Video Lecture on Engineering Fracture Mechanics, Prof. K.

Ramesh, IIT Madras 1

Module No. # 07, Lecture No. # 35: Crack Closure

We have been discussing concepts related to crack initiation and life estimation in this
chapter. And we had looked at in the last class, for the growth of a fatigue crack, Paris came
out with a very elegant empirical relation relating da by dN and delta k. We call that as the
Paris law. And I said, it is a very important contribution to fracture mechanics. In fact, the
designers looked at utilizing concepts of LEFM, only when they came across the contribution
by Paris. Because that provided a mechanism for them to implement in practice and find out
whether the structure would be safe or not, by knowing how the crack propagates due to
fatigue. And I also mentioned, it’s a very complex phenomena.

And environment has a very important role to play in crack growth. The ingenuity of Paris
law was, he got the kernel of how the crack grows. If he had looked at all issues and
attempted to arrive at an empirical relation, he may not have arrived at an empirical relation
at all. And now, we will see in the discussion today, how modifications of Paris law are
utilized to explain various phenomena observed in actual experimentation? And in fact,
before this chapter, we had also discussed modeling of plastic zone at the crack-tip. That
knowledge would be useful to understand the phenomena observed in the experiments.

(Refer Slide Time: 02:03)

Crack Closure: Designer’s Approach

So that is the reason, why that

chapter was covered earlier. And
now, we will take up a very
important contribution by Elber.
This was done in 1970. He brought
out a concept called crack closure.
And what he observed was, a
fatigue crack growing under cyclic
tension can close on itself at about
half the maximum load. And this
was purely an experimental
observation. Based on the
experimental observation, he
arrived at this kind of a very
interesting result. This effect
becomes significant for crack
growth calculations under random
loading.

(Refer Slide Time: 02:45)

See, you will have to keep in mind, fatigue is one of the most difficult type of loading that
acts on the structure. You will invariably have variable amplitude loading. And one of the
difficulties is, finding out the actual load spectrum. It’s not simple. There are methodologies
available. And you have to study the literature. And then only calculate the actual load
spectrum. So invariably, you will have variable amplitude loading. From modeling point of
view, we may study, under constant amplitude loading, how the crack grows? But if you
want to take the knowledge to the field, you will have to account for random loading. And if
you go for random loading, you will have to bring in the observations of crack closure into
your life estimation calculation.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 2

(Refer Slide Time: 03:54)

And what we find is, the crack

closes about half the maximum
load. And how does the crack
open? What is the difficulty in the
experimentation? In practice, the
crack does not open uniformly
along its front. And there is a
continuous transition from closed
to open, making experimental
determination of the opening stress
difficult. See, unless you make
certain observations by experiment
and then identify these aspects
have to be modeled by a numerical
model, you cannot develop a
numerical model. So, initially you
need to have experimental observation for such complex phenomena.

So, one of the difficulties is, how to see whether the crack is fully opened or not. And how do
you judge, whether a specimen displays crack closure phenomena. People also have
methodologies to look at that. And one of the simplest way is, to get the load deflection
curve. So, if you get the load deflection curve for a cracked body experiencing crack closure
is bilinear. This is the observation. And you have the displacement versus applied stress.
And if you look at, the stress displacement relationship is bilinear.

(Refer Slide Time: 05:34)

So, you have slope changes from

the initial to later part. Initially, it is
more stiff. Initial slope is dictated
by the closed crack. So, that
means, crack behaves as if it is
smaller in length. Then, you have a
slope which is different. At higher
loads, because the crack is fully
opened, the slope is also different.
So, if a fatigue cracked specimen
experiences a stress displacement
behavior as bilinear which is the
model again. In practice, you will
have a smooth translation from this
to this. And that’s also depicted here.

And the question here is, how will you identify when does the crack closes? And when does
the crack opens? This is a very difficult to measure experimentally. And people have found
various ways to do that. One way of describing this is, you say at this stress, when there is a
translation of the slope, the crack is fully opened. And when you have a curved portion
becomes straight here, this you say, that crack is closed. Part of the crack is closed. So, this
is one way of estimation. You know, at this stage, it’s also wise to look at, what was the
history in the case of crack closure. Initially, people did not believe that crack closure could
happen. This was done by Elber. There was an article written by him in Current Contents in
1983. And in that, he says, the original discovery of the crack closure phenomenon was
made when I was a graduate student at the University of New South Wales in Sydney,
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 3

Australia. Rather than being the result of years of directed effort, it was a split second
observation made while cutting up a fatigue cracked plate. See, this is very important.

You know, sometimes you do a very systematic study yet you don’t get a result. Here, he
admits it was not a systematic study. He chanced upon an interesting observation. He was a
good experimentalist. He did not miss out the details. See, this is very important.
Experimentalists have to be alert. Look at anything deviating from your basic understanding,
record it. And scrupulously honest. Because if you report what you observe in all its totality,
new theories can develop. That’s what happened in crack closure concept. In fact, when he
reported, it was not accepted by the community. It was rejected also. I will read his quote for
that. He says, because of the geographic location, I was not aware during the ensuing
months, during which I wrote up that discovery into my thesis that I was going to contradict
many of the existing teachings in fracture mechanics. Or how much that phenomenon
contributed to the resolution of the complexity of crack growth under variable amplitude
loading. And what is interesting is, the next few sentences, it says like this. After having my
thesis initially rejected by the first American reviewer, I went to the Federal Republic of
Germany to continue my research’. And the story goes like this. You know, poor fellow, he
found a very nice observation. And when he reported because the conventional thinking was
not able to accept that crack can close. So, they rejected the thesis. Then he carried on. He
joined NASA Langley research center. Now, it is an accepted practice. In fact, what I am
reading is from Current Contents. This says, this week’s citation classic. It was published in
December 12, 1983 where his paper on the significance of fatigue crack closure - damage
tolerance in aircraft structures, published as ASTM- STP- 486 had a very high citation of
240, I think. That’s what I remember, 240 citations. And it is discussed as a citation classic
which was originally rejected. It became a citation classic. And we will further see, what the
crack closure concept is all about.

(Refer Slide Time: 10:56)

So, this is what I had mentioned

which is summarized again. Elber
in 1970 demonstrated through a
set of carefully planned
experimental investigations that a
fatigue crack propagating under
zero to tension loading that is R to
R equal to 0, might be partially or
completely closed at zero load.
And this observation has led to
new concepts in fatigue crack
growth studies. Obviously, after
initial hesitation. People were not
accepting his theory, initially. Once
it was convinced, people said how it could be modeled. The ultimate usage of this
observation is, how do you modify your crack growth calculation. There is a designers’
approach. We will see that designers’ approach first.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 4

(Refer Slide Time: 11:57)

So, what they have done is, to

account for crack closure,
designers employ a modified Paris
law which is as follows. Normal
Paris law will have da by dN equal
to C delta K power m. Here, the
delta K term is put as delta K
effective. And how do you define,
what is delta K effective? Suppose,
I have a sinusoidal variation. And I
have K max and K minimum is 0.
But you have, at a particular stress
only, the crack starts opening.
Afterwards you know, only the
crack becomes effective. The crack closes by this value of K op. So, the effective delta k is
between K max and K op. So, people could embed the phenomena of crack closure by
taking delta K effective. Now the question is, how to find delta K effective? That is where the
challenge lies. It is not a simple task. Conceptually, the crack is effective only part of the
cycle.

So, you look at, what is the effective delta K rather than the actual delta K. So, you have to
find out, how to get delta K effective? Several models have been proposed. People have
done experimentation and calculated it. And you will also have to bring in, the effect of
stress ratio R. As a function of R, how does delta K effective changes. So, this is how
research proceeded. And we will see, how they developed delta K effective. So, delta K
effective is K max minus K opening where K op is the value of KI when the crack opens
completely during the cycle.

(Refer Slide Time: 14:50)

And this is difficult to evaluate. But

we have to do that if I have to
embed crack closure in our crack
growth calculation. There are
empirical relations reported. And in
order to ensure that crack closure
indeed happens, Elber also
provided fractographic evidence to
show that crack closure could
affect the shape of the striation
pattern. See, if you look at the
history also whoever reports the
discovery has to justify ‘yes, his
discovery is correct’. If he doesn’t
do that homework, that concept does not get recognized. We had seen in the case of
Dugdale model, he had performed the experiment but he did not report the result. So, there
was skepticism among the practitioners not to accept Dugdale model in the initial stages.
Once Hahn and Rosenfield provided the experimental evidence, people started using
Dugdale mode. So, in the case of Elber, when he reported the crack closure later he also
provided fractographic evidence to show such thing happens. And the effect is the striation
peaks are flattened due to deformations during crack closure. And as I mentioned, one of
the important issues is, how to estimate the stress intensity factor corresponding to fully
opened crack.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 5

And you have to note, unlike the fracture toughness, K opening is not a material constant
but depends on a number of factors. So, that makes the problem difficult. Different alloys
exhibit different closure behaviors. So, this is one observation. So, you have to do a lot of
tests, to find out. Even for a given alloy, the closure behavior is different in different loading
regimes. So, that makes the problem much more complex. And I think, you have the paper
by Elber. This is the paper published in 1970, ‘Fatigue crack growth under cyclic tension’.
This was published in Engineering Fracture Mechanics. This is volume number 2 between
37 to 45. You can go and have a look at it, how he has presented the work.

(Refer Slide Time: 17:47)

Empirical relations for Keff

So, you have empirical relations to

find out, what is delta K effective.
And you know, for all aerospace
structures, aluminum is very
important. So, you have a 2023
Titanium, that is T3 aluminum.
Elber reported that delta K
effective over delta K can be
written as 0.5 plus 0.4 R where R
is the stress ratio.

And he had obtained the relation

for the range 0, R. R is between 0
and 0.7. And there were also other
researchers who have extended
this range of R. But the point to
note here is, you need to have some way of finding out, what is the value of K effective? If
necessary, I can also enlarge it. So, this is the equation that you have. And another
researcher reported that if negative R is to be considered then one should use the following
formula for the same alloy. This is little more elaborate. We will also have a look at it. Show
here whichever way it goes out. But you can write the expression, delta K effective divided
by delta K equal to 0.55 plus 0.33R plus 0.12R squared. This is from R equal to, between
minus 1 to 0.54. R lies between minus 1 and 0.54. You have another expression. See, the
point here to note is, you have empirical relations available.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 6

(Refer Slide Time: 19:55)

Crack closure – current focus

So, look at the literature, find out

for your given alloy, what is the
empirical relation to use? Then you
will be able to use concept of crack
closure. And what is the current
focus of crack closure? Because
initially people were not ready to
accept crack closure. Once they
have started accepting, people
also have done additional work. A
great deal of effort has been taken
by various researchers to
measure, characterize and predict
crack closure behavior and its
effect on crack growth rates. A
majority of this research has been
experimental in nature. That’s what I said. The phenomena is complex. So, initially you have
to understand, what is happening through experiments? With such inputs, it’s possible for
you to develop numerical models.

And what is the main difficulty in the experiment? One of the main difficulties in experiments
is obtaining the information of displacement histories at mid-section in a thick specimen.
See, I have already mentioned, the crack opening is not uniform along the front. So, you
have to ensure if I have a thick specimen, the in-between portion whether it has opened or
not, I have to have experimental measurement. Only then, it will be more realistic. That
poses a challenge. It is challenging as it is not easily accessible. I suppose, you have been
able to write these points. And the time is sufficient for you to write.

(Refer Slide Time: 21:46)

Having seen that the experiment is

difficult, the current focus is on
developing appropriate numerical
models to study the phenomena.
Not only this, they also want to find
out the effect of T stress. That is
sigma naught x for
Photoelasticians. We have spent
ample time in understanding this
when we discussed crack-tip
stress and displacement fields. So,
they don’t want to restrict only to
the stress intensity factor K. They
also want to go to the second term.
And such an attempt is made
using finite element methods which
is recently reported by Roychowdhury and Dodds. It’s a paper in 2004. I think, I would read
this for you. This is by Roychowdhury and R. H. Dodds – ‘Effect of T stress on fatigue crack
closure in 3D small scale yielding’. This is on International Journal of solids and structures,
published in 2004. You would be able to see the details of the paper.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 7

And before we move further, you know I would also like to share another note which came in
the citation classic which Elber wrote. He mentions like this. The time from the initial
discovery of any phenomenon to a profitable application is often long. In this case, which
means crack closure, the inclusion of crack closure in numerical crack growth calculations
adds significant cost and complexity for gains in accuracy. Only required in the space and
aeronautics industries. Many industries will continue to use simpler and less accurate
theories, leaving the continued work in crack closure as an area of basic research.

(Refer Slide Time: 24:18)

Residual stress ahead of a crack in a cycle: Plasticity induced crack closure

So, you will have to keep in mind,

you know, if you want to employ a
new phenomena, it doesn’t come
cheap. It comes to a added cost.
Now, what we will do is, we will try
to go and understand, what is
crack closure? There are various
categories of crack closure. And
we will first see, plasticity induced
crack closure. And first concept
that you have to learn is, what is
the residual stress ahead of a
crack in a cycle? And what is
shown here in a tiny fashion, is
actually when you have a
sinusoidal load, you are actually
looking at one half of the curve where you are going from 0 to maximum. So, when I go to 0
to maximum, we would see, what happens at the crack-tip?

So, you go from 0 to maximum and then come down. And make neat sketches of this as
much as possible. So, I have this as the crack. I have the axis as r. This is the distance
measured. And in this axis you have, stress is marked. So, you have Tresca yield criteria. I
have marked this as sigma x. So, this will be horizontal. And there will be some variation of
this along the crack axis. And we have already seen, the plastic zone in the case of mode I
will be something like this. And what is shown in blue is the elastically deformed
neighborhood. The slight light brown color shows a material plastically deformed.

So, what you have to recognize is, even though your external loading is much smaller
because of the presence of a crack, when the load is increased gradually from to 0 to peak,
invariably you will have plastic deformation near the crack-tip. So, this happens during
loading. And what is our focus is, we want to see, what is a residual stress ahead of a crack
in a cycle? This also you have to note. We are just looking at a virgin material. We have
loaded it for the first time. And you have a plastic zone surrounded by an elastic region. And
when I reduce the load, what would happen? Unloading takes place. Whatever that has
happened to the elastic region, it will spring back to its original position. If there is no plastic
deformation at the crack-tip, the spring back would be uniform all throughout. But because
you have plastically deformed zone near the crack-tip, the elastic recovery will go and
compress this zone.
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 8

(Refer Slide Time: 27:22)

Let us see, what happens

pictorially. So, you will have a zone
where this is still under plastic
zone condition much smaller in
size because of the elastic
recovery of the neighborhood. And
in essence, you will have a
residual stress formed at the end
of one cycle. You know this
concept, you will have to
appreciate. At the end of one
cycle, invariably you will have a
residual stress pattern which is
compressive near the crack-tip and
tension ahead. And let us see, what are the steps? Elastic zone springs back when load is
released. In the process, introduces a compressive load on the plastically deformed portion.

So, you will have a residual compressive stress and a residual tension. You know, if you
understand this, many other discussions, what happens when there is a overload? What
happens if there is a crack closure? For all of this, you have to recognize because the load
is increased and decreased, you have a residual stress formed ahead of the crack-tip. This
is the first knowledge you should get. So, what I will do is, I will replay this animation again.
You have carefully drawn this sketch as much as possible. And what we will do now is, we
will replay the animation where I am showing, what happens when load is increased from 0
to the peak value? So, you will have a plastic zone. And this is the elastic region. And this is
the approximate shape of the plastic zone at the crack-tip, surrounded by elastic region. And
this is schematic. Don’t think, only a circle around the crack is elastic. No, it is not so.

(Refer Slide Time: 29:55)

Just for visualization purpose.

Because we have to feel,
something goes and compresses
it. You know, that visualization is
very good. When I show something
in blue color and then it works.
Now, what I am going to do is, I am
going to reduce the load. This is
the story when load is increased
from 0 to peak load. And this
plastic zone is subjected to
tension. It is the value of sigma ys.
When there is elastic spring back,
when the load is reduced, you
have a plastic zone where the stress is minus sigma ys. And that is what is shown here.

And you have a compression followed by residual tension. If you understand this, all our
future discussions, it becomes easier to discuss. Now, you have to see, if I have multiple
cycles, the crack will go through a plastic wake. We will see that also. I have very nice set of
animations and we will see through that. In a growing fatigue crack, behind the crack-tip, a
plastic wake is developed. You know, I have initial crack length. And what is shown here is,
this is indicated that you are having a cyclical loading. You are seeing, at the end of one
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 9

cycle what happens. After several cycles, what happens. You know, the crack is growing in
length.

(Refer Slide Time: 31:08)

Plastic wake

So, when the crack grows in

length, the plastic zone also will
grow in shape. It’s not going to
remain the same size. It is going to
increase. That is what is depicted
in the animation. And you observe
it. It is done for sufficiently long
distance. And the crack goes
through plastically deformed
portion which is called as a plastic
wake. And what Elber postulated
was, that crack closure is due to
the existence of residual strain in
the plastic wake.

See, if there is no fatigue loading,

you will not have residual stress formation. Because the load is increased and decreased,
there is a residual stress formation. Because of the residual stress, the crack has to wade
through that. Because of the existence of residual strain, in the plastic wake, crack also
closes. We will see how. I have an animation which explains that. But you have to
recognize, as the crack grows, the plastic zone sizes keeps increasing. And also formation
of plastic wake. And this goes with your cyclical loading. That is what is depicted below.

(Refer Slide Time: 33:23)

Different types of Crack closure

You know, you would have noticed

that there was an oscillation. So, I
think, what I can do is, I can
quickly zoom it. So, you will see,
this only shows that I am looking at
a cracked specimen after this kind
of loading. Okay. So, what you find
here is, a strip of material along the
crack edges though no longer
loaded but has undergone plastic
deformation previously constitutes
the plastic wake. We had seen it in
the previous slide.

And what is depicted here is, I

have a cyclical loading. And I am
discussing, what happens at this point, that is, after few cycles, at the peak load, the story of
the crack would be like this. I would have a crack fully opened. But it would have a plastic
wake. A strip of material along the crack edges though no longer loaded but has undergone
plastic deformation previously constitutes the plastic wake. So, you have the plastic wake
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 10

shown. Suppose, I reduce the load, what happens? When the load is reduced due to
permanent elongation of the crack lips, they close before the load is 0. So, that is what is
shown here. Ideally, only when the load point comes to this, the value of stress is 0.

But even at this stage, you find the crack is closed. So, the effective SIF has to be calculated
from this point to this. It cannot go from 0 to K max. So, that is the issue that is being
discussed. So, what you are really looking at is, because of cyclical loading, if the load is
increased and decreased, there is residual stress formation. If I have multiple cycles like
this, the crack has to go through the plastic wake. And the plastic wake makes the crack
faces to close before the load reaches 0. So that means, only part of the loading is really felt
by the crack. That is why we have K effective. And this was the model proposed by Elbert.
And later similar models have been proposed for various situations.

So, what you have here is, the crack is fully open when the load is at its peak. But the crack
is closed even before the load has reached 0. So, that is the issue that you have to keep in
mind. And this explains, how to understand the phenomena of crack closure. And you can
see the animation again. For this case, you have the plastic wake. And this explains, what is
the plastic wake? And this shows when the load is reduced. You find when the load is still
having some value, the crack is closed. And you also have the typical residual stress pattern
at zero load. And I think, I can magnify this for you.

(Refer Slide Time: 37:14)

So, I have this as closure

stresses. And you have a residual
tension here. And because of this,
the crack surfaces close. And this
needs to be appropriately modeled
in your da by dN calculation. And
that is what is given in the next
slide. In view of this, the crack
faces do not open until K reaches
K opening. You have already
drawn this graph. This is just to
emphasize that you have to use
the delta K effective. And we have
also seen, this becomes significant
when you have random loading.
For constant amplitude loading,
probably you may not want to use
it. But for space structures, you will
have to definitely do, is the
recommendation by Elber. And it
comes with a cost. It doesn’t come
cheap.

(Refer Slide Time: 38:50)

Once you have appreciated this,

there are also other models that
are given which are easy to
understand on a similar logical
frame work. Crack closure can also
happen because of phase transformation. The residual strain developed as a result of a
phase transformation produced by applied stress also causes crack closure. The physical
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 11

appreciation of phenomena is very similar to plasticity induced case. Instead of the plastic
wake, here it would all be dictated by the phase transformation. Otherwise, there is
comparison between the two. And that is what is mentioned here. This is similar to the
plasticity induced crack closure.

Here again, you look at, after few cycles, at the peak load. At the peak load, the crack is
open. Just to distinguish it from plastic wake, here the shading is different. But the pictures
are essentially same. And this is because of phase transformation. And now you see, when
the load is reduced before it reaches 0, the crack is closed. It is because of phase
transformation. So, that is what is indicated here.

(Refer Slide Time: 40:50)

So, once people understood the

concept of crack closure
postulated by Elbert. Several other
mechanisms have been identified.
One is because of plasticity
induced. Here, it is because of
phase transformation. And the third
case is wedge induced crack
closure. It could be due to oxide,
usually associated with aggressive
environment. The corrosion
products become wedged between
crack faces. That is what is shown
here. I can show it little big.

Because of the corrosive action, you have corrosion product shown. And when the load is
reduced, you know, because of corrosion products, it appears as if the crack is closed. It has
a similar effect of crack closure. And this is called oxide-induced closure. So, this is
observed in aggressive environment. In every case, discussing on crack closure, we show
the cyclical loading. You see, what happens at the peak load? And what happens when the
load is reduced? And when the crack faces are closed. So, that is the way you have to
interpret what is given on the left and what is given on the right. Left shows the loading. And
the green point shows in the loading history at which point we are really looking at the crack
faces. This is at the peak load. This is at a reduced load.

(Refer Slide Time: 42:31)

Then, you also have roughness

induced closure. This is another
phenomena. And what is reported
is, microscopic roughness of the
fatigue fracture surface cause
this. At the microscopic level, due
to micro structural heterogeneity,
mixed-mode conditions can exist.
You know, once you come to
here, too much of material
science knowledge you need to
have. It’s all highly exaggerated
pictures, I am going to show. So,
do not come and say, why you put the crack faces as jagged. At such high magnification,
even a straight edge would appear jagged only. So, you have to look at, these are not
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 12

macroscopic pictures; highly magnified microscopic pictures. That’s the way you have to
look at it. And what is mentioned is, the displacement due to mode II can cause mismatch
between upper and lower crack faces contributing to crack closure phenomena.

And we will also see that as a picture. Here again, you see the load history. At the peak
load, you are looking at it. And at the reduced load, you have a picture like this. I think, I
could magnify it for you. And, this is called mode II induced closure. And this mode II loading
comes at the microscopic level. So, what you will have to keep in mind is, crack closure is
an accepted phenomena. Though it was not initially accepted, later people have recognized
its role. And many other mechanisms also, people have observed and categorized.

So, in all those cases, you should have a methodology, how to find out delta K effective. So,
you have to perform more tests and look at the literature and find out, how to get the delta K
effective. Only then, you will be able to use the crack closure model for finding out the da by
dN. See, there is also another important aspect that you have to keep in mind. Suppose, I
have an overload, is it beneficial from fracture mechanics point of view or not? See, if you
recall, in my early discussion, I have said, plasticity is a friend of fracture mechanics.

(Refer Slide Time: 45:06)

Influence of overload on crack Growth

In fact, we would argue based on

that only. And people have
noticed, if you do proof testing
occasionally, the life of the
structure is improved. In proof
testing, what do they do? They
load the structure up to 125
percent of the load for which it is
designed. So, you are really
putting an overload at periodic
intervals. People have observed
from actual practice that it has
helped the life of the structure. And
you will have a proof based on
plastic deformation that overload
really helps. And overload of what
kind? That is another question. Okay. So, we will look at a graph like this. On the x axis, you
have number of kilocycles. On the y axis, the length of the crack is put. It’s a log graph on
the y axis. And initially, you would take a constant amplitude.

And let us see, how the crack growth for this case is? You also have some numbers given to
this. It is varying from 48 MPa to 113 MPa. And you find, the crack growth is something like
this. Suppose, after sometime, I have an overload which goes from positive to negative. It
goes to 188 MPa to minus 28 MPa for the purpose of this graph. And, how does the crack
growth looks like? So, definitely the rate at which crack grows after an overload going from
positive to negative is definitely smaller than what you see when you have constant
amplitude. But has not really benefited the structure. Let me take another case where I have
an overload only in the positive direction.

How does the crack growth rate changes? And what you find here is, when there is overload
only in the positive direction, there is definitely retardation in the crack growth. And I have
another overload here; another overload here. You have to know, why is this so?
Video Lecture on Engineering Fracture Mechanics, Prof. K. Ramesh, IIT Madras 13

Nevertheless, it is beneficial. So, the wisdom, people gained by sheer experience of periodic
proof testing of the structure has helped its life. You have an answer from fracture
mechanics point of view. What way we are going to look at is, the plastic zone would be
much larger. When I have a overload, the plastic zone would be much larger. That has
helped this retardation.

So, your modeling of plastic zone is very important. And recognizing, when you have a
cyclical load, you are going to have a residual stress ahead of the crack-tip. And, when I
have an occasional overload, I will have a very large plastic zone. And this crack has to
wade through that plastic zone. So that, in a sense, retards the crack growth. And you have
a model for that. And this model is proposed by Wheeler. The details of it, we would see in
the next class. Again, he will go back to the basic kernel of how Paris reported da by dN.
There would be modification to that which will accommodate the effect of overload. And this
is very important in the fracture literature - Influence of overload on crack growth.

So, in one graph, you see for constant amplitude, how is the crack growth? For a reversed
overload, how does it retardation take place? For overload in one direction, the retardation is
quite significant. So, in this class, we have essentially looked at concepts of crack closure
proposed by Elber. We also saw, how he arrived at crack closure. How the result was
initially accepted by the community. After sustained effort, he has been able to convenience
a scientific community to accept this phenomena. Now, it has become a routine, in
advanced life estimation calculations. People have embedded that. And the softwares like
NASGRO and such software use this as part of their life estimation calculation. Thank you.