AUSTRALASIAN CORROSION ASSOCIATION CORROSION TECHNOLOGY An Introduction to Corrosion and its Control Version 3.0 Copyright 1998 Australasian Corrosion Association Inc, PO Box 634, , Brentford Square VIC 3131, Australia.Product names mentioned in this document are trademarks or registered trademarks of their respective companies and are herby acknowledged. The use or representation of @ particular brand ot product in this publication is not to be taken to imply ACA or CPC approval or endorsement of the product or the brand, ‘The Australasian Corrosion Association, the Corrosion Prevention Centre, their employees, officers, and directors do not accept any responsibilty forthe use ofthe methods or materials discussed herein, The information is derived from various sourees and was believed to be correct when published. The information is advisory only, provided in good faith and not claimed to be an exhaustive neaiment ofthe relevant subject. The use of materials and methods described is solely at the risk ofthe user. Further professional advice might be needed t0 be obtained before taking any action based on the information providedCONTENTS Chapter J: Introduction 1, What is Corrosion?... ci 2: Costs and Importance of Corrosion : 3. Expressions for Corrosion Rate : Chapter 2: The Corrosion Process 1. The Copper-Zine C 2. Other Galvanic Cell 3, The Galvanic Series 4, Importance of the Cathode Phe Corrosion of Steel sasnurcn sosion Cell Chapter 3: Predicting Corrasion Reactions 1. Oxidation and Reduction .. Electrode Potentials. i seeoneee vase Applications and Limitations of the Electromotive Series Electrochemical Series vs, Galvanic Series Reference Electrodes...... Conversion of Electrode Potentials. Pourbaix Diagrams... Corrosion Rate and Current Flow 9. Polarisation ..... 10. Experimental Determination 11. Passivity 12, Polarisation Curves and the Pourbaix Diagram os £ SUGHNSSSS PNAVALVH Chapter 4: Types of Metallic Corrosion 1! General Corrosion... 2. Galvanic Corrosion 3. Crevice Corrosion 4, Pitting Corrosion... 5, Intergranular Corrosion. 6. Selective Corrosion. 7. Erosion Corrosion... 8. Cavitation... 9, Fretting Corrosion sec... 10. Stress Corrosion Cracking .. 11. Hydrogen Damage... 12. Corrosion Fatigue.. 13. Liquid Metal Embrittlement 14, Stray Current Corrosion Appendix: Summary of Main Forms of Chapter 5: Corrosion in Natural Environments 1. Corrosion in the Atmosphere... 2. Corrosion in Water. : 3, Corrosion in SoilsChapter 6: Chapter 7: Chapter 8: Chapter 9: 2. Life of the Structure and Corrosion AllowanceS.n0 3. Importance of Shape 4. Compatibility of Materials. 5. Mechanical Factors... 6. High Temperatures... saaneae 6 7. Eflects of Fabrication : eee 67 Corrosion Properties of Metals 1, Carbon and Low Alloy Steels 2. Cast Irons 3, Stainless Steels. eceenae 4. Nickel and its Alloys. 5, Titanium and its Alloys. 6. Copper and Copper Alloys. 7. Aluminium Alloys 8, ZINC weno 9. Lead 10, Magnesium 11. Tin. Inhibitors and Water Treatment Types of Aqueous Inhibitors. Inhibitors for Other Environments. Principles of Water Treatment... Formation and Control of Scale. Fouling Deposits Biological Fouling -... Corrosion Control AAUALRS Organic Coatings 1. Surface Preparation ae 2. Liquid Coatings for Atmospheric Exposure 3. Types of Liquid Coatings... 4, Pigments and Additive: 5. Selecting the Coating System . 6. Pipeline Coatings . 7. Quality Assurance and Inspection. Chapter 10: Metallic and Inorganic Coatings 1. Hot Dip Coatings 2. Flame or Metal Spraying... 3. Electroplating 4. Other Metal Coating 5. Anodizing..... 6. Phosphating 7. Chromating : 8. Black Oxide Coatings . 9. Vitreous Enamel Coatings . 10. Cement CoatingsChapter 11: Cathodic and Anodic Protection 1. Principles of Cathodic Protection .....0eenesnsese Vel 2. The Galvanic Method. 1-2 3. The Impressed Current Method ......cnssmnniosannisnnnnnnnsese U3 4. Cathodic Protection Current Requirements .....esem toe LES 5. Cathodic Protection Criteria... : sennnees 116 6. Stray Current Effects. 4-7 7. Cathodic Protection Monitoring .. nae : 11-8 8. Anodic Protection ....... a cae 13-10 Appendix A: Introduction to Metallurgy 1. The Structure of Metals...... I APP A+] 2. The Structure of Real Metals APP A-3 3. Alloying... Seabesan APP A-4 4, Structure and Properties of Steels a coo APP A-S 5. Heat Treatment of Steels... : APP A-7 6, Other Constituents of Metal: APP AS 7. Welding... APP A-8 8. Tensile Properties of Metals... toe APP A-9 9, Fracture and Fracture Testing... “LAPP A-1} References and Further Reading Glossary of Corrosion TermsFOREWORD ‘The Australasian Corrosion Association has devcloped this short course to provide an understanding of how and why corrosion happens, how it manifests itself and how the relevant methods of corrosion prevention and control operate. The course is suitable for many working in a corrosion-related field, This course is an integral part of the certification scheme developed by the Association. A pass in this course (or an equivalent) is necessary to become an ACA-certificated Corrosion Technician or Technologist. The only prerequisite is an understanding of secondary school year 10 level of chemistry and mathematics. This text is designed to complement the course. This course is designed for industry personnel and covers only the background theory necessary {o understand the concepts of corrosion and protection. Unlike many courses and textbooks, these notes concentrate on the forms of corrosion and, more importantly, bow itis dealt with. Anyone wanting more background electrochemistry should seek out the basic (exts in the Further Reading. No book can claim to cover more than a small portion of the subject and the books in the Further Reading chapter should be consulted for more information ‘There have been a number of changes for the 31d edition of these notes. All chapters have been completely re-written, with extra figures and tables wherever possible. Provision of useful information to the corrosion practitioner has continued, There have been major expansions in the sections on forms of corrosion, stainless steels, paint coatings and inhibitors. Background theory explaining polarisation data has also been expanded. Unfortunately, to keep the course to an acceptable Jength, the chapter on metallurgy is no longer an integral part of the course or notes, but has been amended and moved to an Appendix. The book describes corrosion and its mitigation in general terms, applicable to a wide range of industries. The information in this book is largely taken from a number of text books and from the notes to other ACA courses. However, any errors are mine and | would be grateful to readers if they could point them out. Any other comments would be welcome. Rob Francis Melbourne, May 1998.Chapter 1 INTRODUCTION In this chapter, a definition of corrosion, a brief description of the costs and other implications of corrosion and the various expressions used for corrosion rates are given 41. WHAT IS CORROSION? Corrosion can be defined as the destruction or deterioration of a material because of reaction with its environment. The term can refer to a process or the damage caused by such a process. It is important to note: + In some cases of corrosion there is no visible deterioration. The effect of the reaction with the environment is to change the properties, ofien the strength, of the material Such forms of corrosion are not common in day-to-day experience but are of great importance to the corrosion engineer. = Loss due to purely mechanical forces is known as wear. However, there are cases when corrosion and loss by mechanical forces combine and these are described in these notes. + The definition applics to materials other than metals. However, the term is usually applied to metals. (Metallic corrosion is mainly considered in these notes). The mechanisms of corrosion of non-metallic materials are governed by different physical and chemical principles. + Rusting is a term reserved for corrosion of iron and steel. Rusting of iron and steel is perhaps the most common example of metallic corrosion. There are many well-known examples such as the rusting of exhaus! systems and bodies of cars, water pipes and various types of structural steelwork, There are other examples of corrosion in everyday life. Silver tamishes by reacting with sulphur to form black silver sulphide. Copper roofs develop a green patina by reacting with various components of the atmosphere. Boat owners will be aware of a number of corrosion problems that can arise in a marine environment, even in metals such as stainless steels which are often erroneously believed to be corrosion resistant. However, there are many other comosion problems ‘occurring in industry, especially in the petroleum and chemical industries. The average person would be totally unaware of them. LdINTRODUCTION While corrosion is largely a harmful process, there are some desirable and beneficial cases ‘The green patina on copper mentioned above is generally considered to be pleasing to the eye. Chemical or electrochemical machining can be used to machine hard and difficult-to- muchine parts economically. Anodising of aluminium is a beneficial corrosion process used for obtaining a better appearance and corrosion resistance. Corroding ctchants are used in metallurgical investigation to enable examination of metallic structures. Perhaps the most common example of beneficial corrosion is in batteries. Many of the chemical reactions Which occur in batteries are corrosion reactions. 2, COSTS AND IMPORTANCE OF CORROSION Corrosion is a costly and wasteful process and the costs of corrosion to society are enormous. They can be divided into direct costs and indirect costs. Some examples of direct losses due to corrosion + Costs of anti-corrosive painting or other protection. + Costs of repair or replacement of corroded components or equipment + Overdesign to allow for corrosion. + The requirement 10 use more expensive materials, for example stainless steel rather than carbon steel Examples of indirect costs include: + Contamination of a product; for example traces of metal may alter the colour of dyes or food may be spoiled in corroded metal containers. + Loss of product from a container that has corroded + Damage caused by leakage of a material such as oil or water from a corroded vessel + Loss of production because of a corrosion failure or loss of efficiency because of, for ‘example, corrosion product reducing heat transfer. Replacement of a corroded tube in a plant may cost only a few hundred dollars, but shutdown while repairs are underway may mean 2 cost of many thousands of dollars in lost production, + Effect on safety. For example, failure may cause fire, explosion or release of a dangerous product, + Appearance. A corroded material is unpleasant to the eye so commercial establishments may use expensive trim for the sake of appearance. In most developed countries the cost of corrosion has been estimated as around 3 to 4% of the GNP. The costs of corrosion in Australia were estimated to be of the order of $8,000 million at 1982 prices. However, corrosion is a natural process, like water running downhill, and money and effort will always have to be spent to control it. What is perhaps mote important therefore is that around $2,000 million or one quarter of the cost could have been saved each year in Australia if the best corrosion practices were applied. This means using better protective systems, improving design, better materials selection, and other methods discussed in detail in these notes. Corrosion is of increasing importance to scientists and engineers. There are a number of reasons for this. Despite the many developments in plastics and ceramics, metals are being used increasingly in all fields of technology. Furthermore, metals such as cast iron which have been widely used in the past could tolerate a much greater degree of corrosive attack 12INTRODUCTION than the newer metals which have replaced them in many applications, Rarer and more expensive materials are being increasingly used in special applications, such as aerospace, electronics and atomic energy, and it is important that such metals do not deteriorate. There are increasingly stringent environmental and health and safety regulations meaning failures, breakdowns and unplanned emissions are becoming more and more unacceptable Finally, some environments are more corrosive due to new pollutants, although many atmospheres are now less corrosive than they were in the past due to clean air provisions. As well as having technical and economic consequences, corrosion has environmental ramifications. Destruction of material is a waste of valuable natural resources. It has been calculated that in the UK, one tonne of stecl is converted completely into rust every 90 seconds, Manufacture of metals from their ores consumes vast amounts of energy and by making products last longer, energy resources are conserved and there is less pollution, including that from greenhouse gases. Failure of pipelines, ships or other vessels by corrosion ean yesult in pollution of the environment by oil, chemicals or other damaging materials. Corrosion control is of great benefit to the environment. 3. EXPRESSIONS FOR CORROSION RATE Different materials can be compared on the basis of their corrosion resistance and this can be measured in a number of ways. Weight loss experiments are most commonly used. A specimen is cleaned, degreased, measured, weighed, exposed to the corrodent for 2 known time, cleaned to remove the corrosion products and reweighed. There are many expressions for corrosion rate. Milligrams per square decimetre per day is commonly used in the scientific literature and is an expression which includes the effect of exposed area and duration of exposure. However, those that give the rate directly in terms of penetration, such as millimetres per year, are best from an engineering stand-point as they can be used directly to predict the life of a given component. ‘The metric expressions for corrosion penetration are micromeires (microns or 1m) per year for most corrosion rates although millimetres per year may be required for rapid corrosion rates. Mils (thousandths of an inch) per year (mpy) is common in many corrosion books and literature from the USA. The expressions are calculated from weight-loss data using the following equations: ym Ww mm, v — = 87600 —— or = 876 — yr DAT yr DAT where W: weight loss in milligrams, D: density in g em”, A: area in cm? and T: exposure time in hours. ‘As described in later chapters, corrosion rates are not constant and will usvally reduce with time although eventually reach a steady-state rate. Therefore, when looking at corrosion rates, itis important to know if they refer to early years of comrosion, such as the first year, or long-term rates. When investigations are being carried out, experimenters usually only wait a year or so, so will determine first year rates. However, for engineering purposes, long-term or steady-state figures are of more use, ISO 9223 and ISO 9224 gives five 1:3INTRODUCTION categories of corrosion rates for atmospheric corrosion and typical figures for various metals. Table J-1 from these standards compares the corrosion rate (both first year and steady state) for mild steel in the five ISO categories. Higher corrosion rates, up to some millimetres per year for steel, are found in chemical environments and severe natural environments such as splash zones. Table 1-1: Comparison of corrosion rates for mild stee! (in pm/year) in various atmospheric environments 1sO Corrosivity __‘First____‘Steady Typical environment category year state Ci Very tow <13 0.1 Diy indoors C2 Low 13:25 0.1.1.5 Desert to non-polluted urban C3 Medium 25:50 1.56 _ Mild marine or mild industrial ca High 50-80 6-20 Marine (near calm sea) cs Very high 80-200 20-90 _Marine (near surf beach)Chapter 2 THE CORROSION PROCESS In this chapter, the following introductory aspects of corrosion will be covered: + the essential features of the corrosion cell + a description of the Galvanic Series + the importance of the cathode in corrosion reactions + how steel plate corrodes 41. THE COPPER-ZINC CORROSION CELL Corrosion is the reaction between a metal and its envi ronment to form chemical compounds. lis action is known as electrochemical or electmrolytic and arises due to the formation of a cell similar to that of a battery. The simplest caorrosion reaction can be best explained by considering the operation of a simple battery. (Other, more common, corrosion reactions arise in a slightly different manner as explained later). Figure 2-1 shows a battery consisting of a plate of zinc and a plate of copgoer immersed in a solution of salt water. If the plates are not electrically connected io each =>ther no reaction takes place. As soon as they are connected, an electric current flows thro-agh the completed cireuit. At the same time, oxygen dissolved in the solution is consumed at the copper plate while the inc plate dissolves or corrodes to form positively-charged ziac ions. As the zine dissolves into the solution, electrons are released. These flow through “the circuit, lighting the lamp and are consumed by reactions at the copper plate. If a vo“Atmeter replaced the lamp, a voltage or potential difference can be shown to be generated “Wetween the zinc and copper. A corrosion cell has thus been created. The types of reactions occurring in this cell are the same Gas those which occur in any corrosion process. This cxample illustrates the four essential constituents of @ corrosion cell. 1. The Anode ‘This is the zinc plate in the above example. Put simply, the anode is the site at which the metal is corroded. At the same time, electrons are produced. ‘This reaction can be shown as: Zine — Zine ions + electrons 21THE CORROSION PROCESS or, in terms of chemical symbols: Zn > Zn" 4 2e J 2ne | —~ — Copper | (nosey ~~ (Cathode) Figure 2-1: The corrosion of zinc in a simple voltaic cell 2. The Cathode This is the copper plate in the example. This is not consumed but serves to get rid of the electrons produced by the corroding anode. These electrons react with the oxygen in solution, and water molecules, to form negatively charged particles known as hydroxyl ions. This reaction can be shown as: Oxygen + Water + electrons + Hydroxy! ions or in chemical terms as: 40, + H,0 + 2e > 20H" 3. The Electrolyte This is the salt solution in the example. This contains positively and negatively charged ions and will conduct electricity. The corroding anode produces positively charged ions which flow into the electrolyte. At the cathode, positively charged ions or neutral species take up the electrons flowing in the electrical circuit. In the example, oxygen and water take up the electrons to produce hydroxyl ions. The flow of ions in the electrolyte completes the electrical circuit 2-2THE= CORROSION PROCESS: 4, The Conducting Link Electrons need to have a means to travel from the anode to the e==athode. In the example, this is a wire between the two but this could just as easily be provi. ded by connecting the two metals together 2. OTHER GALVANIC CELLS ‘Whenever there is a cathode, an anode, a conducting link and electrolyte, corrosion will occur, Consider a piece of copper joined to a piece of iron in = he presence of water. Here the iron is the anode and corrodes (rusts), the copper is the cath—>de and the water acts as an electrolyte, The anodic reaction in this case is: Fe > Fe" + 2c While the cathodic reaction is the same as previously: YO, + HyO + 2e" > 207° In this case, the hydroxyl ions react with the ferrous ions to foramen ferrous hydroxide: Fe + 20H > Fe(OH) Ferrous hydroxide is a white, milky precipitate which rapid. Ty oxidises in air to ferric hydroxide (Fe(OH),). This is also unstable and reacts with fx_arther oxygen and water to form various oxides and hydroxides of iron, collectively know=m as rust. Rust does not have a single chemical formula as its composition depends on = Hy ‘The two cathodic reactions mentioned so far are by far the mom st common, although others are possible, In a corrosion reaction, the anodic reaction is al==vays the metal dissolving to form metal ions. ‘Whenever two dissimilar metals are connected in the presence of an electrolyte (which can be humidity from the atmosphere), the anode will corrode amd the cathode will not. Of 23THE CORROSION PROCE! course, it is important to know which is which and to determine this the relative tendency for metals to corrode must be known, This can be measured by determining the voltage generated by the clectrode in a given environment against a reference electrode. Different metals will give different voltages depending on their tendency to dissolve. 3. THE GALVANIC SERIES The Galvanic Series is obtained by compiling the potentials of different metals measured against a reference electrode in a list. The Galvanic Series of some important metals is, given in Table 2-1 and gives a measure of their tendency to dissolve. If there is @ pair of metals from the list joined together, it is always the one below (or the more aetive metal) which becomes the anode and dissolves while the one above (or more noble metal) becomes the cathode and stays intact. The further apart the two metals are on the list, the greater the Griving force that affects the corrosion reaction. In some books and literature, especially from the USA, this table is printed with the most active metal at the top and most noble metal at the bottom. Table 2-1: The Galvanic Series for metals in aerated sea-water. [The GALVANIC SERIES in sea water (Cathodic or noble end) platinum gold carbon (graphite) titanium Type 316 or 304 stainless steel (passive) Monel metal (70 % nickel, 30 % copper) silver nickel lead bronze, copper, brass tin leaditin solder ‘Type 316 or 304 stainless steel (active) cast iton, mild steel cadmium aluminium alloys pure aluminium zine magnesium (Anodic or active end) The galvanic series shows why the two cells already described operate in the way they do, Copper and zine and copper and steel are far apart and the corrosion reactions rea occur. If however, copper was connected to silver, then the metal that corrodes would be the copper. Another example of galvanic corrosion, the corrosion arising when different metals are joined together, is mild steel electrically connected to stainless steel, Mild steel is the anode and corrodes while the stainless steel is the cathode and will be protected. While this simple series is often sufficient to predict galvanic behaviour, more sophisticated tables, such as thosc given in AS 4036, give more detailed information on corrosion behaviour of galvanic couples. 24THE CORROSION PROCESS Jn some environments, some metals become covered with a corrosion layer such as an oxide which protects the metal from further corrosion. The metal is said to be passivated, and if now placed in a different corrodent the corrosion potential is different to that exhibited by the unpassivated material. This explains the different positions of the same alloy of stainless steel in the series. Under those conditions where the protective oxide film is not kept intact, such as in crevices or de-aerated water, the stainless steel has a similar position on the table, and similar reactivity, to mild steel. When the oxide film is intact, in “oxygenated environments such as a clean atmosphere, stainless stecl behaves as a more noble metal. Is not only different metals that cause galvanic corrosion. A similar problem occurs with steel covered with mill-scale. Mill-scale, a mixture of various oxides of iron formed when the steel is hot worked, is strongly cathodic to steel. Therefore, corrosion will be concentrated at breaks in the mill-scale which will act as anodes and there will be a very large cathodic area for oxygen reduction to take place. This is one of the reasons why complete removal of mill-scale is so important in surface preparation. 4. IMPORTANCE OF THE CATHODE So far, reactions occurring at the cathode have not been of great concem. Do they matter? As every electron produced at the corroding anode must be taken up by the cathode, the reactions which take place here are indeed of importance and will significantly influence the corrosion rate. In the copper-zine battery, the oxygen in the water takes up the electrons produced by the anodic reaction to form negatively-charged hydroxyl ions. In atmospheric corrosion, oxygen in the air and moisture from the bumidity react in exactly the same way. Under acidic conditions, hydrogen ions (H’) in the acid react with the electrons to form hydrogen gas. In all cases, gases are involved. Under these circumstances, the arca of the cathode is crucial. If there is a large cathode, then the cathodic reaction can easily take place ~ either producing hydrogen or converting oxygen to hydroxy! ions — and electrons will be used up as fast as the anode can produce them. If, however, there is only a small cathode, then only a small amount of gas can get to ot leave the cathode and this will slow down the anodic reaction. Thus, the nature of the cathodic process has a major influence of the rate of many corrosion processes. Compare, for example, a structure with copper rivets holding large stee! sections together with a structure with steel rivets holding large copper sections together. The first is acceptable because, although iron is definitely anodic to copper, the rate at which air can get to the small cathode is very slow and the corrosion rate is negligible. But in the second case there are large areas of copper available for the cathodic reaction to take place. As a result, the iron is going to rust as quickly as it can get info solution, which is very rapid indeed. Failure of steel rivets can occur within a few days in an aggressive environment. A similar problem: if a steel pipe had to be joined a to a copper one, which one should be painted, the steel or the copper? The copper should be painted. The cathode will then be very small; in fect only those regions that the paint has not covered. The anode will be 2-5THE CORROSION PROCESS Jarge and corrosion will be minimised. If dissimilar metals must be joined, the anode should be large and the cathode small. 5. THE CORROSION OF STEEL So far, only one form of corrosion, galvanic corrosion which arises when different metals are joined, has been considered. However, there are other forms of corrosion where the anode and cathode arise due 10 means other than different metals in contact. Exactly the same principles are involved — there must be an anode, cathode, electrolyte and conductor and the area ratio between the anode and cathode is important, for example. However, the anode and cathode are not as immediately obvious as in galvanic corrosion. For example, a piece of steel may appear to be uniformly corroding without anodes and cathodes. In fact, there are a number of ways that anodes and cathodes can arise on a piece of stecl. As a result, there will be areas that are anodic and corrode, and areas that are cathodic and not corrode. Figure 2-2 shows these cells on a piece of steel. The electrolyte must cover both anodic and cathodic regions as for galvanic corrosion but in this case, the electronic conductor arises simply because the metal is a conductor and electrons can easily flow from the anode to the cathode. The anodic and cathodic sites do not have to be side- by-side, but can be any distance apart, as long as there is an electrolyte and conducting, path. peste negations Stee! Figure 2-2: Formation of an anode and cathode on a piece of ste! ‘One way that the celis arise is duc to the fact that steel is not a uniform material. Under the microscope, steel is seen to consist of crystals of different compositions. These crystals may be anodes or cathodes and lots of tiny corrosion cells can form on the surface. Also, any impurities in the steel, such as sulphide and silicate particles arising from steelmaking, will be cathodic sites Another means of creating anodes and cathodes on steel is to weld, cut, drill, grind or bend it. This is because these worked areas tend to become anodic to the unworked areas. Placed in an electrolyte, a corrosion cell arises and the steel corrodes. Another source of anodes and cathodes is different concentrations of chemicals around the metal. The area with the lower concentration of a chemical will be anodic to the area with the higher concentration, This can be a problem in chemical plants and in stagnant flow conditions. One of the most common ways that corrosion cells can occur on homogeneous metals such as steel is because of differences in oxygen concentration. An area with low oxygen 26THE CORROSION PROCESS concentration will become anodic to an area with a higher oxygen concentration forming what is known as a differential aeration cell. This can causc corrosion in @ number of ways. Figure 2-3 shows examples of regions where coxrosion occurs in inaccessible places. For example, in bolted or riveted plates, the exposed regions become cathodes and hidden regions anodes. Similarly, the exposed part of @ pipeline will be cathodic to the buried part The bottom part of a buried pipeline will have a lower oxygen concentration and corrode while the top of the pipe remains cathodic and free fom corrosion. A region of pipeline buried in compact clay will be anodic to a region buried in less compact sand, The anodic and cathodic arcas can be any distance apart, even many kilometres, as Jong as there is an electrical connection between them. ~ cathode a Clean Metal pee Anode Cathode ‘Anode Figure 2-3: Corrosion occurs in inaccessible regions dus to differences in oxygen concentration. In atmospheric corrosion and other uniform environments, the anodic and cathodic sites on stee] are usually not constant but changing all the time. A build up of rust near cathodic sites prevents oxygen reaching these sites and they then tum into anodes. The anodes, which now have access fo the oxygen, in tum change to cathodes. As a result, corrosion over the surface is uniform and rusting continues unabated. The layer of corrosion product is loose and porous and fresh rust will continue to form at the metal-rust interface. This is known as uniform corrosion and is generally the most common form of corrosion, However, in 2 number of situations this change from anode to cathode cannot occur and localised corrosion such as pitting or erevice corrosion will arise. This occurs when the anodes are small compared to the cathodes and this swapping of functions cannot take place, such as in the bolted plates in Figure 2-3. These forms of localised corrosion are far more insidious and of more concem to the corrosion engineer.Chapter 3 PREDICTING CORROSION REACTIONS Metallic corrosion arises because of the presence of anodes znd cathodes. However, it is well known that, for example, steel corrodes in the atmosphere while copper or stainless steel generally do not, The corrosionist needs to know why cestain metals corrode or, more importantly, how to stop them corroding. To understand these fundamental properties of metals there are two main areas of study, Firstly, whether or not a given metal can corrode in @ given environment must be known. It is chemically impossible for some metals to corrode in a given environment. This subject is known as thermodynamics. The second area is how fast will it corrode. A meta] may be able to corrode but its reaction rate is so slow that it is effectively corosion-resistant. This subject is known as kinetics. A basic understanding of the principles of these two subjects is very useful 10 anyone wanting to understand the corrosion behaviour of metals. This chapter introduces these two subjects as they relate to corrosion. 4. OXIDATION AND REDUCTION Metallic corrosion involves conversion of a metal to its ions along with the production of electrons. For example, corrosion of iron (Fe) to ferrous ions (Fe"’) occurs through the reaction: Fe Pe" + 2¢° Fg 3.1 ‘This occurs at the anodic sites. At the same time, the electrons are taken up at the cathode. ‘There are a number of possible cathodic reactions, but in oxygen-free acids for example, the reaction is 2H" + 2e + Hy Eg 3.2 The overall reaction is Fe +2H* > Fe" +H) Eg3.3 These types of reactions are called oxidation-reduction (or redox) reactions. The iron is converted to positive ions with the production of electrons by a process known as oxidation. Whenever an atom (or ion) loses electrons, the reaction is termed oxidation. Reaction with oxygen, such as when a metal bums in air, is an oxidation reaction, thus the name. However, oxygen does not have to be involved for oxidation to occur. Whenever 31PREDICTING CORROSION REACTIO! oxidation takes place, there must be another reaction to take up electrons. In the above example, this was the reaction between hydrogen ions and electrons to form hydrogen gas. If 2 metal ion in solution takes up electrons to form a metal atom (the reverse of the corrosion reaction) as will occur in, for example, electroplating, the metal ions are reduced. Any reaction in which an atom (or ion) gains electrons is termed reduetion. In the overall reaction 3.3 above, therefore, the iron is oxidised to ferrous ions and the hydrogen ions are reduced to hydrogen gas. The hydrogen ions oxidise the iron, so 4 solution containing hydrogen ions, or any other compounds that can be reduced, is termed an oxidising agent In a corrosion cell, oxidation occurs at the anode and reduction at the cathode. 2. ELECTRODE POTENTIALS From the corrosion point-of-view, it is important to know how easily a metal can be oxidised, that is lose its electrons. Experience tells us that metals like gold and platinum are very hard to oxidise while metals such as iron and zine are much easier to oxidise. Some measure of the relative ability of a metal to be oxidised must be known. At the same time, oxidising agents have different ‘strengths’. For example, copper will not corrode in hydrochloric acid, but will in nitric acid. That is, nitric acid has a stronger oxidising power than hydrochloric acid. Determining relative oxidation-reduction power is of great importance and, in fact, relatively casily to quantify. If a metal is placed in a solution, there will be a tendency, which may be weak or strong, for metal atoms to change to ions and go into solution. This tendency will create a potential difference (or voltage) at the surface of the piece of metal between the metal and the solution. It would be very convenient if this tendency to ionise could be directly measured There is, however, no practical method of evaluating the absolute magnitude of this potential difference since completing the electrical circuit to measure the potential would introduce a second metal/solution interface with a second unknown potential difference. So instead, the relative potential is measured against a second electrode which is arbitrarily assigned a potential of zero. If the picce of metal under examination is connected to a voltmeter, and the other voltmeter terminal connected to this reference electrode, the potential developed will be a measure of the ease of oxidation of the metal (M) to metal ions (M’) (see Figure 3-1). Voltmeter Reference Vv electrode Figure 3-1: Measurement of a single electrode potential ‘The observed voltage will be the sum of the potential difference across the reference electrode plus that across the metal/solution interface. Since the potential difference across 3-2PREDICTING CORROSION REACTION the reference electrode has been assigned as ero, the measurement ean be reported as the single electrode potential for the reaction: M+e=M By carrying out such experiments with a series of metals, a list can be drawn up called the Electrochemical Series which shows the relative abilities of metals to ionise. ‘Table 3-1, also called the EMF Series, the half-cell potentials or oxidation-reduction potentials, shows the series with the hydrogen reduction reaction arbitrarily given a potential of zero. The concentration of metal ions in solution influences the potential so this is fixed, as is the temperature. Under these conditions, the potential is known as the Standard Reduction Potential and given the symbol E°. The more negative the potential, the greater the tendency for metal ions to oxidise, or go into seolution, i.e the reaction goes in the reverse direction to that shown in the table, Such metals are called active, base or anodic. The more positive the potential, the greater the tendency for the metal to remain stable, or the reaction proceed in the forward direction, as shown in the table. Such metals are termed noble or cathodic. Thus magnesium, with a reduction potential of -2.37 volts, is highly active and readily corrodes, while gold, with a reduction potential of +1.70 volts, is noble and has very little tendency to corrode. Table 3-1; The Electrochemical Series Electrode Reaction Standard Reduction Potential E® (volt) Cathodic or Noble End Autre + 1.70 Ch, + 20 + 1.36 BO, + 2H" + 2e + 123 Pitt 26 + 492 Ho?" + 26 + 0.85 Ag) +e + 0.80 Fe" +e + 07 YO, + H,0 + 26 + 0.40 Cu + 26° + 0.34 2H + 26° 0.00 {by defintion) Poe +26 - 0.13 NP +20 ~ 0.26 Co +20" = 040 Fer +26 - 0.44 Cr +3 - O74 Zn?" + 20° - 076 Te +26 = 163 AM + 3e ~ 1.66 Mg + 26 ~ 237 Na‘ +e = 271 Active or Anodic End Not only corrosion reactions are shown in the table. For exaraple, the oxidation of ferrous ions (Fe) to ferric ions (Fe) is also shown. So, in a strongly oxidising environment, 3:3PREDICTING CORROSION REACTIONS iron will be oxidised, first to ferrous ions and then to ferric ions, This process occurs in atmospheric rusting. Also shown are reactions involving hydrogen and oxygen gase ‘These commonly occur as reduction reactions, that is reactions accepting electrons. Any of the metals below the hydrogen reaction on the scale, will be oxidised by hydrogen ions. ‘This means those metals will dissolve if placed in an acid solution, As a result, the cathodic reaction will be reduction of hydrogen ions (reaction as written in the table) and the anodic xeaction will be oxidation of the metal (reverse of the reaction as written in the table) Another reaction involving a gas is reduction of oxygen gas. This reaction has a very positive potential in acid solutions, and a less positive potential in alkaline solutions, so ‘oxygen is a powerful electron acceptor and, under the right conditions, can oxidise any of the metals below it on the table. Chlorine gas in water is also seen to be 2 very strong oxidising agent. 3. APPLICATIONS AND LIMITATIONS OF THE ELECTROCHEMICAL SERIES The Electrochemical Series provides a great deal of useful information to the corrosionist. It can be used to determine whether corrosion reactions will occur and what reactions will take place. It can also be used to determine the corrosion potential between the anodic and cathodic reactions, with the greater the potential, the greater the driving force. To use the Series in real applications requires corrections to take account of the fact that the concentration of metal ions significantly affects potentials. These corrections are outside the scope of these notes, but first approximation predictions can be made using the Table. in any electrochemical reaction, the most negative or active half-cell tends to be oxidised, and the most positive or noble half-cell tends to be reduced. This fact can be used to predict electrode reactions and some useful functions are discussed below. @__ The displacement of hydrogen. Noble metals above hydrogen in the Electrochemical Series such as gold, platinum etc. are incapable of liberating hydrogen when the metal is placed in de-aerated acid. Only the base metals, below hydrogen in the series, will liberate hydrogen and therefore corrode in acid solutions. So that zinc or iron, for example will dissolve in de-aerated acid while copper and silver will not Aerated acid, with oxygen present, is a much stronger oxidiser and will dissolve copper and silver. Gi) ‘The displacement of a metal. One metal will displace another above it in the series. The driving force for the displacement increases with increased separation of the metals in the Electrochemical Series. Example: A rod of zinc in copper sulphate solution causes zine ions to form and copper to be plated out. For the copper-zine system, the two reactions are: Cu +2e"= Cu B°=+034V Zn" +2 = Zn E°=~0.76V The more positive reaction proceeds in the forward direction and the more negative reaction in the reverse direction so the overall reaction is: Zn+ Cu" Zn + Cu 34PREDICTING CORROSION REACTIONS, However, no reaction occurs if a copper rod is placed in a solution of zine ions. (iii) Electrolysis. The series can predict the order which metals will plate out from a solution containing a mixture of ions. The more noble metals deposit first. Example: In an acidified solution containing H’ and Cu"* jons, copper is liberated first. But in a solution containing Na’ and H’ only hydrogen is liberated. There are, however, a number of limitations of the Electrochemical Series. (i) The series assumes reversible, equilibrium conditions which means the reactions are allowed to come to a steady state condition no matter how long this takes. This rarely ‘occurs in practice, certainly not under corroding conditions, Gi) Only pure metals are listed because it is not possible to establish a reversible potential for alloys which contain two or more reactive components. However, most ‘engineering applications use alloys rather than pure metals, Gil) The series requires the metal to be in contact with its ions of given concentration. In practice, the ion concentration varies widely. (iv) Surface films form under a variety of conditions. For example, chromium, titanium and aluminium are more active than iron and the series suggests they are more liable to corrode. But it is well-known that these elements are far more corrosion resistant than iron. This is because they form protective surface films. () The series takes no account of the effect of pH. It only takes into account the electrochemical equilibrium involving metals and their simple ions. But there are many other possible reactions which can occur in practice. For example, with iron Fe + 20H > Fe(OH), may occur at high pH and form a protective film. (vi) Temperature is important. The series is applicable only at 25°C. A change in temperature could result in an increase or decrease in the difference between normal electrode potentials. 4, ELECTROCHEMICAL SERIES VS. GALVANIC SERIES. Examination of the tables of the Electrochemical Series and the Galvanic Series shows that these two tabulations are similar. Both can be used to determine corrosion behaviour, although the Galvanic Series is of more practical use. There are however, significant differences between the two: (The Electrochemical Series is used to predict whether corrosion of a single given metal will occur, By contrast, the Galvanic Series is used to predict whether galvanic 3-5PREDICTING CORROSION REACTIONS corrosion will occur when two metals are coupled togethes and, if so, which metal shows increased corrosion. qa) 1c Galvanic Series shows metals and alloys whereas the Electrochemical Series only shows pure metals. Gil) The Galvanic Series shows passive metals such as aluminium and stainless steels whose electrochemical behaviour is based on a passive film rather than the active underlying metal (iv) The Galvanic Series is determined in a given electrolyte, such as sea Water, whereas the Electrochemical Series uses a solution of the metal ions of a fixed concentration. (¥) Corrosion products can modify the behaviour of metals in the Galvanic Series and some experience is needed to predict corrosion rates, 5. REFERENCE ELECTRODES A reference electrode is an electrochemical cell which has a stable and reproducible potential so can be used as a reference point for potential measurement. It is used in an electrochemical circuit to measure the electrochemical potential of a metal-clectrolyte interface. A high impedance voltmeter or potentiometer is required to ensure current flow is minimised which would interfere with the reading. The negative terminal is connected to the reference electrode, the positive terminal to the metal and the potential is read. It was stated earlier that the standard potential for the reaction 2H + 2e° > Hy, is assumed to be equal to zero under standard conditions. The hydrogen electrode is made of an inert platinum electrode with hydrogen gas at one atmosphere bubbled over it, The surface is ‘weated (platinized) to allow rapid electron transfer to occur at the surface. Despite the convenience in potential measurements, the standard hydrogen electrode is rather awkward to use and others are extensively used, particularly in the field (i) Calomel Reference Electrode This has long been used in the laboratory. It consists of a platinum wire in contact with pool of mercury covered with a layer of mercurous chloride (calomel) and immersed in a solution of potassium chloride (KCI) acting as the electrolyte. A porous plug in the end allows passage of ions without causing significant cross contamination of the potassium chloride and the test cell electrolyte. The potential depends on the chloride concentration so there is the ‘saturated’ calomel electrode with a saturated solution of KCI and the ‘normal’ calomel electrode using 1M KCl. The saturated electrode is the most widely used. (ii) Copper - Copper Sulphate Electrode This is commonly used to measure the electrode potential of buried metal structures. It is Jess accurate than the calomel and not used in laboratory work. It consists of a copper rod immersed in a saturated copper sulphate solution in a porous tube or a container with a porous plug (Figure 3-2(a)). 3-6PREDICTING CORROSION REACTIONS Gii) Silver -Silver Chloride Electrode For Jaboratory work, this consists of a platinum wire coated with silver and silver chloride enclosed in a suitable glass tube (Figure 3-2(b)). Such electrodles can be made quite small, more so than with the calomel electrode. Again the potential over silver over Copper sulphate eau pletinum Porous plug Protective ea (omoved tr @) o Figure 3-2: Reference electrodes (a) Copper - Copper Sulphaie electrode, (b) Silver - Siver Chioride electroce. Gv) Other Electrodes A lead/lead chloride electrode has been used as a reference electrode for measuring potential on lead covered cables. A rod of pure zinc may be used as a reference electrode for rough determinations in soil. Without a solution container and porous plug, it is not subject to contamination but is subject to variations of up to 50 millivolts, far greater than with other reference electrodes. Steel of similar composition to a protected structure, sometimes cast in a concrete block, is also occasionally used. It has similar limitations to the zine electrode. 6. CONVERSION OF ELECTRODE POTENTIALS The potentials of the commonly used half-cells versus the Standard Hydrogen Electrode (SHE) and the Copper/Copper Sulphate Electrode (CSE) are listed in Table 3-2. Converting from one to another can be confusing. However, it may help understanding different potentials to think of each electrode in the same way that heights of mountains are 37PREDICTING CORROSION REACTIONS measured. The standard reference point is sea level, which is analogous to the hydrogen electrode. For example, Mt Kosciusko is 2230 metres above sca level. For someone in Canberra, however, they may wish to know the height of Mt Kosciusko from their reference level. Canberra is 580 metres above sea level, so Mt Kosciusko must be (2230 — 580 =) 1650 metres above Canberra, Converting from one reference electrode can be done in a analogous manner, although the signs can become confusing. It may be best to think of negative values as depths below sea level, and positive values as heights above sea level. Table 3-2: Reference Electrodes compared to the Standard Hydrogen Electrode (SHE) and Copper/Copper Sulphate Electrode (CSE) Electrode Potential vs CSE CopperiCopper Sulphate (satd) (CSE) 4 0316 00 Normal Caiomel (NCE) + 0.280 = 0.036 SilveriSilver Chloride in sea water (SSC) + 0.25 approx. = 0.065 Saturated Calomel (SCE) + 0.242 0.074 Standerd Hydrogen (SHE) 0.000 (by defn) - 0.316 Zine 0.80 approx 1.416 Following the above analogy, to determine the potential of a cell measured against a reference electrode other than the SHE, it is easiest to draw a scale showing the reference electrodes and use this to calculate potentials (see Figure 3-3). +0316 CuCuso, + 0.280 Normal Calomel +0.25 Ag/AgCl (sea water) + 0.242 Saturated Calomel 0.0 Hydrogen - 0.80 Zine Figure 3-3: Diagram for conversion of electrode potentials (not to scale). Example 1: Calculate the potential difference across a cell comprising 2 saturated calomel electrode and a standard iron half cell (E? =-0.44 vs SHE). Mark the figure for the iron half cell on the diagram. Looking at the diagram, the magnitude of the potential difference between the iron and calomel cells must be greater by 0.242 volts. The iron is more negative than the calomel electrode, so the sign will be negative. So the potential wil] be Been = By ~E; = (0.440) - (+0.242) =-0.682 volts 3-8PREDICTING CORROSION REACTIONS Example 2: The potential of a pipe relative to the Cu/CuSO, electrode is -0.700 volts (ie the pipe is at a Jower potential than the reference). What is the potential of the pipe relative to the SHE? Mark in the pipe potential 0.700 volts below the copper cell. The copper cell is 0.316 volts more positive than the hydrogen cell. Therefore, the pipe potential versus the SHE will be more positive by 0.316 volts than it is versus the copper electrode. Epipe ~ —0.700 40.316 pipe Eyipe = 0.384 volts, ine 7, POURBAIX DIAGRAMS The Electrochemical Series is very useful to electrochemisis in predicting reactions but there are a number of limitations, especially to the corrosionist. For example, the EMF of the cell depends on the acidity and the concentration of the ionic species. Furthermore, corrosion products may build up on the metal surface and stifle the reaction. For example, aluminium and titanium are both far more corrosion resistant than the series suggests due to the formation of a protective oxide film. These limitations have been overcome to some degree by the use of the Potential-pH diagram, more commonly known as Pourbaix diagrams after their originator. Only one metal can be clearly represented on one diagram but the data is of such value that they have been prepared for most metallic elements. ‘The Pourbaix diagram combines the information in the Electrochemical Series with the effect of pH and the oxides and hydroxides of the metal which may form. As a reminder, PH is simply a measure of the acidity of the solution, with « pH of 7 meaning a neutral solution, a pH less than 7 an increasingly acid solution and a pH of more than seven an increasingly alkaline solution. The Pourbaix diagram indicates the compounds which are stable at various pH values and various potentials. The potential scale can be thought of as @ measure of the relative oxidising power of the environment. More positive values indicate more oxidising conditions or contact with noble metals while reducing conditions, such as lack of dissolved oxygen or contact with a more active metal, bring about a change in potential in a negative direction. A potential-pH diagram will show four regions — acid/oxidising conditions in the top left hand corner, alkaline/oxidising in the top right hand corer, acid/reducing in the bottom lefi and alkaline/reducing at the bottom right hand comer. Figure 3-4 shows a simplified Pourbaix diagram for iron in water. This shows the various species which are stable under different conditions. Under highly reducing conditions, at all pH values, iron is stable and will not corrode. This region is called the immune region. Under acidic conditions, iron is relatively easily oxidised to ferrous ions, or with a stronger oxidant, to ferric ions. These regions are termed the corrosion domain. Iron will also corrode under very alkaline conditions where the HFeO,’ ion will form. Under alkaline conditions, hydroxides (Fe(OH), or Fe(OH); are the stable species. In these regions, iron is not immune from corrosion but protective films means in practice corrosion does not occur. These are called the passive domains, 3-9PREDICTING CORROSION REACTIONS The diagrams will usually show two sloping straight lines which indicate the region of stability of water. At potentials above the oxygen line, water wil] be reduced and oxygen gas given off while at potentials below the hydrogen line, hydrogen ions will oxidise the metal and hydrogen gas will be given off. Line (1) in the iron diagram is below the hydrogen line, so for iron metal in aqueous solutions with pH values less than about 9, iron will be oxidised with the evolution of hydrogen gas. At pH] values from about 9 up to about 12.5, a passivating layer of Fe(OH), is formed. Oxid. 1.2 08 Potential vs SHE (V) 1.2 Red . 4 : Ofte erties tii On atta Acid pH Alkaline Figure 3-4: Pourbaix Diagram for the Fe-H.O system at 25°C. (Simplified version). As well as predicting whether or not corrosion will occur and the corrosion products which will form, the diagrams also enable us to predict environmental changes which will prevent ‘or reduce corrosion attack. For example, consider a piece of iron immersed in an oxygen- free solution at pH 6. The potential of the anodic sites will approximately lie on line (1) while the potential of the cathodic sites will approximate to that on the hydrogen line, both at pH 6. As a result, the potential of the sample will fall somewhere between these two equilibrium potentials, around -0.4 volts on the Hydrogen scale. This is in the corrosion (Fe™*) zone. There are then three ways the iron can be removed from the corrosion region: 1. By lowering the specimen into the region of immunity. This can be achieved by making the specimen the cathode in a galvanic cell or application of an external voltage. This is called cathodic protection and will be discussed in a later chapter. 2. By raising the potential into the region of passivity. There are a number of ways this can be achieved. For example, applying an external voltage to increase the potential in a similar manner to cathodic protection can be used and this is called anodic protection. Adding an oxidising agent causes a passivating film to form on the surface raising its potential. 3-10PREDICTING CORROSION REACTIONS 3. Increasing the alkalinity of the solution (raising the pl) also forms a passivating film. Figure 3-5 shows the simplified Pourbaix Diagram for aluminium-water. This shows that aluminium will corrode at pH] values lower than about 4 and higher than about 8.5 over & wide potential range. Between these limits, the oxide forms a passivating film, The diagram also shows that under strongly oxidising conditions, aluminium is passive, even under acidic conditions. This is the basis for the anodising process where a protective oxide film is grown on the surface by applying a positive potential in a sulphuric acid solution: Figure 3-5: Potential-pH diagram for aluminium. ‘The Potential-pH diagrams are useful means of cisplaying thermodynamic data for corroding systems but they do have inherent limitations which must be kept in mind. + No information is given on corrosion rate. + The term passivity is applied to oxides, hydroxides etc. irrespective of their protective properties. For example, hydroxides may not form on the metal surface and will not therefore be protective. + In corrosion, other ions such as chloride, sulphate etc. are important and will take part in corrosion reactions. + Production of H” and OH’ ions may cause highly localised pH changes at the electrode surface. Since they may not greatly influence the bulk pH, forecasts of behaviour based on bulk pH measurements may be misleading, + Impurities in the solution may influence reactions. + The effect of alloying elements is not considered. + The diagrams are only useful at room temperature, 3-11PREDICTING CORROSION REACTIONS 8. CORROSION RATE AND CURRENT FLOW ‘The potential difference between the anode and cathode is « measure of the driving force for the corrosion reaction but the rate at which the reaction will actually occur is determined by the corrosion cell current which wil) flow as a result of this potential difference. This current is often completely independent of the potential difference. There can be a large potential difference between the anode and cathode, yet only a small current flowing so a low corrosion rate. A study of corrosion currents is therefore of ereat importance in corrosion reactions. When an anodic reaction proceeds, electrons are liberated which are consumed by the cathodic reaction. The total current flowing out of the anodic reaction must be equal to the current flowing into the cathodic reaction to preserve elecirical neutrality. Faraday discovered that the amount of material corroded in an anodic reaction, or deposited in a cathodic reaction, is directly proportional to the quantity of electricity passed according to the equation: Bg. 3.4 Fn where m is the mass of metal dissolved or deposited in grams, ] the current in amperes, t the time of current flow in seconds, A the atomic or molecular weight, n the valency (number of electrons transferred per atom dissolved) and F a constant known as the Faraday (96,500). This equation may also be written: m=ZIt Eq. 3.5 where Z is a constant known as the electrochemical equivalent of the substance dissolved or deposited. Therefore, the current or quantity of electricity per unit time (assuming it can be measured) will be a measure of the rate of corrosion, or amount of metal dissolved per unit time, More importantly, the corrosion rate as weight loss per unit area can be easily determined if the current density, the number of amperes per unit area, can be measured. Current density is widely used in corrosion studies as a measure of corrosion rate. 9. POLARISATION During corrosion, current flows from the anode to the cathode and the greater the current, the greater the corrosion rate. If this current could be easily measured, the corrosion rate could be calculated using Faraday’s Laws as described in the preceding section. Unfortunately, this is not possible because corrosion current usually flows from a large sumber of small anodes on the metal surface to a large number of small cathodes, However, it is possible to indirectly measure current flowing between anodes and cathodes in the laboratory using special equipment as described in the following section, and thus determine corrosion rates. Such measurements mean corrosion processes can be explained in terms of potentials and currents. 3-12PREDICTING CZORROSION REACTIONS In « corrosion cel] with a very large resistance between the anocie and cathode, there will be a potential difference between the two electrodes but negligible cusrent will flow. Using a reference electrode such a calomel electrode, the potential of thm e anode and cathode can be measured separately. The potentials of pure metals measured w Bien negligible current flows arc known as the reversible potentials, The anode will have amnore negative potential than ‘the cathode. Once current starts flowing, the system is no Jonger in equill ibrium and the anodic and cathodic reaction potentials will differ from their reversible potentials. As a result of current flow, the potential of the cathode becomes less positive= (or more negative) and the potential of the anode becomes less negative (or more positive) and the electrode potentials are displaced towards one another. The anodic and cathodic «actions are said to become polarised. Polarisation is of great importance because it determeaines the practical corrosion rate of many metals and alloys. Figure 3-6 shows this behaviour for @ corrosion cell with iron. corroding at the anode and oxygen reduction at the cathode. With infinite resistance between the anode and cathode, cach electrode has a potential equal to the reversible pote=mtial. As current flows, the potentials move toward one another. If the intemal resistance Cf the cell is negligible as in the case of corrosion reactions occurring adjacent to one anothe=r on a piece of steel, the cell current, known in this case as the corrosion current (ieg,), is given by the point of intersection (A) of the anode and cathode polarisation curves. Ai point A, the anode and cathode will be at the same potential E,oq, called the corrosiaw n potential. If the shape of the anode and cathode polarisation curves is known, it is poss ible to predict the corrosion current and hence the rate of corrosion in a given situation. Polarisation diagrams of this type are known as Evans diagrams Foot ~ WTS 0, 49140 He 40H Log (curent density) Figure 3-6: Cathode and anode polarisation curves for a corrosion cell (Evans Diagram). ‘The deviation from the reversible potential is also known as o- erpotential or overvoltage. ‘There are three sources of polarisation which contribute to whe total polarisation of the electrode reactions. These are activation, concentration and re=sistance polarisation. These can affect both reactions but usually polarisation at the cathcede dominates the corrosion process, @ Activation Polarisation is the result of the fact that reacticens occurring at an electrode do not occur instantaneously but rather take a finite time to pro =z curves: re Experimental eae anodic cuve oS JE coer : <—Tatel region ’ i _ Kode oxtation | “MMs 1 Experimental Active : i cathodic curve i oon Log (current density) Figure 3-9; Polarisation curves for a correding metal electrode. Anodic and cathodic polarisation curves can be used to explain corrosion phenomena, such as galvanic corrosion. Figure 3-10 shows anodic and cathodic curves for copper, iron and aluminium in deaerated sulphuric acid. The corrosion potentials on the vertical axis show the positions of the three metals on the galvanic series for this environment, with copper the most noble, aluminium the most active and iron in between. However, the polarisation curves show much more than just potentials. The intersection where the anodic curve for the more active metal intersects the cathodic curve for the more noble metal gives the approximate current density or corrosion rate if the two metals are joined in this environment, assuming approximate equal areas. So in this environment, the corrosion rate of aluminium joined to copper (position x) is roughly equal to the corrosion rate of aluminjum joined to iron (position y), even though the difference in corrosion potential between aluminium and copper is much greater than between aluminium and iron. In this environment, galvanic action on aluminium by connecting to a more noble metal is much less than potential differences would suggest. However, the aluminium anodic curve does not strictly show Tafel behaviour as there is no linear region, and its behaviour may not be as predicted. While such curves would appear to give more information than the simple list of metals in the galvanic series, there are problems with such an approach, Diagrams giving anodic and cathodic curves of all common metals are very cluttered and it is 3-16PREDICTING CORROSION REACTIONS: difficult to interpret. Furthermore, such curves are only meaningful when both reactions are under activation control, while in practice the cathodic reaction is usually under concentration polarisation and single curves cannot be obtained for such conditions, ‘Tafel extrapolation can provide an accurate and rapid determination of corrosion rate, and such plots are widely reported in the corrosion literature. It is especially valuable for measuring very low rates of corrosion. However, it does require both reactions to show Tafel behaviour, at least over one order of magnitude. This is rarely the case where oxygen reduction is the dominant cathodic reaction. Where reduction of hydrogen ions to hydrogen gas is the dominant reaction, such as metals corroding in acid, activation polarisation does usually dominate and Tafel behaviour takes place. The method also requires that only one reduction process takes place and that resistive effects are negligible. In more resistive solutions, a special reference electrode placed very close to the working electrode surface (nown as a Luggin probe) is used 10 reduce resistive effects, Furthermore, the large current densities applied to construct the plots may not be representative of the tue corrosion situation. For example, a large current may cause the surface to passivate altering its surface from the free corroding state, However, by relating experimental current determination 10 weight loss and by comparing Tafel slopes with reported values, the validity of the process can be checked Potential (volts vs SCE) 07 1.0 10 Current density (mafsq cm) Figure 3-10: Potentiostatic polarisation curves for copper, iron and aluminium in sulphuric aca. 11. PASSIVITY The formation of passive films on metals such as aluminium, chromium, titanium and steel may be considered as a special form of resistance polarisation. Under certain environmental conditions, an oxide or other protective layer builds up on some metals and considerably reduces the corrosion rate. This loss of chemical reactivity is known as passivity. For example, aluminium metal is an extremely reactive metal and would corrode in many environments if it was not for a thin protective oxide layer. Similarly, stainless stee! relies on a thin chromium oxide layer for its corrosion resistance, Passive layers can be a problem as they tend to be stable only in certain environments. For example, both aluminium and stainless steel can pit badly in chloride environments. Care must be taken when relying on a passive film for corrosion protection 3-17PREDICTING CORROSION REACTIONS Steel behaves in a passive manner only under very specific conditions, and its passive film is easily destroyed. The nature of passivity on steel can be casily illustrated by observing the its behaviour n nitric acid (originally observed and noted hy Faraday). If a small piece of iron is placed in concentrated nitric acid, no reaction is observed. If the solution is diluted one to one, still no change occurs. But if the sample is scratched with a glass rod, rapid reaction occurs whereas scratching has no effect on the sample in concentrated acid, If the specimen is immersed in more dilute acid, vigorous reaction occurs without scratching, This same sample is very reactive (active) in dilute acid and inert (passive) in concentrated nitric acid The phenomenon of passivation is also studied by means of potentiostatic measurements as, described earlier. Figure 3-11 shows an anodic polarisation curve for 2 metal showing, passivity. Cathodic polarisation curves can also be obtained, but anodic behaviour is generally of most interest to the corrosionist. For both active-passive and non-passive metals, the metal initially shows its reversible potential at A and dissolves as would be expected with normal anodic behaviour obeying Tafel’s Law at B. This is called the active range. A non-passive metal such as zinc or copper does not nosmally become passivated and will only show this active behaviour. When oxidising agents are added to solutions in contact with non-passivating metals, corrosion rate increases linearly. With active-passive metals however, as oxidising power is increased, at a certain point corrosion rate decreases significantly. On the polarisation curve, # critical current density (igy,) is reached at C and the dissolution rate decreases to a very small value (D) and remains essentially independent of potential over a considerable potential range. This is termed the passive region. Here the metal is covered with a layer of an impervious passive film, such as an oxide or hydroxide. Finally, at very noble potentials (G), the dissolution rate may increase again in what is referred to as transpassive behaviour or oxygen may be discharged at these high potentials. For aluminium, anodising occurs at high potentials with the growth of a protective oxide film. = = 2 Flage or eriticar *Perentiat Active ‘Current Density Figure 3-11: Potentiostatic anodic polarisation curve for metals exhibiting passivity There are a number of desirable criteria for selecting a metal exhibiting passivity. The potential at which passivity occurs (known as the Flade potential or more commonly ‘critical potential’ or ‘primary passivation potential’) should be as low as possible. Titanium, for example has a critical potential of -0.24 volts (vs SHE) so can be passivated 318PREDICTING CORROSION REACTIONS. in non-oxidising acids whereas steel has a positive critical potential so requires an oxidising acid such as nitric acid to become passive. The critical current density for passivation (ig) should also be as low as possible as should the corrosion rate in the passive state. Moreover, the range of potentials over which a metal exhibits passivity should be as wide as possible, so that it can retain its passivity under a wide range of environmental conditions, Current potential curves such as those described are a quick and simple means of describing corrosion behaviour. Such curves are widely used to explain behaviour of inhibitors and alloys of differing composition, effect of temperature, pitting probability and many other corrosion phenomena. For example, Figure 3-12(a) shows the effect of chromium content on polarisation behaviour of iron-chromium alloys in 10 per cent sulphuric acid. As the chromium content increases, the critical current density for passivation decreases and the critical potential becomes more negative. In addition, the ‘corrosion rate in the passive state decreases. From these curves, the beneficial effect on passivity of increasing chromium content is clear. Figure 3-12 (a) shows polarisation curves of mild steel corroding in different concentrations of sodium chloride (NaC!) solutions at a pH of 1. In a solution free from NaCl, the steel shows classic passive behaviour with a wide potential range over which the surface remains passive. With NaCl, the chloride ion is extremely aggressive so causes high corrosion rates at all potentials with no passive range if the concentration is high enough. Breakdown of passivity due 10 chloride ions occurs locally rather than generally over the entire surface and, as a result, the steel surface would show significant pitting under such conditions. Applied. Potential Applied Potential Current Density Current Density @) (b) Figure 3-12: Anodic polarisation curves of (a) Fe-Cr alloys of differing chromium content and (b)mitd ste! in solutions of differing NaCI concentration. 412, POLARISATION CURVES AND THE POURBAIX DIAGRAM The polarisation curve showing current and voltage can be combined with the Pourbaix diagram to give a complete picture of a metal or alloy in a specific environment. Obtaining Pourbaix diagrams for all environments would require an excessive amount of experimentation, but using polarisation curves, an idea of the Pourbaix diagram can be 3-19PREDICTING CORROSION CTIONS: obtained. Alternatively, the Pourbaix diagram can be used to give a rough idea of the polarisation behaviour of a metal. Figure 3-13 shows how a qualitative polarisation curve can be built up that from the Pourbaix diagram for iron at a pH of around 7. At very negative potentials, iron is in the immune area and corrosion current is negligible. Making the potential more positive, iron reaches the corrosion zone of the Pourbaix diagram and the corrosion rate increases with increasing potential. In the passivation zone, current falls away again with low corrosion rates, Such a simplified polarisation curve could be drawn for other values of pH and for other metals if the Pourbaix diagram is available Conversely, a region of the Pourbaix diagram could be obtained from a polarisation curve, although a number at different pHs would be required to determine the slope of the sloping lines. This relation between the two methods of presenting corrosion data can assist in explaining corrosion processes. Passivation Protection Passivation Corrosion Protection coats . Ra Ae tat etHE 4 6 8 ‘Current Density pH Figure 3-13: Relation between Pourbaix diagram and polarisation curve. 3-20Chapter 4 TYPES OF METALLIC CORROSION It is convenient to classify corrosion into various types according to the appearance or reason or both. This classification can then be used in det 1g the cause (or causes) of a corrosion failure and in determining methods of minimising future failures. High temperature corrosion is a separate subject and not covered in these notes. Classification according to environment is covered elsewhere. In this chapter, the following types of metallic corrosion are described: += General corrosion or uniform attack (including dew point corrosion) + Galvanic corrosion + Crevice corrosion (including filiform corrosion) + Pitting corrosion + Intergranular corrosion (including exfoliation) + Selective corrosion or dealloying + Erosion corrosion + Cavitation Fretting corrosion + Suess corrosion cracking + Hydrogen damage, including embrittlement, cracking and attack + Corrosion fatigue + Liquid metal embrittlement + Stray current corrosion or electrolysis The appendix at the end of this chapter summarises the main features of these forms. 1. GENERAL CORROSION General or uniform corrosion is the most common form of corrosion and, as its name suggests, it proceeds uniformly over the metal surface. Examples are steel rusting in air, steel immersed in sea water, steel immersed in dilute sulphuric acid, copper or zine corrosion in the atmosphere, etc. It occurs because the anodes and cathodes continuously change position on a metal surface so the corroding metal can be regarded as one single electrode on which anodic and cathodic reactions occur simultaneously. If protective, corrosion products may progressively reduce the rate of corrosion. As examples, Table 4-1 shows the uniform corrosion rates of a number of metals and alloys obtained by immersion tests in sea water. This table is included to give relative corrosivity of alloys, rather than absolute figures. Local conditions, increases in velocity, presence of biological activity, etc can have a marked effect on corrosion rates, 41TYPES OF METALLIC CORROSION Atmospheric corrosion of ferrous and non-ferrous metals is the most common form of corrosion and represents the greatest destruction of metals on a tonnage basis. However, uniform corrosion is not of great concer from the technical stand-point because this type of corrosion is normally readily observed and the life of equipment can be accurately estimated on the basis of comparatively simple tests, such as weight loss or electrical resistance tests. Sudden failures should be rare. Other forms of corrosion are insidious in ature and their behaviour is much more difficult to predict. They are also localised and tend to cause unexpected or premature failures, Table 4-1: Unitorm conosion rate of some metals completely immersed in sti or slowly ‘moving sea water. “Metal — ‘Average corrosion ‘Other possible rate (mmilyear) modes of corrosion Mild & low alloy steel, grey cast iron 01-02 Brass (60Cu-40Zn) 0.08 dezincifcation Brass (70Cu-30Zn) 0.05 dezincification Assenical admiralty brass 0.05 Copper 0.04 ‘Aluminiurn bronze 0.04 pitting Zinc 0.02 Cupro-nickel (90/10, 70/30) 0.01 Lead 0.01 Monel (70Ni-30Cu) 0.01 Stainless steel (304, 316) Nil pitting, crevice corrosion Nickel Nil pitting Titanium Nil Ni-Cr and Ni-Cr-Mo Alloys Nil Uniform comosion can be controlled by the use of corrosion resistant alloys, surface coatings (metallic, inorganic and organic), eathodie protection and the use of inhibitors. Dew point corrosion (or low temperature corrosion or, more correctly, acid dewpoint corrosion) is a special type of general corrosion which occurs when hot gases, in most cases combustion products, are cooled so much that condensation starts to occur and corrosion processes commence. This requires the metal surface in contact with the gas to be at a temperature below the dew point of the gas. The dew point, the temperature at which an atmosphere can no longer contain its moisture so it condenses out, depends on the components of the atmosphere. The problem is common with waste flue gases produced by combustion of fossil fuels which contain several acid gases, such as sulphur trioxide and hydrochloric acid, along with watcr, and therefore such gases show several dew point temperatures at which the various species begin to condense, The problem is found in fossil fuel power stations, refineries and many other industries using hot combustion gases. tis also the common means of failure of automotive exhausts. In flue gases, two dewpoint temperatures are usually identified: + The water dewpoint which is associated with the condensation of water from the flue gas and depends on the amount of water only. + The acid dewpoint which is associated with the condensation of sulphuric or, less commonly, hydrochloric acid solutions. This is determined by the amount of sulphuric 4-2TYPES OF METALLIC CORROSION acid (or sulphur trioxide) or hydrochloric acid and water in the flue gas. Thus, for example, brown coal with a low sulphur content and a high moisture content will have a different acid dewpoint from a black coal with bigher sulphur but half the moisture Small amounts of these contaminants can make significant differences to corrosion behaviour. Even as little as 5 ppm sulphur trioxide in flue gases can raise the dew point by as much as 40°C causing condensation at a much higher temperature than would be expected in the absence of this gas. Figure 4-1 shows the effect of temperature on the corrosion rate of mild steel where it is exposed to a cooling flue-gas stream. As the temperature of the gas drops, corrosion commences at the acid dew point and increases as the quantity of liquid depositing increases. As the flue gas temperature is reduced further, the concentration of acid is reduced and the corrosion rate drops. This results in certain areas of the process failing while nearby areas remain intact. At Jower temperatures, gases such as sulphur dioxide dissolve in the water and the corrosion rate increases again, However, this is not usually a problem as temperatures in the stack and other parts of the process stream are usually above such levels, | Figure 4-1: Dew point corrosion rate variation with temperature Corrosion rate 50 700 150 Surface temperature (C) One of the most important methods to prevent this type of corrosion is to keep the ‘temperature everywhere above the dew point. As well as lagging, good housekeeping, careful plant operation and comprehensive maintenance all assist in minimising the possibility of lowering back-end temperatures. Thermal insulation must be properly designed and maintained, inspection doors kept sealed, leaking expansion joints repaired, and so on. Neutralising additives such as calcium hydroxide/oxide can neuiralise acids but their ability to reduce corrosion rates is not proven. More resistant materials than carbon steel may also help, but these have not always been successful as deposit formation can produce pitting. Coatings such as aluminium metal spray and glass-flake polyesters have been used with some success, but there have been difficulties in surface preparation and application, adhesion under cycling temperatures and wear duc to dust. As with many corrosion problems, modifications to design may also help to overcome such corrosion, 2. GALVANIC CORROSION Galvanic corrosion occurs when two or more dissimilar metals in electrical contact are placed in an electrolyte. This results from a potential difference between the metals which causes a flow of current between them. In an earlier chapter the galvanic series was used to 43TYPES OF METALLIC CORROSION predict the galvanic corrosion tendency of various metals. The more noble metal becomes cathodie while corrosion takes place at the more active metal, Galvanic corrosion is sometimes more accurately known as bimetallic or imilar metal corrosion. ‘The three essential components for galvanic corrosion are Qa) Q) @) two materials possessing different surface potentials, a common electrolyte and a common electrical path, If the two metals are electrically isolated, they will not experience galvanic corrosion regardless of their proximity, relative potential or size. Galvanic corrosion is usually characterised by special attack on one metal, while the other metal is relatively free from attack. Factors which contribute to galvanie corrosion include: a ai) (iii) Potential Difference. As mentioned above, the more active metal of the joined couple on the galvanic series will corrode. The further apart the two metals are on the series gives an indication of the probable magnitude of the corrosive effect. As potential is affected by environmental factors such as formation of surface films, @ galvanic series has to be measured in each environment of interest. However, the most common galvanic series has been constructed from measurements in sea water which has been found to be broadly applicable to other environments such as natural waters and uncontaminated atmospheres. There are some differences in the atmosphere as, for example, copper and copper alloys tend to act slightly more noble than the position given in the sea water series. Temperature can also be important; for example, zinc and iron have been found to show ‘cell reversal” or swapping of their places in the series at elevated temperatures in some water systems. Cathodic polarisation. As polarisation of the cathodic reaction is usually the rate limiting reaction in galvanic corrosion, it is necessary to have some knowledge of polarisation behaviour of the cathode before predicting galvanic behaviour, If the cathode has no oxide film, or any oxide film is readily reduced by the flow of galvanic current, reduetion of dissolved oxygen occurs readily and metals such as platinum, gold, silver, copper, etc. will act as efficient cathodes. However, if reduction takes place on a stable oxide film, the reaction tends to be much more polarised, therefore the cathode potential moves towards the active metal potential, and metals such as titanium, nickel, stainless steel and, to a lesser extent, aluminium tend to act as inefficient cathodes. As a result, such metals can cause less galvanic corrosion than would be expected from their position in the galvanic series. This means that galvanic corrosion of an active metal such as stec! joined to stainless ste! will be less than if it was in contact with copper, at whose surface oxygen reduction can readily take place, Area Effect. The rate of corrosion is directly proportional to the ratio of the arca of the cathodic metal to the anodic metal. Consequently, a small anode/large cathode will cause severe corrosion of the anodic metal. The opposite area ratio — large anode and small cathode — produces only slightly accelerated galvanic effects. The area effect is important in selection of fasteners and Table 4-2 gives a guide to fastener selection ensuring galvanic compatibility. It should be recognised that many other 44TYPES OF METALLIC CORROSION factors, including environmen, design, strength, aesthetics and cost are important in fastener selection. Moreover, use of plastic fasteners, organic-coated fasteners or insulating sleeves can avoid galvanic problems Table 4-2, Fastener selection for avoiding galvanic corrosion. ‘Structural material Fastener Zara | Alumin- | Carbon | Copper | Monel 3047316 coated | ium steel | forass stainless steel (1) foronze steel Galvanised steel v © © * © ° Aluminium ° ra (2) xQ) 0) ¥@) Carbon Stee! v x ° v y v Copper x x * v v 2 Monel x x x x v 2 304, 316 stainless steet | x x * x 2 v4) 7: Compatible - recommended o. Salisfactory = Not recommended 2: More information required or variable behaviour. (4) Amount of protection depends on coating thickness (2) May lead to enlargement of hole in aluminium (3) May lead to enlargement of hole in 2 severe marine environment, (4) May suffer crevice corrosion Gy) Distance and Electrolyte Effects. Galvanic attack decreases with increasing distance from the junction but the effect depends on the conductivity of the solution because the path of current flow is important. In a high conductivity solution, attack on the active metal may spread a Jong distance from the contact region and therefore be less severe. On the other hand, in low conductivity water corrosion can become localised in the contact area where it may be quite severe (see Figure 4-2). In addition, current will not flow readily around comers so geometry of the circuit is also important. Galvanic attack is usually most intense immediately adjacent to the cathodic material. zak code node @ @) Figure 4-2 :Effect of electrolyte conductivity on corrosion distribution in a galvanic couple, (a) igh conductivity electrolyte, (b) Low conductivity electrolyte. Control of galvanic corrosion can be achieved through a number of different methods. Combinations of metals or alloys widely separated on the galvanic series should be avoided, Controlling the environment by, for example, addition of inhibitors or oxygen scavengers, has been found to be effective. Barrier coatings, whether metallic, organic or vitreous, which can isolate the metals from the environment are effective but it must be stressed that it is extremely dangerous to coat only the anode of the couple. Electrical isolators to break electrical continuity, replaceable sections, or other design changes can also be considered, Cathodic protection can also be considered by using anodes of a more active metal than others in the assembly. 4-5TYPES OF METALLIC CORROSION Deposition corrosion is similar to galvanic corrosion in that it requires two dissimilar metals. Unlike galvanic corrosion, however, there is no clectrical connection between the two metals. Jt occurs when a cathodic metal, such as copper or mercury, is plated out of solution onto an anodic metal surface such as aluminium or zinc. For example, soft water passing through copper pipes will dissolve some copper and accumulate copper ions. If the water is admitted to a galvanised or aluminium vessel, particles of metallic copper will plate out causing pitting by local ceil action (see Figure 4-3), This is also the reason copper fasteners are not recommended for galvanised stee! or aluminium in Table 4-2, despite the favourable area ratio for galvanic corrosion. This process is sometimes called cementation, ‘The design of the structure should be such that zinc or aluminium are not downstream of water that may contain copper ions. Deposition corrosion can be avoided by preventing the pick-up of cathodic ions that will enter the equipment or by scavenging them by passing the contaminated product through a tower packed with turnings of a more anodic metal such as steel, on which the ions can deposit. Copper piping ‘Aluminium orzine, LLL. Figure 4-3. Deposition corrosion on aluminium or galvanised steel fram water containing ‘copper ions. 3. CREVICE CORROSION Intense localised corrosive attack often occurs within confined spaces or crevices formed by certain mechanical configurations, such as holes, gasket surfaces, lap joints, surface deposits and crevices under bolt and rivet heads. The corrosion occurring under solid deposits is often called deposit corrosion. The crevice (no matter what the cause) must be wide enough to allow the liquid to enter, but sufficiently narrow to maintain a stagnant zone. Thus it usually occurs in very narrow openings, from less than a micron up to about 1mm, Permeable corrosion products and fibrous gaskets, which have a wick action, will create almost ideal crevice corrosion conditions. The crevice former does not need to be a metal and usually the smaller the crevice, the more readily crevice corrosion is likely, due to capillary action, Crevice corrosion is one of the most dangerous types of localised corrosion processes. It is ‘most prevalent in the easily passivating alloys such as stainless steels, It may also occur in alloys not belonging to this class (e.g. mild steel) under conditions where the corrosive medium is highly oxidising or passivating, Until recently, it was believed that crevice corrosion was simply caused by differences in oxygen concentration (differential aeration) between the crevice and its surroundings. However, more recent work has shown that although differential aeration does play a part 46TYPES OF METALLIC CORROSION in the initiation process, it is the decrease in p}4 within the crevice that plays the major role in the propagation. Yo illustrate the mechanism of crevice corrosion, consider a crevice within a stainless stecl structure which is immersed in an aerated solution containing chloride ions (sce Figure 4- 4(2)). Initially, @ small amount of general corrosion will occur in the passive state, consuming oxygen to produce hydroxy! ions. This will occur both within the crevice and on the external surface due to the reaction: YO, + H,0 + 2e > 20H (Bg. 4.1) Afier all the oxygen is consumed within the crevice (the narrow entrance to the crevice will minimise the diffusion of oxygen into the crevice), the cathodic reaction (Eq 4.1) will occur entirely on the outer surface whereas within the crevice, the anodic reaction will ‘occur: Fe Fe +2e (Eq. 4.2 Hydrolysis reactions (reactions involving break down of water) can the occur within the crevice, for example: 3Fe2* + 4H,0 > Fe,0, + 8H* + 2c (Eq. 4.3) “on Me 0, or RAR SS (a) (b) Figure 4-4: Schematic representation of the reactions occurring in crevice corrosion (2) immediately after contact with the electrolyte, (b) acidification within the crevice The formation of hydrogen ions will result in a lowering of the pH of the crevice solution, as low as a pH of 1, which is strongly acidic. With the increasing cell current, chloride ions can migrate into the crevice forming hydrochloric acid (see Figure 4-4(b)). The acidification of the crevice solution can lead to the crevice becoming active, while the outer surface is still passive. The accumulation of corrosion products in the erevice mouth will further restrict the diffusion of oxygen into the crevice. The corrosion products will also restrict any interchange between the acid solution within the crevice and the bulk solution, thus maintaining the acid conditions within the crevice. The presence of hydrogen ions also means that the cathodic reaction (Eq 4.4): 2H? +20 > H (Bo. 4.4) ATTYPES OF METALLIC CORROSION may also occur within the crevice. ‘The propagation stage of crevice corrosion has been reached when the above reactions have created an active-passive macro-cell between a small active and acid anodic area (the crevice) and a large passive cathode area (the outer surface). This leads to rapid corrosion within the crevice. Crevice corrosion is autocatalytic, that is the corrosion process produces conditions which are both necessary and stimulating for continuing activity. Crevice corrosion is often a problem in stainless steels and nickel-based alloys in sea water and other chloride-containing solutions. Table 4-3 shows the crevice corrosion resistance of a number of alloys in sea water. Generally, the higher the chromium and molybdenum content, the better the resistance to crevice corrosion. This table shows relative resistance for this environment, and performance in other environments may be different. Table 4-3: Resistance of some alloys to crevice corrosion in sea water. Least Resistant 430 Stainless steel 304 Stainless steet 316 Stainless steel Incoloy 825 £04 Stainless steel Alay 2205 Siainless stec! Ferralium 256 Stainless Stee! ZE4SMO Stainless steel “tania | Inconel 625 [Hastelloy C-276 Most Resistant As well as selecting the correct alloy, the best method of controlling crevice corrosion is to avoid crevices in the design and fabrication of the structure. It is also important to avoid conditions that lead to the formation of deposits on a metal surface by removing solids in suspension and maintaining clean surfaces. Filiform Corrosion is a special form of crevice corrosion which consists of thread-like filaments of corrosion products on the surface of a metal under a transparent coating. Filiform corrosion has been observed on steel, zinc, aluminium and magnesium surfaces covered with various types of lacquer and paint coatings. Although the mechanism of filiform corrosion is not fully understood, the mechanism appears to be a special form of crevice corrosion as described above. Corrosion is restricted to the active head where hydrolysis reactions occur to maintain an acidic environment. The cathodic tail consists of oxygen, water and corrosion products (see Figure 4-5). The damage caused by filiform corrosion is very slight with the major objection to this type of attack being the adverse effect on the appearance of the metal, especially where the coating is transparent. Filiform corrosjon occurs mainly in atmospheres between 65% and 95% relative humidity. The only way to prevent filiform corrosion is reduce the relative humidity to below 60 per cent but this is not always possible so inhibitive primers, smooth surfaces or thicker coats may be necessary. 48TYPES OF METALLIC CORROSION Hydrolysis reactions (low pl) Cathode ‘Oxygen readction (high pH) Low oxygen concentration Figure 4-5: Schematic representation of filiform corrosion on steel Poultice corrosion is a special case of crevice corrosion whieh arises when an absorbent material such as paper, dirt, cloth, etc. is in contact with a metal surface that becomes periodically wetted, Examples of poultice corrosion are found in automobile mudguards and doors where mud and other material accumulates causing corrosion of the steel body (see Figure 4-6) and aircraft fuel tanks where mats of bacterial and fungal growth cause damage of the aluminium surfaces. While the assembly is completely wet or completely dry, no action takes place. But during the drying period, wet and dry areas develop leading to localised attack due to differential aeration. Poultice corrosion is prevented by avoiding the contact of absorbent materials with the metal surface, designing to prevent such materials from becoming wet in service, good housekeeping to prevent dirt build-up or painting the surface in contact with the absorbent material. Packed mut causes rusting ~~ mud is thrown into trap Figure 4-6: Poultice corrosion caused by mud inside car mudguard. Water line attack is another case of differential aeration leading to localised corrosion. This is corrosive attack just below the liquid level on metals immersed in water. The meniscus on the metal forms a cathodic region with easy oxygen access and the area just below this forms an anode where attack takes place. Corrosion products are deposited between the two electrodes. Examples are steel poles driven into soil where attack is especially marked just below ground water level making their removal difficult. A steel pole driven into the sea bed is most strongly attacked just below the bed since the part of the pole in contact with the oxygenated sea water serves as a cathode. 49TYPES OF METALLIC CORROSION 4. PITTING CORROSION Pitting is one of the most destructive forms of corrosion as it can cause equipment failures due to perforation, while the Joss of metal due to uniform corrosion is minimal. Generally pitting occurs on oxide-covered metal surfaces such as stainless steels or aluminjum due to the localised breakdown of the oxide film by aggressive anions, especially chloride ions. Pitting can also occur on steel in boilers and other water systems, when the oxygen content increases $0 that the protective magnetite film breaks down locally causing pits or depressions on the stecl surface. If allowed to continue, the pits increase in size and corrosion products form around the pits in nodules, a process sometimes called tuberculation. Pitting also occurs on steels when some types of protective coatings break down, although this is more likely @ form of crevice corrosion. Pitting in mild steels usually results in the formation of shallow pits, whereas stainless materials tend to form deep pits. The initiation stage of pitting corrosion of stainless steels or aluminium usually involves the chloride or other halide ions penetrating the oxide film usually where the film is weakest (such as at sulphide particles or other non-metallic inclusions) causing local destruction of the passive film. Once the pit has formed, corrosion products wil] usually form over the mouth of the pit restricting diffusion between the pit and the bulk solution. Hyarolysis reactions within the pit will lead to acidification of the solution within the pit as with crevice corrosion. ‘The acidification of the pit solution can lead to the pit becoming active while the outer surface is still passive (see Figure 4-7). ‘The propagation stage of pitting corrosion is identical to that of crevice corrosion in that an active-passive micro-cell is formed between a small active and acid anode area (the pit) and a large passive cathode area (the outer surface). This results in rapid corrosion within the pit. Pitting corrosion is also autocatalytic like crevice corrosion. Electrolyte Corrosion products Passive oxide film (cathode) Active ‘anode Figure 4-7: Pitting corrosion mechanism on an active-passive metal. ‘The best way to avoid pitting corrosion is to use metals that are known not to pit in the particular environment. For example, 316 stainless steel, with molybdenum, is more resistant to pitting than 304 stainless steel. Higher nickel alloys such as the Hastelloys, or titanium, are even more pitting-resistant. Pitting resistance is similar to crevice corrosion resistance and the order of metals in Table 4-3 gives approximate resistance to pitting in sea water. To counteract pitting in aluminium, the copper, iron and silicon content should be kept to 2 minimum and annealing at 500°C should be avoided since these alloying 410TYPES OF METALLIC CORROSION elements produce noble precipitates. Pitting of steel pipes in water systems is avoided by Glosely monitoring oxygen content. Other methods can also be considered. Since pitting is electrochemical it can be stopped by cathodic protection. Reducing the aggressiveness of the environment, for example by reducing chloride ion concentration, temperature, acidity, etc or addition of inhibitors can also prevent pitting. Inhibitors must be used with care because if pitting is not completely stopped there may be an increase in pitting intensity. Sacrificial metal coatings, such as zinc on stee} or Alclad aluminium, or protective coatings such as zine-rich paints can also prevent penetration by pitting. In some cases, agitation of the environment to eliminate stagnant solutions will prevent pitting, 5. INTERGRANULAR CORROSION Ina grain boundary, atoms are loosely packed unlike atoms in regular lattice positions and ‘one would expect them to be more prone to attack than the bulk metal. In fact, in most applications, the difference in reactivity is only slight resulting in uniform attack. However, under certain conditions, impurities or precipitates may segregate to the grain boundary resulting in localised attack and catastrophic failure. Intergranular corrosion is a deceptive form of attack in that very little weight loss is evident and a network of disintegration spreads inwards undetected, resulting in a Joss in mechanical properties, with a drop in ductility occurring well before a loss in strength ‘The most common example of this type of corrosion is the intergranular corrosion of the austenitic stainless steels, If these steels are held in the temperature range of S00 to 800°C, or cooled slowly through this range, fine precipitates of chromium carbide grow in the grain boundaries (a process known as sensitization). This depletes the chromium in the nearby regions below the 12 per cent level required for corrosion resistance (see Figure 4- 8(@)). As a result, these regions will no longer be passive and will corrode preferentially. Intergranular corrosion of austenitic stainless steels is often caused by welding, where the term weld decay is used, In such cases, the attack usually occurs in bands on each side of the weld and at a certain distance from it (see Figure 4-8(6)). Chromium 2000 1500 600 500 100 Approx temperature (70) (a) (b) Figure 4-6: (a) Chromium-depleted grain boundary zones resulting in intergranular corrosion, (b) Weld decay due to intergranular corrosion. Intergranular corrosion of austenitic stainless steels can be controlled or minimised by three methods:TYPES OF METALLIC CORROSION (Reduce the carbon level below 0.03% which removes the possibility of chromium carbide precipitation. These low carbon steels are designated "L" e.g, type 3041 (i3)__Use stabilised grades containing titanium (321 grade) or niobium (347 grade) which have a stronger affinity for the carbon than the chromium, thus titanium and niobium carbides are precipitated rather than chromium carbide. Low carbon grades have substantially replaced stabilised grades, except for high temperature service. (iii) Heat the sensitized steels to 1000 10 1200°C to redissolve the chromium carbides then quench in water. This process is expensive, only suited to small objects and quenching may cause distortion or undesirable residual stresses. Stabilised austenitic stainless steels may become susceptible to a localised form of intergranular corrosion called knife-line attack. During welding, the base metal adjacent 10 the fusion line is heated to temperatures high enough to dissolve the stabilising carbides, but cooling is usually rapid enough, especially in thin sections, to prevent carbide precipitation. Subsequent welding passes reheat this area into the critical temperature range in which chromium carbides precipitate, Thus a narrow band adjacent to the fusion zone becomes susceptible to this particular form of intergranular corrosion. Knife-line attack can be avoided by proper choice of welding variables and the use of a heat treatment to 1050°C after welding to dissolve the chromium carbide and re-precipitate the stabilising carbides. A similar welding problem, although not true intergranular corrosion, is grooving corrosion which has been found in ERW (electric resistance welded) carbon steel pipes. The ERW process causes rapid fusion and cooling of the steel in the welded region which results in melting and redepositing of sulphide inclusions along the weld line. This creates a Jocal anode relative to the base metal which can rapidly corrode when exposed to aggressive waters. Grooving corrosion should not be a problem if the sulphur content is kept below 0.02 per cent or if the weld is subject to a post weld heat treatment. There are other, less common, forms of intergranular corrosion. High nickel alloys are attacked by sulphur and sulphur-bearing gases above about 315°C. Low melting sulphides form at the grain boundaries resulting in intergranular corrosion causing brittleness. Age hardenable, high strength aluminium alloys are also susceptible to intergranular corrosion, particularly in a chloride-containing environment. These alloys are the precipitation hardenable Al-Cu, Al-Cu-Mg and Al-Zn-Mg alloys and work hardenable Al- Mg alloys (with more than 3 per cent magnesium), Again, precipitates can form in grain boundaries resulting in a potential difference between the precipitates and adjacent grain, In some alloys such as Cu-Al alloys, the precipitates are more noble resulting in matrix corrosion while in others the precipitates are more active than the adjacent grain so they corrode preferentially. The degree of susceptibility depends on the heat treatment which governs the nature, size and number of intergranular precipitates. Intergranular corrosion has also been observed in nickel-chromium alloys, magnesium and copper-base alloys. Exfoliation or layer corrosion is a particular type of intergranular corrosion that occurs in high strength aluminium alloys, in particular Al-Cu alloys (duralumin-type alloys) and Al- Zn-Mg alloys. If these are forged, rolled or extruded, they will have an elongated grain 412‘TYPES OF METALLIC CORROSION structure with grains aligned to the extemal work surface. Exfoliation initiates typically where the ends of the elongated grains are exposed by machining and proceeds along the parallel grain boundary networks, as shown in Figure 4-9. The corrosion product, which has a greater volume than the aluminium consumed, causes the layers to be pushed out into a mass of thin, loosely-adherent flakes. This type of corrosion can be avoided by selection of un exfoliation resistant alloy or by slightly over-aging the alloy. Grain boundaries Fastener Figure 4-9: Exfoliation corrosion of aluminium panel adjacent to @ fastener. 6. SELECTIVE CORROSION Selective corrosion, leaching or dissolution (also known 2s dealloying or demetallification) is the removal of one element from a solid alloy by a corrosion process. The most common examples are the selective removal of zinc fom brass alloys (dezincification) and the removal of ferrite from grey cast iron (graphitization). Other occurrences include the removal of aluminium from aluminium bronzes, cobalt from cobali-tungsten-chromium alloys, and others. Parting is sometimes used 0 describe selective corrosion but this term is usually restricted to noble metal alloys such as gold- copper or gold-silver where the less-noble metal is selectively removed to purify the gold. Dezincification of brass is readily apparent as the yellow brass colour is replaced by the reddish colour of copper. It may be uniform (layer type) or localised (plug type). With layer type dezincification, zine is uniformly removed from the surface at a constant rate. In plug-type attack, a segion or plug of brass is dezincified locally and a porous residue of copper is left. This form is more serious as failure can occur within months or even weeks. Dezincification results in loss of strength or complete perforation; however in some cases the overall shape of the alloy may not change. Dezincified brass pipe may retain sufficient strength to resist internal water pressures until an attempt is made to uncouple the pipe or water hammer occurs, causing the pipe to split open. In some soft waters, the loss of zine can result in the blockage of a fitting with precipitated zinc salts, producing what is known as ‘meringue dezincification’. This is not a different type of attack. The most commonly accepted mechanism for dezincification involves the dissolution of the entire alloy with the ine ions staying in solution while the copper ions plate out as metallic copper to form the Porous copper mass. An altemative theory is that the zinc is selectively leached from the alloy leaving behind the copper. In general, the rate of dezincification increases as the zinc content increases. Brasses with less than 15 per cent zinc are immune while those with 30 per cent or more of zine are highly susceptible. Dezincification occurs over a range of pH and is favoured by stagnant 4-13TYPES OF METALLIC CORROSION conditions, the presence of chlorides, high temperatures and deposits and scales. It mainly ‘oceurs with sea water, but can occur with fresh water where the chloride is high and the temporary hardness is low. Figure 4-10 (sometimes called the Tumer diagram) relates the likelihood of meringue dezincification in fresh water to chloride content and temporary hardness (hardness due to calcium bicarbonate and magnesium carbonate and bicarbonate) of the water. The diagsam should not be used too literally as other factors such as temperature, pli and other anions present can also be important 100; I pH>8.3, dezinefcation likely : 1 pH18 Inconel 625 No limits Hastelloy C-276 No limits Titanium No limits Erosion-corrosion is best controlled by reducing solution velocity to reduce turbulence (although this is rarely possible in practice) or by designing to avoid turbulence by increasing pipe diameter, using large radius bends and eliminating abrupt changes in section. The problem in condenser tubing can be prevented by inserting short lengths of flared tubing (ferrules) in the inlet ends. The ends must be ‘feathered’ to blend the flow or erosion-corrosion will occur just beyond the end of the ferrule because of the step. It is also important to choose a material which is resistant to this kind of attack. Titanium, Monel, aluminium brass, chromium nickel and chromium cobalt-based alloys can offer good performance in a wide range of corrodents, and polyester fibreglass composites can provide good service, Flexible coatings such as elastomeric polyurethanes or hard ceramic linings have good resistance to abrasive slurries and have been successfully used as linings in pipes and pumps. A small amount of iron, by an addition to the alloy, a FeSO, addition to the water or from an iron sacrificial piece, will enable copper alloy tubing to tolerate higher flow rates 8. CAVITATION Cavitation or cavitation corrosion is a special form of erosion-corrosion which is caused by formation and collapse of vapour bubbles in a liquid near a metal surface. It occurs in ships’ propellers, pump impellers, cylinder liners and steam turbine blades producing 416TYPES OF METALLIC CORROSION 7. EROSION CORROSION Erosion-corrosion is the attack caused by the combined effects of corrosion and wear which results from the relative movement between the corosive environment and the metal surface. The corrosion rate can be greatly inereased when conditions such as liguids moving ai high velocities, solids in suspension (slurries), marked turbulence or impingement are present. Erosion-corrosion is sometimes called impingement. Cavitation and fretting are sometimes categorised as forms of erosion corrosion, but arc better classed as a distinct forms of corrosion Under turbulent flow conditions, the random high velocity motion of the corroding solution on the metal surface removes any protective oxide films. This leads to further oxide film formation and subsequent removal, and an acceleration of the corrosion rate. ‘The damage caused by impingement may be uniform or localised and is often characterised by the metal surface having a horseshoe-shaped pits which are smooth and free from corrosion products (sce Figure 4-11). These depressions follow a paitern indicative of the relative movements of metal and solution. Erosion corrosion can also take the form of grooves, waves, gullies and rounded pits +——— Water flow Figure 4-11: Erosion-corrosion damage of a metal Virtually anything that is exposed to a moving liquid can be susceptible to erosion corrosion. Examples of equipment which can fail due to impingement are agitators, copper pipes, centrifuges, elbows, heat exchangers and condensers, impellers, pumps, propellers and turbine blades and casings. Even straight sections of pipe which may contain weld beads protruding in the pipe bore can be subject to crosion-corrosion. At the inlet ends of tubing in condensers and other shell and tube heat exchangers, erosion-corrosion often causes attack over the first few centimetres of the tube. Non-turbulent flow develops afier the liquid has progressed down the tube a relatively short distance and the problem disappears. The extent of damage by impingement is influenced by a number of factors: (i) Nature of the surface film: A hard, dense adherent film such as that on stainless steels, nickel alloys, etc. provides better protection than one which is easily removed such as that on steel or bronze. (ii) Nature of the environment: Increasing velocity and corrosi exacerbates impingement. ty of the environment (ii) Pressure and size of air bubbles: Bubbles in the corrosive stream increase the erosive effect but their size must be within a critical range to cause maximum erosion. Large and small bubbles appear to be comparatively harmless. (iv) Suspended solids increase erosive effect, 415TYPES OF METALLIC CORROSION savage fissures and pits. The appearance is similar to that produced with pitting but the surface is usually roughened and the pitted arcas are closer together resulting in a spongy appearance Figure 4-12 is a schematic diagram of the mechanism of cavitation, showing how collapsing bubbles destroy the metal surface. 1 occurs in those parts of the item, such as the face of a propeller blade, where the liquid flow produces localised low pressures. Vapour bubbles can form as the pressure drops because the metal moves away from the Jiquid (Figure 4-12(@)). These are essentially negative pressure bubbles. When the metal stops moving, the liquid continues to move collapsing the bubbles causing the pressure to increase rapidly (Figure 4-12(b)). As the metal starts to move toward the liquid, these bubbles will collapse or implode resulting in shock waves with very high pressure (Figure 4-12(c)). This results in a strong hammering of the metal surface which will fracture and remove the surface oxide film and allow corrosion of the metal surface. Liquid rpovernent Liquid movernent ee ie Waly ‘Metal ee metal movernent metal ovement @) © © Figure 4-12: Cavitation mechanism: (a) void nucteates as metal moves away from liquid, (b) Void collapses as metal movement stops, (c) metal starts moving towards liquid and void implodes. Cavitation damage can be prevented by the same means as for impingement attack — design factors and resistant alloys, Smooth finishes which do not provide sites for bubble nucleation have been used successfully to reduce cavitation effects on propellers and impellers. Rubber or plastic coatings have proved beneficial but it is important 10 avoid bonding failures between the metal-coating interface. The choice of the alloy is very important. An alloy with high toughness, ability to work harden and a good protective oxide coating such as stainless steels, titanium and aluminium bronzes have good resistance to cavitation whereas cast irons, most steels and brass have poor resistance. Aluminium bronzes are frequently used for ship’s propellers. Table 4-5 lists a number of engineering materials based on their resistance to cavitation damage. The addition of inhibitors and use of cathodic protection may be of some value, particularly if corrosion is an important part of the process. 9. FRETTING CORROSION Fretting corrosion (sometimes known simply as fretting) takes place where metal surfaces are under load and subject to slight oscillating movement, such as vibration. Initially when metals make contact, the high parts of the surface weld together and, with any motion, these welds break off and rapid oxidation of the freshly broken surface occurs. Further motion causes more abrasion, welding and oxidation. The result is rapid deterioration of the mating surfaces and the accumulation of debris, including metal particles and oxides. The fretted surface has a pitted and grooved appearance, discoloured by corrosion products. Fretting corrosion is differentiated from other types of wear in that the relative velocity of 4.17TYPES OF METALLIC CORROSION the two surfaces is Jow, the surfaces are never brought out of contact and that the amplitude of motion is small, usvally Jess than J00um. Fretting is usually diagnosed by the distinctive appearance of the corrosion product. On stecl, the debris is much redder than normal rust and sometimes called ‘cocoa’. On aluminium, the debris is black as distinct from the usual white corrosion products, When serious, fretting is often associated with fatigue cracking Table 4-5: Ratings of materials for cavitation resistance. Most Resistant Stelltes 18-8 stainless steel (welding rod) Bronze welding rod (Cu-10Al-11 5Fe) 18-8 stainless steel (cast) Nil Bronze (cast) 13% Cr cast iron shee} Manganese bronze (cast) 18-8 stainless steel (spray metallizing) Cast ste! Bronze Rubber Cast iron Aluminium Least Resistant Most materials are susceptible to fretting and this type of corrosion occurs in vibrating machinery particularly in the vicinity of shrink fits, press fits, bolted or riveted assemblies and other components clamped together and not intended to undergo relative movement. It also occurs in surfaces intended to act as bearings, such as tracks of rolier and ball bearings where it is often termed ‘false brinelling’. The deterioration not only causes @ loss in dimensional accuracy and ruins the surface of bearings but increases the susceptibility to fatigue failure. Metals which form passivating films such as aluminium, titanium and stainless steel are particularly prone to fretting damage. Table 4-6 shows some metallic combinations that have low, medium and high resistances to fretting corrosion in air under dry conditions. Table 4-6: Resistance of various metal couples to fretting corrosion under dry conditions, Low Medium High ‘Steel on steel Zinc on steel Lead on steel Nickel on stee! Copper alloys on stee! Silver piate on steel Aluminium on stee! Zine on aluminium Silver plate on aluminium piate Tin on steel Copper plate on aluminium Steel with a conversion, ‘Aluminium on aluminium Nickel plate on aluminium coating on steel Zinc-plated steel on aluminium Silver plate on aluminium Design and consideration of mechanical variables — removing the source of vibration, Jocking a moving joint, increasing greatly the amplitude of vibration — are probably the simplest and most economical means of avoiding fretting problems. Use of a soft metal with a hard metal, for example plating steel with zinc, lead, cadmium or silver, can assist. The use of lubricants may help if the load is not too high. Gaskets such as PTFE with « low co-efficient of friction or elastomeric materials which absorb motion can be considered, 4-18TYPES OF METALLIC CORROSION Phosphating, the parts and impregnating the porous phosphate layer with oil has been found to be effective. Similarly, anodising can reduce wear of aluminium 10. STRESS CORROSION CRACKING Stress corrosion cracking (SCC) refers to cracking caused by the simultaneous presence of tensile stress and a specific corrosive medium, Only certain combinations of alloys and chemical environment lead 10 SCC. However, many of these are common, such as aluminium alloys and sea water, copper alloys and ammonia, mild steel and caustic compounds and austenitic stainless stecls and chioride-containing compounds. The corrodent may only need to be present in parts per million concentration. The stress must be tensile, although residual stresses arc often sufficient. There is usually a threshold temperature that must be exceeded so, for example, stress corrosion cracking of austenitic stainless steels in chlorides is rarely observed below 50°C. However, there is a complex interaction between stress, concentration of corrrodent, temperature and other variables and predicting SCC can be difficult. Table 4-7 gives some alloy/enviroament combinations. Such tables should be treated with caution as service conditions can differ markedly from those in which susceptibility listed in the table was determined, Moreover, new alloy/- environment combinations are being found all the time. Table 4-7: Partial isting of environments causing stress corrosion cracking for some common metals (Those most commonly causing in-service failures are indicated by bold type) Alloy Environments Carbon steel Hydroxyl (caustic) solutions, nitrate solutions, carbonate/bicarbonate solutions, cyanide solutions Austenitic stainless steels Aqueous chloride solutions, other halides, caustic solutions. Aluminium alloys Aqueous chloride solutions Copper alloys Moist ammonia atmospheres, nitrite, amines, steam Stress corrosion cracks give the appearance of brittle, mechanical failure while, in fact, they are the result of local corrosion. Both intergranular and transgranullar are observed (see Figure 4-13) and cracking generally proceeds normal to the applied stress. Cracks also vary in their degree of branching. In some cases they are virtually without branching, in others they exhibit multibranched ‘river delta’ patterns. The presence of branched cracking is probably the most common means of identifying stress corrosion cracking. Because of the complex interplay between the metal, the interface and the environment, it is difficult to propose a general mechanism for SCC. It is widely believed that electrochemical dissolution plays a major role. The first stage in cracking is the localised break down of the passive film by the acting stress or by a specific agent (e.g. chloride ions for stainless steels) or both. Then the joint action of the tensile stress and metal dissolution causes the cracks to propagate along grain boundaries (intergranular cracking) or along slip planes (transgranular cracking). During SCC, the metal surface shows virtually no sign of 4-19TYPES OF METALLIC CORROSION corrosion while cracks propagate Ieading to ultimate fracture of the component. Hydrogen embrittlement may also be caused in some systems (@) (o) Figure 4-13: Modes of stress corrosion cracking - (a) Intergranular end (b) Transgranular. Stress corrosion cracking may be reduced or prevented by application of one or more of the following methods. Changing the alloy to a more resistant species in the given environment is probably the most common solution. Lowering the stress below the threshold value (if one exists) by annealing, thickening the section or reducing the load can also be considered. Altering the environment (c.g. by degasification, demineralisation or using a corrosion inhibitor) are other alternatives. Cathodic protection will reduce the corrosion and SCC but must only be used if hydrogen embrittlement is not a possibility 11. HYDROGEN DAMAGE Hydrogen in very small amounts can cause problems in a number of metals. While corrosion reactions can be the source of hydrogen, and chemical reactions can occur, the problems are not classical ‘anode-cathode’ corrosion. Hydrogen embrittlement occurs with very low concentrations (approximately 1 ppm) and can result in normally strong and tough metals becoming weak and brittle. With slightly higher amounts of hydrogen, hydrogen cracking (also known as hydrogen induced cracking (HIC), flaking, hairline cracking or shatter cracking) causes actual cracks to form in a metal. Related to this, blistering results in large blisters forming on vessels, pipes or other products made from rolled steel. In sulphide environments, commonly encountered in the oil industry, sulphide ions enhance atomic hydrogen penetration into steel resulting in a form of hydrogen cracking known as sulphide stress cracking (SSC). Hydrogen attack is less common and occurs when hydrogen reacts with carbon in the stecl, under conditions of high temperatures and pressures, to form methane resulting in internal cracks and fissures Figure 4-14 shows the range of hydrogen problems which can occur in a metal. Such problems may occur in a wide range of high strength metals but the problem is most common in steels. The hydrogen can be introduced from the presence of moisture during smelting which is entrapped during solidification, or it may be introduced during heat treatment, welding, acid pickling, electroplating or cathodic protection. Even hydrogen produced at the cathode in corrosion reactions has been known to cause hydrogen damage. Microstructural changes are not observed with hydrogen embrittlement but some mechanical properties, such as ductility and tensile strength, can change. The mechanism 4-20OF METALLIC CORROSION TYP of embrittlement is not as well understood as for the other forms of hydrogen damage. It is proposed that embrittlement arises because hydrogen atoms fix dislocations hindering their movement. ee a ee ult LE org 2 oP Po 929 of o a Staten Hoy fc ie eee ee cain \ Tie achat i ape > taaat race pe Hap zone Frese Emoiionen) — [REE [Suphide evens [Frsioaen ‘racking / ‘tack Figure 4-14: Hydrogen embrittlement, cracking, blistering and attack processes. Hydrogen cracking or blistering occur when hydrogen atoms diffuse into a metal and combine to form molecular hydrogen at a void or non metallic inclusion, Since molecular hydrogen cannot diffuse, the concentration and pressure will build up until itis sufficient to rupture the material. Hydrogen cracking is normally parallel to the rolling direction. Sulphide stress cracking was originally thought to be a form of stress corrosion cracking, thus its name, but is now known to be caused by hydrogen. Carbon and low alloy steels are susceptible in wet hydrogen sulphide environments, where the H,S serves to introduce hydrogen into the metal. Unlike hydrogen cracking, a tensile stress (applied, residual or both) is required for sulphide stress cracking, and eracking will proceed perpendicular to the tensile stress direction, often after initiating at inclusions. Hydrogen damage is best avoided by preventing hydrogen from entering the metal in the first place. Dry conditions during welding and use of low-hydrogen welding rods are important. Control of conditions during electrodeposition and cathodie protection can help prevent hydrogen pickup, Coatings and inhibitors to prevent corrosion reactions will reduce the amount of hydrogen available. If the hydrogen is still in atomic form in the metal lattice, it can be removed by baking at a relatively low temperature (100 to 150°C) Using ‘clean’ steels with a low content of non-metallic inclusions will increase resistance to cracking and blistering because of the absence of voids. If hydrogen cannot be avoided, then the use of nickel-base alloys should be considered as these have a far lower susceptibility to hydrogen problems than high strength steels. Sulphide stress cracking susceptibility depends on the strength of the material and most steels are resistant if the hardness is kept below Rockwell C 22, NACE Standard MRO175 documents materials that can be used under different levels of HS, Unlike most forms of corrosion, resistance to sulphide stress cracking and other forms of hydrogen damage (except attack) increases as temperature is raised above room temperature as hydrogen can more easily diffuse from the metal. 4.21TYPES OP METALLIC CORROSION 12, CORROSION FATIGUE Fatigue is the failure of a metal under cyclic stresses below its maximum static strength. A material cannot withstand as high stresses under long periods of cyclic loading as it can under static loading. Fatigue is one of the most common forms of failure of components such as crankshafis under repeated and reversed stresses. Fatigue failure usually commences at the surface of the structure as a small crack which grows slowly into the material. The crack front advances into the material with cach cycle and is marked on the final fracture surface as a small undulation or ripple line. These have a characteristic form of roughly circular, concentric arcs, giving what is ofien called a ‘clamshell appearance’ or “beach marks’. These grow until the cross section can no longer sustain the load and the structure fractures. ‘The stresses which a sample can endure for a given number of cycles are determined by fatigue tests. Specimens are cycled with a given Joad and the number of cycles they can endure is recorded. Data are plotted as stress versus the log of the number of cycles 10 failure; these are known as S-N curves. Stecls usually show a fatigue or endurance limit when tested in air, i. there is a load below which the material will not fail in a fatigue manner regardless of the number of eycles (see Figure 4-15(a)). Non-ferrous metals do not usually show a fatigue limit (see Figure 4-15(b)), although for practical purposes a large number of cycles, such as 10° cycles, is considered as the endurance limit. (a) Steels Stross S Fatigue limit tor steels (b) Non-ferrous metals (c) Steels unde? Corrosion fatigue conditions Number of cycles N (log scale) Figure 4-15: Stress vs. number of cycles for (a) steots, (b) non-ferrous metals under non- corrosive conditions and (c) steeis under corrosion fatigue conditions. When chemical attack is added to the eyclie stress, corrosion fatigue results. In a corrosive environment, the fatigue properties of a material are changed. For example, with a ferrous material, the fatigue limit is no longer observed (see Figure 4-15(c)). When corrosion and fatigue occur simultancously, the chemical attack greatly accelerates the rate at which fatigue cracks propagate, This results in large reductions in fatigue strength after even quite short exposures to a corrosive atmosphere, Some corrosion fatigue data for metals in sea water are given in Figure 4-16, along with fatigue data for similar specimens tested in air. These give a fair comparison of relative fatigue properties in this environment, even though they are based on laboratory tests which is unlikely to show crack-initiating effects of localised corrosion. The important conclusion from the data is that fatigue properties depend on corrosion resistance under test conditions, rather than mechanical properties, The steel alloys all show negligible corrosion fatigue resistance, even though normal fatigue resistance increases with strength. 422TYPES OF METALLIC CORROSION The stainless steels, and more so the high nickel alloys, show improved corrosion resistance and much greater corrosion fatigue resistance. Copper, copper-nickel and titanium alloys do not show any effect of the marine environment and their corrosion fatigue resistance is the same as resistance to fatigue in air Mild steet Brembo, 316 SS Kea. 304 SS SMe _ Copper assay fms Corrosion foigue Monet Fatigue Hastelloy C SS. Ferralium 255) Inconel 625 Ti 6Al4y 0400 700300400 son ooo 700 Fatigue Strength at 10° cycles. (MPa) Figure 4.16: Corrosion fatigue strength in sea water and fatigue strength in air of different metals. Corrosion fatigue fracture surfaces exhibit general characteristics similar to those of ‘normal’ fatigue fractures, However, many more cracks are formed on slightly different planes, whereas there is rarely evidence of more than one crack in fatigue in air. They are transgranular and may show some branching, like stress corrosion cracking. The fracture surfaces are frequently discoloured or damaged by corrosion products, but the presence of corrosion products does not necessarily indicate corrosion fatigue as superficial rusting ean oceur during ordinary fatigue. Unlike SCC, the fracture surface is rippled and shows corrosion products. Corrosion fatigue seems to be most prevalent in media which produce pitting attack, and it is the corrosion pits which act as stress raisers and initiate cracks. However, pitting is not essential for corrosion fatigue and some metals, such as carbon steels, do not show surface pitting although pitting forms on the crack surfaces once the cracks have formed Corrosion fatigue can be prevented by a number of methods. In general, the choice of a material for this type of service should be based on its corrosion-resistant properties rather than its fatigue-resistant properties. It may be reduced by lowering the stress on the component by altering the design, shot pening the surface or stress relieving. Reducing the effects of the corrosive atmosphere, such as by using corrosion inhibitors or metallic or organic coatings, will also lengthen the service life as long as they do not produce tensile stresses in the surface or introduce hydrogen. 4.23TYPES OF METALLIC CORROSION 43. LIQUID METAL EMBRITTLEMENT Liquid metal embrittlement (LME) of alloys is another form of environmentally induced cracking which may occur while a metal is in contact with liquid metals and in the presence of a tensile stress, There have been few service failures due to this cause, although the Flixborough disaster in the UK in 1974 was at least partially due to zinc embrittlement of stainless steel. J.ME arises when a normally ductile material stressed in tension becomes coated with a thin film of liquid metal causing brittle failure well below the yield stress of the metal. Embrittlement is not a corrosion process involving dissolution, but rather penetration of the liquid metal into the grain boundaries of the affected metal weakening the metal Jeading to cracking and fracture, similar to hydrogen embrittlement and cracking Only specific combinations of liquid metals and stressed solid metal result in such failures, as shown in Table 4-8, analogous to that of stress corrosion cracking. Mercury, for example, will embrittle aluminium and copper alloys, but not steel under normal circumstances, s0 large-scale mercury boilers are made of steel. As well as metals which are normally liquid, LME can arise with low melting point metals such as Jead and zinc. Embrittiement of heated leaded steels has been observed, with the degree of damage dependant of strength level of the steel and composition of the lead. Austenitic stainless steels are embrittled by zinc when above 750°C. Below this temperature a harmless zinc alloy layer will form. Embrittlement may be produced by the heat treatment or by cutting ‘or welding of austenitic steels in the presence of zinc or zinc-based paints. Also, there is a risk of cracking where there is a fire if the temperatures are above 750°C and molten zinc can come into intimate contact with stressed, hot austenitic steel. If the fire remains below 750°C, or the tensile stress is very low, a harmless superficial layer only will form. Where austenitic stainless stecls are subject to high temperatures, all steps should be taken to avoid contact with zine. Table 4-8: Susceptibility of solid metals to liquid metal embrittement. Tiguid metal Solid metal Mercury Zinc Tead Steel ¥ (iy * = Copper alloys x v x Aluminium alloys x x x x:Cracking No cracking (1) Gracking observed in noiched samples. 14. STRAY CURRENT CORROSION Stray current corrosion, commonly (incorrectly) referred to as electrolysis or electrolytic corrosion, occurs when the corrosion reactions are driven by an external direct current source, This outside current source forces current through the cell and corrosion occurs at the anode where the conventional current leaves the metal. Because it is externally induced, it is basically independent of environmental factors such as oxygen concentration or pH. Electrolytic corrosion can be severely damaging. A common source is direct-current driven trains or trams which cause damage to pipelines if the rails, the system's intended current retum path, are not completely insulated from the soil. If an underground pipe or other metallic equipment picks up a portion of this current, it will corrode at the point where the (conventional) current leaves to rejoin the original circuit (see Figure 4-17). Other sources 4.24TYPES OF METALLIC CORROSION of direct stray current are electric welding equipment grounded to underground utilities and stray direct current from nearby cathodic protection systems. Altemating current is generally not a cause of corrosion but in rare cases a small portion of the AC may become rectified to direct current which then causes corrosion. overhead cable ee TTT TT TTT TTT IA current Ur current anodic buried structure Bee current flow ——> Figure 4-17: Stray current due to an electric railway. In marine situations, stray current corrosion occurs when a metal hull or underwater meiallic fittings are used as electrical earths, The use of DC welding equipment without proper earthing can also Jead to electrolysis problems. The current finds an easier retum from the hull, through the water and back to the earth. In serious cases, hulls have perforated and ships sunk at the quayside Prevention of electrolysis requires isolation of the structure from DC power sources by ensuring all clecirical circuits, including the earth, are fully insulated. Where the problem is due to unavoidable stray currents, such as pipelines affected by electrified tramways or railways, a drainage bond is placed between the structure and the stray current source. In this way, current is passed back its source electronically and there is no electrolytic current discharge from the pipe, hence no corrosion. An altemative approach is to connect an easily replaced sacrificial anode which cortodes in preference to the structure. Insulating couplings installed along a pipeline can minimise stray current damage but caution is required because current can flow around the insulators and cause corrosion damage at multiple sources if the voltage is large. 4.25TYPES OF METALLIC CORROSION APPENDIX: SUMMARY OF MAIN FORMS OF CORROSION Formof Characteristics Typicalexample Typical unlikely Typical corrosion situation proventative measures Uniform Uniform metal Steel, zincin Stainless slecks Alloy selection, removal atmosphere, fresh and mostnon- coatings, or sea water ferrous metals in inhibitors, cathodic atmosphere protection Galvanic Preferential Metals far apart on Metals close ‘Avoid coupling cortosionon one galvanic series -—_together on, dissimilar metals, metal in a couple galvanic series insulation Grevice Localised attack in Asai, Wide crevice. high Alloy selection, crevice stainless steels in Mo alloy design to avoid marine crevices environments Pitting Localised pitting Aluminium, High Mo alloy in Alloy selection attack stainless steels in sea water, steel in marine atmosphere environments Intergranular Corrosion along —_Sensitised Most other alloys Alley selection grain boundaries stainless steels Selective Preferential High zinc brass, Most other alioys Alloy selection, dissolution of one cast iron environmental metal from alloy modification Erosion Attack from Most soft metais Hard metals, Design for turbulent flow of under flowing metals with slow streamlined flow, liquids conditions flowing fluids alloy selection Cavitation —_Fissures end pits Cast vibrating Wrought metals Design, smooth ‘caused by metals in fluids finish, alloy vibrating metal in selection fluid Fretting Erosion of slightly Vibrafing surfaces Lubricated Lubrication, moving surfaces surfaces decrease load ‘Stress Cracking of Austenitic Stainless steel in Alloy selection, corrosion _— stressed ductile stainless steel in ammonia, copper _reduce stress, cracking metal in specific chloride, copper alloys in chloride reduce environment alloys in ammonia temperature Hydrogen Cracking in High strength steel Most non-ferrous Reduce hydrogen, cracking hydrogen in hydrogen metals using softer alloy, environment environment bake hydrogen out Corrosion Cracking of metal Any metalunder Resistant metal for Alloy selection, fatigue under cyclic cyclic stresses in the environment reduce stress stresses corrosive environment ‘Stray current Corrosion dueto Any metalwhere No direct current Prevent direct corrosion electrical current high current flowing ‘current flow from passes to an metal, bonding electrolyte 4-26Chapter 5 CORROSION IN NATURAL ENVIRONMENTS Corrosion has already been defined as a reaction between a metal and its environment. In an earlier chapter, the various forms in which corrosion can occur were discussed. In this chapter, we will look at the three natural environments of most concern to the corrosionist, via + the atmosphere = water + soil. 1 must be remembered that water is always necessary for corrosion; also that many situations involve more than one of these environments. Despite these difficulties, these categories prove convenient. 4, CORROSION IN THE ATMOSPHERE In the atmosphere, corrosion depends on three factors: oxygen, moisture and the contaminants in the air (dust, salt, pollutants, etc.) Oxygen is required to sustain corrosion processes. It makes up about 20 per cent of the atmosphere and will generally be always available to take part in corrosion reactions in the atmosphere. Corrosion rates are usually controlled by the rate at which oxygen can diffuse to cathodic sites, but this is rarely a problem in atmospheric corrosion as moisture is usually in the form of thin films or droplets. Moisture is also essential for corrosion reactions to proceed, However, a metal does not have to be visibly wet for corrosion to take place. An invisible moisture film will develop at moderate levels of relative humidity. It has been established that there is a critical value of humidity below which corrosion of metals is negligible. In practice, it has been found ‘that corrosion is usually only significant in clean atmospheres wien the relative humidity is greater than about 80 per cent. In the absence of salt or other contaminants, the number of hours per year the humidity is above this figure (known as the time of wetness) has been found to be a rough indicator of corrosion rates. Humidity changes during the day (usually inversely proportional to temperature) and is highest in the early morning and this is when most corrosion takes place. Corrosion is usually low in polar regions and hot deserts where humidity is low. Corrosion rates are much higher when a visible film will form when dew formation occurs or where water arises from rain or splashing.CORROSION ENVIRONMENTS, Contaminants in the atmosphere are also important from a corrosion standpoint, Carbon dioxide (CO;), although not strictly a contaminant as it is found in all atmospheres, has litde, if any, effect on corrosion rates of most metals other than zinc. Other contaminants are more important. Sulphur dioxide (SO;), sulphur trioxide (SO), hydrogen sulphide (4,8), ammonia (N13) are often found in industrial areas. Dust, ash, soot and other particulate matter are found in industrial and urban areas. All these contaminants will have a significant effect on corrosion rates. ‘The interrelationship between some of these factors is shown in Figure 5-1 showing the effects of sulphur dioxide, particulate matter and relative humidity on corrosion rates ‘When the relative humidity is less than about 60 per cent, corrosion rates are low. Sulphur dioxide forms sulphurous and sulphuric acid in the presence of moisture and causes significant increase in corrosion rate. When the atmosphere contains suspended dust particles, corrosion will be greatly accelerated. These particles may absorb moisture or sulphur dioxide from the air, may form galvanic couples (e.g. carbon acts as a cathode) or may obscure part of the metal leading to pitting corrosion Soct and 0.01% sulphur doxide 0.01% sulpnur ‘oxide only orrsion Rate o 20 49 60 80 400 Relate Humicity (94) Figure 6-1: Etfect of relative humidity and various pollutants on the rusting of iron. Salt from the sea has been found to be a major cause of high corrosion rates. Close to the ‘sea, corrosion rates are significantly higher, but the effect depends on the roughness of the sea. Near a surf beach or other rough seas, very high corrosion rates are found right on the coast, but the marine effect can travel inland 20 or more kilometres. On the other hand, near a sheltered bay or other quieter seas, such as regions surrounded by shallow reefs, high corrosion rates are only noted right on the shore and the marine effect will travel inland only a kilometre or so, Figure 5-2, from work carried out in Australia by CSIRO, shows how corrosion rates drop away rapidly moving inland, and the differing effect of rough seas compared to a sheltered bay As mentioned above, the contaminants present vary according to the area ~ whether it industrial, urban, rural or marine. Near the coast, salt is present and carried by the wind to settle on exposed surfaces. In industrial regions, sulphur dioxide, dust and other pollutants produce a corrosive atmosphere. In both marine and industrial regions, corrosion rates are high. Pollutants are less in urban regions and almost non-existent in rural regions so the corrosion rate drops. In the tropics, corrosion rates are usually low, except near the sea. As 5-2CORROSION ENVIRONMENTS 1 factors, there are a range of atmospheric environments, as shown in Pigure a result of the: 53. ‘As well as the effect of the general or ‘macro-environment’, the local, or ‘micro- environment’ is also important. A site just downwind of a large smelter will have # higher corrosion rate than a site upwind of the same smelter, even though there may be a small distance between the sites. The corrosion rate of a site facing the sea will be much higher than one facing the opposite direction, Rural areas generally have low corrosion rates but agricultural chemicals can raise these rates considerably. Therefore, local effects have to be taken into account when considering corrosion rates 200 100 Comoswty (umyear) cain” ° 0 02 04 06 08 10 42 14 Distance trom share (km) Figure 5-2: Relationship between corrosion rate and distance from sea Other factors can also have a major effect on corrosion rates. For example, heavy rain can be detrimental if it cannot drain away or beneficial if it washes away pollutants which have collected on the metal surface. Dew and condensation are especially undesirable from a corrosion point-of-view because they will wet all surfaces, including intemal ones, they rapidly become saturated with salt and other pollutants and allow oxygen easy access to the surface. Time-of-wetness is a critical variable with respect to atmospheric corrosion as it determines duration of electrochemical process. Urban’ Rural/milé Severe Marine Industrial moderate marine wt Figure 5-3: Variations in salt content, pollution and humidity give a range of atmospheric environments. Temperature can affect atmospheric corrosion because of the increase in reaction rate which follows an increase in temperature. However, the indirect affect of temperature on 5-3CORROSION ENVIRONMENTS condensation is probably more important. Below a certain temperature, known as the dew point, condensation will spontaneously occur in the atmosphere. When the relative humidity is high, the dew point is only slightly less than the ambient temperature Therefore, as the ambient temperature rises, the temperature of metal objects tends to lag behind because of their heat capacity. This makes them act as condensers, maintaining a film of moisture on their surfaces, allowing corrosion to continue. As a result, there is no relationship between temperature and atmospheric corrosion, with both high and low rates observes in both hot and cold parts of the planet. The above commenis have mainly concentrated on the effects of the atmosphere on steel, Generally, they will apply also to non-ferrous metals although some metals behave differently under certain conditions. For example, in a sulphur dioxide environment, Jead forms a virtually insoluble sulphate shielding it from attack. Similarly, copper forms a protective sulphate in such conditions (the green patina found on copper), but much more slowly. On the other hand, zinc and cadmium form soluble sulphates which provide no protection, Table 5-1 gives a summary of corrosion rates of various metals and alloys in the major atmospheric types. These metals show far lower general corrosion rates than steel although it must be remembered that aluminium, for example, may show unacceptable pitting in marine atmospheres. Table 5-1: Approximate average long-term corrosion rates for atmospheric corrosion of various metals in ym per year, ‘Metal Urban/industrial Marine Rural “Carbon steel rT) 5-25 Aluminium 1 05 0-01 Copper 42 1 05 Lead 0s 05 0.05 Tin 15 25 05 Nicke! 35 4 02 Monel 45 02 02 Zine 246 058 02-3 2. CORROSION IN WATER ‘The term water, in a general rather than chemical sense, can refer to solutions of mixed composition ranging from distilled water, through natural waters such as rain water, fresh water and sea water, to treated and polluted waters. Over such a wide spectrum, it is not surprising to find an cnormous range of corrosion properties. It is therefore necessary to ook at the various constituents of natural waters and how they affect corrosion properties and then look at the major types of water encountered and the nature of corrosion within these groups. This material wil] focus on natural rather than treated waters. (a) Factors affecting corrosiveness of natural waters Dissolved oxygen is of considerable importance in corrosion by waters as. cathodic polarisation is one of the principal factors controlling corrosion rate. The rate of corrosion in many systems is determined almost entirely by the rate of supply of dissolved oxygen to cathodic sites. The higher the dissolved oxygen content in such systems, the greater the 5-4CORROSION ENVIRONMENTS: corrosion rate. However, in other systems such as aluminium, oxygen is important as passivating agent and it forms passive films protecting the metal from further attack. The Gissolved oxygen content in natural waters can vary from near zero to about 8 milligrams per litre/l, the level for saturated water at ambient temperatures, varying corrosion performanec. These opposite effects explain the shape of the corrosion rate curve in Figure 5-4, For a given pH, corrosion rate increases linearly with increase in dissolved oxygen content, but beyond @ maximum or critical concentration there is a reduction in corrosion rate due to the formation of a protective passive film. The maximum varies with temperature, dissolved salts, pH and water velocity and at pH 10, for example, the maximum is at 4 milligrams per litre of oxygen. ‘lr saturation at 25°C Cortesion Rete oe 4 8 2 6 Dissolved oxygen content (gf) Figure 5-4: Variation of corrosion rate of iron with dissolved oxygen content. Dissolved salts. Pure water free from oxygen is not particularly corrosive, but as soon as the water is adulterated with dissolved salts, the corrosion rate is stepped up greatly. The chemical species making up the dissolved salt content mainly consist of cations of sodium (Na’), magnesium (Mg”), calcium (Ca™*) and minor amounts of potassium (K*) along with anions of chloride (CI), sulphate (0,7), bicarbonate (HCO,) and lesser amounts of nitrate (NO5) and carbonate (CO;”). These ions account for the greater proportion of total dissolved salts found in natural water. Under certain circumstances, for example water polluted with fertilisers, phosphates, ammonium and fluoride may be present and transition metal ions of iron, copper and manganese may be found in low levels in some ground waters. The total dissolved solids content (tds) is of considerable value in determining the corrosion propensity of water and generally speaking, the higher the tds the greater the corrosion rate. Sodium and chloride ions generally constitute the bulk of the total dissolved solids, have the greatest effect on corrosion properties and have been widely studied. As the sodium chloride content increases, the corrosion rate of iron in aerated water increases up to a maximum of about 3 per cent NaCl, This is because conductivity increases so anodes and cathodes on the surface can operate further apart, and the initial Fe(OH), corrosion product forms away from the surface and does not provide a protective barrier layer on the metal surface. Above 3 per cent NaCl, oxygen solubility decreases continually with increasing sodium chloride concentration, explaining the falling away of the corrosion rate at higher concentrations. 5-5CORROSION ENVIRONMENTS ‘The pH of most waters is usually fairly neutral (around 6 to 8) but can range from 4.5 to 9. ‘The more acidic waters will tend to be more corrosive, the more alkaline less cosrosive for steel, with corrosion rate independent of pH between about pH4 and 32, For other materials the effect of pH variation may be different. For example, the corrosion rate of both zinc and aluminium is high in acidic and alkaline conditions, while for noble metals a Jow corrosion rate is found at all pH levels. Dissolved carbon dioxide may form acidic carbonic acid or scale-forming insoluble carbonates, s0 its effect will be variable. It is usually aggressive and an accelerating factor when the pH is low or neutral or dissolved oxygen content is high or both. The amount of dissolved carbon dioxide is usually low, from 0.5 to 2 ppm for rain water and 010 5 ppm for surface waters of lakes and in rivers. Higher figures are found where there is considerable decaying organic matter. Bore waters can contain up to several hundred paris per million carbon dioxide due to the presence of decaying organic matter combined with high pressure. With water in contact with limestone, the free carbon dioxide can be converted to bicarbonate. Above a pH of about 8, free carbon dioxide is entirely converted to protective carbonate by combining with the hydroxyl ions. Other dissolved gases are far less important than oxygen or carbon dioxide. Dissolved sulphur dioxide or hydrogen sulphide may be present in polluted waters which are extremely corrosive. Hydrogen sulphide is present as dissolved hydrogen sulphide unless the pH is high, in which case alkali sulphides will form. Nitrogen is a relatively inert gas and has no corrosive effect on metals. Corrosion rates generally increase with increasing temperature. However, dissolved oxygen content decreases with increasing temperature, therefore, in an open system where oxygen is allowed to escape, it will often be found that corrosion rates decrease at high temperatures. In closed systems, oxygen cannot escape with rise in temperature and corrosion rate increases with rise in temperature until all the oxygen is consumed. Figure 5- 5 shows schematically how corrosion rates vary with temperature in open and closed systems Closed Open Corrosion Rete ° 20 40 60 80 700 Water temperature °C Figure 5-5: Variation of corrosion rate of iron with temperature. Increasing the flow or velocity of the water also increases the corrosion rate. This is mainly due to increasing supply of oxygen to the cathode. There will often be a drop off in 5-6CORROSION ENVIRONMENTS: corrosion rate above a certain flow rate when passivation occurs. With a high chloride concentration, such as sea water, passivity is not established and corrosion rate continues to increase. At very high velocities, erosion effects become significant. Natural waters contain a range of micro-organisms which can play an important role in corrosion processes. In air-frec or anaerobic waters, sulphate-reducing bacteria can cause high rates of corrosion. The action of micro-organisms is especially important in soi] corrosion and the subject is covered in that section. ‘Trace metals, such as copper and mercury, can greaily increase corrosion rates, especially of galvanised and aluminium pipes. In hard waters, copper is Jargely precipitated so has little effect on corrosion (b) Corrosion behaviour in natural waters Sea water contains about 3.5 per cent salt although lower levels occur adjacent to rivers and tidal estuaries. Sodium and chloride ions are the main dissolved constituents but sulphates, carbonates and others are also present. The pH is usually between 8.1 and 8.3 Oxygen is also present and, as this is required for the cathodic reaction, the level of dissolved oxygen is important. In the open sea there is about 4 to 8 milligrams per lite of ‘oxygen, ample 10 allow rapid corrosion. But where the oxygen level is lower, such as in stagnant water, the corrosion rate is lower. This differing oxygen concentration results in widely varying corrosion rates in a marine environment. Figure 5-6 shows the corrosion rate of a steel pile embedded at sea, At the splash zone, oxygen is freely available and the corrosion rate is highest. In the mud, the rate is lower because less oxygen is available. Marine ‘atmosphere Splash Surface zone line Clean mus 0 t00z00 800-400" 606 Corrosion rate (micronstyear) Figure 5-6: Corrosion rates of bare mild steel! in seawater, The average corrosion rate of stee! in sea water is about 125 m/year, normally showing fairly uniform corrosion apart from the factors mentioned above. Stainless steels show excellent corrosion resistance when the passive film is readily maintained, Increasing the alloy content improves corrosion resistance with the austenitic 304 grade corroding at a rate of about 2.5 um/yr. With crevices, or in areas where the passive film is not maintained, rapid pitting occurs. Copper alloys generally corrode at similar rates to austenitic stainless 57CORROSION ENVIRONMENTS stecls except in the presence of sulphur compounds from decaying marine Jife or pollutants. Brasses show dezincification but copper-aluminium alloys show better corrosion resistance than other high-copper alloys. Nickel alloys behave similarly to stainless steels and generally prefer flowing aerated conditions. Aluminium alloys generally show good resistance to sea water but aluminium-copper alloys suffer from severe pitting and are not recommended for sea water usage. Fresh water may come from surface water or ground sources and typically contains less than one per cent sodium chloride. It generally has lower corrosion rate than sea water but dissolved salts have a major effect. The concentration of dissolved calcium and magnesium salts is a major influence on corrosion rate. If the concentration of such salts is high, the water is called hard and deposition of calcium carbonate forms a protective barrier. If the concentration of such salts is low, the water is soft and more corrosive than bard water. However, sodium chloride and other dissolved species along with dissolved oxygen content also play a part. Cast iron, steel and galvanised steel are the most widely used materials for handling fresh water. The inherent corrosion rate of mild steel and cast iron are essentially identical in fresh water. Copper, brass, aluminium, some stainless steels, Monel and cupro-nickel are also used where temperature, contamination ot longer life are factors, Distilled or demineralised water contains extremely low levels of dissolved solids. However, its oxygen content may vary according to its source and, if highly aerated, may be extremely corrosive because of its freedom from scaling compounds. As feedwater for boilers it must be deaerated. 3, CORROSION IN SOILS. ‘The third natural environment is soil. Soils can range from highly corrosive to mildly corrosive. The most common structures associated with soils are buried pipelines, but other structures such as buried tanks, steel piling and well or borehole casings also suffer from soil corrosion. The following factors influence soil corrosion, although their roles are usually interrelated. ‘Oxygen Content and Soil Type: As with water, the reduction of oxygen is usually the predominant cathodic process in the corrosion of buried metals. The corrosion rate is usually controlled by its rate of arrival at the metal surface in most soils, apart from those which are strongly acidic or under the influence of microbiological corrosion. Where oxygen reduction is the rate controlling factor, corrosion rates of steel are usually fairly low, ranging from about 10 to 60 microns per year. Since the oxygen must reach the metal surface by a tortuous route through the soil, the pore content of the soil is important. The pore space of a soil may contain either air or water, s0 the oxygen availability is related to the pore volume and the retained water content. High clay soils for example, have a low pore volume and often a high water content, thus the oxygen content is correspondingly low. There is a range of water contents for a given soil which will promote maximum corrosion. Lack of water will remove the electrolytic path essential to the corrosion mechanism so rates are low. Excess water will retard oxygen diffusion and promote anaerobic conditions at the metal surface, so again rates are low. 5-8CORROSION ENVIRONMENTS Intermediate conditions will provide both the electrolytic path and the cathodic reactant, with corresponding high corrosion rates. The physical and chemical properties of soil may vary along the length of a structure. With differing soil types, parts of the structure may be exposed to an environment of high oxygen content, other parts to low oxygen content. The lower oxygen area will act as an anode and corrode in preference to the higher oxygen area. In a similar manner, it is necessary to distinguish between the corrosion risks in undisturbed and disturbed soils. This formation of macrocells Jeads to a phenomenon is known to pipeline technologists as “long line* corrosion or *hot spot’ corrosion, It also plays an important part in corrosion of steel piling. For a pile driven through a number of different soil layers as shown in Figure 5-7, corrosion rates will vary at different positions on the pile, and show different behaviour from isolated items. In the upper zone, which is usually inherently corrosive for an isolated steel item, corrosion rates are low on a pile as it has the highest oxygen content and becomes cathodic in the macrocell formation. At oxygen deficient areas, the piling becomes anodic and corrosion rates are high, while for isolated items the lack of oxygen reduces corrosion rates. Metal dissolution and absence of oxygen can increase local acidity further intensifying corrosion. This region of highest corrosion rates for piling is usually just below the water table where oxygen content is lowest. For isolated specimens, corrosion rates are low in this region. Below this region, there is little or no corrosion on the pile or for isolated items, again unless there is local acidity or microbiological activity. rm Steel pile lpconertes} Isolated items Moderate 60] [oo oo 3 tonosen 0-| | oo oo | Jo Intense oo] | 0 Pores © corrosion ° oo 0 ao eo] fo mero qd joo ° coool lo oo] |cooccsoo ° 00 O[ | O Moderate Stocion ©] | Safer Gei@ Lo 9 corrosion oretton Water Low salzstes corrosion Uitte corrosion ot Figure 5-7: Mecrocell corrosion of a stee! pile, as compared o corrosion of isolated items, Water Content: The presence of water is fundamental to corrosion in soil. The water in soils is present in three forms: as capillary water retained by the capillaries of the soil particles, as gravitational water present as the result of percolation of water deriving from rainfall, and as ground water formed by accumulation of gravitational water at the water table. Corrosion is governed largely by the capillary water, since the gravitational water is transient (although it may lower the salt content) and the ground water provides a quite separate environment more like that to be expected in the fully immersed condition. Soluble Salts: The capillary water is an aqueous solution of salts derived from the soil ‘The principal cations are sodium, potassium, calcium and magnesium, whilst the anions are most commonly chloride, sulphate and bicarbonate. The salt content of soils can be markedly altered by human activities. Agricultural practices, such as fertiliser use and weed control can introduce many chemicals into the soil as can industrial waste disposal All ions increase the conductivity of the soil and thereby facilitate the corrosion reaction. 5-9CORROSION ENVIRONMENTS However, both calcium and magnesium have a tendency to form insoluble protective carbonates or hydroxides on the metal surface. Thus soils rich in calcium or magnesium (calcareous soils) tend to be less corrosive, High chloride soils tend to be very corrosive. PH: For most soils, the pH ranges from about 5 to 9 and pH is not usually an important factor influencing corrosion rate, Outside this range, the corrosion rate of most metals will increase with a decrease in pH (increase in acidity), For this reason soils rich in natural organic acids are more aggressive than alkaline soils such as limestone. Some metals, such as zinc and aluminium, however, are attacked by alkaline soils. Bacteria: Sulphate-reducing bacteria (SRB) can promote corrosion and are widely present in shallow soils. Their presence however does not necessarily indicate the certainty of microbial corrosion. Their role in corrosion is due to their ability to depolarise the cathodic reaction by using atomic hydrogen to reduce sulphate according to the reaction: $0," + 8H (in presence of SRB) -> S* + 4H,0 (Eq. 5.1) Whilst SRB are viable in aerated soils, they only become active under anaerobic conditions in the presence of organic nutrients. Clay soils are often waterlogged, have a low oxygen content and thus prone to active SRB. The growth of SRB is restricted to the pH range 6.5 to 9.5 and particularly favoured between pH 7 and 8. Thus, clay soils which generally have a pH of 7 to 8, not only provide the requisite anaerobic conditions, but also have a near- optimum pH, so can be extremely corrosive to underground structures. In practice, however, such structures will be protected by cathodic protection and the high alkalinity produced by cathodic protection is bactericidal. Resistivity: Soil resistivity is a function mainly of moisture content, but salt content, salt type and temperature of a soil can also be significant. In general, there is a correlation between soil resistivity and corrosion rate for shallow soils. This is because corrosion is an electrochemical phenomenon; thus a lower soil resistivity produces a lower circuit resistance for the corrosion cell resulting in a higher corrosion current flow, and thus a faster corrosion rate. The specific factors referred to in the above sections prevent a direct correlation between soil resistivity and corrosion rate. There are a number of tables published in the literature, such as Table 5-2, which relate severity of corrosion to soil resistivity and are useful for determining general trends. However, they can be very misleading and should be interpreted with care. Anaerobic and low pH conditions can, for example, significantly change trends Table 5-2: Relationship between soil corrosivity and electrical resistivity. ‘Soil Type Corrosivity Electrical Resistivity Q-om (a-m) ‘Sandy! good drainage Tow 6,000- 50,000 (60 - 500) Loams/ fair drainage Moderate 4,500-6,000 (45-60) Clays! poor drainage High’ 2,000-4,500 (20-45) Marsh’ organic/ wet Severe 02,000 (0-20) 5-10CORROSION ENVIRONMENTS (a) Soil corrosion processes ‘The electrochemical corrosion processes arising on a buried structure can vary from a mechanism similar to atmospheric corrosion in well aerated conditions to a mechanism similar to corrosion under completely immersed conditions. The mechanism will depend on soi] compactness, water content and various other processes discussed above, Of greatest significance in soil corrosion is the differential acration cell, As mentioned, there can be significant variations in oxygen content over very short distances or very Jong distances. Differentials in potential can arise due to factors other than oxygen availability For example, pipes of different material coupled together can lead to galvanic corrosion Table 5-3 gives a galvanic series of some metals in aerated neutral soils. Notice that new pipe is anodic to old pipe, so that replacement lengths of pipe installed in an old line can corrode preferentially. Differences in acidity, salt concentration or soil compositions can also give rise to corrosion cells. In addition, bacteria can cause corrosion. Table 5-3: Galvanic table of metals in air-saturated neutral soils Metal ‘Approx Potential: Volts toa i CulCuS04 Electrode Carbon, Graphite, Coke 40210403 Millscale on mild steel 02 Copper, brass, bronze 02 High silicon cast iron ~02 Mild Stee! (in good concrete) 02 Mild Stee! (in Crcontaminated concrete) 0310-05 Mild Stee! (rusty) 0.3 10-06 Dastile iron ~03t0-08 Lead 05 Cast Iron (not graphitised) ~05 Mild Steel (uncorroded) 06 0-08 ‘Aluminium “08 to-1.1 Zine “14 Magnesium 6 Jn practice, all these factors can play a part in the corrosion of a buried object. Most metal objects are exposed to soil by burying. This produces a substantial disturbance of the soil condition that affects its corrosivity by affecting its oxygen and water content. The excavated hole or ditch can become a drainage feature. The bottom of the excavation usually has a higher moisture content than the undisturbed soil at the same depth and may even contain free-standing water. The soil immediately above the buried object usually does not attain the former level of compaction and thus it permits higher rates of water and air ingress. As a result, pits tend to develop more on the bottom side of the pipeline than on the top side. Furthermore, the leve] of microbial infection in the excavated areas is both higher and more extensive than in the undisturbed soil. What happens to a buried structure such as a pipe in practice therefore can be quite complex and it can be difficult to predict what might happen. Not only will there be variations in the corrosivity of the soil due to changes in soil type, but there will also be differential aeration effects. Figure 5-8 shows a schematic diagram of a pipeline laid from a bill across a valley. The following effects can be expected:CORROSION ENVIRONM) TS 1) Rocky area: The soil is well-drained and of hi low. ii) Clay area: Clay is poorly oxygenated but reasonably corrosive and, in this case, anodic to cathodic oxygensted regions on either side. Corrosion rates are high posed area: An oxygenated, cathodic area with a low corrosion rate, iv) Road: Poorly acrated and anodic to well acrated regions either side. A high corrosion rate. v) River: Well oxygenated, but flowing water depolarises local cells causing a high corrosion rate. vi) Marsh: Poorly oxygenated but a high corrosive rate because a wet area and probably acidic, h resistivity so corrosion rates are Figure 6-8: Eifects of soil and environmental change on corrosivity. (b) Behaviour of other metals in soils ‘The above discussion has concentrated on corrosion of steel in soils. Graphitization of cast irons can occur when they are buried underground. Copper corrodes at about one-sixth the rate of iron, but in tidal marshes, for example the rate is comparatively higher. at about one-half that of iron. Lead corrodes on average at a Jower rate than steel but the rate increases in organic acids, especially in anaerobic conditions. The behaviour of aluminium depends on the nature of the soil and can range from being virtually unaffected to severely pitted. Stainless steels behave similarly to aluminium. Zinc coatings will increase the life of steel pipes except in acid soils. Cinders constitute one of the most corrosive environments for all metals. In one series of tests carried out over four to five years, the costosion rate of steel and zine was five times, copper eight times and lead 20 times higher than average rates in cinder-free soils. 5-12Chapter 6 CORROSION CONTROL BY DESIGN IMPROVEMENT The most effective and cheapest method of dealing with the corrosion problem is at the design stage. A designer will normally take into account function, fabrication and mechanical strength when designing a construction. By also considering corrosion aspects at the design stage, full attention can be paid to details that would otherwise lead to corrosion or interfere with the application or maintenance of protective coatings or other comosion control methods. The cost of corrosion prevention by design alterations is usually small but incorrect decisions can be extremely costly. All corrosion protection work should therefore start with the designer. Some aspects of design which should be considered by the designer are described in this chapter. The most general rule is keep it simple. For example, attempt to make structures with clean and smooth lines, try to keep conditions as uniform as possible, avoid crevices, avoid dissimilar metals and minimise stress raisers. For the reader who is not involved in design, this chapter is useful in highlighting corrosion pitfalls in design that may require special attention. Most of the examples are in terms of simplistic shapes which may not always be applicable to the design of plants or equipment. However, the principles should be apparent and able to be incorporated into more complex designs. Section 4 of AS/NZS 2312 or Part 3 of ISO 12944 provide further exaraples of good and poor design features that influence corrosion contol, especially coatings. 41. BASIC REQUIREMENTS The choice of design, materials and protective method will depend on the basic requirements of the component. The following points should be considered with regard to corrosion aspects: + What environment will the component be exposed to? + What is the type and concentration of corrosive species? + What is the expected life of the construction? + Is the construction accessible for preventative maintenance? + What are the operating conditions — temperature, pressure, velocity, etc? Ase these conditions significantly altered during shutdown, cleaning, ete? + What type or types of corrosion may be expected? Is there a risk of stress corrosion cracking, hydrogen embrittlement or corrosion fatigue? + Is there a risk of product contamination by corrosion products? If so, what level is acceptable? * Can a stoppage or close down be tolerated? 6-1CORROSION CONTROL BY DESIGN IMPROVEMENT Knowing the answers to these questions, or at Jeast asking these questions, means the designer is aware of potential corrosion problems and is taking measures to deal with them. 2. LIFE OF THE STRUCTURE AND CORROSION ALLOWANCES, It is uneconomic for the life of some elements in a plant to be much greater than that of the whole plant if they will be scrapped with it, For example, a stainless steel exhaust on a car may last much longer than the car so would not be cost effective. Sometimes, therefore, it makes most economic sense to accept corrosion and plan for it. Since corrosion is a penetrating action it is necessary to make allowances for reduction in thickness when designing pipes, tanks, ete. In general, wall thickness is usually made twice the thickness that would give the desired life. For example, if a 10 year life is required and the corrosion rate is 0.1mm per year, the thickness should be 2mm. Of course, mechanical considerations such as stress and pressure must also be taken into consideration. The designer must be aware that excessive corrosion allowances add weight and costs. 3. IMPORTANCE OF SHAPE The shape or geometry of a structure can have a marked effect on its likelihood of failure due to corrosion because of factors such as crevices, poor drainage and poor access for coating or maintenance. Some of the more important principles are considered below: (i) Simplify forms. Structures having simple forms, such as tubes, can be protected more easily and efficiently than more complex structures. The more angles, comers, edges and internal surfaces a construction contains, the greater the area that is in contact with the corrosive environment, the greater the likelihood of crevices and the more difficult it is to apply protective coatings. On offshore platforms, for exemple, bracing is eliminated as far as possible in the highly corrosive splash zone. Moreover, tubular piles are used which cut exposed surface area, eliminate sharp edges and comers and lend themselves to many different coating systems such as tape wrappings and Monel sheathing. Gi) Avoid crevices. Crevices can permit accumulation of stagnant liquid or dirt and may accelerate corrosion due to differential aeration. They are also difficult to coat. If crevices are present they should be welded or filled (Figure 6-1), Crevice Filler i i Undesirable Desirabie Figure 6-1: Corrosion prevention by avoiding crevices. ii) Avoid residual moisture. Corrosion cannot occur without the presence of moisture. A structure should therefore be designed so that moisture is not retained. Profiles should be designed so that condensed moisture can drain away (Figure 6-2(a)). In closed structures, ventilation is as important, Storage containers should be raised above the ground so they can be drained completely and so that moisture will not collect below them (Figure 6-2(b)). 6-2CORROSION CONTROL BY DESIGN IMPROVEMENT Furthermore, such a structure can be properly paimted and maintained. Roofs should have sufficient slope to allow rapid drainage of rainwater and areas where condensation occur should have provision for draining off condensed moisture. Proper heat insulation may sometimes prevent condensation (but see High Temperatures below). Typical structures susceptible to condensation include chimneys or exhausts to high temperature plants, boilers or incinerators. Undesirabio Desirable : Desrabie /; \ Incomplete Serle srinage Conosion ea Drain hole fiscas SFEEion fa) (b) Figure 6-2: Structures should be designed to avoid residual moisture, Gv) Avoid sharp corners. Sharp comers result in a thin paint coating (Figure 6-3) and should be avoided if the structure is to be coated. Furthermore, rounded corners are e: to clean. Undesirable Desirable sas ier pei Uniform paint Figure 6-3: ‘Sharp corners resull in @ thin paint film. (v) Consider access and paintability. Coating, inspection and maintenance require adequate access to all parts of the structure. Poor access makes surface preparation and coating difficult. As a general rule of thumb, the distance between two items should be one-third of the height of the items, with a minimum of 45 mm (Figure 6-4). Furthermore, it is necessary to ensure that areas which are accessible during shop fabrication do not become inaccessible on installation, making inspection and maintenance difficult. Undesiable Desirable JtItiL Lt. CLIC (vii) Provision for cathodic protection. If cathodic protection is to be applied, or may be applied at some time in the future, some initial design provisions may be necessary, or desirable on economic grounds. For example, electrical continuity or isolation of the 6-3CORROSION CONTROL BY DESIGN IMPROVEMENT structure may be impractical if not incorporated in the initial design. A cathodic protection specialist should be consulted to determine the requirements. (viii) Avoid deposit traps. Debris which will absorb or bold water, such as dirt, loose corrosion product, leaves, etc, can shield surfaces setting up differential aeration cells. Corrosion in motor cars is often due to build up of dirt in various traps within the vehicle. Furthermore, debris can influence heat transfer. The corrosion occurs out of sight beneath the deposit. The design should be such that debris cannot accumulate. Rounding comers and edges, providing access and using sloping rather than horizontal surfaces will minimise the risk of damage. 4. COMPATIBILITY OF MATERIALS Metals may not be compatible with one another, with the environment or with other materials with which they may come into contact. (i) Avoid galvanic corrosion. Different metals and alloys should not be in electrical coniact when moisture is present, especially when they are far apart on the galvanic series. The relative areas of the two metals in electrical contact will influence corrosion rates and galvanic action is particularly serious if the anodic metal has a smaller area than the cathodic metal, There are many examples of dissimilar metal couplings which have given many years of useful service, usually because of the absence of the electrolyte. However, when two metals are joined, galvanic action is always a possibility. If joining of dissimilar metals cannot be avoided, the following steps can be taken to minimise damage + Keep the anode as large as possible in the particular component + Use electrical insulators between two metals. However, ensure the material is not porous or crevice corrosion may occur. Glued joints between the metals are helpful. Remember that the insulation must be complete; a bolt requires a sleeve as well as washers of an insulating material (Figure 6-5). Incomplete insulation may be worse than none at all. + Use an easily exchangeable corroding piece (a sacrificial waster) which will protect other metals in the system. Wasters should be located in positions where they are accessible and there is room to replace them. + Apply cathodic protection in the form of galvanic anodes (must be of a metal more active than either component) or impressed current, ‘+ Use paints. Either paint the whole assembly or just the noble member of the galvanic couple. Painting only the active metal may aggravate the situation. + Exclude oxygen from the system. This will inhibit oxygen reduction which is the major cathodic reaction. If oxygen is excluded from closed water systems, even metals far apart on the galvanic series (e.g. steel and copper) can be safely joined. Figure 6-6: Insulating sleeve and washers used to avoid galvanic corrosion Ss Insulating washers ‘and socket 64CORROSION CONTROL BY DESIGN IMPROVEMENT (ii) Consider the complete environment. The designer will generally consider the normal operating environment which the structure has to withstand, and use appropriate materials or corrosion preventive methods. Ofien however, commissioning, cleaning or operation will introduce species which were not taken into consideration during the design stage. For cxample, commissioning a pressure vessel will normally require hydrostatic testing, and water (often polluted or salty) used for this purpose may wreak havoc with the vessel Cleaning of cooling systems ofien requires acids or chlorine-containing biocides which can destroy many materials, Proximity to electrical supplies can result in damage if stray currents pass between adjacent structures. It is important for the designer to consider all aspects of the environment, not just that of normal operation. (ii) Less noble metals should not be used downstream from more noble metals. For example, small amounts of copper salts can arise from corrosion of copper by water which can seriously corrode aluminium and its alloys and zine. (iv) Consider other materials. Care may be necessary when a metal is in contact with plastics, wood, concrete, etc. Organic materials, such as wood, plastics and paint, may give off aggressive vapours which can initiate and contribute to corrosion damage on adjacent metal surfaces. Acetic acid vapour, preservatives and fire retardants from wood can all pose serious problems, Plastics are generally not liable to evolve corrosive vapours and are safe to use in contact with metals. However volatile residues from manufacture can be released from some polymers when the material is new or heated. For example, poly vinyl chloride (PVC) can give rise to hydrochloric acid. Concrete is alkaline and will generally protect steel, but may cause corrosion in contact with aluminium, 5. MECHANICAL FACTORS. If mechanical stresses are involved, corrosive environments can be considered even more damaging. Stress corrosion cracking, corrosion fatigue and fretting all oceur under the action of stresses. Other forms of corrosion can be expedited by stress, Flow of fluid can likewise increase corrosion rate or result in erosion corrosion or cavitation, (i) Avoid mechanical stresses. Stresses, either residual or applied, static or altemating, should be minimised to avoid the possibility of stress-assisted corrosion. Design should take account of the time factor and the effects of general corrosion and consequent loss of section in raising stress levels. Avoid stress raisers (sharp comers, keyways, threaded screw roots, ete), excessive overhang without support and poor fit assemblies that would lead to fretting. Proper fabrication, correct startup, shutdown and operation procedures and smooth surfaces and weld beads will further assist in reducing stresses. (ii) Avoid combinations of metal and environment which are known to have failed by stress corrosion. Different metals have different environments in which stress corrosion cracking becomes a problem. The most common are austenitic stainless steels in chloride environments, copper alloys in ammonia or amine environments and steel in caustic environments. Many corrosion textbooks contain lists and should be consulted. Be aware that new metal-environment combinations are being discovered all the time. 6-5CORROSION CONTROL BY DESIGN IMPROVEMENT (iii) Be aware of flowing fMuids. Fluid movement need not damage a material. However, disturbances to flow, such as protruding parts, poorly fitting gaskets, burrs or weld metal can Jead to turbulence resulting in erosion corrosion or cavitation damage (see Figure 6-6) Sharp bends and other areas where fluid direction is changed rapidly should also be avoided. Deflectors or baffles can prevent direct impingement, Some metals have threshold velocities for a specific fluid and these should not be exceeded, Profiles should be designed to permit smooth flow. Areas experiencing local boiling effects and dead-end locations can be sites for cavitation damage. Corrosion fatigue can also be caused by vibrations in a tank or pipe produced by adjacem machinery or pumps. Pipe runs, valves and pumps should be well clamped to minimise vibration, Undesirable Desirable Undesirable Desirable nangerent Ss \x (st amar Figure 6-6: Avoid protruding sections and sharp bends in pipes. 6. HIGH TEMPERATURES Higher temperatures generally increase corrosion rates, however, raising temperature may be beneficial if it results in drying. If operating temperatures are too high, the material will scale. Another problem which can arise due to high temperatures is condensation which arises when hot gases cool. (i) Avoid hot spots, Heat exchangers and other heat transfer devices should be designed to ensure uniform temperature gradients. Uneven temperature distribution leads to local heating and high corrosion rates along with stresses that may produce stress corrosion cracking. In addition, as scale increases efficiency of the heat transfer operation decreases. Gi) Avoid condensation. Condensed moisture at the contact point of a hot gas and a cold metal surface can cause acid dew point corrosion. The condensing moisture is often not pure water, but rather aggressive solutions such as sulphurous and sulphuric acid. These have a higher dew point than water and condense at a higher temperature, earlier in the process stream. The factors for minimising residual moisture in the section on shape above are applicable for minimising problems caused by condensing moisture. (ii) Be aware of problems associated with insulation. Many heat insulation materials contain soluble salts, usually chlorides and sulphates. Condensed moisture dissolves these salts and they may then deposit on the wall of the insulated object leading to rapid attack, even of stainless steels. Lagging and other porous material can also act as a sponge and conduct water for long distances. Water-absorbent materials should be eliminated from areas of high humidity, or the outer surface should be fully sealed to prevent the ingress of water. Zinc is corroded rapidly by water at around 60°C, so galvanizing, zine-rich paints or other zine coatings should not be used where moisture may condense at this temperature. 6-6CORROSION CONTROL BY DESIGN IMPROVEMENT 7. EFFECTS OF FABRICATION ‘The various processes involved in the fabrication of structure will influence properties. Such activities as machining, cutting, stamping, drilling and heat treatment will have relatively little effect except where mentioned above, but joining can significantly influence corrosion resistance. cormosion () Fasteners should be more noble than the jointed material. In this way the favourable combiniation of small cathode - large anode is created cathodically protecting the joint (see Figure 6-7) Rivetted joint ey Store noble Figure 6.7: More noble fasteners Bolted used in fabrication joint 73 (ji) Use welding in preference to bolted joints. Joints and fastenings should be arranged to give clean, uninterrupted lines. Welds are therefore preferable to bolted or riveted joints, and butt welds are preferable to lap welds (see Figure 8). If lap joints have to be used, the appropriate welding or filling may be necessary to avoid the entrapment of moisture and dirt VN L227 ASS Best Good 1 eazeee ee = Y : Lt Poor Foo! Figure 8: Examples of joining methods. (ii) Welds should be even and well-shaped, without crevices. A smooth weld will not contain holes or crevices, and will be easy to coat. The welds should be clean and free from slag and spatter. The joint should be such that there are no exposed crevices. Continuous welds are more expensive than intermittent welds but are preferable (see Figure 9) Undesirable Desirable Figure 9: Continuous welds are preferable to skip welds. 6-7CORROSION CONTROL BY DESIGN IMPROVEMENT (iv) Weld metal should be cathodic to the base metal. For mechanical and metallurgical reasons, weld filler metal normally is more cathodic than the base metal so the favourable combination of small cathode and large anode is obtained. The problem of weld metal attack due to galvanic factors therefore is not normally a problem. However, the heating of the region cither side of the weld can cause local anodes to form and preferential attack is sometimes observed in the heat affected zone (not to be confused with intergranular corrosion of stainless steels). Stresses caused by welding can also cause stress-related corrosion, such as stress corrosion cracking or hydrogen cracking. (v) Soldered joints have few corrosion problems, Solder is usually more noble than the base metal so soldered joints have the favourable small cathode - large anode. Solder has good fluidity and crevices are usually well filled. However, fluxing agents are corrosive and should be thoroughly removed. (vi) Threaded, riveted and screwed joints have many potential problems. Such a joint will contain crevices and, as they are usually heavily worked during manufacture, corrode faster than surrounding metal. Nuts and bolts should be galvanized, or coated with another protective metal. For constructions that are dismantled often, a rust-preventive grease should be used. (vii) Glued joints have few corrosion problems. Gluing has the great advantage that it automatically gives an electrically insulated joint, which means galvanic corrosion is excluded. While glue is becoming increasingly important industrially, it has limited applicability in corrosive environments 6-8Chapter 7 CORROSION PROPERTIES OF METALS Mild steel, the most common material for engincering structures, corrodes easily and rapidly. There are many other materials which have far better corrosion resistance. However, because of its combination of mechanical properties and cost, stec] is widely used even in corrosive atmospheres. Corrosion resistant materials can be selected when the superior mechanical properties of steel are not required or other means of corrosion protection cannot be used. The corrosion properties of some common metals and alloys are described in this chapter. 1. CARBON AND LOW ALLOY STEELS Carbon steels, despite their limited corrosion resistance, are the most widely used engineering materials. The specific corrosion properties of such materials, and the measures taken to mitigate corrosion are discussed throughout these notes and will not be repeated here. Carbon, or mild, steels usually have Jess than 2 per cent by weight of alloy additions. The most important of these elements is carbon, which has the greatest effect on its structure and mechanical properties. Other elements present in carbon steels include manganese, silicon, phosphorus, sulphur and traces of oxygen and nitrogen, Generally, the levels of addition do not produce any changes in corrosion behaviour. There are a large number of Australian and overseas standards to which steel can be supplied, most of which specify mechanical properties. For example, there are standards for boiler and pressure vessel steel (AS 1548), reinforcing bars (AS 1302) and spring steels (AS 1447) ASINZS 3678 and AS/NZS 3679 are often used for general structural steelwork and specify grades such as 200, 250 and 350 based on yield strength. ‘The resistance to corrosion of carbon steels is somewhat limited. Corrosion is more rapid in industrial and marine environments than in country air. It is little affected by neutral water, but corroded by brines and seawater although the metal can often be used for these purposes if iron contamination is not objectionable. Steel should not be used with most acids, although can be used for sulphuric acid of concentrations above 90%. Steel can be used to handle alkaline materials, but there will be some corrosion and the possibility of caustic embrittlement. Steel is little affected by organic chemicals except organic chlorides. Allloy steels contain more than 1% of elements (but Jess than levels in stainless steels) deliberately added for modifying properties such as strength and hardenability. Elements added include chromium, nickel, molybdenum, phosphorus and vanadium. High strength TlCORROSION PROPERTIES OF METALS. low alloy GISLA) or microalloyed steels contain small amounts (0.005% to 0.01%) of elements such as niobium, vanadium and titanium to produce high strength steels which are used for high pressure pipelines. Small amounts of alloying elements will enhance corrosion resistance in moderately corrosive environments but in severe environments their corrosion resistance is often no better than that of carbon steel. Weathering steels have superior atmospheric corrosion resistance compared to plain carbon stecls due to the presence of the alloying elements phosphorus, chromium, nickel and copper. These cause the rust that forms on the surface to be dense and adherent so preventing further significant corrosion. These can be used in the unpainted condition but ase not suitable in an environment which is continually damp or contains ageressive chemicals. A further problem arises with rainwater draining from weathering steel which causes staining on adjacent materials, AS/NZS 3678 and AS/NZS 3679 specify steels with enhanced corrosion resistance with a yield strength of 350 MPa. Figure 7-1 compares the corrosion rate of a copper steel and HSLA steel to a carbon steel in an industrial environment. 400 ‘Structural ‘carbon steel 8 Structural copper stea! Figure 7-1: Atmospheric corrosion rates of different steels in an industrial High strength, environment. fow alloy steel 0 5 70 i820 Exposure time (yeas) Alloys containing 5 per cent chromium and ¥ per cent molybdenum are widely used in the petroleum and chemical industries in both wrought and cast forms for heat exchangers, pipes, fittings, and a wide range of process plant operating up to 500 deg C. Such steels have improved corrosion resistance compared to normal steels, but selection is based on creep resistance rather than corrosion resistance. 2. CAST IRONS Alloys of iron that have more than about 2% of carbon are called cast irons. They are relatively inexpensive and can be cast into intricate shapes because of their good fluidity and low melting points. They can be alloyed for improvement of corrosion resistance and strength Grey cast irons contain from 2 to 4% carbon and 1 to 3% silicon and are the least expensive of the engineering metals. The carbon is in the form of graphite flakes and, when fractured, the exposed surface has a grey appearance. They possess excellent compressive strength and machinability, good wear resistance and vibration dampening characteristics, Their main disadvantage is their brittleness. Grey cast irons generally have the same 72CORROSION PROPERTIES OF METALS corrosion resistance as steels, and because heavy sections are generally used, corrosion is not normally a problem. Under flowing conditions, a graphite-iron corrosion product forms on the surface giving grey ison an advantage over carbon steel under such conditions. Grey iron is subject to a form of corrosion known as graphitisation where the iron matrix is selectively leached leaving only a graphite network. Grey iron castings for engineering applications are covered by AS 1830 specified by tensile strength or hardness. White cast irons have practically all their carbon in the form of iron carbide rather than graphite and, as a result, give a white fracture appearance. These alloys are extremely hard, brittle and difficult to machine and rarely used for engineering purposes. They are used for rolls, grinding equipment and other purposes requiring abrasion resistance. Corrosion resistance is the same as for grey irons but they do not suffer from graphitisation. They are covered by AS 2027. Malleable irons are produced by heat treatment of white irons which results in graphite forming in the form of clusters or rosettes, instead of flakes. This results in @ material with similar properties to grey irons but much improved ductility. The corrosion resistance is, again, similar to that of the grey irons. Because of the additional processing, they are expensive and not widely used for engineering purposes. Malleable irons are covered by AS 1832. Ductile iron (also known as spheroidal graphite (SG) or nodular iron) contains graphite present as spheroids or nodules as a result of special treatment of the molten metal with cerium or magnesium. This results in a stronger and more ductile material. The mechanical properties can be altered by heat treatment similar to ordinary steel. The combination of properties makes ductile iron @ rather desirable engineering material and it has replaced grey and malleable cast iron and steel castings in numerous applications such as water pipes and automobile crankshafts. The corrosion properties are not significantly different from grey irons although graphitisation is less of a problem as the spheroids are not interconnected. AS 1831 covers ductile iron castings. High silicon cast irons (such as Duriron) contain 14.5% silicon, 1% carbon and the balance iron. They have unexcelled corrosion resistance to a wide range of chemicals due to the formation of a SiO, layer on the surface. They can handle sulphuric acid at all concentrations up to and including the boiling point. They will also handle nitric acid at most concentrations and temperatures and, when alloyed with 4.5% chromium, are resistant to hydrochloric acid at room temperature and severe chloride-containing solutions, such as ferric chloride. However, they are very hard and brittle and susceptible to thermal stressing and impact, and can only be machined by grinding, Austenitic cast irons contain between 12 and 36% nickel (with or without chromium and copper) and have improved corrosion, abrasion and heat resistance and mechanical properties. They are produced in flake and spheroidal graphite types. Austenitic irons (also known as the MResist series) have greatly increased corrosion resistance over the grey irons; for example, in flowing sea water the corrosion resistance of a NiResist is ten times that of a grey or low alloy iron. Austenitic irons are used for valve and pump bodies, and for condenser water boxes in refinery, marine and electricity generating systems. They are almost immune to graphitisation. AS 1833 covers austenitic cast irons. 73CORROSION PROPERTIES OF METALS, High chromium irons are basically white cast irons alloyed with 12 to 30% chromium. where the chromium confers greatly enhanced corrosion resistance These grades are intended to provide abrasion resistance under corrosive environments at high temperatures, and are used for impellers and other moving parts of pumps handling corrosive and abrasive slurries. AS 2027 specifics four grades of high chromium irons by composition 3. STAINLESS STEELS Stainless steels are iron alloys which depend on a very thin, transparent, passive surface film of chromium oxide to provide resistance to corrosion. This film will form, stay intact and repair in air at room temperature as long as two conditions are met: the matrix material must have at least 11 per cent chromium dissolved in it and + there must be a ready supply of oxygen. ‘The UNS Numbering System ‘The Unified Numbering System (UNS) was established as a means of correlating many different numbering and other proprietary designations, abbreviations and other systems for describing alloys. A UNS number consists of a single letter prefix followed by five digits, In most cases the letter 1s suggestive of the family of alloys, with A for aluminium, C for copper alloys, N for nickel alloys, for stainless stee! and P fot precious metals. Wherever feasible, existing designation humbers are incorporated into the UNS number. For ‘example, Carbon steel 1020 is given UNS G10200, 304 stainless steel is covered by $30400 and nickel alloy Inconel 625 is given UNS NO6625. Numbers are only given to metals and alloys which have @ commercial standing, and numbering is contiolled by SAE and ASTM in the USA’ The numbering system hes not has not been widely accepted outside the USA, but is increasingly used in technical literature. The numbers are not easy to recognise and memorise and the better known designation is often used, after initially identifying the alloy by the UNS number in brackets when the name first appears in the text or tables. If these conditions are met, stainless steels will exhibit excellent resistance to many corrosive environments. There arc, however, a number of problems which can arise that can cause the protective surface film to break down resulting in different types of corrosion. Pitting and crevice corrosion can arise due to localised loss of the passive film, The former occurs in the presence of chlorides, bromides and fluorides which produce deep cavities as a result of a local anodic area on the surface. The latter occurs in narrow fissures where access to oxygen is poor. Stress corrosion cracking occurs in certain types of stainless steels due to the combined influence of a chloride-containing environment, a tensile stress and temperatures above about 60°C. Intergranular corrosion occurs when the grain boundaries become depleted in chromium due to the formation of chromium carbides at temperatures between 550°C and 850°C, a process known as sensitization. The most common occurrence is due to welding where the problem is known as weld decay Erosion corrosion is not usually a problem with stainless steels for cooling water. Water velocity should be kept above about 1 metre per second to minimise deposits. Other problems which can arise when using stainless stee! include galvanic corrosion when stainless steel is joined to a more active metal, such as mild steel and high temperature corrosion in certain environments Dealing with these potential problems, along with the need to provide stainless steels with a range of mechanical properties that can be fabricated by all types of manufacturing methods, has led to development of numerous grades of stainless steels. In this section, 14,'S OF METALS CORROSION PROPERTI only the major types and common environments can be discussed. Stainless steels are generally divided into five categories according to their metallurgical structure. The most common method of designating different steels in Australia is according to a three figure AISI number (refer to AS 1449 and AS 2837). However, new proprietary alloys are being continually developed which do not use this numbering system. Ferritie stainless steels are the simplest type containing from 11 10 20 per cent chromium and only small amounts of other alloying elements, if any. Table 7-1 lists the chemical composition and major uses of some of the common ferritic stainless steels on the market. These steels are magnetic and not hardenable by heat treatment. They tend to be more resistant to stress corrosion cracking than other types, but Jess resistant to general corrosion than the more common austenitic types. In the past they were rather prone to brittle fracture and susceptible to embrittlement during welding but modern manufacturing techniques to remove carbon and nitrogen have reduced these problems somewhat. The most common grade was Type 430 containing 16 to 18% chromium. It is suitable for mildly corrosive conditions and moderate temperature heat resisting applications and is widely used for motor vehicle trim and parts, interior decoration, stove facings, kitchen utensils, etc, Low chromium stainless steels have been developed for a variety of less corrosive conditions at a lower cost and are sometimes referred to as ‘utility’ stainless steels. Type 409 was introduced in the 1970s for automobile emission control equipment. A new steel which is designed to compete with low alloy and galvanized steels and provide reasonable corrosion resistance is 3CR12 (trade name). It provides the fabrication ease of ordinary steels with the corrosion, abrasion and temperature resistance of ferritic stainless steels, Table 7-1: Common ferritic stainless stee! grades and applications Common UNS Typical Analysis (%) ‘Typical Properties and Name c Cr TTNi__ Other _ Applications ‘408 $40900 0.06 14.5 ~~ ~—~—~«OSTi__ Heal resistant steel, easily formed and welded. Mainly used for automotive exhausts (of welded applications where superior performance to galvanised steel is required 4108 S41006 008 1250 = Used for heal resistant applications up to {650°C in power plant and oil refineries, where high strength at elevated temperatures not require 430 843000 0.08 170 + Basic, ‘workhorse’ ferric grade. interior architectural components, stove and ‘automotive tim. Welds tend to be brite, Fairly good resistance to oxidising media such as nifrie acid, 444 S44400 0.02 185 = 2.0 Mo, Heat exchanger and hot water tanks, and in ‘0.4 Ti’ chloride containing waters. Not prone to chloride stress corrosion and superior resistance to pitting, crevice and intergranular ‘corrosion. Possesses excellent deep drawing Properties. 3cr12, 0.02 116 0.65 —25Ti__ Mildly corrosive and wet abrasion applications where weldabiity up to 25mm is required but esthetics not important. Ti prevents hardening during welding, Used for sugar industry, coal processing and rail transportation Sake de SORA Sere TSCORROSION PROPERTIES OF MI Higher alloyed ‘super-ferritic’ stainless steels with more than 25% chromium and 3.5% or more molybdenum are available which provide much better corrosion resistance, although austenitic or duplex grades are usually used in critical, highly corrosive applications, Martensitic stainless steels ean be hardened by heat treatment like ordinary carbon steels. ‘They contain 12 to 18% chromium, but little or no other major alloying elements However, to enable hardening carbon contents are higher than for ferritic stecls. Like ferritic steels, they are magnetic, They have high strength, but this is at the expense of corrosion resistance. They are used in applications such as steam turbine blades, valve bodies and seats, bolts and screws, springs, knives, surgical instruments and chemical engineering equipment. They are not usually used in welded fabrications, The most ‘common grade is Type 410, and there is a lower carbon version 410S (UNS $41008). Type 420 is used for cutlery, surgical instruments and general engineering applications. There are a number of type 440 grades with high chromium, high carbon, which are used where high hardness and wear resistance is required, bearings for example. Table 7-2 gives the chemical composition and applications of some martensitic stainless steels. Table 7-2: Common martensitic stainless steel grades and applications ‘Gommen UNS Typical Analysis (%) Typical Properties and Name Co Gr’ Ni Other Applications ‘410 $41000 010 125 - = Basic martensitic type. General purpose grade for use in mildly corrosive Environments, 416 S41600 0.10 125 - 02S __ Free machining variant of 410. 420 s42000 025 125 =~ General engineering uses, such as pump and valve shafts. az0c : 030 725 | Higher hardness obtained than 420. Used for ccuting tools, eurgical knives, etc. aat $4300 0.18 160 20 Higher Cr and Ni give improved corrosion resistance and mechanical properties. Harcenable stee! with corrosion resistance appraaching 02. Used for pump shafts, et Should be double tempered after hardening. aoc. 844004 1.10 17.0 - —0.40Mo_ Highest hardness due to high © content (up to 60 R,) and abrasion resistance of all the glans stools. Corrosion resitance sitar Austenitic stainless steels are the most commonly used stainless steels. They are non- magnetic and cannot be hardened by heat treatment. Chemical compositions range from the basic 18% chromium 8% nickel types to the highly alloyed types which contain up to 26% chromium and 22% nickel. Addition of molybdenum improves general and pitting corrosion resistance and increases high temperature strength. Nitrogen is sometimes added to improve pitting resistance. Carbide precipitation is prevented in austenitic stainless steels by either lowering the carbon content below 0.03% (the "L" grades) or by adding carbide-forming elements such as niobium and titanium. Austenitic stainless steels are characterised by good ductility, reasonable strength, ease of formability, good weldability and good corrosion resistance. They are, however, susceptible to stress corrosion cracking ‘Type 304 (also known as 18/8) is the ‘basic’ type with other Austenitic steels being derived from it. It is used for a wide variety of applications such as kitchen sinks, tubs, piping, vessels for food and chemical processing and indoor and outdoor architectural work. Type 16CORROSION PROPERTIES OF METALS 316 contains 2-3% molybdenum and is used in more severely corrosive conditions such as outdoors at the seaside and food and chemical processing applications. Both these grades Jable in low carbon variants (i.e. 3041, and 3161.) where welding is required. Table 7-3 gives the chemical composition and applications of some austenitic stainless steels. Table 7-3. Common austenitic stainless steel grades and applications “Common UNS Typical Analysis (&) Typical Properties and Name Cr Ni Other Applications 301 '$30100 0.08 170 76 ~ General purpase with good corrosion resistance ‘for most applications. Used for structural ‘applications such as rail carriages. Higher strength than 304 302 30200 0.08 180 90 Used for architecture, food processing, domestic sinks and tubs and deep drawing applications. Higher C then 304 giving higher strength on cold rolling, 303 $30300 0.12 18.0 9,0 258 _ Free machining stee! where extensive machining is required, Corrosion resistance and welding inferior to 302 304 §30400 0.05 185 9.0 — - Widely used, general purpose grade. Similar cotiosion resisiance to 302. Used where higher resistance to weld decay is needed in food processing, process plant, elc, 3041, 830403 0.03 185 9.5 = _—_Lewcarbon version of 304. Used in chemical Reecdom om senaizaton feared i plate 304N $3045 0.08 185 85 0.1N__Nitogen added o enhance strength 30H 90400 008 185 95. Higher C fr bigh temperature etvenath 08 30800 0.08 200 110 Used for welding rods a10 $8100 0.12 250 200 —-_High temperature siainless steel. Fumace parts and equipment. Resisiant to scaling and temperatures 900°C to 1100°C. 316 $31600 0.06 17.0 12.0 2.25Mo Molybdenum improves pitting and general ‘cortosion resistance, Used where higher Ccoitosion resistance is required, ie marine ‘equipment, chemical industry. Can be welded up fo 3mm without subsequent heat treatment. 316L $31603 0.03 17.0 14.0 2.25Mo Alow carbon mosiication af 316 where heavy Secon weldments are requred witout risk of intergranular corrosion. 316N $31651 0.08 17.0 11.0 2.25Mo Nitrogen added to enhance strength oi 3t6Ti $31635 0.05 17.0 11.0 2.25Mo Titanium stablised 316 used where resistance to (0.571 intergranular corrosion is required as well as high ‘temperature strength, iL 'S31703 0.03 19.0 13.0 3.25Mo For chemical plant. Highet Mo gives greater ‘corrosion resistance than 316 in certain ~epplications, notably in contact with brines and halogen salts, 321 S32100 0.05 18.0 10.0 STi Titanium stabilised 304, Heavy weldments in ‘chemical and other industries. Suitable for heat resisting applications to 800°C. Not sultable for right polishing. 347 $34700 0.06 18.0 10.0 O.6Nb _Nicbium-stabilised 304, Similar to 321. Higher alloyed austenitics contain less than 50% iron and have better resistance to general, pitting and crevice corrosion and stress corrosion cracking. Superaustenitics, such as 254SMO are known as “6% Mo stainless steels’. The high chromium and 6% molybdenum 77 CU 6 Peon = 4 oh 2 Doe. lect fpon BIC S/SCORROSION PROPERTIES OF MI ETALS and 0.2% nitrogen give these alloys high pitting and crevice corrosion resistance. Alloys such as 9041 are sometimes referred to as ‘4.5% Mo stainless steels’, and have a high nickel content making them resistant 10 chloride stress corrosion cracking. Such alloys oficn contain between 1 and 2 per cent copper which improves corrosion resistance in acid solutions, They often also have a nitrogen addition which improves strength and enhances resistance to pitting and crevice corrosion, There is an overlap between these alloys and nickel alloys described later. Some examples are given in Table 7-4 Table 7-4: Some highly-alloyed austenitic staintess steel grades and applications Common UNS Typical Analysis (%) Typical Properties and Name Ctr Other Applications ‘254SMO _S31254 0.02 20.0 180 6.0 Mo _Nitrogen-alloyed, high molybdenum stainless 1.0Cu steel with general, piting, crevice and SCC 0.2N resistance similar'to nickel alloys. Used in ‘seawater heal exchangers, pulp and paper industry, 9aL NOB904 0.02 200 25.0 42Mo For use under very severe conditions in process. industries. High resistance fo general, pitting anc crevice corrosion and SCC. 253MA —-$30815 1 —-2.0 11.00.15 N__ Cerium provides oxidation resistance and nitrogen 04Ce provides creep strength. High temperature Fesistant to 1 150°C, Used for furnace parts, radiant shields, fuidised beds, Sanioro 28 NOBOZ8 0.01 27 31_--3.8Mo_ High Crand Ni content gives excellent resistance 1.0.Cu to pitting and SCC. Used for handling phosphoric ‘acid and in sour oll and gas production. 654SMO —-$32654 0.01 24 © 227.3 Mo_High chromium, molybdenum and nitrogen give G.5N excellent resistance to crevice and pitting 05 Cu corrosion. Used for critical seawater handling systems, pulp and paper and chemical processing equipment © Saree Ange - The Pitting Resistance Equivalent number (PRE or] Duplex (austenitic-ferritic) PREN) is @ means of assessing resistance to pitting (and crevice corrosion resistance) when handling sea water and other process fluids. It takes into account the chromium, molybdenum and nitrogen content and the higher the figure, {the greater the pitting resistance. The calculation is made on the basis of PRE = %Cr+ 3.3%Mo + 16%N (in weight %) This provides a simple ranking and is applicable to most grades. Tungsten is sometimes included when a value of 1.%6W has been used and the number is referred to PREW. For super-austenitc steels, the nitrogen factor is sometimes 30%N. Highr Fnite 206 218 T9e atonsteele i wo 1 2 8 30 3 4045 Increasing Piting Resistance Equvatent Number (PRE) —» ‘super se ‘Soper aoe stainless steels combine the best features of the austenitic and ferritic grades. They have the toughness, weldability and good corrosion resistance of the Aus- tenitic types combined with the strength and resistance fo stress corrosion cracking of the fertitic types. Figure 7-2 shows the pitting resistance and stress corrosion cracking resistance at various chloride concentrations and temperatures of two proprietary grades compared with the two common Austenitic types. They have high chromium contents, low amounts of nickel, low carbon contents and generally contain molybdenum. The presence of ferrite in these alloys means they become brittle above 300°C, unlike austenitic grades, and are not recommended for use above this temperature. Examples are 78