Modern Construction Materials

Prof. Ravindra Gettu

Department of Civil Engineering
Indian Institute of Technology, Madras

Module - 3
Lecture - 8
Failure Theories

Welcome to the eighth lecture of Modern Construction Materials. In the previous lecture,
we had looked at how materials respond to stress and today we will talk about failure
theories.

(Refer Slide Time: 00:38)

We look at what kind of criteria we have to use for determining when a material will fail
under multi-axial stresses. I start with this picture of a temple (Refer Slide Time: 00:40)
in Halberd, Karnataka which has survived since the 12th century.

We see in this picture a lot of nice stone columns which has survived, taken the load for
many hundreds of years and have not failed. What do we consider as failure?
(Refer Slide Time: 00:59)

A structural material is considered to have failed when it can no longer perform its
design function and in this case we are talking about mechanical aspects. Failure may
occur in two ways a) complete fracture i.e brittle failure or rupture, when an element of a
structure break apart. b) Failure can also be considered as an excessive deformation that
is the structural element deform so much that though it does not rupture or break, the
deformation is such that it cannot be used for what it has been designed for. (i.e)
something that we can call ductile failure as supposed to brittle failure in the first case.
Now, when we have to study failure and determine when a material fails under a uni-
axial loading, it is pretty simple.

We look at the stress strain curve and this stress strain curve when drawn up to failure
can tell us how a material will respond to stress and when the material will fail and how
it will fail. Under multi-axial stresses, it is more complicated to visualize and therefore
failure theories are needed for representing the material behaviour and this is usually
done for metals based on plasticity or yielding and in other materials plasticity may be
combine with fracture.
(Refer Slide Time: 02:43)

When we look at the uniaxial behaviour of a metal say in tension we can get different
types of behaviours, when we compare different materials. The top curve here (Refer
Slide Time: 02:58) is for mild steel where we have a linear elastic response culminating
in what is called the proportionality limit or the elastic limit. In this case both coincide.
Then we have the yield Plateau then there is hardening and at point ‘d’ (Refer Slide
Time: 03:19) which is the maximum stress we have necking occurring. Now the neck
starts to develop and finally, you have failure. In the case of aluminum, we find that there
is no definite point where we can identify the yielding or the proportionality, yielding to
start or proportionality to complete it is somewhere ‘x’ in the top most curve.

So, in order to define yield point objectively what is done for materials such as aluminum
where we do not have a clear, well defined yield point like we had in mild steel. In mild
steel it is that a 0.2 percent offset is taken. A line is drawn parallel to the initial slope at a
strain of 0.2 percent and wherever it intersects is taken as the yield point. This is classical
of poly crystalline materials which have different grains yielding at different points in
time and this we have discussed in detail in some other previous lectures. We can also
have some metals which do not undergo significant yielding or necking and failure, like
in the case of cast iron (again we saw in the previous lecture that cast iron was a brittle
material and we have a slightly curved behaviour instead of a linear elastic behaviour, we
can say that the proportionality limit is somewhere here in the region (Refer Slide Time:
04:54) the curves stops being linear and then again an elastic limit could be somewhere
here (last point of the iron stress-strain curve) and then failed. In this case, we do not
really define the different reasons like we in the case of mild steel.

(Refer Slide Time: 05:21)

So, what is clear is that, between metals itself the behaviour can vary significantly. The
tensile behaviour is also used to determine how a material should be processed. Going
back to the behaviour of mild steel say in the case of hot rolled steel, we have this type of
behaviour that we just discussed, we have elastic behaviour signified by a linear elastic
range followed by the onset of yielding, the upper yield limit, then we have a lower yield
limit a stable necking occurring, strain hardening, we have the peak here (Refer Slide
Time: 06:02) and then we have failure (complete rupture) occurring at a strain which can
be given a name of epsilon which can be called epsilon f (𝜀𝑓 ) for failure.
Now, instead of looking at the hot rolled steel behaviour, if you think of a cold rolled
steel what has happened in the cold rolled steel is that this material that we have in the
upper diagram (Refer Slide Time: 06:33) was loaded to a certain point and then unloaded
that is what we see here (Refer Slide Time: 06:38). So, the hot rolled behaviour is this
(Refer Slide Time: 06:43) but what has happened is that in cold rolling we have
deformed the material to a certain extent unloaded and this new material now is the cold
rolled steel and if we were to determine the tensile behaviour of the cold rolled steel we
will now has an elastic part starting from the zero reference stress strain and then here
(Refer Slide Time: 07:13) we will have yield, then necking and then rupture. So, we need
this behaviour(Refer Slide Time: 07:20) to determine how much of cold working has to
be done, so that we get the desire yield value and also have a sufficient amount of
elongation before failure.
(Refer Slide Time: 07:41)

So, the knowledge of this behaviour is not only needed for design but also for process.
The complete stress strain behaviour under all sorts of loading conditions is difficult to
model and it is quite complex. In this diagram shown above, (Refer Slide Time: 07:58)
we put together different types of loading conditions and how the material can respond.
This is to understand that the behaviour is quite complex. A single simple model may not
be sufficient to explain all the possible scenarios. In this case (Refer Slide Time: 08:21)
we have linear elastic behaviour then if we continue and unload, you will have a certain
loading, unloading generally this will have the same slope of the initial part of the curve
and then if we go on say the whole stress say from this point (Refer Slide Time: 08:38)
you are holding stress constant now due to creep that we have studied before, this strain
will continue to increase.

So, here (Refer Slide Time: 08:46) you need a model which looks at creep also, then if we
continue and we come to a certain point and we hold this strain constant (Refer Slide Time:
08:54) that is your holding strain constant we will see now that the stress drops (Refer Slide
Time: 08:58) this is called relaxation. This has mechanisms similar to creep but here what
we are doing is looking at the behaviour when strain is a held constant instead of stress being
held constant in creep, in relaxation, we have strain being held constant and now this stress
relaxes of the stress drops. In along this part (Refer Slide Time: 09:31) in any case that we
unload behind the elastic region we will have a permanent strain. So, whichever model is
needed for the plastic part has to correctly look at how permanent strain occurs. Then we
can also have fatigue we looked at cyclic loading in the previous lecture and we have to
see what sort of stress strain diagram we have when the stress is cyclic this is called a
hysteresis loop. (Refer Slide Time: 09:59) So, the behaviour in elastic region is quite
complex. Many phenomena can intervene.
(Refer Slide Time: 10:19)

And therefore we need complex models and simple models are not sufficient. Let us look
at a material that does not have ductile failure, say rock or concrete. We would have
stress strain diagram this time may be looking more at compression (Refer Slide Time:
10:35) rather than tension because these material are generally weak in tension and we
use them more in compression. So, if we were to test a specimen of concrete under
uniaxial compression, we will have initially a small non-linear part which can be
attributed to the closure of pores to the settling of the boundaries under the load. Then we
have an elastic part this is where (Refer Slide Time: 11:09) we will expect the material to
be under service conditions the slope of this (Refer Slide Time: 11:13) is the young’s
modulus. Then damage is initiated defects which occur in the material slowly propagate
there is micro-cracking and this (Refer Slide Time: 11:23) is called the pre peak non-
linear region where we have distributed micro cracking occurring. At a certain point of
strain the micro cracks coalesce, localize and larger crack starts forming, the cracks link
and the load carrying capacity comes down (Refer Slide Time: 11:41). In a rock crack
joins and cracks starts opening up then you have major cracking or faulting, large strains
occur and then you have failure. The part of the curve (Refer Slide Time: 12:00) where
failure occurs may also depend on the stiffness of the testing machine or of surrounding
in general and getting a unique objective post peak part in these materials is sometimes
very difficult. This behaviour where the stress drops after a peak is called softening, as
supposed to what we saw in the metals when the curve goes up and it call its strain
hardening this is called strain softening (Refer Slide Time: 12:31).

(Refer Slide Time: 12:34)

So, any behaviour that is below a plastic plateau is brittle and is called the strain softening
and what is above a plastic flat behaviour is called strain hardening and material such as
concrete, rocks, ceramics exhibit such type of nonlinearity,where we have a post peak drop
in stress as strain increases and this is called strain softening.

(Refer Slide Time: 13:08)

Now plastic stress strain curves can be idealized for modeling and these are the different
models (Refer Slide Time: 13:18) that often used on the left top, we have a rigid
perfectly plastic model, where instead of a finite elastic modulus we have rigid material
the elastic modulus is infinite then we have the plastic part and unloading, reloading is
now vertical and the plateau start at the stress equal to the yield strength. This may be
more realistic the curve on the top right where we have an elastic part both in
compression and in tension, then we have the plateau occurring that is yield strength and
initial slope is given by the young’s modulus and unloading reloading follows the slope
of the young's modulus.
Hardening behaviour can be modelled by these two curves at the bottom
(Refer Slide Time: 14:16) with a bi linear behaviour signifying represented by two
slopes ‘e’ and ‘h’ or you can have a power law type hardening where we have curved
behavior.

(Refer Slide Time: 14:34)

When we look at multiaxial loading we have to remind ourselves of principal stresses

and principal directions and if you go back to strength of materials you would have
studied that under every stress state there are three principal directions, orthogonal to
each other (that is right angles to each other) along which the principal stresses act. Each
principal stress represents the maximum or minimum normal stress for one set of plane
stresses. Along those principal planes we do not have shear stresses and we have only
axial stresses.
(Refer Slide Time: 15:25)

The principal stresses are usually designated by 𝜎1 ,𝜎2 , 𝜎3 considering tension as

positive. ‘𝜎1 ’ would be the maximum principal stress, ‘ 𝜎3 ’ would be the minimum
principal stress. The maximum shear stress in the body is given by the difference
between the maximum and minimum principal stresses divided by 2.
𝜎1 −𝜎3
So, 𝜏𝑚𝑎𝑥 is equal to .
2

(Refer Slide Time: 15:57)

In the cases hydrostatic pressures when we have hydrostatic compression, the pressure
on the body is the same in all directions that is, it is always normal to any surface on

which it acts or 𝜎1 = 𝜎2 = 𝜎3 = - p (minus p ) which is the pressure and therefore,

no shearing is possible. So, we will not be able to have any shear stress when the
material is under the hydrostatic compression. So, this means failure due to shearing will
not occur under hydrostatic compression

(Refer Slide Time: 16:40)

And this goes back to when we started looking at the condon morse diagram and we said
that failures cannot occur under pure compression because the atoms as they are pushed
together have very high repulsive forces and we have a behaviour like what we see
(Refer Slide Time: 17:00) as the pressure is increased as is the hydrostatic pressure is
increased the atoms are pushed together. So, that is as strain is increasing atoms are push
together the repulsive force is increased and at very high pressures we find that the curve
between hydrostatic pressure and volumetric strain actually starts going up ward due to
the nonlinearity that we saw in the inter atomic bond going back to the condon morse
diagram.

So, here there is a lot of repulsion and then we enter into the rigid way the bonds are
behaving in a non-linear manner. The slope of this line (Refer Slide Time: 17:50)
is the bulk modulus (k). In some cases; however, we can have this curve (Refer Slide
Time: 17:55) instead of going up going down were this would indicate the collapse of
the microstructure. That is if you have a lot of pores in the material, these pores can
collapse due to the very high hydrostatic pressures that is being applied.
(Refer Slide Time: 18:13)

Coming back to biaxial stress states and principal stresses and principal planes. Suppose
we have a point ‘o’ (Refer Slide Time: 18:23) around at which we are applying stress this
is where in generic stress state given by the different normal stresses and the shear
stresses, we would find that there will be a set-of planes on which the principal stresses

act 𝜎1 , 𝜎2 . This is the case of plane stress we are looking at something like a thin plate

where 𝜎3 is equal to 0 for simplicity, but the same ideas apply to all multiaxial

behaviours. So, the principal stresses are now 𝜎1 and 𝜎2 and you see that on the

principal planes we do not have any shear stresses.

(Refer Slide Time: 19:04)

Now how do we use this concept to determine failure. There are different failure criteria.
The first one and most simple is the maximum principal stress criterion of the Rankine
theory. According to the Rankine theory, yielding begins when the maximum principal stress
reaches a value equal to the tensile or compressive yield stress under uniaxial tension or
compression that is yielding in any state will occur when the principal stress reaches the
value corresponding to yield in the uniaxial case. And in the uniaxial case now we know that
yielding will occur when the stress reaches the yield strength. So, what we are saying is that
under a biaxial stress state as soon as the maximum principal stress reaches the yield strength
yielding will occur under yield stress. So, according to Rankine theory yielding occurs
when the maximum principal stress reaches the value of the yield stress.

(Refer Slide Time: 20:23)

And this is represented by the yield surface which says that as long as the stress state lies

within the surface defined by 𝜎1 ,𝜎2 equal to y on the four sides. And when we have a
stress state determined by 𝜎1 ,𝜎2 .Inside this yield surface (Refer Slide Time: 20:52)

failure does not occur. Failure occurs as soon as the stress state reaches or touches this
yield surface (Refer Slide Time: 21:00) that is when yielding will start to occur, that is

the meaning of the yield surface. This is in two-dimensions with 𝜎3 equals to 0.

(Refer Slide Time: 21:10)

In the case of 𝜎3 being nonzero, in the case a multiaxial stress state instead of a square

yield surface we have a cubic yield surface. Again the surfaces are defined by the yield

strength. So, now again as long as the stress state is such that with 𝜎1 , 𝜎2, 𝜎3 known,
the point is inside the cube then failure does not occur and yielding is not occurred. As
soon as the point as stress increases, reaches yield surface, the failure is said to occur.

(Refer Slide Time: 21:56)

In the tresca criterion or what is called the maximum shear stress criteria we give more
importance to shear stress. This theory is based on the observation and this we have
discussed extensively in previous lectures that in ductile materials slip occurs due to
yielding.

This slip is provoked by shearing and therefore the tresca criterion gives the maximum
important to the maximum shear stress. The criterion says that the yielding occurs when
the maximum shear stress under in arbitrary stress state reaches the value of the
maximum shear stress at yield in uni-axial stress. So, we have to find out at what value
the maximum shear stress will have at yielding and at the uni-axial tension and when the
same maximum shear stress occurs under any arbitrary condition yielding is said to
occur.

(Refer Slide Time: 23:03)

Under uni-axial condition this will be the case(Refer Slide Time: 23:17) 𝜎1 is equals to
y, at yield 𝜎2 , 𝜎3are 0 (remember this is uni-axial loading). So, now, the shear stress
are the maximum shear stress is (𝜎1 − 𝜎3 )/2 (that is half the yield stress). So, what we
have said is when the maximum shear stress under any conditions reaches this value
(Refer Slide Time: 23:37) then yielding will occur under multiaxial loading now there
are three possible shear stresses which can occur given by the half the differences
between the different principal stresses. The maximum now will govern failure. So,
linking these two (Refer Slide Time: 23:58) we have the tresca criterion which says that
failures will occur if any one of these conditions are satisfied.

𝜎1 − 𝜎2 = + + +
−𝑌 , 𝜎2 − 𝜎3 = −𝑌 , 𝜎3 − 𝜎1 = −𝑌

(Refer Slide Time: 24:17)

And this is represented by now by this yield surface (Refer Slide Time: 24:21) for the

tresca criterion this is again for 𝜎3 is equals to 0 (plane stress situation). This surface
comes from 𝜎2 equals to Y , 𝜎1 equal to Y .So, again as long as stress state is inside
this yield surface failure (Refer Slide Time: 24:43) does not occur. And when stress state
reaches this yield surface, yielding is set to start.

.
(Refer Slide Time: 24:52)
The tresca yield criterion gives good agreement with experimental results for most
ductile materials and it is very simple and therefore it is the most often used yield theory.
The main objection or the limitation of this theory is that its ignores the effect of the
intermediate principal stress if you remember the shear stress only depends on two of the
principal stresses. What is better is that we look at now, the maximum distortional strain
energy theory which predicts yielding slightly better than the tresca theory,
but the differences are not generally more than fifteen percent.

(Refer Slide Time: 25:46)

Tresca is more easy to apply and more exact theory would be, the maximum distortional
strain energy theory. This maximum distortional strain energy theory is called also the
von Mises theory. Other names are the octahedral shear stress theory or the Huber-
Hencky-von mises theory. In all these cases we consider that yielding occurs when the
distortional energy density reaches a value equal to the distortional energy density at
yield in a uni-axial case. So, we find out what is the distortional energy density at
yielding under uni-axial tension and when that same distortional energy density occurs in
any other arbitrary state we consider that yielding occurs. The total strain energy can be
divided always into two parts- a volumetric part and a distortional part. 𝑈0 is the total

strain energy given by above equation (Refer Slide Time: 26:57) where 𝜎1 , 𝜎2 , 𝜎3 are

the principal stresses, E the young modulus and 𝑈0 can be divided into the volumetric
energy and the distortional energy. The volumetric energy is now what is giving a
hydrostatic situation. Distortional energy is that corresponds to the change on a distortion
in the shear.
(Refer Slide Time: 27:17)

𝑈𝑉 as I said this is the volumetric energy related to volume change under mean
hydrostatic pressure. Under any arbitrary stress state there will always be a component
that can be attributed to hydrostatic pressure, the remaining other part is the distortional
energy which is related to the change in shape.

U v is equal to the ((sigma 1 +sigma 2+ sigma 3)2 /18 k) where k again is the bulk
modulus given by the right corner equation(Refer Slide Time: 27:58). The distortional
energy is given by left bottom equation (Refer Slide Time: 27:59) i.e the sum of the
squares of the differences between the different principal stresses divided by 12 G where
G is the shear modulus. So, we can see that this clearly from the hydrostatic part that we
know hydrostatic pressure will not cause failure, it only causes volumetric compression,
as the stress increases and we saw that shearing leads to slip and yielding and that is
what is behind the distortional energy.
(Refer Slide Time: 28:47)

So, let us see how to apply this criteria to failure. So, what we said is the distortional
energy density at uni-axial loading must first to be determined. So, under uni-axial
tension we have 𝜎2 , 𝜎3 is 0 ,at yielding 𝜎1 equal to y. So, using the equation of the
previous slide the distortional energy is given by y2 /(6 G). So, what we as said is the
distortional energy and any arbitrary stress state when it reaches this value yielding will
occur. So, in a multi-axial loading we say that this would be the equation. This is set
equal to this value(Refer Slide Time: 29:35) and we get finally this equation.(shown in
the Refer Slide Time: 28:47)

(Refer Slide Time: 29:44)

So, the von mises yield criteria failure through yielding is represented by red line. when
we draw the failure surface we get these red line which circumscribes the yield surface
that we had in the tresca condition shown as the blue line. So we find that they are very
similar except that there is a slightly higher stress that can be taken before failure occurs
given by the difference between the red line and blue line. This is because we have not
taken in to consideration the volumetric part and we are only looking at the shear part.
So, actually the stress that can be endured is slightly higher in some cases than what is
given by the tresca criteria.

(Refer Slide Time: 30:31)

We can also have failure that is occurring by other mechanisms such as fracture or
cracking rather than just by yielding. For this we have to look at failure theory called the
Mohr-coulomb theory and for this again you have to revise the Mohr’s circles which
your studied in the strength of materials and we know that Mohr’s circle is a graphical
representation of stress state if we consider an element given in figure (a) (Refer Slide
Time: 31:10) with 𝜎𝑥 , 𝜎𝑦 and 𝜏𝑥𝑦 as the known normal and shear stresses along a
certain plane, we can now draw a Mohr’s circle fixing point a and b from known values
of 𝜎𝑥 , 𝜎𝑦 and 𝜏𝑥𝑦 and we can now draw this circle (Refer Slide Time: 31:42)

where this circle cuts x axis the normal stress axis are the principal stresses 𝜎1 would

be the maximum principal stress 𝜎2 would be the minimum principal stress, and the

value given by the top of the Mohr’s circle is the maximum shear stress 𝜏𝑥𝑦 . Y axis
gives shear stress and the X-axis gives the normal stresses.
(Refer Slide Time: 32:11)

So, this would be the Mohr’s circle corresponding to that element in this body (Refer
Slide Time: 32:13). Failure is now taken to occur when we look at the Mohr’s circle at
failure at a certain stress state. When we draw all the different Mohr’s circle at failure for
this material and we take the envelope this is called the coulomb failure envelope (Refer
Slide Time: 32:41). We now assume that failure will occur when any Mohr’s circle
touches this failure envelope. When the circle is smaller failure has not occurred, stress
state is such that failure is far from occurring and as the stress increases this Mohr’s
circle will become larger and larger and finally, it will touch the failure envelope. The
coulomb failure envelope which is taken linear and failure is then set to occur.
(Refer Slide Time: 33:17)

We can determine failure envelope through test for example, we have a material that is
subjected to different confinement pressures say we take a cylinder or a core of some
material and subjected to different confinement pressures (i.e) 𝜎3 . In first case have no
confinement pressure, second case you have1000 mega pascal confinement pressure and
third case 5000 Mpa (Refer Slide Time: 33:41) and when we increase the 𝜎1 the axial
pressure, 𝜎3 is confinement pressures and now we have increasing axial pressures under
different confinement pressure in three different cases.

We have in first case a linear part and then the peak occurring at around 27 Giga pascals
and then failure and this would be the 𝜎1 (Refer Slide Time: 34:15) at failure under this

value of 𝜎3 =0. So, we repeat on an identical specimen a test with 𝜎3 equal to 1000

mega pascals and we again load axially keeping 𝜎3 constant and by increasing 𝜎1 and at

failure we observe that 𝜎1 is around 36 Giga pascals. Again another test is done as we
want to get at least 3 points to confirm the failure envelope we now apply a confinement
pressure of 5000 mega pascals and we keep increasing 𝜎1 , the axial part and we find

now that with higher confinement the failure now occurs at 𝜎1 equal to 55 Giga pascals.

So, we have now 3 sets of 𝜎1 , 𝜎3 values at failure for this material .We have also seen
that, as confinement increases we have a higher failure axial stress and this is the concept
that we use a lot in civil engineering. We confine a material to provide effectively for a
higher load carrying capacity. A simple example is a concrete column with hoops or ties
which confines the concrete and therefore, increase its effective load carrying capacity.

(Refer Slide Time: 36:09)

So from tests like this we construct this failure envelope and we have done huge tests of
the same material under different confinement pressures and axial stresses. We now
draw the Mohr’s circles for the cases represented by each of these tests. The failures
stress is 𝜎1 . So, this is 𝜎1 and this would be 𝜎3 for the last test (represented in red color)

(Refer Slide Time: 36:30). For the test that we did 𝜎3 equal to thousand mega pascals
(middle circle) this would be the Mohr’s circle and this would be the case (smallest
circle) where there is 0 confinement(Refer Slide Time: 36:44) i.e 𝜎3 is equal to 0.Now
we draw an envelope tangent to this Mohr’s circle and this is now the failure envelope
for the material that we have tested. (Refer Slide Time: 36:59)
(Refer Slide Time: 37:04)

So, this is the way that the failure can be constructed for different materials. What we
generally see is that instead of a straight line the failure envelope is curved. Coulomb
originally defined the failure envelope as a straight line, but later Mohr showed that the
slope decreases as the confinement pressure increases, there is the flattening out of this
curve (Refer Slide Time: 37:26) and so, you get a pinched response initially and a
parabolic failure envelope and this is now called the Mohr’s coulomb failure envelope.
What we also observe from this diagram(Refer Slide Time: 37:42) is that the angle of the
fracture plane remember that this angle (Refer Slide Time: 37:45) , now tell us their how
the failure will occur and as the confinement changes this angle indicating the failure
plane also changes. You see that when we look at the angle of line joining the point
which coincides with the failure envelope and the centre of the Mohr’s circle, we find
that this angle changes (Refer Slide Time: 38:07), this indicates that the failure plane
changes its direction as the stress state increases. Lastly we should understand that there
are different materials which have a varying failure mode as the stress state increases or
changes, say in a brittle material like concrete or rock we have to employ lot of varying
failure theories sometimes these are empirical based or modified failure theories and one
such failure theory or a combination of failure theories is shown on the figure below (on
the left top) (Refer Slide Time: 39:01) which is a composite failure envelope starting of
with tensile failure criteria.
(Refer Slide Time: 38:21)

So, we could have a material failing under tension under very little confinement seen this
case (Refer Slide Time: 39:14), we have a specimen which is subjected to compression
and with little bit of lateral tension and you can have splitting type failure and tensile
fractures occur. Then with little bit more confinement you can have the behaviour that
we discussed earlier the more parabolic failure criteria this is the spark (Refer Slide
Time: 39:41) where we have a change from the vertical splitting to shear type failure.
This is called the transitional tensile fracture. You have the crack that is sliding and
opening due to this stress state, some amount of compression or confinement then we
have an area which is corresponding to the coulomb or straight line failure criteria. This
is where you should have purely shear dominating the failure there is some amount of
confinement here, until in these cases the confinement was very low here and as you go
right on the curve you have most significant confinement and you have shearing
occurring. Then we have a brittle transfers brittle plastic transition going towards the von
mises yield criteria given by this diagram (Refer Slide Time: 40:37) may have shearing
and some amount of a plasticity indicated by the bulk. Then finally, we will have very
high confinement leading to plastic type behaviour where we can apply the von mises
plastic yield criteria. So, you have here (Refer Slide Time: 40:53) the shear lines forming
at 45 degrees (at E) and failure occurring with plastic yield. So, this obviously is very
complicated complex failure envelope which will require lot of parameters to develop
and something like this is needed when we want to cover different types of failures in
brittle material and other materials. We also see that the mode of failure in other
materials such as concrete, rocks, ceramics would change the type of failure. The failure
mode changes with the confinement.
(Refer Slide Time: 41:30)

So to conclude we have looked in the previous lectures on how materials respond to stress
and today we have looked at what sort of criteria we need to explain failure, to design for
failure and to understand failure. And with this we will look at different mechanical
properties of continuous media. So, in all these cases we are looking at materials which do
not have any discontinuity. What we will do in the next lecture is look at fracture of
materials where we look at discontinuities , defects becoming cracks and then the concept of
stresses strain sort of breakdown. If you remember from strength of materials and mechanics
and what we look at in previous lecture ,stress and strain are defined at a point where strain
was the change in length divided by the original length and from that definition, we can
understand why the concept of stress and strain breakdown when there is a crack. If you
can a imagine a point which where you suddenly have a crack the original length is 0
between the plane defined by palms now (Refer Slide Time: 42:56) when you have a
crack you suddenly have a displacement between them. So, there is a change in length
whereas, the original length was almost 0. So, if you divide the finite change in length
divided by a 0 original length you have a infinite strain. So, whenever there is a
discontinuity or a crack you have infinite strains and what you will also see in next
lecture is you sometime have infinite stresses. So, therefore, the concepts that we have
discussed in the previous lecture and in this, where we looked at continuous media
breaksdown and we have to bring in the concepts of fracture mechanics that will deal
with in the next lecture.
Thank you very much
v