Lecture 02 : Part I

IIT Kanpur July 201
20 followers
Jul 19, 2024
[Music]
students welcome to lecture number two
of this course on a IML in materials
engineering so first we'll see some
recap of whatever we learned in the last
lecture that is lecture number one you
know I introduced the course or I made a
small uh 10-minute presentation on the
introduction to the course that's
basically for very simple thing there's
no need of recap of that but then I told
you why we need to use a IML in
materials engineering the requirement
data and informatics is very important
in design and discovery of materials
that is exactly I told you the last
lecture but you know again let us talk
something small time on what is ai ai is
the ability of machines and computers to
mimic human decision making as well as
problem solving
capability and this ability as you know
is very limited at this moment but we
are happy what abilities it has and
that's why we can do all kinds of fancy
things using Ai and know machine
learning and deep learning are part of
the AI and and AI is
basically meant to use machines who can
understand and achieve some specific
goals therefore machine learning uses
structured data and deep learning uses
unstructured data which we will discuss
a lot in the different parts of this
course so now ai is only whatever AI we
use today is only narrow AI that is it
can only do specific task but it is not
cannot do General task that means
general AI is still a dream to achieve
and it's a concept it is still not close
to whatever we can even think of and AI
routinely does speech recognitions as
you know I told you about customer
service AI can answer customer service
questions or even provide
recommendations AI uses computer Visions
quite a lot your in a mobile phone you
have a Google lens and that actually
uses a computer you take a photograph
and then you can detect what is the type
of flower what type of fruits or what
the type of what kind of faces if Google
has all the databases it can easily
detect it so that's nothing but use of
computer vision universities use
recommendation engines many of them they
are actually all based on AI and today's
world everybody knows about stock
trading stock trading is also automated
using thanks to AI you know but why you
know why is important for Material
Science that is what is the basic
purpose of this course so I showed you
this chart this is very important chart
taken from Michael asb's book which I'll
refer in next couple of slides is
basically tells you use of human
materials in human civilization starting
from 10,000 BC to till
2020 and the time is counting on right
so as you can see the importance of the
materials
vary as a function of time it's a metal
or it's a composite or Ceramics and
glasses polymers came M letter that's
why it has not been discussed here but
these three four sets of materials or
are three sets of materials are very
important in this world uh no sorry
polymer is there you can clearly see
polymer larts are very important like
bealite nylon polythylene PMA acrylics I
still today high temperature polymers
are getting used in various applications
so polymers are becoming more important
in today's world as far as the relative
importance in concern well let's talk
about how the materials are selected for
a particular application and I am very
happy to select this example because
this is a day-to-day instrument which
you use a vacuum cleaner and thanks to
Professor Michael aspie which has
described this application very nicely
in this book called material selection
in mechanical design it was published so
back 1992 almost like 30 years ago more
than 30 years ago but still it's very
relevant as you see here vacuum cleaners
are wiely used to clean various kinds of
things especially uh you know carpets or
walls or you know sometime computers or
devices but you see these are the
pictures 1905 more than 120 years almost
120 years ago 1950 19685 in
1997 I have not used the modern one
which is available right which is even
more fancy so first one was hand powered
Bellow cleaner okay it was just like a
bleno cleaner largely made up of wood
and leather in those days metals and
polymers were not widely used Metals
could have been used but they didn't use
it was wood basically which was
naturally available so you can see till
about last century we have using wood
naturally available material for various
kinds of instruments and then in case of
1950s vacuum clearer become very
sophisticated we have used quite a lot
of polymeric material you can see here
these tubes and these things and also
metallic and then came in 1985 which are
very lightweight because this metallic
plot which is used here was Heavy in
1985 it are using lightweight vacuum
cleaners made of mostly polymers little
bit of metal like aluminum
and last one is a centrifugal dust
extraction cleaner of
1997 today's V are much Sleek much
sophisticated some of them actually are
computer controll you may have heard of
room cleaners which are AI driven so
they are the modern avatars of this if I
put together all the data which is again
taken from this book of Michael asby in
1900s oods canvas leather was used power
was very small 100 only 50 wats because
it's Hand p driven weight was massive 10
kgs and cost was something about
$250 but in 1950 cylinders were used but
these are all not electrical plow
powered and they made of M Steels which
we can get get up to 300 wats of power
weights have been almost Haled and cost
has also reduced 1985 again it is
lightweight that's why you know molden
Plastics and polyline was used power
increased power means sucking power
actually so that's about 800 wats and
weight was further reduced but then
1995 they are actually made of Dyson I
you can see the polythylene
polycarbonate they are used power was
again increased so 5050 wats become 1200
wats in about 90 years that's what is
the weights reduced
substantially and you know this cost are
basically obviously taken cost 1998 cost
might have changed now so do not so the
use choice of material depends on time
and design of the product how you want
another important aspect you could have
seiz aesthetic of the products is also
very important when you design something
so therefore these design
parameters are basically was used
manually to get this kind of products so
now this one I showed you right the case
study in the last class a device is to
open a cocked bottle if you want to take
out a liquid from the cor bottle you
have to open the Cork and what did I
show you okay don't use this these are
accidents either you can break it
obviously any bottle can be broken by
using some kind of a by hitting it or
you can actually have explosion on the
top side that also possible or you can
put some acid it will also corot the
cork but I doubt many acids do not
corrode the cork so best way is to
remove this cork by putting a screw
inside it or a blade inside it that's
what is shown here you can see here so
selection of the device selection of the
material for the device changes if you
want to use a screw or something you see
that that's why so the screw remains
always metallic you can't change it
because it has to be hard about these
other things whether you're using a
spring assistance pull or a g de liever
or a direct pull or even other kinds of
things that will dictate what kind of
material you need to choose so design of
this kind of devices requires selection
of materials what kind of M use that's
what I'm talking about it so these are
the four things I showed you please go
through this literature his book is
available you can read it through very
interesting one and design is very
important this one also I showed you
again taken from Michael as's book this
is basically heat SN and I showed you
why we have to use metals or polymers
mostly Metals for heat sink uh material
so that it can stand about 12 about
1200° sorry 200° C temperature well I'll
talk about another new example now you
have seen probably high tension cables
right because that is what is used to
hang high power High current capacity
high voltage cables as you know you must
have seen in various
Villages or even near the cities so have
you ever think think about it what kind
of materials will be used for over
transmission lines obviously you have a
limited choice because you have to have
a material which can conduct electricity
uh that means it should have very low
resistivity or rather very high
conductivity so but you know you always
like to increase this span between the
two high tension to the structures and
they are very expensive to make the
steel structures to be fixed on the
ground so they're very expensive to make
also they're going to occupy some space
on the ground so therefore idea is to
create as long span as possible so
constant is the length specified but if
you have a long span the weight of the
wire will be sufficiently high to allow
it to Sag or allow it to bend only
weight it will fall and it might be very
dangerous because if it sag quite low
some person may get affected by this so
that is what it is and obviously the
material should be strong that means it
should have a strength about more than
80 MPS M sorry megap pascals so that's
limits your choices so therefore you
have to have a minimum electric
resistivity you have to have a span well
defined and you have to have a material
whose strength is more than 8 TMP these
conditions are imposed what choices you
have so if you put it together
resistivity this is the resistivity
conductivity plots you can see here
thermal conductivity resistivity so
resistivity should be as slow as
possible right so resistivity should be
something about less than this
value right am I correct and your
choices are very limited there must be
Metals only because strength is also
very high that's why I use copper wires
okay copper is the best choice out of
all these is you can use aluminium also
but then aluminium has a problem of the
impurities so therefore conductivity
will
be compromised that's why High peric
opar virus are used actually you must be
knowing that in the electrical
applications electrical wires they're
mostly made of copper you can't replace
it because of its high electrical and
thermal conductivity it can conduct
electricity very nicely and if at all
there's a resistive heating it can also
take a heat can be delete taken away so
that's why they are choices so after
giving all this example which you can
always read more about from Michael as's
book how design of materials can be done
or design of objects can be done and
based on the design how the selection of
material can be done so what is the
problem statement finding a new material
with Target properties is of tremendous
importance for technological you know
development in the fields like energy
National Security resilience ures human
welfare everything or rather we talk
everything in this world that finding
new materials with with the target
properties is the main concern if you
want to send a Rover on a martian soil
or if you want to show send an uh you
know kind of a Rob on moon surface what
kind of Target properties are required
and based on the target properties can
we design new materials
so classical approach of medial
Discovery is basically primarily based
on trial and error and it depends on
what large number of costly and
time-taking experiments the moment
you're going to try and find out you're
going to make lot of mistakes and learn
from the mistakes and design it that's
going to be very costly and timec
consuming and in this stud's world you
don't have that much of time available
to you normally let me tell you the
material cycle from Discovery to
application is about 25 years and that's
very long according to many countries so
hence most of the technological experts
they want to reduce this cycle time
almost half that means from 25 years to
12.5 years or at Max 15 years so how do
you reduce it that means you have to
reduce number of Trials and experiments
that's why the AI and ml comes into
picture can you design a material using
this concept so that number of Trials
are reduced cost is saved that's the
question we are going to ask but you
know what is the best St and err
approach you can think about it is
basically what is known as edisonian
approach you must have heard of Edison
SAS ala Edition that's his full name
okay you can get it all this information
from Wikipedia link I provided at the
bottom so thas Al Edition was one of the
finest scientists world has ever
produced he has largest number of
patents he discovered electrical bulp
first time something around 1895 or 96 I
can remember that and that was a game
changer and these are the different
pictures of the trials he
made as you can see it is written here
Edition and his team tested more than
3,000 designs of bulbs
between 1878 and 1880 that means between
within 2 years they worked so hard they
tested 3,000 different types of designs
that is what is called best example I
found of trial and error I don't think
we can do it easily 3,000 different
types of design make the bulb test it oh
my God that's a massive job so Edison
has a large team of people working with
him and you could have you read this
Wikipedia you learn about more about him
so this kind of approach was good in
those time 1800 1900 because there was
no computers in those days that's why it
was very famine and this approach is
characterized by trial and Discovery
other than a systematic theoretical
approach that is what is you know known
today's world that is exactly what is
known is today's world but you know Ed
approach is never always dealt with the
problem of failure there will be lot of
failures and you're going to learn from
failures that's why it is not
recommended nowadays but you know then
how to discover materials one of the
ways is intuition based Discovery it
intuition means it's nothing based on
any strong scientific foundations it can
be synthesis of material or synthesis
preparation it can be some test you do
like mechanical test or test of any
functional properties elal conductivity
thermal conductivity or you can test it
under certain conditions and find out
based on that either you can choose the
correct one or you can discard the wrong
one okay that's intuition based that's
not trial and error tal error has no
intuition you simply have certain kind
of a design in your mind make it and
test it it doesn't work again change the
design this doesn't work again you
change something else here that's not
the case it's basically intuition based
it's more like a you know you think over
a lot and then got some idea which is
prob to work and many cases this
actually works because our intentions
are always you know happens to be quite
useful but you know there's something
called
Serendipity I don't know whether you
know it or not Serendipity is very very
important role in our material science
engineering why do Material Science
every every aspects okay so before I
come back to this I just want to give
you this one this is very important I've
taken from the from one of the best
scientist in the history of humankind
sir Alexander Fleming you know that he
discovered penicillin the first
antibiotic in the world in
1928 and he wrote this on his after his
Discovery he gave this lecture when he
received his Nobel Prize when I wake up
I just saw I when I woke up just after
Dawn on September 8 28
1928 look at it September 28 1928 the
date is very interesting I certainly did
not plan to ruize all medicines by
discovering the world first antibiotic
call or bacteria clear but I suppose
that was exactly what I did this is a
picture of the petrius on which you
doing experiment you can see this
actually stapy cocas bacteria the small
small dots and this is a fungus which is
growing and near the fungus there's no
very less number of bacteria most of
them are dead because the fungus was
producing penicillin so he separated the
penicillin that's why he discovered the
first antibiotic so that also a Serenity
and you know how it happened well he was
doing experiments for a long time and
one fine day or even evening he was very
frustrated he basically left this
petridis along with all this thing
outside his room it was a cold day
September is basically cold month in UK
and he thought everything is over is
nothing will happen but next day morning
or something
different so it's very clear SAS
discoveries happen when you know the
person concerned works hard to achieve
something that's what is what is said in
Material Science sedia Discovery happens
in various place this paper is very
important paper published in Indian
Journal bluo Material Science by one of
our classic scientist Professor Rober W
Khan see name of the paper is challenges
challenge response and Serendipity in
design of material what all has been
cipus the most important cipus material
you know what is aluminium copper alloy
what is known as a duraline
and it was discovered in Germany by
Alfred Wim okay his name is Alfred wi
Alfred what he was trying to do I let me
explain Alfred head of a newly
established mological section in a
German Government research laboratory in
Berlin in
1906 almost 120 years ago he did not
publish anything for almost an year
that's what Professor Khan is saying
this is from his paper now what happened
he was working on a commercial alloy or
basically alloy contained 48% copper 58%
magnesium and
58% manganese along with aluminium in
the cast and Ro conditions and he
examined its indentation harness on a
Saturday and went off for a lierly
weekend yes he wrote Metalist basically
has lizer in this far of days and on
returning Monday next Monday he repeated
the
measurements and just to be sure the
alloy was now 40% harder before what it
was he knew better to Shug off this
finding as a result of mere careless
measurements he repeated the experiment
again that's what happens for the best
of the best people and determine the
first age hardening allo what happened
discover yourself
pus but not sufficiently so because
there was no basis available in finding
any AAL allo in those
days but you know it was discovered in
1914 and it was used in 1918 to make
aircrafts in the first world war that's
what is the so therefore it was used by
US services in 1918 and did not knew
this aspects so therefore this is an
serendipia disc Discovery Marite is a
send deput
Discovery I can't keep on listenting
them quasi is a CP
Discovery Mar discover by Rudolph Mars
that's why discovered by Danny shman and
similarly metallic
Glass by Paul du so all there are many
actually I I don't want to list all of
them it will be very largest in material
sence 7p discover is very high let's
talk about this one which we left behind
you know this is basically glass ceramic
there ceramic material in which you have
glassy phase along with some
nanocrystals discovered by Dr scoot in
1953 and how it discovered
furnace was not working properly that's
how it discovered so it has been told to
me and I was looking at literature from
various literature I was trying to find
out what is it you know this person Dr
Donald skutki is actually did phg from
MIT and then he was working as a glass
research chemist and some company and
then what happened he was doing
experiments and one fine day to furnace
malfunctions a malfunction and then he
discovered that there are some crystal
in the glass Ceramics in the glass phas
of ceramic that's saw the glass ceramic
was discovered if you're lucky you can
discover this also there are many things
you can discover but obviously C
Discovery cannot happen every day it
happens sometimes there are few reports
so as you can understand this is another
way of discovering material the first
one was obviously trial and error then
intuition based then sendia Discovery
what else possible in this world can you
think about it well let's talk a little
bit more on this aspect in the next
lecture I'll start from these aspects
thank you
[Music]
[Music]
[Music]
