oxygenated Buffer Solutions .......................................................................... 42

20. Potentiodynamic Anodic Polarization Curve of High Chromium Alloy in

Nitrogenated Buffer Solutions......................................................................... 43 21. Potentiodynamic Anodic Polarization Curve of High Chromium Alloy in

Oxygenated Buffer Solutions .......................................................................... 44 22. Potentiodynamic Anodic Polarization Curve of 1018 Steel in

Nitrogenated Pond Water Solutions ................................................................ 47 23. Potentiodynamic Anodic Polarization Curve of 1018 Steel in

Oxygenated Pond Water Solutions.................................................................. 48 24. Potentiodynamic Anodic Polarization Curve of High Chromium Alloy in

Nitrogenated Pond Water Solutions ................................................................ 49 25. Potentiodynamic Anodic Polarization Curve of High Chromium Alloy in

Oxygenated Pond Water Solutions.................................................................. 50 26. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in

Oxygenated Buffer Solutions .......................................................................... 52 27. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in

Nitrogenated Buffer Solutions......................................................................... 53 28. Potentiodynamic Cathodic Polarization Curve of High Chromium Alloy in

Oxygenated Buffer Solutions .......................................................................... 54 29. Potentiodynamic Cathodic Polarization Curve of High Chromium Alloy in

Nitrogenated Buffer Solutions......................................................................... 55 30. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in

Oxygenated Pond Water Solutions.................................................................. 58 31. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in

Nitrogenated Pond Water Solutions ................................................................ 59 32. Potentiodynamic Cathodic Polarization Curve of High Chromium Alloy in

Oxygenated Pond Water Solutions.................................................................. 60

xv


Figure Page

33. Potentiodynamic Cathodic Polarization Curve of High Chromium Alloy in Nitrogenated Pond Water Solutions ................................................................ 61

34. Effect of Grinding Time on Weight Loss .............................................................. 66 35. Effect of Solution pH on Wear Rate...................................................................... 67 36. Effect of Rotation Speed on Wear Rate................................................................. 68 37. Effect of Mill Crop Load on Wear Rate ................................................................ 69 38. Effect of Feed Solids Percentage on Wear Rate.................................................... 70 39. Response for pH and Rotation Speed at Crop Load 50% and

Solid Percentage 64%...................................................................................... 76 40. Response for pH and Crop Load at Rotation Speed 70 RPM and

Solid Percentage 64%...................................................................................... 77 41. Response for pH and Solid Percentage at Rotation Speed 70 RPM and

Crop Load 50% ............................................................................................... 78 42. Response for Rotation Speed and Crop Load at pH 7.0 and

Solid Percentage 64%...................................................................................... 79 43. Response for Rotation Speed and Solid Percentage at pH 7.0 and

Crop Load 50% ............................................................................................... 80 44. Response for Crop Load and Solid Percentage at pH 7.0 and

Rotation Speed 70 RPM .................................................................................. 81 45. Normal Probability Plot of Residual for 1018 Carbon Steel ................................. 82 46. Response Surface and Contours for Desirability Function for Carbon Steel ........ 83 47. Normal Probability Plot of Residual for High-Chromium Alloy .......................... 88 48. Response Surface and Contours for Desirability Function for High-Chromium

Alloy................................................................................................................ 89

xvi


Figure Page

49. Effect of Polarization Potential on Corrosion Rate of Carbon Steel at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ........... 92

50. Effect of Polarization Potential on Corrosion Rate of Carbon Steel at pH 6.8,

Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ........... 93 51. Effect of Polarization Potential on Corrosion Rate of Carbon Steel at pH 9.2,

Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ........... 94 52. Effect of Polarization Potential on Corrosion Rate of High-Chromium Alloy

at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percent- age 64% ........................................................................................................... 95

53. Effect of Polarization Potential on Corrosion Rate of High-Chromium Alloy


54. Effect of Polarization Potential on Corrosion Rate of High-Chromium Alloy


55. Effect of Polarization Potential on Corrosion Current Density of Carbon

Steel at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ............................................................................................. 100





58. Effect of Polarization Potential on Corrosion Current Density of High-

Chromium Alloy at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%............................................................................. 103

59. Effect of Polarization Potential on Corrosion Current Density of High-

Chromium Alloy at pH 6.8, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%............................................................................. 104

xvii


Figure Page

60. Effect of Polarization Potential on Corrosion Current Density of High- Chromium Alloy at pH 9.2, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%............................................................................. 105

61. Effect of Impressed Current Density on Corrosion Rate of Carbon Steel at

pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ............................................................................................. 109





64. Effect of Impressed Current Density on Corrosion Rate of High-Chromium

Alloy at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64% ............................................................................................. 112





67. Effect of Sodium Sulfate Concentration on Wear Rate of 1018 Carbon Steel

Without Cathodic Protection ......................................................................... 117 68. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of 1018

Carbon Steel at pH 3.1 When -1.0 V Was Applied....................................... 118 69. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of 1018

Carbon Steel at pH 6.8 When -1.0 V Was Applied....................................... 120 70. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of 1018

Carbon Steel at pH 9.2 When -1.0 V Was Applied....................................... 122

xviii


Figure Page

71. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of High-Chromium Alloy at pH 3.1 When -0.7 V Was Applied....................... 124

72. Photograph of Conductivity Meter ...................................................................... 126 73. Effect of Sodium Sulfate Concentration on Phosphate Slurry Conductivity ...... 127 74. Schematic Example of Eh-pH Diagram for Iron, All Dissolved Species at

Activities of 10-6 M........................................................................................ 129 75. Eh-pH Diagram for Fe-O-H System, Assuming Fe(OH)3 as Stable Fe3+ Phase

and Activity of Dissolved Fe = 10-6 M.......................................................... 130 76. Eh-pH Diagram for Fe-Cr-H2O System, Assuming Total Concentration of

10-2 M and 5×10-3 M Cr ................................................................................. 131 77. Schematic Illustration of SEM ............................................................................ 134 78. Photograph of SEM ............................................................................................. 135 79. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

Without Cathodic Protection ......................................................................... 136 80. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

With Polarization Potential -1.0 V ................................................................ 137 81. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 6.8


With Polarization Potential -1.0 V ................................................................ 139 83. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2


With Polarization Potential -1.0 V ................................................................ 141 85. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 Without Cathodic Protection ............................................................. 143

xix


Figure Page

86. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 3.1 With Polarization Potential -0.7 V .................................................... 144

87. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 6.8 Without Cathodic Protection ............................................................. 145

88. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 6.8 With Polarization Potential -0.7 V .................................................... 146

89. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 9.2 Without Cathodic Protection ............................................................. 147 90. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 9.2 With Polarization Potential -0.7 V .................................................... 148 91. Photograph of EDS.............................................................................................. 151

92. EDS Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

Without Cathodic Protection ......................................................................... 152 93. EDS Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

With Polarization Potential -1.0 V ................................................................ 153 94. EDS Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 6.8


With Polarization Potential -1.0 V ................................................................ 155 96. EDS Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2


With Polarization Potential -1.0 V ................................................................ 157 98. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 Without Cathodic Protection ............................................................. 159 99. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 With Polarization Potential -0.7 V .................................................... 160

xx


Figure Page

100. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 6.8 Without Cathodic Protection ............................................................. 161

101. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 6.8 With Polarization Potential -0.7 V .................................................... 162 102. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 9.2 Without Cathodic Protection ............................................................. 163 103. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 9.2 With Polarization Potential -0.7 V .................................................... 164 104. Photograph of XRD ........................................................................................... 167 105. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

Without Cathodic Protection ......................................................................... 168

106. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1 With Polarization Potential -1.0 V ................................................................ 169

107. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 6.8

Without Cathodic Protection ......................................................................... 170 108. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 6.8

With Polarization Potential -1.0 V ................................................................ 171 109. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2

Without Cathodic Protection ......................................................................... 172 110. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2

With Polarization Potential -1.0 V ................................................................ 173 111. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 Without Cathodic Protection ............................................................. 175 112. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 With Polarization Potential -1.0 V .................................................... 176 113. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 6.8 Without Cathodic Protection ............................................................. 177 114. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at


xxi


Figure Page

115. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 9.2 Without Cathodic Protection ............................................................. 179

116. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at


xxiii

LIST OF TABLES Table Page

1. Compositions of 1018 Carbon Steel ....................................................................... 26 2. Compositions of High Chromium Alloy ................................................................. 26 3. Phosphate Size Analysis Results ............................................................................. 28 4. Pond Water Composition Analysis.......................................................................... 30 5. Chemical Analysis Results of Phosphate Sample ................................................... 30 6. Test Levels of Parameters for BBD......................................................................... 62 7. Factorial Design and Experimental Response for 1018 Carbon Steel..................... 72 8. Analysis of Various for 1018 Carbon Steel............................................................. 73 9. Analysis of Various for Reduced Quadratic Model of 1018 Carbon Steel ............. 74

10. Factorial Design and Experimental Response for High-Chromium Alloy ............. 86 11. Analysis of Various for High-Chromium Alloy...................................................... 87 12. Required Current Density for Effective Protection............................................... 108 13. Effect of Sodium Sulfate Concentration on Wear Rate of

1018 Carbon Steel at pH 3.1 ........................................................................... 119 14. Effect of Sodium Sulfate Concentration on Wear Rate of



High-Chromium Alloy at pH 3.1 .................................................................... 125 17. Ball Mill Size and Operating Constants ................................................................ 185 18. Economic Evaluation for Different Materials for Two Mills................................ 186

1

EXECUTIVE SUMMARY

Approximately half of the grinding media wear results from corrosion or oxidation dissolution of metal surfaces in grinding. Corrosion may play a greater role in wear of phosphate grinding mills used in Florida fertilizer plants due to the acidic slurry (pH 2-4). At acidic pHs, protective passivation films of metal oxides and/or hydroxides do not exist on the surface of grinding balls and mill liners, exposing fresh metal surface that is more susceptible to corrosion.

Metal corrosion reactions are electrochemical in nature and their reaction rates are controlled by the electrochemical potential at the surface. A reduction in potential with excess electrons (e-) will depress the anodic dissolution reaction of metal (e.g., M = M2+ + 2e-). This can be accomplished by supplying an impressed current to the object to create a negative potential change called cathodic polarization which reduces the rate of metal reaction. In the present laboratory studies the impressed current was supplied by an EG&G PARC potentiostat. Fundamental corrosion analysis indicated that a 0.1 V reduction in potential of the iron or steel electrode can reduce the corrosion rate by more than 99%.

This research project was intended to develop an effective cathodic protection system that is capable of minimizing the corrosive wear of the mill liner and grinding media for the grinding mills used in Florida phosphate industry by creating a reducing environment. The cathodic system protects the mill from corrosion by impressing a cathodic current to the mill, lowering its electrochemical potential to a value at which its oxidation reaction is negligible. The research project involved the following major tasks: SPECIALLY DESIGNED BALL MILL Unlike earlier mill corrosion studies that used devices that merely simulated mill motion, this research program utilized a specially designed ball mill which closely resembled the actual operating conditions. The potential and current applied to the coupons were controlled using an EG&G PARC, Model 273 potentiostat. The results from this study closely resembled those to be expected from the industrial ball mills. PHOSPHATE ACQUISITION AND CHARACTERIZATION

Phosphate samples and pond water samples were acquired from CF Industries,

Plant City, Florida. The phosphate sample was thoroughly mixed and wet screened into seven different size fractions. Each size fraction and the head sample were analyzed for their compositions. The composition analysis for phosphogypsum pond water was also carried out.

2

DETERMINATION OF POLARIZATION DIAGRAM FOR FUNDAMENTAL ASSESSMENT

The polarization diagram gives a fundamental quantitative assessment of the decrease in corrosion rate caused by cathodic polarization. The effects of solution pH, electrode material, gaseous environment, and solution composition on polarization curves were performed. Polarization diagrams indicate that the current density is higher in nitrogenated solution than in oxygenated solution at the same pH value. This is mainly because the fresh surface of electrode is readily oxidized in oxygenated solution than in nitrogenated solution, which passivates metal surface. The current density of 1018 carbon steel was higher than that of high-chromium alloy under the same condition. This is attributed to the fact that the addition of chromium to iron increases the ease of passivation by reducing critical anodic current density. The polarization diagrams also indicated that the current density is higher in buffer solution than that in pond water under the same operating conditions. This behavior is attributed to the fact that the pH of a buffer solution resists changes caused by small amounts of base produced during oxidation reactions; the pH of a pond water solution increased as reactions took place.

CHARACTERIZATION OF CORROSION RATE UNDER VARIOUS OPERATING CONDITIONS

To characterize the total wear, corrosive wear and abrasive wear of the ball mill under various operating conditions, a statistical Box-Behnken Design (BBD) of experiments was performed to evaluate individual variables and their interactions. The most significant variables and optimum conditions were determined from statistical analysis of the experimental results using response surface methodology. Solution pH had the most significant effect on the wear rate for both 1018 carbon steel and high-chromium alloy. The optimum process parameters for minimum wear rate were solution pH at 7.36, rotation speed at 70.31 RPM, solid percentage at 75.50, and crop load (% total volume of grinding media, ore, and water relative to volume of the mill ) at 71.94% for 1018 carbon steel; solution pH at 8.69, rotation speed at 61.13 RPM, solid percentage at 64.86, and crop load at 57.63% for high-chromium alloy. EFFECT OF POLARIZATION POTENTIAL ON CORROSION RATE Anodic polarization represents a driving force for corrosion by the anodic reaction. When surface potential measures more positive, the oxidizing (or corrosive) power of the solution increases. To investigate effects of polarized potential on corrosion rate of the ball mill and grinding media, three-hour grinding tests were conducted under controlled electrochemical conditions. Experimental results indicate that polarization potential of -1.0 V was sufficient to reduce the wear rate of 1018 carbon steel. The total wear rate was reduced by 42% to 46 % for 1018 carbon steel when a potential of -1.0 V was applied. Similar results were obtained with the high-chromium alloy, except that the required polarization potential was -0.7 V.

3

DETERMINATION OF REQUIRED CURRENT DENSITY FOR EFFECTIVE PROTECTION

The ball mill can be protected from corrosive wear using the impressed current technique which utilizes an external power source to force current out of the dimensionally stable anode (DSA), into the electrolyte, and eventually onto the ball mill. The impressed current cathodically polarizes the ball mill and reduces its corrosion rate. The magnitude of impressed current determines the effectiveness of corrosive wear rate reduction. The required current density to effectively reduce the wear rate of 1018 carbon steel was 210 mA/m2 in pH 3.1 solution, 180 mA/m2 in pH 6.8 solution, and 160 mA/m2 in pH 9.2 solution. While the required current density for high-chromium alloy was 150 mA/m2 in pH 3.1 solution, 125 mA/m2 in pH 6.8 solution, and 95 mA/m2 in pH 9.2 solution. CORROSION MECHANISM INVESTIGATION

To understand the cathodic protection process, scanning electron microscopy (SEM), energy dispersive spectrometer (EDS), and X-ray diffraction (XRD) methods were used to investigate the corrosion products, surface morphology, and composition of metals during phosphate grinding with and without cathodic protection in different solutions. SEM results indicated that pitting corrosion was the main corrosion type. EDS images revealed the element wt% and atom% changes of 1018 carbon steel and high-chromium alloy after they were exposed to phosphate slurry in different solutions with and without cathodic protection. XRD images suggested that Fe2O3 was the main corrosion product for 1018 carbon steel in pH 3.1 solution without cathodic protection, while Fe2O3, Fe3O4 and Fe(OH)3 were the corrosion products in neutral and alkaline solutions. For high-chromium alloy, Fe2O3, Cr2O3 and FeCr2O4 were the main corrosion products in pH 3.1 solution; Cr2O3 and FeCr2O4 were the corrosion products in pH 6.8 and pH 9.2 solutions. XRD images also suggested that the corrosion was significantly reduced since most of the corrosion products disappeared from the coupon surface when polarization potential was applied.

5

INTRODUCTION The United States is the largest phosphate rock producer in the world. About 30% of the world production in 1990 was produced by the United States (Bartels and Gurr 1994). This level of production has continued to the present. The State of Florida generates approximately 70% of the U.S. annual phosphate production of about 40 million tons (Harben 1980). A typical phosphate beneficiation process in central Florida includes washing, classification, fatty acid flotation, acid scrubbing, and amine flotation. The upgraded phosphate concentrate is used for toothpaste, detergents, food, fertilizers, etc. Approximately 95% of produced phosphate is consumed in fertilizer plants where phosphorus, together with nitrogen and potassium, is utilized as the major nutrient for plants.

In fertilizer plants, phosphate is used to produce phosphoric acid by reacting phosphate with sulfuric acid. The phosphogypsum crystals formed during this chemical reaction are discarded and the phosphoric acid is further processed to give the finished fertilizer product. To increase the chemical reaction rate, phosphate rock is ground to fine particle sizes prior to the chemical reaction process. It usually takes 10-15 minutes of grinding to reduce phosphate size to more than 40% -200 mesh. The direct operating costs in grinding are mainly the energy consumed and the metal lost through wear and corrosion. Fine grinding of phosphate is energy intensive and more than 60% of the electrical energy consumed in the phosphate fertilizer plant is used in grinding phosphate rock. Thousands of tons of grinding media is consumed in phosphate chemical plants every year. Reducing wear and corrosion rates will significantly reduce the grinding costs and improve the grinding efficiency by maintaining optimal grinding ball size.

Previous efforts to reduce grinding media consumption were mainly on finding

the most wear-resistant metal, e.g., high-chromium alloys, and significant improvements have been achieved. However, high-chromium alloys are rather expensive. Reduced consumption rate of metal alloy will result in significant savings for the Florida phosphate industry.

This research program was intended to minimize the corrosive wear of the mill

liner and grinding media by developing an effective cathodic protection process based on the impressed current principle. The cathodic process protects the mill from corrosion by impressing a cathodic current to the mill, lowering its electrochemical potential to a value at which its oxidation reaction or corrosion rate is negligible. This also helps to reduce the abrasive wear rate by eliminating the synergistic effects between corrosion and abrasion.

7

LITERATURE REVIEW

Corrosion is the destructive attack of a metal by chemical or electrochemical reaction with its environment. Deterioration by physical causes is not called corrosion, but is described as erosion, galling, or wear. In some instances, chemical attack accompanies physical deterioration as described by terms of corrosion-erosion, corrosive wear, or fretting corrosion. Nonmetals are not included in the present definition. Plastics may swell or crack, wood may split or decay, granite may erode, and Portland cement may leach away, but the term corrosion is presently restricted to chemical attack of metals (Uhlig 1971). To the great majority of people, corrosion means rust, an almost universal object of hatred. Rust is, of course, the name which has more recently been specifically reserved for the corrosion of iron, while corrosion is the destructive phenomenon which affects almost all metals. Although iron was not the first metal used by man, it has certainly been the most used, and must have been one of the first with which serious corrosion problems were obtained. It is not, therefore, surprising that the terms corrosion and rust are almost synonymous (Trethewey and Chamberlain 1988). IMPORTANCE OF CORROSION STUDIES The importance of corrosion studies is threefold. The first area of significance is economics including the objective of reducing material losses resulting from the corrosion of piping, tanks, metal components of machines, ships, bridges, machine structures, etc. The second area is improved safety of operating equipment which, through corrosion, may fail with catastrophic consequence. Examples are pressure vessels, boilers, metallic containers for radioactive materials, turbine blades and rotors, bridge cables, airplane components, automotive steering mechanisms. Third is conservation, applied primarily to metal resources--the world’s supply of these is limited, and the wastage of them includes corresponding losses of energy and water reserves associated with the production and fabrication of metal structures. Currently the prime motive for research in corrosion is provided by the economic factor. Losses sustained by industry, by the military, and by municipalities amount to many billions of dollars annually (Trethewey and Chamberlain 1988). Economic losses are divided into direct losses and indirect losses. By direct losses are meant the costs of replacing corroded structures and machinery or their components, such as condenser tubes, mufflers, pipelines, and metal roofing, including necessary labor. Other examples are repainting of structures where prevention of rusting is the prime objective and the capital costs plus upkeep of cathodically protected pipelines. Sizeable direct losses are illustrated by the necessity to replace several million domestic hot water tanks each year because of failure by corrosion or, similarly, the need for annual replacement of millions of corroded automobile mufflers. Direct losses

8

include the extra cost of using corrosion-resistant metals and alloys instead of carbon steel where the latter has adequate mechanical properties but not sufficient corrosion resistance; there are also the costs of galvanizing or nickel plating of steel, of adding corrosion inhibitors to water, or of dehumidifying storage rooms for metal equipment. The total combined loss to the United States alone has been estimated to be about $126 billion per year, according to the Department of Commerce (Jones 1996). Indirect losses are more difficult to assess, but a brief survey of typical losses of this kind compels the conclusion that they add several billion dollars to the direct losses already outlined. Indirect losses may occur as a result of any of the following: lost production during a shutdown or as a result of a failure, high maintenance costs, loss of efficiency, loss of product quality in a plant owing to contamination from corrosion of the materials used to make the production line, loss of customer confidence and sales, and extra cost to redesign the corroded parts etc. (Klein and Rice 1966; Uhlig 1971; Trethewey and Chamberlain 1988). TYPE OF CORROSION DAMAGE

There are many forms of corrosion: uniform corrosion, crevice corrosion, stress corrosion, pitting corrosion, intergranular corrosion, dezincification and parting corrosion, etc. Corrosion is often accelerated in certain places where cracks appear as a result of the weakening of the material or fatigue after it is subjected to mechanical forces periodically. This is particularly significant for grinding media that are constantly under the influence of impact and attrition (Uhlig 1971).

In general, metals or alloys are covered with oxide or hydroxide films, resulting in

surface passivation which reduces corrosion rate. Formation of cracks and fissures can destroy the passivation. The depth of crevices increases rapidly because it is only there that the metal is not covered with a protective layer of oxide/hydroxide. The result is an increase in surface roughness and possible problems due to reduction in mechanical strength. With the grinding mill crevice corrosion could be a significant problem for the mill shell. For grinding media (balls and rods), the passivation layer is removed during grinding and fresh surface is always exposed. As a result, they are more susceptible to corrosion. A uniform, regular removal of metal from the surface is the usually expected mode of corrosion for grinding media. This includes the commonly recognized rusting of iron. Fogging of nickel and high-temperature oxidation of metals are also examples of uniform attack. This is especially true for grinding operations under acidic solutions, such as in Florida phosphate fertilizer plants.

Crevice corrosion is shown in Figure 1a. Corrosion is often greater in the small

sheltered volume of the crevice created by contact with another. The second material may be part of a fastener (bolt, rivet, washer) of the same or a different alloy, a deposit of mud, or a nonmetallic gasket. Corrosion within a crevice may be caused in atmospheric exposures by retention of water, while the outer surfaces can drain and dry.

9

(a) (b) (c) Figure 1. Forms of Corrosion: (a) Crevice Corrosion, (b) Stress Corrosion, and (c)

Pitting Corrosion.

Stress corrosion (shown in Figure 1b) is caused by the mechanical forces applied to metals. When the corrosion reaction occurs with hydrogen evolution, hydrogen atoms, owing to their small size, can enter the metallic lattice and thus reduce the strength of the interatomic bonds. This is known as hydrogen embrittlement. If a mechanical force is applied to the metal, there is a great possibility that it will rupture.

Pitting corrosion (shown in Figure 1c) results from localized attack in an

otherwise resistant surface. The rate of corrosion is greater at some areas than at others. The pits may be deep, shallow, or undercut. If appreciable attack is confined to a relatively small fixed area of metal, acting as anode, the resultant pits are described as deep. If the area of attack is relatively larger and not so deep, the pits are called shallow. Pitting corrosion is often caused by the presence of chloride ions that manage to pass through the passive film and initiate corrosion, resulting in rupture of the passive film. It is one of the most destructive forms of corrosion because it progresses rapidly. The stainless steels and nickel alloys with chromium that depend on a passive film for corrosion resistance are especially susceptible to pitting by local breakdown of the film at isolated sites.

Fretting corrosion, which results from slight relative motion (as in vibration) of

two substances in contact, one or both being metals, usually leads to a series of pits at the metal interface. Metal-oxide debris usually fills the pits so that only after the corrosion products are removed do the pits become visible.

Cavitation-erosion resulting from formation and collapse of vapor bubbles at a

dynamic metal-liquid interface (as in rotors of pumps or on tailing faces of propellers) causes a sequence of pits, sometimes appearing as a honeycomb of small relatively deep fissures.

Intergranular corrosion is a localized type of attack at the grain boundaries of a

metal, resulting in loss of strength and ductility. Grain-boundary material of limited area, acting as anode, is in contact with large areas of grains acting as cathode. The attack is often rapid, penetrating deeply into the metal, and sometimes causing catastrophic

10

failures. Improperly heat-treated 18-8 stainless steels or Duralumin-type alloys (4% Cu-Al) are among the alloys subject to intergranular corrosion.

Dezincification is a type of attack occurring with zinc alloys (e.g., yellow brass)

in which zinc corrodes preferentially, leaving a porous residue of copper and corrosion products. The alloy so corroded often retains its original shape, and many appear undamaged except for surface tarnish, but its tensile strength and especially ductility are seriously reduced. Dezincified brass pipe may retain sufficient strength to resist internal water pressures until an attempt is made to uncouple the pipe, or a water hammer occurs, causing the pipe to split open.

Parting is similar to dezincification in that one or more reactive components of the

alloy corrode preferentially, leaving a porous residue that may retain the original shape of the alloy. Parting is usually restricted to such noble metal alloys as gold-copper or gold-silver and is made use of practically in refining of gold. FACTORS AFFECTING CORROSION

Comminution is one of the most important unit operations in the phosphate beneficiation and fertilizer industry. It is well known that the direct operating costs in crushing and comminution circuits are mainly the energy consumed and the metal lost through wear and corrosion (Allen 1993; Shishodia and others 1993). More than 60% of the electrical energy consumed in the phosphate fertilizer plant is used in grinding phosphate rock. It is estimated that more than 1 million tons of grinding rods and balls were consumed annually by the U.S. mineral industry alone. An additional 1.5 million tons of iron and steel was consumed as mill liners (National Materials Advisory Board 1981). These losses essentially represent metal worn away by abrasion, removed by corrosion, and lost as scrap because of wear and corrosion (Iwasaki and others 1985).

Factors that affect metal loss in comminution operations include the composition

and metallurgical properties of the grinding media, mineral properties (particularly hardness and particle size), properties of the pulp such as solid/liquid ratio and pH, and mill operating conditions. Abrasion occurs through direct impact of falling balls, indirect impact propagated through other balls, and contact between spinning balls. Abrasion rate depends on pulp viscosity, rheological properties (Klimpel 1982, 1983), surface tension (Meloy and Crabtree 1967), and the distribution and characteristics of micro constituents in the metal or alloy (Norman and Loeb 1948). Corrosion rate depends on the chemical composition of the metal, electrochemical potential developed by each component (rods or balls, liners, and mineral charges) and galvanic contacts between them. The electrochemical potentials are determined by dissolved ions, dissolved oxygen, solution pH, and the presence of oxidizers such as oxygen and ferric ions. Corrosive wear increases significantly when galvanic coupling exists between the mineral surface and the grinding media or mill liner (high carbon steel, manganese steel, porcelain or rubber) that generates a corrosion current. Abrasion and corrosion have been found to show pronounced synergistic effects that significantly increased metal loss in aqueous

11

environments (Dun 1985; Abuzriba and others 1992). This suggests that techniques for corrosion prevention and control will significantly reduce abrasion as well.

The solution pH has significant effects on the corrosion of metal in aerated water

since it changes the cathodic reduction reaction (Muller and Gampper 1994). Uhlig and Revie (1985) have found that in the intermediate pH 4 through 10 range, the corrosion of iron produced a loose, porous ferrous oxide deposit that shelters the surface and maintains the pH at about 9.5 beneath the deposit. The corrosion rate is nearly constant and is determined by uniform diffusion of dissolved oxygen through the deposit in this intermediate range of pH.

Dissolved oxygen in aqueous solution has profound influences on the corrosion of

iron or steel since oxygen reduction is the dominant cathodic reaction that drives the anodic reaction. Any factors affecting dissolved oxygen thus affects the corrosion of metal. Solution agitation or stirring increases the transport rate of dissolved oxygen, increasing corrosion rate.

Hydrogen is another factor that affects the metal loss. The influence of hydrogen

on corrosion potential and anodic dissolution in neutral NaCl solution has been studied by Wallinder and others (2001). The results show that hydrogen in chromium decreases the corrosion potential and increases the anodic dissolution rate.

The influence of hydrogen sulfide (H2S) on corrosion of iron under different

conditions was carried out by Ma and others (1998, 2000). It was found that H2S can either accelerate or inhibit corrosion of iron under different experimental conditions. In most cases, H2S has a strong acceleration effect on both the anodic iron dissolution and cathodic hydrogen evolution, causing iron to be seriously corroded in acidic medium. However, H2S can also have a strong inhibition on the iron corrosion under certain special conditions where H2S concentration is below 0.04 mmol dm-3, pH value of electrolyte solution is within 3-5 and the immersion time of the electrode is over 2 hour. The inhibition effect of H2S on the iron corrosion is attributed to formation of ferrous sulfide (FeS) protective film on the electrode surface.

An increasing temperature initially increases corrosion rates, but temperature

higher than about 80oC tends to reduce corrosion rate as a result of decreased solubility of dissolved oxygen. Other dissolved gases such as chlorine, ammonium, and carbon dioxide may also affect corrosion rate of metals, but their effects are relatively small in comparison of those of dissolved oxygen (Oesch 1996). The influence of temperature to the crevice corrosion of AISI 316 steel has been investigated by Jakobsen and Maahn (2001). The experiments in acidified environments indicate that crevice corrosion at low temperatures results from acidification in the crevice. At higher temperatures crevice corrosion is believed to be the result of metastable pitting stabilized by the crevice.

It is reported that alloying elements affect the corrosion behavior of materials.

Chromium and molybdenum contents resist pitting and uniform corrosion because they form dense and secure protective films of Cr2O3 and MoO2 on the surface (Salih 2002;

12

Wang and others 2002). Nickel alone cannot resist corrosion, but with the presence of chromium and molybdenum it has good corrosion resistance (Liu and others 2001). Chromium by itself is very corrosion resistant. It is therefore used as a coating for base metals to maintain their integrity. An example is decorative trim for appliances. Of even greater importance is the use of chromium in alloys. It is the key component of stainless steel. McBee and Kruger (1972) have shown that the addition of chromium to iron in an alloy causes the oxide film to go from polycrystalline to noncrystalline as the amount of chromium increases.

It is estimated (Jang and others 1989) that well over half of the grinding media

wear results from corrosion or dissolution from the active nascent metal surfaces continuously being exposed in grinding. This is particularly true for phosphate grinding mills widely used in Florida fertilizer plants where slurries are strongly acidic. At acidic pH′s, protective passivation films of metal oxides/hydroxides do not exist on the surface of balls and liners, exposing fresh metal surface that is more susceptible to corrosion. The significance of corrosion in media wear is also indicated by the fact that metal loss in wet grinding where corrosion is present is about 7 times as much as in dry grinding where corrosion is absent (Bond, 1964). Hoey and others (1975) demonstrated that a reduction in ball wear of up to 50% was achieved through the use of corrosion inhibitors. Pazhianur and others (1997) showed cathodic protection of ball mills from corrosion reduced mill wear by 30 to 60%. These studies clearly indicate that the wear rate of the ball mill and grinding media can be reduced considerably by controlling the corrosion chemistry inside the grinding mill. TECHNIQUES USED TO PROTECT AGAINST CORROSION

There are a number of techniques that can be used to prevent metal corrosion. They include use of different coatings, inhibitors, biofilm, alloy substitution, design, cathodic protection, etc. (Munger 1984; Elliot 1987; Charles and Crane 1989; Fedrizzi and Bonora 1997; Davis 2000; Liu and others 2001).

Coatings maintain the functional ability of structures and equipment to bear loads,

maintain dimensions, store and transport liquids and gases, and so on. However, in addition, coatings also act as a corrosion resistant physical barrier to isolate the underlying metal from the corrosive media (Jones 1996; Davis 2000). Souto and Alanyali (2000) found that steel coated with TiN and TiAlN had a lower corrosion rate (current density), about three orders of magnitude lower than the untreated steel substrates. It was found that the phosphating with a subsequent chromate passivation and oil impregnation significantly improved the corrosion resistance of the nitrided steel, imparting the resistance up to an order of magnitude higher than that of the unnitrided stainless steel (Rausch 1990; Flis and others 2001).

It is known that a protective film formed on the metal surface using different

methods can reduce the corrosion rate significantly. Giddey and co-workers (2001) found that in the industries involving alkaline solutions in different process streams, the nature

13

and stability of oxide films formed on the metallic surfaces determine the rate of erosion-corrosion of the equipment. It was reported that alkylamines were strongly chemisorbed on iron surface to form an ultrathin protective film for iron corrosion in acid solutions (Tsuji and others 2000; Senkevich and others 2000). An alkanethiol self-assembled monolayer adsorbed on copper has been modified with alkylchlorosilanes to form protective films of two-dimensional polymer against copper corrosion (Haneda and Aramaki 1998; Taneichi and others 2001). Nozawa and Aramaki (1999) found that one- and two-dimensional polymer films of modified alkanethiol monolayers seem to be promising in protecting against the aqueous and atmospheric corrosion of iron.

The involvement of microorganisms in metal corrosion has proposed by a number

of investigators (Kobrin 1976; Tatnall 1981; Pope and others 1984; Stoecker and Pope 1986; Scotto and others 1986). It was reported that in the presence of a bacterial biofilm produced by bacillus subtilis pitting corrosion was observed initially. However, after about three days the pitting stopped (Ornek and others 2001; Salvago and Magaginin 2001). Corrosion control using biofilms was also achieved by Geesey and others (1996).

Another role of microorganisms in metal corrosion is to repair the coatings. It

was reported that damaged or partly corroded vivianite (Fe3(PO4)2·8H2O) coatings on mild steel could be rephosphated in cultures of pseudomonas putida. The corrosion resistance of bacterially repaired steel during three weeks of exposure to a corrosive medium was as good as of bacterially phosphated steel (Jayaraman and others 1997; Volkland and others 2001).

Inhibitors are chemical compounds that deposit on exposed metal surfaces from

the corrosive environment. The inhibitor may form a uniform film, which acts as a physical barrier (Jones 1996). Chromates, silicates, and organic amines are common inhibitors. The mechanisms of inhibition can be quite complex. In the case of the organic amines, the inhibitor is adsorbed on anodic and cathodic sites and stifles the corrosion current. Some inhibitors specially affect either the anodic or cathodic process. Others promote the formation of protective films on the metal surface (Davis 2000). Bentiss and others (2000) found that 1,3,4-oxadiazoles is a new class of corrosion inhibitors of mild steel in acidic media. Results obtained reveal that the inhibition efficiency value increases with the inhibitor concentration and reaches a maximum at 80 mg.l-1 in acidic media. Another recently synthesized compound 1(benzyl)1-H-4,5-dibenzoyl-1,2,3-triazole (BDBT) has an notable inhibition effect on mild steel. The corrosion rate of mild steel in 1% HCl was reduced by more than 95% in the presence of 50 ppm of BDBT (Abdennabi and others 1996). It was reported that thiocarbamide and vanillin act as good inhibitors for the corrosion of mild steel and aluminum in acidic solution, respectively. The inhibiting effect increases with increase of inhibitor concentration and temperature of the corrosion medium (Stoyanova and others 1997; El-Etre 2001).

Polarization and weight loss studies showed that 2-mercaptopyrimidine is

effective for the inhibition of low carbon steel over a wide concentration range of aqueous phosphoric acid solutions (Wang 2001). The inhibitor retards the anodic and

14

cathodic corrosion reactions with emphasis on the former. Weight loss studies also showed that both 2-mercaptothiazoline and cetyl pyridinium chloride are effective for the inhibition of low carbon steel over a wide concentration range of aqueous phosphoric acid solution (Wang and others 2001).

The action of some organic compounds from the group of surfactants and

polyethylene glycols (PEG) on zinc corrosion in alkaline media was investigated by Dobryszycki and Biallozor (2001). The investigations were carried out by use of electrochemical and nonelectrochemical methods. The effectiveness of the inhibitors was compared and it was found that the PEG of average molar weight 400 was especially effective. Some polymers can also be used as inhibitors. It is reported that the corrosion rate of zinc can be inhibited by addition of styrene-maleic acid-acrylic ester copolymers (Muller and others 2000).

It was also found that natural honey exhibited a very good performance as

inhibitor for steel corrosion in high saline water. The inhibition efficiency increases with an increase in neutral honey concentration. After some time, the inhibition efficiency decreased due to the growth of fungi in the medium. The adsorption of natural honey on the carbon steel was found to follow the Langmuir adsorption isotherm (Gonzalez and others 1996; El-Etre 1998; El-Etre and Abdallah 2000).

Yasuda and co-workers presented the “barrier-adhesion” principle for corrosion

protection of aluminum alloy. It was found that if a good barrier is adhered to the metal surface by the tenacious water-insensitive adhesion, the corrosion-induced delamination could be prevented without corrosion inhibitors in the premier. The coatings based on this principle could prevent both damaged surface corrosion and pitting corrosion, which are difficult to simultaneously prevent with coatings solely on electrochemical corrosion protection, such as chromium conversion coating or chromated primers (Yasuda and others 2001).

The effect of adatoms on the corrosion rate of copper was studied by Hourani and

Wedian (2000). The results showed that the nonmetals (F, S, and I) markedly enhanced the rate of corrosion; Cr, Ta, Sb, Bi, Ti, and Sn slightly enhanced the rate of corrosion of copper. Zinc, however, was the only element which decreased the rate of corrosion. Inhibition of zinc corrosion in aerated 0.5 M NaCl with multivalent cations, Al3+, La3+, Ce3+ and Ce4+ was investigated by polarization techniques and XPS (Aramaki 1999, 2000, 2001). It was found that Ce3+ and La3+ can reduce zinc corrosion by more than 90% but those of other cations stimulate zinc corrosion. Gunasekaran and his co-workers (1997, 2001) found that corrosion of steel in oxygen saturated environment can be inhibited by insoluble products with the metal ions exiting on the solution, which precipitate on the surface to form a three-dimensional protective layer.

It was reported that surface treatment of stainless steel is an effective method to

reduce the corrosion rate (Habazaki and others 1996; Breslin and others 1997; Bozec and others 2001). The influence of stainless steel surface treatment on the oxygen reduction reaction in seawater was investigated by Bozec and co-workers (2001). Their results

15

showed that the oxygen reduction is controlled by the properties of the surface, which are governed by the composition of the alloy and the surface preparation. Their work has also shown that the rate of O2 reduction increases in the following sequence: chemically treated surface < passivated surface < polished surface < prereduced surface.

Itoh and others (2000) investigated the effects of various surface states on the

corrosion rates of copper in air containing water vapor and sulfur dioxide by in situ IR-RAS. It was found that the initial corrosion rate greatly depends on the surface state. A fresh oxide layer slightly inhibits the corrosion. On the other hand, an aged oxide layer formed in one day considerably delays the corrosion rate. The corrosion products are the same in spite of different surface states and corrosion rates.

The use of lasers in metal surface processing is an effective method to increase its

corrosion resistance (Lu and Aoyagi 1994; Cottam and Emmony 1998; Conde and others 2001). The surface of 304 steel was melted in argon and nitrogen atmospheres using different laser beam scan rates and gas flows. The results obtained show that in both cases the laser surface melting (LSM) induced improvements in pitting resistance with respect to the base steel. Electrochemical studies showed that the improvements are greater when the LSM is carried out in a nitrogen environment because of the incorporation of nitrogen into solid solution.

The application of rational design principles can eliminate many corrosion

problems and greatly reduce the time and cost associated with corrosion maintenance and repair. Corrosion often occurs in dead spaces or crevices where the corrosive medium becomes more corrosive. These areas can be eliminated or minimized in the design process. Where stress-corrosion cracking is possible, the components can be designed to operate at stress levels below the threshold stress for cracking (Davis 2000).

Some new techniques and equipment for corrosion control have been developed

in recent years. Microelectrochemical measurements are a new method to determine the influence of small precipitates and depletion zones on the corrosion behavior of a single phase (Perren and others 2001). Acoustic emission (AE) technique was first used to study the development of pitting corrosion on austenitic stainless steels by Fregonese and others (2001). The corrosion step is characterized by the emission of resonant signals.

An optical corrosion-meter has been developed for materials testing and

evaluation of different corrosion phenomena. The idea of the optical corrosion-meter was established based on principles of 3D-holographic interferometry for measuring microsurface dissolution, i.e. mass loss, and on those of electrochemistry for measuring the bulk electronic current, i.e. corrosion current of metallic samples in aqueous solutions (Habib 2000).

Cathodic protection is one of the most widely used methods of corrosion

prevention. Cathodic protection suppresses the corrosion current that causes damage in a corrosion cell and forces the current to flow to the metal structure to be protected. Thus, the corrosion or metal dissolution is prevented. In practice, cathodic protection can be

16

achieved by two application methods, which differ based on the source of the protective current. An impressed-current system uses a power source to force current from inert anodes to the structure to be protected. A sacrificial-anode system uses active metal anodes, for example, zinc or magnesium, which are connected to the structure to provide the cathodic-protection current. In principle, it can reduce or prevent the corrosion of any metal or alloy exposed to any aqueous electrolyte. Corrosion can be reduced to virtually zero, and a properly maintained system will provide protection indefinitely.

The cathodic protection principle has been used in mineral beneficiation to

control mineral oxidation and ball mill corrosion. Tao and others (1997) used the galvanic coupling principle for control of pyrite oxidation, which significantly improved pyrite rejection in coal flotation. Tomashov and Chernova (1967) found that the corrosion rate of steel balls reduced drastically in contact with a cathodic protector even in sulfuric acid. Hoey and others (1975) reduced grinding ball wear of up to 50% through corrosion prevention. Pazhianur and others (1997) reported a 40-60% of reduction in corrosion rate of the mill liner as a result of cathodic protection.

Studies of ball mill corrosion are often conducted using a device that simulates the

ball mill (Chenje and others 2003). Kotlyar and others (1987) used a rotating cylinder/anvil apparatus and Pozzo and Iwasaki (1989) a rotating ball/pin setup to investigate the weight loss of grinding media. The mechanisms of corrosive and/or abrasive wear of grinding media were studied by measuring galvanic current on a three-electrode electrochemical system (Nakazawa and Iwasaki 1985; Pozzo and Iwasaki 1987; Pozzo and others 1990), carrying out ball wear tests in a porcelain mill (Iwasaki and others 1985), and performing spectroscopic surface characterization (Nakazawa and Iwasaki 1985). A more recent study by Pazhianur and others (1997) used a small ball mill constructed of a mild steel pipe that resembled an industrial ball mill. The corrosion environment of the ball mill shell was controlled using a potentiostat that can apply a predetermined potential to the mill for cathodic protection.

However, a practical approach to protecting both grinding media and the mill

shell remains yet to be developed. In view of the characteristics and operating conditions of grinding ball mills used in the Florida phosphate industry, cathodic protection using impressed current is considered to be the most practical and effective method for wear minimization.

17

SCIENTIFIC DISCUSSION FUNDAMENTALS OF ELECTROCHEMICAL CORROSION

Corrosion of metals in aqueous environments is electrochemical in nature, involving two or more electrochemical reactions taking place on the metal surface. A change in electrochemical potential or the electron activity or availability at a metal surface has a profound effect on the rate of corrosion reaction. Thus, corrosion reactions are said to be electrochemical.

As a result of corrosion reaction, some of the elements of the metal or alloy change from a metallic state into a non-metallic state. The products of corrosion can be dissolved species or solid corrosion products. In either case, the energy of the system is lowered as the metal converts to a lower-energy form. Rusting of steel is the best known example of conversion of a metal (iron) into a nonmetallic corrosion product (rust). The change in the energy of the system provides the driving force and controls the spontaneous direction for a chemical reaction according to the laws of thermodynamics. The electrochemical nature of corrosion of metal M in acidic solutions can be characteristically described by two half-cell reactions: .22 −+ += eMM (1) .22 2HeH =+ −+ (2) Reaction (1) is referred to as the anodic half-cell reaction. It is an oxidation process in which metal valence increases from 0 to +2, liberating electrons, e. Reaction (2) is referred to as the cathodic half-cell reaction. It represents a reduction process in which the oxidation state of hydrogen decreases from +1 to 0, consuming electrons. The overall reaction for the corrosion process is: .2 2

2 HMHM +=+ ++ (3)

There is a change of free energy, ∆G, associated with any such chemical reaction. When the reaction products have a lower energy than the reactants, ∆G is negative in the spontaneous reaction. The free-energy change, ∆G, may be associated with an electrochemical potential, E, at equilibrium, by the fundamental relationship ∆G = -nFE (4) where n is the number of electrons (or equivalents) exchanged in the reaction, and F is Faraday’s constant, 96,500 coulombs per equivalent. Thus, we have the fundamental relationship in which a charge, nF, taken reversibly at equilibrium through a potential, E, corresponds to an energy change, ∆G. The half-cell reactions (1) and (2) also have free-energy changes analogous to ∆G and corresponding potentials ea and ec. That is,

18

).(22 reactionanodiceMM −+ += ea (5) ).(22 2 reactioncathodicHeH =+ −+ ec (6) The algebraic sum of these potentials is equal to E in equation (7). That is, E = ea + ec (7) The potentials ea and ec have been called half-cell, single electrode, or redox (reduction/oxidation) potentials for the corresponding half-cell reactions.

H+

H+

H+

H+

H+

H+

H2

H+

M2+

Metal M

e -

e -

Figure 2. Schematic of Simultaneous Metal Dissolution and Hydrogen Evolution.

This electrochemical reaction process involving charge transfer or exchange of electrons can be shown schematically in Figure 2. The metal dissolves by reaction (1), liberating electrons into the bulk of the metal which migrate to the adjoining surface, where they react with H+ in solution to form H2 by reaction (2). The aqueous solution, called the electrolyte, is required to carry ions such as M2+ and H+ in order for the electrochemical reaction to occur.

It has been found that when an excess of electrons are supplied to the metal in

Figure 2, the rate of corrosion, expressed by the anodic reaction (1), is reduced, while the rate of hydrogen evolution reaction (2) is increased. Therefore, an application of a

19

negative potential with attendant excess electrons always decreases the corrosion rate. This is the basis of cathodic protection of metal structure for the mitigation of corrosion.

It is known that for corroding metals the anodic reaction invariably is the form of

reaction (1) while the cathodic reduction reactions can be one or more of the following reactions in addition to reaction (2):

),(222 22 solutionsalkalineandneutralOHHeOH −− +=+ (8) ),(244 22 solutionsacidOHeHO =++ −+ (9) ),(442 22 solutionsalkalineandneutralOHeOHO −− =++ (10) .23 +−+ =+ FeeFe (11)

Reactions (2) and (8) represent the evolution of H2 from acid and neutral solutions, respectively. Reactions (9) and (10) show the reduction of dissolved oxygen in acid and neutral/alkaline solutions, respectively. Reaction (11) typifies the reduction of a dissolved oxidizer (e.g. Fe3+) in a redox reaction.

+2/ FeFee

+HHe

/2

Fe 2+ + 2e = Fe

Fe = Fe2+ + 2e

H2 = 2H+ + 2e

2H+ + 2e = H2

icorr, Fe

10-12 10-10 10-8 10-6 10-4 10-2

Current Density (A/cm2)

Pote

ntia

l (V

) vs.

SHE

corre

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

Figure 3. Anodic and Cathodic Half-Cell Reactions for Iron in Acid Solution.

When a metal is corroding, both the anodic and cathodic half-cell reactions, e.g., reactions (1) and (2), take place simultaneously on the surface. Each has its own half-cell electrode potential and exchange current density, as shown in Figure 3. As reactions (1) and (2) proceed on the surface of metal, the anodic electrode potential (ηa) and cathodic electrode potential ηc change respectively, according to equations (12) and (13) until they become equal at the mixed corrosion potential Ecorr (Fontana 1986),

20

o

aaa i

ilogβη = (12)

o

ccc i

ilogβη = , (13)

where βa and βc are the Tafel constants for the anodic and cathodic half-cell reactions, respectively, which can be considered to be about +0.1 and -0.1 V for many applications; ia and ic are the anodic and cathodic current density; io is the exchange current density equivalent to the reversible rate at equilibrium. The anodic Tafel constant βa is the cathodic overvoltage required to reduce corrosion rate by one order of magnitude. The exchange current io is the most important factor determining the rate of the reaction. The greater the βa value, the higher is the potential change needed for the protection of the anode (Jones 1971).

Electrochemical reactions either produce or consume electrons. Thus, the rate of electron flow to or from a reacting interface is a measure of reaction rate. Electron flow is conveniently measured as current, I, in amperes. The proportionality between I and mass reacted, m, in an electrochemical reaction is given by Faraday’s Law:

,nFItam = (14)

where F is Faraday’s constant, n the number of equivalents exchanged, a the atomic weight, and t the time. Dividing equation (14) through by t and the surface area, A, yield the corrosion rate, r:

.nFia

tAmr == (15)

where i, defined as current density, equals I/A. Equation (15) shows a proportionality between mass loss per unit area per unit time (e.g., mg/dm2/day) and current density (e.g., µA/cm2). Current density rather than current is proportional to corrosion rate because the same current concentrated into a smaller surface area results in a larger corrosion rate.

The effect of pH on the corrosion of metal in aerated water is caused by the cathodic reduction reaction. A ferrous oxide deposit will be formed on the metal surface. At the metal surface under the deposit, oxygen is reduced cathodically by reaction (10). In more acidic solutions below pH 4, the oxide is soluble and corrosion increases, due to availability of H+ for reduction by reaction (2) which is under activation control. In this pH range, metallurgical variables affecting the anodic reaction (1) have significant effects on the corrosion rate, which implies that the selection of the appropriate material is more important for ball mills operating at acidic pHs than at intermediate pHs. The absence of the surface deposit also enhances access of dissolved oxygen, which, if present, further increases corrosion rate through the cathodic reaction (9). At pH above 10, corrosion rate

21

is relatively low, due to formation of the passive ferric oxide film in the presence of dissolved oxygen. However, at pH above about 14 without dissolved oxygen, corrosion rate may increase when the soluble ferrite ion, HFeO2

-, forms. Dissolved oxygen affects the corrosion of iron or steel in aqueous solution by

oxygen reduction (9) or (10), which is the dominant cathodic reaction, that drives the anodic reaction (1). This is particularly true for neutral and alkaline conditions where reaction (2) is absent. Any factors affecting dissolved oxygen thus affect the corrosion of metal. Solution agitation or stirring increases the transport rate of dissolved oxygen, increasing corrosion rate. CATHODIC PROTECTION PRINCIPLES

Cathodic protection can be used to protect metals exposed to any aqueous electrolyte from corrosion and has been widely applied to pipe lines, bridges, oil platforms, etc. Cathodic protection reduces the corrosion rate by cathodic polarization of a corroding metal surface. Consider iron corroding in a dilute aerated neutral electrolyte solution. The respective anode and cathode reactions are equations (1) and (10), respectively. Cathodic polarization reduces the rate of the half-cell reaction (1) with an excess of electrons, which also increases the rate of oxygen reduction and OH- produced by reaction (10).

Using this method corrosion can be reduced to virtually zero, and a properly

maintained system will provide protection indefinitely. There are two possible ways to achieve cathodic protection. The first is to use a rectifier to supply an impressed current by converting alternating current to direct current. The second is to utilize a sacrificial anode which has a more active corrosion potential (i.e., more negative potential). By connecting the sacrificial anode to the metal to be protected, which is called galvanic coupling, the metal is cathodically polarized while the sacrificial anode is dissolved. The sacrificial anode must be replaced periodically as they are consumed by anodic dissolution.

22

+2/ FeFee

+HHe /2

Fe2+ + 2e = Fe

Fe = Fe2+ + 2e

H 2 = 2H

+ + 2e2H + + 2e = H

2

icorr

Current Density (A/cm2)

Pote

ntia

l (V

) vs.

SHE

corre

10-9 10-8 10-7 10-6 10-5 10-210-4 10-3 10-1

-0.4

-0.3

-0.2

-0.1

0.0

0.1

-0.5

-0.6+2/, FeFeoi

+HHoi /,2

βc= -0.10 V

βa= 0.04 V

ia ic

appi

Figure 4. Cathodic Polarization Protection of Steel by an Impressed Current in an

Acid Solution.

Figure 4 shows a polarization diagram that illustrates the effect of cathodic protection using an impressed current on corrosion rate for steel in acid solution. Without cathodic protection the corrosion potential is about -0.25 V and the corrosion rate is about 1.0 mA/cm2. Cathodic polarization of about 0.12 V, i.e., decreasing the corrosion potential from -0.25 V to -0.37 V by applying current iapp, will reduce the corrosion rate by three orders of magnitude, to 1.0 µA/cm2. The extent of reduction of corrosion rate is governed by the value of the anodic Tafel constant, βa. If βa = 40 mV as shown in Figure 4, each 40 mV of cathodic polarization decreases corrosion rate by an order of magnitude. The same principle can be applied to sacrificial anode cathodic protection where electrons flow from the active sacrificial anode to lower the potential of the noble cathode.

The cathodic polarization curve gives an indication of the impressed-current

density, iapp, for cathodic protection to any required level of cathodic polarization.

23

METHODOLOGY CONSTRUCTION OF A SPECIALLY DESIGNED BALL MILL Specially Designed Ball Mill

Unlike earlier mill corrosion studies that used devices that merely simulated mill motion (Kotlyar and others 1987; Pozzo and Iwasaki 1989), the present study utilized a specially designed ball mill under operating conditions. The results from this study were more closely resemble those to be expected from the industrial ball mills. The mill was constructed of a stainless steel pipe of 8½” in diameter and 12” in length with a wall thickness of ¼”, as illustrated in Figure 5. Three lifter bars were installed at identical intervals for effective grinding. The fabricated mill is shown in Figure 6. The mill resting on two rollers was driven by a one-horse power, three-phase motor, as shown in Figure 7.

To characterize the weight loss of the ball mill shell or its liner, three coupons

made of the same material as the mill shell or liner were installed flush with the interior surface of the mill. These coupons, 1” in diameter and ¼” in height, were electrically isolated from the mill. The potential or current of coupons was controlled via copper brushes riding on the concentric copper ring. The cathodic protection of the ball mill was accomplished using a ¼’’ diameter and 12’’ length titanium-indium oxide, dimensionally stable anode (DSA) acquired from Corrpro Companies Inc., Medina, Ohio that is located in an anode compartment and isolated from the slurry with a high-density polyethylene membrane (HDPE) to prevent particulates from contaminating the anode. Gases such as oxygen and hydrogen produced at the anode were released from the anode chamber through a vent shown in Figure 5. The potential and current applied to the coupons were controlled using an EG&G PARC, Model 273 potentiostat as shown in Figure 8. An industrial standard calomel electrode (SCE) was used as the reference electrode. The applied potential vs. SCE was converted to the potential with respect to Standard Hydrogen Potential (SHE) by adding 0.245 V.

24

Figure 5. Schematic of the Specially Designed Mill for Grinding Tests.

Figure 6. Photograph of the Specially Designed Mill for Grinding Tests.

DSA

Gas vent

Reference electrode

Slurry

Three copper ring brushes for mill potential control

Coupon

Copper ring brush for coupon potential control

Ring insulator

Sitting on roller Sitting on roller Not rotating

bearing

HDPE

25

Figure 7. Photograph of the Mill on Rollers.

Figure 8. Photograph of the EG&G PARC, Model 273 Potentiostat. Coupons and Dimensionally Stable Anode

The electrochemical condition of the ball mill and coupons was controlled by a potentiostat. The coupons used to represent the grinding balls and the mill shell have been made of 1018 mild carbon steel purchased from McMaster and high-chromium alloy acquired from Magotteaux, Nashville, Tennessee, a major grinding media supplier

26

for Florida phosphate industry, to represent different grinding media which are known to show different corrosion and abrasion behavior. The compositions of 1018 carbon steel and high-chromium alloy are shown in Table 1 and Table 2, respectively. The coupon assembly contains the metal coupon in the center and the insulator around it. The design of the coupon assembly took it into consideration that the metal coupon has to be taken out after each grinding experiment to determine the weight loss. Since the high-chromium alloy is extremely difficult to machine, two different designs of coupon assemblies were used, as shown in Figure 9. The first design (a) is suitable for 1018 mild carbon steel and the second for high-chromium alloy which cannot be threaded.

In addition to the reference electrode the dimensionally stable anode (DSA) that is

necessary for the potential control has been acquired from Corrpro Companies Inc., Medina, Ohio. Three different types of high density polyethylene (HDPE) membranes have been received from Daramic Inc. They were used to prevent solid particles from reaching the DSA. Table 1. Composition of 1018 Carbon Steel.

Sample Fe(%) C(%) Mn(%) P(%) S(%)

1018 Carbon Steel 98.81-99.16 0.15-0.20 0.60-0.90 00.04 0.50

Table 2. Composition of High-Chromium Alloy.

(a) (b)

Figure 9. Two Different Designs for Coupon Assembly for (a) Mild Steel and (b)

High-Chromium Alloy.

Sample Fe(%) Cr(%) C(%) Mn(%) Si(%) P(%) S(%) Ni(%) Mo(%)

High Cr Alloy

61.92-65.92

26.0-30.0 0.50 1.00 2.00 0.04 0.04 4.00 0.50

Mild steel

Insulator

Mill shell

AlloyCopper wire

Mild steel

Insulator

Mill shell

AlloyCopper wire

27

The constructed mill is shown in Figure 10 with the coupons, DSA, and the reference electrode. The electrochemical condition of the mill and the coupon was controlled by the potentiostat through the copper rings and ring brushes.

Figure 10. Photograph of the Specially Designed Mill with Coupons and Electrodes. PHOSPHATE ACQUISITION AND CHARACTERIZATION

Twelve 5-gallon buckets of phosphate samples and three 5-gallon of pond water samples were acquired from the CF Industries, Plant City, Florida. The phosphate sample was thoroughly mixed and split into small lots of about 2 kg each for storage. A representative sample was taken for size distribution analysis.

The phosphate sample was wet screened into seven different size fractions. The

size distribution is shown in Table 3. The cumulative weight percent vs. particle size for phosphate sample is shown in Figure 11. It is clear that the particle is quite evenly distributed in all size fractions except in -200 mesh. Few particles are smaller than 200 mesh. Each size fraction and the head sample were analyzed for the ingredients. The acidic phosphogypsum pond water was used as part of feed water, which is a standard

28

industrial practice. The water analysis and phosphate analysis results are shown in Tables 4 and 5, respectively. Table 3. Phosphate Size Analysis Results.

Sample Wt (%) ∑Wt (%)

+10 mesh 15.74 15.74

-10+20 27.67 43.41

-20+40 13.88 57.29

-40+60 21.44 78.73

-60+100 13.20 91.93

-100+200 7.87 99.80

-200 0.20 100.00

29

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Particle Size (mm)

0

10

20

30

40

50

60

70

80

90

100C

umul

ativ

e W

eigh

t (%

)

0.00 0.02 0.04 0.06 0.08 0.10 0.12

0

200

400

600

800

100

120

140

160

180

200

Figure 11. Cumulative Weight Percent vs. Particle Size for Phosphate Sample.

30

Table 4. Pond Water Composition Analysis.

Component Weight (%)

Water 96

Phosphoric acid <4

Sodium potassium fluosilicate <1.5

Calcium sulfate <1

Monoammonium phosphate <0.4

Fluosilicic acid <1

Sulfuric acid <1

Dissolved metallic impurities <1

Fluorides, as F <1 Table 5. Chemical Analysis Results of Phosphate Sample.

Sample P2O5 (%) Insol (%) MgO(%) Al2O3(%) Fe2O3(%) CaO(%)

Feed 27.35 13.08 1.06 0.81 0.88 41.45

+10 mesh 24.65 12.67 2.30 0.91 1.04 40.41

-10+20 26.96 12.27 1.23 0.81 0.92 41.93

-20+40 26.18 18.05 0.70 0.68 0.96 39.50

-40+60 27.87 13.93 0.66 0.68 0.78 41.31

-60+100 28.96 11.07 0.67 0.63 0.66 43.14

-100+200 29.21 11.25 0.61 0.59 0.64 42.75

-200 16.87 23.83 4.18 0.85 1.30 31.33 DETERMINATION OF POLARIZATION DIAGRAM FOR FUNDAMENTAL ASSESSMENT Polarization Diagram

The polarization diagram such as the one shown in Figure 4 gives a fundamental quantitative assessment of the decrease in corrosion rate caused by cathodic polarization. It can be established using a standard three-electrode electrochemical cell controlled by a

31

potentiostat which applies a sweeping potential to the working electrode made of the same metal as the mill or grinding medium (Tao and others 1994). The polarization potential of the working electrode was controlled in this study using an EG&G PARC potentiostat which applies to the working electrode a potential sweeping from -1.0 to 0.5 V at a rate of about 5 mV/min. A computerized data collection system automatically displays the polarization diagram and saves the data for analysis.

Polarization η is the potential E change from the equilibrium half-cell electrode

potential Eo, i.e., η = E - Eo. For iron dissolution process, Eo is -0.447 V (SHE). When electrons are supplied to the electrode surface, the electrode potential E becomes negative to Eo due to an electron buildup. This process is referred to as cathodic polarization. For anodic polarization, electrons are removed from the electrode, an electron deficiency results in a positive potential change due to the slow liberation of electrons by the surface reaction. There are two types of polarization: activation and concentration. Activation polarization occurs when a step in the electrode reaction controls the rate of electron transfer. Concentration polarization happens when the half-cell electrode potential E changes as the dissolved species being reduced or oxidized is depleted at the surface. For corrosion, concentration polarization is significant primarily for cathodic reduction processes. Concentration polarization for anodic oxidation during corrosion can usually be ignored due to an unlimited supply of metal atoms is available at the surface.

The relationship between activation polarization or overpotential η and the rate of

the reaction represented by current density is Equations (16) and (17) for anodic and cathodic polarization, respectively:

o

aaa i

ilogβη = (16)

o

ccc i

ilogβη = , (17)

where βa and βc are the Tafel constants for the half-cell reaction, and they are approximately 0.1 and -0.1 V respectively in most cases. Polarization Cell and Electrodes

The electrochemical cell used for polarization study is shown in Figure 12(a) and is illustrated in Figure 12(b). It has three electrodes: the working electrode (WE) made of mild steel or alloy, the counter electrode made of platinum wire, and the reference electrode. The working electrodes made of different metals are shown in Figure 13. The working electrode is centrally located in the cell with a pair of auxiliary electrodes on either side for better current distribution. The test specimen for the working electrode is embedded in an insulating epoxy resin in heavy-wall Pyrex tubing. The reference electrode, RE, is placed near the working electrode for potential measurement of the working electrode. Potentials were measured against a saturated calomel reference

32

electrode (SCE) and can be transformed to the standard hydrogen electrode (SHE) scale by adding 0.245 V.

(a) (b)

Figure 12. Photograph (a) and Illustration (b) of Polarization Cell.

The counter electrode (CE) made of platinum wire, which is inert to the electrolyte even under strong anodic polarization, was used to minimize the ohmic electrolyte resistance between the working electrode and the reference electrode. The ohmic resistance polarization is undesirable because it masks the other components of overpotential, which are of interest to determine corrosion rate and mechanism.

The composition of the buffer solutions used in the present study is given below: pH 4.6: 0.5 M CH3COOH and 0.5 M CH3COONa pH 6.8: 0.05 M KH2PO4 and 0.0224 M NaOH pH 8.0: 0.05 M KH2PO4 and 0.0461 M NaOH pH 9.2: 0.05 M Na2B4O7

All chemicals were of reagent grade and were acquired from Sigma-Aldrich Corp,

St. Louis, MO. Metal specimens were obtained from McMaster-Carr Supply Company, Cleveland, OH.

The working electrode was polished with 600 grit silicon carbide paper prior to

each test and cleaned with acetone, hydrochloric acid and double distilled water. Solutions were bubbled with nitrogen for at least 10 minutes before experiments were started to remove dissolved gas which will interfere with the electrochemical reactions of the working electrode. Nitrogen was kept flowing over the solution surface during experiments to prevent the diffusion of air into the cell. The polarization potential of the working electrode was controlled in the study using an EG&G PARC potentiostat which applies to the working electrode a potential sweeping from -1.0 to 0.5 V at a rate of about

REWE

CEREWE

CE

33

5 mV/min. A computerized data collection system automatically record and display the polarization diagram and save the data for analysis. Figure 14 shows the experimental setup, which included the electrochemical cell, the potentiostat, and the computer system.

Figure 13. Photograph of the Working Electrodes.

Figure 14. Photograph of the Electrochemical System.

34

Polarization Study

This task was performed to study the effects of solution pH, electrode material, environment gas, and solution composition on polarization curves.

Figure 15 shows the polarization curve for 1080 steel in 1 N H2SO4 solution. It

can be determined that βc is about -0.1 V. The corrosion current Icorr is about 1.2 mA/cm2. During the cathodic polarization the cathodic current was only slightly greater than the anodic current at low cathodic overpotentials. As the cathodic potential increased the cathodic current increased whereas the anodic current decreased. The change was quite sharp and soon the anodic current was negligible compared to the cathodic current. The linearity or the Tafel behavior of the cathodic polarization was limited only to the range of one order of magnitude. Nonlinearity was apparent at current densities greater than 40 mA/cm2. This may be caused by near surface depletion of the oxidizer H+ (concentration polarization) or ohmic resistance in the solution. Higher electrolyte concentration reduces concentration polarization, increases conductivity, and thus extend Tafel behavior.

Polarization in the anodic (positive) direction is analogous to cathodic

polarization. Electrons were drawn out of the working electrode and the current flowed in the opposite direction compared to the cathodic polarization. The deficiency of electrons made the potential change positive with respect to Eo. The anodic current was increased and the cathodic current was decreased and the difference rapidly increased as the potential increased. The anodic polarization curve did not show a linear region. It is not well understood why the linearity did not occur. However, it has been suggested that anodic dissolution of metal is irreversible in dilute corroding solutions. Rapid anodic dissolution can cause unacceptable solution contamination before the anodic polarization curve is completed. The surface may be roughened or otherwise changed as liberated corrosion products accumulate and precipitate or form oxide/hydroxide films on the surface. In weakly or moderately corrosive solutions, the anodic overpotential is often higher than would be expected from cathodic data, possibly due to formation of inhibiting surface films.

The effects of solution pH on cathodic polarization are shown in Figure 16. As

pH was increased, concentration polarization began much sooner and the limiting current density decreased quickly. In the higher pH solutions hydrogen evolution by reduction of water (reaction 8) occurs above the limiting current density for hydrogen reduction reaction (2): ).(22 2 reactioncathodicHeH =+ −+ (2) ),(222 22 solutionsalkalineandneutralOHHeOH −− +=+ (8) It is very obvious from Figure 16 that solution pH has very significant effects on metal corrosion rate. More acidic pH increases cathodic reaction rate, which in turn increases anodic metal corrosion rate.

35

Figure 17 shows the anodic dissolution curve of iron, 18 Cr stainless steel (type 430), and 18Cr-8Ni stainless steel (type 304L) in 1 N H2SO4 solution. Iron is difficult to passivate in acidic solutions due to its large critical anodic density and relatively noble primary passive potential. It is obvious that a limiting oxygen diffusion current density of 0.1 mA/cm2 is not sufficient to cause its passivation. Figure 17 indicates that the addition of chromium to iron increases the ease of passivation by reducing critical anodic current density. Although the critical anodic current density of this alloy is too high to be passivated by dissolved oxygen, it is more easily passivated by oxidizing agents than is pure iron. Figure 17 also suggests that the addition of both chromium and nickel to iron markedly increases the ease of passivation. This steel is widely used in the chemical industry. The critical current density of about 0.1 mA/cm2 indicates that it will be passivated by dissolved oxygen in acid solutions.

36

0.1 1 10 100

Current Density (mA/cm2)

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1Po

tent

ial (

V, S

HE)

10-1.0 100.0 101.0 102.02 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 7

-0

-0

-0

-0

-0

-0

-0

-0

Anodic

Cathodic

Figure 15. Polarization Curves for 1080 Steel in 1 N H2SO4.

37

0.0001 0.001 0.01 0.1 1 10


-1.1

-0.9

-0.7

-0.5

-0.3

-0.1

0.1Po

tent

ial (

V, S

HE)

10-4.0000 10-3.0000 10-2.0000 10-1.0000 100.0000 101.00002 3 4 2 3 4 2 3 4 2 3 4 2 3 4

-1.1

-0.9

-0.7

-0.5

-0.3

-0.1

0.1

pH 1.5pH 2.0pH 4.0

Figure 16. Effect of Solution pH on Cathodic Polarization of Iron.

38

0.0001 0.001 0.01 0.1 1 10 100 1000


-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4Po

tent

ial (

V, S

HE)

0.0001 0.001 0.01 0.1 1 10 100 1000

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Iron18Cr SS18Cr-8Ni SS

Figure 17. Potentiostatic Anodic Polarization Curve of Iron, 18Cr SS, and 18Cr-8Ni

SS in 1 N H2SO4.

Figure 18 shows potentiodynamic anodic polarization curves of 1018 mild steel specimen in nitrogenated buffer solutions at pH 4.6, 6.8, and 9.2, respectively. The buffer solution was bubbled with nitrogen for 10 minutes prior to each experiment. As can be seen, current density decreases with increasing solution pH. For example, the maximum current density is about 30 mA/cm2 at pH 4.6, which is about 300 times greater than the maximum current density of 0.1 mA/cm2 at pH 9.2. This is mainly because iron hydroxide is more readily formed at higher pH as a result of surface oxidation, which passivates metal surface. The specimen showed significant passivation at potential of about 0.3 V. However, the current density of passivated surface at pH 4.6 is one order of

39

magnitude greater than that at pH 9.2. The more interesting characteristic of the polarization curves is that potential has more significant effects on current density at lower pH, which suggests that the cathodic protection technique to be developed in this program has more significant impacts on corrosion rate at acidic pHs.

In order to evaluate the potentiodynamic anodic polarization characteristic under

different conditions, additional potentiodynamic polarization studies were also carried out in oxygenated buffer solutions. Figure 19 illustrates potentiodynamic anodic polarization curves of 1018 mild steel specimen in oxygenated buffer solutions with pH value of 4.6, 6.8, and 9.2, respectively. The buffer solution was bubbled with oxygen for 10 minutes prior to each experiment. It is clear that current density at acidic pH 4.6 is much greater than that at other pHs. For example, the current density is about 20 mA/cm2 at pH 4.6, which is about 2000 times greater than the current density of 0.01 mA/cm2 at 0.5 V in pH 9.2 solution. This is mainly because iron hydroxide is more readily formed in oxygenated solution at higher pH as a result of surface oxidation, which passivates metal surface. The specimen showed significant passivation at potential of about 0.3 V. However, the current density of passivated surface at pH 4.6 is about 20 times greater than that at pH 9.2. These results show that potential has more significant effects on current density at lower pH, which again indicates that the cathodic protection technique to be developed in this program has more significant impacts on corrosion rate at acidic pHs.

Comparing Figure 18 with Figure 19 indicates that the current density is higher in

nitrogenated buffer solution than that in oxygenated buffer solution at the same pH value. For example, the maximum current density is about 0.1 mA/cm2 in nitrogenated buffer solution at pH 9.2, which is about 10 times greater than the maximum current density of 0.01 mA/cm2 in oxygenated buffer solution at pH 9.2. This is mainly because ferric oxide is more readily formed in oxygenated solution than in nitrogenated solution as a result of surface oxidation, which passivates metal surface.

In order to evaluate the potentiodynamic cathodic polarization characteristics of

different materials, additional potentiodynamic polarization studies were also carried out with high-chromium alloy electrodes to compare their polarization characteristics with 1018 mild steel. Figure 20 shows potentiodynamic anodic polarization curves of high-chromium alloy specimen in nitrogenated buffer solutions with pH value of 4.6, 6.8, and 9.2, respectively. The result indicates that current density decreases with increasing solution pH. For example, the limiting current density is about 10 mA/cm2 at pH 4.6, which is about 100 times greater than that of 0.1 mA/cm2 at pH 9.2. The specimen showed significant passivation at potential of about 0.3 V. However, the current density of passivated surface at pH 4.6 is about 0.1 mA/cm2, which is about 3 times greater than the current density of passivated surface at pH 6.8.

Comparing Figure 20 with Figure 18 suggests that the current density of 1018

carbon steel is higher than that of high-chromium alloy in nitrogenated buffer solution at the same pH value. For example, the limiting current density is about 20 mA/cm2 of 1018 carbon steel at pH 4.6, which is about twice as great as the limiting current density

40

of 10 mA/cm2 of high-chromium alloy at pH 4.6. This is because the addition of chromium to iron increases the ease of passivation by reducing critical anodic current density.

The following potentiodynamic cathodic polarization tests of high-chromium

alloy electrodes were carried out in oxygenated buffer solutions. Figure 21 shows potentiodynamic anodic polarization curves of high-chromium alloy specimen in oxygenated buffer solutions at pH 4.6, 6.8, and 9.2, respectively. It is clear that current density decreases with increasing solution pH. For example, the maximum current density is about 9 mA/cm2 at pH 4.6, which is about 450 times greater than the maximum current density of 0.02 mA/cm2 at pH 6.8. This is mainly because electrode surface is more easily oxidized at higher pH solution and iron hydroxide was produced, which passivates metal surface.

Comparing Figure 21 with Figure 20 indicates that the current density is higher in

nitrogenated buffer solution than in oxygenated buffer solution at the same pH value. For example, the maximum current density is about 0.2 mA/cm2 in nitrogenated buffer solution at pH 6.8, which is about 10 times greater than the maximum current density of 0.02 mA/cm2 in oxygenated buffer solution at pH 6.8. This is because the fresh surface of electrode is easily oxidized in oxygenated solution than in nitrogenated solution. The result of this oxidation is that ferric oxide is formed, which passivates metal surface.

41

0.01 0.1 1 10 100 1000 10000 100000

Current density (µA/cm2)

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

Pote

ntia

l (V

, SH

E)

pH 6.8

pH 4.6

pH 9.2

Figure 18. Potentiodynamic Anodic Polarization Curve of 1018 Steel in Nitrogen-

ated Buffer Solutions.

42

0.001 0.01 0.1 1 10 100 1000 10000


-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

Pote

ntia

l (V

, SH

E)

pH 4.6pH 6.8

pH 9.2

Figure 19. Potentiodynamic Anodic Polarization Curve of 1018 Steel in Oxygenated

Buffer Solutions.

43

0.01 0.1 1 10 100 1000 10000 100000


-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Pote

ntia

l (V

, SH

E)

pH 4.6

pH 6.8

pH 9.2

Figure 20. Potentiodynamic Anodic Polarization Curve of High-Chromium Alloy in

Nitrogenated Buffer Solutions.

44

0.001 0.01 0.1 1 10 100 1000 10000


-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

1.2

Pote

ntia

l (V

, SH

E)

pH 4.6pH 6.8

pH 9.2


Oxygenated Buffer Solutions.

Since phosphate grinding takes place in solutions which is a mixture of gypsum pond water and fresh water, polarization studies were also carried out in simulation solution made of gypsum pond water and tape water. The pH of undiluted gypsum pond water was measured to be 1.31. The pond water and sodium hydroxide (NaOH) were used to make solutions of pH 4.6, 6.8 and 9.2. The potentiodynamic anodic polarization curves of 1018 steel specimen in nitrogenated pond water solutions at these pHs are shown in Figure 22. It is very obvious that metal corrodes much faster in lower pH solution than in higher pH solution, as indicated by higher current density. The limiting current density is about 3 mA/cm2 at pH 4.6, which is about 3 times greater than the limiting current density of 1.0 mA/cm2 at pH 6.8. This conclusion is in good agreement with that obtained with the buffer solutions. If potential is reduced to about -0.3 V,

45

current density can be reduced to 0.01 mA/cm2 from the activation limiting current of 20 mA/cm2.

Additional potentiodynamic anodic polarization studies of 1018 steel were also

carried out in oxygenated pond water solution. Figure 23 shows potentiodynamic anodic polarization curves of 1018 steel specimen in oxygenated pond water solutions with pH value of 4.6, 6.8, and 9.2, respectively. It is clear that metal corrodes much faster in lower pH solution than in higher pH solution, as indicated by higher current density. The limiting current density is about 2 mA/cm2 at pH 4.6, which is about twice as high as the limiting current density of 1.0 mA/cm2 at pH 6.8. This conclusion is in agreement with that obtained with the buffer solutions.

Comparing Figure 23 with Figure 22 reveals that the current density is higher in

nitrogenated pond water solution than that in oxygenated pond water solution at the same pH value. For example, the maximum current density is about 3 mA/cm2 in nitrogenated pond water solution at pH 4.6, which is about 1.5 times greater than the maximum current density of 2 mA/cm2 in oxygenated pond water solution at pH 4.6. This is mainly because high-chromium alloy is more easily passivated by oxygen in oxygenated pond water solution.

In order to evaluate the corrosion rate of different materials in pond water

solutions, additional potentiodynamic anodic polarization studies were also carried out with high-chromium alloy electrodes to compare their corrosion characteristics with 1018 mild steel in pond water solutions. Figure 24 shows potentiodynamic anodic polarization curves of high-chromium alloy specimen in nitrogenated pond water solutions at pH 4.6, 6.8, and 9.2, respectively. As can be seen, limiting current density decreases with increasing pond water solution pH. For example, the limiting current density is about 2.5 mA/cm2 at pH 4.6, which is about twice as high as the limiting current density of 1.2 mA/cm2 at pH 6.8.

Comparing Figure 24 with Figure 20 indicates that the results obtained with

buffer solutions are in agreement with those obtained with pond water solution, i.e., current density decreases with increasing solution pH. But the current density in buffer solution is higher than that in pond water solution at the same pH value. This is because the pH value can remain the same in buffer solution, but will increase in pond water solution with time due to formation of OH– in reactions (8) and (10).

Additional potentiodynamic anodic polarization studies of high-chromium alloy

electrodes were also carried out in oxygenated pond water solutions. Figure 25 shows potentiodynamic cathodic polarization curves of high-chromium alloy specimen in oxygenated pond water solutions with pH value of 4.6, 6.8, and 9.2, respectively. It is clear that the limiting current density decreases with increasing pond water solution pH. For example, the limiting current density is about 2.0 mA/cm2 at pH 4.6, while the limiting current density is 1.0 mA/cm2 at pH 6.8. This conclusion is in good agreement with that obtained with the buffer solutions.

46

Comparing Figure 25 with Figure 24 reveals that the limiting current density is higher in nitrogenated pond water solution than in oxygenated pond water solution at the same pH value. For example, the limiting current density is about 2.5 mA/cm2 in nitrogenated pond water solution at pH 4.6, while the limiting current density is only 2.0 mA/cm2 in oxygenated pond water solution at pH 4.6. This is believed to result from easier passivation by oxygen in oxygenated pond water solution.

47

0.001 0.01 0.1 1 10 100 1000 10000


-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

1.2

Pote

ntia

l (V

, SH

E)

pH 4.6

pH 6.8

pH 9.2

Figure 22. Potentiodynamic Anodic Polarization Curve of 1018 Steel in Nitrogen-

ated Pond Water Solutions.

48

0.001 0.01 0.1 1 10 100 1000 10000


-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

Pote

ntia

l (V

, SH

E)

pH 4.6

pH 6.8

pH 9.2

Figure 23. Potentiodynamic Anodic Polarization Curve of 1018 Steel in Oxygenated

Pond Water Solutions.

49

0.001 0.01 0.1 1 10 100 1000 10000


-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

1.2

Pote

ntia

l (V

, SH

E)

pH 4.6

pH 6.8

pH 9.2


Nitrogenated Pond Water Solutions.

50

0.001 0.01 0.1 1 10 100 1000 10000


-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

1.2

Pote

ntia

l (V

, SH

E)

pH 4.6pH 6.8

pH 9.2


Oxygenated Pond Water Solutions.

In order to investigate the potentiodynamic polarization characteristics, cathodic polarization studies were carried out under different conditions. Figure 26 shows potentiodynamic cathodic polarization curves of mild steel 1018 specimen in oxygenated buffer solutions at pH 4.6, 6.8, and 9.2, respectively. The buffer solution was bubbled with oxygen for 10 minutes prior to each experiment. It is very obvious from Figure 26 that solution pH has very significant effects on metal corrosion rate. More acidic pH increases cathodic reaction rate, which in turn increases anodic metal corrosion rate. As can be seen, current density decreases with increasing solution pH. For example, the maximum current density was about 42 mA/cm2 at pH 4.6, which is about 14 times greater than the maximum current density of 3 mA/cm2 at pH 6.8. It is interesting to note that the potential has more significant effects on current density at lower pH, which

51

suggests that the cathodic protection technique has more significant impacts on corrosion rate at acidic pHs.

In order to evaluate the potentiodynamic polarization characteristics under

different conditions, additional polarization studies were also carried out in nitrogenated buffer solutions. Figure 27 shows potentiodynamic cathodic polarization curves of 1018 mild steel specimen in nitrogenated buffer solutions at pH 4.6, 6.8, and 9.2, respectively. The buffer solution was bubbled with nitrogen for 10 minutes before each experiment was started to remove dissolved gas which interferes with electrochemical reactions of the working electrode. As can be seen, current density decreases with increasing solution pH. For example, the limiting current density is about 65 mA/cm2 at pH 4.6, while the limiting current density is 4.3 mA/cm2 at pH 6.8.

Chromium by itself is very corrosion resistant. It is therefore used as a coating for

base metals to maintain their integrity. Of even greater importance is the use of chromium in alloys. It is the key component of stainless steel and widely used in the chemical industry. In order to evaluate the potentiodynamic cathodic polarization characteristics of different materials, potentiodynamic polarization studies were also carried out with high-chromium alloy electrodes to compare their polarization characteristics with mild steel. Figure 28 shows potentiodynamic cathodic polarization curves of high-chromium alloy specimen in oxygenated buffer solution with pH value of 4.6, 6.8, and 9.2, respectively. As can be seen, current density is much higher at acidic pH than neutral or alkaline pHs and higher pHs give consistently lower current density.

The potentiodynamic polarization characteristics of chromium alloy electrodes

were also carried out in nitrogenated buffer solutions. Figure 29 shows potentiodynamic cathodic polarization curves of high-chromium alloy specimen in nitrogenated buffer solutions at pH 4.6, 6.8, and 9.2, respectively. It is clear that current density decreased with increasing solution pH. For example, the maximum current density was about 22.5 mA/cm2 at pH 4.6, which is about 7 times greater than the maximum current density of 3.2 mA/cm2 at pH 6.8. Similar results have also been obtained for 1018 carbon steel, except that the current density was higher for 1018 carbon steel than the chromium alloy.

Comparing Figure 26 with Figure 27 indicates that the current density was higher

in nitrogenated buffer solution than in oxygenated buffer solution at the same pH value. For example, the maximum current density was about 65 mA/cm2 in nitrogenated buffer solution at pH 4.6, which is about 1.5 times greater than the maximum current density of 42000 µA/cm2 in oxygenated buffer solution at pH 4.6. This is mainly due to the fact that ferric oxide is more readily formed in oxygenated solution than in nitrogenated solution as a result of surface oxidation, which passivates metal surface. Similar results were obtained from comparing Figure 28 with Figure 29.

Comparing Figures 28, 29 with Figures 26 and 27 indicates that the current

density was higher for 1018 mild steel electrode than that of high-chromium alloy electrode. This is mainly because the addition of chromium to iron in an alloy increased the ease of passivation by reducing critical cathodic current density. Although the critical

52

cathodic current density of this alloy is too high to be passivated by dissolved oxygen, it is more easily passivated by oxidizing agents than pure iron.

0.01 0.1 1 10 100 1000 10000 100000Current density (µA/cm2)

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0Po

tent

ial (

V, S

HE

)

pH 9.2

pH 6.8pH 4.6

Figure 26. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in Oxygen-


53

0.01 0.1 1 10 100 1000 10000 100000


-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 9.2

pH 6.8

pH 4.6

Figure 27. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in Nitrogen-


54

0.001 0.01 0.1 1 10 100 1000 10000 100000


-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 9.2

pH 6.8pH 4.6

Figure 28. Potentiodynamic Cathodic Polarization Curve of High-Chromium Alloy

in Oxygenated Buffer Solutions.

55

0.01 0.1 1 10 100 1000 10000 100000


-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 9.2

pH 6.8pH 4.6


in Nitrogenated Buffer Solutions.

Since phosphate grinding takes place in solution mixture of gypsum pond water and fresh water, polarization studies were also carried out in simulation solution made of gypsum pond water and distilled water. The pH of undiluted gypsum pond water has been measured to be 1.31. The pond water and NaOH were used to make solutions of pH 4.6, 6.8, and 9.2. The potentiodynamic cathodic polarization curves of 1018 steel specimen in oxygenated pond water solutions at these pHs are shown in Figure 30. The pond water solution was bubbled with nitrogenated for 10 minutes prior to each experiment. It is obvious that metal corroded much faster in lower pH solution than in higher pH solution, as indicated by higher current density. The limiting current density was about 20 mA/cm2 at pH 4.6, which is about 12 times greater than the limiting current

56

density of 1.6 mA/cm2 at pH 6.8. This conclusion is in agreement with that obtained with the buffer solutions.

Additional potentiodynamic polarization studies of 1018 steel were also carried

out in nitrogenated pond water solution. Figure 31 shows potentiodynamic cathodic polarization curves of 1018 steel specimen in nitrogenated pond water solutions with pH value of 4.6, 6.8, and 9.2, respectively. It is clear that metal corrodes much faster in lower pH solution than in higher solution, as indicated by higher current density. The limiting current density was about 35 mA/cm2 at pH 4.6, which is about 17 times greater than the limiting current density of 2.04 mA/cm2 at pH 6.8. Similar results were obtained with the buffer solutions.

In order to evaluate the corrosion rate of different materials in pond water

solutions, potentiodynamic cathodic polarization studies were carried out with high-chromium alloy electrodes to compare their corrosion characteristics with those of mild steel in pond water solutions. Figure 32 shows potentiodynamic cathodic polarization curves of high-chromium alloy specimen in oxygenated pond water solutions with pH value of 4.6, 6.8, and 9.2, respectively. As can be seen, limiting current density decreased with increasing pond water solution pH. For example, the limiting current density is about 6.2 mA/cm2 at pH 4.6, which is about six times greater than the limiting current density of 1.0 mA/cm2 at pH 6.8.

Potentiodynamic cathodic polarization tests of high-chromium alloy electrodes

were also carried out in nitrogenated pond water solutions. Figure 33 shows potentiodynamic cathodic polarization curves of chromium alloy specimen in nitrogenated pond water solutions with pH value of 4.6, 6.8, and 9.2, respectively. As can be seen, limiting current density decreased with increasing pond water solution pH. For example, the limiting current density was about 9.5 mA/cm2 at pH 4.6, while the limiting current density was 1.5 mA/cm2 at pH 6.8. This conclusion is in agreement with that obtained with the buffer solutions.

Comparing Figure 32 with Figure 33 reveals that the limiting current density is

higher in nitrogenated pond water solution than in oxygenated pond water solution at the same pH value. For example, the limiting current density was about 9.5 mA/cm2 in nitrogenated pond solution at pH 4.6, while the limiting current density is only 6.2 mA/cm2 in oxygenated pond water solution at pH 4.6. This is believed to be result from easier passivation by oxygen in oxygenated water solution. Similar results were obtained with comparing Figure 29 with Figure 30.

Comparing Figures 32, 33 with Figures 30 and 31 indicates that the current

density was higher for 1018 mild steel electrode than that of high-chromium alloy electrode. This is because the addition of chromium to iron in an alloy increased the ease of passivation by reducing critical cathodic current density. This conclusion is in agreement with that obtained with the buffer solution.

57

Comparing Figures 30, 31, 32, 33 with Figures 26, 27, 28, and 29 reveals that the current density was higher in buffer solution than that in pond water solution under same operating conditions. This is mainly because the buffer solution is a mixture of a weak acid or weak base and its salt. The pH of a buffer solution resists change when small amounts of acid or base are added. But the pH of a pond water solution will increase as the reaction proceeds. The corrosion rate decreased with increasing the pH value of pond water solution.

58

0.01 0.1 1 10 100 1000 10000 100000


-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 6.8

pH 4.6

pH 9.2

Figure 30. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in Oxygen-


59

0.01 0.1 1 10 100 1000 10000 100000


-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 9.2

pH 4.6

pH 6.8

Figure 31. Potentiodynamic Cathodic Polarization Curve of 1018 Steel in Nitrogen-


60

0.01 0.1 1 10 100 1000 10000 100000Current density (µA/cm2)

-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 9.2

pH 6.8pH 4.6


in Oxygenated Pond Water Solutions.

61

0.001 0.01 0.1 1 10 100 1000 10000100000


-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

Pote

ntia

l (V

, SH

E)

pH 4.6

pH 6.8

pH 9.2


in Nitrogenated Pond Water Solutions. CHARACTERIZATION OF CORROSION RATE UNDER VARIOUS OPERATING CONDITIONS

The purpose of this task was to characterize the total wear, corrosive wear, and abrasive wear of the ball mill under various operating conditions. The wear rate of coupons was measured in mils (1mils = 0.001 in.) penetration per year (MPY). The operating parameters studied included solution pH, grinding time, rotation speed, feed solids percentages, mill crop load, etc. The critical speed for the ball mill is 95 RPM. The grinding time was fixed at three hours and feed solids at 64%, unless specified otherwise. A statistical Box-Behnken design (BBD) of experiments was performed to evaluate individual variables and their interactions. The levels of each parameter studied

62

are shown in Table 6. The most significant variables and optimum conditions were determined from statistical analysis of the experimental results using response surface methodology (RSM). The BBD design and subsequent RSM analysis were conducted using a sophisticated software Design-Expert 6.8® acquired from Stat-Ease Inc., Minneapolis, MN.

The total wear rate of metal (W) determined from the weight loss measurement

has three components, as shown in Equation (18): ,SCA WWWW ++= (18) where WA represents abrasive wear, WC corrosive wear, and WS stands for the synergistic effect of abrasive and corrosive wear. Abrasive wear (WA) is assumed to be the weight loss obtained when the coupon was cathodically protected with large negative potential (e.g., -1.0 V) or large impressed current (e.g., 0.1 mA/m2). Corrosive wear (WC) was determined experimentally using an electrochemical cell in the same slurry as used in the grinding tests (Jang and others 1989; Rajagopal and Iwasaki, 1992; Pazhianur and others 1997). The synergistic effect (WS) was then calculated from Equation (18). The relative importance of individual components and the contribution of corrosive wear can thus be determined for different conditions.

The coupons used to represent the grinding balls and mill shell were made of

1018 mild steel and high-chromium alloy to represent different grinding media which are known to show different corrosion and abrasion behavior. The coupon assembly contains the metal coupons in the center and the insulator around it. Three coupons were inserted and set flush with the interior surface of the mill. Total wear rate was estimated from the weight loss of the coupons after grinding. The effects of individual variables on the metal loss and their interactions were investigated. Table 6. Test Levels of Parameters for BBD.

Levels Factor Units Coded

parameter -1.0 0.0 +1.0

Solution pH A 3 7 9

Rotation speed RPM B 60 70 80

Crop load % C 40 50 60

Solid concentration % D 54 64 74

63

Parametric Study of Corrosion Rate

The effects of individual variables on corrosion rate were investigated in this subtask. The operating parameters studied included grinding time, solution pH, rotation speed, feed solids percentage, and mill crop load. Effect of Grinding Time

The purpose of this study was to investigate the effects of grinding time on the corrosion rate. The solution pH was fixed at 6.8, rotation speed at 70 RPM, crop load at 60%, and the feed solids percentage at 64% in this study. Grinding time was varied at 0.5, 1, 2, 3, and 4 hours, respectively. Figure 34 clearly shows that there is a linear relationship between the weight loss and the grinding time. The longer the grinding time, the more weight loss. After 0.5 hour, the weight loss was 0.00043 g. However, the weight loss increased to 0.0032 g after four hours. The wear rate was determined to be 0.00082 g/hr or 167.8 MPY. The wear rate in mils penetration per year (MPY) was calculated from Equation (19):

,534DAT

WMPY = (19)

where W is weight loss in milligrams, D is density in grams per cubic centimeter, A is area in square inches, and T is time in hours (Jones 1996). Effect of Solution pH

The purpose of this subtask was to determine the effects of solution pH on wear rate. Since there is a linear relationship between the grinding time and the weight loss, grinding time was kept at three hours to observe a significant weight loss. Rotation speed was fixed at 70 RPM, crop load at 60% and the feed solids percentage at 64%, respectively. The wear rate of ball mill was measured as a function of solution pH and is given in Figure 35. It is obvious that wear rate decreased with increasing solution pH. For example, the wear rate was about 678.8 MPY at pH 3.0, which is about twice greater than the wear rate of 347.6 MPY at pH 10.0. The effect of pH on the corrosion rate in aerated water is related to the cathodic reduction reaction. A ferrous oxide deposit is known to form on the metal surface. Ferrous oxide is an insulator that prevents electron transfer, hindering corrosion. In more acidic solutions, the oxide is soluble. Furthermore, corrosion increases due to increased availability of H+ for reduction by reaction (2). The absence of the surface deposit also enhances access of dissolved oxygen, which further increases corrosion rate through the cathodic reaction (9). At pH above 10, corrosion rate was relatively low, due to formation of the passive ferric oxide film in the presence of dissolved oxygen. These results are in excellent agreement with those reported by Muller and Gampper (1994).

64

Effect of Rotation Speed

The speed of rotation necessary to achieve most grinding for least power consumption has been considered to be that which generates enough centrifugal force to make the rods lift up on the shell lining to a height of about two-thirds of the diameter of the shell with least sliding and cascading (Lowrison 1974). The economic value of speed of rotation for tumbling mills is about 75% of the critical speed. The critical speed in terms of revolutions per minute is if the radii are expressed in centimeters (Lowrison 1974). The radii of grinding balls used in this research are ½’’, 3/8’’, ¼’’ and 1/8’’, respectively. The internal radius of the mill used in this study is 4.25’’ and therefore,

)(21.9727.125.454.2

3001 RPMNc =

−×= (21)

)(62.959525.025.454.2

3002 RPMNc =

−×= (22)

)(12.94635.025.454.2

3003 RPMNc =

−×= (23)

)(68.923175.025.454.2

3004 RPMNc =

−×= (24)

So the average critical speed is

4

4321 ccccc

NNNNN +++=

)(91.944

68.9212.9462.9521.97 RPM=+++

= (25)

In this study, grinding time was fixed at three hours, solution pH at 6.8, crop load

at 60% and the feed solids percentage at 64%, respectively. Rotation speed was varied at 30, 50, 60, 70, and 80 RPM, respectively. The effects of rotation speed on wear rate are shown in Figure 36. It is clear that when the rotation speed is less than 50 RPM, the wear rate decreased with increasing rotation speed. This is mainly because at low rotation speed, balls were sliding against the wall of the mill. Abrasive wear mainly caused the weight loss. Increasing the rotation speed in this speed range changed the ball motion

300Nc = (20) (internal radius of mill – radius of media)1/2

65

from sliding to cascading and the corrosion type from abrasive wear to pitting corrosion. If the rotation speed was higher than 50 RPM, the wear rate increased with increasing rotation speed. This may be caused by the fact that the balls were lifted higher at high rotation speed and the balls had stronger impacts on the mill shell upon falling on to the coupon, which increased the abrasive wear. However, if the rotation speed was close to or higher than the critical speed, the wear rate was reduced significantly because the media were held against the shell by centrifugal force and no abrasive wear happened at this speed range. This result is in good agreement with that reported by Lowrison (1974). Effect of Mill Crop Load

The purpose of this study was to characterize the effects of mill crop load on the wear rate. Grinding time was fixed at three hours, pH at 7.0, rotation speed at 70 RPM, and the feed solids percentage at 64%. Mill crop load was varied from 40, 50, 60, to 70%. The wear rate of ball mill was measured as a function of mill crop load and results are given in Figure 37. It is obvious that wear rate decreased with increasing mill crop load. For example, the wear rate was about 593.3 MPY at mill crop load 40%, which is about 50% more than the wear rate of 409.3 MPY at mill crop load 70%. This behavior is attributed to the fact that increasing crop load reduced the motion space of the balls, reducing the contact chance between the balls and the mill shell. As a result, abrasive wear and fretting corrosion were reduced significantly. This result is consistent with earlier studies that showed sufficient loading of balls reduced the corrosion rate, prevented surging, gave maximum rate of production of new surface (Lowrison 1974). Effect of Feed Solids Percentage

Feed solids percentage is another important factor that affects metal loss in comminution operations. In this study, grinding time was fixed at three hours, pH value at 7.0, rotation speed at 70 RPM, and the mill crop load at 60%. Feed solids percentage was varied at 44%, 54%, 64%, 74%, and 84%. The effects of feed solids percentage on wear rate are given in Figure 38. It is clear that wear rate decreased with increasing feed solids percentage. For example, the wear rate was about 636.6 MPY at feed solids percentage 44%, which is about 3 times the wear rate of 225.4 MPY at feed solids percentage 84%. This is attributed to the fact that increasing feed solids percentage increased the pulp viscosity which affected the abrasion rate. At higher pulp viscosity, the motion of the balls and the phosphate samples were reduced significantly. If the pulp viscosity was too high, the balls and the phosphate samples would be stuck on the mill shell, significantly reducing the abrasion rate. Another reason is that at lower feed solids percentage, more oxygen will be dissolved in the pulp. Dissolved oxygen in aqueous solution has profound influence on the corrosion of metal since oxygen reduction is the dominant cathodic reaction that drives the anodic reaction. This conclusion is in good agreement with previous results reported by Oesch (1996).

66

0 1 2 3 4 5

Grinding Time (hour)

0.000

0.001

0.002

0.003

0.004W

eigh

t Los

s (g

)0 1 2 3 4 5

0.0

0.0

0.0

0.0

0.0

Figure 34. Effect of Grinding Time on Weight Loss.

67

2 3 4 5 6 7 8 9 10 11 12

pH

0

100

200

300

400

500

600

700

800W

ear R

ate

(MP

Y)

2 3 4 5 6 7 8 9 10 11 12

0

100

200

300

400

500

600

700

800

Figure 35. Effect of Solution pH on Wear Rate.

68

20 30 40 50 60 70 80 90 100

Rotation Speed (RPM)

0

100

200

300

400

500

600

700W

ear R

ate

(MPY

)20 30 40 50 60 70 80 90 100

0

100

200

300

400

500

600

700

Figure 36. Effect of Rotation Speed on Wear Rate.

69

30 40 50 60 70 80

Crop Load (%)

200

300

400

500

600

700W

ear R

ate

(MPY

)30 40 50 60 70 80

200

300

400

500

600

700

Figure 37. Effect of Mill Crop Load on Wear Rate.

70

30 40 50 60 70 80 90

Feed Solids Percentage (%)

0

100

200

300

400

500

600

700W

ear R

ate

(MP

Y)

30 40 50 60 70 80 90

0

100

200

300

400

500

600

700

Figure 38. Effect of Feed Solids Percentage on Wear Rate. Statistical Box-Behnken Design of Experiments

The Box-Behnken Design is an independent quadratic design that does not contain an embedded factorial or fractional factorial design. In this design the treatment combinations are at the midpoints of edges of the process space and at the center. The designs are rotatable (or near rotatable) and require three levels of each factor. The Box-Behnken Design has limited capability for orthogonal blocking compared to the central composite designs. However, it requires fewer treatment combinations than a central composite design in cases involving three or four factors.

71

Design-Expert 6.8 software offers multilevel factorial screening designs to identify the critical factors that can lead to breakthrough improvements. Design-Expert 6.8 provides in-depth analysis of process variables or mixture components to locate ideal process settings to achieve peak performance or discover the optimal product formulations. Design-Expert 6.8 offers rotatable 3D plots to help visualize the response surface. The 2D contours can be explored with setting flags along the way to identify coordinates and predict responses. The optimum spot where all the requirements are met can be found via the program's numerical optimization function, which finds the most desirable factor settings for up to 12 responses simultaneously. INTERACTIONS OF INDIVIDUAL VARIABLES Determination of Optimal Conditions

The purpose of this task was to characterize the interactions of individual variables on corrosion rate. The individual variables include solution pH, rotation speed, mill crop load, and solids percentage. Since wear rate increased linearly with grinding time, grinding time was fixed at three hours. The Box-Behnken Design (BBD) fractional factorial experiment was conducted to determine the interactions between different variables and the optimum condition. The fractional factorial design is particularly useful in screening a large number of variables and helps to identify the most significant parameters and optimum condition with a very small number of experiments (Box and others 1978; Montgomery 1991; Sung and Parekh 1996). Table 6 presents four operating variables examined in this study as well as their three levels. The factorial design and total wear rate in MPY unit for each experiment was determined by coupon weight loss and results are shown in Table 7.

72

Table 7. Factorial Design and Experimental Response for 1018 Carbon Steel.

Std Run Block Factor A: Solution

pH

Factor B: Rotation

speed (rpm)

Factor C: Crop load

(%)

Factor D: Solid

Concentration (%)

Response: Wear rate

(MPY)

23 1 Block 1 0.0 -1.0 0.0 +1.0 378.66 18 2 Block 1 +1.0 0.0 -1.0 0.0 471.08 6 3 Block 1 0.0 0.0 +1.0 -1.0 460.72 10 4 Block 1 +1.0 0.0 0.0 -1.0 412.63 17 5 Block 1 -1.0 0.0 -1.0 0.0 777.93 3 6 Block 1 -1.0 +1.0 0.0 0.0 920.52 8 7 Block 1 0.0 0.0 +1.0 +1.0 363.52 27 8 Block 1 0.0 0.0 0.0 0.0 430.34 2 9 Block 1 +1.0 -1.0 0.0 0.0 327.21 5 10 Block 1 0.0 0.0 -1.0 -1.0 548.24 12 11 Block 1 +1.0 0.0 0.0 +1.0 287.37 25 12 Block 1 0.0 0.0 0.0 0.0 450.64 19 13 Block 1 -1.0 0.0 +1.0 0.0 593.31 1 14 Block 1 -1.0 -1.0 0.0 0.0 654.42 29 15 Block 1 0.0 0.0 0.0 0.0 446.31 13 16 Block 1 0.0 -1.0 -1.0 0.0 409.97 22 17 Block 1 0.0 +1.0 0.0 -1.0 681.85 15 18 Block 1 0.0 -1.0 +1.0 0.0 401.16 9 19 Block 1 -1.0 0.0 0.0 -1.0 855.45 16 20 Block 1 0.0 +1.0 +1.0 0.0 511.83 20 21 Block 1 +1.0 0.0 +1.0 0.0 388.33 7 22 Block 1 0.0 0.0 -1.0 +1.0 470.22 11 23 Block 1 -1.0 0.0 0.0 +1.0 601.99 24 24 Block 1 0.0 +1.0 0.0 +1.0 476.84 4 25 Block 1 +1.0 +1.0 0.0 0.0 450.71 26 26 Block 1 0.0 0.0 0.0 0.0 467.01 21 27 Block 1 0.0 -1.0 0.0 -1.0 538.08 28 28 Block 1 0.0 0.0 0.0 0.0 458.21 14 29 Block 1 0.0 +1.0 -1.0 0.0 696.44

73

Table 8. Analysis of Variance for 1018 Carbon Steel.

Source Sum of Squares

Degree of Freedom

Mean Square F Value Prob>F

Model 6.358E+05 14 45413.44 28.39 <0.0001

A 3.558E+05 1 3.558E+05 222.40 <0.0001

B 92991.90 1 92991.90 58.13 <0.0001 C 38852.97 1 38852.97 24.29 0.0002 D 75143.16 1 75143.16 46.97 <0.0001 A2 6149.21 1 6149.21 3.84 0.0701 B2 17837.71 1 17837.71 11.15 0.0049 C2 422.72 1 422.72 0.26 0.6152 D2 227.73 1 227.73 0.14 0.7116 A×B 5822.93 1 5822.93 3.64 0.0771 A×C 3252.72 1 3252.72 2.03 0.1758 A×D 5405.92 1 5405.92 3.38 0.0873 B×C 7726.41 1 7726.41 4.83 0.0453 B×D 519.61 1 519.61 0.32 0.5778 C×D 91.97 1 91.97 0.057 0.8140

Residual 43679.33 22 1985.42

Lack of fit 42923.31 18 2384.63 12.62 0.0123

Pure error 756.03 4 189.01

Total 6.582E+05 28

74

Table 9. Analysis of Variance for Reduced Quadratic Model of 1018 Carbon Steel.


Degree of Freedom


Model 6.145E+05 6 1.024E+05 51.58 <0.0001

A 4.022E+05 1 4.022E+05 202.60 <0.0001

B 88183.59 1 88183.59 44.42 <0.0001 C 35753.18 1 35753.18 18.01 0.0003 D 7.283.62 1 70283.62 35.40 <0.0001 B2 15134.60 1 15134.60 7.62 0.0114 B×C 7726.41 1 7726.41 3.89 0.0612

Residual 43679.33 22 1985.42

Lack of fit 42923.31 18 2384.63 12.62 0.0123

Pure error 756.03 4 189.01

Total 6.582E+05 28

Design-Expert 6.8 program was used to analyze each response of above factors to the regression model using the following methodology: (a) analysis of variance (ANOVA) was conducted to determine the adequacy of linear, quadratic and cubic models; (b) one model was then chosen for an in-depth regression analysis; (c) diagnostic evaluation of the robustness of the model was determined; and (d) response surface analysis was conducted to optimize the wear rate.

The lack of fit tests and model summary statistics for the wear rate of carbon steel

showed that F values of 17.40, 11.45 and 9.07 were obtained for the linear, quadratic, and cubic models with corresponding probability values of 0.0066, 0.0156 and 0.0326, respectively. A large F value and a small probability value indicate the model validity. The predicted residual sum of square (PRESS), which indicates how well the model fits the data, was 1.017E+05 for the linear model, 1.225E+05 for the quadratic model, and 6.282E+05 for the cubic model. The adjusted R-squared was 0.8821 for linear model, 0.9319 for quadratic model, and 0.9703 for cubic model. Maximized adjusted R-squared should be focused on the model selection. For this design, the cubic model was aliased because of insufficient design points to estimate the coefficients. Thus, the quadratic model was chosen for this design.

The ANOVA calculations were conducted and results are shown in Table 6. It is

obvious based on the value of F that the solution pH has the most significant effect on the wear rate of 1018 carbon steel, followed by the rotation speed. At a given crop load,

75

higher solution pH, lower rotation speed, higher solid percentage tend to produce significantly lower wear rate. The effects of crop load on wear rate were not as significant as solution pH, rotation speed or solid percentage.

The analysis of interactions between different variables was carried out using the

same software. The interactions of each two variables affecting carbon steel wear rate while other variables were kept at the middle levels are shown in Figures 39 to 44, respectively. As shown in these figures, the interactions of higher rotation speed and lower crop load had the most significant effect on the wear rate of 1018 carbon steel, followed by the interactions of lower solution pH and higher rotation speed. The interactions of higher crop load and lower solid percentage had the least significant effect on the carbon steel wear rate. As shown in Table 8, the influence of variable interactions on the wear rate was in the order of rotation speed × crop load > solution pH × rotation speed > solution pH × solid percentage > solution pH × crop load > rotation speed × solid percentage > crop load × solid percentage.

The analysis results indicated that solution pH, rotation speed, solid percentage,

crop load, rotation speed2, and rotation speed × crop load were significant model terms. The effects of other interactions on the wear rate were not significant. It is known that if there are many insignificant model terms (not counting those required to support hierarchy), model reduction will improve the model. Therefore, insignificant terms were excluded from the model. The ANOVA calculations for reduced quadratic model were conducted and results are shown in Table 9. The model for wear rate yielded an F value (i.e., the comparison of the treatment variance with the error variance) of 51.58 and a probability value of 0.01% (i.e., the probability that the model terms are null). Since large value of F (i.e., F>>1) for the model indicates that the error was relatively small for selecting the model term, these statistics indicate that the model was robust for the wear rate for 1018 mild carbon steel, yielding the regression equation for coded variables: Wear rate = 530.34 – 177.50×A + 85.72×B – 54.58×C – 76.53×D + 46.41×B2 – 43.95×B×C (26)

Diagnostic evaluation of the robustness of the model was determined with graphical means. The most important diagnostic is studentized residuals. This is illustrated in Figure 45 for wear rate in this model. Departures from a straight line show non-normality of the error term. The diagnosis of residuals did not reveal any statistical problems in the regression analysis (i.e., predicted is close to actual).

76

Figure 39. Response for pH and Rotation Speed at Crop Load 50% and Solid

Percentage 64%.

3 5 2 .0

4 8 7 .7

6 2 3 .5

7 5 9 .2

8 9 5 .0

Wea

r rat

e (M

PY

)

3 .04 .56 .07 .59 .0

6 0

6 5

7 0

7 5

8 0

p H

Rot

atio

n sp

eed

(RP

M)

77

Figure 40. Response for pH and Crop Load at Rotation Speed 70 RPM and Solid

Percentage 64%.

342.3

458.7

575.2

691.6

808.1W

ear r

ate

(MP

Y)

3 .04.56.07.59.0

4045

50

5560

pH

Cro

p lo

ad (%

)

78

Figure 41. Response for pH and Solid Percentage at Rotation Speed 70 RPM and

Crop Load 50%.

324.4

452.7

581.0

709.3

837.6 W

ear r

ate

(MP

Y)

3 .04.56.07.5 9.0

54 59

64

69 74

pH Sol

id p

erce

ntag

e (%

)

79

Figure 42. Response for Rotation Speed and Crop Load at pH 7.0 and Solid

Percentage 64%.

324.4

452.7

581.0

709.3

837.6W

ear r

ate

(MP

Y)

6065707580

4045

505560

R ota tion speed (RP M )

Cro

p lo

ad (%

)

80

Figure 43. Response for Rotation Speed and Solid Percentage at pH 7.0 and Crop

Load 50%.

324.4

452.7

581.0

709.3

837.6W

ear r

ate

(MP

Y)

6065707580

5459

6469

74

Rotation speed (RPM)So

lid p

erce

ntag

e (%

)

81

Figure 44. Response for Crop Load and Solid Percentage at pH 7.0 and Rotation

Speed 70 RPM.

324.4

452.7

581.0

709.3

837.6

Wea

r rat

e (M

PY

)

4045505560

5459

6469

74

Crop load (%) Solid

per

cent

age

(%)

82

Figure 45. Normal Probability Plot of Residual for 1018 Carbon Steel.

Studentized Residuals

Nor

mal

Pro

babi

lity

(%)

-1.55 -0.60 0.35 1.30 2.25

1

510

2030

50

7080

9095

99

83

Actual factors: crop load = 71.94, solid percentage = 75.50% Figure 46. Response Surface and Contours for Desirability Function for Carbon

Steel.

p H

Rot

atio

n sp

eed

(RPM

)

3 . 0 4 . 5 6 . 0 7 . 5 9 . 06 0

6 5

7 0

7 5

8 00 . 7 3 5

0 . 7 8 8

0 . 7 8 8

0 . 8 4 1

0 . 8 9 4

0 . 9 4 7

0 . 9 8 0

P r e d i c t i o n 1 . 0 0X 7 . 3 6Y 7 0 . 3 1

A c tu a l fa c to rs : c ro p lo a d = 7 1 .9 4 % , s o lid p e rc e n ta g e = 7 5 .5 0 %

0 .6 8 2

0 .7 6 2

0 .8 4 1

0 .9 2 1

1 .0 0 0

Des

irabi

lity

3 .04 .5

6 .07 .5

9 .06 0

6 5

7 0

7 5

8 0

p H

Rot

atio

n sp

eed

(RPM

)

84

The ball mill grinding tests have also been carried out at different conditions using high-chromium alloy as coupons to determine the most significant variables and optimum conditions from analysis of experimental results based on the same experimental design methodology as described earlier. The ranges of experimental parameters were the same as those for 1018 carbon steel. The four operating variables examined in this study and their three levels are shown in Table 6. The factorial design and the total wear rate for high-chromium alloy are shown in Table 10.

The lack of fit tests and model summary statistics for the wear rate of high-

chromium alloy indicated that F values of 18.46, 15.19 and 7.11 were obtained for the linear, quadratic, and cubic models with corresponding probability values of 0.0059, 0.0092 and 0.0482, respectively. As mentioned earlier, a large F value and a small probability value indicate the model validity. The PRESS was 10758.96 for the linear model, 17194.08 for the quadratic model, and 48428.36 for the cubic model. The adjusted R-squared was 0.8079 for linear model, 0.6930 for quadratic model, and 0.1354 for cubic model. For this design, the cubic model was aliased because of insufficient design points to estimate the coefficients. Since maximum adjusted R-squared should be focused on for the model selection, the linear regression model was chosen for the wear rate of high-chromium alloy.

The ANOVA calculations for high-chromium alloy were conducted and results

are shown in Table 11. It is clear that solution pH had the most significant effect on the wear rate of high-chromium alloy. At a given condition, higher solution pH tends to lower the wear rate of high-chromium alloy significantly. The rotation speed was the second most significant factor for the wear rate. However, the effects of crop load on wear rate were not as significant as that of solid percentage. These results were similar to those obtained with carbon steel.

The fact that linear regression model was chosen for this design indicates that the

interactions between different variables on the wear rate of high-chromium alloy were not significant. Therefore, variable interactions were excluded the model. As shown in Table 11, the model for wear rate of high-chromium alloy yielded an F value of 42.48 and a probability value of 0.01%. As we know that large value of F (i.e., F>>1) for the model indicates that the error was relatively small for selecting the model term, these statistics indicate that the model was robust for the wear rate of high-chromium alloy. The regression equation for coded variable was: Wear rate = 156.24 – 50.31×A + 24.81×B – 14.63×C – 23.75×D (27)

Diagnostic evaluation of the robustness of this model was also determined with graphical means. As described earlier, the most important diagnostic is studentized residuals, which is shown Figure 47. Departures from a straight line show non-normality of the error term. The diagnosis of residuals did not reveal any statistical problem in the regression analysis.

85

Numerical optimization was used to determine the optimum process parameters for minimum wear rate. In simultaneous numerical optimization, an objective function was set up which was zero outside of the limits and one at the goal. The desirability function, which gives some idea how well the goals were met was utilized to combine the wear rate response. In order to find the optimum conditions in a wide range, the solution pH range was changed to 1-12, rotation speed 50-90 RPM, crop load 30-80%, and solid percentage 44-84%. The response surface and contours for desirability function for 1018 carbon steel are shown in Figure 46. As shown in the figure, the maximum value of the function was 1.00. The optimum process parameters were solution pH at 7.36, rotation speed at 70.31 RPM, solid percentage at 75.50, and crop load at 71.94%.

Numerical optimization was also performed to determine the optimum condition

for minimum wear rate of high-chromium alloy. Again, the desirability function was utilized to combine the wear rate response. In order to find the optimum conditions in a wide range, the solution pH range varied between 1-12, rotation speed 50-90 RPM, crop load 30-80%, and solid percentage 44-84%. Figure 48 illustrated the response surface and contours for desirability function for high-chromium alloy. The optimum process parameters were solution pH at 8.69, rotation speed at 61.13 RPM, solid percentage at 64.86, and crop load at 57.63%.

86

Table 10. Factorial Design and Experimental Response for High-Chromium Alloy.

Std Run Block

Factor A: Solution

pH

Factor B: Rotation

Speed (rpm)

Factor C:

Crop Load (%)

Factor D: Solid

Concentration (%)

Response: Wear Rate

(MPY) 23 1 Block 1 0.0 -1.0 0.0 +1.0 110.03 18 2 Block 1 +1.0 0.0 -1.0 0.0 140.61 6 3 Block 1 0.0 0.0 +1.0 -1.0 136.26 10 4 Block 1 +1.0 0.0 0.0 -1.0 118.54 17 5 Block 1 -1.0 0.0 -1.0 0.0 213.31 3 6 Block 1 -1.0 +1.0 0.0 0.0 264.84 8 7 Block 1 0.0 0.0 +1.0 +1.0 108.17 27 8 Block 1 0.0 0.0 0.0 0.0 123.45 2 9 Block 1 +1.0 -1.0 0.0 0.0 90.07 5 10 Block 1 0.0 0.0 -1.0 -1.0 152.75 12 11 Block 1 +1.0 0.0 0.0 +1.0 75.79 25 12 Block 1 0.0 0.0 0.0 0.0 135.21 19 13 Block 1 -1.0 0.0 +1.0 0.0 162.77 1 14 Block 1 -1.0 -1.0 0.0 0.0 176.14 29 15 Block 1 0.0 0.0 0.0 0.0 128.77 13 16 Block 1 0.0 -1.0 -1.0 0.0 119.66 22 17 Block 1 0.0 +1.0 0.0 -1.0 197.28 15 18 Block 1 0.0 -1.0 +1.0 0.0 120.72 9 19 Block 1 -1.0 0.0 0.0 -1.0 255.15 16 20 Block 1 0.0 +1.0 +1.0 0.0 148.26 20 21 Block 1 +1.0 0.0 +1.0 0.0 103.44 7 22 Block 1 0.0 0.0 -1.0 +1.0 126.74 11 23 Block 1 -1.0 0.0 0.0 +1.0 174.66 24 24 Block 1 0.0 +1.0 0.0 +1.0 131.95 4 25 Block 1 +1.0 +1.0 0.0 0.0 122.24 26 26 Block 1 0.0 0.0 0.0 0.0 127.67 21 27 Block 1 0.0 -1.0 0.0 -1.0 152.36 28 28 Block 1 0.0 0.0 0.0 0.0 126.74 14 29 Block 1 0.0 +1.0 -1.0 0.0 202.15

87

Table 11. Analysis of Variance for High-Chromium Alloy.


Degree of Freedom


Model 49081.23 4 12270.31 42.48 <0.0001

A 32355.44 1 32355.44 112.02 <0.0001

B 7387.43 1 7387.43 25.58 <0.0001 C 2569.61 1 2569.61 8.90 0.0065 D 6768.75 1 6768.75 23.43 <0.0001

Residual 6932.34 24 288.85

Lack of fit 6858.05 20 342.90 18.46 0.0059

Pure error 74.30 4 18.57

Total 56013.57 28

88

Figure 47. Normal Probability Plot of Residual for High-Chromium Alloy.

Studentized Residuals

Nor

mal

Pro

babi

lity

(%)

-1.95 -0.90 0.15 1.19 2.24

1

510

2030

50

7080

9095

99

89

Actual factors: crop load = 57.63%, solid percentage = 64.86% Figure 48. Response Surface and Contours for Desirability Function for High-

Chromium Alloy.

0 .0 7 0

0 .3 0 2

0 .5 3 5

0 .7 6 7

1 .0 0 0 D

esira

bilit

y

1 .0 03 .7 5 6 .5 0 9 .2 5

1 2 .0 06 0

6 5

7 0

7 5

8 0

p H

Rot

atio

n sp

eed

(RP

M)

A c tu a l fa c to rs : c ro p lo a d = 5 7 .6 3 % , s o lid p e rc e n ta g e = 6 4 .8 6 %

p H

Rot

atio

n sp

eed

(RPM

)

1 . 0 3 . 7 5 6 . 5 0 9 . 2 5 1 2 . 0 0

6 0

6 5

7 0

7 5

8 0

0 . 3 7 3

0 . 4 9 80 . 6 2 4

0 . 7 4 90 . 8 7 5

0 . 1 8 9

0 . 2 7 1

0 . 9 5 3

P r e d i c t i o n 1 . 0 0X 8 . 6 9Y 6 1 . 1 3

90

EFFECT OF POLARIZATION POTENTIAL ON CORROSION RATE

Anodic polarization represents a driving force for corrosion by the anodic reaction (1). When surface potential measures more positive, the oxidizing (or corrosive) power of the solution increases. To investigate effects of polarized potential on corrosion rate of the ball mill and grinding media, three-hour grinding tests were conducted using the experimental setup described earlier under controlled electrochemical conditions. The polarization potential of the ball mill was varied using the EG&G PARC potentiostat which applied a constant potential at -1.5, -1.2, -1.0, -0.8, -0.6, -0.4, -0.2, 0 V to the mill during the entire grinding period. The change in the total wear, corrosive wear, abrasive wear, and synergistic effect of abrasive wear and corrosive wear were determined for each test and correlated with applied potential. The correlation between wear rate and polarization potential was established for different conditions of solution at pH 3.1, 6.8, and 9.2. The difference in the correlation at varying pHs indicates the effect of solution pH on corrosion rate, and possibly corrosion mechanism as well. The information generated in this task was used for determining the optimal conditions for ball mill protection. 1018 Carbon Steel Corrosion Rate

In all grinding tests, the rotation speed was fixed at 70 RPM which is 75% of the critical speed. The solid percentage was chosen to be at 64% which is close to the industrial value. The load percentage was kept at 50% which is the middle level of this factor. The effects of polarization potential on wear rate of 1018 carbon steel in different solutions are shown in Figures 49, 50, and 51, respectively. It is obvious that wear rate decrease with increasing cathodic polarization potential or making the potential more negative. For example, the wear rate was 760.4 MPY without polarization potential in pH 3.1 solution, while the wear rate was 418.2 MPY at -0.9 V polarization potential in the same solution. These three figures clearly show that the wear rate changed very little when the polarization potential was changed from -1.0 to -1.5 V. This indicates that potential of -1.0 V was sufficient to effectively reduce the wear rate of 1018 carbon steel. These three figures also suggest that the total wear rate was reduced by 42% to 46 % when a potential of -1.0 V was applied.

The total wear rate of metal determined from the weight loss measurement has

three components, which are abrasive wear, corrosive wear and synergistic effect of abrasive and corrosive wear. The polarization potential was mainly used to reduce the corrosive wear. The corrosive wear rate was determined using the approach described earlier and the results are also shown in Figures 49, 50, and 51. These three figures clearly indicate that the corrosive wear of carbon steel decrease significantly with increasing polarization potential. As shown in Figure 49, the corrosive wear rate was 290.91 MPY without polarization potential in pH 3.1 solution, which was reduced to 39.21 MPY at -1.0 V polarization potential in the same solution. The corrosive wear rate of 1018 carbon steel was reduced by 86% in pH 3.1 solution. Furthermore, cathodic protection was also effective for reducing the synergistic effect of abrasive wear and

91

corrosive wear. For example, the synergistic effect of abrasive wear and corrosive wear was reduced by 51% in pH 3.1 solution by applying -1.0 V polarization potential. But the effect of polarization potential to the abrasive wear was negligible. Figures 50 and 51 indicate that the corrosive wear rate of 1018 carbon steel was reduced by 84% in pH 6.8 and 85% in pH 9.2; the synergistic effect of abrasive wear and corrosive wear was reduced by 42% in pH 6.8 and 54% in pH 9.2, respectively. High-Chromium Alloy Corrosion Rate

In order to evaluate the effect of polarization potential on different materials, the mill grinding tests have also been carried out using high-chromium alloy as the coupons. The experimental conditions were the same as those for 1018 carbon steel. The wear rate of ball mill was measured as a function of polarization potential and results are shown in Figures 52, 53, and 54, respectively. It is clear that wear rate decreased with increasing polarization potential for high-chromium alloy. As shown in Figure 52, the wear rate was 223.72 MPY without polarization potential at solution pH 3.1, which was about 50% more than the wear rate of 138.94 MPY obtained at -0.7 V polarization potential in the same solution. Figures 52, 53, and 54 also show that the wear rate changed very little when the polarization potential was changed from -0.7 to -1.3 V. This suggests that -0.7 V potential was required to effectively reduce wear rate for high-chromium alloy coupons. These three figures clearly indicate that for high-chromium alloy the total wear rate was reduced by 36% to 38% when a potential of -0.7 V was applied.

As described earlier, the polarization potential was mainly used to reduce the

corrosive wear and the synergistic effect of abrasive wear and corrosive wear. These three figures suggest that the corrosive wear of high-chromium alloy decrease significantly with increasing polarization potential. As shown in Figure 52, the corrosive wear rate was 82.85 MPY without polarization potential in pH 3.1 solution, the corrosive wear rate reduced to 15.74 MPY at -0.7 V polarization potential at the same pH solution. The corrosive wear rate of high-chromium alloy was reduced by 81% in pH 3.1 solution. The synergistic effect of abrasive wear and corrosive wear was 71.29 MPY without polarization potential at pH 3.1, which was 30% more than that of 51.17 MPY at -0.7 V polarization potential. Similar results were obtained for experiments conducted at pH 6.8 and 9.2. These three figures indicate that the synergistic effect of abrasive wear and corrosive wear was reduced by 23% to 30% when polarization potential -0.7 V was applied. These results are in good agreement with that reported by Pazhianur and others (1997).

92

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

Potential (V, SHE)

0

100

200

300

400

500

600

700

800

Wea

r Rat

e (M

PY)

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

0

100

200

300

400

500

600

700

800

Total WearAbrasive WearSynergistic EffectCorrosive Wear

Figure 49. Effect of Polarization Potential on Corrosion Rate of Carbon Steel at pH

3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%.

93

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

Potential (V, SHE)

0

100

200

300

400

500W

ear R

ate

(MPY

)0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

0

100

200

300

400

500




94

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

Potential (V, SHE)

0

50

100

150

200

250

300

350

400W

ear R

ate

(MP

Y)

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

0

50

100

150

200

250

300

350

400




95

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

Potential (V, SHE)

0

25

50

75

100

125

150

175

200

225

250W

ear R

ate

(MPY

)


0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

0

25

50

75

100

125

150

175

200

225

250

Figure 52. Effect of Polarization Potential on Corrosion Rate of High-Chromium

Alloy at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%.

96

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

Potential (V, SHE)

0

25

50

75

100

125

150

Wea

r Rat

e (M

PY)

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

0

25

50

75

100

125

150




97

0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

Potential (V, SHE)

0

25

50

75

100

125

Wea

r Rat

e (M

PY)


0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4

0

25

50

75

100

125



98

DETERMINATION OF CORROSION CURRENT DENSITY

The ball mill can be protected from corrosive wear using cathodic polarization potential which utilizes an external power source to force current out of the dimensionally stable anode (DSA), into the electrolyte, and eventually onto the ball mill. The impressed potential cathodically polarized the ball mill and reduced its corrosion rate. The magnitude of corrosion current density represents the corrosion rate. 1018 Carbon Steel

The purpose of this study was to determine the corrosion current density when different cathodic polarization potential was applied to the coupons. The corrosion current density was directly recorded from the ammeter connected to the circuit. Lower corrosion current density means lower corrosion rate. The corrosion current density may depend on solution pH, rotation speed, crop load, and solids percentage in the ball mill which affect the diffusion rate of oxygen and composition and stability of surface reaction species. The experimental conditions were kept the same as those described earlier. The effects of polarization potential on corrosion current density of 1018 carbon steel at pH 3.1, 6.8, and 9.2 are shown in Figures 55, 56, and 57, respectively. As shown in these figures, the corrosion current density decreased significantly with increasing polarization potential. For example, Figure 55 shows the corrosion current density was 27.8 µA/m2 at 0.0 V polarization potential; it reduced to 21.0 µA/m2 at -0.4 V polarization potential; at -0.7 V polarization potential, the corrosion current density was further reduced to 15.8 µA/m2. The corrosion current density changed very little when the polarization potential was changed from -1.0 V to -1.3 V. This result proved that the polarization potential of -1.0 V was sufficient to effectively reduce the wear rate of 1018 carbon steel and the corrosion current density was 15.9 µA/m2 at polarization potential -1.0 V. Experimental results also indicate that the wear rate was 760.41 MPY at 27.8 µA/m2 corrosion current density; 608.34 MPY at 21.0 µA/m2 corrosion current density, and 494.67 MPY at 15.8 µA/m2 corrosion current density for 1018 carbon steel at pH 3.1. These data are consistent with those obtained from effect of polarization potential on corrosion rate.

Comparing Figure 55 with Figures 56 and 57 reveals that the corrosion current

density was higher in acidic solution than in neutral or alkaline solutions. As shown in these three figures, at polarization potential -1.0 V, the corrosion current density was 9.3 µA/m2 in pH 6.8 solution and 6.3 µA/m2 in pH 9.2 solution compared to 15.9 µA/m2 in pH 3.1 solution. This conclusion is in good agreement with that reported by Jones (1996).

99

High-Chromium Alloy

In order to investigate the corrosion current density for different materials, the corrosion tests have also been carried out at different conditions using high-chromium alloy coupons. The conditions for high-chromium alloy were the same as those for 1018 carbon steel as described earlier. Figures 58, 59, and 60 show the effects of polarization potential on corrosion current density of high-chromium alloy in different solutions. These figures indicate that the corrosion current density decreased with increasing cathodic polarization potential. For example, Figure 58 reveals the corrosion current density was 12.0 µA/m2 at 0.0 V polarization potential, while it was 7.5 µA/m2 at -0.6 V polarization potential. At -0.7 V polarization potential, the corrosion current density was further reduced to 6.5 µA/m2. This conclusion is similar to that obtained for 1018 carbon steel.

Comparing Figures 55, 56, 57 with Figures 58, 59, and 60 indicates that the

corrosion current density was higher for 1018 carbon steel than for high-chromium alloy under the same condition, which is in good agreement with the results obtained from polarization diagram study. This may be caused by the fact that the addition of chromium to iron in an alloy increased the ease of passivation by reducing critical cathodic current density.

100

0 200 400 600 800 1000 1200

Time (Second)

0

5

10

15

20

25

30

35

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0

5

10

15

20

25

30

35

-0.0 V

-0.4 V

-0.7 V-1.0 V -1.3 V

Figure 55. Effect of Polarization Potential on Corrosion Current Density of Carbon

Steel at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%.

101

0 200 400 600 800 1000 1200

Time (Second)

0.0

2.5

5.0

7.5

10.0

12.5

15.0

17.5

20.0

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0.0

2.5

5.0

7.5

10.

12.

15.

17.

20.

-0.0 V

-0.4 V

-0.7 V-1.0 V -1.3 V



102

0 200 400 600 800 1000 1200

Time (Second)

0.0

2.5

5.0

7.5

10.0

12.5

15.0

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0.0

2.5

5.0

7.5

10

12

15

-0.0 V

-0.4 V

-0.7 V-1.0 V

-1.3 V



103

0 200 400 600 800 1000 1200

Time (Second)

0.0

2.5

5.0

7.5

10.0

12.5

15.0

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0.0

2.5

5.0

7.5

10

12

15

-0.0 V

-0.3 V

-0.5 V

-0.7 V-1.0 V

Figure 58. Effect of Polarization Potential on Corrosion Current Density of High-

Chromium Alloy at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%.

104

0 200 400 600 800 1000 1200

Time (Second)

0.0

2.0

4.0

6.0

8.0

10.0

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0

2

4

6

8

10

-0.0 V

-0.3 V

-0.5 V

-0.7 V-1.0 V



105

0 200 400 600 800 1000 1200

Time (Second)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Cur

rent

Den

sity

(µA

/m2 )

0 200 400 600 800 1000 1200

0

1

2

3

4

5

6

-0.0 V

-0.3 V

-0.5 V

-0.7 V-1.0 V



106

DETERMINATION OF REQUIRED CURRENT DENSITY FOR EFFECTTIVE PROTECTION

The ball mill can be protected from corrosive wear using the impressed current technique which utilizes an external power source to force current out of the dimensionally stable anode (DSA), into the electrolyte, and eventually onto the ball mill. The impressed current cathodically polarizes the ball mill and reduces its corrosion rate. The magnitude of impressed current determines the effectiveness of corrosive wear rate reduction. For example, the unpolarized corrosion rate is 10–3 mA/cm2 for steel in an acid solution in Figure 4. Applying a 1.5×10–2 mA/cm2 impressed current density reduced corrosion rate by three orders of magnitude, from 10–3 to10–6 A/cm2.

The purpose of this task was to determine the optimum current density, which is

the lowest current density that provides sufficient protection from corrosive wear. Lower current density means lower energy consumption, and therefore, lower operating cost. 1018 Carbon Steel

To investigate effects of impressed current density on corrosion rate of the ball mill and grinding media, three-hour grinding tests were conducted using the experimental setup described earlier under controlled electrochemical conditions. The impressed current density of the ball mill was varied using the EG&G PARC potentiostat which applied a constant current density at 10, 50, 100, 150, 160, 180, 210, 250 mA/m2 to the mill during the entire grinding period. The change in the total wear, corrosive wear, abrasive wear, and synergistic effect of abrasive wear and corrosive wear were determined for each test and correlated with applied current density. The correlation between wear rate and impressed current density was established for different conditions of solution at pH 3.1, 6.8, and 9.2. The difference in the correlation at varying pHs indicates the effect of solution pH on corrosion rate, and possibly corrosion mechanism as well. Thus, the optimum magnitude of impressed current density was readily identified. The information generated in this task was used for determining the optimal conditions for ball mill protection.

In the following grinding tests, the rotation speed was fixed at 70 RPM which is

75% of the critical speed. The solid percentage was chosen to be at 64% which is close to the industrial value. The load percentage was kept at 50% which is the middle level of this factor. The effects of impressed current density on wear rate of 1018 carbon steel in different solutions are shown in Figures 61, 62, and 63, respectively. It is obvious that the total wear rate decreased with increasing impressed current density. For example, the total wear rate was 760.4 MPY without impressed current density in pH 3.1 solution, while the wear rate was 412.4 MPY at 210 mA/m2 impressed current density in the same solution. Figure 61 clearly shows that the wear rate changed very little when the impressed current density was changed from 210 to 250 mA/m2. This indicates that impressed current density 210 mA/m2 was sufficient to effectively reduce the wear rate of

107

1018 carbon steel in pH 3.1 solution. Figure 61 also suggests that the total wear rate was reduced by 45 % in pH 3.1 solution when current density of 210 mA/m2 was applied.

The total wear rate of metal determined from the weight loss measurement has

three components, which are abrasive wear, corrosive wear and synergistic effect of abrasive and corrosive wear. The impressed current density was mainly used to reduce the corrosive wear. The corrosive wear rate was determined using the approach described earlier and the results are also shown in Figures 61, 62, and 63. These three figures clearly indicate that the corrosive wear of 1018 carbon steel decrease significantly with increasing impressed current density. As shown in Figure 61, the corrosive wear rate was 292.9 MPY without impressed current density in pH 3.1 solution; it reduced to 39.2 MPY at 210 mA/m2 impressed current density in the same solution. The corrosive wear rate of 1018 carbon steel was reduced by 85% in pH 3.1 solution. Furthermore, cathodic protection was also effective for reducing the synergistic effect of abrasive wear and corrosive wear. For example, the synergistic effect of abrasive wear and corrosive wear was reduced by 47% in pH 3.1 solution by applying 210 mA/m2 current density. But the effect of impressed current density to the abrasive wear was negligible. Figures 62 and 63 indicate that the corrosive wear rate of 1018 carbon steel was reduced by 83% in pH 6.8 solution and 85% in pH 9.2 solution; the synergistic effect of abrasive wear and corrosive wear was reduced by 40% in pH 6.8 solution and 54% in pH 9.2 solution, respectively. These results are consistent with those obtained from potential control. For example, the total wear rate was reduced by 51% and corrosive wear rate was reduced by 86% in pH 3.1 by applying -1.0 V polarization potential.

Comparing Figure 61 with Figures 62 and 63 reveals that the required current

density for effective protection was higher for acidic solutions than for neutral or alkaline solutions. For example, an impressed current density of 210 mA/m2 provided sufficient protection from corrosive wear for pH 3.1 solution, while the required current density was 180 mA/m2 for pH 6.8 solution and 160 mA/m2 for pH 9.2 solution. High-Chromium Alloy

The mill grinding tests have also been carried out using high-chromium alloy coupons. The experimental conditions were the same as those for 1018 carbon steel. The wear rate of ball mill was measured as a function of impressed current density and results are shown in Figures 64, 65, and 66, respectively. It is clear that the total wear rate decreased with increasing impressed current density for high-chromium alloy. As shown in Figure 64, the total wear rate was 223.7 MPY without impressed current density at solution pH 3.1, which was about 50% more than the wear rate of 138.9 MPY obtained at 150 mA/m2 impressed current density in the same solution. Figure 64 also show that the wear rate changed very little when the impressed current density was changed from 150 to 190 mA/m2. This suggests that 150 mA/m2 impressed current density was required to effectively reduce wear rate for high-chromium alloy coupons in pH 3.1 solution. Figure 64 clearly indicates that for high-chromium alloy the total wear rate was reduced by 38% in pH 3.1 solution when a current density of 150 mA/m2 was applied.

108

As described earlier, the impressed current density was mainly used to reduce the corrosive wear and the synergistic effect of abrasive wear and corrosive wear. These three figures show that the corrosive wear of high-chromium alloy decreased significantly with increasing impressed current density. As shown in Figure 64, the corrosive wear rate was 82.8 MPY without impressed current density in pH 3.1 solution, it reduced to 15.7 MPY at 150 mA/m2 current density at the same pH solution. The corrosive wear of high-chromium alloy was reduced by 80% in pH 3.1 solution. The synergistic effect of abrasive wear and corrosive wear was 70.2 MPY without impressed current density at pH 3.1, which was 30% more than that of 51.2 MPY at 150 mA/m2 impressed current density. Similar results were obtained for experiments conducted at pH 6.8 and 9.2 solutions. These data are in good agreement with those obtained from potential control.

Comparing Figure 64 with Figures 65 and 66 indicates that the required current

density was 125 mA/m2 for pH 6.8 solution and 95 mA/m2 for pH 9.2 solution, while the required current density was 150 mA/m2 for pH 3.1 solution. These results are in good agreement with that obtained from 1018 carbon steel and that reported by Gartland and others (1983).

Comparing Figures 61, 62, 63 with Figures 64, 65, and 66 suggests that the

impressed current density was higher for 1018 carbon steel than that for high-chromium alloy at the same condition, which is in good agreement with the results obtained from polarization diagram study. This is because the addition of chromium to iron in an alloy increases the ease of passivation by impressed current. The required current density for effective protection at different pHs is summarized in Table 12. Table 12. Required Current Density for Effective Protection.

Required Current Density (mA/m2)

Solution 1018 Carbon Steel High-Chromium

Alloy

pH 3.1 210 150

pH 6.8 180 125

pH 9.2 160 95

109

0 40 80 120 160 200 240 280

Current Density (mA/m2)

0

100

200

300

400

500

600

700

800W

ear R

ate

(MP

Y)

0 40 80 120 160 200 240 280

0

100

200

300

400

500

600

700

800


Figure 61. Effect of Impressed Current Density on Corrosion Rate of Carbon Steel

at pH 3.1, Rotation Speed 70 RPM, Crop Load 50% and Solid Percentage 64%.

110

0 40 80 120 160 200 240 280


0

50

100

150

200

250

300

350

400

450

500W

ear R

ate

(MP

Y)

0 40 80 120 160 200 240 280

0

50

100

150

200

250

300

350

400

450

500




111

0 40 80 120 160 200 240 280


0

50

100

150

200

250

300

350

400

Wea

r Rat

e (M

PY)

0 40 80 120 160 200 240 280

0

50

100

150

200

250

300

350

400




112

0 20 40 60 80 100 120 140 160 180 200


0

25

50

75

100

125

150

175

200

225

250

Wea

r Rat

e (M

PY

)0 20 40 60 80 100 120 140 160 180 200

0

25

50

75

100

125

150

175

200

225

250


Figure 64. Effect of Impressed Current Density on Corrosion Rate of High-


113

0 20 40 60 80 100 120 140 160 180 200


0

25

50

75

100

125

150

Wea

r Rat

e (M

PY

)

0 20 40 60 80 100 120 140 160 180 200

0

25

50

75

10

12

15




114

0 20 40 60 80 100 120 140 160 180


0

25

50

75

100

125

Wea

r Rat

e (M

PY

)

0 20 40 60 80 100 120 140 160 180

0

25

50

75

100

125




115

EFFECT OF SLURRY CONDUCTIVITY ON WEAR RATE Effect of Sodium Sulfate on Wear Rate Without Cathodic Protection

The electrochemical reactions are closely related to the electric conductivity of the solution. It was reported that the electrochemical reaction speed increases with increasing solution conductivity (Koryta and others 1993). Theoretical analysis indicates that corrosive wear rate can be reduced by 99.9% if sufficient cathodic polarization is applied, as shown in Figure 4. However, experimental data show that the corrosion rate of 1018 carbon steel was reduced by 83% to 85% in different pH solutions when -1.0 V polarization potential was applied. This may be attributed to relatively poor electric conductivity of the phosphate slurry.

The effect of sodium sulfate concentration, which determines the electrical

conductivity of slurry, on wear rate of 1018 carbon steel was carried out without cathodic protection. Figure 67 illustrates the effect of sodium sulfate concentration on wear rate of 1018 carbon steel at different pHs. This figure suggests that the wear rate of 1018 carbon steel changed a little after sodium sulfate was added into the slurry. For example, the wear rate of 1018 carbon steel at pH 3.1 was 749.60 MPY without sodium sulfate after grinding 3 hours, while it was 765.23 MPY at 0.001 M and 795.21 MPY at 0.1 M sodium sulfate concentration. Thus, the effect of sodium sulfate concentration on wear rate without cathodic protection was not remarkable. Effect of Sodium Sulfate on Wear Rate With Cathodic Protection

The purpose of this task was to investigate the effect of electrolyte concentration on wear rate reduction with cathodic protection. The electrolyte used in the tests was sodium sulfate and the concentration was changed from 0, 5×10-4, 10-3, 5×10-3, 10-2, to 5×10-2 M.

In all grinding tests, the rotation speed was fixed at 70 RPM which is 75% of the

critical speed. The solid percentage was chosen to be at 64% which is close to the industrial value. The load percentage was kept at 50% which is the middle level of this factor. The grinding time was fixed at three hours for observing a significant weight loss. Figure 68 and Table 13 show the effect of sodium sulfate concentration on wear rate reduction of 1018 carbon steel at pH 3.1 when -1.0 V was applied. As can be seen, the total wear reduction and corrosive wear reduction increased with increasing sodium sulfate concentration. For example, the total wear reduction was 41.1% and corrosive wear reduction was 83.0% without sodium sulfate in pH 3.1 solution, while the total wear reduction and corrosive wear reduction increased to 47.4% and 92.4% in the same solution after 10-2 M sodium sulfate was added into the phosphate slurry. Figure 68 and Table 13 clearly show that the total wear reduction and corrosive wear reduction changed very little with further increasing sodium sulfate concentration from 10-2 to 5×10-2 M.

116

This indicates that part of the applied electric current was consumed by the phosphate slurry, grinding medium, and the ball mill to produce heat.

The effect of sodium sulfate concentration on wear rate reduction of 1018 carbon

steel at pH 6.8 was shown in Figure 69 and Table 14. The results indicate that wear reduction significantly increased with increasing sodium sulfate concentration when -1.0 V polarization potential was applied. For example, the total wear reduction increased from 43.2% to 48.2% and the corrosive wear reduction increased from 84.5% to 93.4% with increasing sodium sulfate concentration from 0 to 10-2 M.

Figure 70 and Table 15 illustrate the effect of sodium sulfate concentration on

wear rate reduction of 1018 carbon steel at pH 9.2 when -1.0 V polarization potential was applied. They suggest that total wear reduction increased from 46.1% to 49.1% and the corrosive wear reduction increased from 85.1% to 94.2% after 10-2 M sodium sulfate was added into the phosphate slurry. Comparing Figure 70 with Figures 69, 68 reveals that corrosion was more serious in acidic solution than that in neutral and alkaline solutions even sodium sulfate was added into the slurry.

The effect of sodium sulfate concentration on wear rate reduction of high-

chromium alloy has been carried out under the same conditions as those for 1018 carbon steel. Figure 71 and Table 16 show the effect of sodium sulfate concentration on wear rate reduction of high-chromium alloy at pH 3.1 when -0.7 V polarization potential was applied. Figure 71 and Table 16 indicate that wear reduction increased with increasing sodium sulfate concentration. For example, the total wear reduction increased from 38.1% to 43.4% and the corrosive wear reduction increased from 81.2% to 92.1% after 10-2 M sodium sulfate was added into the slurry. This conclusion is in good agreement with that obtained for 1018 carbon steel.

117

0.00 0.02 0.04 0.06 0.08 0.10 0.12

Sodium Sulfate Concentration (M)

100

200

300

400

500

600

700

800

900W

ear R

ate

(MP

Y)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

10

20

30

40

50

60

70

80

90

pH 3.1pH 6.8pH 9.2

Figure 67. Effect of Sodium Sulfate Concentration on Wear Rate of 1018 Carbon

Steel Without Cathodic Protection.

118

0.00 0.01 0.02 0.03 0.04 0.05 0.06

Sodium Sulfate Concentartion (M)

30

40

50

60

70

80

90

100W

ear R

educ

tion

(%)

0.00 0.01 0.02 0.03 0.04 0.05 0.06

30

40

50

60

70

80

90

100

Corrosive WearTotal Wear

Figure 68. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of 1018

Carbon Steel at pH 3.1 When -1.0 V Was Applied.

119

Table 13. Effect of Sodium Sulfate Concentration on Wear Rate of 1018 Carbon Steel at pH 3.1.

Total Wear Corrosive Wear

Without Protection

With Protection (-1.0 V)

Without Protection

With Protection (-1.0 V) Concentration

(M) Conductivity

(µS/cm) Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

0 1680 760.41 428.87 43.6 292.91 39.21 86.5

5×10-4 2692 763.16 422.03 44.7 297.63 38.39 87.1

10-3 3510 765.23 418.58 45.3 302.26 32.65 89.2

5×10-3 6537 771.06 410.20 46.8 308.42 27.14 91.2

10-2 10340 779.36 409.94 47.4 311.74 23.70 92.4

5×10-2 16520 783.26 408.32 47.8 313.30 21.93 92.9

120

0.00 0.01 0.02 0.03 0.04 0.05 0.06


30

40

50

60

70

80

90

100W

ear R

ate

Red

uctio

n (%

)


0.00 0.01 0.02 0.03 0.04 0.05 0.06

30

40

50

60

70

80

90

10



121



Without Protection


Without Protection


(M) Conductivity

(µS/cm) Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

0 153 450.50 249.31 44.5 173.98 27.68 84.1

5×10-4 267 453.13 248.21 45.2 174.96 21.01 88.0

10-3 342 458.12 246.01 46.3 176.92 15.75 91.1

5×10-3 1141 460.23 240.24 47.8 177.91 3.17 92.6

10-2 1743 463.24 239.96 48.2 178.90 11.81 93.4

5×10-2 7840 465.21 239.58 48.5 179.66 11.14 93.8

122

0.00 0.01 0.02 0.03 0.04 0.05 0.06

Sodium Sulfate Concentration

30

40

50

60

70

80

90

100W

ear R

educ

tion

(%)


0.00 0.01 0.02 0.03 0.04 0.05 0.06

30

40

50

60

70

80

90

100



123



Without Protection


Without Protection


(M) Conductivity

(µS/cm) Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

0 702 366.89 198.23 46.1 114.19 17.13 85.1

5×10-4 1413 369.26 196.45 46.8 114.95 12.65 89.0

10-3 1902 373.23 195.57 47.6 116.16 9.08 92.2

5×10-3 4157 376.71 195.53 48.1 117.25 7.75 93.4

10-2 6870 384.12 195.48 49.1 119.55 6.95 94.2

5×10-2 13286 387.21 195.15 49.6 120.51 6.63 94.5

124

0.00 0.01 0.02 0.03 0.04 0.05 0.06


30

40

50

60

70

80

90

100W

ear R

educ

tion

(%)


0.00 0.01 0.02 0.03 0.04 0.05 0.06

30

40

50

60

70

80

90

100

Figure 71. Effect of Sodium Sulfate Concentration on Wear Rate Reduction of High

Chromium Alloy at pH 3.1 When -0.7 V Was Applied.

125

Table 16. Effect of Sodium Sulfate Concentration on Wear Rate of High-Chromium Alloy at pH 3.1.


Without Protection


Without Protection


(M) Conductivity

(µS/cm) Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

Wear Rate

(MPY)

Wear Rate

(MPY)

Wear Rate

Reduced (%)

0 1680 223.72 138.94 38.1 82.85 15.74 81.2

5×10-4 2692 224.15 135.61 39.5 83.02 13.24 84.1

10-3 3510 226.16 132.53 41.3 83.76 11.56 86.2

5×10-3 6537 228.26 131.03 42.6 84.53 8.29 90.2

10-2 10340 229.41 130.52 43.1 84.96 7.99 91.6

5×10-2 16520 230.32 130.36 43.4 85.29 6.75 92.1

Slurry Conductivity Measurement

Phosphate slurry conductivity measurement was necessary to determine the effect of sodium sulfate addition on wear rate with cathodic protection. The apparatus used in the tests was Conductivity and TDS Meter (Model TC102) acquired from Control Company, Friendswood, TX, as shown in Figure 72. Solutions to be measured were held in a 0.5 L beaker equipped with a magnetic stirrer. Different sodium sulfate concentrations in the range 0 to 0.1 M were prepared. Figure 73 shows the effect of sodium sulfate concentration on phosphate slurry conductivity in different pH solutions. Figure 73 indicates that slurry conductivity increased with increasing sodium sulfate concentration. At the same concentration, the slurry conductivity was higher in pH 3.1 solution than that in pH 6.8 and pH 9.2 solutions. This is attributed to the fact that there are more free ions in acid and basic solutions than in neutral solution. The conductivity decreases with increasing pH in acidic solutions, reaches the minimum at the neutral pH and increases with increasing pH in basic solutions. This result is consisted with previous work reported by Muccitelli and Diangelo (1994).

126

Figure 72. Photograph of Conductivity Meter.

127

0.00 0.02 0.04 0.06 0.08 0.10 0.12


0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000C

ondu

ctiv

ity (µ

S/cm

)0.00 0.02 0.04 0.06 0.08 0.10 0.12

0

200

400

600

800

100

120

140

160

180

200

pH 3.1pH 9.2pH 6.8

Figure 73. Effect of Sodium Sulfate Concentration on Phosphate Slurry Conductivity. Eh-pH Diagram

The Eh-pH diagrams have been proven useful for many aspects. They are widely used for metallurgical purposes, for predicting species in solutions, for converting products of ores and in hydrothermal systems etc.

Eh is defined as the electromotive force between a half-cell reaction in any state,

and the SHE (Standard Hydrogen Electrode). Potentials are given versus SHE. The Eh-pH diagrams illustrate the fields of stability of mineral or chemical species in terms of the activity of hydrogen ions (pH) and the activity of electrons (Eh). Consequently, the

128

reactions illustrated on Eh–pH diagrams involve either proton transfer (e.g., hydrolysis) or electron transfer (oxidation or reduction) or both.

Eh-pH diagrams are thermodynamic diagrams which show the possibility of

metals to be attacked, dissolved or corroded at various pH and oxidation conditions. To determine the most stable phase of a metal in aqueous environments at a particular potential and pH value, it is necessary to refer to Eh-pH diagram.

The schematic Eh-pH diagram for iron was plotted using HSC Chemistry

acquired from Buffalo, NY. The plot is shown in Figure 74. The diagram is a map of the regions of immune, passive, and corrosion behavior for iron as a function of potential and pH. Four distinct regions are shown in this figure. The region at the bottom of the diagram represents the conditions where iron is immune and no corrosion occurs. Under these conditions, iron is thermodynamically stable. This region covers reducing conditions (negative potentials) across the entire pH range, from acid to alkaline. For any potential-pH combination in this region, iron is stable and does not corrode.

Figure 74 has two corrosion areas where iron corrodes. In both the region to the

left (oxidizing and acidic) and the small region to the extreme right (reducing and highly alkaline), iron reacts to form soluble products, and corrosion continues. The passivity area represents a region of passivity for iron. Under oxidizing conditions in neutral to alkaline solutions, iron reacts to form insoluble products and further corrosion reaction is blocked by a protective film. An assumption of the Eh-pH diagram is that any insoluble product will be protective.

Eh-pH diagrams allow presentation of a large amount of information in a compact

and efficient format. For instance, a single diagram (Figure 74) describes the general corrosion behavior of iron. Iron is a relatively active metal and corrodes under reducing, moderately oxidizing, and strongly oxidizing conditions in strong acids. In weak acids, iron again corrodes under reducing and moderately oxidizing conditions; however, iron can become passive if the oxidizing conditions are increased to a sufficiently high level. Under neutral and mildly alkaline conditions, iron does not corrode, because it either is immune under strongly reducing conditions or is in a passive state for more oxidizing conditions. In strong alkaline environments, iron is free from corrosion except for the small region of potentials and pHs where a soluble, alkaline, corrosion product forms.

The construction of Eh-pH diagrams is a useful way to summarize redox

properties and chemical speciation. The Eh-pH diagrams of Fe-O-H system and Fe-Cr-H2O system were constructed and results are shown in Figures 75 and 76, respectively. Figure 75 shows that iron was easily attacked in a wide pH range. Corrosion experiments show that lower pH values attack the iron at a faster rate than at high pH values. This figure also suggests that iron can be polarized when polarization potential was applied. For example, Fe was the most stable phase when polarization potential -0.8 V was applied. This will reduce the corrosion rate of iron. Figure 76 is the Eh-pH diagram for Fe-Cr-H2O system. This figure indicates that the chromium alloy was easily attacked at pH values below 3. Similarly, the corrosion rate of high-chromium alloy was reduced when polarization potential was applied.

129

Figure 74. Schematic Example of Eh-pH Diagram for Iron (All Dissolved Species at

Activities of 10-6 M).

0 2 4 6 8 10 12 14

pH

-1.6

-1.2

-0.8

-0.4

-0.0

0.4

0.8

1.2

1.6

2.0

Eh (V

, SH

E)

0 2 4 6 8 10 12 14

-1.6

-1.2

-0.8

-0.4

-0.0

0.4

0.8

1.2

1.6

2.0

Corrosion

Passivity

Immunity Corrosion

130

Figure 75. Eh-pH Diagram for Fe-O-H System, Assuming Fe(OH)3 as Stable Fe3+

Phase and Activity of Dissolved Fe = 10-6 M.

0 2 4 6 8 10 12 14

pH

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4Eh

(V, S

HE)

0 2 4 6 8 10 12 14

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Fe3+

Fe(OH)3Fe2+

Fe

Fe(OH)2

FeO22-

Fe2O3

O2

H2O

H2O

H2

131

Figure 76. Eh-pH Diagram for Fe-Cr-H2O System, Assuming Total Concentrations

of 10-2 M Fe and 5×10-3 M Cr. CORROSION MECHANISM INVESTIGATION

To understand the cathodic protection process, scanning electron microscopy (SEM), energy dispersive spectrometer (EDS), and X-ray diffraction (XRD) methods were used to investigate the corrosion products, surface morphology, and composition when grinding phosphate rock with and without cathodic protection. The information

0 2 4 6 8 10 12 14

pH

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0Eh

(V, S

HE)

0 2 4 6 8 10 12 14

-1.0

-0.8

-0.6

-0.4

-0.2

-0.0

0.2

0.4

0.6

0.8

1.0

FeOH2+

Cr2O3

Fe3+

CrO42-

Cr(OH)4-

Fe3O4

FeCr2O4

Fe2O3

Fe2+

Fe

O2

H2O

H2O

H2

132

generated in this task was used for determining the optimal conditions for ball mill protection. Scanning Electron Microscopy (SEM)

The scanning electron microscopy is a microscope that uses electrons rather than light to form an image. A beam of electrons is produced at the top of the microscope by heating a metallic filament as shown in Figure 77. The electron beam follows a vertical path through the column of the microscope. It makes its way through electromagnetic lenses which focus and direct the beam down towards the sample. Once the beam strikes the sample, other electrons (backscattered or secondary) are ejected from the sample. Detectors collect the secondary or backscattered electrons, and convert them to a signal that is sent to a viewing screen similar to the one in an ordinary television. SEM produces detailed photographs that provides important information about the surface structure and morphology of almost any kind of samples.

There are many advantages of using the SEM rather than light microscope. The

SEM is one of the most versatile and widely used tools of modern science as it allows the study of both morphology and composition of biological and physical materials. By scanning an electron probe across a specimen, high resolution images of the morphology or topography of a specimen, with great depth of field, at very low or very high magnifications can be obtained. Compositional analysis of a material may also be obtained by monitoring secondary X-rays produced by the electron-specimen interaction. Thus detailed maps of elemental distribution can be produced from multi-phase materials or complex, bio-active materials. Characterization of fine particulate matter in terms of size, shape, and distribution as well as statistical analyses of these parameters, may be performed.

The machine used for SEM tests was Hitachi S-3200. Figure 78 shows the

photograph of SEM. Before any work was done with the SEM, the sample was properly mounted for viewing. The sample's dimensions should not exceed 150 mm in diameter. If the sample was non-conductive, conductive graphite should be sputtered on the sample’s surface for proper viewing. After the sample was properly attached to the mount, the height-restriction marker was used to ensure that the sample was short enough to clear the SEM's working parts. If the sample was too tall, the z-axis should be adjusted appropriately to ensure the clearance before you tilt the sample.

SEM Analysis for 1018 Carbon Steel

To obtain the coupon sample for SEM analysis, the grinding time was fixed at three hours to produce significant weight loss. The microstructures of the corroded surfaces of the coupons obtained at different pH values under different impressed currents were investigated by scanning electron microscopy. Figures 79 and 80 show the SEM images of 1018 carbon steel after it was exposed to phosphate slurry at pH 3.1 with

133

and without cathodic protection, respectively. In Figure 79, the gouges and scratches on the surface indicate wear due to corrosion and abrasion. The darker spots might be due to the formation of an oxide. The figure clearly shows that there are a lot of deep and shallow pits on the surface. This suggests that pitting corrosion was the main corrosion mechanism. Pitting corrosion was caused by localized attack in an otherwise resistant surface. The rate of corrosion was greater in some areas than others. Pitting corrosion was caused by the presence of phosphoric acid, fluosilicic acid, and sulfuric acid that manage to pass through the passive film and initiate corrosion, resulting in rupture of the passive film. Figure 80 shows the SEM image when polarization potential -1.0 V was applied. This SEM image shows fewer dark spots, indicating a reduction of corrosion on the surface.

Figures 81 and 82 illustrate the SEM images of 1018 carbon steel exposed to

phosphate slurry at pH 6.8 with and without cathodic protection, respectively. These two figures suggest that corrosion was significantly reduced after polarization potential -1.0 V was applied. Comparing Figures 81, 82 with Figures 79, 80 reveals that the corrosion was much more serious in acidic solution than in neutral solution.

Figures 83 and 84 indicate the SEM images of 1018 carbon steel after exposing to

phosphate slurry at pH 9.2 with and without cathodic protection. These two figures clearly show that the corrosion was not so significant in alkaline solution as in acidic and neutral solutions. This is mainly because the passive film was readily formed in alkaline solutions and the corrosion was considerably reduced by -1.0 V polarization potential.

134

Figure 77. Schematic Illustration of SEM.

135

Figure 78. Photograph of SEM.

136

Figure 79. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1

Without Cathodic Protection.

137

Figure 80. SEM Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH

3.1 With Polarization Potential -1.0 V.

138



139


With Polarization Potential -1.0 V.

140



141


With Polarization Potential -1.0 V. SEM Analysis for High-Chromium Alloy

The corrosion tests have been carried out using high-chromium alloy as coupons. The conditions for high-chromium alloy were the same as those for 1018 carbon steel. Figures 85 and 86 show the SEM images of high-chromium alloy after exposing to phosphate slurry at pH 3.1 with and without cathodic protection. These two figures indicate that few scratches were found on the high-chromium alloy surface. The corrosion was significantly reduced after polarization potential -0.7 V was applied.

142

Comparing Figures 85, 86 with Figures 79, 80 indicates that the corrosion of 1018 carbon steel was much more serious than that of high-chromium alloy at the same conditions.

Figures 87 and 88 illustrate the SEM images of high-chromium alloy exposed to

phosphate slurry at pH 6.8 with and without cathodic protection, respectively. These two figures suggest that corrosion was significantly reduced after polarization potential -0.7 V was applied. Comparing Figures 87, 88 with Figures 85 and 86 reveals that the corrosion was much more serious in acidic solution than in neutral solution.

Figures 89 and 90 indicate the SEM images of high-chromium alloy after

exposing to phosphate slurry at pH 9.2 with and without cathodic protection. These two figures clearly show that the corrosion in alkaline solution was not as significant as that in acidic and neutral solutions. This is mainly because the passive film was easily formed in neutral and alkaline solutions. The corrosion was considerably reduced by -0.7 V polarization potential.

Comparing SEM images of high-chromium alloy with those of 1018 carbon steel

reveals that the corrosion of 1018 carbon steel was much more serious than that of high-chromium alloy under the same conditions. The carbon steel that depends on a passive film for corrosion resistance is especially susceptible to pitting by local breakdown of the film at isolated sites. The SEM images indicate that there are fewer scratches and pits on high-chromium alloy surface than on 1018 carbon steel surface under the same condition. These SEM images also suggest that -0.7 V potential was required to effectively reduce corrosion rate for high-chromium alloy coupons, while the required potential for 1018 carbon steel was -1.0 V.

143

Figure 85. SEM Image of High-Chromium Alloy Exposed to Phosphate Slurry at

pH 3.1 Without Cathodic Protection.

144


pH 3.1 With Polarization Potential -0.7 V.

145



146



147



148


pH 9.2 With Polarization Potential -0.7 V. Energy Dispersive X-Ray Spectrometer (EDS)

Energy Dispersive X-ray Spectrometer (EDS) is a standard procedure for identifying and quantifying elemental composition of sample areas as small as a few square micrometers. The characteristic x-rays are produced when a material is bombarded with electrons in an electron beam instrument, such as a scanning electron microscope (SEM). Detection of these x-rays can be accomplished by an energy dispersive spectrometer, which is a solid state device that discriminates among x-ray energies.

149

SEM micrographs were collected using the conventional electron detector that shows the topography of the surface of a bulk specimen. Surface and microscopic analysis of materials can be performed using the energy dispersive spectrometer (EDS). This includes collecting elemental spectra from the surface that simultaneously displays peaks that correspond to the concentrations of the elements carbon through uranium. These spectra can be used for qualitative or quantitative analysis. Also, x-ray line scans and maps can be collected that show the distribution of elements across a surface for magnifications that range from 30 to 20,000 times. The precise elemental composition of materials can be obtained with high spatial resolution.

One of the most outstanding features of the SEM-EDS is that it allows elemental

analysis and observation from an ultra microarea to a wide area on the specimen surface without destroying the specimen. This feature is very important in fields of research including electronics and material sciences, chemistry and physics, biology and medicine, industrial engineering.

The machine used for EDS tests was a Noran Voyager System (Hitachi

Company). Figure 91 shows the photograph of EDS. Samples selected for EDS analysis are first cleaned using toluene and methanol to remove hydrocarbon residue, and then dried in a vacuum oven. Carbon paint is used to cover the base of the sample and the epoxy to improve conduction and reduce charging effects for the test. Samples are sputter coated with conductive graphite for 5 minutes. EDS Analysis for 1018 Carbon Steel

To produce the coupon sample for EDS, the grinding time was fixed at three hours to cause a significant weight loss. The element content of the corroded surfaces of the coupons obtained at different pH value under different impressed current was investigated by energy dispersive spectrometer. Figures 92 and 93 show the EDS images of 1018 carbon steel after it was exposed to phosphate slurry at pH 3.1 with and without cathodic protection, respectively. Figure 92 indicates that iron atom content is 32.34% and the element weight percent is 55.72%. Figure 93 shows the EDS image when polarization potential -1.0 V was applied. This figure indicates that iron atom content increased to 38.48% and the element weight percent increased to 62.91%. The impressed current cathodically polarizes the coupon and reduces its corrosion rate. The Eh-pH diagram of Fe-O-H system (Figure 75) shows that Fe2+ was the most stable phase in solution pH 3.1 without polarization potential, while Fe was the most stable phase when polarization potential -1.0 V was applied due to the polarized surface. Cathodic polarization increased iron atom content and element weight percent. Furthermore, phosphoric acid existed in phosphate slurry. The general reactions involved in phosphate slurry are as follows.

2

22 HFeHFe +→+ ++ (28) ++− +→+ HPOFeFePOH 4)(32 243

242 (29)

150

The summation gives 224342 )(232 HPOFePOHFeH +→++ −+ (30)

The solubility constant (KSP) of ferrous phosphate (Fe3(PO4)2) is 1.0×10-36 at 25oC. According to the Taggart Principle, a chemical reaction theory, the ferrous phosphate was difficult to dissolve in the solution and then precipitated on the coupon surface. Ferrous phosphate formed a protective film covered on the coupon surface to act as a corrosion resistant physical barrier. This was also helpful to increase the iron element content on coupon surface.

Figures 94 and 95 illustrate the EDS images of 1018 carbon steel exposed to

phosphate slurry at pH 6.8 with and without cathodic protection, respectively. These two figures suggest that iron atom content increased when polarization potential -1.0 V was applied. Iron atom content was 22.60% and element weight percent was 43.28% without polarization potential. After polarization potential -1.0 V was applied, iron atom content increased to 59.67% and element content increased to 81.63%. The Eh-pH diagram of Fe-O-H system indicates that Fe2+ was the most stable phase in neutral solution without cathodic protection. Fe2+ was polarized to stable phase Fe when polarization potential -1.0 V was applied. This increased the iron elemental content.

Figures 96 and 97 indicate the EDS images of 1018 carbon steel after exposing to

phosphate slurry at pH 9.2 with and without cathodic protection. These two figures clearly show that iron atom content increased significantly when polarization potential -1.0 V was applied. Without cathodic protection, iron atom content was 31.99% and element weight percent was 59.32%. Iron content and element weight percent increased to 37.58% and 64.45% when polarization potential -1.0 V was applied. The Eh-pH diagram of Fe-O-H system indicates that Fe(OH)3 was the most stable phase in solution pH 9.2 without polarization potential. The phase Fe(OH)3 changed to stable phase Fe(OH)2 when polarization potential from -0.25 to -0.55 V was applied. The reactions in the solution are as follows:

2

22 HFeHFe +→+ ++ (28) ++ +→+ HOHFeOHFe 2)( 22

2 (31) −− +→+ OHOHFeeOHFe 23 )()( (32)

Fe2+ cations hydrolyze to form a soluble weak base, Fe(OH)2, leaving behind excess H+. The Eh-pH diagram of Fe-O-H system suggests that the coupon surface was polarized to iron with increased impressed polarization potential, increasing the iron atom content and element weight percent.

151

Figure 91. Photograph of EDS.

152

Element Atom, % Element Wt, % Fe 32.34 55.72 O 22.38 11.05 C 21.87 8.10 Ca 11.18 13.83 P 5.79 5.53 Si 5.12 4.44

Mn 0.31 0.53 Na 0.31 0.22 Al 0.28 0.23 Mg 0.26 0.20 S 0.16 0.16

Figure 92. EDS Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1


153


Mn 0.35 0.56 Na 0.26 0.18 Al 0.26 0.20 Mg 0.32 0.23 S 0.16 0.15



154


Mn 0.19 0.35 Na 0.18 0.14 Al 0.30 0.28 Mg 0.22 0.18 S 0.16 0.17



155

Element Atom, % Element Wt, %

Fe 59.67 81.63 O 12.08 4.73 C 18.13 5.34 Ca 3.17 3.11 P 1.38 1.05 Si 4.39 3.02

Mn 0.51 0.68 Na 0.08 0.05 Al 0.18 0.12 Mg 0.26 0.15 S 0.14 0.11



156

Element Atom, % Element Wt, % Fe 31.99 59.32 O 21.43 11.38 C 31.61 12.61 Ca 5.98 7.96 P 3.61 3.72 Si 5.09 4.75 Al 0.29 0.26



157


Mn 0.34 0.57 Al 0.35 0.29 Mg 0.19 0.14


With Polarization Potential -1.0 V. EDS Analysis for High-Chromium Alloy

The corrosion tests have been carried out using high-chromium alloy as coupons. The conditions for high-chromium alloy were the same as those for 1018 carbon steel. Figures 98 and 99 show the EDS images of high-chromium alloy after exposing to phosphate slurry at pH 3.1 with and without cathodic protection. Figure 98 indicates that

158

iron atom content was 70.91% and the element weight percent was 84.63%. Figure 99 shows the EDS image when polarization potential -0.7 V was applied. This figure indicates that iron atom content increased to 79.34% and the element weight percent increased to 92.04%. The impressed current cathodically polarized the coupon and reduced its corrosion rate. The Eh-pH diagram of Fe-Cr-H2O system (Figure 76) indicates that Cr2FeO4 was the most stable phase in solution without cathodic protection. The phase Cr2FeO4 was polarized to stable phase Fe when polarization potential -0.7 V was applied, which reduced the corrosion rate and increased the iron element content.

Figures 100 and 101 illustrate the EDS images of high-chromium alloy exposed to

phosphate slurry at pH 6.8 with and without cathodic protection, respectively. These two figures suggest that iron atom content increased when polarization potential -0.7 V was applied. Iron atom content was 54.47% and element weight percent was 79.67% without polarization potential. After polarization potential -0.7 V was applied, iron atom content increased to 61.39% and element weight percent increased to 84.12%. The Eh-pH diagram of Fe-Cr-H2O system indicates that in neutral solution without cathodic protection Cr2FeO4 was the most stable phase, which was polarized to stable phase Fe when polarization potential -0.7 V was applied. This increased the iron elemental content on the surface.

Figures 102 and 103 illustrate the EDS images of high-chromium alloy after

exposure to phosphate slurry at pH 9.2 with and without cathodic protection. These two figures clearly show that iron atom content increased significantly when polarization potential -0.7 V was applied. Without cathodic protection, iron atom content was 40.28% and element weight percent was 70.40%. Iron content and element weight percent increased to 48.27% and 75.88% when polarization potential -0.7 V was applied.

159

Element Atom, % Element Wt, % Fe 70.91 84.63 O 14.86 6.73 Ca 8.43 4.11 P 2.88 2.88 C 1.02 0.98 Si 0.96 0.36

Mn 0.94 0.31 Figure 98. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at


160

Element Atom, % Element Wt, % Fe 79.34 92.04 O 4.75 1.58 Ca 1.36 1.14 P 0.81 0.52 C 12.01 3.00 Si 0.45 0.26

Mn 1.28 1.46 Figure 99. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at


161

Element Atom, % Element Wt, % Fe 54.47 79.67 C 23.96 7.54 O 14.77 6.19 Ca 2.37 2.49 P 1.70 1.38

Mn 0.88 1.26 Si 1.66 1.22 Cr 0.19 0.26

Figure 100. EDS Image of High-Chromium Alloy Exposed to Phosphate Slurry at


162

Element Atom, % Element Wt, % Fe 61.39 84.12 C 21.47 6.33 O 11.52 4.52 Ca 2.00 1.97 Si 1.75 1.21 P 1.09 0.83

Mn 0.54 0.72 Cr 0.24 0.30



163

Element Atom, % Element Wt, % Fe 40.28 70.40 C 24.53 9.22 F 14.99 8.91 O 17.04 8.53 Si 2.30 2.02 S 0.35 0.35 Al 0.34 0.29 Cr 0.17 0.28



164

Element Atom, % Element Wt, % Fe 48.27 75.88 C 16.99 5.74 F 24.54 13.12 O 8.57 3.86 Si 1.35 1.07 S 0.14 0.13 Cr 0.14 0.20


pH 9.2 With Polarization Potential -0.7 V. X-Ray Diffraction (XRD)

X-ray diffraction (XRD) is a versatile, non-destructive analytical technique for identification and quantitative determination of the various crystalline forms, known as phases, of compounds present in powdered and solid samples. XRD has proved to be one of the most useful techniques for structure analysis. It is widely used in research, production and quality control environments to analyze crystalline materials.

165

The x-rays are generated by electrons as they accelerate through electromagnetic fields situated at various points of the storage ring. After interacting with the sample, the intensities of the diffracted x-rays are recorded using single photon counters and image-plate detectors. XRD can be used to reveal the structure of a complex arrangement of atoms in a molecule, to study structural changes resulting from subjecting samples to extremes of temperature and pressure etc. Identification is achieved by comparing the x-ray diffraction pattern of diffractogram obtained from an unknown sample with an internationally recognized database containing reference patterns for more than 70,000 phases.

In our experiments, x-ray diffraction was used to identify the structure of the

corrosion products and the surface chemistry of each coupon at different pH value under different impressed current. x-ray diffraction in the conventional θ-2θ setup was used to identify the structure of the corrosion products. The results obtained using XRD were used to determine the main corrosion product.

The machine used for XRD tests was Philips X’Pert as shown in Figure 104.

Samples selected for XRD analysis are first cleaned using toluene and methanol to remove hydrocarbon residue, and then dried in a vacuum oven. XRD Analysis for 1018 Carbon Steel

To generate the coupon sample for XRD analysis, the grinding time was fixed at three hours to result in a significant weight loss. The structure of the corrosion product of the coupons obtained at different pH value under different impressed current was investigated by x-ray spectrometer. Figures 105 and 106 show the XRD images of 1018 carbon steel after it was exposed to phosphate slurry at pH 3.1 with and without cathodic protection, respectively. Figure 105 indicates that there exited iron (Fe) and ferric oxide (Fe2O3) on the coupon surface without cathodic protection. This suggests that Fe2O3 was the main corrosion product in pH 3.1 solution (Wu and Kim, 1999). Figure 106 shows the XRD image when polarization potential -1.0 V was applied. This figure reveals that corrosion was considerably reduced since Fe2O3 disappeared from the coupon surface. This conclusion is in good agreement with the Eh-pH diagram of Fe-O-H system (Figure 75). Figure 75 shows that Fe2+ and Fe2O3 were the most stable phases in solution pH 3.1 without polarization potential, while Fe was the most stable phase when polarization potential -1.0 V was applied due to the polarized surface. The reactions involved in acidic solutions are as follows:

322 234 OFeOFe →+ (33) 3224 OFeOFeO →+ (34)

Figures 107 and 108 illustrate the XRD images of 1018 carbon steel exposed to phosphate slurry at pH 6.8 with and without cathodic protection, respectively. Figure 107 suggests that ferric oxide (Fe2O3), magnetite (Fe3O4) and ferric hydroxide (Fe(OH)3) were the corrosion products in pH 6.8 solution (Shan and others 2002; Raymahashay and

166

Khare 2003). Figure 108 shows the XRD image when polarization potential -1.0 V was applied. This figure indicates that corrosion was essentially eliminated since Fe2O3, Fe3O4 and Fe(OH)3 did not appear on the coupon surface. This conclusion is consistent with that obtained from the Eh-pH diagram of Fe-O-H system.

Figures 109 and 110 indicate the XRD images of 1018 carbon steel after exposing

to phosphate slurry at pH 9.2 with and without cathodic protection. Figure 109 clearly shows that ferric oxide (Fe2O3), magnetite (Fe3O4) and ferric hydroxide (Fe(OH)3) were the corrosion products in pH 9.2 solution without cathodic protection, which is similar to that obtained from pH 6.8 solution. Figure 110 shows the XRD image when polarization potential -1.0 V was applied. This figure suggests that corrosion was substantially reduced since Fe2O3, Fe3O4 and Fe(OH)3 disappeared from the coupon surface. This conclusion is in good agreement with that obtained from the Eh-pH diagram of Fe-O-H system. The reactions involved in neutral and alkaline solutions are as follows:

322 234 OFeOFe →+ (33) 222 )(OHFeOHOFe →++ (35) 2

2 )(2 OHFeOHFe →+ −+ (36) 322 )(4634 OHFeOHOFe →++ (37) OHOFeOOHFe 23222 42)(4 +→+ (38)

The formation of magnetite (Fe3O4) involves a slow oxidation step:

−+−+ +→+ OHFeOHOHFe 22 (39) 2)(OHFeOHFeOH →+ −+ (40) OHOFeOOHFe 24322 62)(6 +→+ (41)

167

Figure 104. Photograph of XRD.

168

Figure 105. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1 Without Cathodic Protection.

A

A B B

B

B

B

A: Fe B: Fe2O3

2θO

Cou

nts/

s

169

Figure 106. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 3.1 With Polarization Potential -1.0V.

A

A

A: Fe

2θO

Cou

nts/

s

170

Figure 107. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH

6.8 Without Cathodic Protection.

A

A

B B B C

C

B, D

C D

B

A: Fe B: Fe2O3 C: Fe3O4 D: Fe(OH)3

2θO

Cou

nts/

s

171

Figure 108. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH

6.8 With Polarization Potential -1.0V.

2θO

A

A

A: Fe

Cou

nts/

s

172

Figure 109. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2 Without Cathodic Protection.

2θO

A

A

B B B

B, D

B

C C

C C

D

A: Fe B: Fe2O3 C: Fe3O4 D: Fe(OH)3

Cou

nts/

s

173

Figure 110. XRD Image of 1018 Carbon Steel Exposed to Phosphate Slurry at pH 9.2 With Polarization Potential -1.0V.

XRD Analysis for High-Chromium Alloy

The XRD tests have been carried out using high-chromium alloy as coupons. The conditions for high-chromium alloy were the same as those for 1018 carbon steel. Figures 111 and 112 show the XRD images of high-chromium alloy after exposing to phosphate slurry at pH 3.1 with and without cathodic protection. Figure 111 indicates that there exits iron (Fe), ferric oxide (Fe2O3), chromium oxide (Cr2O3) and chromium ferrite (FeCr2O4) on the coupon surface without cathodic protection. This suggests that Fe2O3, Cr2O3 and FeCr2O4 were the main corrosion products in pH 3.1 solution (Khan and Kelebek, 2002; Xu and others 2002). Figure 112 shows the XRD image when polarization potential -0.7 V was applied. This figure indicates that corrosion was completely reduced since these corrosion products disappeared from the coupon surface when polarization potential -0.7 V was applied. The Eh-pH diagram of Fe-Cr-H2O system (Figure 76) reveals the similar conclusion since Fe was the stable phase after applying polarization potential -0.7 V.

2θO

A

A

A: Fe

Cou

nts/

s

174

Figures 113 and 114 illustrate the XRD images of high-chromium alloy exposed to phosphate slurry at pH 6.8 with and without cathodic protection, respectively. Figure 113 suggests Cr2O3 and FeCr2O4 were the corrosion products in pH 6.8 solution without cathodic protection. Figure 114 shows the XRD image when polarization potential -0.7 V was applied. This figure suggests that corrosion was almost completely reduced since the corrosion products did not appear on the coupon surface. This conclusion is consistent with that obtained from the Eh-pH diagram of Fe-O-H system.

Figures 115 and 116 indicate the XRD images of high-chromium alloy after

exposing to phosphate slurry at pH 9.2 with and without cathodic protection. Figure 115 clearly shows that Cr2O3 and FeCr2O4 were the corrosion products in pH 9.2 solution without cathodic protection, which is similar to that obtained from pH 6.8 solution. Figure 116 shows the XRD image when polarization potential -0.7 V was applied. This figure suggests that corrosion was significantly reduced since most of the corrosion products disappeared from the coupon surface except for FeCr2O4. The Eh-pH diagram of Fe-O-H system shows that the stable phase should be Fe or FeCr2O4, or both at pH 9.2 when polarization potential -0.7 V was applied.

175

Figure 111. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 3.1 Without Cathodic Protection.

A

A

C C

A

B B

D

B, D D C C

B

B

A: Fe B: Fe2O3 C: Cr2O3 D: FeCr2O4

2θO

Cou

nts/

s

176

Figure 112. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at


A

A

A B

B

A: Fe B: Cr

2θO

Cou

nts/

s

177


A

A

A: Fe B: Cr2O3 C: FeCr2O4

A

B B

B

B C

C C

2θO

Cou

nts/

s

178

Figure 114. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 6.8 With Polarization Potential -0.7 V.

A

A

A B

B

2θO

A: Fe B: Cr

Cou

nts/

s

179


A

A

A

A: Fe B: Cr2O3 C: FeCr2O4

B

B B

B C C

C

2θO

Cou

nts/

s

180

Figure 116. XRD Image of High-Chromium Alloy Exposed to Phosphate Slurry at pH 9.2 With Polarization Potential -0.7 V.

A

A: Fe B: Cr C: FeCr2O4

A A

C C B

B C

2θO

Cou

nts/

s

181

ECONOMIC EVALUATION

Corrosion is basically an economic problem. In the national economy, the annual costs due to corrosion are calculated to be ten billion dollars fifteen years ago (Trethewey and Chamberlain 1988). Industry is continuously seeking ways to save money in the purchase or repair of equipment. Any steps taken that will reduce cost or increase effectiveness will in turn enhance the economic benefits of a corrosion control program. Previous efforts to reduce grinding media consumption were mainly concentrated on development of more sophisticated corrosion resistant alloys, better design of protection equipment, more extensive use of non-corrosive materials, and availability of more effective inhibitors and coatings. The most economical approach for the protection of a grinding ball mill is to use an effective cathodic protection process based on the impressed current principle.

In the grinding tests, the rotation speed was fixed at 70 RPM which is 75% of the

critical speed. The solution pH was 3.1 and solid percentage was chosen to be at 64%, which are close to the industrial values. The load percentage was kept at 50% which is the middle level of this factor. Given the ball mill size (14×24 ft) and the grinding media size (2’’, 1½” and 1’’), the economic evaluation was performed as follows.

Interior surface area for one mill:

S = hrr ••+• ππ 22 2

= )3048.024(23048.0142

23048.0142

2

××

×

+

× ππ

= 126.84 m2

Since there are two mills at the CF fertilizer plant, the total interior surface area for two mills is:

ST = 2×126.84 = 253.68 m2

The volume of one mill is:

hrV ••= 2π = )3048.024(23048.014 2

××

×π

= 104.82 m3 So the total volume for two mills is:

VT = 2×104.82 = 209.64 m3

Assume the media load percentage is 43%, the media volume can be drawn:

VM = 43%×209.64 = 90.15 m3

182

The volumes for three different balls are as follows:

3''2 3

4 rV •= π 3

20254.02

34

×

= π = 0.00006864 m3

3''2

11 34 rV •= π

3

20254.05.1

34

×

= π = 0.00002896 m3

3''1 3

4 rV •= π 3

20254.01

34

×

= π = 0.00000858 m3

So the needed ball numbers for different sizes are:

43779100006864.0

315.90

''2 ==n

1037638000028958.0

315.90

''211

==n

350233100000858.0

315.90

''1 ==n

The surface areas for three different balls are:

22''2 )

20254.02(44 ×

=•= ππ rS = 0.008107 m2

22''2

11)

20254.05.1(44 ×

=•= ππ rS = 0.004560 m2

22''1 )

20254.01(44 ×

=•= ππ rS = 0.002027 m2

Then the total surface areas for three different balls are:

S2’’T = 437791×0.008107 = 3549.17 m2

S11/2’’T = 1037638×0.004560 = 4731.63 m2

S1’’T = 3502331×0.002027 = 7099.22 m2

183

The total corroded surface area is:

Scorrosion = ST+ S2’’T + S11/2’’T + S1’’T = 253.68 + 3549.17 + 4731.63 + 7099.22 = 15633.70 m2

The coupon surface area is:

SCoupon 22 )2

0254.083

(×

=•= ππ r = 0.000071256 m2

As reported earlier, the wear rate of 1018 carbon steel coupon is 3.7151 mg

without cathodic protection and 1.7925 mg with applied potential for grinding three hours. Assuming working hours are 24 hour/day and operating days are 350 day/year, the total saved material can be calculated.

WCarbon Steel 1000

7925.17151.3000071256.0

70.15633324350 −

×××=

= 1181101725 g = 1181.10 ton

Since the wear rate of high-chromium alloy coupon is 1.0930 mg without

cathodic protection and 0.642 mg with cathodic protection for grinding three hours, the total saved material can be determined:

WChromium Alloy 10006420.00930.1

000071256.070.15633

324350 −

×××=

= 277060687.7 g = 277.06 ton

Given the price $817/ton for 1018 carbon steel and $1250/ton for high-chromium

alloy, the benefit from saved material can be calculated:

MCarbon Steel = 817 × 1181.10 = $964,958/year MChromium Alloy = 1250 × 277.06 = $346,325/year

The applied polarization potential was -1.0 V for 1018 carbon steel and -0.70 V

for high-chromium alloy. The applied current was 210 mA/m2 for 1018 carbon steel and

184

150 mA/m2 for high-chromium alloy. Given the total corroded surface area, the used electricity power can be calculated:

PowerCarbon Steel hrVmmA

dayhr

yearday 1170.15633

100021024350 2

2 ×××××=

= 27577.85 KW.hr

PowerChromium Alloy hrVmmA

dayhr

yearday 17.070.15633

100015024350 2

2 ×××××=

= 13788.92 KW.hr

Since the electricity price is $0.035/kw.hr, the electricity cost is:

MElectricity for Carbon Steel = 0.035 × 27577.85 = $965/year MElectricity for Chromium Alloy = 0.035 × 13788.92 = $483/year

So the total benefits for different materials are:

BenefitCarbon Steel = 964958 – 965 = $963, 993/year BenefitChromium Alloy = 346325 – 483 = $345,842/year

Based on the above calculations, the operating constants and economic

performance for different materials are shown in Table 17 and Table 18, respectively. The results indicate that wear rate reduction can be reduced 1181.10 ton/year for 1018 carbon steel and 277.06 ton/year for high-chromium alloy. It is clear that $963,993/year can be saved for 1018 carbon steel and $345,842/year for high-chromium alloy after cathodic protection is applied to the ball mill.

185

Table 17. Ball Mill Size and Operating Constants.

Factor Constant

Ball Mill Size (ft) 14 × 24

Volume for Two Mills (m3) 209.64

Interior Surface Area for Two Mills

(m2) 253.68

Operating hours (hr/day) 24

Operating Days (day/year) 350

Mill Capacity (ton/hr) 240

Media Load Percentage (%) 43

1/3 2’’

1/3 1½” Media Size

1/3 1’’

Grinding Media Volume for Two mills

(m3) 90.15

Grinding Media Surface Area for Two Mills

(m2) 15,380.02

Total Corroded Surface Area for Two Mills

(m2) 15633.70

186

Table 18. Economic Evaluation for Different Materials for Two Mills.

Material

Factor

1018 Carbon Steel High-Chromium Alloy

Wear Rate Reduction (g/m2·hr) 8.994 2.110

Wear Rate Reduction (ton/year) 1,181.10 277.06

Grinding Media Price ($/ton) 817 1,250

Benefit from Media ($/year) 964,958 346,325

Electricity Price ($/Kw-hr) 0.035 0.035

Electricity Cost ($/year) 965 483

Total Benefit ($/year) 963,993 345,842

Media Cost for Company ($/year)

3,187,733 1,144,080

187

CONCLUSIONS

Based on the results obtained on the CF Industries samples, the following conclusions can be drawn:

(1) Unlike earlier mill corrosion studies that used devices that merely simulated

mill motion, the results from the specially designed ball mill more closely resemble those to be expected from the industrial ball mills.

(2) Polarization diagrams indicated that the current density was higher in

nitrogenated solution than in oxygenated solution at the same pH value. This is mainly because the fresh surface of electrode was easily oxidized in oxygenated solution than in nitrogenated solution, which passivates metal surface.

(3) The current density of 1018 carbon steel was considerably higher than that of

high-chromium alloy under the same condition. For example, the maximum current density was about 42 mA/cm2 at pH 4.6 for 1018 carbon steel in oxygenated buffer solution, while it was 22.5 mA/cm2 for high-chromium alloy. This is attributed to the fact that the addition of chromium to iron increased the ease of passivation by reducing critical anodic current density. Addition of both chromium and nickel to iron markedly increased the ease of passivation.

(4) The current density was higher in buffer solution than that in pond water

solution under the same operating conditions. This behavior may be caused by the fact that the pH of a buffer solution resisted changes when small amounts of acid or base were added. But the pH of a pond water solution was increased by ongoing reactions. The corrosion rated was decreased with increasing the pH value of pond water solution.

(5) The influence of individual variables and their interactions on the wear rate of

1018 carbon steel was in the order of solution pH > rotation speed > solid percentage > crop load > rotation speed2 > rotation speed × solid percentage. The effects of these variables on the wear rate of high-chromium alloy was in the order of solution pH > rotation speed > solid percentage > crop load. The effects of interactions of these variables on the wear rate of high-chromium alloy were not significant. Solution pH had the most significant effect on the wear rate for both 1018 carbon steel and high-chromium alloy.

(6) The regression equation for 1018 carbon steel wear rate as coded variables

was: Wear rate=530.34–177.50×A+85.72×B–54.58×C–76.53×D+46.41×B2–43.95×B×C The regression equation for high-chromium alloy wear rate was: Wear rate=156.24–50.31×A+24.81×B–14.63×C–23.75×D

188

where A represents solution pH, B rotation speed, C crop load, and D stands for solid percentage.

(7) The optimum process parameters for minimum wear rate were: for 1018

carbon steel, solution pH at 7.36, rotation speed at 70.31 RPM, solid percentage at 75.50, and crop load at 71.94%; and for high-chromium alloy, solution pH at 8.69, rotation speed at 61.13 RPM, solid percentage at 64.86, and crop load at 57.63%.

(8) Polarization potential of -1.0 V was sufficient to effectively reduce the wear

rate of 1018 carbon steel and -0.7 V for high-chromium alloy. The total wear rate was reduced by 42% to 46 % and corrosive wear rate was reduced by 84% to 86% for 1018 carbon steel when potential of -1.0 V was applied. For high-chromium alloy, 36% to 38% was reduced for total wear rate and 80% to 81% was reduced for corrosive wear when polarization potential of -0.7 V was applied.

(9) The required current density to effectively reduce the wear rate of 1018

carbon steel was 210 mA/m2 in pH 3.1 solution, 180 mA/m2 in pH 6.8 solution, and 160 mA/m2 in pH 9.2 solution. The required current density for high-chromium alloy was 150 mA/m2 in pH 3.1 solution, 125 mA/m2 in pH 6.8 solution, and 95 mA/m2 in pH 9.2 solution.

(10) The effect of slurry conductivity on wear rate reduction demonstrated that

the wear reduction increased with increasing sodium sulfate concentration. For example, the total wear reduction was 41.1% without sodium sulfate in pH 3.1 solution, while it increased to 47.4% in the same solution after 10-2 M sodium sulfate was added into the slurry. Similar results were obtained with the high-chromium alloy.

(11) SEM analysis indicated that there were a lot of deep and shallow pits on the

surface. This suggests that pitting corrosion was the main corrosion mechanism. Pitting corrosion was caused by the presence of phosphoric acid, fluosilicic acid, and sulfuric acid that manage to pass through the passive film and initiate corrosion, resulting in rupture of the passive film. SEM photographs suggested that corrosion was significantly reduced after polarization potential was applied.

(12) XRD images suggested that Fe2O3 was the main corrosion product for 1018

carbon steel in pH 3.1 solution without cathodic protection, while Fe2O3, Fe3O4 and Fe(OH)3 were the corrosion products in pH 6.8 and pH 9.2 solutions. For high-chromium alloy, Fe2O3, Cr2O3 and FeCr2O4 were the main corrosion products in pH 3.1 solution; Cr2O3 and FeCr2O4 were the corrosion products in pH 6.8 and pH 9.2 solutions. XRD images also suggested that the corrosion was significantly reduced since most of the corrosion products disappeared from the coupon surface when polarization potential was applied.

(13) Economic evaluation indicates that for the CF fertilizer plant at Plant City,

Florida where two 14’×24’ ball mills used for size reduction the economic benefit will be $963,993/year if 1018 carbon steel balls are used and $345,842/year when high-chromium alloy balls are employed.

189

REFERENCES Abdennabi AMS, Abdulhadi AI, Abu-orabi ST, Saricimen H. 1996. The inhibition action of 1(benzyl)1-H-4,5-dibenzoyl-1,2,3-triazole on mild steel in hydrochloric acid media. Corrosion Science 38 (10): 1791-1800. Abuzriba MB, Dodd RA, Worzala FJ, Conrad JR. 1992. Wear corrosion: separation of the components of corrosion and wear. Technical Note. Corrosion 48: 2-4. Allen MP. 1993. The Vernal phosphate rock mill. In: Beneficiation of phosphate: theory and practice. El-Shall H, Moudgil BM, Wiegel R, editors. Littleton (CO): SME. Aramaki K. 2001. Effects of organic inhibitors on corrosion of zinc in areated 0.5 M NaCl solution. Corrosion Science 43(10): 1985-2000. Bartels JJ, Gurr TM. 1994. Phosphate rock. In: Carr DD, editor. Industrial minerals and rocks, 6th ed. Littleton (CO): SME, p 751-64. Bentiss F, Traisnel M, Lagrenee M. 2000. The substitute 1,3,4-oxadiazoles: a new class of corrosion inhibitors of mild steel in acidic media. Corrosion Science 42(1): 127-46. Bond FC. 1964. Metal wear in crushing and grinding. Chem. Eng. Prog. 60(2): 90-100. Box GEP, Hunter WG, Hunter JS. 1978. Statistics for experiments. New York: John Wiley. Bozec NL, Compere C, L’Her M, Laouenan A, Costa D, Marcus P. 2001. Influence of stainless steel surface treatment on the oxygen reduction reaction in seawater. Corrosion Science 43(4): 765-86. Breslin CB, Chen C, Mansfeld F. 1997. The electrochemical behavior of stainless steels following surface modification in cerium-containing solutions. Corrosion Science 39(6): 1061-73. Charles JA, Crane FAA. 1989. Selection and use of engineering materials. 2nd edition. Sevenoaks, Kent (England): Butterworths. p 11. Chenje TW, Simbi DJ, Navara E. 2003. The role of corrosive wear during laboratory milling. Minerals Engineering 16: 619-24. Conde A, Garcia I, Damborenea JJ. 2001. Pitting corrosion of 304 stainless steel after laser surface melting in argon and nitrogen atmospheres. Corrosion Science 43(5): 817-28.

190

Cottam CA, Emmony DC. 1998. TEA-CO2 laser surface processing of corroded metals. Corrosion Science 41(8): 1529-38. Davis JR, editor. 2000. Corrosion: understanding the basics. Materials Park (OH): ASM International. p 1-9. Dobryszycki J, Biallozor, S. 2001. On some organic inhibitors of zinc corrosion in alkaline media. Corrosion Science 43(7): 1309-19. Dun DJ. 1985. Metal removal mechanisms comprising wear in mineral processing. In: Ludema KC, editor. Wear of materials. New York: American Society of Mechanical Engineers. p 501-8. El-Etre AY. 2001. Inhibition of acid corrosion of aluminum using vanillin. Corrosion Science 43(6): 1031-9. Elliot P. 1987. Metals handbook. Vol. 13, Corrosion. 9th ed. Metals Park (OH): ASM International. p 339. Fedrizzi L, Bonora PL. 1997. Organic and inorganic coatings for corrosion prevention – research and experiences. London: The Institute of Materials. Flis J, Mankowski J, Zakroczymski T, Bell T. 2001. The formation of phosphate coatings on nitrided stainless steel. Corrosion Science 43(9): 1711-25. Fontana MG. 1986. Corrosion engineering. 3rd ed. New York: McGraw-Hill. p 1-5, 457. Fregonese, M, Idrissi H, Mazille H, Renaud L, Cetre Y. 2001. Initiation and propagation steps in pitting corrosion of austenitic stainless steels: monitoring by acoustic emission. Corrosion Science 43(4): 627-41. Gartland, PO, Strommen R, Bardal E. 1983. Current density requirements for cathodic protection of steel structures in the North Sea. Materials Performance 6, p 40-8. Geesey GG, Gillis RJ, Avci R, Daly D, Hamilton M, Shope P, Harkin G. 1996. The influence of surface features on bacterial colonization and subsequent substratum chemical changes of 316L stainless steel. Corrosion Science 38(1): 73-95. Giddey S, Cherry B, Lawson F, Forsyth M. 2001. Stability of oxide films formed on mild steel in turbulent flow conditions of alkaline solutions at elevated temperatures. Corrosion Science 43(8): 1497-1517. Habazaki H, Paez MA, Shimizu K, Skeldon P, Thompson GE, Wood GC, Zhou X. 1996. The importance of surface treatment to the anodic oxidation behavior of Al-Cu alloys. Corrosion Science 38(7): 1033-42.

191

Habib K. 2000. Detection of crevice corrosion by optical interferometry. Corrosion Science 42(3): 455-67. Haneda R, Aramaki K. 1998. Protective films on copper by multistep modification of an alkanethiol monolayer with chlorosilanes and alkanediol. J. Electrochem. Soc. 145(8): 2786-91. Harben P. 1980. Where is Florida phosphate industry going? Industrial Minerals nr 148, p 48-55. Hoey GR, Dingley W, Freeman C. 1975. Corrosion inhibitors reduce ball wear in grinding sulfide ore. CIM Bulletin 68: 120-3. Itoh J, Sasaki T, Ohtsuka T. 2000. The influence of oxide layers on initial corrosion behavior of copper in air containing water vapor and sulfur dioxide. Corrosion Science 42(9): 1539-51. Iwasaki I, Reimer SC, Orlich JN, Natarajan KA. 1985. Corrosive and abrasive wear in ore grinding. Wear 103: 253-67. Jakobsen PT, Maahn E. 2001. Temperature and potential dependence of crevice corrosion of AISI 316 stainless steel. Corrosion Science 43(9): 1693-1709. Jang JW, Iwasaki I, Moore JJ. 1989. The effect of galvanic interaction between martensite and ferrite in grinding media wear. Corrosion 45(5): 402-7. Jones DA. 1971. Application of electrode kinetics to theory and practice of cathodic protection. Corrosion Science 11(6): 439-51. Jones DA. 1996. Principles and prevention of corrosion. 2nd edition. Upper Saddle River (NJ): Prentice Hall. p 3-31. Khan A, Kelebek S. 2002. Oxygen mass transfer into flotation slurries – Part I: Effect of mild steel corrosion during grinding of quartz. Trans. Instn. Min. Metall. Sect. C: Mineral Process. Extr. Metall. 110: 115-22. Klimpel R. 1982. Slurry rheology influence on the performance of mineral/coal grinding circuits. Part 1. Mining Engineering 34: 1665-72. Klimpel R. 1983. Slurry rheology influence on the performance of mineral/coal grinding circuits. Part 2. Mining Engineering 35: 21-6. Kobrin G. 1976. Corrosion by microbiological organisms in natural waters. Materials Performance 15(7): 38-43.

192

Koryta J, Dvorak J, Kavan L. 1993. Principles of electrochemistry. 2nd ed. New York: John Wiley. p 90-93. Kotlyar D, Pitt CH, Wadsworth ME. 1987. Simultaneous corrosion and abrasion measurements under grinding condition. Corrosion-NACE 44(4): 221-7. Liu GQ, Zhu ZY, Ke W, Han EH, Zeng CL. 2001. Corrosion behavior of stainless steels and nickel-based alloys in acetic acid solutions containing bromide ions. Corrosion Science 57(8): 730-8. Liu PS, Liang KM, Gu SR. 2001. High-temperature oxidation behavior of aluminide coatings on a new cobalt-base superalloy in air. Corrosion Science 43(7): 1217-26. Lowrison GC. 1974. Crushing and grinding. London: Butterworth & Co. p 231-2. Lu YF, Aoyagi Y. 1994. Laser-induced dry cleaning in air--a new surface cleaning technology in lieu of carbon fluorochloride (CFC) solvents. Jpn. J. Appl. Phys. 33, Pt. 2(3B): L430-L433. Ma HY, Cheng XL, Chen SH, Wang C, Zhang JP, Yang HQ. 1998. An AC impedance study of the anodic dissolution of iron in sulfuric acid solutions containing hydrogen sulfide. J. Electroanal. Chem. 451(1-2): 11-17. Ma HY, Cheng XL, Li GQ, Chen SH, Quan ZL, Zhao SY, Niu L. 2000. The influence of hydrogen sulfide on corrosion of iron under different conditions. Corrosion Science 42(10): 1669-83. McBee CL, Kruger J. 1972. Nature of passive films on iron-chromium alloys. Electrochem. Acta 17(8): 1337-41. Meloy TP, Crabtree D. 1966. Surface tension and viscosity in wet grinding. DECHEMA Monographien 57(993-1026): 405-26. Montgomery DC. 1991. Design and analysis of experiments. New York: John Wiley. Muccitelli JA, Diangelo NA. 1994. Electrolytic conductivity of aqueous solutions of potassium and sodium phosphate to 325oC. Journal of Chemical and Engineering Data 39(1): 131-3. Muller B, Gampper M. 1994. Corrosion and inhibition of corrosion of aluminum pigments in alkaline aqueous medium. Werkst. und Korros. 45(5): 272-7. Muller B, Oughourlian C, Schubert M. 2000. Amphiphilic copolymers as corrosion inhibitors for zinc pigment. Corrosion Science 42(3): 577-84.

193

Munger CG. 1984. Corrosion prevention by protective coatings. Houston: NACE. p 174. Nakazawa H, Iwasaki I. 1985. Effect of pyrite-pyrrhotite contact on their floatabilities. Minerals and Metallurgical Processing 2: 206-11. National Materials Advisory Board. 1981. Comminution and energy consumption. report of the Committee on Comminution and Energy Consumption. Washington (DC): National Academy Press. Publication nr NMAB-364. Norman TE, Loeb CM Jr. 1948. Wear tests on grinding balls. Trans. AIME 176: 249-86. Nozawa K, Aramaki K. 1999. One- and two-dimensional polymer films of modified alkanethiol monolayers for preventing iron from corrosion. Corrosion Science 41(1): 57-73. Oesch S. 1996. The effect of SO2, NO2, NO and O3 on the corrosion of unalloyed carbon steel and weathering steel--the result of laboratory exposures. Corrosion Science 38(8): 1357-68. Ornek D, Jayaraman A, Wood TK, Sun Z, Hsu CH, Mansfeld F. 2001. Pitting corrosion control using regenerative biofilms on aluminum 2024 in artificial seawater. Corrosion Science 43(11): 2121-33. Pazhianur R, Adel GT, Yoon R-H, Richardson PE. 1997. Cathodic protection to minimize corrosive wear in ball mills. Minerals and Metallurgical Processing 14(4): 1-7. Perren RA, Suter TA, Uggowitzer PJ, Weber L, Magdowski R, Bohni H, Speidel MO. 2001. Corrosion resistance of super duplex stainless steel in chloride ion containing environments: investigations by means of a new microelectrochemical method. I. Precipitation-free states. Corrosion Science 43(4): 707-26. Perren RA, Suter TA, Solenthaler C, Gullo G, Uggowitzer PJ, Bohni H, Speidel MO. 2001. Corrosion resistance of super duplex stainless steel in chloride ion containing environments: investigations by means of a new microelectrochemical method. II. Influence of precipitates. Corrosion Science 43(4): 727-45. Pope DH, Duquette DJ, Johannes AH, Wayner PC. 1984. Microbiologically influenced corrosion of industrial alloys. Mat. Perform. 23(4): 14-18. Pozzo RL, Iwasaki I. 1987. Effect of pyrite and pyrrhotite on the corrosive wear of grinding media. Minerals and Metallurgical Processing 4: 166-71. Pozzo RL, Iwasaki I. 1989. Pyrite-pyrrhotite grinding media interactions and their effects on media wear and flotation. Journal of Electrochemical Society 136(6): 1734-9.

194

Pozzo RL, Malicsi AS, Iwasaki I. 1990. Pyrite-pyrrhotite-grinding media contact and its effect on flotation. Minerals and Metallurgical Processing 7: 16-21. Rajagopal V, Iwasaki I. 1992. Corrosion properties of cast iron mall materials in wet grinding. Corrosion 48(2): 124-31. Raymahashay BC, Khare AS. 2003. The arsenic cycle in late quaternary fluvial sediments: mineralogical considerations. Current Science 84(8): 1102-4. Rausch W. 1990. The phosphating of metals. Finishing Publications. Salih SA. 2002. Corrosion behavior of some stainless steel alloys in molten potassium bisulphate. Corrosion Prevention & Control 3: 32-47. Salvago G, Magagnin L. 2001. Biofilm effect on the cathodic and anodic processes on stainless steel in seawater near the corrosion potential. Part 1--Corrosion potential. Corrosion Science 57(8): 680-92. Salvago G, Magagnin L. 2001. Biofilm effect on the cathodic and anodic processes on stainless steel in seawater near the corrosion potential. Part 2--Oxygen reduction on passive metal. Corrosion Science 57(8): 759-67. Scotto V, Alabiso G, Marcenaro G. 1986. An example of microbiologically influenced corrosion. The behavior of stainless steels in natural seawater. Bioelectrochem. Bioenerg. 16(2): 347-55. Senkevich JJ, Jones DA, Chatterjee I. 2000. Measuring the growth films on low carbon steel at 500oC by impedance spectroscopy. Corrosion Science 42(2): 201-10. Shan GB, Liu HZ, Xing JM, Zhang GD, Ma ZY, Liu XQ, Liu JG. 2002. Surface modification of nano-magnetic Fe3O4 particles and its applications. In: Proceedings of the China-EU Forum on Nanosized Technology; 2002 Dec 17-18; Beijing, China. p 68-76. Shishodia V, Rastogi S, Pitchumani B. 1993. Effect of ball size distribution on rock phosphate grinding in industrial ball mill. In: El-Shall H, Moudgil BM, Wiegel R, editors. Beneficiation of phosphate: theory and practice. Littleton (CO): SME. p 157-62. Souto RM, Alanyali H. 2000. Electrochemical characteristics of steel coated with TiN and TiAlN coatings. Corrosion Science 42(12): 2201-11. Stoecker JG, Pope DH. 1986. Study of biological corrosion in high temperature demineralized water. Mat. Perform. 25(6): 51-6.

195

Stoyanova AE, Sokolova EI, Raicheva SN. 1997. The inhibition of mild steel corrosion in 1M HCl in the presence of linear and cyclic thiocarbamides--effect of concentration and temperature of the corrosion medium on their protective action. Corrosion Science 39(9): 1595-1604. Taneichi D, Haneda R, Aramaki K. 2001. A novel modification of an alkanethiol self-assembled monolayer with alkylisocyanates to prepare protective films against copper corrosion. Corrosion Science 43(8): 1589-1600. Tatnall RE. 1981a. Case histories: bacteria induced corrosion. Mat. Perform. 20(8): 41-8. Tatnall RE. 1981b. Fundamentals of bacteria induced corrosion. Mat. Perform. 20(9): 32. Tao DP, Yoon R-H, Lu MX, Richardson PE, Luttrell GH. 1997. Improving pyrite rejection by galvanic control. Coal Preparation 18: 53-68. Tomashov ND, Chernova GP. 1967. Passivity and protection of metals against corrosion. New York: Plenum. p 151-8. Trethewey KR, Chamberlain J. 1988. Corrosion for students of science and engineering. New York: Wiley. p 4-9. Tsuji N, Nozawa K, Aramaki K. 2000. Ultrathin protective films prepared by modification of an N,N-dimethylalkylamine monolayer with chlorosilanes for preventing corrosion of iron. Corrosion Science 42(9): 1523-38. Uhlig HH, Revie RW. 1985. Corrosion and corrosion control: an introduction to corrosion science and engineering. 3rd edition. New York: Wiley. p 96-111. Uhlig HH. 1971. Corrosion and corrosion control: an introduction to corrosion science and engineering. 2nd ed. New York: Wiley. Wallinder D, Hultquist G, Tveten B, Hornlund E. 2001. Hydrogen in chromium: influence on corrosion potential and anodic dissolution in neutral NaCl solution. Corrosion Science 43(7): 1267-81. Wang F, Itoh K, Watanabe T. 2002. Relationship between the crystallographic structure of electrodeposited Fe-Cr alloy film and its thermal equilibrium diagram. Materials Transactions 43(3): 439-42. Wang L. 2001. Inhibiting effect of 2-mercaptopyrimidine on the corrosion of a low carbon steel in phosphoric acid. Corrosion Science 43(9): 1637-44.

196

Wang L, Yin GJ, Yin, JG. 2001. 2-mercaptothiazoline and cetyl pyridinium chloride as inhibitors for the corrosion of a low carbon steel in phosphoric acid. Corrosion Science 43(6): 1197-1202. Wu X, Kim SB. 1999. Synthesis and electrode properties of α-Fe2O3 from iron phthalocyanine. Electrochemical and Solid-State Letters 2(4): 184-6. Xu Q, He YD, Qi HB, Wang DR, Li ZW, Gao W. 2002. Oxidation behavior of micro- and nano-crystalline coatings deposited by series double-pole electro-pulse discharge. Materials Letters 56: 85-92. Yasuda HK, Yu QS, Reddy CM, Moffitt CE, Wieliczka DM, Deffeyes JE. 2001. “Barrier-adhesion” principle for corrosion protection. Corrosion Science 57(8): 670-9.

corrosion protection of grinding mills in...

Documents