Higgins R.A. Engineering Metallurgy: Applied Physical Metallurgy

Подождите немного. Документ загружается.

Metallic Corrosion and

its Prevention

21.10 Of all metallurgical problems which confront the engineer, few can

be economically more important than the prevention of metallic corrosion.

In Great Britain alone it is estimated that the cost of metallic corrosion is

of the order of 3% of the annual gross national product which is equivalent

to £10 billion or £150 per capita. This is perhaps less surprising when one

considers, as an example, the team of painters permanently employed in

protecting the steel of the Forth Railway Bridge from the ravages of air,

rain and sea water. Nevertheless even more significant is the amount of

steel which is allowed to rust away for lack of adequate protection. This

amounts to some 1000 tonnes every single day—surely an expensive way

of 'saving jobs'. But enough of this cynicism.

Other metals, in addition to iron and steel, corrode when exposed to

the atmosphere. The green corrosion-product which covers a copper

roof,

or the white, powdery film formed on some unprotected aluminium alloys

is clear evidence of this.

Metals do not corrode where there is no atmosphere. The expensive

Hasselblad camera left behind on the Moon by the American astronauts

will remain in perfect condition as far as its metal parts are concerned,

though radiation may damage some of its non-metallic components. More-

over little corrosion of most engineering alloys, including carbon steels,

occurs in a dry atmosphere at ambient temperatures; but when the atmos-

phere is also moisture laden it is a very different matter and 'rusting' of

steel is rapid. Fossil fuels, and in particular coal, contain varying quantities

of sulphur and nitrogen compounds. On combustion these compounds

yield sulphur dioxide and some oxides of nitrogen which, if not extracted

from the flue gases, will escape to the atmosphere where they will combine

with moisture to form the corrosive agents sulphurous, nitrous and nitric

acids.

Over much of Europe, both West and East, this accelerates the

corrosion of metals, destroys ancient and modern buildings and kills large

tracts of our forests. To extract the acidic oxides from flue gases is an

expensive process which, in turn, will create some of its own environmental

problems but it must be done. The only alternative seems to be the use of

nuclear-derived power in our current state of technology.

The Mechanism of Corrosion

21.20 Oxidation or 'Dry' Corrosion The metals of Group I and II of

the Periodic Classification react readily with oxygen so that, apart from

beryllium and magnesium which are useful because of their very low spe-

cific gravities combined with good strength, they are of little use as

materials in constructional engineering. Most of the engineering metals are

to be found in the 'transition' groups where affinity for atmospheric oxygen

is rather less. Oxidation of many of these metals is extremely slow at

ambient temperatures but occurs much more rapidly as the temperature

rises as is demonstrated by the scaling of steel at red heat. When iron is

heated in an atmosphere containing available oxygen it becomes coated

with a layer of black oxide scale, FeO:

2Fe 4- O

-> 2FeO

The above equation is only a simplified way of expressing the reaction. In

fact atoms of iron have been oxidised whilst atoms of oxygen have been

reduced. These processes are related to a transfer of electrons from atoms

of iron to atoms of oxygen:

—>

" + 2 electrons (e~) (Oxidation)

O + 2 electrons (e") -> O~~ (Reduction)

The terms 'oxidation' and 'reduction' have a wider meaning in chemistry

and involve reactions in which oxygen takes no part. Thus, sulphur gases

will attack nickel alloys at high temperatures giving rise to intercrystalline

corrosion by the formation of nickel sulphide films. In this instance nickel

has been 'oxidised' even though the element oxygen is not involved:

Ni -> Ni

+ 2e" (Oxidation)

S + 2e~ -> S~~ (Reduction)

Consequently, in its wider sense, oxidation can be described as a process

where the atom involved loses electrons; whilst reduction can be described

as a process where the atom gains electrons.

21.21 Affinity for oxygen is not the sole criterion affecting the rate at

which a metal oxidises. Although aluminium has a very high affinity for

oxygen it is nevertheless corrosion resistant. This is due largely to the

dense and impervious nature of the oxide film which forms on the surface

and so protects it from further attack. The extent to which an oxide film

will protect the metal beneath depends upon two factors:

(i) the continuity of the film and how effectively it bonds to the metallic

surface. Some films are porous and offer poor protection whilst others

crack or peel away with repeated heating and cooling of the metal. Such

metals and alloys oxidise progressively.

(ii) the mobility of metal and non-metal ions within the oxide film.

Iron at a high temperature will continue to oxidise even though coated

with an oxide skin. Iron atoms at the metal interface ionise (Fig. 21.1),

releasing electrons which travel quickly to the surface of the scale. There

the electrons react with oxygen from the air forming ions, O . These

oxide ions are then attracted towards the oppositely-charged Fe

ions.

Although the equilibrium composition of the oxide skin is Fe

O~~ that

near the metal surface will contain an excess of Fe

ions and that

adjacent to the atmosphere will contain more O ions. The rate at

which new ions form depends upon their mobilities within the oxide

film. That is, new Fe

ions will form as existing ones diffuse away from

the metal surface towards O ions. The more rapidly Fe

ions move

away, the more rapidly will iron atoms ionise to produce new Fe

ions.

Prolonged heating of large steel ingots for hot rolling produces a very

thick layer of scale. The concentration gradients of Fe

and O ions

within this layer are often so great that three separate crystalline zones

are found in the layer. The thin zone adjacent to the metal surface is of

basic composition, FeO, whilst that in contact with the atmosphere is of

composition Fe

. Between the two is a zone containing the intermedi-

ate oxide Fe

—or FeO-Fe

21.22 When two steel surfaces are in contact under fairly high pressure

and at the same time subjected to alternating or vibrational stresses, fretting

corrosion may occur. Differences in elastic properties between the surfaces

may lead to localised welding of contacting high spots. The welds sub-

sequently rupture and local high temperatures, set up as a result of friction,

cause oxidation of the surface. Fretting is indicated by the presence of a

reddish-brown powdery deposit of oxide, Fe

, sometimes referred to as

cocoa powder. This is highly abrasive and so accelerates the attack. Fret-

ting is prevalent in wire ropes and occurs in splines and other press-fitted

components which are subjected to alternating stresses. In addition to

general wear, such corrosion can lead to the initiation of fatigue cracks

iron

air

Rg.

21.1

oxide scale

particularly since the affected components are subjected to vibrational

stresses. Though difficult to eliminate completely it can be minimised by

excluding air by the use of a high-pressure grease or by the use of a solid

high-pressure lubricant such as molybdenum disulphide.

21.23 As indicated above oxidation in industrial atmospheres is rarely

a simple process involving oxygen only. Contaminants such as carbon diox-

ide,

carbon monoxide, sulphur dioxide, oxides of nitrogen and water vap-

our are frequently responsible for the very rapid deterioration of metals

at high temperatures. Under varying conditions carbon dioxide can pro-

mote both oxidation and carburisation in alloys. This phenomenon leads

to green rot in Nimonic Alloys—precipitation of &23C6 leaves the matrix

so depleted in protective chromium that extensive oxidation of the nickel

to green NiO occurs rapidly.

Nimonic and other nickel-base superalloys are rapidly attacked in the

presence of sulphur-rich combustion gases. This leads to the formation of

black NiS so that the effect is aptly described as black plague. Solid ash

from burning fuel, carried in the combustion gas stream, can also react

chemically with surface oxidation products. As a result a fluxing action

may take place resulting in the formation of fluid products at temperatures

above 650

C. Fluid glassy slags so formed dissolve oxides rapidly so that a

metallic surface is exposed to further oxidation. The very rapid corrosion

of Fe-Ni-Cr alloys which occurs under such conditions is known as

catastrophic oxidation.

21.30 Electrolytic Action or Wet Corrosion Involving Two Dissimi-

lar Elements Electrolytic action in one form or another is responsible

for the bulk of corrosion which occurs in metals at ambient temperatures.

In this particular instance it will occur when two dissimilar metals of differ-

ent 'electrode potential' are in electrical contact with each other and with

an 'electrolyte'. The term 'electrolyte' describes some substance which

contains both positively- and negatively-charged ions, able to move about

freely within it.

Much of this chemical action is similar to that which occurs in a simple

Galvanic cell (Fig. 21.2), consisting of a copper plate and a zinc plate,

immersed in dilute sulphuric acid (the electrolyte). When the external

circuit is closed a current begins to flow through the ammeter. This current

is composed of electrons which are released in the zinc plate and, as their

concentration builds up there, are forced to flow to the copper plate. As

a result of the loss of electrons zinc atoms become zinc ions (Zn

) and

pass into solution in the electrolyte:

Zn -> Zn

+ 2e"

As the electrons, which have been forced round the external circuit by

pressure of numbers, collect on the copper plate it becomes negatively

charged so that hydrogen ions (H

), present in the electrolyte from ionised

sulphuric acid, are attracted to the copper plate where they combine with

the available electrons to form ordinary atoms and, hence, molecules of

hydrogen so that bubbles of the gas form on the copper plate:

Fig.

21.2 The

chemical reactions

in the

simple

cell.

The conventional direction ascribed

to the

current

in the

early days

electrical technology

is opposite

that

which electrons

fact flow.

+ 2e~

—>

(bubbles

copper plate)

ions 'pair

up'

with

ions

in the

electrolyte

that

as the

concentration

sulphuric acid falls that

zinc sulphate rises. Thus

obtain 'electrical energy'

at the

expense

of a

loss

chemical 'potential

energy'

by the

zinc.

This electrolytic action occurs

as the

result

of a

difference

'electrode

potential' between copper

and

zinc (Table 21.1).

The

electrode potential

metal

to the

amount

energy required

remove

the

valence electrons from

its

atoms. Thus zinc loses

its

valence electrons more

readily than does copper

and

zinc

said

to be

anodic towards copper.

The

electrode which supplies electrons

to the

external circuit

called

the

anode

whilst

the

electrode which receives electrons

via the

external circuit

called

the

cathode.

21.31

The

reader

may

have been puzzled

by the

fact that

Fig. 21.2

shows electrons

flowing

in the

opposite direction

that indicated

for

current

most elementary text books

electrical technology.

The

Greek

derivation

of the

term anode suggests something which

being built

and

fact

the

anode

was so

named

in 1800

because

it was

thought

that

time that positively-charged 'particles

electricity' were passing through

ELECTRONS

CONVENTIONAL

'CURRENT'

ammeter

ZINC

(ANODE)

ELECTROLYTE

(dilute

sulphuric

acid,

)

COPPER

(CATHODE)

(zinc

sulphate)

bubbles

hydrogen

Table 21.1 The Electrochemical Series (Electrode Potentials for Some Metals)

Metal (ion) Electrode Potential (volts) Eh

(Noble) Gold (Au

+ + +

) +1.50 (Cathodic)

Platinum (Pt

+ + + +

) +0.86

Silver (Ag

) +0.80

Copper (Cu

) +0.52

(Cu

) +0.34

Hydrogen (H

) 0.00 (Reference)

Iron (Fe

+ + +

) -0.05

Lead (Pb

) -0.13

Tin (Sn

) -0.14

Nickel (Nr

) -0.25

Cadmium (Cd

+ +

) -0.40

Iron (Fe

+ +

) -0.44

Chromium (Cr

+ + +

) -0.74

Zinc (Zn

+ +

) -0.76

Aluminium (Al

+++

) -1.66

Magnesium (Mg

+ +

) -2.37

(Base) Lithium (Li

) -3.04 (Anodic)

the external circuit from the cathode to the anode. We now know that the

electric current in the external circuit is in fact a flow of

negatively-charged

particles (electrons) in the opposite direction. The original convention of

current direction is, however, still retained.

21.32 The bulk of metallic corrosion is due to electrolytic action of this

type.

Electrolytic action is possible when two different metals or alloys are

in electrical contact with each other and are also in common contact with

an electrolyte. Action such as this occurs when a damaged 'tin can' is left

out in the rain. If some of the tin coating has been scratched away so that

the mild steel beneath is exposed, electrolytic action takes place between

the tin and the mild steel when the surface becomes wet with rain water

or condensation (Fig. 21.3A). Water containing oxygen or other dissolved

gases is ionised to the extent that it will act as an electrolyte, and electrons

flow from the iron (mild steel) to the tin, leading to the release of Fe

ions into solution. Thus, once the coating is broken, the presence of tin

accelerates the rusting of iron it was meant to protect, and the electrolytic

action follows the same general pattern as that which prevails in the simple

cell. The iron of the can corresponds to the zinc plate, the tin coating to

the copper plate and the solution of oxygen in water to the dilute sulphuric

acid used as the electrolyte in the simple cell. Industrial atmospheres accel-

erate corrosion because of the sulphur dioxide they contain. This gas is

present due to the combustion of sulphur in coal and other fuels. Sulphur

dioxide dissolves in atmospheric moisture forming sulphurous acid which

has a much stronger electrolytic action than has dissolved oxygen. How-

ever, in the pages which follow electrolytic corrosion related only to the

anion, -OH", is discussed; but it should be appreciated that in industrial

atmospheres containing the anions -SO

, —NO2~ and -NO

", electro-

lytic corrosion will be accelerated by the presence of these strongly electro-

negative ions.

21.33 Metallic coatings are frequently used to protect steel from cor-

rosion, but it is necessary to consider in each case how corrosion will be

Fig.

21.3 The

mechanisms

electrolytic

corrosion.

affected

if the

coating becomes scratched

broken. Thus metals used

protect steel

can be

divided into

two

groups:

(a) Metals which

are

cathodic towards iron, such

copper, nickel

tin. These metals

can be

used only

if a

good-quality coating

assured,

since protection offered

purely mechanical

that

the

coating isolates

the surface

the steel from

the

corrosive medium. Suppose that

some

point

in a

tin-coated iron sheet

the tin

film

broken

(Fig

21.3A). Here

the

tin

film acts

as a

cathode

and the

steel

the

anode. Fe

ions

into

solution

and

electrons,

released, travel

to the tin

cathode where they

react with water

and

dissolved oxygen:

ELECTROLYTE e.g. MOISTURE

OH IONS Fe

IONS

TIN COATING (CATHODE)

PITTING

STEEL (ANODE)

ELECTRON

FLOW

ELECTROLYTE

ZrI

+ IONS

OH-

IONS

ZINC COATING (ANODE)

STEEL (CATHODE)

ELECTRON

FLOW

Thus hydroxide ions (OH ) are released at the cathode and, away from

the region of the cell, these combine with Fe

ions:

+ + 2OH- -> Fe(OH)

Iron(II) hydroxide, Fe(OH)

, so formed quickly oxidises to iron(III)

hydroxide, Fe(OH)

, the basis of the reddish-brown substance we call

rust. Thus corrosion of the steel is actually accelerated when a coating

of some substance which is cathodic to steel becomes damaged. Fortu-

nately, in the case of both tin and nickel, metals commonly used to coat

iron and steel, the difference in electrode potential in the cell so formed

is small, so that the acceleration of attack is not great. The further apart

metals are in the electro-chemical series (Table 21.1) the greater the rate

of corrosion of the anode.

(b) Metals which are anodic towards iron, such as zinc and aluminium.

These metals will go into solution in preference to steel (Fig

21.3B)

which they will therefore offer some chemical, as well as mechanical,

protection. Since they protect the steel by going into solution themselves,

such protection is generally referred to as 'sacrificial'. Zinc is used in

this way for galvanising steel products, whilst aluminium is effective as

a protective paint partly for the same reason. It must be emphasised

however that the object of galvanising should be to produce a sound

continuous coating of zinc on the surface because sacrificial protection

will only be temporary should the coating be porous.

21.34 Having dealt briefly with the principles of electrolytic corrosion,

we are now in a position to give an adequate explanation of the rusting of

mild steel (Fig 21.4). Here electrolytic action takes place between the iron

and the oxide film which will be present on the surface—assuming of

course that the oxide film is broken and that the gap is covered by an

electrolyte (moisture containing dissolved oxygen).

Fig.

21.4 The electrolytic corrosion of iron due to the presence of a surface film of

oxide.

IRON

(ANODE)

ELECTRON

FLOW

MOISTURE

IONS

OH"

IONS

OXIDE

FILM

(CATHODE)

Since iron is anodic to its oxide, it will go into solution as ions:

Fe -> Fe

+ 2e"

The electrons which are released travel to the cathode and take part in the

reaction:

O + '/2O

+ 2e" -> 2OH-

(from

atmosphere)

As a result, Fe

ions go into solution at the anode and hydroxide ions

(OH~) form at the cathode. When these two different sorts of ions meet

away from the region of

electrolytic

action the following reaction occurs:

+ 2OH- -> Fe(OH)

This iron(II) hydroxide, Fe(OH)

, is quickly oxidised by atmospheric oxy-

gen to iron(III) hydroxide, Fe(OH)

, which is precipitated as a reddish-

brown substance, the main constituent of rust.

21.35 So far we have been dealing only with electrolytic action between

some surface coating and the metal supporting it. The same type of cor-

rosion may take place at the surface of a metal in which particles of an

impurity are present as part of the microstructure. In Fig 21.3c it is assumed

that the particle of impurity is cathodic towards the metallic matrix, thus

causing the latter to dissolve. Since impurities are often segregated at

crystal boundaries, this will lead to gradual intercrystalline corrosion of

the material as more particles of impurity become exposed by electrolytic

action.

21.36 Those metallic alloys which have a crystal structure consisting of

two different phases existing side by side are also prone to electrolytic

corrosion. If one phase is anodic with respect to the other it will tend to

dissolve when the surface is wetted by a suitable electrolyte. Patches of

pearlite in steel will tend to corrode in this manner (Fig 21.5). Ferrite is

Fig.

21.5 The electrolytic corrosion of pearlite.

MOISTURE

ELECTRON

FLOW

CEMENTITE

(CATHODE)

FERRITE

(ANODE)

anodic to cementite, consequently ferrite goes into solution (as Fe

ions)

and the electrolytic action is similar to that which takes place between

ferrite and an oxide film during the rusting of mild steel mentioned above.

This leaves the brittle cementite platelets standing in

relief,

but these will

ultimately break off as corrosion proceeds and increases their fragility. It

is therefore apparent that in order to have a high corrosion-resistance, a

metal or alloy should have a structure consisting of one type of phase only,

so that no electrolytic action can take place. Similarly, as engineers will

know, it is unwise to rivet or weld together two alloys of widely different

electrode potentials if they are likely to come into contact with a substance

which can act as an electrolyte. In such circumstances the alloy with the

lower electrode potential may corrode heavily due to electrolytic action.

An example of bad practice of this type is illustrated in Fig. 21.6. Here a

steel pipe has been connected to a copper pipe. If the system carries water

which is not chemically pure, then electrolytic corrosion of the steel, which

is anodic to copper, can be expected. This corrosion will be much more

rapid than if the whole pipe were of steel.

Those readers who are householders will probably be aware that the

galvanised cold-water tank corrodes more quickly in that region adjacent

to the copper inlet-valve mechanism, zinc being strongly anodic to copper.

This form of plumbing is doubtless very good for trade, particularly as

failure of the tank usually occurs when the room beneath has been newly

decorated. The obvious remedy is to match the copper plumbing with a

copper tank—which is less expensive than it sounds. Fortunately most

modern domestic plumbing systems now used 'plastics' (polypropylene)

cold-water tanks which are of course impervious to electrolytic corrosion

and are also relatively inexpensive.

21.37 The extent to which electrolytic action between two dissimilar

electrodes takes place depends not only upon the difference in potential

between them but also upon the chemical nature of the electrolyte, and in

particular the concentration of hydrogen ions (pH). For example, our

motor cars rot away even more rapidly as a result of the salt used on roads

during winter. When electrovalent compounds dissolve in water molecules

Fig. 21.6 The

accelerated corrosion

of a

steel

pipe

due to the

presence

of an

adjacent

copper

pipe.

COPPER

(CATHODE)

STEEL

(ANODE)

FLOW

WATER