Professional Documents
Culture Documents
A review is given of the various ways in which humidity and liquid water
can influence the friction and wear of metals, polymers, carbons and
graphite, lamellar solid lubricants and ceramics. Compared with dry
sliding, water usually reduces friction of materials to a limited extent,
except for MoS2, but wear rates change depending on the materials
concerned. Most of the fundamental mechanisms involved appear to be
reasonably well understood, but uncertainties still remain concerning the
role of humidity on the formation of third-body layers of aggregated
wear debris.
6!
%rh
simiIar to that frequently observed between the ~miId'
and 'severe' regimes 7. The transition load increased
with increasing relative humidity, suggesting that water
vapour facilitates the formation of a more protective
oxide. Above the transition, however, the wear rates
were greater at high humidities; to account for this it
was suggested that the adsorption of water vapour may
impede the aggregation and transfer of metallic debris. b
This interpretation conflicts with other work ~, however, c
where the prevention of debris aggregation and transfer
of a copper-zinc alloy by water was found to reduce
Ag I
the rate of wear. 2
o/
d 5 I
oL
6 /
/
/ o /
/ IO40 E040
{
~2 Itq
H20
1 t
o 0 50 I00
-- o o"" %rh
b
0 i i ~ [ I i z i i I
0 i0 20
PH20 , mm Hg
be of interest to repeat all these experiments using
6 0 6 3 ~T6 A1203 abrasives for which water-accelerated fracture
/ / is more probable.
/
/ / zn
7 // Lubricated conditions
TRIBOLOGY I N T ER N A TI O N A L 373
J. K. Lancaster--The influence of environmental humidity and w a t e r on friction, lubrication and wear
depend on the type of base stock and are much greater, relatively large fragments (compared with asperity
for example, in synthetic diesters than in hydrocarbons. dimensions) are detached from the surface via fatigue
Some types of additives may also increase the solubility after a large number of repeated stress cycles. Pitting
of water 22. Because of its high polarity, water competes occurs mainly in high stress conditions where the ,~,
directly with other surface-active species in adsorption ratio (elastohydrodynamic film thickness ro surface
processes and, in conjunction with oxygen, also contrib- roughness) is small and there is consequently a~
utes to the interracial chemistry of surface film appreciable amount of solid-solid corttac~ occurring
formation to provide protection against intermetailic through the lubricant film. [n this situation, crack
contact. The whole topic of additive interaction chemis- formation is believed to be initiated a~ the surface, in
try is too broad and complex to summarize in the contrast to that occurring with high )~ ratios where
present context and a review has been given, for sub-surface initiation is more probable 29. [t is now welt
example, by Fein 23. It must suffice here merely to known that the presence of water in the lubricant plays
mention one or two examples of the rote of water. a very important role in pitting failures and this has
Appeldoorn et at 24 have examined the lubricated wear been demonstrated both in simulation testing---rolling
four bal?°, 3~ and in comptete bearings 22, _~2.Examples
of steels in controlled atmospheres where the conditions
from both types of test are shown in Fig 3. As can be
of sliding were insufficiently severe to lead to scuffing.
seen from Fig 3(a) failure is accelerated even by the
They postulate that in these conditions, wear is
very small amounts of water wkich dissolve in a
primarily of the corrosive type in which oxide is
lubricant during exposure ~o humid air. It ha~ been
continuously being removed from the surfaces. As in
suggested 33 that such dissolved water may concert:rate
the case of dry sliding, mentioned earlier, the rate of
in developing cracks bv capillary condensation and
oxidation and the type of oxide can be affected by
the presence of an immiscible water phase is thus not
water, and thus higher wear rates were observed in
an essential requirement. Pitting Iife decreases wit~
wet air and argon than in dry air and argon. These
increasing water content (Fig 3(b)) but tends to approach
trends appeared to be largely independent of the type
a limiting value beyond the point of satv.ration
of lubricant but other work conflicts with this. Fein 23
solubility-;2o The presence of oxygen in addition to
has shown that in hexadecane + 0.43% stearic acid,
water aiso influences pitting life. but the reported
for example, the wear of steel increases with increasing
environmental humidity whereas it decreases in a
neutral solvent. The interesting observation was made
that these humidity effects appeared to be associated 8 i
more with the atmospheric humidity present at the 90- t t o el /
& A
time of cleaning the surfaces rather than with the
humidity of the air used to saturate the oils during
testing. 'Water-conditioning' of the surfaces may there- 70-
I l i
,,t
o
I
®I t
t
"s
5O
i l
water alone are seldom sufficient to provide adequate 3O
protection against scuffing, although they may increase "5
the scuffing load to a limited extent 25. It is therefore 20
common practice to incorporate anti-wear additives in
lubricants to provide additional protection. The two CO
ing sliding and film formation occurs via reactions a Fatigue life, rnil lions of revolutions
between the decomposition products and the metal.
Decomposition studies 26 have shown that the presence
of water increases the rate of breakdown of ZDPs at
any one temperature and also increases the rate of
film formation on steel. With TCP, however, the
situation is more ambiguous. The wear-reducing action
of this additive is commonly attributed to hydrolysis27
leading to an acid phosphate which then reacts with
d
~5
o2
"2_
:f @ @
Pitting of rolling elements Fig 3 Variation of pitting fatigue life with humidly~, :(a)
Rolling 4-ball tests in mineral oil (from Ref 29). (b)
One of the main limitations in the life of rolling Tapered roller bearings in SAE 20 oil containing rust
element bearings is the incidence of pitting in which and oxidation inhibitors (from Ref 22)
374 December 90 VoR 23 No 6
J. K. Lancaster--The influence of environmental humidity and water on friction, lubrication and wear
trends are conflicting. Some experiments -+~ suggest humid air than in vacuum and attribute this to a
oxygen is beneficial and reduces the fatigue-inducing greater adhesion of the resins to the oxidized surface.
effects of water, whilst others indicate the opposite > . A similar increase in adhesion has been reported 4~- for
PTFE sliding on steel in moist air, but in this instance
The causative agent believed to be mainly responsible
the increased adhesion leads to the formation of a
for water-accelerated crack growth during pitting is
transfer film and the friction decreases rather than
atomic hydrogen which penetrates into the steel,
increases.
leading to embrittlement of the highly stressed surface
layers ~4. Direct evidence of reduced pitting life follow- Changes in the friction of polymers m the presence of
mg hydrogen penetration has been obtained ++`' bv liquid water also depend on the type of counterface
electrolytically pretreating balls before testing, and the against which they slide. Polymers fire not 'wetted' by
penetration of hydrogen during pitting has been water so that when sliding against themselves, the
followed using tritium as a radioactive tracer -~. There friction is little different from that observed in dry
arc several ways in which hydrogen can be liberated conditions. However, against hydrophilic surfaces, such
during lubricated sliding and it is still not certain which as glass or clean metals, friction usually decreases. The
is to he considered the most important. Interactions left-hand side of Table 1 shows some results from two
between water, oxygen find freshly exposed metal sources 4~44. Tanaka 44 attributes the reduced friction
surfaces lead to peroxide and hydrogen formation ~'' and in water to a sharing of load support between contact
electrochemical corrosion will also produce hydrogen in through the water film and that between the solids
the atomic state tit cathodic regions within a crack ~. themselves, i.e. to a 'mixed" lubrication process. This
In addition, the dissociation of organic lubricants, is supported by the fact that the extent of the friction
irrespective of their water content, also liberates reduction in water increases with increasing speed.
hydrogen, as shown by tracer experiments with deu- Because polymers, in general, have relatively low
terium-labelled paraffins :~r. Hydrocarbons dissociate elastic moduli, hydrodynamic effects can become
during chemisorption on clean metals but not on oxides, significant at much lower speeds than with metals ~-4~'.
whilst fatty acids can be catalytically decomposed on There are some complications in behaviour, however,
both metals find oxides -~*. It follows that there is likcl\ with those polymers, such as nyhm 6.6, which absorb
to bc a wide variability in the extent to which hydrogen appreciable amounts of water and become plasticized
can be liberated during rolling contact, depending on with a consequent reduction in strength. Cohen and
the type of lubricant, the additives present and the Tabor 47 have shown that when water is added to nylon
water content. In these circumstances, it is hardly 6.6 sliding against itself, there is an immediate reduction
surprising that large differences have beet) found in in friction as a rcsult of absorbed water reducing the
pitting fatigue lives with different oils a''. shear strength of the outermost surface layer. With
increasing time, however, the water penetrates further
One other aspect should also be mentioned. It was
into the polymer and reduces the sub-surface strength
suggested 4" many years ago that lubricants might
also: the real area of contact then increases and the
accelerate pitting failure bv penetrating into a
friction rises again.
deveh'ping crack and inducing a wedging action by
hydrostatic stresses when the crack mouth is tempor- Not a great deal of information is awulablc concerning
arily closed during contact. This 'hydraulic effect' is the influence of atmospheric humidity on the wear of
now no longer believed to be a major factor in pitting unfilled polymers but, as with friction, the effects do
fatigue of rolling elements, but it could well be not seem to be particularly large. Results from one
significant during lubricated sliding wear, as will be
seen l a t e r .
Table 1 Effects of w a t e r on the friction and wear of
unfilled polymers. A: On 0.15 # m Ra stainless steel,
L = 10 N, V = 0.2 m s 1 (Ref 43). B: On 0.05 g m Ra
Polymers
stainless steel, L = 10 N, V = 0.01 m s 1 (Ref 44)
Unlike metals, the friction of polymers is not, in
general, particularly sensitive to the presence of surface Polymer Friction
contamination. Semicrystalline polymers, such as poly- coefficient, W e a r rate,
ethylene, acetal, nylons etc. usually show only small wet/dry wet/dry
differences in friction coefficients in vacuum, air or
inert gases. On the other hand, amorphous thermoplas-
tics, such as polystyrene or polymethylmethacrylate, A B A B
and cross-linked thermosets, such as epoxies or phe-
nolics, arc rather more susceptible to the environment Poly(phenylene sulphide) 0.72 -- 2.9 --
and tend to show somewhat higher friction in vacuum Poly(phenylene oxide) 0.6 -- 1.25 --
than in air. Hydrogen bonding may play a part in Acetal 0.58 0.65 1.1 1.6
increasing the friction of polymers containing OH Nylon 6 -- 0.66 -- 3
groups in humid environments. When sliding against N y l o n 6.6 0.73 -- 0.8 --
metal counterfaces, however, the effects of water
PTFE 0.8 0.9 0.6 2.36
vapour on friction can become more significant L.D. p o l y ( e t h y l e n e ) 0.42 0.18 0.4 0.74
because, as already mentioned, water is able to modify H.D. p o l y ( e t h y l e n e ) I 0.58 -- 1.7
the state of oxidation of the metal. McLaren and UHMWPE 0.48 -- 0.35 --
Tabor at. for example, have found the friction of Polyimide 0.85 1.08 8 7.5
silicone and epoxy resins against steel to be greater in
TRIBOLOGY I N T E R N A T I O N A L 375
J. K. Lancaster--The influence of environmental h u m i d i t y and water on friction, lubrication and wear
[
J
10-SL 10-5
/ \ Materials I- II on
XI5% Cr steel
N
9 \
10-6 __ 10-6
z z \ \ ~ ~ " Air
w l O
2
b
~. 10-7 / cJ
10-7
._____.__ N 2
He
10-8 10-8 _
-- C3H6
.. C3H8,H4HIo
C2H6
I
~0 -9 ____ I L J I J to -9 L J I 1 J ] I
10-2 I0 -I I iO 102 I03 104 10-2 I0 -t I I0 I02 I03 104
a ppm H20 ~ ppm H20
Fig 4 Variations of wear rate with humidity for PTFE composites (cd'ter Schuhert~-'). (a) Materials 1-II on cast iron
in N>. (b) Material 1 on cast iron ill various atmospheres
SLIDING ENVIRONMENT
06
g Increased severity of
sliding conditions
F_
4,
"6 0 4 I Higher contact temperature Reduction in partial pressure
"E I of water vapour, p
I
I
I
Po increases
~ Q2 B l
x x
iO -2 Increased friction
f A
4,
Disruption of surface films
4,
Increased roughness
4,
Fewer contact areas
'E ~0 -B I __7,
4,
Higher contact temperatures Higher contact stresses
E
aJ
• I Brittle fracture of asperities
o
i0 -4
Increased s u r f a c e r o u g h n e s s
X
,,,X-- X--
--x- X
1 STATEDUST,NG I
l~ig 8 Sequence of events involved in dusting (/?ore Ref
71)
i0 5 _ ~___ [ I~ I I I _ _
0 8 12 24
Speed, ms -I
contamination present in carbons, which may volatilize
under the influence of frictional heating, could greatly
Fig 7 Variation qf friction and wear rate with speed for
affect their friction and wear behaviour in dry environ-
an electrographite sliding on carbon-graphite in air
ments. Most manufactured carbons and graphites are
(~50% rh, 20°0; load = 22 N. (A) Clean. (B)
porous and organic contamination is neither easy to
Containing -~7°/i, wt of (NH4)e HP04 (from Re]" 74)
detect nor to remove. One interesting example of the
role of organic materials in preventing high wear in
the absence of water vapour is given by some results
for high-strength carbon fibre-reinforced polymers
appreciably lower than that needed in vacuum (0.1)
sliding in air and vacuum. Table 2 shows that, as would
because the adsorption of oxygen from the air sup-
be expected, the wear rates of polyimide and PTFE
plements that of water vapour 7". The attainment of a
containing carbon fibre are both appreciably greater
critical temperature depends on the interactions
in vacuum than in ordinary (moist) air. With the
between two groups of parameters TM. On the one hand
acetal composite, however, the reverse occurs. In the
are those which influence the rate of heat generation
particular conditions of sliding used, the localized
and dissipation--load, speed, coefficient of friction,
surface temperatures were sufficiently high ( > 180°C)
thermal properties of the materials and ambient
to degrade the acetal, liberating formaldehyde which
temperature--whilst on the other are the factors
could then substitute for the absent water. The greater
influencing the number, size and distribution of the
thermal stabilities of PTFE and polyimide effectively
localized contacts--surface roughness, apparent con-
preclude the formation of sufficient quantities of
tact area, and third-body formation. In general, dusting
occurs most readily for rough surfaces with small organic vapour. Dusting can also be inhibited by
the incorporation into carbons of certain inorganic
apparent contact areas sliding against counterfaces of
compounds which tire widely used as anti-oxidants.
low thermal conductivity. The sequence of events
involved is shown in Fig 8. These compounds, for example POCI3 or
(NH4)2HPO4, function by adsorbing on dislocations
The onset of dusting can be inhibited in a number of exposed on basal planes which tire the nuclei for
ways. Organic vapours can substitute for water, and oxygen attack 73. One would therefore also expect them
the critical ratios of P/Po for hydrocarbons and alcohols to be effective substitutes for water in preventing
decrease monotonically with increasing chain dusting and this turns out to be so TM. Line B in Fig 7
length ~'~-72. With heptane, for example, the critical shows that the incorporation of 7 wt% of (NH4)2HPO4
P/P,, at 30°C is only of the order of 10 5 compared into a graphitic carbon eliminates the transition to
with that of 10 ~ for water vapour at the same dusting within the range of speeds, and hence contact
temperature. It follows that small traces of organic temperatures, examined.
TRIBOLOGY INTERNATIONAL 379
J. K. Lancaster--The influence of environmental humidity and water on friction, lubrication and wear
Under very severe conditions of sliding for which the formation on a metal counterface and, by analogy with
critical temperatures reach values well in excess of what has already been described for P T F E composites,
those needed to initiate the transition to high friction water is likely to lead to increased wear by disrupting
and wear, dusting wear rates increase with increasing these films. This has been observedS6; the wear rates
partial pressure of water vapour, as shown in Fig 9. of a partially graphitized carbon and a non-graphitic
This rather unexpected trend has been attributed 75 to one sliding against 0.15 pm Ra stainless steel increased
the idea that the mechanism of particle detachment in in water by factors of 830 and 25 times, respectively.
dusting conditions involves crack propagation leading The lower factor for the non-graphitic carbon is
to brittle fracture on a localized, asperity scale. Water attributable to its greater ability to polish the steel.
vapour then increases the rate of crack propagation in counterface in wet conditions. In view of the fact that
proportion to its availability, ie pressure. In this these materials were porous~ there could also have
respect, therefore, carbons appear to behave in a very been some cot~tribution to the increased wear in water
similar way to some ceramics, as will be described from the previously mentioned hydratflic effect. An
later. Crack propagation and fatigue are also believed additional feature of significance in the wear of
to be the predominant mechanism of particle detach- carbons ,in water is electrochemical corrosio~ of metal
ment in the normal, low-wear regime of carbons 7~ but counterfaces because of the electronegativity of carbon
here there is no significant effect of environmental with respect to metals. A limited amount of corrosion~
vapour pressure 7"~. The difference arises because in however, can sometimes be beneficial to wear by
ordinary environments sufficient amounts of adsorbed reducing the initial roughness of the counterface-~%
vapour are already available to migrate to a developing
crack tip77; the crack growth rate is therefore already
at its maximum value irrespective of the surrounding Lamellar solid lubricants
vapour pressure providing, of course, that it always
remains above the critical minimum to avoid dusting. Most of the information concerning the influence of
water and water vapour on the performance of lame!far
Some particular grades of manufactured carbons, solid lubricants is confined to McS2. Some limited dat
usually incorporating metals or synthetic resins to are available on related dichalcogenides of Mo. W
minimize porosity, are widely used for seals and Nb. and Ta 7s but. in general, their behaviour parallels
bearings in water-lubricated applications. Their wear that of MoS2. Unlike graphite. MoS2 is an intrinsic soIid
rates thus, presumably, remain adequately Iow for this lubricant whose low friction is a direct consequence of
purpose. There appears to be no information available, its lamellar structure and does not depend critically on
however, which directly compares these wear rates in the presence of adsorbed vapours. This is not to say,
water with those in air. Most carbons give reIadv-eiy however, that vapour adsorption is irrelevant: on the
low wear rates in air as a consequence of transfer fiIm contrary, as will be seen° ~t can significantly affect
both the magnitude of the friction coefficient and the
wear life of deposited films.
(x 10-4") 5
It was first shown by Peterson and Johnson 7') that the
friction coefficient of a rubbed film of MoS2 powder
on a metal substrate in air containing water vapour
increased up to a relative humidity of 70% ann
then decreased again. The increase was subsequently
confirmed on numerous occasions 8°-u3 and a review of
some of this early work has been given by Wirier s~
An example of the scale of the humidity effect on
E
friction is shown in the top half 0 ~ Fig
" 10 w,h ere vabaes
(3
master curve. It has also been suggested 8° that tb.e
L
reductions in friction of MoS2 films which usually occur
with increasing load and speed can be who~iy explained
I by the removal of adsorbed water from :he surfaces
as a consequence of frictional heating. This idea
appears to be supported by Barry and Binkelman s-5
who found the friction of rubbed MoS2 films in dry
atmospheres to be independent of load and speed (Fig
11). The absence of any variation with load. however.
VGCUGJm seems a little surprising because other experiments s'~
(-IO-I Tort) have shown that. irrespective of any humidity effects.
the friction of thin-film lubricants tends ~0 decrease
O IO 20
Pressure ,Torr
with increasing load because the constraining influence
of the substrate prevents the real area of contac~ from
Fig 9 Variation of the dusting wear rate of electrograph-
~ncreasing proportionately with load.
ite on non-graphitic carbon with pressure of water
vapour. Load = 20 N, speed = 27m s ~, temperature The increased friction in humid environments of rubbed
= 300°C (from R e f 72) MoS~ films is generally associated with an appreciable
380 December 90 Vo] 23 No 6
J. K. Lancaster--The influence of environmental humidity and water on friction, lubrication and wear
06
0.5 _ o o
O.IC
A =L
0 0 5 0 % rh
0.05
0.4 Dry
g
_o A Z~ O
I I I _ I _ I I I
I 2 3 4 5 6 7
o Speed, I0 -2 m s -I
g0.3
o
0.15
_o
~ ~ - - ~ , j ~ 6 0 % rh
0.2 OiO
:=L
V = 7 . S m m s -I
,oc
0 05
0.1 Dry
I I I I I_ I
0 200 4-00 600 800 1000 1200
Load, N
0 _ _ ~ J__ I I I I
0 I0 20 30 40 50 60 70 Fig 11 Variation or friction with load and speed Jot a
a Relativehumidifyofspecimenfemperoture,% rubbed M o S 2 film on stainless ,steel in wet and dry air
(after Barry and Binkehnan '~-~)
o 4-5 - ( specimen femperofure,°C )
9000
Table 4 S o m e examples of the trends in friction and w e a r caused by increasing humidity or the addition of
liquid w a t e r for various ceramic combinations
/
/
A / 4
/ E q E
/ 7z ~,/ / / / Soda-lime }J
10-5 [ /
/
& / J° ~Io-
! / / 10-6 /
/// i
-I
I
g
£ io-6 /
/ Z Vacuum- [
( Soda - l i me) "]
IO-8~ , , , I ~ , ~ ,I __}
I I0 I00
iO-7 Relative humidity, %
i0 -4 _
10-8 I I " I I I
4 5 6 7 8
Stress intensity Ki, N m-s/e (x 105)
Fig 14 Schematic variation of the rate of crack growth E
with stress intensity for soda-lime glass at different
humidities (after Wiederhorn n7)
;
-~: -5
'\
Experiments on the wear of glass, sliding against
smooth (0.02 Ixm Ra) tungsten carbide in conditions O %NkN X~
where the localized contact stresses remain well below \% \ weo~
T
the elastic limit, have shown that the wear rate E X
W
z \
increases with humidity in exactly the same way as the \
rate of slow crack growth in regime I (Fig 15(a)). % N
E \
Crack growth rates for glass in liquid water are also ¢o \
known to increase exponentially with temperature t19 ~10-6 -- N
®N
and Fig 15(b) shows that the wear rate again exhibits N
a similar trend, it is therefore of interest to examine \ I
N i
the extent to which ceramics might show similar \ x Crack growth
behaviour.
\ \o !
Al203 is known to be susceptible to water-accelerated
crack growth in an almost identical manner to glass !4
I0-7
80
I
60 40 20 ' 0 r,°C I
and increased wear rates have been observed in humid
environments (for example Fig 13(b)) and in water m3, 0.0030 0.0052 0.0034 0,0056 0.00:58
ms,ira,us. In support of this correspondence between b //fig)
wear and crack growth rates, it has also been found a2° Fig I5 Variation of the wear rate of glass against smooth
that the susceptibility of A1203 to increased wear in (0.02 # m Ra) tungsten carbide-cooak' " ' (a) rdative
w,,tn
water can be reduced by incorporating small pro- humidity in air; (b) temperature in water (from Ref
portions of ZrO2 into the Al203. These improve the zig)
fracture properties by limiting the extent of crack
propagation. However, the opposite trend of reduced
wear in water is also well documented m2,m4,n~. observed 121 when water is added to a fluid labricant0
Although A1203 has commonly been supposed in the as shown in Fig 16(a), and again attributed to
past to be essentially inert with respect to water, this the consequences of a tribochemical reaction, h is
does not seem to be correct. It has recently been interesting to note from Fig 16(b), however, that when
s h o w n m4 that tribochemical reactions can occur during sliding against rough surfaces the trend in the w e a r
sliding in water leading to the formation of an rates of A1203 with water content of the fluid is
atuminium hydroxide film which has a layer-lattice reversed: the higher stresses imposed by the rough
structure and is able to reduce both friction and wear. surface now lead to water-accelerated crack growth
Reductions in the wear rate of A1203 have also been becoming the dominant wear mechanism.
I I I I 10-8 !
0
0 25 ,50 75 I00
Soturc]tion of diester with woter. %
b o DISC in Ar+ H20
Fig, lO Variation of wear rates of Al,O¢ and PSZ I ~ ~ • PIN in Ar+ H20
with water content in di-2-ethylhexylsebacate. (a) on E] DISC in Air+ H2O
"Syndite", (b) on 35 tzm diamond-impregnated poly- • PINin Air +H20
DISC in H20
ester strip (from Ref 121) • PIN in H2O
10-9
o
"vE
There is also some ambiguity about the role of water 'z
in the wear of ZrO2. Wear rates, or the level of surface
damage, have been observed to increase in water "'s d
.m,~,~Js,,z2 and when water is added to a lubricant
(Fig 16) or to decrease m2'u'~. In general, the number
of reported increases outweigh the decreases. Much
less seems to be known about the effects of water on 10-t0
crack-propagation and tribochemical reactions with
Z r O : than for other ceramics. It has been stated ~2-~ O0 Z~
following microscopic examination of worn surfaces, •
that water increases the rate of crack propagation and o
recent evidence, also from wear experiments, suggests
that tribochemical reactions can occur as well. Sasaki m2
has shown that the wear rate of Z r O , first of all
increases with relative humidity up to about the 20% 10-11 I_ I I I I
20 40 60 80 I O0
level, then decreases again. This trend is qualitatively Humidity,%
explained in terms of the competing effects of water-
accelerated crack growth and tribochemical reaction. Fig 17 Wear rates of S(¢N4 sliding against itself in
air and argon as a function of relative humidity.
The wear behaviour of SigN4, SiC, and Sialon in water Hemispherical pin on disc, load = 10 N, speed =
or humid environments is much more consistent than 1 rnm s i (from Re,[ 123)
disrupting, or preventing the formation of, these third- prevents the formation of third-body debris layers or,
body layers. This was observed by Ajayi and Ludema ~26 if introduced after such layers have formed during dry
during slow-speed sliding of a diamond hemisphere sliding, leads to their rapid disruption via the 'hydraulic
o v e r A 1 2 0 3 , Si~N4,SiC and ZrO2, and has also been effect'. The presence of water in the vapour phase
noted with similar ceramics during sliding against a does not seem to inhibit the formation of third-body
metal-bonded diamond disc--'Syndite' ~Zl. Some results layers but it must almost certainly affect their properties
from the latter experiments are given in Table 5. The as, for example, with burnished MoSz films9.. Little
steady-state wear rates in air were all associated with is known so far about this aspect, the main problem
third-body formation, but such layers were no longer being that there is no convenient way in which to
obtained when sliding in water and the diester lubricant. characterize the properties of debris layers, such as
It is interesting to note that the wear rates in water their cohesive strength or stiffness. One way of
are all appreciably lower than those in the dried approaching the problem might be via some form
diester, in agreement with the contention that water of 'scratch' test on layers produced in different
can promote a tribochemical reaction. environmental conditions. This, however, remains a
challenge for the future.
Concluding remarks
Acknowledgements
The information presented in this review has revealed
a wide variety of different roles played by water in This paper was prepared during a period spent at the
friction and wear processes. Many of the effects of Institut National des Sciences Appliqu6es de Lyon.
varying humidity on the magnitudes of the coefficient The author is greatly indebted to Prof. R. Hamelin.
of friction and the rate of wear are relatively small, Director of the [NSA and Prof. M. Godeto Director
but they are nevertheless still likely tobe of significance of the Laboratoire de Mdcanfq~e des Contacts for the
when considering power requirements and life predic- provision of facilities, and to the Reglo~ Rh6ne-Alpes
tions for components in service. Measurements of the and the Fondation Scientifique de Lyon e~ Sud-Est for
environmental humidity should, therefore, always be financial support.
included in any materials-testing or research pro-
gramme concerned with friction and wear. There are
three effects of water, however, which stand out as References
being of major importance, namely its influence on
i. Bowden F.P. and Tabor D. The friction and wear of solids.
the following: Oxford University Press. ]950
(1) crack propagation; 2. Sikorski M.E. Correlation of the coefficient of adhesion with
(2) reactions with the sliding surfaces; various physical and mechanical properties of melals. Yra~1s.
(3) modifications to third-body layers of aggregated ASME. Ser. D. 1963. 85. 279-285
wear debris. 3. Peterson MoB° and Murray F. Requiremems of materials for
sliding and roiling contacts. Ch. 9. in Boundary lubrication.
When the friction and wear processes of particular A n appraisal of world literature A S M E . 1909, t45-/95
materials include any of these elements, some form of 4. Kubaschewski O. and Hopkins geE. Oxidation of metans and
humidity control during testing would appear to be alloys° 2nd Edn, Butterworths. 1962
highly desirable. The first two are affected by water 5. Barnes D.j.. Wilson JOE., 5tott F.H. and Weed G,C. The
in both its liquid and vapour phases. The fundamental influence of oxide fitms on the friction and wear of F e - 5 % C r
mechanisms involved appear to be reasonably welt alloy m controlled environments, Wear. ]977. 45, 16]-]76
understood, although there is still some uncertainty 6. Vaessen G.H.G. and deGee A.W.J. Influence ol wa~cr vapour
about how vapour adsorption leads to increased friction on the wear of lightly loaded contacts. Wear. 197], ~8. 325-332
with MoS2 films. The last of the three effects above, 7. Lancaster J.N. The formation of surface films ar the Lransition
however, is most important with liquid water and is between mild and severe metallic wear. Prec. R. See. f963
common to all material groups. Water effectively A 273. 466-483
8, Lancaster J.K. Severe metallic wear, Prec. ]m. Co~y~ ()~ Lubr.
and Wear. lust. Mech. Eng, t957. 694-700
9 Uh[ig H.° 1-Ming Feng, Tierney W° D. and McClellan A. A
fundamemal investigation of fretting corrosion. N A C A TN-
Table 5 Wear rates of four ceramics sliding against 3029 ]953
a 'Syndite' diamond counterface (~ 0.10 i~m Ra) in
10, Nield g.J. and Griffin D.G. Relation between wear rate and
air {~ 50% rh) water, and dried di-2-ethylhexylse- debris composition in wear of wrought iron and mild sleel
bacate (from Ref 121} Wear. ]964. 4, 111-122
11. Gotn H., Ashida M. and Endo K The influence of oxygen and
Ceramic Wear rates, 10 5 mm a N-~m water vapour on the friction and wear of an alumipAL!m £1oy
under fretting conditions, Wear 1987. 116 14]-[55
12. Gore H. and guckiey D,H. The influence of water vapour J~
Air Water Di-2-ethyihexylsebacate
air on the friction behaviour of pure metals during fretting.
Trib. Int. 1985. 18, 237-245
AI203 1.5 5 40 "t3. Larsen-Basse J. Influence of atmospheric humidity on abrasive
SiC 6 !10 2OO wear I 3-body abrasion. Wear. ]975. 31. 373-379
Sialon 0.5 6 105 14. Micha~ske T.A.. Bunker B.C. and Freimaa S,W. Stress
PSZ 0.8 1.5 8 corrosio~ of ionic and mixed ionic/covalent so]ids. J. A m
Ceram. Soc.. 1986. 69, 721-724
31. Schatzberg P. and Felsen i.M. Effects of water and oxygen 55. Lancaster J.K. Carbons and graphite in tribology. Proc. 5"oc.
during rolling contact lubrication. Wear, 1968, 12, 331-42 Chem. Ind. 4th London ('on.[~ on ('arbons and Graphite, 1976,
663-6 77
32. Felsen, I.M., McQuaid R.W. and Marzini J.A. Effect of sea-
water on the fatigue life and failure distribution of flood- 56. Lancaster J.K. Lubrication of carbon fibre-reinforced poly-
lubricated angular contact ball bearings A S L E Trans. 1972, mers. Part 1 Water and aqueous solutions Wear, 1972, 20,
15, 8 - 1 7 315-333
33. Schatzherg P. Inhibition of water-accelerated contact fatigue. 57. B r a m h a m R.W., King R.B. and Lancaster J . K The wear of
Trans. A S M E , J, Lubr. Tech. 1971, 91, 231 235 PTFE-containing dry-bearing liners contaminated by fuids.
A S L E Trans. 1981, 24, 479-489
34. Petch N.J. The lowering of fracture stress due to surface
adsorption. Phil. Mag. 1956. I, 331 337 58. Evans D.C. and Lancaster J.K. The wear of polymers. Treatise
on Materials Science and Techttolo~,,y, Vol. 13, Wear, 85-139,
35. Grunberg, L., Jamieson D. and Scott D. Hydrogen penetration
Academic Press, 1979
in water-accelerated fatigue of rolling surfaces. Phil. Mag.
1963, 8, 1553-1568 59. Evans D.C. and Lancaster J.K. Improvements in or relating
to bearing materials. Brit. Pat. I 552 603. 1979
36. Grunberg, L The formation of hydrogen peroxide on fresh
metal surfaces. Proc. Phys. Soc, B, 1953, 66, 15,7-161 60. Dobson J.V. The effect of humidity on brush operation Flee.
J. 1935, 32, 5 2 ~ 5 2 8
37. Swets D.E. and Frank R.C. Hydrogen from a hydrocarbon
lubricant absorbed by ball bearings. Trans. A I M E , 1961, 221, 61. Ramadanoff D. and Glass S.W. High altitude brush problem.
1082 1083 Trans. A I E E , 1944, 63. 825-829
38. Eischens R.P. Catalysis studies related to boundary lubrication. 62. Savage R.H. Carbon-brush contact films. (;en. Elec. Rev, 1945,
Ch. 5 in Boundao' lubrication: an appraisal o f world literature, 48, 1.7 20
63. Campbell W.Eo and Kosak R. Studies in boundary lubrication. 88, Fusare RoL. Effect of substrata surface finish on the lubrication
HI. The wear of carbon brushes in dry atmospheres. Trans. and failure mechanisms of m o l y b d e n u m disulphide films. A S L E
A S M E . 1948. 70. 491-498 Trans.. 1982, 25. 141-56
64. Plu~lova L.A. and Panyusheva Z.A. Intensive wear of graphite 89. Ross S. an{] Sussman A. Surface oxidation of molybdenum
materials. Mashinov. 1970. (4), FDT-MT-24-293-72. 112-110 disulphide. J. Ph)s. Chem. 1955. 5% 889-92
65. Lancaster 3.K Transitions in the friction and wear of carbons 90. deGee, A°W.Jo, Salomen G. and Zaat J.H. O n the mechanisms
and graphites sliding against themselves. A S L E Trans. 1975. of MoS2 fl!m failure in slidino friction. A S L E Trans.. 1965. g.
18, i87-201 156-63
66. Porgess P.V.K. and Wilman H. Surface re-orientation, friction 91. johnstone R.R.M. and Moore A.J.W° The burnishing of
and wear. in the unidirectional abrasion of graphite. Proc. m o l y b d e n u m disulphide on to metal surfaces. Wear. ]964. 7
Phys. Soc. 1960. 76. 513-525 498-512
67. Roselman I.C. and Tabor D. The friction of carbon fibres. ]. 92. Fusare R.L. Graphite fluoride lubrication: the effect of fluorine
Phys. D: Appl. Phys, 1976. 9. 2517-32 content, atmosphere and burnishing technique. A S L E Trans,.
1977. 29. 15-24
68. Savage R.H. and Schaeffer D.L. Vapour lubrication of graphite
sliding contacts J. AppL Phys. 1956. 27. 136-138 93. Briscoe B.J. and Smith A.C. The interfacial shear s~reneth of
molybdenum disulphide and graphite films A S L E Trans,.
69. Lepage J. and Zaida H. Influence of water vapour adsorption
!982. 25. 349-54
on the boundary conditions in tribology. Proc. 14th Leeds-Lyon
Syrup. on Interface Dynamics. Elsevier. 1988. 259-66 94. Bowden F.P. and Tabor D. The friction and lubrication of
solids Part II Oxford UniversiO, Press. 1964, p. 198
70. Pardee R.P. Moisture dependence of silver-graphite brushes
m air. nitrogen, helium and carbon dioxide Trans. I E E E 95. Haltner A.J. and Oliver C.S. Chemical atmosphere effects in
1967. PAS-86, 616-625 the frictional behaviour of m o l y b d e n u m disulphide, Proc, A CS
Div. Pet. Chem. Syrup. on Chemistry in friction and wear
71. Lancaster J.K. The wear of carbons and graphites, Treatise on
1958. 3 (4}. A 7 7 ~ 4
Materials Science and Technology, Vnl, 13, Wear. 141-174,
Aademic Press. 1979 96 Matsunaga, M.. H o m m a T. a~d T a n a k a A. Investigation vf
vapour adsorption on molybdenum disulphide by A n g e r
72. Lancaster J.K. a n d Pritchard J.R, The influence of environment
electron spectroscopy. A S L E Trans.. [982, 25. 323-28
and pressure, on the dusting wear of graphite. Y. Phys. D.:
Appl. Phys.. 1981, 14. 747-62 97. Pardee R.P. The effect of humidity on low-load properr, les of
73. McKee D,W°. Savage R.H. and Gunnoe G. Chemica! factors a bonded solid film lubricant. A S L E Trans. I972. 15. 130-I42
in carbon brush wear. Wear. 1972. 22. 193-214 98. Molybdenurn disulphide powder, lubricating. UK. D E F - S T A N
68-62 1 France. A I R 4223: W. Germany, VTL-6810-O]5:
74. Lancaster J.K. Additive effects on the friction and wear of
Canada. 3-GP-806a
graphitic carbons. Proc. 3rd. Leeds-Lyon Symlo on Wear o f
non-metallic materials'. MEP. 1978. I87-195 99. DiSapie A. Bonded coatings lubricate metal Darts. Prod. Er~g.,
5th Soy. 1960, 48-5,?
75. Lancaster J.K. and Pritchard J.R Or~ the dusting wear regime
of graphite sliding against carbon. J, Phys. D: Appl. Phys, 100. Sastry, V.R.K., Phatak S.D. and Sethuramiah A Experimenta~
!980. 13. I551-64 study of the influence of Iiquid lubricants on burnished MoS-
films. Wear. 1983. 86. 213-18
76. Clark WoT. and Lancaster J.K. Breakdown and surface fatigue
of carbons during repeated sliding, Wear. 1963, 6, 4d7-482 10i. Young R,L. Effect of conventional lubricams upon resin-
bonded solid film lubricants. Ll*,br. Eng. 1964, 20, 222-24
77. F r d m a n S.W. and Mecholsky n.J. Effect of temperature and
envlronmen~ on crack propagation in graphite. J.Mater. Sci. 102. Sasaki S. The effects of surrounding atmosphere on the friction
t97& 13. 1249-60 and wear of alumina, zircoma, silicon carbide and silicon
nitride. Wear o f Materials. i989 i l l 407-17, A S M E
78. Lansdewn A.R, M o l y b d e n u m disulphide lubrication. ESRO
Rap. [974. CR-402: ESA Rap. 1976: CR-764: ESA Rap. I977. 103. Kapelski G. Etudes des propri6tds tribologiques de c4ramiques
CR(P)-1045; 198I. ESA TRIB/4: 1983: ESA TRIBI7: Swansea thermo-m6caniques en fonction de la temp4ratare et pour
Trib. Centre Rap. 1984. 841549 diffdrents e n v i r o n n e m e m s Thesis, University o f Limoges, 1989
79. Paterson M.B. and Johnson R.L. Friction and wear investi- 104 Gates R.S., Hsn S.M. and Klaus E.E. Tribochemical m e c h a n i s m
gauons of m o l y b d e n u m disulphide: I. Effect of moisture. of A1203 with w a t e r Trib. Trans. i989. 32, 357-03
N A C A TN-3055 1953 105. Wal|briflge N., Dnwsen D. and Roberts E.W. The wear
80. Midgdy J . W . The frictional properties of m o l y b d e n u m disul- characteristics of sliding pairs of high density, polycrystalline
ohide. J. Inst. Pet. 1956. 42. 316-I9 aluminium oxide under both dry and wet conditions, Wear o:
Materials. 1983. 202-220, A S M E
81. Haltner A..L and Oliver C.S. Effect of water on the friction
of m o l y b d e n u m disulphide. Ind. Eng. Chem. Fund. 1966. 5. !06. ][shigaki H.. Kawaguchi [., lwasa M. and Teabana Y, Friction
348-55 and wear of hot-oressed silicon nitride and other ceramics.
Wear o f Materials. 1985. t3-21. A S M E
82. Gansheimer a. Effect of certain vapours and liquids on the
frictional properties of m o l y b d e n u m disulphide. A S L E Frans.. 107. Klaffke Do and Habig K.H, Fretting wear ~ests of silicon
!96Z 1 0 , 3 9 0 - 9 9 carbide. Wear o f Materials. I98Z 361-70. A S M E
83. Pritchard C. and Midgely J . W . The effect ol humidity on the 108. Ishigaki H., Nagata R, and ~wasa M. Friction ann wear of
friction and life of unbonded m o D b d e n u m disulphide films. partially stabilized zirconia. Proc. Con q on Tdbo[ogy: yq*iegon
Wear. 1969 1 3 . 3 9 - 5 0 lubrication and wear. 50 years on, t987. VoL H. 609-6]4. Inst.
84. Wirier W.O. M o l y b d e n u m disulphide as a lubricant: a review Mech. Eng., London
of fundamental knowledge. Wear, 1967. 10. 422-52
109, Kiaffke D° Tribological surface layers in the ~retting of ceramics.
85. Barry H.F. and Binkelman J,P. MoS 2 lubrication of various Proc. 14th. Leeds-Lyon Syrup. on Interface Dynamics, Elsevier,
metals. Lub. Eng. 1960. 22. 139-45 1988, 237-44
86. Karpe S.A, The effects of load on the frictional properties of
molybdenum disulphide. A S L E Trans., 1965, 8, 164-78 110. Scott N.G. Friction and wear of zirconia at very low sliding
speeds. -Wear o f Materials, 1985, 8-!2, A S M E
87. deGee A . W . J . , Begelinger A. and Solomon G. Influence of the
atmosphere on the endurance of some solid lubricants compared 111. Z u m G a h r K.-H. Sliding wear of ceramic/ceramic, ceramic/steel
at constant layer thickness. Proc. L Mech. E., 1968/9, 183(Pt.3), and stee!/steel pairs in lubricated and unlubricated contact.
18-27 Wear o f Materials, 1989 (t), 431-39, A S M E
i12. Rehbinder P.A. and Shchukin E.D. Surface p h e n o m c n a in and wear resistance of alumina-zirconia ceramic composites.
solids during deformation and fracture processes. Prog. SurJbce Trib. Trans., 1989, 32, 77-84
Sci., 1972, 3, 97-188 121.
Lancaster J.K., Mashal Y. and Atkins A.G. The role of water
113. Dufrane K.F. and Glaeser W.A. Influence of adsorbed fluids in the wear of ceramics. In press
on the rolling contact deformation of MgO singlc crystals.
Wear, 1978, 46, 55-64 122. Fischer T.E., Anderson M.P., Jahanmir S. and Sahler R.
Friction and wear of tough and brittle zirconia in nitrogen,
114. Braza J.F., Cheng H.S., Fine M.E., Gangopadhyay A.K., Keer water, hexadccane and hexadecane containing stearic acid.
L.M. and Worden R.E. Mechanical failure mechanisms in Wear. 1988. 124, 133-48
ceramic sliding and rolling contacts. Trib. Trans,, 1989, 32,
1-8 I23. Fischer T.E. and Tomizawa H. Interaction ot tribochemistry
115. Ajayi O.O. and Ludema K.C. Surface damage of structural and microfracture on the friction and wear of silicon nitridc.
ceramics: implications for wear modelling. Wear, 1988, 124, Wear of Material.s, 1985, 22-32, ASME
237-57
124. Kimura Y., Okada K. and Enomoto Y. Sliding damage of
116. Atkins A.G. and Mai Y-W. Elastic and plastic fracture. Ellis- silicon nitridc in phme contact. Wear of Materials, 1989 (1),
Horwood, ('hichcsler, 1985 .761-68, ASME
117. Wiederhorn S.M. Intlucnce of water vapour on crack propa-
gation in soda-lime glass. J. Am. ('eram. Sot., 1967, 50,407 14 125. Sugita T., Ueda K. and Kanemura Y. Material removal
mechanism of silicon nitridc during rubbing in water. Wear,
i18. Lancaster J.K. (:rack propagation and particle detachment 1984, 97, / - 8
during the wear of glass under elastic contact conditions. Proc.
14th Leeds-Lyon Syrup. on Interfiwe Dynamics, Elsevier, 126. Ajayi O.O. and Ludema K.C. Formation of transfer film dnring
1988, 111-119 ceramics/ceramics repeat pass sliding. Wear Of Materials, 1989
119. Wiederhorn S.M. and Boltz L.H. Stress corrosion and static (1), 349-59, ASME
fatigue of glass. J. Am. ('cram. Sot., 1970, 53, 543 48
127. Trabelsi R. and Berthier Y. Vclocit_~ accommodation between
12ft. Trabelsi R., Treheux D., Orange G., Fantozzi G., Homerin P. rubbing ceramics. Proc. 5th. Int. ('ongresv on Trihologv,
and Thevenot F. Relationship between mechanical properties Hel,sinki, 1989, 4, 48-53