Corrosion Science 80 (2014) 191–196

Contents lists available at ScienceDirect

Corrosion Science

journal homepage: www.elsevier .com/ locate /corsc i

Galvanic corrosion of metal/ceramic coupling

0010-938X/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.corsci.2013.11.024

⇑ Corresponding author. Tel.: +49 351 2553 7793; fax: +49 351 2554 108.E-mail address: Michael.Schneider@ikts.fraunhofer.de (M. Schneider).

M. Schneider ⇑, K. Kremmer, C. Lämmel, K. Sempf, M. HerrmannFraunhofer Institute for Ceramic Technologies and Systems, Winterbergstr. 28, 011277 Dresden, FRG

a r t i c l e i n f o a b s t r a c t

Article history:Received 4 June 2013Accepted 19 November 2013Available online 26 November 2013

Keywords:Galvanic corrosionA. Ceramic

The galvanic corrosion risk of metal/SiC-based ceramic coupling in 3.5 wt% NaCl aqueous solution isinvestigated. Electrochemical measurements as linear sweep voltammetry and galvanic corrosion testsaccording to DIN-50919 were used. The results clearly show that the resistivity of the SiC-based ceramicplays the key role in the understanding of galvanic corrosion between metal and ceramic. In all cases theceramic represents the cathodic site of the coupling. The corrosion rate is dominated by the electricalresistance of the ceramic.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Modern engineering structures are frequently created as mixedmaterial constructions with various couplings of materials. Cou-plings of two or more dissimilar metals or metals with electronconductive non-metals are highly susceptible to corrosion underthe following conditions [1–3]:

(i) The materials are in electrical contact.(ii) The materials are exposed in the same electrolyte.

(iii) A potential difference (usually >50 mV [3]) has to existbetween the different materials.

Under these conditions the material coupling represents a gal-vanic cell where the material with the more negative electrode po-tential forms the anode and its corrosion can be severelyaccelerated. This so called galvanic corrosion is well known and re-ported in an abundance of papers over many decades (see e.g. [4–19]). A brief but impressive introduction into the theory of galvaniccorrosion and the most important influencing variables is given byOldfield [33]. The examples of galvanic corrosion are very wide-spread from orthopedic implants [13] and welds [19] to automo-tive manufacturing [12,14], aircraft construction [11] andshipbuilding [10]. Extended investigations and theoretic treat-ments of galvanic corrosion between couplings of dissimilar struc-tural materials are reported by Mansfeld and co-workers [34–37].However, the same corrosion mechanism takes place on microscalein the case of heterogeneous microstructures of materials, e.g.,

age-hardenable aluminum alloys [20,21], duplex steels [22] andcermets [23,24] also designated as ‘‘internal bimetallic corrosion’’[23]. Sometimes cermets are used as protective coatings to en-hance hardness and wear resistance [25]. Thereby, a galvanic cou-pling between the ceramic coatings and the substrate material(e.g., steel) can be created, depending on the defects of the coatings(e.g., pores, cracks). The special case of metal ceramic couples onthe microscopic scale, so-called metal matrix composites (MMC),is relatively intensively investigated (see e.g. [26,27]). In contrastto this, the corrosion behavior of the macroscopic material com-pound consisting of a metal ceramic coupling is de facto next tonothing reported in the electrochemical focused literature. In thelate 1950, Podbreznik [28] tried to define the term ‘‘contact corro-sion’’ generally as corrosion of a material in contact with anotherand distinguished between the corrosion of metal/metal contacts,the corrosion between non-metallic materials and the corrosionof a metal/non-metal coupling. This classification he reported onthe corrosion of aluminum in contact with concrete and mortar.However, nowadays the term ‘‘galvanic corrosion’’ is usually usedas corrosion between dissimilar metals which are in electrical con-tact. By definition ‘‘galvanic corrosion’’ is the corrosion betweenvarious electronic conductive or semi-conductive materials whichare in contact and exposed in the same electrolyte.

In this work the authors just present the results of the electro-chemical investigation of a metal/ceramic coupling. The ceramic isa semi-conductive SiC-based material. Therefore, the term galvaniccorrosion is correctly used in view of the aforementioneddefinition.

Owing to their extended behaviors in view of tribology, wearand corrosion protection, SiC-based materials are in widespreaduse under high chemical and abrasive stress, e.g., in bearings, sea-lings in chemical industrial facilities, automotive industry andpower plant engineering.

Fig. 2. FESEM micrograph of SiC II showing the SiC (gray area), residual graphite(dark area) and residual Si (white area).

Table 1Resistivity of the used material. The data of the metals originated from [32].

Material SiC I SiC II Ni DC01a

Resistivity (X cm) 250 0.01 7 � 10�6 1 � 10�5

a The resistivity of DC01 is assumed to iron.

192 M. Schneider et al. / Corrosion Science 80 (2014) 191–196

2. Materials and methods

2.1. Materials

The material couple consists of a metallic part and a ceramicpart. The metallic parts were used in both cases and were purenickel (99.99 wt%; Chempur) as well as unalloyed steel DC01(0.12 wt% C; Georg Martin GmbH) with hypoeuthectoide micro-structure. Two different SiC-based ceramics were used in the pres-ent work. One of them, EKasic�D (ESK Ceramics GmbH) later termedSiC I, is a solid phase sintered silicon carbide (SSiC) containing asmall amount of aluminum (0.5 wt%) as a sintering additive. Theincorporation of aluminum into the silicon carbide grains resultsin a high doping level and increases the electric conductivity. Themicrostructure of SiC I is dominated by a relatively fine grainedSiC-phase (Fig. 1). Additionally, a small amount of residual carbonand very few small pores are visible in the micrograph.

The other one was the material SiC30 (Schunk Kohlenstofftech-nik), later termed as SiC II. SiC30 is declared as a SiC-graphite com-posite material (�35 wt% C, �3.0 wt% free Si) [38]. Themicrostructure of SiC II consists of large areas of un-reacted carbonsurrounded by a three-dimensional network of SiC and free silicon(Fig. 2). The material is produced by the reactive Si infiltration ofcarbon. Resistivities of the materials are shown in Table 1.

Before electrochemical experiments the metal samples wereonly cleaned in ethanol and rinsed with deionized water. The sur-face area amounts to 10 cm2.

The SiC-based materials were electrically contacted and embed-ded in epoxy resin. The free surface was 1 cm2. All ceramic sampleswere mechanically ground, polished with a diamond suspensionup to 1 lm diamond grain size and rinsed with deionized waterprior to the electrochemical investigation.

Fig. 3. Schematic drawing of the experimental setup investigating galvaniccorrosion following the DIN-norm 50919 [29].

2.2. Methods

The electrochemical experiments on the individual materialswere carried out in a conventional 3-electrode design using a flatcell. A computer controlled potentiostate Autolab 30 (Metrohm)was used for the linear sweep voltammetry (LSV) as well as thegalvanic corrosion measurements. As a reference, electrode workson a saturated calomel electrode (Sensortechnik Meinsberg). A plat-inum sheet was used as counter electrode. All measurements werecarried out in aerated sodium chloride solution (3.5 wt%, pH 6.5).

The galvanic corrosion experiments were carried out followingDIN-norm 50919 [29]. The norm requires a measuring arrange-ment as schematically shown in Fig. 3. The distance between both

Fig. 1. FESEM micrograph of SiC I revealing the relatively fine grained structure(various gray areas) and some C-inclusions and pores (dark area).

electrodes amounts to 3 cm. The relation of the surface betweenthe ceramic samples and the metallic materials amounts to 1:10.This relation considers that in technical constructions the ceramicis often embedded in a distinctly larger metallic design.

3. Results and discussion

Fig. 4 shows the open circuit potentials (OCPs) of nickel andsteel (Fig. 4a) as well as the SiC-based ceramics (Fig. 4b). TheOCP is stable after only few minutes. In the case of the pretreatedSiC-based ceramic the OCP is identical to the unpretreated ceramicstate after 30 min. This is caused by the formation of the oxide filmwhich was removed by the pretreatment in hydrofluoric acid.According to the discussion in Section 1 the potential differenceDE between the material SiC I–Ni (DE � 165 mV), SiC I–DC01(DE � 475 mV) and SiC II–DC01 (DE � 400 mV) fulfills the thermo-dynamic precondition for galvanic corrosion. The potential differ-ence between SiC II–Ni (DE � 90 mV) is close to the criteriadiscussed in Section 1. With respect to the usual variation of thepotential of technical materials it seems possible that the potentialdifferences can be too small in a given case.

0 10 20 30 40 50 60-550

-500

-450

-400

-350

-300

-250

-200

-150

Ni

DC 01

E SCE /

mV

t/min

Fig. 4a. OCP of the used metallic samples in aerated 3.5 wt% NaCl (pH 6.5).

-300

-250

-200

-150

-100

-50

0

50

SiC I without pretreatmentSiC I with pretreatmentSiC II without pretreatmentSiC II with pretreatment

0 10 20 30 40 50 60t/min

E SCE /

mV

Fig. 4b. OCP of the used ceramic samples in aerated 3.5 wt% NaCl (pH 6.5) (the greycolored graphs represent the OCP of samples which were pretreated in 1 wt% HF-solution before measurements).

-600 -400 -200 0 20010-5

10-4

10-3

10-2

10-1

100

101

102

103

Ni

SiC II

SiC II j I

/ µA

cm

-2

ESCE / V

Fig. 5a. LSV on Ni, SiC I and SiC II in aerated 3.5 wt% NaCl (dE/dt = 1 mV s�1).

-600 -400 -200 0 20010

-5

10-4

10-3

10-2

10-1

100

101

102

103

DC 01

SiC II

SiC II j I

/ µA

cm

-2

ESCE / V

Fig. 5b. LSV on DC01, SiC I and SiC II in aerated 3.5 wt% NaCl (dE/dt = 1 mV s�1).

M. Schneider et al. / Corrosion Science 80 (2014) 191–196 193

In a further experiment linear sweep voltammetry is carried outto characterize the individual electrochemical behavior of theinteresting materials in the sodium chloride electrolyte (Fig. 5). Re-cently, Sydow et al. reported on the electrochemical behavior ofSSiC ceramics in acidic as well as in alkaline aqueous solutions[30].

The SiC-based ceramic indicates a very small corrosion currentdensity in the range of a few nanoamps per square centimeter forSiC I (jcorr � 1 nA cm�2) and SiC II (jcorr � 10 nA cm�2). Neverthe-less, it should be emphasized that the corrosion current densitiesof SiC I and SiC II are significantly different. The oxygen reductionis the dominating reaction in the cathodic sweep (Eq. (1)):

O2 þ 2H2Oþ 4e� ! 4OH� ð1Þ

The cathodic Tafel-factor of the both SiC-based ceramic(approximately bc � �30 mV dec�1) indicates that the oxygenreduction is charge-transfer controlled. Whereby, a very thin na-tive film on the surface must be assumed, despite the HF-pretreat-ment. The charge transfer takes place trough the interface SiC/native oxide/electrolyte. At higher cathodic over potentials the cur-rent density reaches the level of the oxygen diffusion limited cur-rent density (jlimit � 20 lA cm�2 in an aerated and non-stirredsolution). In the anodic sweep the SiC-based ceramics form a pas-sive layer consisting of SiO2 [30]. The anodic slope of the curveamounts to more than 450 mV dec�1. The calculation of an anodicTafel-factor is not serious because the number of involved elec-trons it depends on is not really known, because various reactionsof the formation of SiO2 from SiC are possible (see [30]). However,it seems beyond dispute that the passive layer that is formed hasdetermined the kinetic of the anodic sweep. It must be emphasizedagain that the current density in the range of passive layer (jpass) isclearly smaller for SiC I (�5 nA cm�2) than for SiC II (�50 nA cm�2).

Nickel shows a corrosion current density of jcorr � 0.5 lA cm�2.The oxygen reduction is charge-transfer controlled at low cathodicover potentials (bc � �30 mV dec�1) and diffusion-controlled athigher over potentials. The anodic sweep is firstly determined bythe native oxide film on nickel. At potentials higher than 100 mVSCE

the current density rises, because the active dissolution of nickelbegins which leads to a further passive layer formation (out ofthe measured range) according to [31]:

Ni! Ni2þ þ 2e� ð2Þ

Ni2þ þ 2H2O! NiðOHÞ2 þ 2Hþ ð3Þ

The behavior of DC01 in the cathodic sweep (Fig. 5b) is compa-rable with nickel. The jcorr � 3 lA cm�2 is slightly higher than onnickel and mainly determined by an oxygen diffusion limited cur-rent density. The anodic sweep is characterized by charge con-trolled dissolution (bc � �35 mV dec�1) of iron according to:

Fe! Fe2þ þ 2e� ð4Þ

0 5 10 15 20 25 30

-600

-500

-400

-300

-200

-100

0

100 Ecouple jcouple ESiC I ENi

t/min

E SCE /

mV

0,00

0,02

0,04

0,06

0,08

0,10

IjI /

µA c

m-2

Fig. 6a. Galvanic current and the contact potential Ecouple of the SiC I–Ni coupling(in 3.5 wt% NaCl). Additionally, the individual OCP of Ni and SiC I before and afterthe experiment is presented. (The absolute value of the current is normalized on thearea of the SiC sample.)

0 5 10 15 20 25 30

-600

-500

-400

-300

-200

-100

0

100 Ecouple jcouple ESiC I EDC01

t/min

ESC

E / m

V

0,0

0,5

1,0

1,5

2,0

2,5

3,0

IjI /

µA c

m-2

Fig. 6b. Galvanic current and the contact potential Ecouple of the SiC I–DC01coupling (in 3.5 wt% NaCl). Additionally, the individual OCP of DC01 and SiC I beforeand after the experiment is presented. (The absolute value current is normalized onthe area of the SiC sample.)

0 5 10 15 20 25 30

-600

-500

-400

-300

-200

-100

0

100 Ecouple jcouple ESiC II ENi

t / min

ESC

E / m

V

0,00

0,02

0,04

0,06

0,08

0,10

IjI /

µA c

m-2

Fig. 7a. Galvanic current and the contact potential Ecouple of the SiC II–Ni coupling(in 3.5 wt% NaCl). Additionally, the individual OCP of Ni and SiC II before and afterthe experiment is presented. (The absolute value of the current is normalized on thearea of the SiC sample.)

0 5 10 15 20 25 30

-600

-500

-400

-300

-200

-100

0

100

E SC

E / m

V

Ecouple jcouple ESiC II EDC01

t/min

0

5

10

15

20

25

30

IjI /

µA c

m-2

Fig. 7b. Galvanic current and the contact potential Ecouple of the SiC II–DC01coupling (in 3.5 wt% NaCl). Additionally, the individual OCP of DC01 and SiC IIbefore and after the experiment is presented. (The absolute value of the current isnormalized on the area of the SiC sample.)

194 M. Schneider et al. / Corrosion Science 80 (2014) 191–196

Fe2þ ! Fe3þ þ e� ð5Þ

Due to the fact that the solubility of Fe3+ is only 10�8 mol l�1 atpH 7, iron (III) hydroxide is precipitated on the surface and influ-ences the dissolution kinetic at higher over potentials (Eq. (6)).

Fe3þ þ 3H2O! FeðOHÞ3 þ 3Hþ ð6Þ

With the knowledge of the electrochemical behavior of the sin-gle components the experiments of galvanic corrosion began. Toinvestigate the tendency of calvanic corrosion the authors carriedout measurements of the galvanic current and the contact poten-tial following the DIN-norm [29]. Fig. 3 represents the experimen-tal setup according to the described procedure in [29].

The OCP of both materials were measured before the galvaniccoupling was closed by a switcher and after switch off. Duringthe contact measurement the galvanic current was recorded andthe electrode potential was measured. The results are shown inFigs. 6 and 7. Although the potential difference between SiC I andnickel is large enough, the galvanic current is very small (Fig. 6a).The passive layer on both materials inhibits the galvanic corrosion.In contrast to this result the galvanic current between SiC I andDC01 is approximately two orders of magnitude higher. Unalloyedsteels such as DC01 do not form a passive layer under these condi-tions and therefore galvanic corrosion takes place. The contact po-tential is Ecouple � �500 mV (Fig. 6b). In this case the SiC I ceramicforms the cathode on which the oxygen is reduced. The galvaniccurrent density on the cathode (A = 1 cm2) is about 1.5 lA cm�2.This complies with the current density on SiC I at E = �500 mVshown in Fig. 5. In other words, the galvanic corrosion is deter-mined by the cathodic process (oxygen reduction). The kinetic ofthe cathodic reaction is again charge transfer controlled and de-pends on the specific electrical conductivity of the ceramic ascathode.

The situation shown in Fig. 7a is comparable to Fig. 6a. The gal-vanic corrosion between SiC II and Ni can be ignored althoughthe higher specific electrical conductivity of SiC II would enhancethe charge -transfer through the cathode. However, in contrast tothe results in Fig. 6a the galvanic corrosion is not inhibited by passivefilms or electrode resistances. Moreover, the potential differencesare too small to be sufficient for the thermodynamic driving force.

The highest galvanic current in this set of experiments is mea-sured between SiC II and DC01 (Fig. 7b). The potential difference isbig enough to take effect as a driving force. Although the potentialdifference between SiC II and DC01 is smaller than between SiC Iand DC01, the galvanic current is one order of magnitude higher

than the galvanic current shown in Fig. 6b. Also in this couplingthe ceramic acts as a cathode. However, the specific electrical

0,00

0,01

0,5

1,0

1,5

2,0

2,5

3,0 140mg/cm²y

14mg/cm²y

0.3mg/cm²y0.2mg/cm²y

SiC I - DC01

SiC II - Ni

SiC II - DC01

q / m

C c

m-2

SiC I- Ni

Fig. 8. The consumed electrical charge during the galvanic corrosion experiments(shown in Figs. 6 and 7) and calculation of the mass loss per year.

-600 -500 -400 -300 -200 -100 0-5

-4

-3

-2

-1

0

1

2

3

4

5

mixed potentialSiC I

j / µ

A c

m-2

ESCE / mV

DC01

Fig. 9. Details of the LSV on DC 01 and SiC I in 3.5 wt% NaCl solution. The graphsbase on the diagram shown in Fig. 5b.

Fig. 10a. Equivalent circuit of the measuring configuration of galvanic corrosionshown in Fig. 3.

10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103

0,5

0,4

0,3

0,2

0,1

0,0

ρSiC= 0.01...250 Ω cm

calculated Ecouple vs. SCE

ESiC I= -24 mV vs. SCE

EDC 01 = -512 mV vs.SCE

ρMe= 10-5 Ω cm

pote

ntia

l E

/ m

V

specific resistance ρ / Ω cm

Fig. 10b. Simulated course of Ecouple between a metal (high electrical conductivity)and materials of assuming various specific resistances. The simulation bases on themeasured OCP of SiC I and DC01 and the specific resistance of iron. The range of theceramics used in this work is marked by the bar (according to electricalmeasurements shown in Table 1).

M. Schneider et al. / Corrosion Science 80 (2014) 191–196 195

conductivity is clearly higher than that of SiC I and therefore it is nolonger rate determining. Related to the SiC II surface (A = 1 cm2)the current density of the galvanic corrosion is in the range ofthe oxygen diffusion-limited current density and only this still im-pedes higher corrosion rates.

Based on the measured contact current the consumed chargecan be calculated according to:

q ¼Z tend

tstart

idt ð7Þ

whereby q represents the normalized charge, tstart and tend to be thetime of the galvanic coupling experiment and i the current density.The results for the various couples are shown as a bar chart in Fig. 8.The values on the top of the bars are the mass loss of metal per yearcalculated by using Faradays law.

In all presented experiments the potential Ecouple is almost iden-tical to the OCP of the metals. Applying the mixed potential theoryto the galvanic corrosion as described by Oldfield [33] this fact canbe expected due to the LSV in Fig. 5. Fig. 9 exemplarily illustratesthe expected contact potential Ecouple as well as the expected gal-vanic current with respect to the mixed potential theory. The gal-vanic corrosion is controlled by the kinetic of the oxygen reductionon the ceramic according to Eq. (1). The metal represents the anodein the couple. A potential drop DE ¼ jRelectrolyte can be neglected be-cause the resistivity of the electrolyte is only q = 18 X cm.

According to the cell geometry and the current density the poten-tial drop in the electrolyte amounts to DE � 0:5 mV. It is easy to seethat the expected values (j and E) are close to the measured resultsas aforementioned and discussed.

A completely other approach to explain the potential Ecouple ofcoupled materials is considered as the big differences in the resis-tivity of the electrodes. The measuring configuration shown inFig. 3 can be transformed into an electric equivalent circuit pre-sented in Fig. 10a. If the resistances of the electrolyte and theammeter are very small in comparison with the resistance of theelectrodes, then both the first-mentioned resistances (Ramperemeter

and Relectrolyte) can be neglected. Furthermore, if the resistance ofthe ceramic is much higher than the resistance of the metal (RSiC -� RMe) then the couple potential is identical to the potential of themetal (Ecouple = EMe). Fig. 10b shows the simulation of the potentialEcouple dependence on the resistance of the ceramic.

4. Conclusions

The present work has shown that galvanic corrosion can occurin mixed material constructions between metallic materials andsemi conductive ceramics. The phenomena can suitably be investi-gated and tested by the common experimental techniques accord-ing DIN-50919 and well explained by the mixed potential theory.The criteria of the risk of corrosion are the same as for the more

196 M. Schneider et al. / Corrosion Science 80 (2014) 191–196

frequent bimetallic corrosion. In the present case the galvanic cor-rosion is determined by the cathodic oxygen reduction. The poten-tial Ecouple is close to the OCP of the metals which can be explainedby the mixed potential theory as well as by the different resistivityof the electrode material. Therefore the use of two reference elec-trodes as suggested in the DIN-50919 during the galvanic corrosionis not absolutely necessary. The comparison of the coupling sys-tems shows once again that the galvanic corrosion is determinednot only by the potential difference, but, moreover, by the corro-sion kinetic. It can be shown that the resistivity of the ceramicmainly determines the corrosion reaction. The higher the resistiv-ity, the lower the galvanic corrosion rate. With respect to the mosttechnical applications, where the area ratio metal/ceramic is usu-ally high, the damage risk is low (the area rule). Moreover, theceramic is cathodic protected by the metal which can be an advan-tage because ceramic components are often used at very criticalsites of the construction (e.g., sealings).

Acknowledgements

The authors gratefully acknowledge financial support from theAiF/Federal Ministry of Economics and Technology under contractnumber 390 ZBG and the company ESK Ceramics GmbH for sup-porting with the material SiC I.

References

[1] G. Frankel, D. Landolt, in: M. Stratmann, A.J. Bard (Eds.), Fundamental ofCorrosion in Encyclopedia of Electrochemistry, vol. 4, Wiley-VCH, 2003.

[2] H. Kaesche, Corrosion of Metals, Springer, 2003.[3] K.-H. Tostmann, Korrosion, Wiley-VCH, 2001.[4] W. Machu, W. Köhler, Die korrosion des systems aluminium/stahl, Mater.

Corros. 13 (1962) 601–605.[5] T. Günther, Schadensfalle an Apparaten und deren Berücksichtigung für neue

Konstruktionen, Chem. Ing.-Techn. 42 (1970) 774–780.[6] H. Gräfen, Berücksichtigung der Korrosion bei der konstruktiven Gestaltung

von Chemieapparaten, Mater. Corros. 23 (1972) 247–254.[7] F. Mansfeld, Galvanic corrosion of Al alloys, Mater. Corros. 25 (1974) 578–586.[8] K.A. van Oeteren, Schadensfälle durch falsche Werkstoffwahl und falsche

Werkstoffkombinationen, Mater. Corros. 26 (1975) 357–358.[9] D. Kuron, R. Kilian, H. Gräfen, Measuring system for the compilation of galvanic

corrosion tables, Z. Werkstofftech. 11 (1980) 382–386.[10] W. Huppatz, K. Mechsner, Fügeverbindungen im Schiffbau -

Korrosionsverhalten und Korrosionsschutz, Mater. Corros. 49 (1998) 421–427.[11] G. Tober, Th. Meier, Korrosionsinspektion im Flugzeugbau, Mater. Corros. 46

(1995) 405–409.[12] R. Feser, G. Klose, Mischbauweise im Automobilbau Oberflächenvorbe-

handlung und Korrosionsschutz, Mater. Corros. 50 (1999) 470–474.[13] L. Reclaru, R. Lerf, P.-Y. Eschler, A. Blatter, J.-M. Meyer, Pitting, crevice and

galvanic corrosion of REXstainless-steel/CoCr orthopedic implant material,Biomaterials 23 (2002) 3479–3485.

[14] G. Song, B. Johannesson, S. Hapugoda, D. StJohn, Galvanic corrosion ofmagnesium alloy AZ91D in contact with an aluminium alloy, steel and zinc,Corros. Sci. 46 (2004) 955–977.

[15] E. Tada, K. Sugawara, H. Kaneko, Distribution of pH during galvanic corrosionof a Zn/steel couple, Electrochim. Acta 49 (2004) 1019–1026.

[16] B. Hu, L.M. Richardson, Mechanism and possible solution for transversesolidification cracking in laser welding of high strength aluminium alloys,Mater. Sci. Eng. A 429 (2006) 287–294.

[17] A.B.M.M. Rahman, S. Kumar, A.R. Gerson, Galvanic corrosion of laserweldments of AA6061 aluminium alloy, Corros. Sci. 49 (2007) 4339–4351.

[18] I. Bakos, S. Szabo, Corrosion behaviour of aluminium in copper containingenvironment, Corros. Sci. 50 (2008) 200–205.

[19] M. Schneider, U. Langklotz, A. Michaelis, B. Arnold, Microelectrochemicalinvestigation on aluminium–steel friction welds, Surf. Interface Anal. 42(2010) 281–286.

[20] P. Campestrini, E.P.M. van Westing, H.W. van Rooijen, J.H.W. de Wit, Relationbetween microstructural aspects of AA2024 and its corrosion behaviourinvestigated using AFM scanning potential technique, Corros. Sci. 42 (2000)1853–1861.

[21] J.H.W. de Wit, New knowledge on localized corrosion obtained from localmeasuring techniques, Electrochim. Acta 46 (2001) 3641–3650.

[22] W.-T. Tsai, J.-R. Chen, Galvanic corrosion between the constituent phases induplex stainless steel, Corros. Sci. 49 (2007) 3659–3668.

[23] D. zur Megede, E. Heitz, Das Korrosionsverhalten von Hartmetall-Verbundwerkstoffen in chloridhaltigen wäßrigen Lösungen, Mater. Corros.37 (1986) 207–214.

[24] V.A.D. Souza, A. Neville, Linking electrochemical corrosion behaviour andcorrosion mechanisms of thermal spray cermet coatings (WC–CrNi and WC/CrC–CoCr), Mater. Sci. Eng. A 352 (2003) 202–211.

[25] A. Collazo, X.R. Novoa, C. Perez, Corrosion behaviour of cermet coatings inartificial seawater, Electrochim. Acta 44 (1999) 4289–4296.

[26] C.A. Nunez-Lopez, P. Skeldon, G.E. Thompson, P. Lyon, H. Karimzadeh, T.E.Wilks, The corrosion behaviour of Mg alloy ZC71/SiC, metal matrix composite,Corros. Sci. 37 (1995) 689–708.

[27] G. Kiourtsidis, S.M. Skolianos, Corrosion behavior of squeeze-cast siliconcarbide-2024 composites in aerated 3.5 wt.% sodium chloride, Mater. Sci. Eng.A 248 (1998) 165–172.

[28] F. Podbreznik, Die Kontaktkorrosion zwischen Metallen und nichtmetallischenMaterialien in der Atmosphare und in einigen Elektrolyten, Mater. Corros. 10(1959) 495–498.

[29] DIN 50 919, Korrosionsuntersuchungen der Kontaktkorrosion inElektrolytlösungen.

[30] U. Sydow, M. Schneider, M. Herrmann, H.-J. Kleebe, A. Michaelis,Electrochemical corrosion of silicon carbide ceramics Part 1: Electrochemicalinvestigation of sintered silicon carbide (SSiC), Mater. Corros. 61 (2010) 657–664.

[31] M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions,Pergamon Press, 1966.

[32] Tabellenbuch Chemie (team of authors) Dt. Verlag für Grundstoffindustrie,ninth ed., Leipzig, 1982.

[33] J.W. Oldfield, Electrochemical theory of Galvanic Corrosion, in: H.P. Hack (Ed.),Galvanic Corrosion ASTM STP 978, American Society for Testing and Materials,Philadelphia, 1988, pp. 5–22.

[34] F. Mansfeld, D.H. Hengstenberg, J.V. Kenkel, Galvanic corrosion of Al alloys. I.Effect of dissimilar metal, Corrosion 30 (1974) 343–353.

[35] F. Mansfeld, E.P. Parry, Galvanic corrosion of bare and coated Al-alloys coupledto stainless-steel 304 or Ti–6Al–4V, Corros. Sci. 13 (1973) 605–621.

[36] F. Mansfeld, J.V. Kenkel, Galvanic corrosion of Al-alloys 2. Effect of solutioncomposition, Corros. Sci. 15 (1975) 183–198.

[37] F. Mansfeld, J.V. Kenkel, Galvanic corrosion of Al-alloys 3. Effect of area ratio,Corros. Sci. 15 (1975) 239–250.

[38] Product Information 03.35, SiC30-Siliziumcarbid-Graphit-Verbundwerkstoff,Schunk Kohlenstofftechnik GmbH Heuchelheim, 2009.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具