UPTEC K11 018
Examensarbete 30 hp April 2012
Evaluation of corrosion in crevices where materials with different
surface coatings are combined
Ida Siggelkow
Teknisk- naturvetenskaplig fakultet UTH-enheten
Besöksadress:
Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0
Postadress:
Box 536 751 21 Uppsala
Telefon:
018 – 471 30 03
Telefax:
018 – 471 30 00
Hemsida:
http://www.teknat.uu.se/student
Abstract
Korrosion i spalter där material med olika ytbeläggningar kombinerats
Evaluation of corrosion in crevices where materials with different surface coatings are combined
Ida Siggelkow
In modern-day automobile industries the predominant type of corrosion is crevice corrosion located on door hem flanges and fasteners. Today these areas are protected by anti-corrosive agents. The appliance of these anti-corrosive agents is a costly process and could be excluded if a more corrosion resistant surface coating were to be used. In this thesis work new surface coatings are tested in an attempt to reduce the use of anti-corrosion agents in the concealed spaces. The coatings used were two types of zinc-magnesium alloys and a thin organic coating. Tests were performed in an accelerated corrosion testing-chamber with three different types of crevice models. The analytical methods used were mainly image analysis, X-ray diffraction and Fourier transform infrared spectroscopy. The results show that the thin organic coating gives the best corrosion protection according to this test method. It was also discovered that the surfaces with pure zinc coating corrodes less when in contact with a dissimilar material. Furthermore it was seen that the test method does not apply to all materials and environments.
Sponsor: Scania AB
ISSN: 1650-8297, UPTEC K11 018 Examinator: Karin Larsson
Ämnesgranskare: Leif Nyholm Handledare: Bengt Johansson
1
Table of contents
1 Introduction ... 2
2 Theory... 3
2.1 Corrosion in the automobile structure 3 2.1.1 Crevice corrosion 3 2.1.2 Corrosion products of zinc in crevices 6 2.1.2 Corrosion products of magnesium in crevices 8 2.2 Bimetallic corrosion 9 2.3 General corrosion 10 2.4 Corrosion protection of the automobile structure 10 2.4.1 Zinc-magnesium surface coatings 11 2.4.2 Thin organic coatings 14 3 Experimental ... 15
3.1 Instrumentation and methods of use 15 3.1.1 Image analysis 15 3.1.2 Electrochemical corrosion measurement 15 3.1.3 Scanning electron microscope 15 3.1.4 X-ray diffraction 16 3.1.5 Fourier transform infrared spectroscopy 17 3.1.6 Accelerated corrosion testing chamber 18 3.2 Test materials 20 3.3 Testing methods 21 3.3.1 Crevice coupons 21 3.3.2 Mini-doors 21 4 Results... 22
4.1 Image analysis 22 4.1.1 Crevice coupons (type A) 22 4.1.2 Crevice coupons (type B) 23 4.1.3 Mini-doors 24 4.2 Electrochemical corrosion measurement 25 4.3 Scanning electron microscope 25 4.4 X-ray diffraction 25 4.4.1 KSP/KSP 26 4.4.2 ZEMg/ZEMg 26 4.4.3 ZM90/ZM90 27 4.4.4 Zn/KSP 27 4.4.5 Zn/ZEMg 28 4.4.6 Zn/ZM90 28 4.4.7 Zn/Zn 29 4.5 Fourier transform infrared spectroscopy 29 5 Discussion ... 30
5.1 Image analysis 30 5.1.1 Crevice coupons (type A) 30 5.1.2 Crevice coupons (type B) 30 5.1.3 Mini-doors 30 5.2 Electrochemical corrosion measurement 31 5.3 Scanning electron microscope 31 5.4 X-ray diffraction 31 5.5 Fourier transform infrared spectroscopy 31 5.6 Other 32 6 Conclusions ... 32
7 Future work ... 32
8 Acknowledgements ... 33
9 References ... 34 10 Appendix ... Fel! Bokmärket är inte definierat.
2
1 Introduction
In modern-day automobile industries the predominant type of corrosion is crevice corrosion located on the lower parts of the body; it is also here that the metal comes in contact with the road environment. The corrosion can mainly be assigned to door hem flanges and fasteners where crevices and confined areas are formed. Today these areas are protected by anti-corrosive agent. The appliance of these anti- corrosive agents is a costly process and could be excluded if a more corrosion resistant surface coating were to be used.
Steel surfaces of automobile parts are in most cases galvanized. Galvanization offers a good protection against corrosion. Nevertheless, in the more vulnerable areas mentioned earlier, an additional corrosion protection is needed. The additional
corrosion protection can come from the appliance of anti-corrosive agents or from the use of corrosion-resistant coatings. Included in this thesis is the study of the crevice corrosion behavior of zinc-magnesium coatings and thin organic corrosion protective primers. The addition of magnesium into the zinc layer of galvanized surfaces has been shown to give an improved corrosion resistance. The magnesium can be added to the alloy either by the use of physical vapor deposition (PVD) of a thin magnesium layer on top of a zinc layer followed by diffusion through heat treatment or by adding magnesium into the melt in hot-dip galvanization. In the latter case, the maximum addition of magnesium in the zinc coating for large scale productions will be limited to a few weight percent, whilst the PVD-technique can offer a higher percentage of magnesium in the layer. As the PVD technique is more complicated, it is also a more costly option. The coating made with PVD may be more homogeneous than the hot- dipped. The thin organic coating used in this experiment is made of an epoxy matrix with additions of zinc particles.
In crevices different types of surface coated materials can meet, thus the corrosion properties of crevices with combined materials should be studied. The materials can be used with the same coating on both sides, alternatively, with different coatings on each side e.g. conventional galvanized surface on one side and zinc with organic corrosion protective primer on the other side. In this way, high surface finish on one side can be combined with enhanced corrosion protection on the other. On the
exposed surfaces Scania will probably continue to use zinc coatings for some time, in other areas the new materials might come in use. The advantages of the new
materials are in addition to corrosion properties that pre-treatment and painting can be avoided thus increasing productivity and in concealed spaces increasing quality.
If these new surface coatings are to be used extensive testing has to be done to assure its corrosion resistance and corrosion behavior, particularly in the mentioned areas of the automobile body. The aim of this thesis work is to evaluate corrosion in crevices were combinations of material with different coatings have been used.
3
2 Theory
2.1 Corrosion in the automobile structure
In metal constructions crevices can be found between joint parts and in flanges. The crevice geometry gives rise to a confined environment, where the diffusion of
corrosion reactants is limited; especially the low concentration of O2 is of importance.
In larger crevices stagnation of water is possible, making way for waterline corrosion, where the electrolyte fluid becomes depleted of oxygen. Other types of corrosion of the automobile body are galvanic corrosion and general corrosion. Bimetallic
corrosion arises from the difference in corrosion potential when two different metals are connected while in contact with an electrolyte. In atmospheric corrosion the whole metallic surface is evenly corroded with cathode and anode areas all over the
surface.
2.1.1 Crevice corrosion
Crevices are formed, among other places, behind spot-welded overlays, under the rim of sheet metal which has been folded to give a smooth outer edge or beneath loose inert fastenings (Trethewey & Chamberlain,1988). Crevice corrosion is consequently found in these areas as the circulation of the electrolyte is strictly limited. As electrolyte enters the crevice it enables a conducting path between the metals creating the crevice. The geometry of the crevice may allow the fluid to stay until it evaporates and also limits the diffusion of reactants, such as oxygen and carbon dioxide, into the crevice. The rate of transport of reactants into the crevice can then determine the corrosion rate.
Studies have shown that the crevice gap distance is of great importance to the corrosion rate inside the crevice. A larger gap distance more resembles an open surface and therefore the conditions and reactions are more similar to the way open surfaces corrode. A smaller crevice gap has reduced diffusion possibilities of
reactants and enables fluid to stagnant. A smaller crevice also increases the capillary forces and so prolongs the drying time (Zhu et al., 2000). Where crevice corrosion is found, the gap has a typical width range from 25 to 100 µm, in a crevice smaller than 25µm water cannot penetrate and in larger crevices other mechanisms of corrosion are expected (Trethewey & Chamberlain, 1988).
The initiation step of crevice corrosion is either the creation of a differential cell or galvanic corrosion. Similarly to general corrosion, crevice corrosion starts with the reduction of oxygen and the oxidation of metal. Since the reduction of oxygen in crevices is most often occurring faster than the diffusion of oxygen into the crevice, the electrolyte becomes oxygen deficient. This oxygen deficiency leads to the
formation of an oxygen concentration cell (also known as differential aeration cell). As the higher oxygen concentration is found at the crevice opening, this is where the cathodic reaction (1) takes place. The corresponding anodic reaction (2) takes place on the inside of the crevice.
4
Negative chloride ions, migrates to the anodic areas forming complexes with the positive metal ions and water molecules. These complexes undergo hydrolysis
creating metal hydroxides and hydrogen ions reducing the pH. This can be described by the simplified reaction:
However, not all metals undergo hydrolysis. Small ions or ions of low charge are more likely to undergo hydrolysis than for example a large divalent ion such as Zn.
An active corrosion cell is self-sustaining because the metal lost through corrosion forms additional deposits which in turn create new differential aeration cells, and result in more corrosion (Trethewey & Chamberlain, 1988). A schematic picture of crevice corrosion can be seen in figure 1.
(a) (b)
Figure 1. (a) Initial conditions of crevice corrosion. Corrosion proceeds evenly throughout the area.
(b) Final conditions of crevice corrosion. An oxygen differential cell is formed.
The liberation of hydroxide ions due to the cathodic reaction leads to an increased pH at the cathodic sites, located outside or at the wider parts of the crevice. The metal dissolution at the anodic sites produces cations, which, as mentioned earlier, in some cases undergo hydrolysis. The hydrolysis reaction creates hydrogen ions and metal hydroxides as shown in reaction (3). The hydrogen ions cause the pH-value to drop that further facilitates the metal dissolution. The formation of metal ions further decreases pH inside the crevice and at the same allows more chloride ions to diffuse into the crevice.
5
Another phenomenon found in crevices is waterline corrosion, also known as partial immersion cells, illustrated in figure 2. Waterline corrosion occurs in larger crevices where water is stagnant for a longer period. Corrosion depletes the electrolyte of dissolved oxygen and a differential aeration cell is formed at the waterline. The anode is formed under the waterline and cathode just at the waterline, see reaction (1) and (2).
Figure 2. Formation of waterline corrosion of zinc, seen from above (Stoltz Länta, 2010).
Inside a crevice the waterline is in constant movement due to the varying speed of evaporation of water. The speed of evaporation decreases with time because of precipitation of corrosion products. As the evaporation of water proceeds the more concentrated the electrolyte gets, in combination with slow evaporation this makes the edged and corners of the crevice more exposed and therefore also more corroded (Persson et al., 2007).
6
2.1.2 Corrosion products of zinc in crevices
The three main corrosion products within confined zinc surfaces are: zinc oxide (ZnO), hydrozincite (Zn5(OH)6(CO3)2) and simonkolleite (Zn5(OH)8(Cl2)∙H2O). Their relative appearance depends on the surrounding atmospheric environment and the degree of confinement (Zhu et al., 2000). On a newly produced zinc surface a thin film of zinc oxide can be seen as a grayish layer. This layer is a few nanometers thick and somewhat protective against further corrosion. As in reaction (1) and (2), where M is Zn, hydroxide anions and zinc cations are produced. We can expect that the zinc cation and hydroxide anion reacts into zinc oxide in the overall reaction:
(Hosking et al., 2007)
The formation of zinc hydroxide is favored by a neutral or alkaline wet environment and Zn(OH)2 is not a commonly found product. Second to the formation of oxides is the formation of zinc carbonates. The formation of zinc hydroxyl carbonate
(hydrozincite) is favored by the high pH values found at the cathodic sites. The formation of carbonates is shown in reaction (6). Hydrozincite formation can be explained by the following reaction:
As the electrolyte inside the crevice evaporates carbon dioxide more easily diffuse trough and dissolves in the electrolyte, thus inducing the following pH decreasing reactions:
(Zhu et al., 2000)
A lowered pH combined with high concentration of chloride ions enables the
formation of simonkolleite. In simple terms simonkolleite is formed by replacing the carbonate in hydrozincite with chloride (Prosek et al., 2008). The pH and chloride dependence is shown in figure 3 where it can be seen that simonkolleite is unstable at high pH values.
Figure 3. Stability diagram for zinc showing the formation of zinc oxide and simonkolleite at varying pH and chloride concentrations. (Lindström et al., 2000)
7
The formation of simonkolleite can chemically be described with this reaction:
Since the simonkolleite is formed in the lower pH-regions (see figure 4), it is naturally found at the anodic sites (Hosking et al., 2007). Both reaction (6) and (7) change the composition and properties of the electrolyte. Volovitch et al. propose that with growing carbonate-ion concentration simonkolleite can transform back into hydrozincite or smithsonite. The hydrozincite then decompose into zinc oxide (Volovitch et al., 2009).
As the crevice geometry limits the transport of carbon dioxide into the corroding areas, this affects the corrosion products. In the presence of carbon dioxide, the carbonates will react with the hydroxide ions thus lowering the pH. If no carbon dioxide is allowed to enter the crevice the pH will consequently rise, making way for different kinds of corrosion products. In the absence of carbon dioxide, zincite is the dominant corrosion product, while in presence of carbon dioxide zinc hydroxy
carbonates and simonkolleite are the dominate products (Lindström et al., 2000). The passivating properties of hydrozincite make it more protective than other zinc
corrosion products, most likely lowering the corrosion rate. In summary one can say that a low concentration of carbon dioxide in chloride containing environments increases the corrosion rate (Zhu et al., 2000).
As mentioned earlier the formation of corrosions products is pH dependent due to their chemical properties, also the crevice geometry creates a condition of varying pH. Combined, this leads to a theoretical distribution of corrosion products as shown in figure 4.
Figure 4. Schematic drawing of the theoretical distribution of corrosion products of zinc inside a crevice.
8
2.1.2 Corrosion products of magnesium in crevices
In figure 5 we can see that magnesium is more active than zinc in a NaCl-solution (sea water) and therefore the dissolution of magnesium on the surface will precede the dissolution of zinc. Magnesium oxide is formed according to the following reaction:
Magnesium oxide is, unlike zinc oxide, an insulator and is insoluble in water which means that it hinders oxygen reduction and therefore also lowers the corrosion rate.
In the presence of water magnesium hydroxide is more thermodynamic stable than magnesium oxide, therefore in wet environments we can assume the following reaction:
Magnesium hydroxide can also be formed by the reaction between magnesium ions and hydroxide ions released in reaction (1). This reaction is shown below.
(Hosking et al., 2007)
The hydroxide concentration is high at the cathodic site and consequently the reaction will take place there. In contrast to zinc, magnesium does not form soluble corrosion products at high pH which implies that the cathodic areas will be covered with the precipitated magnesium hydroxide. In atmospheric environments,
magnesium hydroxide may be transformed into magnesium carbonate (magnesite):
The formed magnesite is then hydrated and nesquehonite is formed as in reaction (12).
Further reaction will lead to formation of hydromagnesite:
The released hydroxyl anions may be consumed in reaction (6) creating more
carbonates. The hydrated magnesium hydroxide carbonate is stable at pH≥7 and it is therefore considered quite protective against further corrosion as it extends the
passive region into more alkaline conditions (Hosking et al., 2007).
9
2.2 Bimetallic corrosion
For bimetallic corrosion to occur two different metals have to be in electrical contact and immersed in an electrolyte, if so a galvanic cell may be formed. Every metal and alloy has its unique corrosion potential. The force driving the corrosion arises from the difference between the corrosion potentials of the two metals, where a higher difference implicates a larger driving force. However, a difference in the corrosion potential of minimum 50 mV has been set as practical limit for bimetallic corrosion to occur. The metal with the lowest corrosion potential serves as anode and the metal with the highest serves as cathode causing the anodic metal to corrode. Even two metals that otherwise have good corrosion resistance may in contact with each other and an electrolyte suffer from severe corrosion.
The cathode-to-anode ratio of surface area greatly influences the corrosion rate. If the cathode area is larger than the anode area the corrosion rate is high (Jones, 1996). Not all dissimilar metal couples corrode. Corrosion only occurs if a galvanic cell is formed. If it is possible to avoid either the metals from being connected or the presence of a conducting path, a galvanic cell cannot be established. When
evaluating the risk of bimetallic corrosion it is of great importance that the correct potential values are used, determined for the right solution and temperature
(Trethewey & Chamberlain, 1988). The galvanic series, shown in figure 5, list metals and alloys in the order of the corrosion potential they exhibited in flowing sea water.
Figure 5. The galvanic series in sea water.
10
2.3 General corrosion
The presence of water is of great importance for atmospheric corrosion to occur. A thin water film on a metal surface is sufficient for the corrosion to begin. The
temperature affects the rate of atmospheric corrosion, higher temperature increases the reaction rate but also reduces the solubility of oxygen. A higher temperature may also dry the surface and slow down the corrosion. The temperature also affects the relative humidity, if the temperature falls below dew point, the air becomes saturated with water and water droplets condense on exposed surfaces. Many metals, for example, steel, iron, copper and zinc corrode if the relative humidity exceeds 60%.
To minimize atmospheric corrosion the most effective way is to seal the components in impervious coatings after removing the atmosphere (e. g. painting) (Trethewey
&Chamberlain, 1988).
2.4 Corrosion protection of the automobile structure
The most common measure for reducing corrosion of the car body is metallic coating of steel surfaces and especially galvanizing with zinc. The metallic coating serves two purposes; it acts like a barrier to the environment and often serves as cathodic
protection, meaning that the coating is sacrificially corroded. Metal coatings for corrosion protection are usually applied by hot dipping, electroplating and clad coating. The two most used methods for zinc depositing on metal is
electrogalvanizing and hot dip zinc coating. In electrogalvanizing the steel surface is immersed into an electrolyte containing zinc salts. The steel surface serves as the cathode in an electrolytic cell in which a thin layer of zinc is deposited. The thickness of the coating is limited to a maximum of 20-25 µm. In terms of hot dipping the
structure is dipped into a bath of molten zinc creating a layer consisting of an alloy of zinc and iron containing several different phases, followed by a top surface layer of pure zinc. It is possible to make coatings with a thickness of up to hundreds of µm.
The time to the appearance of red rust is linearly proportional to the thickness of the zinc coating (Jones, 1996). To further improve the corrosion protection most
automobile parts are both phosphated and painted.
11
2.4.1 Zinc-magnesium surface coatings
Previously research has shown that zinc alloy coatings containing magnesium on steel have markedly superior corrosion resistance than conventional zinc coatings.
The magnesium layer is either applied with physical vapor deposition (PVD) or by traditional hot dipping. The PVD coating is obtained by applying a layer of
magnesium on top of an electrogalvanized steel surface. The temperature is then increased until it exceeds 250 oC making the magnesium diffuse into the zinc layer, thus obtaining the intermetallic phase Zn2Mg (Prosek et al., 2008).The hot dipping method is performed by mixing magnesium into the melt, often with addition of
aluminum to stop the diffusion of zinc between the melt and the steel. Cross sections of the zinc-magnesium coatings applied with the different techniques can be seen in figure 6.
Figure 6. Cross sections of the ZnMg-coatings applied by hot dip galvanizing and PVD on a steel (CRS) sheet (Stoltz Länta, 2010).
A variety of different theories to why zinc-magnesium alloys give improved protection can be found in the literature. Adding magnesium to the alloy changes the galvanic properties, due to the more electroactive properties of magnesium (Volovitch et al., 2009). Prosek et al. detected a layer of oxidized magnesium on untreated zinc magnesium and showed that this oxide layer grew thicker with increasing amount of magnesium in the alloy. Magnesium oxide is an insulator while zinc oxide is an n-type semiconductor. The insulating character of the magnesium oxide film hinders oxygen adsorption on the surface thus decreasing the oxygen reduction rate. The decrease in oxygen reduction rate slows down the total rate of corrosion. This means that at a zinc-magnesium surface has a lower oxygen reduction rate than a pure zinc surface (Prosek et al., 2008).
According to Nishimura et al. (Nishimura et al., 2000) one of the mechanisms behind the good corrosion resistance of zinc-magnesium coatings is that the addition of Mg makes the corrosion product simonkolleite stay stable for a long period. The
corrosion product simonkolleite is also packed more densely thus probably hindering the diffusion of oxygen and therefore also hindering the cathodic reaction, the latter in agreement with the results of Volovitch et al. (Volovitch et al., 2009). Prosek et al.
claims that the corrosion resistance stems from the formation of a magnesium-based surface film with reduced conductivity due to the low band gap of magnesium oxides.
As the prolonged presence of simonkolleite is said to be one of the factors resulting in increased corrosion protection one should have in mind that the magnesium also promotes the formation of hydrozincite which is considered to be less protective than simonkolleite (Volovitch et al., 2009). Volovitch et al. and Hosking et al. (Hosking et al., 2007) have both found shifts in the peaks of simonkolleite in the XRD-data, suggesting the uptake of magnesium.
12
Volovitch et al. has taken the help of Hydra-Medusa© software program to illustrate the equilibria of the corrosion, see figures 8-10. The calculations were made for an aqueous solution of Zn2+, Mg2+, Cl- and at 25oC. In figure 7a we can see that Mg2+ ions will preferentially precipitate with carbonate, thereby lowering the
carbonate concentration and stabilizing the protective simonkolleite phase (Fig. 8b and c). In figure 7 it is shown that the critical carbonate concentration to form zinc carbonates and hydro carbonates is significantly increased in the presence of magnesium due to the formation of more stable magnesium carbonates.
Figure 7. Example of the zinc precipitation product composition as a function of total Mg2+
concentration.
13 Figure 8. Example of evolution of the settlement product composition, simulation: (a) magnesium products as s function of pH; (b) zinc products as a function of the total carbonate concentration in the absence of Mg2+; (c) zinc products as a function of the total carbonate concentration in the presence of
equal concentrations of Mg2+ and Zn2+.
Figure 9. Examples of pH regions for precipitation of different zinc compounds (a) in absence of Mg2+, (b) in presence of equal concentrations of Mg2+ and Zn2+.
14
It may also be noted that in the absence of Mg2+, zinc carbonate is stable even in neutral and low basic pH solutions (up to pH 8, Fig 9a) while in the presence of Mg, this product can only be formed in the acidic region (pH <7, Fig 9b). Volovitch et. al.
concludes that this means that the zinc salts should stay stable for more dry-wet cycles with zinc-magnesium coatings than with the zinc coatings. This indirect protection from Mg is important for a long time dry-wet corrosion resistance during which larger amount of carbonate can be accumulated (Volovitch et al., 2009).
Prosek et al. discovered that alloying Zn with more than 16 wt% magnesium cancel the beneficial effects of the coating. This is probably due to the fact that the corrosion potential of these alloys in chloride solution is much lower than that for ZnMg-alloys with lower magnesium content. At such low potentials hydrogen evolution takes place and the water decomposition, consuming a greater number of electrons than normal oxygen reduction. Together, this leads to a faster degradation of the metal. Prosek et al. concludes that the addition of magnesium to zinc is positive as long as the
corrosion potential is high enough to avoid hydrogen evolution (Prosek et al., 2008).
2.4.2 Thin organic coatings
Organic coatings act primarily as physical barriers between the substrate and the corrosive environment. The coating consists of a polymer film often applied as a liquid, by brushing, rolling or spraying. Organic coatings may consist of several layers having different functions against corrosion. However a polymer film is not ideal under all conditions. Inadequate adhesion, absorption of water leaving the film swollen, oxidation that decomposes the film and hydrolysis, which especially at elevated temperatures can decompose and reduce the strength of the film, are problems that could arise. It is known that organic zinc-rich primers, such as the one used in this thesis, are not suitable for environments with pH<5 or >10, where the zinc is more readily dissolved. A polymer film often has a lower hardness than a metallic surface layer making it possible for damage to occur more easily (Jones, 1996). In figure 10, a cross-section of the organic corrosion protective film used in this experiment is shown. The protective film consists of an epoxy matrix with addition of zinc particles.
Figure 10. Cross-section of the organic corrosion protective film (Supplier meeting; Scania-VoestAlpine, December 2008).
15
3 Experimental
3.1 Instrumentation and methods 3.1.1 Image analysis
The AxioVision Rel. 4.8.1 software program for image analysis was used to study the relative coverage of rust on the coupons. The program recognizes preselected colors in digital photographs and calculates the percentage of these areas in proportion to the whole test area. After full time exposure, for each batch, the coupons were disassembled and photographed. The reference coupons were evaluated visually every week by photographing the surfaces. The relative coverage of rust was in both cases determined through digital image analysis. After eight weeks some materials showed signs of red rust. The red rust areas were marked in the same way as the white rust areas and the areas were added together.
3.1.2 Electrochemical corrosion measurement
The open circuit potential (OCP) was measured versus a saturated calomel electrode reference in 0,1M NaCl-solution. The OCP is the electrical potential difference
between two terminals of a device whit out external load. When the OCP is measured versus a reference electrode, e.g. a saturated calomel electrode, the measured
potential is the value of the materials corrosion potential.
3.1.3 Scanning electron microscope
A scanning electron microscope equipped with energy-dispersive X-ray spectroscopy (SEM-EDS) was used to confirm the absence of sodium in the white corrosion
products, to not confuse these with NaCl crystals from the ACT-chamber.
In a SEM, the sample is scanned with a high-energy beam of electrons produced by heating a tungsten filament. When the atoms in a sample are bombarded with electrons backscattered or secondary electrons are ejected from the sample. An electron detector measures the number of electrons that spread from any point on the sample surface. The detector's output signal, which is proportional to the number of detected electrons, is converted to a digital signal stored by the control computer.
The computer then creates an image using the different intensities of the output signal.
By using a detector for characteristic X-rays one can combine SEM with EDS. When the atoms in a sample are bombarded with electrons other bound electrons are lost from the different electron shells of the atoms. The energy state of the atom is now higher than normal. Because of nature’s struggle for the lowest possible energy state, electrons move from the outlying electron shells into the excluded electrons orbits.
The rebound from the higher unstable energy state to the lower stable state releases energy, this involves emission of X-ray photons. The energy of the photon is unique for each element which allows the identification of the elements in the sample. The released photons are detected by a detector consisting of silicon or germanium.
When a photon hits the detector crystal a current pulse is formed whose value is determined by the photon energy creating a spectrum of elements (Åsa Gustafsson, poster at Scania).
16
3.1.4 X-ray diffraction
To detect the crystalline corrosion products, X-ray diffraction (XRD) was performed.
The instrument was an Bruker D8 advanced, with a parallel beam through a Göbbel mirror, with an 1.2 mm divergence slit, and a long soller slit on the detector. The radiation was Cu K with the energy dispersive detector, SolX. A goniometer in the θ-θ layout was used which means that the sample position was fixed and horizontal.
The diffraction patterns were matched against ICCD PDF 4+ (version 2008) with filter elements of Ca, S, O, P, Al, F, Cl, Na and H. The matching was performed with the EVA software.
X-ray Diffraction is a non-destructive analytical technique which provides information about the internal lattice of crystalline substances in the form of a diffraction pattern.
Nearly all crystalline solids have a unique X-ray diffraction pattern making it possible to characterize the compounds of the sample.
X-ray diffraction is based on constructive interference of monochromatic X-rays and a crystalline sample. The X-ray radiation commonly used is that emitted by copper. The X-ray beam is filtered to produce monochromatic radiation and collimated to make the beam aligned and directed towards the sample. The interaction of the incident rays with the sample produces constructive interference (and a diffracted ray) when the conditions satisfy Bragg's Law (nλ=2d sinθ), see figure 11.
Figure 11. Bragg's law of reflection. The diffracted X-rays exhibit constructive interference when the distance between the upper beam path and the lower beam path differs by an integer number of
wavelengths (λ) (Wikipedia).
Bragg’s law relates the wavelength of the electromagnetic radiation to the diffraction angle and the lattice spacing in a crystalline sample. These diffracted X-rays are then detected, processed and counted resulting in a unique diffraction pattern (Atkins et al., 2006).
17
3.1.5 Fourier transform infrared spectroscopy
The amorphous products, which are hard to detect with XRD, where characterized with Fourier transform infrared spectroscopy (FTIR). Fourier transform infrared spectroscopy (FTIR) is used to obtain information of the presence of several functional groups. A spectrum can be contained from a liquid, gas or solid sample.
Figure 12. Schematic diagram of a Michelson interferometer, configured for FTIR (Wikipedia).
In the FTIR spectrometer the beam of an IR radiation source is split by a Michelson interferometer, shown in figure 12, so that it reflects simultaneously from a moving mirror and a fixed mirror leading to interference. After the beams recombine, the combined beam passes through the sample making the molecules vibrate. The beam then reaches the detector and is recorded as raw data. To convert the raw data into a spectrum Fourier transformation is required. The frequency of a given stretching vibration in an IR spectrum can be related to both the masses of the bonded atoms and the relative stiffness of the bond. The location of the IR absorption band is given by the specific vibration frequencies of the compound. An IR spectrum contains several peaks, even for a relative simple compound, making the possibility that two different compounds having the same IR spectra very small (Solomon & Fryhle, 2004).
FTIR spectra were recorded using the standard potassium bromide (KBr) tablet technique. Tablets were obtained by grinding the sample in a mortar with KBr. The concentration of the sample in the KBr tablets was ≈ 0.5 wt%.
18
3.1.6 Accelerated corrosion testing chamber
The corrosion testing was performed using Accelerated Corrosion Testing (ACT) following the Scania standard STD4319. The latter standard specifies the test
procedure to simulate atmospheric corrosion conditions in a controlled way. It is also set to assess the corrosion resistance of metals in environments where there is a significant concentration of chloride ions, in attempt to mimic the conditions in a mainly marine environment and on the roads in presence of de-icing salts. The ACT is a climate chamber that cycles the humidity (%RH) and temperature as shown in figure 13. The major cycle, which is run the better part of the week, is called sub- cycle 1 and a rain cycle, which is run twice a week, is called sub-cycle 2.
Figure 13. The ACT cycles (Scania standard STD4319).
Sub-cycle 1 can be divided into four different steps, 1:1-1:4. During the first step (1:1) the temperature and the humidity are kept at constant values at 35OC and 95 %RH, respectively, for four hours. During the second step (1:2), the temperature is
increased to 45 OC whilst the relative humidity is decreased linearly to 50 % for a period of two hours. In step three (1:3), the temperature is kept at 45 OC and the relative humidity at 50 % for four hours. During step four (1:4), the temperature decreases from 45 to 35OC simultaneously with an increase of the relative humidity from 50 % to 95 %, for a period of two hours. Then the cycle is repeated again. The second sub-cycle contains five steps (2:1-2:5) where the first consists of a 15 minute spray with a 1.0 % NaCl-solution. In step (2:2) the conditions are kept constant at 35
OC with a relative humidity of 95-99 % for one hour and 45 minutes. Step one and two are then repeated two times giving a total run time of six hours. Step (2:3) is a drying period with a relative humidity of 50 % and the temperature increasing from 35 to 45OC over a two hour time period. For step (2:4) constant conditions are kept at 45
OC and 50 % RH during two hours. Step (2:5) consists of a temperature decrease from 45OC to 35 OC with an accompanying increase in relative humidity from 50 % to 95 %, again over a two-hour period. In figure 14, sub-cycle 1 and sub-cycle 2 are visually described.
19 Figure 14. Sub-cycle 1 and 2 (Scania standard STD4319).
The rain in the chamber is composed of a 1 ± 0.1 wt % sodium chloride solution acidified with 1 ml of 0.5 M sulphuric acid for every 10 liters of sodium chloride solution resulting in a pH of 4.2 ± 0.1.
Inside the chamber, the coupons and mini-doors were placed in plastic holders at a 20 degree angle vertically with a mutual distance of approximately 30 mm. Figure 15 shows the placement of the coupons in the test chamber and the spreading of the rain from the nozzles while figure 16 shows the placement of the crevice coupons and mini-doors in this experiment.
Figure 15. Example of climate chamber (ACT) (Scania standard STD4319).
Figure 16. Picture of the inside of the ACT-chamber after positioning the test material.
20
3.2 Test materials
The tested materials were galvanized steel in combination with two kinds of zinc- magnesium surface coatings and one kind of organic surface coating, see table 1.
Table 1. Test materials.
Name Coating material Coating thickness
ZEMg Zn,16%Mg (PVD) Zn: 6.0 μm ZnMg: 2.2 μm
ZM90 Zn,2% Mg, 2% Al (HDG) ZnMg: 6.5 μm
Zn Zn (HDG, Z100) Zn: 7.5 μm
KSP Epoxy and zinc particles (ZE 75/75) Zn: 7,5 μm Org: 2.5/4.1 μm The magnesium PVD-coating was applied on electro galvanized steel and thereafter heat-treated giving an alloy containing 16% of magnesium. The PVD-coating was supplied by Arcelor Mittal. The 2 % zinc-magnesium coating, supplied by
Voestalpine, was applied on a steel substrate with a 6.5 μm layer of zinc on each side. The thin organic coating named KSP (Korrosionsschutzprimer) was built up on a steel coupon based on an electro galvanized surface (ZE75/75) with a zinc layer of approximately 10 μm on each side. The KSP-coating is made of epoxy matrix with additions of zinc particles. It had a thickness of 2.5 μm on the front side and 4.1 μm on the backside. The KSP-coating was supplied by Voestalpine.
Seven different combinations of the materials (see table 2) in 5 batches with triple test of each combination were put in the ACT chamber. The samples were exposed to the corrosive environment for 1, 2, 4, 8 and 16 weeks. The galvanized steel was set to be the backpiece, when possible, in an attempt to resemble a real case of automotive corrosion where the galvanized steel would be in surplus.
Table 2. Test matrix of crevice coupons type A respectively mini-doors.
Backpiece / lintel resp.
external plate / inner plate
• Zn / Zn
• Zn / ZM90
• Zn / ZEMg
• Zn / KSP
• ZM90 / ZM90
• ZEMg / ZEMg
• KSP / KSP
To compare the crevices with possible galvanic corrosion to those without, a few reference coupons with glass lintels were added to the test matrix. These were photographed every week.
Table 3. Test matrix of crevice coupons type B
Backpiece / lintel
Zn/glass
ZM90/glass
KSP/glass
21
3.3 Test methods 3.3.1 Crevice coupons
To simulate a confined situation, such as spot-welded overlays or hem flanges, coupons with confined metal surfaces were used. Each sample coupon was cleaned with degreasing agent and ethanol before masking and assembly. The backpiece of crevice type A had a size of 10x9 cm and the lintel of metal was 2,5x9 cm. The corresponding values for crevice type B were ~10x7,5 cm and 2,5x7,5 cm. After masking the top of the backpiece with anti-corrosion tape, a Teflon spacer were attached to the backpiece and the backpiece and lintel were assembled with plastic clips. The Teflon spacer separates the two pieces by a controlled gap reaching from 0 to 250 µm, as is illustrated in figure 17. In figure 18 both types of crevice coupons are illustrated.
Figure 17. Schematic drawing of the contolled gap inside the crevice.
Figure 18. Photographs depicting type A and B crevices respectively.
3.3.2 Mini-doors
More realistic testing was done by using mini-doors. These are harder to reproduce than crevice coupons as small variations between the mini-doors easily arise during production. The mini-doors were assembled according to Swedish standard SS 18 60 25. The crevice gap inside the mini-doors varied between ~30 and 200 µm. In figure 20, a close up picture of the mini-door crevice is shown.
Figure 19. Close up of mini-door crevice.
22
4 Results
4.1 Image analysis
4.1.1 Crevice coupons (type A)
In figure 20 and 21 the area percentage of rust inside the crevice is plotted against time. The vertical bars represent the standard deviation. Photographs of some of the coupons can be seen in Appendix 1.
Figure 20. Area percent corrosion on the lintels plotted against time.
From figure 20 it is seen that the lintels of Zn/Zn and ZEMg/ZEMg are the ones with the highest area percent corrosion and that KSP/KSP and Zn/KSP are the ones with the least area percentage of corrosion.
Figure 21. Area percent corrosion on the backpiece plotted against time.
From figure 21 it is seen that the backpiece of Zn/Zn and ZEMg/ZEMg are the ones with the highest area percent corrosion and that KSP/KSP and Zn/ZEMg are the ones with the least area percentage of corrosion. After eight weeks, red rust could be observed in the coupons of ZEMg/ZEMg, Zn/KSP (at the zinc surface) and also on the Zn/Zn coupons.
0 20 40 60 80 100
1 3 5 7
Area percent corrosion
Weeks
Comparison of the material combinations, lintel
KSP/KSP ZEMg/ZEMg ZM90/ZM90 Zn/KSP Zn/ZEMg Zn/ZM90 Zn/Zn
0 20 40 60 80 100
1 3 5 7
Area percent corrosion
Weeks
Comparison of the material combinations, backpiece
KSP/KSP ZEMg/ZEMg ZM90/ZM90 Zn/KSP Zn/ZEMg Zn/ZM90 Zn/Zn
23
4.1.2 Crevice coupons (type B)
Figure 22. Area percent corrosion of the glass references.
In figure 22 it is seen that the Zn surface had a full coverage of corrosion products after 8 weeks of exposure, whilst ZM90 had approximately 50% coverage. The one with the least area percentage of corrosion was KSP. Photographs of some of the coupons can be seen in Appendix 2.
Figure 23. Area percent corrosion of zinc with different lintels.
In figure 23 the effect of combining the materials is shown. There is a demonstrable difference in area distribution of corrosion of Zn combined with glass, Zn and KSP.
One can see that zinc has the least surface area percent of corrosion when it is combined with KSP and the most when combined with the glass.
0 20 40 60 80 100
3 4 5 6 7 8
Area percent corrosion
Weeks
Glass lintel
KSP ZM90 Zn
0 20 40 60 80 100
1 2 3 4 5 6 7 8
Area percent corrosion
Weeks
Zinc comparison
Zn/KSP Zn/Zn Zn/Glass
24
4.1.3 Mini-doors
In figure 24 and 25 the area percentage of rust inside the crevice is plotted against time.
Figure 24. Area percent corrosion on the external plate plotted against time.
From figure 24 it is seen that the Zn/ZM90 and Zn/Zn external plate were the ones with the highest area percent corrosion and that KSP/KSP and Zn/KSP were the ones with the least area percentage of corrosion.
Figure 25. Area percent corrosion on the internal plate plotted against time.
From figure 25 it is seen that the ZM90/ZM90, Zn/ZM90 and Zn/Zn internal plate were the ones with the highest area percent corrosion and KSP/KSP and Zn/KSP were the ones with the least area percentage of corrosion. After eight weeks, no red rust was observed on the latter.
0 10 20 30 40 50 60
1 2 3 4 5 6 7 8
Area percent corrosion
Weeks
Comparison of the material combinations, external plate
KSP/KSP ZEMg/ZEMg ZM90/ZM90 Zn/KSP Zn/ZEMg Zn/ZM90 Zn/Zn
0 10 20 30 40 50 60
1 2 3 4 5 6 7 8
Area percent corrosion
Weeks
Comparison of the material combinations, internal plate
KSP/KSP ZEMg/ZEMg ZM90/ZM90 Zn/KSP Zn/ZEMg Zn/ZM90 Zn/Zn
25
4.2 Electrochemical corrosion measurements
Table 4. The open circuit potential (OCP) measured versus a saturated calomel electrode reference electrode (+0.242 V vs SHE) in 0.1 M NaCl-solution.
Material Initial OPC (ca 5 min)/V OCP after ca 1-2 h/V
ZEMg -1.25 -1.01
ZM90 -0.99 -1.03
KSP -0.960 -0.99
Zn -0.97 -0.99
4.3 Scanning electron microscopy
Figure 26. EDS-analysis of the white corrosion products.
The EDS-analysis confirms the absence of sodium in the white corrosion products, see figure 26.
4.4 X-ray diffraction
By comparing the results from the XRD analysis of batch one and four (1 week and 8 weeks exposure) one can see that the zinc peak was higher in the case of the zinc- magnesium coated surfaces than for a zinc coating not containing magnesium. This means that the residual coating weight of ZM90 and ZEMg is greater than that for the zinc coating. This is also consistent with that observed for the organic coating. The weight of the pure zinc phase decreases more for ZM90 than for ZEMg. Furthermore, the peaks for the zinc corrosion products were more intense in the case of ZM90 and ZEMg than in the case of pure zinc coatings, showing that the corrosion product of these has a higher tendency to stay on the surface. The corrosion products on the ZM90 and ZEMg surfaces contained mixed cation carbonates (hydrotalcites). In the diffraction patterns, the presence of hydrotalcites is seen as a vague broadening of the lower parts of the peaks. All the diffraction patterns can be seen in Appendix 3.
Also the surfaces of pure zinc or those combined with pure zinc contained more hydrozincite than the other surfaces.
In figure 27-33, the relative percentage of the corrosion products is plotted. The notation “ZnMg” refers to the phase MgZn2. The oxides ZnO, MgO and (Zn1-xMgx)O all have similar diffraction angles, which means that they cannot be distinguished from each other in these diffraction patterns. In the charts below they are all referred to as oxides.
26
4.4.1 KSP/KSP
Figure 27. Summary of the XRD-analysis of KSP/KSP; batch one and four.
4.4.2 ZEMg/ZEMg
Figure 28. Summary of the XRD-analysis of ZEMg/ZEMg; batch one and four.
27
4.4.3 ZM90/ZM90
Figure 29. Summary of the XRD-analysis of ZM90/ZM90; batch one and four.
4.4.4 Zn/KSP
Figure 30. Summary of the XRD-analysis of Zn/KSP; batch one and four.
28
4.4.5 Zn/ZEMg
Figure 31. Summary of the XRD-analysis of Zn/ZEMg; batch one and four.
4.4.6 Zn/ZM90
Figure 32. Summary of the XRD-analysis of Zn/ZM90; batch one and four.
29
4.4.7 Zn/Zn
Figure 33. Summary of the XRD-analysis of Zn/Zn; batch one and four.
4.5 Fourier transform infrared spectroscopy
FTIR was used to confirm the presence of hydrotalcites, no additional information came from this analysis.
30
5 Discussion
5.1 Image analysis
The results from image analysis show unequivocally that the KSP surface coating gives the best corrosion protection. When interpreting these diagrams one should have in mind that the uncertainties in the image analysis are ± 5 %, which means that it is difficult to draw any direct conclusions regarding small percentage differences.
5.1.1 Crevice coupons (type A)
The KSP surface coating withstands the corrosion test best according to the results of the image analysis. Figure 21 and 22 show that the ZEMg surfaces had a lager surface coverage of corrosion products than e.g. ZM90; this may be due to the effects of the decreased corrosion potential described by Prosek et al. (Prosek et al., 2008).
5.1.2 Crevice coupons (type B)
The results from the image analysis of the glass references correspond well with the results from the image analysis of the type A crevice coupons. After eight weeks the ZM90 type B crevice had corroded slightly less than the type A crevice of the same type. The corresponding trend was also seen for the KSP crevices. The Zn crevices corroded more in the type B than in type A crevice. This may be due to that water stays on the glass for a longer period of than then on a zinc surface thus making corrosion possible for a longer time during the dry cycles in the ACT-chamber.
5.1.3 Mini-doors
It was observed that the corrosion rate inside the mini-doors crevice was lower than that for the crevice coupons. The differences between the results for the crevice coupons and the mini-doors may be due to the difference in crevice geometry and the position of the crevice relative to the rain in the ACT-chamber, see figure 16.
Making the water stay a longer or shorter time inside the crevice clearly affects the corrosion process. To find out which of the results that are most consistent with reality, field-testing should be done.
For both mini-doors and crevice coupons it was observed that the corrosion process of the open surfaces was more rapid than that inside the crevice, although crevice corrosion is said to be much more aggressive. The reason for this is that the ACT- chamber is designed to accelerate corrosion of open surfaces and not crevices. A more suitable test method may be to only wet the crevices.
31
5.2 Electrochemical corrosion measurements
Table 5, below, was created by taking the difference between the open circuit potentials for the different materials in table 4.
Table 5. Difference in OCP between materials.
Material combination Difference after 5 min/V Difference after 1-2 h/V
Zn/KSP 0,01 0
Zn/ZEMg 0,28 0,02
Zn/ZM90 0,02 0,04
In table 5, it is shown that the difference in corrosion potential between Zn/ZEMg exceeds 50 mV, making it theoretically possible for bimetallic corrosion to occur. The results from the image analysis, however show no indications of bimetallic corrosion effects. As mentioned earlier, a difference in the corrosion potential of minimum 50 mV is generally needed for bimetallic corrosion, to occur. One should, however, have in mind that the 50 mV limit is set for practical reasons and does not mean that no corrosion is possible whatsoever.
5.3 Scanning electron microscopy
Analyzing the white corrosion product with SEM was a simple way to confirm that they in fact were corrosion product and not originating from NaCl crystals. By doing this simple analysis it became clear that saline deposits from the salt in the ACT- chamber would not affect the result of the image analysis. An analysis with XRD also confirmed the absence of sodium.
5.4 X-ray diffraction
The results from the XRD analysis showed that residual coating amount of ZM90 and ZEMg was greater than that for the zinc coating. This means that a smaller amount of zinc reacted with the environment to create corrosion products in the former cases. It was also observed that the amount of pure zinc decreased more for ZM90 than for ZEMg. Drawing the conclusion that ZEMg had a better corrosion protection than ZM90 from this result is probably not correct. The differences in the results between ZEMg and ZM90 are probably mainly due to that these surfaces had different
structures. The organic surface coating had the lowest zinc content and one can assume that the zinc particles near the surface were rapidly oxidized. X-ray diffraction has a limit of detection of ~1 wt% and cannot detect amorphous
substances without difficulties. Because of the amorphous characters of hydrotalcites and hydrozincite their relative percentages should not be compared with those of other studies. Due to lack of time only one of the three coupons of each batch were analysed with XRD, resulting in uncertainty in the result which made it harder to see specific trends.
5.5 Fourier transform infrared spectroscopy
The FTIR results reinforced the results of the XRD analyses and also further
demonstrated the presence of hydrotalcites. In which way the hydrotalcites affect the corrosion is not yet fully understood, and should be further investigated.
Unfortunately the FTIR instrumentation suffered from malfunction and only a few coupons could be studied.
32
5.6 Other
At the start of this thesis work it was planned to weigh all the assembled coupons before and after exposure and thus determine the mass of corrosion products inside the crevice. After the first batch was taken out, it however became clear that the corrosion products outside the crevice weighed much more than those inside the crevice the results of this measurement was therefore discarded.
On cut edges of the crevice coupons and mini-doors, we could see a trend towards cut edge corrosion. Cut edge corrosion is corrosion at cut edges of overlapping plates and plate ends and is due to the exposure of unprotected steel. Corrosion at cut edges can spread if not treated in time. In aggressive environments, like the one in the ACT-chamber, it is therefore necessary to protect the exposed cut edges before exposure.
6 Conclusions
The image analysis results show unequivocally that the KSP surface coating gives the best protection against crevice corrosion. It also was shown that no bimetallic corrosion could be detected and that zinc is positively affected by being combined with ZM90, ZEMg and KSP. The results also show that mini-door crevices corrode at a lower rate than the crevices of crevice coupons and that the test method does not fit all materials and environments.
7 Future work
If all coupons and mini-doors of this study could be analyzed further the results would become more reliable, also new interesting findings could be done.
By combining the results from the image analysis with the results from the characterisation of the corrosion products and depth profiles of the corrosion products it would be possible to predict the rate of crevice corrosion in the future.
Field test should be done to find out which of the test methods (crevice coupons or mini-doors) that is most consistent with reality.
Further development of alternative test methods for crevice corrosion.
33
8 Acknowledgements
I would like to thank my supervisors at Scania, Bengt Johansson and Karolina Soltz Länta for all the help and support.
Thanks to Hannu Rantala, Tina Holmgren and Thomas Edenström for participating in my thesis work and contributing with helpful points of view.
Special thanks to the department of UTM, especially UTMR, at Scania for giving inspiration and knowledge and lending me equipment
Thanks to Dan Jacobsson, Dan Persson and David Lindell at Swerea KIMAB for instructions and helpful points of view.
Also thanks to my supervisor at Uppsala University, Professor Leif Nyholm.
34
9 References
P. Atkins, T. Overton, J. Rourke, M. Weller, F. Armstrong, Inorganic Chemistry, (4th ed.). (2006) Oxford: Shriver & Atkins.
N. C. Hosking, M.A. Strom, P.H. Shipway, C.D. Rudd, Corrosion resistance of zinc- magnesium coated steel, Corrosion Science 49 (2007) p. 3669-3695.
Åsa Gustafsson, Poster at Scania.
D. A. Jones, Principles and prevention of corrosion, 2nd edition, N.J: Prentice Hall, Englewood Cliffs (1996).
R. Lindstrom, J-E Svensson, L-G Johansson, The influence of zinc in the presence of NaCl – the influence of carbon dioxide and temperature, Journal of the
Electrochemical Society 147 (2000) p.1751-1757.
K. Nishimura, K. Kato, H. Shindo, Highly corrosion-resistant Zn-Mg alloy galvanized steel sheet for building construction materials, Nippon steel technical report No. 18 (2000) p. 85- 88.
D. Persson, A. Mikhailov, D. Thierry, In situ studies of the corrosion during drying of confined surfaces, Materials and Corrosion 58 (2007), p.452-462.
T. Prosek, A. Nazarov, U. Bexell, D. Thierry and J. Serak, Corrosion mechanism of model zinc-magnesium alloys in atmospheric conditions, Corrosion Science 50 (2008), p 2216-2231.
Scania standard STD4319, Accelerated corrosion test - atmospheric corrosion (2008), Södertälje.
T.W.G. Solomon, C.B. Fryhle, Organic chemistry, (8th ed.). (2004) New Jersey: John Wiley & Sons, Inc.
K. Stoltz Länta, Evaluation of crevice corrosion and penetration of paint into crevices on powder painted material, Scania technical report (2010), Södertälje.
K. R. Trethewey, J Chamberlain: Corrosion for students of science and engineering.
Longman Scientific & Technical. Essex, England (1988).
P. Volovitch, C. Allely, K. Ogle, Understanding corrosion via corrosion product characterization: I. Case study of the role of Mg alloying in Zn-Mg coating on steel, Corrosion Science 51 (2009) p.1251-1262.
F. Zhu, Atmopheric corrosion of precoated steel in a confined environment (Doctoral Thesis), Royal Institute of Technology, Stockholm (2000)
10 Appendix
10.1 Appendix 1: Photos of type A crevice coupons After 1 week in the ACT
broad crevice/narrow crevice
After 2 weeks in the ACT After 4 weeks in the ACT After 8 weeks in the ACT KSP/KSP
ZEMg/ZEMg
ZM90/ZM90
Zn/KSP
Zn/ZEMg
Zn/ZM90
Zn/Zn
10.2 Appendix 2: Photos of type B crevice coupons
3 weeks KSP ZM90 Zn
4 weeks
5 weeks
6 weeks
7 weeks
8 weeks
10.3 Appendix 3: Diffraction patterns
