Transactions of the IMF

The International Journal of Surface Engineering and Coatings

ISSN: 0020-2967 (Print) 1745-9192 (Online) Journal homepage: http://www.tandfonline.com/loi/ytim20

Advances in Replacements for Cadmium Plating in

Aerospace Applications

K. R. Baldwin & C. J. E. Smith

To cite this article: K. R. Baldwin & C. J. E. Smith (1996) Advances in Replacements for
Cadmium Plating in Aerospace Applications, Transactions of the IMF, 74:6, 202-209, DOI:
10.1080/00202967.1996.11871127

To link to this article: https://doi.org/10.1080/00202967.1996.11871127

Published online: 08 May 2017.

Submit your article to this journal

Article views: 2

View related articles

Citing articles: 33 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ytim20
 Advances in
Replacen1ents for
Cadn1ium Plating in
Aerospace Applications

K. R. Baldwin and SUMMARY- This paper examines some of the problems associated with the replacement of cad-
mium plating for the protection of aircraft components and fasteners manufactured from steel. It
C. I. E. Smith reviews the range of coatings which are available commercially and are currently being considered as
alternatives to cadmium plating. Results of research being undertaken into the corrosion behaviour
Structural Materials Centre, Defence and galvanic compatibility of electrodeposited zinc alloy coatings and aluminium-magnesium alloy
Research Agency, Farnborough, coatings are presented. Other aspects of coating performance are also considered including frictional
Hampshire GU14 6TD properties, effects of coatings on fatigue strength, hydrogen embrittlement of steels and the repair of
damaged coatings. It is concluded that although the corrosion performance of cadmium plating may
be achieved with several coating systems, no one coating is likely to be acceptable as a replacement for
cadmium for all aerospace applications.

INTRODUCTION mium-plating is its high lubricity. This

Cadmium plating is widely used on air- means that the effort expended in over-
craft for the corrosion protection of air- coming friction between the bolt and nut
frame components and fasteners surface is minimised. Cadmium's low
manufactured from steel. The advantages coefficient of friction ensures that only
of cadmium plating are well known and moderate torques need to be applied to
are outlined below. The main disadvan- generate the required bolt loading even in
tage associated with the use of cadmium is the absence of lubricants such as cetyl
the high toxicity of the metal and its com- alcohol.
pounds. Because of the poisonous nature The main disadvantages of cadmium
of cadmium salts, the disposal of effluent are also listed in Table 1. The toxicologi-
from plating operations is subject to strin- cal problems associated with cadmium
Paper given at INTERFINISH '96, gent controls and the levels of cadmium plating have already been outlined above.
Birmingham, UK, 10--12 September 1996 that may be discharged into the environ- A further drawback with the electrolytic
ment are extremely low. Under European process is that cyanide baths are
legislation, the use of cadmium plating for employed. Despite their many advan-
general engineering purposes is no longer tages, the use of cyanides is increasingly
permitted. The need for an acceptable viewed as undesirable from a heath and
alternative to cadmiu_J»-plating is widely safety viewpoint'. Apart from toxicity
recognised and severe restrictions on the
use of cadmium exist in many other coun-
tries including the USA and Japan. Some
properties of cadmium-plating are sum-
marised in Table 1.
Cadmium-plating functions as a very 0.0
effective barrier coating, particularly in Nickel
~ -0.2
the marine environments frequently expe- u
rienced by aircraft. Cadmium-plating cor-
rodes at a low rate in chloride bearing
.. -0.4
tJl

> Cobalt
~
media and provides long-term protection.
The application of chromate passivation -;;; -0.6 Mild s tee 1
treatments further extends the life of ....c: Al alloy 2014-16
Al alloy 7075-16
plated components. ....0 -0.8
QJ

Cadmium-plating
Alongside good barrier properties, cad- C>

mium-plating is also a sacrificial coating.

Figure 1 compares the open-circuits
-...."
:::>
-1.0 Zinc-plating

potentials of a range of metals and alloys. -"

t.

The potential of cadmium is more nega-

tive than that of steel so that if the coating 0
.. -1.8
I
c:
C>
Magnesium
is damaged, the cadmium-plating will cor-
rode instead of the steel. Figure 1 further -2.0
indicates that the potentials of aluminium
alloys are similar to that of cadmium so
that the risk of damaging galvanic interac- Figure 1. Open-circuit potential of various
K. R. Baldwin and tions occurring between them is small. metals and alloys in quiescent 600mmoll- 1 NaCI
C. J. E. Smith, Trans IMF, 1996, 74(6), 202 Another attractive property of cad- solution at 25°C.

202
Table 1. Some advantages and disadvantages of cadmium plating cannot be electrolytically deposited from
aqueous solutions. Since the various depo-
Advantages Disadvantages sition methods do not involve exposure to
aqueous electrolytes, the risk of hydrogen
Good barrier against corrosion Toxicity
Electrodeposition from cyanide baths
up-take is avoided (other than at the
Sacrificial coating
Galvanically compatible with cleaning stage). Since the 1970s, alu-
aluminium alloys Embrittlement of steel at elevated temperatures minium coatings have been commercially
Good surface lubricity Embrittlement of titanium available from baths based on non-aque-
In-situ repair by brush-plating Hydrogen embrittlement associated with plating process ous solvents 13· 14 . An advantage of elec-
trodeposited aluminium is that the
coatings are very dense and exhibit good
adhesion to the substrate. Aluminium
Table 2. Some commercially available coatings for steel may also be sprayed but this process has
Method of application
found limited application in the aerospace
Coating type
industry, partly due to difficulties in
Zinc-plating (unalloyed) Electrodeposition achieving fine thickness control.
Mechanical plating The more usual method of aluminium
Zinc-nickel plating Electrodeposition coating is by physical vapour deposition
Zinc-cobalt plating Electrodeposition (PVD) e.g ion-vapour deposition (IVD).
Zinc-tin/tin-zinc plating Electrodeposition The IVD process has been commercially
Aluminium plating Physical vapour deposition available for many years. For example,
Magnetron-sputtering the McDonnell Douglas IV ADIZE
Electrodeposition from organic electrolytes
Electrodeposition from fused salt baths
process was developed to coat landing
Metal spraying gear parts, engine mountings and fasten-
Metallic-ceramic coatings Dip-spin procedure ers with aluminium 15 . Coatings produced
Dip-drain procedure by PVD methods show good substrate
adhesion but tend to be porous and glass-
bead peening is required to compact the
structure. Aluminium coatings, regardless
of their mode of deposition, are usually
issues, cadmium plating does have several may experience periods at elevated tem- chromated to provide additional corrosion
other drawbacks. One of these is that dur- peratures. protection and to promote paint adhesion.
ing the electroplating of cadmium, dan- The last decade has seen a large growth. Aluminium coatings may also be
gerous levels of hydrogen may be particularly in the automotive industry, of formed by another vacuum technique,
introduced into the steel component, lead- the use of zinc alloys containing additions unbalanced magnetron sputtering (UMS),
ing to a drastic reduction in strength. De- of nickel or cobalt. These alloys appear to which has been developed for the plating
embrittlement heat treatments must be offer a marked improvement in corrosion of aerospace parts and other applications,
employed with steels with high tensile resistance over unalloyed zinc and arc at the University of Salford 16•17 • The steel
strengths in order to minimise the risk of deposited from cyanide-free electrolytes parts are cleaned, placed in a vacuum
hydrogen embrittlement failure in service. based on environmentally benign chlo- chamber and then sputter-cleaned under
In the case of ultra-high strength steels, rides and/or sulphates. Zinc alloys such as an argon atmosphere before coating with
cadmium is often vacuum deposited to zinc-nickel (Zn-Ni) may also be deposited aluminium. Recent work 1H has also shown
avoid the risk of hydrogen uptake. from cyanide-free alkaline baths. The use that aluminium-magnesium (Al-Mg)
A range of commercially available coat- of alkaline baths avoids the use of ammo- alloys with improved corrosion resistance
ings currently being assessed by the aero- nium salts, which are often present in can readily be deposited by the UMS
space industry are given in Table 2. acidic baths, and further simplifies effluent technique. An advantage of the UMS
The coatings are either metallic or con- management. A range of post -plating process is that the aluminium coatings are
sist of metallic particles bonded with an chromating solutions has been developed very dense and do not require shot-peen-
inorganic matrix, often based on phos- specifically for zinc alloy coatings, which ing.
phates and/or chromates. An important further enhance corrosion resistance. A different way of utilising sacrificial
requirement is that the coating should Previous research at ORA metals such as aluminium or zinc is to
provide sacrificial corrosion protection if Farnborough 2·3 on Zn-Ni alloys has estab- incorporate them in powder or flake form
the substrate becomes exposed. Figure 1 lished that the optimum level of protec- in an inorganic slurry, often based upon
shows that the coatings listed above, tion in neutral salt fog tests is obtained for phosphates and/or chromates. The slurry
which are either based on pure zinc or alloys containing approximately 14 wt-% may be applied by dip-spin or dip-drain
pure aluminium, will be more electroneg- nickel. Commercial electroplating solu- procedures. The slurry is then stoved at
ative than steel under most conditions, tions marketed in the UK generally pro- elevated temperatures to produce coat-
and hence should provide sacrificial pro- vide this level of nickel in the deposits. In ings consisting of metal particles in an
tection. Figure 1 further shows that the case of zinc-cobalt (Zn-Co) alloys, the inorganic matrix. Coatings that fall into
nickel and cobalt would be unsatisfactory optimum cobalt concentration is thought this category are SermeTel and
since, although they would provide good to be in the range 4-8 wt-%4, however, for Dachromet. SermeTel type coatings have
barrier protection, the steel substrate commercial reasons, the level of cobalt found use on landing gear components in
would corrode as soon as the coating was generally used in practice is lower, at the place of cadmium and are often fin-
breached. between 0.6-1.2 wt-%. A number of stan- ished by applying a layer of organic resin.
A drawback with zinc plating is that its dards now cover the electrodeposition of This part of the paper has reviewed
corrosion products are voluminous and both Zn-Ni'" and Zn-Co' alloys and which some of the corrosion-resistant coatings
can cause the seizure and wedging of fas- define compositional limits. Zinc alloy currently available. In the following sec-
teners. Figure 1 shows that zinc is also plating technology is continuing to tions, the results of a ORA Farnborough
considerably more electronegative than advance. Zinc alloy electrodeposits research programme into cadmium-substi-
steel. Thus, although zinc may initially formed with tinK, manganese"· 10 and tution are presented. The corrosion
provide good protection to the substrate if chromium 11 · 12 are either commercially behaviour of the coatings, galvanic com-
the coating is damaged, rapid galvanic available or under development. and hold patibility with aluminium alloy, coating
corrosion of the zinc means that, in prac- out the prospect of a wider range of zinc- lubricity, effect of fatigue and hydrogen
tice, the protection is often short lived. base coatings being available to the embrittlement problems etc, have been
Like cadmium, zinc may be electrode- designer in the future. considered. Where information is unavail-
posited (ED) and is recommended for Aluminium coatings may be produced able from the current programme, litera-
applications on aerospace parts which by a variety of methods, although they ture data has been reviewed.

203
COATINGS EVALUATED density found for a given coating will be as-plated UMS pure AI coatings has been
The following range of coatings have been related to its corrosion rate in a given attributed to their dense nature as com-
evaluated at DRA Farnborough; environment, subject to various theoreti- pared with the more porous PVD
cal constraints' 2- 24 • Given that the coatings coatings 18 . Table 3 shows that, for the
a) Three electrodeposited zinc-base coat-
examined in the present work showed UMS AI coatings, the addition of magne-
ings:
similar cathodic reactions (oxygen reduc- sium had a detrimental effect on barrier
Zinc-to Def. Stan 03-19' 0
tion), then in general terms, the lower the properties, although the corrosion current
Zinc-14 wt-% nickel*
value of the corrosion current density, the values obtained for the alloy coating
Zinc-0.8 wt-% cobalt*
better the coating's barrier properties will remained comparatively low.
*prepared using commercially avail-
be.
able solutions.
The data obtained after 1 hr show that Sacrificial properties
b) Three aluminium-base coatings pre-
the alloy zinc coatings had higher Icorr val- The sacrificial properties of metal coatings
pared by Physical Vapour Deposition:
ues than found for cadmium, suggesting can be investigated by measurement of
PVD aluminium-deposited by a wire
that they will provide less efficient barrier the open-circuit corrosion potential
fed evaporative source and then
protection, although the data also show (OCP). In the present work, OCP values
peened.
that they afforded significant improve- of coatings have been measured in quies-
UMS aluminium and UMS alu-
ments over zinc alone. After 504 hrs, the cent NaCI solution after 1 hr and 504 hrs,
minium-20 wt-% magnesium alloy-
situation was reversed for Zn-Ni and Zn- and the data obtained are given in Table
both deposited using an unbalanced
Co alloys compared with cadmium, and 4.
magnetron sputtering source at the
the corrosion current values for the alloys Research at DRA Farnborough 18-26 has
University of Salford.
fell appreciably. Previous work at DRA shown that the sacrificial behaviour of
(c) Electrodeposited cadmium- to Def.
Farnborough 22 ·25 has established that the coatings can be predicted with reasonable
Stan 03-20 19 .
decrease in corrosion current seen for Zn- accuracy by comparing their potentials
The above coatings have been Ni alloys, was mainly due to dealloying, or with that of the substrate material e.g
deposited onto standard mild steel tests more specifically, a form of dezincifica- steel. The OCP of mild steel has previ-
panels of 1mm thickness. The metallic- tion. The dezincification process leads to ously been measured in 600 mmol 1- 1
ceramic type coatings were not considered surface enrichment of beneficial nickel (or NaCI solutiott- and a value of -0.67 V
in this study and data on their perfor- cobalt) and hence gives rise to improved (SCEt-was obtained 26 • Table 4 shows that
mance and that of electrodeposited alu- barrier properties 2225 • The chromating of all the coatings evaluated were negative to
minium have been published elsewhere 2·21 • Zn-Ni and Zn-Co alloys stabilised the cor- this value and hence would be expected to
The coatings listed above were tested in rosion currents. however, the values afford a degree of sacrificial protection to
the as-plated and chromated conditions. obtained after 504 hrs were slightly higher steel. This has been confirmed at ORA
with the exception of the sputter-deposits, than found for the as-plated coatings that Farnborough 26 for pure Zn and Zn-Ni
which were tested in the as-plated condi- had undergone dezincification. alloys through the use of a novel scratch-
tion only. In each case, the coating thick- In the case of the aluminium coatings, model technique. The scratch-model
nesses were measured prior to testing. the corrosion current data were generally approach 26 has established that, although
using an eddy current probe. comparable with cadmium, which initially the more negative pure zinc coatings
suggested that they may have similar bar- afford much higher levels of sacrificial
CORROSION STUDIES rier properties. The particularly low cor- protection to steel than the more noble
The ability of metal coatings to protect rosion current values obtained for the Zn-Ni alloy, pure zinc suffers appreciably
steel is mainly dependent on two proper-
ties of the coating, firstly its barrier prop-
erties, and secondly, its ability to Table 3. Corrosion current density values for selected metal coatings determined in 600mmoii-•
sacrificially protect steel once the coating NaCI at 25°C.
has been breached by the corrosive envi-
ronment. The steel substrate will only cor- Coating Surface condition ''"" (~A.cm·')
rode and form the familiar iron-oxide
I hr 504 hrs
(red-rust) corrosion products once both of ED cadmium as-plated 0~2 0.13
these two properties have been lost. In the chroma ted
first part of the study, the barrier and sac- ED zinc (unalloyed) as-plated 10.6 10.2
rificial properties of the coatings have chromated
been considered in isolation, through the Zn-14 wt-% Ni as-plated 2.10 0.10
use of electrochemical techniques on thick chroma ted 0.31 0.30
(>20Jlm) coatings. The overall corrosion Zn-0.8 wt-% Co as-plated 3.17 0.11
chroma ted 0.41 0.32
resistance of each coating type has then 2.05
PVDAl peened only 1.13
been considered using neutral salt fog peened and chroma ted 0.10 0.10
tests and marine exposure trials on coated UMSAl as-plated 0.01
steel panels. UMS Al-20 wt-% Mg as-plated 0.12

Barrier properties
Previous work at DRA Farnborough 22 has Table 4. Open-circuit potential values for selected metal coatings determined in 600mmo11· 1 NaCI
established that the barrier properties of at 25°C.
metal coatings can be conveniently stud-
ied with cathodic potentiodynamic polari- OCP (V vs SCE)
sation techniques. In the present work. Coating Surface condition 1 hr 504 hr
cathodic sweeps have been performed on
the range of metal coatings in aerated ED cadmium as-plated -0.80 -0.75
chroma ted -0.78 -0.75
NaCI solutions, at a sweep rate of -1.03
ED zinc (unalloyed) as-plated -1.03
0.125mV s- 1• Corrosion current density, chroma ted
Icorr• values were determined from the Zn-14wt-% Ni as-plated -0.95 -0.75
cathodic sweeps using a Tafel extrapola- chroma ted -0.90 -0.88
tion technique 22 •23 • Each cathodic sweep Zn-0.8 wt-% Co as-plated -1.02 -0.89
was carried out on a fresh specimen in chroma ted -0.94 -0.93
each case and data obtained for each coat- PVDAl peened only -0.73 -0.75
ing type after two immersion times; 1 hr peened and chroma ted -0.77 -0.73
UMSAl as-plated -0.90 -0.72
and 504 hrs. are shown in Table 3. UMS Al-20 wt-% Mg as-plated -0.92 -0.93
In simple terms, the corrosion current

204
 higher rates of galvanic corrosion. Table 4 environment. barrier properties tend to localised attack would have led to the pre-
also indicates that the potential of pure assume dominance over sacrificial proper- mature loss of barrier properties and
zinc was stable at approximately -1.03 V ties. This hypothesis explains the good exposure of the steel substrate at an ear-
(SCE) whereas. in contrast. the as-plated performance of cadmium-plating in the lier stage than that predicted solely on the
Zn-Ni and Zn-Co alloys ennobled during salt fog environment since the electro- basis of corrosion current measurements.
immersion. due, it is thought. to the chemical data indicated that it had good The as-plated UMS Al-20 wt-% Mg
effects of dezincification22. 25 . The chro- barrier properties whilst providing a alloy coatings were found to delay the
mate treatments appeared to suppress the degree of sacrificial protection once the onset of red-rust formation for twice as
dezincification of both alloy types. coating was breached. long as cadmium-plating in the salt fog
Table 4 shows that the aluminium coat- In contrast to cadmium-plating, the per- environment (Figure 3). A more detailed
ings. regardless of their mode of deposi- formance of pure zinc was very poor in analysis of the corrosion behaviour of AI-
tion. were consistently negative with the neutral salt fog tests. This can be Mg coatings has been given elscwhcre 1".
respect to mild steel. Whilst the potential attributed to the poor barrier properties Briefly. the high levels of protection found
of PVD AI was relatively stable in the of zinc (Table 3). Previous work at ORA with Al-Mg coatings may be attributed to
chloride solution, the potential of UMS AI Farnborough 2 K~Y and elsewhere·" 1 -- ~ has 1
their good barrier properties coupled with
was erratic, possibly due to the effects of shown that, despite its excellent sacrificial improved sacrificial protection. The UMS
micro-pitting 1". As expected, the addition properties, zinc coatings arc rapidly con- Al-Mg alloy deposits were also found to
of the more active magnesium component sumed in salt fog. This is due. in part. to be much less susceptible to pit formation
to aluminium, had the desired effect of the presence of high concentrations of dis- than the pure AI coatings. The PVD AI
lowering the potential and hence improv- solved oxygen and chloride ions and the coatings were less corrosion resistant than
ing the sacrificial properties. rapid generation of an essentially non- the UMS coatings. presumably due to the
protective zinc oxide corrosion product more porous nature of the former.
layer (white-rust) on the zinc metal sur-
Corrosion resistance of coated steel
face. Figs. 2 and 3 show that the addition Marine exposure trials
Neutral salt fog tests of Ni or Co greatly improved the perfor- The range of coatings listed in section 2
Neutral~alt fog tests have been conducted mance of zinc, due mainly. it is suggested. have also been exposed to a natural
on coated steel panels in accordance with to an improvement in barrier properties. marine atmosphere at ORA Fraser.
ASTM B 11727 . For each coating type, a The nature of the corrosion product layer Portsmouth. UK. and the test panels have
number of panels were prepared to give a on Zn-Ni and Zn-Co alloys was also com- been inspected at monthly intervals. As
range of coating thicknesses. In the pre- pletely different from that on zinc alone with the neutral salt fog tests. the corro-
sent work, the time taken for red-rust to and was tenacious and non-voluminous in sion resistance of the coatings was deter-
appear on the panel surfaces, T RR• was nature. mined by estimating TRR.
used as a means of assessing corrosion The aluminium coatings were found to The TRR values obtained for a single
performance. The results obtained are afford high levels of protection in the neu- coating thickness are shown by Figure 4.
shown in Figure 2 and data for a single tral salt fog environment. particularly The pure zinc coatings were found to
coating thickness are presented in Figure those which had been sputter-deposited. afford very high levels of protection in the
3. As indicated above, T RR will be a func- On the basis of the data presented in marine environment. which is in accor-
tion -.{)f both the barrier and sacrificial Table 3, the UMS pure AI platings might dance with published data 2 "-~Y.JJ.:4_ In the
properties of the coatings. be expected to out-perform cadmium. marine atmosphere. it is thought that the
The data show that cadmium-plating since the corrosion current of the former sacrificial properties of metal coatings
provided long-term corrosion protection was over ten times lower than that of cad- assume dominance over barrier proper-
to the steel substrate in the highly aspi- mium. A possible explanation for the ties. provided that the corrosion rate of
rated salt fog environment. Previous stud- apparent discrepancy is that the alu- the coating remains relatively low 2 "-~Y. This
ies~"-~Y on sacrificial coatings have minium coatings were found to be suscep- is opposite to the situation found in the
established that, in the neutral salt fog tible to pitting attack. This form of neutral salt fog environment and accounts
for the reversal in behaviour seen for pure
zinc (Figs. 3 and 4 ). The high level of per-
formance observed for the pure zinc coat-
ings can. therefore. be attributed to its
highly sacrificial nature when compared
ED Pure zinc ED Zn-Ni alloy with, for example. cadmium. i\lthough the
corrosion rate of zinc is approximately
-30
twice that of cadmium in the marine envi-
"
.;;.
ronment. even thin zinc coatings will be
~ 20~ present for many years2s.~Y.
The ED Zn-Co alloy also provided high
~ 10 ~ levels of corrosion protection in the
marine environment. Table 4 indicates
01
~a~_.--~2~0~Da~-*--~4~DD~O~ 2000 4000 2000 4000 that the presence of the small amount of
Time lhaursl Time lhoursl Time (hours) cobalt (0.8 wt-%) would not have caused
a significant loss of sacrificial properties.
Although sacrificial properties are
ED Pure cadmium PVD Aluminium UMS Aluminium thought to be the dominant factor in the
- 30 marine environment. clearly any improve-
Pliltea G Peened
"
:- Al ment in the integrity of the barrier layer
:po
~~e1n1a
1;1"
- G
would also enhance performance. It is
known that the presence of nickel can sta-

~"~ Chro111tea illlOY

bilise the corrosion product layer on zinc
alloy coatings10..15 • Thus. in the present
work. the presence of cobalt may have
0
~0--~--~20~0~0--._~4~00~0~ a 2oao 4aoo 2000 4000 stimulated the formation of the more pro-
Time (hours) Time lhoursl Time (hours)
tective surface film on the alloy surface.
thereby accounting for the modest
improvement in performance seen over
Figure 2. Relationship between coating thickness and time to first red-rust for variOII.I' metal mating,· pure zinc.
on steel, exposed to continuous neutral salt fog. Closed circles: as-plated coatings. open circles: passi- In contrast to the Zn-Co alloy. the cor-
vclled coatings. rosion performance of the Zn-Ni alloy in

205
 Time to rea-rust ~ours) Time to red-rust (days)
200 400 600 800 1000 1200 1400 200 400 600 800 1000 1200 1400

Unp~ Unp~

PVD Al Chroma ted Chroma ted

PVO Al

UMS Al UMS Al
UMS Al-20~Mg········

Figure 3. Comparison of the performances of coated steel sheet exposed to Figure 4. Comparison of the performance of coated steel sheet exposed to a
continuous neutral salt fog. Coating thickness: 8 microns. marine atmosphere. Coating thickness: 8 microns.

the marine atmosphere was poor despite mental arrangement described else- obtained for the coatings coupled to 2014-
the high levels of protection found in the where36. After a 168 hr coupling period, T6 and 7075-T6 alloys are shown by
neutral salt fog test. The poor perfor- the aluminium alloy electrodes were Figure 5. This indicates that, except for
mance of the Zn-Ni alloy has been attrib- removed from the solution and chemically the--Y:n-Ni alloy, the zinc-base and alu·
uted2x·29 to the relatively high cleaned in a hot chromic/phosphoric acid minium-base coatings corroded preferen-
concentrations of nickel in the alloy, bath. Following chemical cleaning, the tially when coupled with the two
which significantly reduce the ability of weight change due to corrosion was calcu- aerospace aluminium alloys. For the Zn·
the coatings to sacrificially protect steel. lated. Ni alloy, the current was found to
Thus at any pores in the coating, the steel The galvanic current density traces decrease during coupling with 2014-T6
will only be protected for a limited time
before the formation of red-rust occurs.
Figure 4 indicated that the pure alu-
minium coatings did not match the level
of corrosion protection afforded by cad-
mium-plating in the marine environment, ED Zn-Co a 11 oy ED Zn-Nl alloy
regardless of the method of coating appli-
cation, which is broadly in-line with the
results obtained from the neutral salt fog
tests (Figs. 3). However, encouraging lev-
els of performance were obtained in the
marine environment for the sputtered Al-
20 wt% Mg alloy coatings. The high level ~ "' 0
of protection obtained for the alloy coat-
100 200 100 200 100
ing is entirely consistent with the above Time (hours) Time (hours) Time (hours)
hypothesis in that the main benefit of
adding magnesium was an improvement
in sacrificial properties.
15 ~~ED~P~u~r~e_c~a~d~mri~u=m_,~
Galvanic compatibility with aluminium
alloys N"!O
e
Figure 1 showed that there was a signifi-
cant potential difference between alu- ~ 51-. 1 .J

minium alloys and steel. Thus, if a bare .a- :;:.................... - -.... .J..···· .....~~
steel fastener was inserted into an alu- ~ 0 ---------------------
minium component, there would be a 100 200
large driving force for galvanic corrosion. Time (hours)
In contrast, the potential of cadmium is
very close to that of aerospace aluminium
alloys so that the use of cadmium-plating !Sr---~P~V~D~A~l~u~m~i~n~iu~m~,_~ UMS Aluminium UMS Al-20wtl Mg
on steel fasteners and other parts serves to
minimise galvanic interactions with alu-
N"!O
minium components. e
u
In the present work, the compatibility '3-5
between the range of metal coatings (as-
plated condition only) and two aerospace
aluminium alloys has been investigated
using an electrochemical approach. The tOO 200 0 100 200 100 200
aluminium alloys selected for this study Time (hours) Time (hours) Time (hours)
were an aluminium-copper alloy, 2014-T6,
and an aluminium-zinc-magnesium alloy,
7075-T6. A zero-resistance ammeter Figure 5. Variation in galvanic current density with time for couples formed between various as-
(ZRA) was used to measure the current plated metal coatings and aluminium alloys, during immersion in quiescent600mMII NaC/. Solid
flow between the couples, using an experi- line-vs 2014- Tn, dotted line-vs 7075- T6.

206
 alloy and eventually, a reversal in the Table 6 shows that Zn-Co alloy and Al- bolts did not change the torque-tension
direction of current flow occurred. This Mg alloy coatings were less compatible characteristics of cadmium-plating, after
effect has been attributed to the dezincifi- than cadmium whereas Zn-Ni alloy and approximately 15 nut runs, the perfor-
cation of the zinc alloy, which has the UMS pure AI were more compatible than mance of the PVD aluminium coatings
effect of reducing the driving force for gal- cadmium. The Zn-Ni alloy was the only declined sharply and high torques were
vanic corrQ.<;ion by ennobling the alloy coating to actually lower the rate of alu- needed to achieve even small bolt loads42 .
surface 36 . minium alloy corrosion, despite the cur- Table 7 indicates that zinc does not pos-
Although the above approach has been rent reversal effect. The PVD AI coatings sess the same self-lubricating properties as
useful in determining if a coating will showed similar compatibility behaviour to cadmium. Only a limited amount of fric-
interact galvanically with an aluminium cadmium-plating. tional data is available for the zinc alloy
alloy, it does not indicate the extent to coatings. Hsu43, reported torque-tension
which coupling may affect the corrosion Frictional properties of coatings data for Zn-Ni plated fasteners, which
behaviour of the aluminium alloy itself. Table 1 showed that cadmium-plating has indicated that the torque required for fas-
One way of determining the effect of gal- several other critical properties apart from teners plated with a Zn-Ni alloy was about
vanic coupling on aluminium alloy corro- corrosion resistance. It is important that the same as that needed for cadmium-
sion is to measure its corrosion rate. In the alternative coatings have acceptable fric- plated fasteners when installed in interfer-
present work, the corrosion rates of 2014- tional properties, particularly if they are ence-fit holes in aluminium panels. These
T6 and 7075-T6 alloys were established to find use on bolts and other threaded findings were supported by those of
from the weight-losses of the aluminium items. This is because a large amount of Altmayer33 who measured the breaking-
alloys determined after coupling and the torque used in tightening bolts is torques of coated parts after 12 month's
cleaning, using a method described else- expended in overcoming frictional con- exposure to a marine atmosphere and
where36. These corrosion rate data repre- straints. The frictional properties of metal found that the values measured for a Zn-
sent the total corrosion rate, RA- The RA coatings are often assessed through a Ni alloy were similar to those obtained for
of a coupled metal will be equal to the comparison of coefficients of friction and cadmium. However, other assessments of
sum of its self (uncoupled) corrosion rate, their torque-tension behaviour. The coef- the lubricity of Zn-Ni coatings by Ko et
R0 , and its galvanic corrosion rate, R 0 . ficients of friction determined for a range al. 44 and of Zn-Co alloys by Hadley45 sug-
The self corrosion rates, R 0 , of the two of coatings are shown in Table 7. gested that both of these alloys have infe-
aluminium alloys have previously been In the absence of a lubricant, the coeffi- rior lubricating properties when compared
determined at DRA Farnborough 36 and cients of friction were found to be depen- with cadmium.
the values obtained are reproduced in dent on coating type. In the degreased
Table 5. condition, none of the coatings Effect of coatings on fatigue
For each galvanic couple investigated, approached the low frictional coefficient There are only a limited amount of data
the value of R 0 was subtracted from each obtained for cadmium, although in the available on the effects of metal coatings
value of RA, to show the effect of galvanic case of the PVD aluminium coatings, the on the fatigue properties of substrate
coupling on the corrosion of 2014-T6 and glass-bead peening and chromate filming materials such as steels, although a DRA
7075-T6 alloys, and the values of (RA-R 0 ) treatments both improved the situation. funded research programme has recently
obtained in each case are shown in Table Although the data for lubricated coatings started to address this issue. Several stud-
6. A vera.ge galvanic current density (1 0 ) was limited, Table 7 shows that the coeffi- ies have examined the effect of PVD alu-
values calculated for the 168 hr period are cients of friction can be drastically minium coatings on the fatigue properties
also shown for each couple. reduced by, for example, cetyl alcohol, of steels. For example, Nevill 46 has
A positive value of (RA-R 0 ) indicates and the values obtained were largely inde- reviewed the effect of PVD aluminium
that galvanic coupling has accelerated cor- pendent of coating type. coatings on the fatigue properties of vari-
rosion of the aluminium alloy, whereas a In previous DRA work 42 where torque- ous substrate materials including steels
negative value shows that coupling has tension measurements were conducted, it and titanium alloys. In the case of steels, it
had a beneficial effect by reducing the cor- was found that degreased PVD aluminium was found that PVD aluminium coatings
rosion rate of the aluminium alloy. An required the application of higher torques had either no effect or a slight effect on
(RA-R 0 ) value of zero shows that galvanic to achieve the same loadings as degreased fatigue properties, depending on the
coupling has had no effect on aluminium cadmium-plating, which is consistent with experimental conditions and materials
alloy corrosion. the higher coefficient of friction of alu- employed. In the same study46 , tension-
Table 6 shows that cadmium-plating minium (Table 7). Furthermore, whereas tension fatigue tests were used to show
caused a moderate increase in the corro- repeated installation of non-lubricated that cadmium-plating caused a slight
sion rate of the two aerospace aluminium
alloys which is in-line with publistred
data3 7•3H. Thus despite the preferential cor-
rosion behaviour of the cadmium-plating Table 6. Average galvanic current densities and (Ra-Ro) parameters
toward the aluminium alloys, coupling
Coating vs 2014-T6 vs 7075-T6
caused an increase in the corrosion rate of 10 (jlAcm 2) RA (mdd) 10 (!lAcm- 2 )
RA-Ro (mdd) RA (mdd) RA-Ro (mdd)
both of the members of the couple. This Cadmium 4.5 13.5 7.6 3.0 12.0 7.8
effect has been attributed to the genera- Zinc 13.6 41.1 35.2 11.8 24.1 19.9
tion of alkaline conditions on the surface Zn-Coalloy 7.2 28.3 22.1 5.7 10.2 14.0
of the aluminium alloy during coupling in Zn-Ni alloy 3.2 3.8 -2.1 3.5 3.2 -1.0
near-neutral saline solutions36·37 • PVDAl 8.5 14.6 8.7 4.3 12.4 8.2
Pure zinc, the most electrochemically UMSAl 3.5 8.2 2.3 2.7 5.1 0.9
active of the coatings, was found to be the UMSAI-Mg 4.7 26.7 20.8 5.1 14.0 9.8
least galvanically compatible, causing
severe attack on the aluminium alloys.
Table 7. Coefficients of friction for selected coatings

Table 5. Self (uncoupled) corrosion rates of Coating Surface condition Reference Coefficient of friction Lubricant
two aerospace aluminiumaUoys in 600mMII De greased Lubricated
NaCI solution Cadmium chroma ted 39 0.27 0.10 Cetyl alcohol
PVDAl as-plated 39 0.79
Aluminium alloy Self corrosion rate, R 0 , peened 39 0.62
(mg.dm- 2.day- 1) peened and chroma ted 39 0.56 0.10 Cetyl alcohol
UMSAl as-plated 40 0.70
2014-T6 5.9 UMS Al-Mg as-plated 40 0.57
7075-T6 4.2 Zinc bulk material 41 0.60 0.16 Paraffin mix

207
reduction in the number of cycles to fail- agents were omitted from Zn-Ni elec- ers and constructors and also by operators
ure but that their overall effect was similar trolytes, or if the solution was contami- and maintenance engineers. There is also
to that of PVD aluminium. McLoughlin 42 nated with certain types of brighteners, or a large, gradually accumulated, bank of
reported similar trends from tension-ten- if baking after plating was omitted. knowledge, experience and data on the
sion fatigue tests, although the effects A survey of the literature would suggest behaviour of cadmium-plating under
were generally more pronounced in his that there is little or no data on the effects operational and service environments as
study than the effects reported by Nevill 46 • of other zinc alloy coatings on the hydro- well as in artificial laboratory conditions.
gen embrittlement of steels. However, There is clearly less experience and data
Hydrogen embrittlement by electroplated there is some data available for tin-base available for many of the coatings being
coatings coatings. Recently, Vasantha et a/. 51 found considered as possible cadmium-substi-
Aqueous electroplating processes are, in that tin and tin-zinc alloy deposited from tutes.
general, less than 100% efficient and dur- gluconate baths onto steel substrates did This paper has outlined recent progress
ing the deposition process, some hydrogen cause some hydrogen uptake but the sus- made in identifying possible substitutes
is invariably evolved. As indicated above, ceptibility to hydrogen embrittlement was for cadmium-plating in aerospace applica-
evolved hydrogen can enter the steel and lower than found with either zinc-plating tions and to highlight their properties as
cause embrittlement. Hydrogen embrittle- or cadmium-plating. broadly as possible. Despite the lack of
men! during cadmium-plating is an impor- information in some areas, considerable
tant consideration in the aerospace Repair of coatings progress has been made over the past few
industry and post-plating heat treatments The last criteria for cadmium replace- years. Many studies, including this one,
are routinely employed to de-embrittle ments considered in this paper is the have compared coatings in terms of their
steel parts. Since the 1960s, Iow-hydrogen- repair of metal coatings. Brush (or selec- resistance to neutral salt fog. Whilst this
embrittlement (LHE) cadmium-plating tive) plating is the usual method information is useful, this study has indi-
solutions have been commercially avail- employed to repair damage to cadmium- cated that salt fog tests alone do not pro-
able to minimise hydrogen uptake during plating. The technique has been devel- vide a complete picture and should, where
plating. For ultra-high strength steels, oped commercially and processes are possible, be complemented with natural
some manufacturers prefer to use vacuum available for plating a wide range of met- exposure trials, at least in the absence of a
deposited coatings. In the case of alu- als including, copper, zinc, cadmium and more representative laboratory test. A
minium coatings, aqueous solutions are nickel. In simple terms, the brush-plating resetH-chJ' programme at DRA
not employed during the plating processes process establishes a plating cell between Farnborough is currently directed at
and the risk of hydrogen embrittlement is an anode, usually of graphite, wrapped in developing a more representative acceler-
minimal. an absorbent material saturated with ated corrosion test for metal coatings.
There have been several detailed inves- electroplating solution, and the work- The tiJse of electrochemical techniques
tigations into the effect of Zn-Ni plating piece, which is the cathode in the circuit. in interpreting the corrosion behaviour of
on the embrittlement characteristics of The passage of direct-current through the coated steel in aqueous chloride environ-
steels, and some of the more recently pub- cell deposits the dissolved metal, and ments was explored in this study.
lished work has been reviewed in Table 8. since deposition occurs only where the Information obtained from the use of
There appears to be a general consen- anode and cathode are in contact, the open-circuit potential measurements and
sus on the effect of Zn-Ni alloy electro- process is often known as selective plat- polarisation techniques enabled the corro-
plating on the hydrogen embrittlement ing. The solutions used do not cause sion of coated steel to be consistently
characteristics of high strength steels. Carr hydrogen embrittlement problems and resolved by identifying the balance
and Robinson 49 examined the effect of de-embrittlement treatments are not nec- between the barrier and sacrificial proper-
plating Zn-Ni alloys from acidic baths on essary so that parts may be repaired in- ties of the coatings. It is clear that the
the hydrogen embrittlement of an AISI situ. Brush-plating is also a suitable electrochemical approach provides an
4340 high strength steel. Baking was car- method for the repair of zinc-base coat- added dimension to the interpretation and
ried out after plating at 200°C for 24 hrs. ings and electrolytes for the deposition of prediction of the corrosion behaviour of
It was found that the Zn-Ni alloy caused Zn-Ni alloys have been available in the metal coatings.
less severe hydrogen embrittlement than UK for many years. The commercial elec- The past five years have seen major
cadmium-plating but, in both cases, com- trolytes tend to yield zinc alloys contain- advances in plating technology as the fin-
plete recovery occurred as a result of bak- ing less nickel ( -10 wt-% Ni) than those ishing industry has responded to the
ing the plated specimens. Importantly, deposited from baths. Brush-plated Zn- demands of increasingly stringent envi-
Carr and Robinson 49 also found that the Ni alloys can be readily chromated. ronmental regulations. The movement of
amount of hydrogen uptake that occurred Aluminium coatings are not usually zinc alloy technology into main-stream
during Zn-Ni plating was dependent on directly repairable using in-situ tech- surface finishing has been one obvious
the electrolyte pH and that the more niques. Instead, the rectification of dam- advance as has the continued develop-
acidic baths employed actually introduced aged areas may involve the use of ment of aluminium coating equipment
less hydrogen into the substrate material. chemical treatments such as chromate and technology. The use of sputter-depo-
The above findings were supported by an filming solutions. sition technology to produce aluminium-
earlier investigation by Hsu 43 who exam- magnesium alloy coatings with controlled
ined the effect of a Zn-Ni coating on the DISCUSSION AND CONCLUSIONS compositions and thicknesses is one par-
hydrogen embrittlement of high strength Over many decades, the use of cadmium- ticularly promising approach.
steels and, more recently, by plating has become firmly entrenched in This study has shown that the corrosion
Pushpavanum and Balakrishnan 50 who the aerospace industry and in manufactur- performance of cadmium-plating can be
used a hydrogen permeability approach. ing industries where metal finishing plays matched or exceeded by several different
In particular, Hsu43 found that failures a significant role. Its limitations are well alternative coating types. However, no
were more likely to occur when addition known and understood by aircraft design- one coating would appear to offer the
same broad range of properties as cad-
mium-plating and hence there is unlikely
Table 8 . The effect of electroplating on hydrogen embrittlement of steels to be, as yet, a single direct substitute.
Areas where further information is
Coating Surface condition Ref. Test Comments required include resistance of coatings to
aircraft fluids, paint adhesion, electrical
Cadmium bright-as plated 47 Constant load Some embrittlement properties, environmental aspects of new
Cadmium bright-as plated 48 Slow-strain rate Significant embrittlement coating systems and cost. These issues and
LHECadmium as plated 48 Slow strain-rate No significant effect others are currently being addre~sed by a
Baked Zn-Ni as-plated 43 Sustained load Zn-Ni plating non-embrittling
LHEZn-Ni as-plated 48 Slow strain rate LHE process non-embrittling
working group on cadmium-substitution
Baked Zn-Ni chromated 48 Slow strain rate Some embrittlement within the Group for Aerospace Research
and Technology in Europe (GARTEUR).

208
REFERENCES Arnell and C. J. E. Smith, Corrosion 35. D. W. Siitari, M. Sagiyama and T.
1. D.R.Gabe, Surface Engineering, 1994, Science, 1996,38 (1), 155-170. Hara, Trans. Iron and Steel lnst.
10 (1), 41-45. 19. Electro-deposition of cadmium. Def. Japan, 1983,23,959-967.
2. K. R. Baldwin and C. J. E. Smith, Stan 03-19, Ministry of Defence, 36. K. R. Baldwin, M. J. Robinson and C.
Product Finishing, 1992,45 {6), 12-18. Directorate of Standardization, J. E. Smith, British Corrosion Journal,
3. K. R. Baldwin, M. J. Robinson and C. Kentigern House, Brown St., 1994, 29 (4 ), 293-298.
J. { Smith, Corrosion Science, 1993, Glasgow, UK (1994). 37. F. Mansfeld, D. H. Hengstenberg and
35 (5-8), 1267-1272. 20. Electro-deposition of zinc, Def. Stan J. V. Kenkel, Corrosion, 1974,30 (10),
4. N. R. Short, A. Abibsi and J. K. 03-19, Ministry of Defence, 343-353.
Dennis, Trans. lnst. Metal Finishing, Directorate of Standardization, 38. M. S. Walker, Materials Performance,
1991,67,73-77. Kentigern House, Brown St.. 1979,18,9-16.
5. Plating, zinc-nickel alloy, SAE Inter- Glasgow, UK (1994). 39. P.L. Lane and C. J. E. Smith,
national, Warrendale, PA, Aerospace 21. K. R. Baldwin and C. J. E. Smith, Development of aluminium-alloy
Material Specification 2417E, 1st Developments in metal coatings for coatings for the protection of steel, in:
January 1993. the protection of steel, in: Proc. Conf. Surface Engineering Practise, eds. K.
6. Standard specification for electrode- Mat-Tee 91, eds. A. Niku Lari and L. N. Strafford, P. K. Datta and J. S.
posited coatings for zinc-nickel alloy Vincent, I.I.T.T International, Paris, Gray, Ellis-Harwood Ltd, Chichester,
deposits, ASTM B841, American France, 1991, 153-158. UK, 1990, 566-577.
Society for Testing and Materials, 22. K. R. Baldwin, M. J. Robinson and C. 40. K. R. Baldwin, R. I. Bates, R. D.
Race St, Philadelphia, USA (1994). J. E. Smith, Corrosion Science, 1994, Arnell and C. J. E. Smith, Aluminium-
7. Standard specification for electrode- 36 (7), 1115-1131. magnesium corrosion resistant coat-
posited coatings for zinc-cobalt alloy 23. F. Mansfeld in: Electrochemical ings, to be published.
deposits, American Society for Techniques for Corrosion. ( ed. R. 41. F. P. Bowden and D. Tabor, The fric-
Testing and Materials, Race St, Baboian), NACE, Houston, Texas, tion and lubrication of solids, Oxford
Philadelphia, USA, ASTM B840 1977, 18-26. Clarendon Press, UK, 1958, p201.
(1994). 24. M. Fontana and N. Greene. Corrosion 42. V. C. R. McLoughlin, An evaluation
8. International Tin Research Institute, Engineering, McGraw-Hill Book Co .. of coatings for steel and titanium alloy
Product Finishing, 1994, 47 (2), 15. New York, USA, 1967. fasteners for aircraft applications,
9. M.Sagiyama, T. Urawaka, T. Adaniya 25. K. R. Baldwin, C. J. E. Smith and M. AGARD CP-256, 47th meeting of
and T. Hara, in: Proc. Conf. Coatings J. Robinson, Corrosion. 1995, 51 (12), AGARD Structures & Materials
and Steel, American Electroplaters 932-940. Panel, Florence, Italy, 26-28th
Soc., Detroit, USA, 24-28 February 26. K. R. Baldwin. M. J. Robinson and C. September 1978, 6.1-6.11.
1986. J. E. Smith, British Corrosion Journal. 43. G. F. Hsu, Plating and Surface
10. M. Eyraud, A. Garnier, F. Mazeron 1994,29 (4), 299-304. Finishing, 1984, 71 (5), 52-55.
and J. Crousier, Plating and Surface 27. Standard Method of Salt Spray 44 C. H. Ko, C. C. Chang, L. C. Chen and
Finishing, 1995,82 (1), 63-70. (Fog) Testing, ASTM B117, T. S. Lee, Plating and Surface
11. M. R. El-Sharif, Y. J. Su, C. U. American Society for Testing and Finishing, 1991,78 (10), 46-50.
Ch~holm and A. Watson, Corrosion Materials, Race St, Philadelphia, 45. J. S. Hadley, Trans. lnst. Metal
Science, 1993,35 (5-8), 1259-1265. USA (1984). Finishing, 1991,62 (2). 10-12.
12. A. Watson, Y. J. Su, M. R. El-Sharif 28. K. R. Baldwin, M. J. Robinson and C. 46. B. T. Nevill, Plating and Surface
and C. U. Chisholm, Trans. lnst. J. E. Smith, in: Proc. Conf. UK Finishing, 1993,80 (1), 14-19.
Metal. Finishing, 1993, 71 (1), 15-20. Corrosion 94, Institute of Corrosion. 47. D. A. Berman, Materials Per-
13. W. Kautek, Corrosion Science, 1988, 31st October-3 November 1994. formance, 1985,24 (11), 36-41.
28 (2), 173-199. Bournemouth, U.K, The Chameleon 48. D. A. Wright, N. Gage and B. A.
14. A. Mayer, J. Electrochem. Soc., 1990, Press, London, Vol.3, 165-175. Wilson, Plating and Surface Finishing,
137 (9), 2806-2809. 29. K. R. Baldwin, C. J. E. Smith and M. 1994,81 (7), 18-21.
15. E. P. Fanin, AGARD CP-256, Ion J. Robinson, Corrosion Reviews, 1994, 49. M. Carr and M. J. Robinson, Trans.
vapor deposited aluminium coatings 12 (1-2), 141-155. lnst. Metal Finishing, 1995, 73 (2), 58-
for improved corrosion protection, 30. E. Almeida, D. Pereira and 0. 64.
47th meeting of AGARD Structures Figueiredo, Progress in Organic 50. M. Pushpavanum and K. Bala-
& Materials Panel, Florence, Italy, 26- Coatings, 1989,17,175-189. krishnan, Trans. lnst. Metal Finishing.
28th September 1978, 5.1-5.8. 31. D. M. Drazic and V. Vascic, 1996, 74 (1 ), 33-35.
16. D. P. Monaghan, D. G. Teer, ."P:- A. Corrosion Science, 1989.29 (10), 1197- 51. V. S. Vasantha, M. Pushpavanum
Logan, K. C. Laing, R. I. &ites and R. 1204. and V. S. Muralidharan, Plating and
D. Arnell, Surface Coatings 32. B. Boelen, B. Schmitz, J. Defourney Surface Finishing, 1995, 82 (6), 81-
Technology, 1993,60,592-602. and F. Blenkkenhorst, Corrosion 83.
17. R.I. Bates and R. D. Arnell, A novel Science, 1993, 34 (11 ), 1923-1931.
coating method for the corrosion pro- 33. F. Altmayer, Plating and Surface ©British Crown Copyright 1996/DRA
tection of aircraft fasteners, to be pub- Finishing, 1993,80 (3), 32-37. Published with the permission of the
lished in Corrosion Management. 34. E. W. Brooman, Plating and Surface Controller of Her Britannic Majesty's
18. K. R. Baldwin, R. I. Bates, R. D. Finishing, 1993, 80 (3), 32-38. Stationary Office.

209