You are on page 1of 31

Friction

For the two solid surfaces that are placed together the contact generally occur at
some discrete points of the nominal area of contact called real area of contact

The forces are exerted through these localized regions of contact between two
bodies and are responsible for friction.

Friction is defined as the resistance encountered by one body while sliding over
another and this definition contains two important classes of relative motion :
rolling and sliding.
Coefficient of Friction, = Frictional force (F) / Normal load (W)
May vary from 0.001 for lightly loaded rolling element bearings to 10 for clean metal surfaces in
vacuum. For most common metals sliding in air it lies between 0.1to 1.
Fig.1 A frictional force F, needed to cause Fig.2 Variation of the friction coefficient,, with normal
motion by (a) rolling or (b) sliding load W, for unlubricated sliding of steel on Aluminium in
air (Bowden and Tabor, The friction and lubrication of
solids, Clarendon Press, Oxford, 1950)

Fig.3 Variation of the friction coefficient,, with apparent area


of contact for wooden sliders on an unlubricated steel surface ,
(From Rbinowicz E, Friction and wear of Materials, John Wiley,,
1965
Laws of Friction
Coeff. Of friction for a given pair of materials and fixed conditions (or absence) of lubrication may be
almost. This led to the formulation of two empirical laws of sliding friction, known as Amontons law
(1699) , however, Leonardo da Vinci described them almost 200 years back.
1. Friction force is proportional to the normal force in sliding;
Implies that is independent of normal load, true for may materials under conditions of lubricated or
unlubricated sliding )
Often obeyed for unlubricated sliding
Most metals and other materials obey this law but polymers do not.
2. Friction force is independent of the apparent contact area;
Well obeyed by most materials except polymers.
The third law is added by Coulomb (1785)
3. Friction force is independent of the sliding speed.
Less well founded. At high sliding velocities of the order of tens or hundreds of m/s, coefficient
of friction falls with increasing velocities.
Theories of Friction
Early investigators Amontons and Coloumb envisaged that major contribution to
the frictional force came from mechanical interaction between rigid or elastically
deforming asperities, known as coloumb model.
In this model action of wedge shaped asperities cause the two surfaces to move
apart as they slide from A position to B.It can be equated by the work done by the
frictional force against the normal load.
In next phase of the motion from B to C most fundamental defect of this model
appears; now the normal load does work on the system.
No net work done or energy dissipation occurs in complete cycle therefore no
frictional force should be observable at macroscopic scale.
Sliding friction
Most current theories stem from Bowden and Tabor model for sliding friction
which assumes that frictional force arises from two sources : an adhesion force
developed at the areas of real contact between the surfaces(the asperity
junctions) and a deformation force needed to plough the asperities of the harder
surface through the softer.
Later it became clear that these two contributions cannot be treated as strictly
independent, but for convenience resultant frictional force can be taken as the
sum of these two contributing terms Fadh due to adhesion and F def due to
deformation.
Adhesion term arises from the attractive forces which are assumed to operate at
the asperity contacts, from common experience when two metals are press
against each other they do not generally adhere. However , if the surfaces are
clean and oxide free and free from other surface films significant adhesion is
observed between metals under ultra high vaccum.
Fig.4 Schematic diagram illustrating principle behind Fig.5 Illustration of adhesion between metals. A clean
Coulomb model for sliding friction. A surface roughness steel or brass pressed with slight twisting motion on to
is assumed to have a saw tooth geometry. As sliding the freshly-scraped surface of an indium. Appreciable
occurs from position A to B work is done against the force is needed to detach the rod from the block and
normal load W. The normal load then does an equal fragments of indium adhere to steel surface.
amount of work as the surfaces move from B to C.
a b
Cont.
Lead and Indium are ductile metals which can be demonstrated under laboratory
conditions.
Rounded end of brass or steel are degreased and abraded to remove surface then
presses on the freshly-scraped surface of indium strong adhesion will occur.
Detaching the rod from fragments of indium adhere to the rod showing adhesive
forces at the junctions are stronger than the cohesive strength of indium.
The crucial assumption of Bowden and Tabor model that adhesive forces will be
present is therefore justifiable.
The reason significant adhesion is not observable when two metals are placed
together are twofold: in the first place the surface will be covered with oxides and
absorbed films which will weaken the adhesion and second the elastic strains
around the asperity when under load generate enough stress to break the
asperity junction during the unloading process, unless the material is particularly
ductile.
Cont
Thus only in soft ductile metals and where oxide films are least partially removed
can appreciable adhesion be demonstrated under ordinary condition.

If we denote true area of contact, sum of cross sectional area of all asperity
junctions by A and assume all junctions have the same shear strength s, then
frictional force due to adhesion will be:
Fadh = As
Real area of contact is almost linearly proportional to the applied normal load.
Initial asperity contact will be effectively plastic and can be written as
W AH
Where H is the Hardness of softer material.
Cont
Contribution to the coefficient of friction from the adhesive forces is therefore
adh = Fadh/W s/H
For metals indentation hardness is about three times the uniaxial yield stress
H 3Y
Yield stress Y will be about 1.7 or 2 times the yield stress in pure shear, s,the precise
factor depending on yield criterion, thus
H 5 and thus

adh 0.2

The frictional force due to ploughing of harder asperities through the
surface of a softer material, the deformation term may be estimated by considering
a simple asperity of idealized shape.
Fig.7 Model for deformation component of friction,
in which a conical asperity of semi-angle ,, indents
and slides through the surface of a plastically
deforming material.
Cont
A conical asperity of semi angle slides over a plane surface the tangential force
needed to displace it will be some flow pressure
Fdef= H ax= H x2 tan
Normal load supported by the asperity is given by
1
W=H a2/2 = H 2 tan2
2
Coefficient of friction due to the ploughing term will be
def= Fdef /W= (2/ ) cot
If asperity is taken as wedge shape(plain strain model)
def= cot
Slopes of real surfaces will always be less than 100 (i.e> 800) ,therefore we would
expect def to be less than about 0.1.
Junction Growth
In simple models discussed so far material is assumed to have constant flow stress .
However all the materials strain harden to some extent although the normal load is
supported by plastic flow some distance from the immediate vicinity of the asperity
junctions, the junction will themselves work- harden significantly, which will raise the
value of s in comparison of H significantly.
In other models we have assumed that the true area of contact is determined solely by
normal load and is unaffected by tangential load this is in fact an oversimplification.
Normal and shear stresses should be also taken in account.

Let us assume a slab material loaded against a rigid plane surface, representing
in a very idealized form an asperity contact.
Fig.8 Stress acting on an element within an idealized asperity pressed against a
counterface (a) with no tangential force and (b) with a tangential applied force.
Cont
The elements of material inside the slab at a) is subjected to uniaxial compression
by a normal stress p0 and can be assume it to be at point of yielding, all asperity
contacts between metals are plastic, b) tangential stress is then applied to the
asperity junction, the element of material experiences an additional shear stress .
For material to remain at point of yielding the normal stress on the element must
be reduced to a value p1. If normal load remains constant ,then the area of contact
must grow the phenomenon is therefore known as junction growth.
Relation between p0, p1 and is determined by yield criterion. For
Trescas criterion, in which plastic flow occurs at a critical value of maximum shear
stress,
p1+ 42 =p02
For Von Mises criterion
p12+ 32=p02
Cont..
both equation leads to same conclusion. Let us examine for Equation obtained
from Tresca criterion. Normal and shear stresses are given by
p1=W/A
and
= F/A ; where A is true area of contact .F here
denote the tangential force and doesn't imply that sliding is actually occurring. We
can now substitute in equation obtained from Tresca
W2+4 F2=A2 p02
In a typical sliding experiment under dead-weight loading, W is constant,
while p0 is a property of the material(its yield stress in compression).The real area
of contact will therefore increase with increasing tangential force, and the ratio
F/W ,instantaneous value of will also increase steadily.
Cont
In this model growth process is not limited it could continue until whole
area of specimen was actually in contact and the coefficient of friction would reach
a very high value. However, practically it is limited by ductility of material and by
the presence of weak interfacial films. We can model effect of weak interface if we
assume that it will fail at some shear stress i ,less than the shear strength of the
bulk material .The maximum possible tangential force is then given by
Fmax= i Amax
And coefficient of friction is
= Fmax/W
If shear yield stress of the bulk asperity of material is 0, now from Tresca yield
criterion
p0=2 0
Now expression for
Cont..
1
= Fmax = 2 1/2
W 2((0/i) 1)
When the interface has the same shear strength as the bulk material (0/i)=1 ,
above equation shows becomes infinite, since junction growth is unlimited. For
weaker interface it is finite and drop rapidly as i/0 falls.

Fig.9 Variation of with the ratio between the shear strength


of the interface and that of the bulk material (i / 0 )
An interface only 10% weaker than bulk is sufficient to reduce to about unity.
For an interface having half the bulk strength, reduces to 0.3. For 1/10th of bulk
strength, = 0.5.

Weak interface can lead to extremely low value of ,and limited junction growths.
If one assumes the idealized asperity to be separated from the substrate by a weak film of
shear strength i ,then the frictional force is determined by the shear strength of the film:
F = Ai
and the normal load is supported by the plastic flow stress of the bulk of the asperity:
W= Ap0
Hence,
= F/W= i/p0

Above equation is important since it suggest a way to reduce the friction., i.e., if a film of material of
low shear strength can be interposed between two surfaces, then the coefficient of friction can be
lowered.
Better model for asperity deformation
Slab like picture is an idealization as asperities involved may be spherical, conical or
wedge shaped protuberances

Hence, for all such cases, the normal stress needed to cause plastic deformation will be
higher than that of frictionless compression of a slab due to plastic constraint.

Bowden and Tabor modified the earlier one to incorporate the general shape of
asperity
p12 + a2 = p02

Where a is a numerical factor, determined empirically, with a value of the order of 12.
However, this approach also leads to the same conclusions as envisaged by earlier
models.
Friction of Metals
Coefficient of friction depends critically on the experimental conditions under which
it is measured

Clean metal surfaces in vacuum form strong metallic bonds across the interface and
when these are pulled apart , transfer of metal occurs from one to another.

With no or limited contamination the junctions growth in vacuum is limited by the


ductility of the asperity material.

Sliding friction under Ultra high vacuum occurs in space engineering and special
measures must be taken to in designing components to work under such
environments.
Self-mated metal pairs
Coefficient of friction in air are much lower than those in vacuum and lie between 0.5 to 1.5

Frictional behavior also depends composition and microstructure of materials apart from
the conditions of measurement

As evident from table, Gold has a high because it does not form any oxide film although there may be some
adsorbed gases present in air, hence asperity junctions are strong. Another reason is the ductility which helps
in junction growth. Both these factors contribute to high friction of gold in air.

All the other metals oxidize in air and form oxide films typically between 1 to 10 nm thick in
a few minutes exposure. These oxide films significantly affect the sliding behavior. Films are
beneficial in reducing friction.

Effect of oxide film can be mitigated if the film is penetrated.


Self-mated metals in air

Friction coefficient drops with the admission


of oxygen as more and more oxygen is
introduced, the value of reached the value
normally observed in air at ambient pressure.

Fig. The effect of oxygen on the friction of pure iron. In vacuum the
value of is very high; as the oxygen concentration is increased, the
coefficient of friction falls (Buckley, D H, Surface effects in Adhesion,
Friction, Wear and Lubrication, Elsevier, 1981)
Self-mated metals in air

Transition from low to high friction as the load is increased.


At low loads oxide films effectively separate the two bodies
and electrical resistance at the interface is high, track formed
is smooth.

Coeff. Of friction is low because oxide acts as low shear


strength film and mostly because of its low ductility resulting
in limited junction growth.

Transitions are common in other metals also. Oxide films


effectively separate the two surfaces and avoid metal metal
Fig. The variation of coefficient of friction with applied
contact normal load for Copper sliding against copper in air,
unlubricated (from Whitehead J R, Proc. Roy. Soc. Lond. A
201, 109-124, 1950).
Self-mated metals in air
For some soft and ductile metals (tin and indium), metallic contact occurs even at the lightest
loads and friction is therefore high shows no change with load. In such cases soft substrates offer
little mechanical support.

To summarize, friction of pure metals in air and unlubricated conditions is often determined by
the presence of oxide films on the surface
If the oxide film is retained during sliding, surface damage is less and the coeff. of friction is governed by
surface oxide
The value of friction is often lower than at higher loads where oxide film is penetrated and metallic
contact takes place. Under this condition surface damage and rapid wear occur.

In sevaral metals no transition occurs because oxide film is able to retain itself over a wide
range of load, whereas in others, where oxide layer is penetrated even at lighter loads, the
contact is always metallic.
Dissimilar metals and alloys
Effect of temperature
With the increase of temperature
(i) Mechanical properties get changed (ii) Rate of oxidation increases
(iii) Phase transformation may take place

Fig. The variation of coefficient of friction with temperature for


various pure metals sliding against themselves in ultra-high vacuum:
(a) c.c.p. metals; (b) b.c.c. metals; (c) h.c.p.metals (from Bowden F
P and Childs T H C, Proc. Roy. Soc. Lond. A 312, 451-466, 1969).
Friction of Metals

Fig. The variation of coefficient of friction with sliding speed for


Pure Bismuth and Copper sliding against themselves (from Bowden
F P and Tabor d, The Friction and Lubrication of Solids, Part II,
Clarendon Press, Oxford, 1964)
Friction of Metals
Friction of Metals

You might also like