WEBVTT

00:00:00.000 --> 00:00:01.991
[SQUEAKING]

00:00:01.991 --> 00:00:03.979
[RUSTLING]

00:00:03.979 --> 00:00:05.967
[CLICKING]

00:00:05.967 --> 00:00:17.700


00:00:17.700 --> 00:00:18.600
PROFESSOR: So OK.

00:00:18.600 --> 00:00:20.720
So in this lecture
here, we are going

00:00:20.720 --> 00:00:23.240
to continue our discussion
about, essentially, glass

00:00:23.240 --> 00:00:24.560
viscosity.

00:00:24.560 --> 00:00:26.680
I'm starting out with
showing you this slide here

00:00:26.680 --> 00:00:29.080
that shows the different
viscosity reference

00:00:29.080 --> 00:00:31.160
points that are
of importance when

00:00:31.160 --> 00:00:33.960
it comes to glass processing.

00:00:33.960 --> 00:00:38.560
So horizontal axis is basically
the viscosity in log scale.

00:00:38.560 --> 00:00:40.920
And there are a few
important viscosity points

00:00:40.920 --> 00:00:42.680
that I want to highlight here.

00:00:42.680 --> 00:00:44.920
First one is being
this working point.

00:00:44.920 --> 00:00:49.120
It corresponds to a viscosity
of 10 to the 3 Pascal-second.

00:00:49.120 --> 00:00:51.520
So below the working
point, the glass

00:00:51.520 --> 00:00:54.420
basically flows just
like a regular fluid,

00:00:54.420 --> 00:00:55.685
so like water-like.

00:00:55.685 --> 00:00:57.560
So it becomes very
difficult to actually hold

00:00:57.560 --> 00:01:00.200
any shape for a
given period of time.

00:01:00.200 --> 00:01:03.200
And that's why below that it's
pretty much just considered

00:01:03.200 --> 00:01:04.700
like a fluid.

00:01:04.700 --> 00:01:06.520
It's a super cool liquid.

00:01:06.520 --> 00:01:10.280
So you need to have a
sufficiently high viscosity

00:01:10.280 --> 00:01:12.140
to enable any kind
of glass processing.

00:01:12.140 --> 00:01:14.240
So that's what the
working point means.

00:01:14.240 --> 00:01:16.440
The viscosity has to be
higher than the working

00:01:16.440 --> 00:01:19.720
point for actual glass shaping.

00:01:19.720 --> 00:01:23.320
Then the next part is
called softening point.

00:01:23.320 --> 00:01:25.920
So if you go above this
softening point, which

00:01:25.920 --> 00:01:31.040
corresponds to 10 to the 6.6
Pascal-second viscosity value,

00:01:31.040 --> 00:01:33.840
then the glass melts,
becomes so stiff

00:01:33.840 --> 00:01:37.400
that it becomes difficult to
actually start to deform it.

00:01:37.400 --> 00:01:41.100
So between this working
point and softening point,

00:01:41.100 --> 00:01:46.080
this is what's defined as the
working range for glass melt. So

00:01:46.080 --> 00:01:48.080
just to give you an
intuitive example about what

00:01:48.080 --> 00:01:50.680
the viscosity is like, I
have actually marked here

00:01:50.680 --> 00:01:54.580
this little arrow right in the
middle of the working range.

00:01:54.580 --> 00:01:57.960
This is the typical
viscosity of chewing gum.

00:01:57.960 --> 00:02:00.240
So intuitively, the
glass will behave

00:02:00.240 --> 00:02:03.040
like a chewing gum within
that viscosity range,

00:02:03.040 --> 00:02:05.278
and you can easily essentially
shape it, change it

00:02:05.278 --> 00:02:07.320
to any kind of geometry,
and you'll hold it there

00:02:07.320 --> 00:02:10.039
for some period of time that
allows you to essentially

00:02:10.039 --> 00:02:11.900
sustain the shape you want.

00:02:11.900 --> 00:02:14.580
And once you cool it down,
the viscosity increases,

00:02:14.580 --> 00:02:17.880
then you're sort of freezing
the geometry you need.

00:02:17.880 --> 00:02:19.265
So I have listed here--

00:02:19.265 --> 00:02:20.640
in the next lecture,
we are going

00:02:20.640 --> 00:02:22.160
to talk about different
glass-shaping techniques,

00:02:22.160 --> 00:02:22.990
specifically.

00:02:22.990 --> 00:02:25.240
But you can see there are a
lot of these glass-shaping

00:02:25.240 --> 00:02:27.560
techniques, they all have
a viscosity value that

00:02:27.560 --> 00:02:29.220
falls within this window here.

00:02:29.220 --> 00:02:31.872


00:02:31.872 --> 00:02:35.440
The next reference point
is annealing point.

00:02:35.440 --> 00:02:36.960
So annealing point
corresponds to 10

00:02:36.960 --> 00:02:39.380
to the 12 Pascal-second
viscosity,

00:02:39.380 --> 00:02:42.900
which basically coincides
with the definition of Tg.

00:02:42.900 --> 00:02:46.560
So annealing point is basically
glass transition temperature.

00:02:46.560 --> 00:02:50.920
Annealing point is where, once
the glass product is made,

00:02:50.920 --> 00:02:54.800
this is the temperature where
you do a post-shaping annealing

00:02:54.800 --> 00:02:58.230
to release any internal stress.

00:02:58.230 --> 00:02:58.910
OK.

00:02:58.910 --> 00:03:01.950
So at the annealing
point, microscopic changes

00:03:01.950 --> 00:03:03.050
can still take place.

00:03:03.050 --> 00:03:05.790
You can relax any internal
stress and strain,

00:03:05.790 --> 00:03:10.030
but macroscopic shape change
no longer takes place.

00:03:10.030 --> 00:03:11.870
Because it requires
low viscosity

00:03:11.870 --> 00:03:14.820
value to induce
macroscopic movement.

00:03:14.820 --> 00:03:17.070
So that's basically the
temperature where you actually

00:03:17.070 --> 00:03:19.550
anneal the glass products
without distorting the shape

00:03:19.550 --> 00:03:23.030
while ensuring that they don't
have any kind of internal stress

00:03:23.030 --> 00:03:23.742
there.

00:03:23.742 --> 00:03:25.450
Now, if you look at
this viscosity value,

00:03:25.450 --> 00:03:29.240
by the way, for pitch, which
is a very viscous fluid,

00:03:29.240 --> 00:03:30.990
you see this viscosity,
you can measure it

00:03:30.990 --> 00:03:33.282
from the pitch drop experiment,
it's of the order of 10

00:03:33.282 --> 00:03:33.950
to the 8.

00:03:33.950 --> 00:03:37.130
So 10 to the 12 is
very high viscosity.

00:03:37.130 --> 00:03:39.670
That's why you don't see any
kind of macroscopic deformation

00:03:39.670 --> 00:03:41.350
there.

00:03:41.350 --> 00:03:47.990
Now, finally, when the glass
exits the annealing furnace,

00:03:47.990 --> 00:03:51.110
its viscosity will be higher
than the straining point.

00:03:51.110 --> 00:03:54.310
So straining point is the
viscosity value where,

00:03:54.310 --> 00:03:58.270
above which you don't even have
microstructural relaxation.

00:03:58.270 --> 00:04:00.990
So annealing essentially takes
place between the annealing

00:04:00.990 --> 00:04:02.290
point and straining point.

00:04:02.290 --> 00:04:06.690
This is where you can have
a strain-stress relaxation.

00:04:06.690 --> 00:04:09.710
So we'll see this
reference chart again

00:04:09.710 --> 00:04:13.042
when we actually talk about
glass processing, in fact,

00:04:13.042 --> 00:04:15.310
in this lecture, the later
part of this lecture.

00:04:15.310 --> 00:04:17.510
But these are the
important reference points

00:04:17.510 --> 00:04:22.989
that's important to know
in glass processing.

00:04:22.989 --> 00:04:27.630
So to control the viscosity,
we have a very simple knob

00:04:27.630 --> 00:04:28.850
here for glass materials.

00:04:28.850 --> 00:04:30.790
You can simply just
change temperature.

00:04:30.790 --> 00:04:34.770
From the previous slide, when we
talked about the Angell's plot,

00:04:34.770 --> 00:04:36.612
we know that the
viscosity changes

00:04:36.612 --> 00:04:39.070
by many orders of magnitude
because its thermally activated

00:04:39.070 --> 00:04:40.510
phenomena.

00:04:40.510 --> 00:04:42.290
So yeah.

00:04:42.290 --> 00:04:44.430
So then it gives
us a very easy way

00:04:44.430 --> 00:04:46.430
to essentially tune
viscosity and push

00:04:46.430 --> 00:04:48.730
to the desired value for
specific processes there.

00:04:48.730 --> 00:04:51.150
So then, depending
on the specific kind

00:04:51.150 --> 00:04:53.330
of amorphous material
you're looking at here,

00:04:53.330 --> 00:04:55.310
they can exhibit different
kinds of behavior

00:04:55.310 --> 00:04:56.710
in terms of viscosity there.

00:04:56.710 --> 00:04:58.270
Specifically polymers.

00:04:58.270 --> 00:05:03.025
So for amorphous polymers
or-- lots of polymers,

00:05:03.025 --> 00:05:04.650
they are actually
not purely amorphous.

00:05:04.650 --> 00:05:07.870
They are a mixture of amorphous
phase and crystalline phase.

00:05:07.870 --> 00:05:10.910
But either way, as long as it
contains amorphous phase, then

00:05:10.910 --> 00:05:14.310
from a rheological perspective,
they all behave pretty similar.

00:05:14.310 --> 00:05:17.950
They exhibit this kind of a
temperature-dependent viscosity

00:05:17.950 --> 00:05:20.990
curve, like what I'm showing
on the left-hand side here.

00:05:20.990 --> 00:05:24.110
At low temperature,
they basically

00:05:24.110 --> 00:05:25.870
behave like hard plastic.

00:05:25.870 --> 00:05:26.950
They don't easily deform.

00:05:26.950 --> 00:05:29.670
It's rigid and stiff.

00:05:29.670 --> 00:05:32.070
As you start to heat it up,
you go into, essentially,

00:05:32.070 --> 00:05:35.910
this sort of rubbery state.

00:05:35.910 --> 00:05:39.850
This is beyond, basically, the
glass transition temperature.

00:05:39.850 --> 00:05:44.230
The molecular chains, they
start to have enough energy

00:05:44.230 --> 00:05:45.272
to move around.

00:05:45.272 --> 00:05:46.730
They can change
their conformation,

00:05:46.730 --> 00:05:48.270
they can move past each other.

00:05:48.270 --> 00:05:51.390
So then they go into a state
that behaves more like rubber.

00:05:51.390 --> 00:05:55.150
So that's a rubbery state
in the polymer language,

00:05:55.150 --> 00:05:56.590
typically what people call it.

00:05:56.590 --> 00:05:58.670
And then if you
continue to heat it up,

00:05:58.670 --> 00:06:01.890
eventually it becomes a
liquid, glass-forming liquid.

00:06:01.890 --> 00:06:06.730


00:06:06.730 --> 00:06:12.430
So you can engineer this
kind of viscosity behavior

00:06:12.430 --> 00:06:14.330
like changing microstructure
of the polymers.

00:06:14.330 --> 00:06:19.910
So for example, if you want to
change the Tg in polymers, so

00:06:19.910 --> 00:06:23.950
that you change the
softening temperature,

00:06:23.950 --> 00:06:26.810
you can add some stiff
chemical groups like benzene.

00:06:26.810 --> 00:06:28.650
Benzene is a ring.

00:06:28.650 --> 00:06:29.570
It's a planar ring.

00:06:29.570 --> 00:06:30.703
It doesn't easily deform.

00:06:30.703 --> 00:06:32.870
So if you add benzenes into
the polymer chain there,

00:06:32.870 --> 00:06:34.870
it increases the stiffness
of the polymer chain.

00:06:34.870 --> 00:06:37.410
It becomes much more
difficult to deform it.

00:06:37.410 --> 00:06:39.910
And as a result, the
viscosity is going to rise.

00:06:39.910 --> 00:06:42.350
And because Tg is defined
with respect to viscosity,

00:06:42.350 --> 00:06:45.230
essentially, that means
you also raise the Tg.

00:06:45.230 --> 00:06:49.070
So that's one way to
essentially change the Tg.

00:06:49.070 --> 00:06:51.310
If you want to decrease
Tg, on the other hand, what

00:06:51.310 --> 00:06:52.870
you can do--
typically, people put

00:06:52.870 --> 00:06:54.910
in what's called a plasticizer.

00:06:54.910 --> 00:06:56.790
Plasticizer, these
are essentially

00:06:56.790 --> 00:06:58.990
small molecule solvents.

00:06:58.990 --> 00:07:02.110
They go into the
polymer network and they

00:07:02.110 --> 00:07:06.350
serve as molecular lubricants.

00:07:06.350 --> 00:07:09.350
They decrease this internal
friction between the polymer

00:07:09.350 --> 00:07:12.190
chains, so that it becomes
a lot easier to move around.

00:07:12.190 --> 00:07:14.870
And the result is
that you can actually

00:07:14.870 --> 00:07:17.828
decrease Tg in this case,
which could be useful

00:07:17.828 --> 00:07:19.870
if you're trying to do
some processing at greatly

00:07:19.870 --> 00:07:22.110
reduced temperatures.

00:07:22.110 --> 00:07:26.390
One example of this kind of
a [INAUDIBLE] low-Tg polymer

00:07:26.390 --> 00:07:29.850
is this kind of polymer there.

00:07:29.850 --> 00:07:31.350
You can actually
buy it from Amazon.

00:07:31.350 --> 00:07:33.710
It has a glass
transition temperature

00:07:33.710 --> 00:07:36.710
that's engineered to be
around 70 degrees Celsius.

00:07:36.710 --> 00:07:40.510
So if you put it into
hot water, it softens.

00:07:40.510 --> 00:07:43.390
It becomes something you
can mold by your hand

00:07:43.390 --> 00:07:44.930
to any kind of geometry.

00:07:44.930 --> 00:07:46.930
But when it cools down
back to room temperature,

00:07:46.930 --> 00:07:48.510
it becomes stiff again.

00:07:48.510 --> 00:07:50.410
This you can search
for on Amazon.

00:07:50.410 --> 00:07:52.315
This is called foam cards.

00:07:52.315 --> 00:07:54.190
So this would be quite
handy if you're trying

00:07:54.190 --> 00:07:55.610
to just make something quickly.

00:07:55.610 --> 00:07:58.350
And again, this all goes
into how you actually

00:07:58.350 --> 00:08:01.070
engineer the polymer Tg, so that
you can change its softening

00:08:01.070 --> 00:08:02.350
behavior in this case.

00:08:02.350 --> 00:08:05.910
So that's for organic polymers.

00:08:05.910 --> 00:08:09.030
Now, for inorganic
glasses, in this slide

00:08:09.030 --> 00:08:11.510
here, I'm summarizing the
temperature dependence

00:08:11.510 --> 00:08:13.630
of viscosity for many
different compositions

00:08:13.630 --> 00:08:15.050
of inorganic glasses.

00:08:15.050 --> 00:08:18.430
The general rule of
thumb is that viscosity

00:08:18.430 --> 00:08:20.390
is a parameter
that links closely

00:08:20.390 --> 00:08:21.970
with the network connectivity.

00:08:21.970 --> 00:08:25.070


00:08:25.070 --> 00:08:28.310
And intuitively, it
makes a lot of sense

00:08:28.310 --> 00:08:30.470
if you have a network
that's connected together

00:08:30.470 --> 00:08:32.590
by many rigid
chemical bonds that

00:08:32.590 --> 00:08:34.289
is very difficult to deform it.

00:08:34.289 --> 00:08:36.350
And therefore, you should
have a high viscosity.

00:08:36.350 --> 00:08:39.429
So typically, high
network connectivity

00:08:39.429 --> 00:08:45.230
would correspond to high
viscosity and hence high Tg.

00:08:45.230 --> 00:08:47.310
So if you look at
this plot here,

00:08:47.310 --> 00:08:55.820
on the upper right
corner, this orange curve,

00:08:55.820 --> 00:09:00.940
this corresponds to a
silica glass, just SiO2.

00:09:00.940 --> 00:09:06.520
Now, you see, for SiO2,
its working range, in fact,

00:09:06.520 --> 00:09:07.820
is extremely high.

00:09:07.820 --> 00:09:10.420
It's beyond this chart
here, its working range.

00:09:10.420 --> 00:09:13.740
So that's part of the
reason why silica glass

00:09:13.740 --> 00:09:16.280
products is much more rare.

00:09:16.280 --> 00:09:17.920
It's actually a
lot more expensive.

00:09:17.920 --> 00:09:20.620
Typical glass workshops, they
cannot work with silica glass,

00:09:20.620 --> 00:09:23.078
because their furnaces are not
at a high enough temperature

00:09:23.078 --> 00:09:25.860
to just soften silica glass so
they can process it in different

00:09:25.860 --> 00:09:27.220
shapes.

00:09:27.220 --> 00:09:30.027
Now, the next group of glasses
we're going to look at here,

00:09:30.027 --> 00:09:31.860
look at these two curves
here, this blue one

00:09:31.860 --> 00:09:33.300
and this green one.

00:09:33.300 --> 00:09:35.700
These are essentially
borosilicate glasses.

00:09:35.700 --> 00:09:37.100
So you notice that
they still lie

00:09:37.100 --> 00:09:38.660
above the conventional
soda-lime.

00:09:38.660 --> 00:09:41.160
This black line is a typical
soda-lime-- windowpane glasses.

00:09:41.160 --> 00:09:43.280
So we have already
talked about it.

00:09:43.280 --> 00:09:46.600
When you add boron
into the glass network,

00:09:46.600 --> 00:09:48.855
you can trigger this
P3 to P4 conversion,

00:09:48.855 --> 00:09:50.980
and therefore, you essentially
increase the network

00:09:50.980 --> 00:09:51.860
connectivity.

00:09:51.860 --> 00:09:54.980
And that's why
borosilicate glasses,

00:09:54.980 --> 00:09:57.940
they tend to have a
higher viscosity compared

00:09:57.940 --> 00:09:59.233
to soda-lime glasses.

00:09:59.233 --> 00:10:01.400
In other words, if you look
at all the temperatures,

00:10:01.400 --> 00:10:03.180
all the viscosity
reference points--

00:10:03.180 --> 00:10:08.060
working range, softening point--

00:10:08.060 --> 00:10:12.020
borosilicate glasses tend to
have much higher temperatures

00:10:12.020 --> 00:10:13.560
compared to soda-lime glasses.

00:10:13.560 --> 00:10:17.180


00:10:17.180 --> 00:10:20.660
And that's why, in fact, it's
more challenging for glass

00:10:20.660 --> 00:10:23.020
workshops to actually
process borosilicate glasses.

00:10:23.020 --> 00:10:26.780
And then finally, you see that
if you add a lot of alkaline,

00:10:26.780 --> 00:10:28.380
then this is the
way you actually

00:10:28.380 --> 00:10:30.255
use to actually decrease
network connectivity

00:10:30.255 --> 00:10:32.660
and hence the glass
processing temperatures.

00:10:32.660 --> 00:10:35.660


00:10:35.660 --> 00:10:37.220
Now, by adding
alkaline, as I said,

00:10:37.220 --> 00:10:40.460
you can decrease the network
connectivity and hence

00:10:40.460 --> 00:10:41.580
the viscosity.

00:10:41.580 --> 00:10:43.900
One very important fact I
also want to mention here

00:10:43.900 --> 00:10:45.900
is that if you-- instead
of adding just one

00:10:45.900 --> 00:10:48.820
kind of alkaline atom-- if
you add two types or more

00:10:48.820 --> 00:10:51.620
than two types of alkali
atoms simultaneously,

00:10:51.620 --> 00:10:55.700
the depression of the viscosity
becomes even more significant.

00:10:55.700 --> 00:10:59.660
So this is known as the
Mixed Alkali Effect.

00:10:59.660 --> 00:11:04.980
So to illustrate this effect,
what I'm showing on this plot

00:11:04.980 --> 00:11:07.260
here on the left-hand
side, vertical axis

00:11:07.260 --> 00:11:09.540
is basically a
logarithm viscosity

00:11:09.540 --> 00:11:13.240
of some silicate glasses
mixed with either sodium oxide

00:11:13.240 --> 00:11:14.600
and potassium oxide.

00:11:14.600 --> 00:11:20.060
Horizontal axis, this is the
molar fraction of sodium oxide.

00:11:20.060 --> 00:11:23.000
So you keep the total amount
of alkaline ion constant,

00:11:23.000 --> 00:11:25.340
you are essentially changing
the relative fraction,

00:11:25.340 --> 00:11:29.740
the relative concentration
of sodium and potassium ions.

00:11:29.740 --> 00:11:32.020
Now, notice that this
curve, they always

00:11:32.020 --> 00:11:35.420
have this kind of
a concave shape.

00:11:35.420 --> 00:11:40.100
So the viscosity decreases by
a couple of orders of magnitude

00:11:40.100 --> 00:11:44.220
if you go into
intermediate composition.

00:11:44.220 --> 00:11:48.340
So basically says the glass
becomes a lot more deformable

00:11:48.340 --> 00:11:50.080
if you mix two kinds
of alkali ions,

00:11:50.080 --> 00:11:53.780
even if you keep the total
concentration the same.

00:11:53.780 --> 00:11:56.700
I want to mention that this
effect exists not only for just

00:11:56.700 --> 00:11:59.403
viscosity, it actually exists
for many physical properties

00:11:59.403 --> 00:12:01.820
of glass materials, especially
in terms of their transport

00:12:01.820 --> 00:12:02.360
properties.

00:12:02.360 --> 00:12:05.900
So only this middle
plot here, I'm

00:12:05.900 --> 00:12:08.420
now plotting diffusion
coefficient-- again, similarly

00:12:08.420 --> 00:12:11.820
as the composition of
these two alkali ion atoms.

00:12:11.820 --> 00:12:14.820
And you see similar kind
of a dip behavior there.

00:12:14.820 --> 00:12:17.700
You have the minimum diffusion
at some intermediate composition

00:12:17.700 --> 00:12:19.020
here.

00:12:19.020 --> 00:12:22.360
So the origin of the
mixed alkaline effect--

00:12:22.360 --> 00:12:26.513
this is still an issue still
of intensive investigation.

00:12:26.513 --> 00:12:28.180
People have come up
with different kinds

00:12:28.180 --> 00:12:30.638
of explanations with respect
to different specific physical

00:12:30.638 --> 00:12:31.820
effects.

00:12:31.820 --> 00:12:34.100
The most commonly
assumed reason is

00:12:34.100 --> 00:12:37.220
that this comes from mutual
blocking of transport

00:12:37.220 --> 00:12:38.580
pathways between ions.

00:12:38.580 --> 00:12:41.340
If you imagine, for example,
you have a glass network that

00:12:41.340 --> 00:12:44.230
contains, say, only one kind
of ion, let's say sodium,

00:12:44.230 --> 00:12:45.980
all the sodium ions,
they have essentially

00:12:45.980 --> 00:12:48.980
their respective size that they
can comfortably be sitting in.

00:12:48.980 --> 00:12:51.540
And when any kind of transport
phenomena like diffusion

00:12:51.540 --> 00:12:55.540
happens, what is required
is that the sodium ion,

00:12:55.540 --> 00:12:58.680
they have to hop from one sodium
side to another sodium side.

00:12:58.680 --> 00:13:03.100
That's how you perform ion
conduction or perform diffusion.

00:13:03.100 --> 00:13:04.700
Now, if you only
have sodium, then you

00:13:04.700 --> 00:13:06.783
can imagine that all the
sides, more or less, they

00:13:06.783 --> 00:13:08.940
have similar energy, so
it's quite easy for them

00:13:08.940 --> 00:13:11.260
to move around without
experiencing a large energy

00:13:11.260 --> 00:13:13.340
barrier.

00:13:13.340 --> 00:13:17.060
If you go to now
mix alkalines, you

00:13:17.060 --> 00:13:21.460
get two kind of sides
for this kind of ions.

00:13:21.460 --> 00:13:24.387
And this creates large
energy differences

00:13:24.387 --> 00:13:26.720
when they are hopping between
the different sides there.

00:13:26.720 --> 00:13:29.900
For example, if you're trying
to squeeze especially potassium,

00:13:29.900 --> 00:13:32.215
a larger ion, into smaller
ions like sodium sides,

00:13:32.215 --> 00:13:33.840
then there's a lot
of steric hindrance.

00:13:33.840 --> 00:13:35.280
And therefore, it
increases the energy

00:13:35.280 --> 00:13:36.900
you need to actually
move these ions.

00:13:36.900 --> 00:13:40.020
And in the end, that's why it
actually inhibits this transport

00:13:40.020 --> 00:13:40.740
behavior.

00:13:40.740 --> 00:13:43.170
And that's the
explanation that people

00:13:43.170 --> 00:13:47.950
are giving for this kind
of mixed alkaline effect.

00:13:47.950 --> 00:13:51.970
So I just want to mention that
this mixed alkali effect, this

00:13:51.970 --> 00:13:54.850
is something that people have
actually discovered long ago.

00:13:54.850 --> 00:14:00.050
So this is accounts of
some historic records that

00:14:00.050 --> 00:14:03.250
mention that, in the early days,
when people were actually making

00:14:03.250 --> 00:14:05.250
thermometers out of glass--

00:14:05.250 --> 00:14:07.490
so some early
thermometers, they put

00:14:07.490 --> 00:14:09.548
something like mercury
into a glass tube,

00:14:09.548 --> 00:14:11.090
and looking at the
thermal expansion,

00:14:11.090 --> 00:14:12.910
and use that to
determine temperature.

00:14:12.910 --> 00:14:17.490
Now, what they have noticed
here is that you have freshly

00:14:17.490 --> 00:14:19.010
made glass.

00:14:19.010 --> 00:14:21.250
You directly put
mercury in there.

00:14:21.250 --> 00:14:26.570
You then put it into,
for example, ice water,

00:14:26.570 --> 00:14:28.410
it would show 0
degrees C. But then

00:14:28.410 --> 00:14:30.850
if you put it into boiling
water and then put it back

00:14:30.850 --> 00:14:33.690
into ice water, instead
of showing 0 degrees C,

00:14:33.690 --> 00:14:35.450
it would show actually
negative 5 degrees

00:14:35.450 --> 00:14:38.530
C, because when you actually
put the glass in boiling water,

00:14:38.530 --> 00:14:42.550
because of high temperature and
then it decreases the viscosity,

00:14:42.550 --> 00:14:44.610
you start to get shape
deformation and relaxation

00:14:44.610 --> 00:14:45.750
of internal stress.

00:14:45.750 --> 00:14:49.050
As a result, this leads
to macroscopic deformation

00:14:49.050 --> 00:14:52.090
that then compromises the
accuracy of this thermometer.

00:14:52.090 --> 00:14:55.450
So they have noticed
there that if you actually

00:14:55.450 --> 00:15:00.610
store these glasses at
just an ambient condition

00:15:00.610 --> 00:15:04.110
for several months before you
make it into a thermometer,

00:15:04.110 --> 00:15:06.090
then this phenomena goes away.

00:15:06.090 --> 00:15:08.250
Now, that's actually
easy to explain,

00:15:08.250 --> 00:15:11.970
because now we understand this
is coming from relaxation.

00:15:11.970 --> 00:15:14.690
So if you allow all
the internal stress,

00:15:14.690 --> 00:15:17.410
shape deformation take place
in over the several months

00:15:17.410 --> 00:15:20.890
of period, then you no
longer have this effect.

00:15:20.890 --> 00:15:22.810
However, what they
have also noticed here

00:15:22.810 --> 00:15:26.290
is that if you heat it up
again and cool it down,

00:15:26.290 --> 00:15:28.330
then that phenomenon
comes in again.

00:15:28.330 --> 00:15:30.372
You are essentially going
through the same cycle.

00:15:30.372 --> 00:15:32.890
You heat it up, back to like
a super cool liquid, quenches

00:15:32.890 --> 00:15:36.092
back down, it erases any
memory of the relaxation.

00:15:36.092 --> 00:15:37.550
So you have the
same problem again.

00:15:37.550 --> 00:15:39.210
Relaxation comes in again.

00:15:39.210 --> 00:15:42.890
And finally, what
they have noticed

00:15:42.890 --> 00:15:44.668
there is that this
relaxation effect

00:15:44.668 --> 00:15:46.210
could be diminished
by a factor of 10

00:15:46.210 --> 00:15:48.710
if only one type of
alkaline ion is used

00:15:48.710 --> 00:15:50.610
rather than using mixed ions.

00:15:50.610 --> 00:15:52.890
So that's basically
the first account

00:15:52.890 --> 00:15:55.432
for this kind of
mixed alkaline effect.

00:15:55.432 --> 00:15:56.890
You're mixing two
kind of ions, you

00:15:56.890 --> 00:15:59.210
decrease the viscosity
and therefore

00:15:59.210 --> 00:16:03.010
the working temperatures.

00:16:03.010 --> 00:16:06.170
So far, we have been discussing
about the temperature

00:16:06.170 --> 00:16:09.890
dependence of viscosity
in amorphous materials.

00:16:09.890 --> 00:16:12.612
The next topic I want to
mention here is that viscosity,

00:16:12.612 --> 00:16:14.070
it not only depends
on temperature,

00:16:14.070 --> 00:16:16.370
it also depends on shear rate.

00:16:16.370 --> 00:16:17.210
OK.

00:16:17.210 --> 00:16:20.730
And this is quite common in
a lot of fluids, including

00:16:20.730 --> 00:16:23.520
glass-forming liquids.

00:16:23.520 --> 00:16:25.770
Now, if you look at the
equation we have been starting

00:16:25.770 --> 00:16:27.950
to talking about here
in the very first slide,

00:16:27.950 --> 00:16:32.130
we said that if you assume a
liquid has a Newtonian behavior,

00:16:32.130 --> 00:16:34.907
then its viscosity is constant.

00:16:34.907 --> 00:16:36.490
It does not depend
on the strain rate,

00:16:36.490 --> 00:16:38.970
and therefore the shear
stress should be proportional

00:16:38.970 --> 00:16:41.210
to the shear strain rate.

00:16:41.210 --> 00:16:45.250
In reality, you can
have different kind

00:16:45.250 --> 00:16:48.870
of a dependence of the
viscosity on the shear rate.

00:16:48.870 --> 00:16:51.250
So what I'm plotting
on this plot here

00:16:51.250 --> 00:16:55.090
is some of this kind of
typical fluids, when they

00:16:55.090 --> 00:16:56.650
exhibit different behaviors.

00:16:56.650 --> 00:16:59.090
So the Newtonian
fluid, this would

00:16:59.090 --> 00:17:03.970
be a straight line here for the
stress and strain rate curve.

00:17:03.970 --> 00:17:06.973
And basically, it says
that the slope of this line

00:17:06.973 --> 00:17:08.890
will give you the
viscosity, and it's constant

00:17:08.890 --> 00:17:10.369
regardless of strain rate.

00:17:10.369 --> 00:17:14.329
You can have shear thinning.

00:17:14.329 --> 00:17:17.977
As you increase the shear rate,
the liquid becomes thinner.

00:17:17.977 --> 00:17:20.310
It becomes easier to deform,
it becomes lower viscosity.

00:17:20.310 --> 00:17:21.630
It's more flowable.

00:17:21.630 --> 00:17:23.390
Or it can be shear thickening.

00:17:23.390 --> 00:17:26.630
As you increase strain rate,
it actually becomes thicker.

00:17:26.630 --> 00:17:27.970
It becomes more viscous.

00:17:27.970 --> 00:17:31.730
Now, you can also have a
so-called Bingham plastic.

00:17:31.730 --> 00:17:34.010
Bingham plastic has this
interesting behavior

00:17:34.010 --> 00:17:37.710
that if your stress is
below a certain threshold,

00:17:37.710 --> 00:17:39.810
it behaves more like a solid.

00:17:39.810 --> 00:17:42.490
It does not exhibit any
kind of flow behavior.

00:17:42.490 --> 00:17:44.150
Now, once you exceed
the threshold,

00:17:44.150 --> 00:17:47.850
now it starts to flow, now
it behaves like a liquid.

00:17:47.850 --> 00:17:51.430
In fact, the common example of
Bingham plastic is toothpaste.

00:17:51.430 --> 00:17:54.168
A lot of toothpaste is
actually Bingham fluid plastic.

00:17:54.168 --> 00:17:55.710
If you think about
toothpaste, right,

00:17:55.710 --> 00:17:59.890
if you just leave it there,
it will hold its shape.

00:17:59.890 --> 00:18:02.390
But once you start to brush,
of course, it starts to deform.

00:18:02.390 --> 00:18:04.265
That's when you actually
apply enough stress,

00:18:04.265 --> 00:18:08.430
exceed the threshold to
induce the viscous flow.

00:18:08.430 --> 00:18:08.930
OK.

00:18:08.930 --> 00:18:11.650


00:18:11.650 --> 00:18:13.408
For amorphous
materials, depending

00:18:13.408 --> 00:18:15.450
on what kind of material
you're looking at there,

00:18:15.450 --> 00:18:20.623
they can either exhibit
stress or shear thickening.

00:18:20.623 --> 00:18:22.290
And I'll give you
some specific examples

00:18:22.290 --> 00:18:26.405
about this different
kind of fluids there.

00:18:26.405 --> 00:18:28.530
And then on this next slide,
on the left-hand side,

00:18:28.530 --> 00:18:30.610
I'm basically now, in this case,
just plotting the viscosity

00:18:30.610 --> 00:18:32.105
as a function of shear rate.

00:18:32.105 --> 00:18:33.230
Same idea as the last plot.

00:18:33.230 --> 00:18:35.813
We are just now looking at the
slope from the last plot there.

00:18:35.813 --> 00:18:36.650
OK.

00:18:36.650 --> 00:18:41.250
Microscopically,
what's the mechanism

00:18:41.250 --> 00:18:43.650
for this different
kind of fluids?

00:18:43.650 --> 00:18:46.410
Viscosity, as we have
mentioned before,

00:18:46.410 --> 00:18:48.930
is internal friction in liquids.

00:18:48.930 --> 00:18:52.090
Viscous flow is when one part of
liquid, some of the molecules,

00:18:52.090 --> 00:18:54.010
has to move past the others.

00:18:54.010 --> 00:18:56.158
And viscosity is the
resistance of the material

00:18:56.158 --> 00:18:58.450
to this kind of relative
motion between different parts

00:18:58.450 --> 00:18:59.730
of molecules.

00:18:59.730 --> 00:19:04.490
So for shear
thickening, what happens

00:19:04.490 --> 00:19:07.010
there is that, as you
increase the strain rate,

00:19:07.010 --> 00:19:09.610
the molecules start to
self-assemble together

00:19:09.610 --> 00:19:11.810
to form larger clusters.

00:19:11.810 --> 00:19:13.730
And with larger
clusters, it becomes

00:19:13.730 --> 00:19:15.590
more difficult to
actually deform it.

00:19:15.590 --> 00:19:17.650
And as a result, you
see apparent increase

00:19:17.650 --> 00:19:20.090
of the viscosity.

00:19:20.090 --> 00:19:23.617
So this is a phenomenon
called hydroclustering.

00:19:23.617 --> 00:19:24.950
It's basically illustrated here.

00:19:24.950 --> 00:19:27.770
When there is no shear stress
or the molecular fragments

00:19:27.770 --> 00:19:29.630
stay separated.

00:19:29.630 --> 00:19:31.830
But once you start to
apply large shear rate,

00:19:31.830 --> 00:19:34.920
then they start to organize,
self-assemble together to form

00:19:34.920 --> 00:19:38.720
larger structures, which
impedes viscous flow.

00:19:38.720 --> 00:19:42.320
So lots of liquids actually
exhibit a shear thickening

00:19:42.320 --> 00:19:44.280
behavior.

00:19:44.280 --> 00:19:47.160
I'm showing you this
little video here.

00:19:47.160 --> 00:19:53.440
It basically shows someone
that's running on liquid.

00:19:53.440 --> 00:19:56.285
This liquid, this
is called oobleck.

00:19:56.285 --> 00:19:57.660
You can actually
make it at home.

00:19:57.660 --> 00:19:58.420
It's quite straightforward.

00:19:58.420 --> 00:19:59.940
It's basically starch and water.

00:19:59.940 --> 00:20:01.880
You mix it with about
one-to-two ratio,

00:20:01.880 --> 00:20:03.585
and then you get
liquid like this.

00:20:03.585 --> 00:20:05.960
It is quite remarkable because
if you do it, what happens

00:20:05.960 --> 00:20:09.360
is that if you just try to use
a spoon or something to screw it

00:20:09.360 --> 00:20:11.520
up, it behaves just like liquid.

00:20:11.520 --> 00:20:14.140
But if you try to change its
shape, you put it in your hand,

00:20:14.140 --> 00:20:15.960
try to mold it into
some shape, as long

00:20:15.960 --> 00:20:19.520
as you do it quick enough,
it behaves more like a solid.

00:20:19.520 --> 00:20:20.020
OK.

00:20:20.020 --> 00:20:22.602
So that's one good example of
this shear thickening behavior.

00:20:22.602 --> 00:20:24.560
And that's why you see
that people can actually

00:20:24.560 --> 00:20:25.980
run on this kind of liquid.

00:20:25.980 --> 00:20:28.320
As long as you do
it fast enough,

00:20:28.320 --> 00:20:30.930
you are not risking
sinking into the liquid.

00:20:30.930 --> 00:20:32.680
Actually, you can find
some YouTube videos

00:20:32.680 --> 00:20:34.860
where people actually play
soccer on this oobleck,

00:20:34.860 --> 00:20:37.440
like a big swimming pool
filled with oobleck.

00:20:37.440 --> 00:20:39.460
But you have to keep it
running all the time.

00:20:39.460 --> 00:20:45.240
As soon as you stop, pssh,
you sink into liquid.

00:20:45.240 --> 00:20:48.960
Some other examples.

00:20:48.960 --> 00:20:51.260
Any of you know Silly Putty?

00:20:51.260 --> 00:20:51.760
Oh, yeah.

00:20:51.760 --> 00:20:53.260
Silly Putty.

00:20:53.260 --> 00:20:54.960
This is similar.

00:20:54.960 --> 00:20:57.623
Silly Putty, you can
shape, mold into a ball.

00:20:57.623 --> 00:21:00.040
And if you throw it on the
floor, it actually bounce back.

00:21:00.040 --> 00:21:02.060
It's basically like
a rock, a solid.

00:21:02.060 --> 00:21:04.333
But if you try to let
it flow a little bit,

00:21:04.333 --> 00:21:06.000
you put it into a
container, it actually

00:21:06.000 --> 00:21:07.760
can flow and conform
to the shape.

00:21:07.760 --> 00:21:11.580
So as long as you
have a low stress,

00:21:11.580 --> 00:21:14.240
then its viscosity decreases
and then you allow it to flow.

00:21:14.240 --> 00:21:17.640
So these are the examples of
this kind of fluids there.

00:21:17.640 --> 00:21:19.400
Shear thickening.

00:21:19.400 --> 00:21:20.580
Shear thinning.

00:21:20.580 --> 00:21:23.560


00:21:23.560 --> 00:21:27.800
The common example of
shear thinning is paint--

00:21:27.800 --> 00:21:28.700
house paint.

00:21:28.700 --> 00:21:32.200


00:21:32.200 --> 00:21:34.720
So shear thinning is
quite desirable for paint

00:21:34.720 --> 00:21:36.602
because when you're
trying to brush it,

00:21:36.602 --> 00:21:38.060
you want to make
sure it's a fluid,

00:21:38.060 --> 00:21:40.160
you can actually apply
it uniformly on the wall.

00:21:40.160 --> 00:21:41.640
So you need low viscosity.

00:21:41.640 --> 00:21:44.260
But once you put it on the wall,
you want it to stick there.

00:21:44.260 --> 00:21:46.360
You don't want it
to continue to flow.

00:21:46.360 --> 00:21:48.500
So that's why people
actually engineer it

00:21:48.500 --> 00:21:50.880
so it has shear thinning
characteristics such

00:21:50.880 --> 00:21:53.720
that when you stop
applying any stress,

00:21:53.720 --> 00:21:55.340
its viscosity
actually increases.

00:21:55.340 --> 00:21:56.240
And that's why you
can actually have

00:21:56.240 --> 00:21:58.240
a uniform coating on the
wall without continuing

00:21:58.240 --> 00:22:00.100
to flow before it dries.

00:22:00.100 --> 00:22:02.600
Another more state-of-the-art
example where people are using

00:22:02.600 --> 00:22:05.240
this kind of shear thinning
characteristic to some benefits

00:22:05.240 --> 00:22:08.680
here is when people
are doing 3D printing.

00:22:08.680 --> 00:22:13.440
So 3D printing involves--
you have some polymer.

00:22:13.440 --> 00:22:16.280
You start to squeeze
it out of a nozzle.

00:22:16.280 --> 00:22:19.460
And as you translate
this nozzle in 3D space,

00:22:19.460 --> 00:22:23.760
you then leave behind a trace
of this polymer in 3D space.

00:22:23.760 --> 00:22:26.040
Now, shear thinning
is, again, advantageous

00:22:26.040 --> 00:22:28.960
here, because when you're
trying to squeeze polymer out

00:22:28.960 --> 00:22:32.480
of a nozzle, this is basically
a viscous flow process.

00:22:32.480 --> 00:22:34.340
If you have a very
high viscosity,

00:22:34.340 --> 00:22:37.160
it becomes very difficult
just to force the polymer out

00:22:37.160 --> 00:22:37.960
of the nozzle.

00:22:37.960 --> 00:22:40.360
And again, once you
are printed or you

00:22:40.360 --> 00:22:42.360
have finished your printing,
you want your shape

00:22:42.360 --> 00:22:43.520
to actually stay static.

00:22:43.520 --> 00:22:45.812
You don't want it to continue
to flow to deform things.

00:22:45.812 --> 00:22:48.520
So that's why, if you have
shear thinning fluids, then

00:22:48.520 --> 00:22:51.020
at the nozzle, this is where
you have very large shear rate,

00:22:51.020 --> 00:22:52.620
you have a decrease
of viscosity.

00:22:52.620 --> 00:22:54.620
But once it comes out,
you relax all the stress,

00:22:54.620 --> 00:22:57.600
you release all the stress,
then you can have high viscosity

00:22:57.600 --> 00:23:00.540
to hold the shape.

00:23:00.540 --> 00:23:04.520
So now coming back to glass.

00:23:04.520 --> 00:23:07.680
So what I'm plotting here, this
is actually for soda-lime glass.

00:23:07.680 --> 00:23:10.440
It plots how the
viscosity actually

00:23:10.440 --> 00:23:13.480
changes as a function
of the strain rate.

00:23:13.480 --> 00:23:15.440
And according to
this plot, you can

00:23:15.440 --> 00:23:18.560
see that soda-lime glass
actually exhibits a fairly clear

00:23:18.560 --> 00:23:21.840
shear thinning behavior.

00:23:21.840 --> 00:23:23.840
So viscosity of
soda-lime glasses

00:23:23.840 --> 00:23:26.220
decreases as you
increase the strain rate.

00:23:26.220 --> 00:23:29.480


00:23:29.480 --> 00:23:32.000
Now, why is this important?

00:23:32.000 --> 00:23:34.440
When it comes to
glass processing,

00:23:34.440 --> 00:23:37.680
especially in glass processes,
where, in the same process,

00:23:37.680 --> 00:23:40.720
you have large
differences of shear rate,

00:23:40.720 --> 00:23:44.160
then this is something that
has to be taken into account.

00:23:44.160 --> 00:23:47.920
One example there would
be an optical fiber join.

00:23:47.920 --> 00:23:50.680
So we will be talking
about optical fiber

00:23:50.680 --> 00:23:52.620
joining processes
later on in the class.

00:23:52.620 --> 00:23:56.200
But essentially, what you do is
you have a macroscopic preform,

00:23:56.200 --> 00:23:58.040
which is about a
couple of centimeters

00:23:58.040 --> 00:23:59.740
in diameter, a cylinder.

00:23:59.740 --> 00:24:01.760
You then put it
into a furnace, so

00:24:01.760 --> 00:24:03.700
that you start to
induce viscous flow,

00:24:03.700 --> 00:24:07.160
and you apply tensile
stress on one end, which

00:24:07.160 --> 00:24:13.780
then deforms this cylinder into
a very thin strand of the fiber.

00:24:13.780 --> 00:24:15.880
Now, if you look at
the process here,

00:24:15.880 --> 00:24:20.280
the deformation rate, or shear
rate, is very non-uniform.

00:24:20.280 --> 00:24:22.760
At this part, at this
macroscopic preform

00:24:22.760 --> 00:24:26.250
here, the shear rate is
actually very, very small.

00:24:26.250 --> 00:24:27.250
It barely deforms.

00:24:27.250 --> 00:24:29.590
But the shear rate
scales, actually,

00:24:29.590 --> 00:24:32.570
inversely with the
diameter of the glass.

00:24:32.570 --> 00:24:35.470
So when you come to
the fiber part here,

00:24:35.470 --> 00:24:37.910
you actually have a
very large shear rate

00:24:37.910 --> 00:24:40.190
during the formation process.

00:24:40.190 --> 00:24:41.710
So when you're
actually developing

00:24:41.710 --> 00:24:46.510
any kind of a modeling
process for this fiber drawing

00:24:46.510 --> 00:24:49.950
technique, you need to account
for the viscosity change.

00:24:49.950 --> 00:24:52.350
The actual viscosity--
apparent viscosity of the glass

00:24:52.350 --> 00:24:55.450
is much lower here
compared to up here,

00:24:55.450 --> 00:24:58.710
even if you are at
the same temperature.

00:24:58.710 --> 00:25:02.742
Otherwise, you may
risk breaking the fiber

00:25:02.742 --> 00:25:04.950
or having non-uniformity
there, if you don't properly

00:25:04.950 --> 00:25:07.070
account for this
kind of shear rate

00:25:07.070 --> 00:25:08.450
dependence of the viscosity.

00:25:08.450 --> 00:25:13.310


00:25:13.310 --> 00:25:18.292
So before I talk about
how to measure viscosity,

00:25:18.292 --> 00:25:19.750
I just want to show
you this slide.

00:25:19.750 --> 00:25:21.230
When I was looking
for literature

00:25:21.230 --> 00:25:26.470
on non-Newtonian fluids, people,
obviously, have identified cats

00:25:26.470 --> 00:25:28.590
as non-Newtonian fluids.

00:25:28.590 --> 00:25:30.030
They actually published a paper.

00:25:30.030 --> 00:25:33.630
They tried to fit all kinds
of fluid dynamic equations

00:25:33.630 --> 00:25:35.560
and used that to
describe cat behavior.

00:25:35.560 --> 00:25:36.810
And in the illustration here--

00:25:36.810 --> 00:25:38.435
I don't know if you
can see it clearly,

00:25:38.435 --> 00:25:40.590
but it shows a few examples
why they consider cat

00:25:40.590 --> 00:25:43.710
as being a fluid
instead of a solid.

00:25:43.710 --> 00:25:45.870
Obviously, they can conform
into different shapes.

00:25:45.870 --> 00:25:48.132
They can handle
uneven substrates.

00:25:48.132 --> 00:25:49.590
You can see there
they can actually

00:25:49.590 --> 00:25:52.550
have non-wetting behavior
on water surface.

00:25:52.550 --> 00:25:54.690
And this one, they can
glide on a smooth surface.

00:25:54.690 --> 00:25:56.690
They can also hang
onto vertical walls.

00:25:56.690 --> 00:25:59.990
So all those behaviors
are unique to liquids,

00:25:59.990 --> 00:26:01.050
not to solids.

00:26:01.050 --> 00:26:03.695
So that's why they conclude
that cat is actually a liquid.

00:26:03.695 --> 00:26:05.070
And they have some
equations that

00:26:05.070 --> 00:26:07.750
consider the relations to
actually describe cats.

00:26:07.750 --> 00:26:10.610
So quite an entertaining paper
if you ever want to read it.

00:26:10.610 --> 00:26:13.970
And this is also won the Ig
Nobel Prize, this research.

00:26:13.970 --> 00:26:18.350
AUDIENCE: So they're like
the author affiliations?

00:26:18.350 --> 00:26:19.310
PROFESSOR: Oh, yeah.

00:26:19.310 --> 00:26:23.890
AUDIENCE: [INAUDIBLE]
superior radiologist?

00:26:23.890 --> 00:26:27.170


00:26:27.170 --> 00:26:29.670
PROFESSOR: I was hoping they
should include some of the cats

00:26:29.670 --> 00:26:31.670
as coauthors.

00:26:31.670 --> 00:26:33.520
But they are important
experimental subjects

00:26:33.520 --> 00:26:34.770
that contributed to the study.

00:26:34.770 --> 00:26:35.530
But whatever.

00:26:35.530 --> 00:26:38.870


00:26:38.870 --> 00:26:40.750
Anyway.

00:26:40.750 --> 00:26:41.250
All right.

00:26:41.250 --> 00:26:44.550
So the last point I want to
mention here about viscosity

00:26:44.550 --> 00:26:47.490
is how do people characterize
viscosity in glass melts.

00:26:47.490 --> 00:26:50.390
Now, viscosity is one of
the phenomena that can span

00:26:50.390 --> 00:26:52.150
many, many orders of magnitude.

00:26:52.150 --> 00:26:53.870
So it's not surprising
at all that you

00:26:53.870 --> 00:26:56.037
have to use different
techniques to actually measure

00:26:56.037 --> 00:26:57.430
viscosity in different ranges.

00:26:57.430 --> 00:26:59.990
So if you're trying
to measure viscosity

00:26:59.990 --> 00:27:02.430
in the low-viscosity
regime, it behaves more

00:27:02.430 --> 00:27:04.590
like liquid, flowable liquid.

00:27:04.590 --> 00:27:06.390
So what you're
typically supposed to do

00:27:06.390 --> 00:27:08.670
is use this kind of
rotating cylinder approach.

00:27:08.670 --> 00:27:11.670
You essentially put the liquid
in between the rotating cylinder

00:27:11.670 --> 00:27:13.110
and the container.

00:27:13.110 --> 00:27:15.190
And you can then
measure what torque

00:27:15.190 --> 00:27:18.350
you need to apply
depending on the spin rate.

00:27:18.350 --> 00:27:20.830
So from that, you
can actually infer

00:27:20.830 --> 00:27:23.910
what is the viscosity, the
friction the liquid exerts

00:27:23.910 --> 00:27:25.210
on the spinning cylinder.

00:27:25.210 --> 00:27:28.530
Now, if you go to
a higher viscosity,

00:27:28.530 --> 00:27:31.130
this is somewhere around
the working range.

00:27:31.130 --> 00:27:34.190
So chewing gum, like glasses.

00:27:34.190 --> 00:27:36.390
Instead of using this
kind of rotating cylinder,

00:27:36.390 --> 00:27:38.870
you actually use a
parallel plate approach.

00:27:38.870 --> 00:27:42.550
So you put some glass materials
between two parallel plates.

00:27:42.550 --> 00:27:44.590
And then you start to
rotate one of the plates

00:27:44.590 --> 00:27:46.030
with respect to another.

00:27:46.030 --> 00:27:49.430
And similarly, you
measure what is the torque

00:27:49.430 --> 00:27:54.590
you need to apply as a function
of the rotational speed.

00:27:54.590 --> 00:27:57.710
Now, this only applies to this
relatively high viscosity range,

00:27:57.710 --> 00:28:02.350
because if the glass
melt flows like a liquid,

00:28:02.350 --> 00:28:04.110
it's going to leak
out from the edges,

00:28:04.110 --> 00:28:05.690
so you cannot hold the shape.

00:28:05.690 --> 00:28:11.230
So that's why this applies only
to intermediate viscosity range.

00:28:11.230 --> 00:28:13.230
Now, if you go to
even higher viscosity,

00:28:13.230 --> 00:28:16.602
like a round glass
transition temperature,

00:28:16.602 --> 00:28:18.810
the glass will behave more
like a solid in this case.

00:28:18.810 --> 00:28:21.550
The deformation actually
gets pretty small.

00:28:21.550 --> 00:28:24.330
So to measure the viscosity
at this high viscosity range,

00:28:24.330 --> 00:28:27.710
you use techniques that you
borrow from solid mechanics,

00:28:27.710 --> 00:28:28.550
like beam bending.

00:28:28.550 --> 00:28:30.390
This is like three-point
beam bending.

00:28:30.390 --> 00:28:33.430
But the difference there, that
in a conventional three-point

00:28:33.430 --> 00:28:36.030
beam bending, where we're trying
to measure elastic material,

00:28:36.030 --> 00:28:37.870
you are trying to
measure the force

00:28:37.870 --> 00:28:41.790
as a function of
deformation, of displacement.

00:28:41.790 --> 00:28:42.630
That's what you do.

00:28:42.630 --> 00:28:44.630
That's how you infer
the elastic constants.

00:28:44.630 --> 00:28:46.150
But in this case,
if you're trying

00:28:46.150 --> 00:28:48.750
to measure viscosity, what
we're trying to measure

00:28:48.750 --> 00:28:54.110
is the force as a function
of deformation rate.

00:28:54.110 --> 00:28:56.110
It's not the absolute
deformation that matters,

00:28:56.110 --> 00:28:58.590
it's how fast the beam sets.

00:28:58.590 --> 00:29:04.730
It's the movement speed is
what really matters there.

00:29:04.730 --> 00:29:06.830
So this kind of a
beam bending approach

00:29:06.830 --> 00:29:09.810
allows you to measure viscosity
up to about 10 to the 13.

00:29:09.810 --> 00:29:13.412
So somewhere around,
like, straining point.

00:29:13.412 --> 00:29:14.870
What if you want
to study something

00:29:14.870 --> 00:29:19.230
with even higher viscosity?

00:29:19.230 --> 00:29:22.110
You may feel like that's
not important, but turns out

00:29:22.110 --> 00:29:23.830
it can be quite
relevant if you are

00:29:23.830 --> 00:29:26.790
trying to develop precision
glass instruments,

00:29:26.790 --> 00:29:28.473
like big telescope mirrors.

00:29:28.473 --> 00:29:29.890
So a lot of big
telescope mirrors,

00:29:29.890 --> 00:29:32.510
they obviously are
quite heavy and they

00:29:32.510 --> 00:29:34.450
need to achieve
extremely high precision.

00:29:34.450 --> 00:29:37.350
The surface has to have
a shape deviation that's

00:29:37.350 --> 00:29:41.310
less than a fraction of the
wavelength of the light.

00:29:41.310 --> 00:29:43.870
Otherwise, you risk distortion
of optical images there.

00:29:43.870 --> 00:29:46.290
So over time, a time
period of several years,

00:29:46.290 --> 00:29:50.110
these big mirrors,
they can slowly deform,

00:29:50.110 --> 00:29:52.470
mostly due to
gravitational force.

00:29:52.470 --> 00:29:56.910
So how do you measure this
kind of a viscous flow

00:29:56.910 --> 00:29:59.830
in an instrument like this,
where the viscosity is obviously

00:29:59.830 --> 00:30:00.870
very high?

00:30:00.870 --> 00:30:04.150
So the technique they use is
basically optical interference.

00:30:04.150 --> 00:30:09.050
So what they will do,
they can, for example,

00:30:09.050 --> 00:30:13.910
put the piece of glass
you want to measure,

00:30:13.910 --> 00:30:16.790
you hold it on
the side, and then

00:30:16.790 --> 00:30:19.080
you put it against
the reference plate.

00:30:19.080 --> 00:30:21.720
So the bottom one, this will
be your reference plate.

00:30:21.720 --> 00:30:22.800
And this is your sample.

00:30:22.800 --> 00:30:25.460


00:30:25.460 --> 00:30:31.140
What you do, you then
shine a light from on top.

00:30:31.140 --> 00:30:33.580
So then when light
passes through this gap

00:30:33.580 --> 00:30:35.220
in between these
two samples here,

00:30:35.220 --> 00:30:37.140
it can bounce back
and forth many times

00:30:37.140 --> 00:30:39.660
between the sample and
the reference plate.

00:30:39.660 --> 00:30:41.780
And then depending
on whether you

00:30:41.780 --> 00:30:44.380
get integral, multiple
wavelengths or maybe

00:30:44.380 --> 00:30:46.520
half-integral or
multiple wavelengths,

00:30:46.520 --> 00:30:49.420
you can get constructive versus
destructive interference.

00:30:49.420 --> 00:30:53.180
So if you use a white light to
shine through this plate here,

00:30:53.180 --> 00:30:55.860
some wavelengths constructively
interfere, some wavelengths

00:30:55.860 --> 00:30:57.220
destructively interfere.

00:30:57.220 --> 00:31:01.380
So you actually see color
in the reflected image.

00:31:01.380 --> 00:31:04.020
So this is very similar
to when you have

00:31:04.020 --> 00:31:05.235
oil films spreading on water.

00:31:05.235 --> 00:31:07.360
And then you see this kind
of color oil film there.

00:31:07.360 --> 00:31:07.860
Same idea.

00:31:07.860 --> 00:31:09.500
This is basically
called interference.

00:31:09.500 --> 00:31:11.980
So from the color
fringe, you can now

00:31:11.980 --> 00:31:16.420
infer what is the gap width
between these two pieces

00:31:16.420 --> 00:31:18.180
here with very high
precision, down to,

00:31:18.180 --> 00:31:21.340
like, single-digit-- a few
nanometers of precision.

00:31:21.340 --> 00:31:23.660
So now with this
technique, you can then

00:31:23.660 --> 00:31:26.500
carry out this measurement
over an extended period of time

00:31:26.500 --> 00:31:30.100
to determine what is the
viscosity at this high viscosity

00:31:30.100 --> 00:31:30.807
range.

00:31:30.807 --> 00:31:31.640
So I'm quoting here.

00:31:31.640 --> 00:31:33.700
This is basically
from one paper.

00:31:33.700 --> 00:31:35.620
The authors perform
the measurement

00:31:35.620 --> 00:31:37.940
over a period of six years.

00:31:37.940 --> 00:31:39.740
And they were able to
measure deformation

00:31:39.740 --> 00:31:42.220
of the order of a few
nanometers to calculate

00:31:42.220 --> 00:31:46.120
viscosity of the order of 10
to 15, 16, 17 Pascal-second.

00:31:46.120 --> 00:31:49.040


00:31:49.040 --> 00:31:49.540
Yes.

00:31:49.540 --> 00:31:52.340
Again, not a very
good topic for PhD

00:31:52.340 --> 00:31:55.320
thesis, but you know, six
years, maybe tolerable.

00:31:55.320 --> 00:31:59.540
Better than the pitch
drop experiment.

00:31:59.540 --> 00:32:01.712
You get one paper
after six years?

00:32:01.712 --> 00:32:02.920
That's actually a true story.

00:32:02.920 --> 00:32:06.093
I was actually at
a glass conference.

00:32:06.093 --> 00:32:08.260
Someone was-- the thesis
topic was actually studying

00:32:08.260 --> 00:32:10.705
glass aging or relaxation.

00:32:10.705 --> 00:32:12.580
So they actually inherited
a bunch of samples

00:32:12.580 --> 00:32:15.295
from previous PhD students.

00:32:15.295 --> 00:32:17.420
And then they actually
measured all the properties.

00:32:17.420 --> 00:32:20.920
And then studying why you
have this kind of relaxation.

00:32:20.920 --> 00:32:23.260
So yeah, some of these
long-term experiments

00:32:23.260 --> 00:32:24.820
can be legit PhD topics.

00:32:24.820 --> 00:32:27.740
But it requires
transition of samples

00:32:27.740 --> 00:32:29.760
between different
generations of PhD students.

00:32:29.760 --> 00:32:32.900


00:32:32.900 --> 00:32:36.700
So now if we are now able
to measure or characterize

00:32:36.700 --> 00:32:39.537
the viscosity of glasses, even
to this very high range there,

00:32:39.537 --> 00:32:41.620
we can now come back to
the question we have asked

00:32:41.620 --> 00:32:42.920
in the beginning of the class.

00:32:42.920 --> 00:32:48.540
So is glass a solid
or a viscous fluid?

00:32:48.540 --> 00:32:51.740
Is the flow of glass
completely negligible

00:32:51.740 --> 00:32:54.083
under ambient conditions?

00:32:54.083 --> 00:32:55.500
So this question
has actually been

00:32:55.500 --> 00:32:57.460
investigated by a few authors.

00:32:57.460 --> 00:32:59.620
This is one of the
most recent attempts

00:32:59.620 --> 00:33:02.040
that people from a group
of scientists from Corning,

00:33:02.040 --> 00:33:03.940
they actually performed
some calculations.

00:33:03.940 --> 00:33:07.340
Essentially, they experimentally
measured the viscosity of glass

00:33:07.340 --> 00:33:10.000
using whatever technique
we mentioned there.

00:33:10.000 --> 00:33:11.780
Then they can extrapolate
this viscosity

00:33:11.780 --> 00:33:14.718
to room temperature using
the [INAUDIBLE] equation.

00:33:14.718 --> 00:33:16.260
Remember the equation
we talked about

00:33:16.260 --> 00:33:18.552
for how to describe viscosity
dependent on temperature?

00:33:18.552 --> 00:33:20.415
So then they could
use that to then infer

00:33:20.415 --> 00:33:22.540
what's the viscosity of
glass, normal glass at room

00:33:22.540 --> 00:33:23.340
temperature.

00:33:23.340 --> 00:33:24.940
And then from
that, they can then

00:33:24.940 --> 00:33:27.780
estimate what is the
time that's needed

00:33:27.780 --> 00:33:32.460
to deform the glass, for
example, for a few nanometers.

00:33:32.460 --> 00:33:36.560
And the conclusion there is
that at room temperature,

00:33:36.560 --> 00:33:39.100
the viscosity is of the
order of 10 to 24 to 10

00:33:39.100 --> 00:33:41.220
to the 25 Pascal-second.

00:33:41.220 --> 00:33:45.220
That's 12 orders of magnitude
higher than glass transition

00:33:45.220 --> 00:33:46.100
temperature.

00:33:46.100 --> 00:33:47.940
And if you think
about pitch, this

00:33:47.940 --> 00:33:50.340
is like 16 orders of magnitude
higher viscosity compared

00:33:50.340 --> 00:33:52.060
to pitch.

00:33:52.060 --> 00:33:54.580
So it's not surprising
that it would

00:33:54.580 --> 00:33:56.460
take some-- the
conclusion here is

00:33:56.460 --> 00:33:59.380
that it takes a
billion years to just

00:33:59.380 --> 00:34:03.900
have a maximum flow of
order of about 1 nanometer.

00:34:03.900 --> 00:34:07.820
So conclusion-- glasses
at room temperature--

00:34:07.820 --> 00:34:09.280
I mean soda-lime glass.

00:34:09.280 --> 00:34:11.460
You're not talking about
these low-Tg glasses.

00:34:11.460 --> 00:34:13.040
For things like
soda-lime glasses,

00:34:13.040 --> 00:34:15.780
you can pretty much treat them
just like [? electric ?] solids.

00:34:15.780 --> 00:34:19.540
The viscous behavior is pretty
much completely negligible

00:34:19.540 --> 00:34:21.520
because it's extremely
high viscosity.

00:34:21.520 --> 00:34:23.968


00:34:23.968 --> 00:34:25.260
Now, you may have the question.

00:34:25.260 --> 00:34:28.120
For those medieval
cathedral color windows,

00:34:28.120 --> 00:34:31.500
why the bottom tend to
be wider than the top?

00:34:31.500 --> 00:34:33.060
The answer is
actually quite simple.

00:34:33.060 --> 00:34:35.760
Because when the construction
workers, when they assemble it,

00:34:35.760 --> 00:34:37.840
they know at the time the
window is not uniform,

00:34:37.840 --> 00:34:38.960
thickness is not uniform.

00:34:38.960 --> 00:34:40.880
So in order to ensure
mechanical stability,

00:34:40.880 --> 00:34:43.219
they tend to put the thicker
side at the bottom, then

00:34:43.219 --> 00:34:46.380
the thinner side on top.

00:34:46.380 --> 00:34:49.020
Now, one of the
examples to support this

00:34:49.020 --> 00:34:50.840
is that people do find mistakes.

00:34:50.840 --> 00:34:53.340
They do sometimes
see these window

00:34:53.340 --> 00:34:56.155
pane glasses where the thicker
side is actually on top.

00:34:56.155 --> 00:34:57.780
That's just because
it's assembly error

00:34:57.780 --> 00:34:58.960
of the construction workers.

00:34:58.960 --> 00:35:01.500


00:35:01.500 --> 00:35:05.140
So just to conclude what
we have talked about here,

00:35:05.140 --> 00:35:07.660
I decided to put this quote--

00:35:07.660 --> 00:35:11.802
"chocolate smells, flowers
wilt, diamonds are forever."

00:35:11.802 --> 00:35:13.510
You can obviously see
where this is from.

00:35:13.510 --> 00:35:16.450
This is from a shop
that sells diamonds.

00:35:16.450 --> 00:35:18.070
But from what we
have learned here,

00:35:18.070 --> 00:35:20.605
you see that glasses
can do equally well.

00:35:20.605 --> 00:35:22.730
In fact, diamonds are not
thermodynamically stable.

00:35:22.730 --> 00:35:24.605
Graphite is a
thermodynamically stable state.

00:35:24.605 --> 00:35:27.690
So glasses are forever.

00:35:27.690 --> 00:35:29.230
So forget about diamond rings.

00:35:29.230 --> 00:35:31.330
I think glass rings
can work equally well.

00:35:31.330 --> 00:35:33.970
OK.

00:35:33.970 --> 00:35:35.150
Think twice about it.

00:35:35.150 --> 00:35:39.170


00:35:39.170 --> 00:35:40.290
That's the next chapter.

00:35:40.290 --> 00:35:42.415
We're going to talk about
mechanical property next.

00:35:42.415 --> 00:35:43.242
So hold on to that.

00:35:43.242 --> 00:35:45.450
You are going to see some
surprising numbers there as

00:35:45.450 --> 00:35:45.950
well.

00:35:45.950 --> 00:35:50.530
Glasses are also pretty good in
terms of mechanical properties.

00:35:50.530 --> 00:35:51.750
You should be more precise.

00:35:51.750 --> 00:35:54.770
It's a hardness that
really makes a difference.

00:35:54.770 --> 00:35:57.210
Mechanical properties--
glasses are quite good as long

00:35:57.210 --> 00:36:00.290
as they are defect-free.

00:36:00.290 --> 00:36:01.170
OK.

00:36:01.170 --> 00:36:03.850
So just to quickly summarize
what we have talked about here,

00:36:03.850 --> 00:36:06.076
we have discussed
about two kinds

00:36:06.076 --> 00:36:09.230
of contrasting phenomena--
elasticity and viscosity.

00:36:09.230 --> 00:36:11.770
We talked about elasticity
being instantaneous.

00:36:11.770 --> 00:36:15.450
Viscosity, the key point here
is that it's rate-dependent.

00:36:15.450 --> 00:36:18.210
It's permanent but
then the shear stress

00:36:18.210 --> 00:36:20.270
depends on the
shear strain rate.

00:36:20.270 --> 00:36:24.610
So these come from different
microscopic mechanisms

00:36:24.610 --> 00:36:26.250
of viscosity versus elasticity.

00:36:26.250 --> 00:36:29.050
Elasticity comes from the
repulsive and attractive forces

00:36:29.050 --> 00:36:30.030
between molecules.

00:36:30.030 --> 00:36:32.570
And that's why they become
very temperature-insensitive.

00:36:32.570 --> 00:36:37.410
Versus viscosity always
involves relative motion

00:36:37.410 --> 00:36:40.650
of molecules or atoms with
respect to each other.

00:36:40.650 --> 00:36:45.270
So once you involve motion
of atoms or molecules,

00:36:45.270 --> 00:36:47.550
this tends to be a
thermally activated process.

00:36:47.550 --> 00:36:50.810
And that's why you tend to see
a large change of viscosity

00:36:50.810 --> 00:36:52.790
as a function of temperature.

00:36:52.790 --> 00:36:54.410
You describe the
phenomena depending

00:36:54.410 --> 00:36:55.827
on whether you
have a strong fluid

00:36:55.827 --> 00:36:57.590
or you have a fragile fluid.

00:36:57.590 --> 00:36:59.690
You can either use the
Arrhenius equation or maybe

00:36:59.690 --> 00:37:03.530
the [INAUDIBLE] equation, or
VFT equation for those kinds

00:37:03.530 --> 00:37:04.530
of behaviors there.

00:37:04.530 --> 00:37:08.890
And finally, we talked about
this simple Newtonian model does

00:37:08.890 --> 00:37:10.510
not always hold for glasses.

00:37:10.510 --> 00:37:14.410
You can have a
non-Newtonian viscous flow.

00:37:14.410 --> 00:37:17.530
So the viscosity actually
depends on strain rate as well.

00:37:17.530 --> 00:37:20.090


00:37:20.090 --> 00:37:24.050
So that's pretty much about the
basic knowledge about viscosity

00:37:24.050 --> 00:37:24.550
there.

00:37:24.550 --> 00:37:26.650
Any questions before
we move on to talk

00:37:26.650 --> 00:37:29.710
about how to apply
the [? phantom ?] side

00:37:29.710 --> 00:37:33.850
we have learned into
shaping glass products?

00:37:33.850 --> 00:37:36.950
Questions?

00:37:36.950 --> 00:37:38.450
All good?

00:37:38.450 --> 00:37:39.030
All right.

00:37:39.030 --> 00:37:40.750
So now glass shaping.

00:37:40.750 --> 00:37:43.225


00:37:43.225 --> 00:37:45.350
Before I go in to talk
about any technical details,

00:37:45.350 --> 00:37:46.930
I just want to show you
this little paragraph,

00:37:46.930 --> 00:37:49.290
which I exerted from an
online forum, where people

00:37:49.290 --> 00:37:52.370
are essentially asking
questions about why people like

00:37:52.370 --> 00:37:54.730
to use glasses to make
all different kinds

00:37:54.730 --> 00:37:56.270
of everyday products.

00:37:56.270 --> 00:37:58.570
What's the benefits there
in terms of processing?

00:37:58.570 --> 00:38:00.770
So I quite like this quote here.

00:38:00.770 --> 00:38:03.410
It says that viscosity makes
things happen essentially

00:38:03.410 --> 00:38:04.970
in slow motion.

00:38:04.970 --> 00:38:06.570
And this is a
unique behavior that

00:38:06.570 --> 00:38:09.610
only exists in glass, where you
have this continuous viscosity

00:38:09.610 --> 00:38:10.950
temperature behavior.

00:38:10.950 --> 00:38:13.090
So you can tune viscosity
precisely with respect

00:38:13.090 --> 00:38:14.570
to temperature.

00:38:14.570 --> 00:38:18.050
So to illustrate
the difference here

00:38:18.050 --> 00:38:21.930
between a crystalline material
and glass, let's think about--

00:38:21.930 --> 00:38:24.430
let's say this is, for example,
volume versus temperature.

00:38:24.430 --> 00:38:27.970
This is a curve we have
seen many, many times.

00:38:27.970 --> 00:38:32.530
So if it's a crystalline
material, then what you will get

00:38:32.530 --> 00:38:34.850
is a curve like this.

00:38:34.850 --> 00:38:38.770
You have a liquid state, but
when you get to melting point,

00:38:38.770 --> 00:38:43.690
abruptly, you suddenly change
from a solid into a liquid.

00:38:43.690 --> 00:38:47.610
So there's a very large switch
of the deformation behavior

00:38:47.610 --> 00:38:48.510
of the material.

00:38:48.510 --> 00:38:52.010
Liquid tend to have
a low viscosity.

00:38:52.010 --> 00:38:54.150
So as a result, things
start to chip away.

00:38:54.150 --> 00:38:55.530
That's what I'm showing there.

00:38:55.530 --> 00:38:58.490
Essentially, you cannot hold the
shape when it goes through this

00:38:58.490 --> 00:39:00.170
abrupt transition there.

00:39:00.170 --> 00:39:03.130
Glass, on the other hand, you
have a continuous viscosity

00:39:03.130 --> 00:39:03.985
temperature curve.

00:39:03.985 --> 00:39:05.610
As you go through
this glass transition

00:39:05.610 --> 00:39:08.230
here, by controlling
temperature,

00:39:08.230 --> 00:39:10.410
you can quite
precisely determine

00:39:10.410 --> 00:39:15.410
the viscosity of the glass melt.
And by choosing a glass melt

00:39:15.410 --> 00:39:18.170
with appropriate viscosity, you
can ensure that you can easily

00:39:18.170 --> 00:39:20.360
deform it, but yet
it can also hold

00:39:20.360 --> 00:39:22.110
its shape, in that
it's not dripping away,

00:39:22.110 --> 00:39:23.970
it's not collapsing.

00:39:23.970 --> 00:39:25.230
That's what is put here.

00:39:25.230 --> 00:39:29.130
So if it's a crystalline
solid, as soon

00:39:29.130 --> 00:39:31.130
as you reach the melting
point, a drop of liquid

00:39:31.130 --> 00:39:34.090
forms and falls away
from the melting surface.

00:39:34.090 --> 00:39:36.290
Glass, on the other
hand, can gradually

00:39:36.290 --> 00:39:39.770
transform from a high solid
to a slowly softening liquid.

00:39:39.770 --> 00:39:42.410
And the liquid now
allows you to--

00:39:42.410 --> 00:39:45.170
it gradually stiffens as it
cools, because the viscosity

00:39:45.170 --> 00:39:45.930
increases.

00:39:45.930 --> 00:39:49.368
And it gives you a window
to allow glass workers

00:39:49.368 --> 00:39:51.410
to then process it into
different kinds of shapes

00:39:51.410 --> 00:39:53.250
by manipulating it.

00:39:53.250 --> 00:39:55.810
So the key takeaway
message here is

00:39:55.810 --> 00:39:57.770
that this continuous
viscosity temperature

00:39:57.770 --> 00:39:59.870
curve is really important.

00:39:59.870 --> 00:40:02.290
That underlies the
process versatility

00:40:02.290 --> 00:40:04.190
of amorphous glass materials.

00:40:04.190 --> 00:40:08.490


00:40:08.490 --> 00:40:12.430
I just also want to mention
that in a couple of weeks,

00:40:12.430 --> 00:40:15.250
we're going to have a demo
session at the MIT Glass Lab.

00:40:15.250 --> 00:40:21.330
You can find the date on the
first slide deck in the class

00:40:21.330 --> 00:40:21.830
there.

00:40:21.830 --> 00:40:23.550
It's basically the
first lesson in March.

00:40:23.550 --> 00:40:26.510
So we are going to meet
directly at the MIT Glass Lab.

00:40:26.510 --> 00:40:28.330
We are going to show
some demos about how

00:40:28.330 --> 00:40:32.770
to use this viscous
flow behavior to shape

00:40:32.770 --> 00:40:34.090
different glass products.

00:40:34.090 --> 00:40:37.830


00:40:37.830 --> 00:40:40.388
So as I promised you, I'm going
to show you this plot again,

00:40:40.388 --> 00:40:42.930
because this is actually quite
important to glass processing.

00:40:42.930 --> 00:40:46.690
Key takeaway message is that
most of the glass processes

00:40:46.690 --> 00:40:49.010
takes place in this
working range, where

00:40:49.010 --> 00:40:51.730
you have a chewing gum-like
fluids that you can easily

00:40:51.730 --> 00:40:53.450
shape and deform.

00:40:53.450 --> 00:40:55.490
You do notice here
that this float glass

00:40:55.490 --> 00:40:58.137
goes in the relatively
low viscosity end.

00:40:58.137 --> 00:40:59.970
And that's because in
a float glass process,

00:40:59.970 --> 00:41:01.845
we're going to talk
about it in a few slides,

00:41:01.845 --> 00:41:04.760
is when you have molten
glass floating on a tin bath

00:41:04.760 --> 00:41:07.900
to get large, flat sheets,
because in this case,

00:41:07.900 --> 00:41:09.240
the glass needs to flow.

00:41:09.240 --> 00:41:12.440
So that's why people like to do
it at slightly lower viscosity

00:41:12.440 --> 00:41:13.040
points.

00:41:13.040 --> 00:41:17.720
Now, this one, this is
precision glass molding.

00:41:17.720 --> 00:41:19.560
So precision glass
molding takes place

00:41:19.560 --> 00:41:21.920
at higher viscosity
ranges, sometimes right

00:41:21.920 --> 00:41:23.520
outside the working range.

00:41:23.520 --> 00:41:26.960
Because unlike other processes
here, precision glass molding,

00:41:26.960 --> 00:41:30.060
you apply very large pressure.

00:41:30.060 --> 00:41:31.800
If you apply large
pressure, we know

00:41:31.800 --> 00:41:33.880
that the shear rate
depends on-- it linearly

00:41:33.880 --> 00:41:35.532
scales with shear stress.

00:41:35.532 --> 00:41:37.240
So high stress
essentially means that you

00:41:37.240 --> 00:41:39.920
can afford to have low viscosity
while still maintaining

00:41:39.920 --> 00:41:41.880
a decent stress rate.

00:41:41.880 --> 00:41:45.020
So basically, pressure
expedites the process.

00:41:45.020 --> 00:41:49.400
That's why you can afford to
have a higher pressure there.

00:41:49.400 --> 00:41:52.280
Now, last one.

00:41:52.280 --> 00:41:54.440
Annealing Lehr.

00:41:54.440 --> 00:41:57.273
Lehr is basically just
another name of "furnace."

00:41:57.273 --> 00:41:59.440
So when we talk about glass
shaping in this chapter,

00:41:59.440 --> 00:42:01.440
you're going to see a
lot of terminologies

00:42:01.440 --> 00:42:03.945
that's pretty unique
only to glass processing.

00:42:03.945 --> 00:42:05.320
"Lehr" is one of
them, but "Lehr"

00:42:05.320 --> 00:42:07.278
basically is just a
furnace, annealing furnace.

00:42:07.278 --> 00:42:10.520
In a Lehr, you anneal
the glass, so it releases

00:42:10.520 --> 00:42:11.700
any kind of internal stress.

00:42:11.700 --> 00:42:14.480
So prevent any premature
cracking of the glass products

00:42:14.480 --> 00:42:17.280
during usage.

00:42:17.280 --> 00:42:21.400
So now, in this
slide, I'm summarizing

00:42:21.400 --> 00:42:23.440
the typical viscosity
reference points

00:42:23.440 --> 00:42:25.040
for different
conditional glasses--

00:42:25.040 --> 00:42:27.880
soda-lime, borosilicate,
fused silica.

00:42:27.880 --> 00:42:31.960
Takeaway message-- fused silica,
you see this temperature,

00:42:31.960 --> 00:42:34.060
it's softening
point is very high,

00:42:34.060 --> 00:42:37.360
working point is
basically absent.

00:42:37.360 --> 00:42:38.580
Its temperature is too high.

00:42:38.580 --> 00:42:42.960
So that's why, as I said, nobody
really processes fused silica.

00:42:42.960 --> 00:42:45.200
You can get them, they
get quite expensive.

00:42:45.200 --> 00:42:47.495
Borosilicate, due to a
higher network connectivity,

00:42:47.495 --> 00:42:49.120
all of this temperature
reference point

00:42:49.120 --> 00:42:51.960
is actually higher than the
corresponding soda-lime glass.

00:42:51.960 --> 00:42:55.600
They also have a much
lower coefficient

00:42:55.600 --> 00:42:58.040
of thermal expansion.

00:42:58.040 --> 00:42:59.920
So that's why,
when you're trying

00:42:59.920 --> 00:43:01.840
to make borosilicate
products, you

00:43:01.840 --> 00:43:03.760
can make it in the
glass workshop,

00:43:03.760 --> 00:43:06.000
but it becomes very
difficult if you're

00:43:06.000 --> 00:43:10.705
trying to mix soda-lime glass
with borosilicate glass.

00:43:10.705 --> 00:43:12.580
Because when you're
processing these glasses,

00:43:12.580 --> 00:43:14.520
when you're shaping them, you
do it at a high temperature.

00:43:14.520 --> 00:43:15.937
If you're trying
to fuse something

00:43:15.937 --> 00:43:18.000
together, then,
as you cool down,

00:43:18.000 --> 00:43:20.000
due to the large
CTE mismatch, you

00:43:20.000 --> 00:43:24.080
accumulate a huge thermal stress
in between the soda-lime glass

00:43:24.080 --> 00:43:25.280
and borosilicate glass.

00:43:25.280 --> 00:43:27.400
And that tends to
actually fracture.

00:43:27.400 --> 00:43:29.350
In fact, when you go
to a glass workshop,

00:43:29.350 --> 00:43:31.100
you're going to do one
of the demos there,

00:43:31.100 --> 00:43:34.620
we're going to draw canes out of
both borosilicate and soda-lime,

00:43:34.620 --> 00:43:36.560
essentially long glass,
thin glass wires.

00:43:36.560 --> 00:43:38.820
We are then trying to
fuse two canes together,

00:43:38.820 --> 00:43:40.945
borosilicate with the soda-lime.

00:43:40.945 --> 00:43:43.320
And once it cools down, you'll
notice that it immediately

00:43:43.320 --> 00:43:45.880
bends, much like a
bimetallic strip,

00:43:45.880 --> 00:43:48.360
and it eventually cracks
into small pieces.

00:43:48.360 --> 00:43:52.800
So you're going to see that when
we go to the glass shop demo.

00:43:52.800 --> 00:43:54.640
All right.

00:43:54.640 --> 00:43:56.660
Questions?

00:43:56.660 --> 00:43:57.280
OK.

00:43:57.280 --> 00:43:59.837
So in the next few
slides in this part here,

00:43:59.837 --> 00:44:01.920
I'm going to show you a
few processes-- how people

00:44:01.920 --> 00:44:04.003
use to actually shape
glasses into different kinds

00:44:04.003 --> 00:44:05.240
of products.

00:44:05.240 --> 00:44:08.720
The earliest process that people
do for the common glass form

00:44:08.720 --> 00:44:11.840
factor, which is flat glass,
like window pane glass.

00:44:11.840 --> 00:44:13.520
The earliest process
people developed

00:44:13.520 --> 00:44:17.560
is so-called crown glass.

00:44:17.560 --> 00:44:21.600
So crown glass, you have a
big bubble of molten glass.

00:44:21.600 --> 00:44:24.520
You open the hole
there at the end,

00:44:24.520 --> 00:44:26.640
so it becomes a big, flat piece.

00:44:26.640 --> 00:44:28.200
And then you start to spin.

00:44:28.200 --> 00:44:31.560
You put it onto a
little steel rod.

00:44:31.560 --> 00:44:33.240
At the end, this is
your glass piece.

00:44:33.240 --> 00:44:35.680
You start to spin it,
so then this glass piece

00:44:35.680 --> 00:44:36.780
starts to rotate.

00:44:36.780 --> 00:44:40.140
And due to the centrifuge force,
it forms this thin, flat sheet.

00:44:40.140 --> 00:44:43.400
So like a cylindrical,
big, thin, flat sheet.

00:44:43.400 --> 00:44:45.440
So you can then use
that to actually create

00:44:45.440 --> 00:44:48.260
this-- cut it and create this
kind of flat glass panels there.

00:44:48.260 --> 00:44:50.200
But then you notice,
obviously, this process

00:44:50.200 --> 00:44:51.840
is not very well controlled.

00:44:51.840 --> 00:44:55.120
You see that there's actually a
big defect right in the center.

00:44:55.120 --> 00:44:57.900
This is where your
glass rose, or pontil,

00:44:57.900 --> 00:45:00.240
as the glass workers call
it, comes into contact

00:45:00.240 --> 00:45:01.640
with the glass piece.

00:45:01.640 --> 00:45:03.590
OK.

00:45:03.590 --> 00:45:07.128
AUDIENCE: Is it
like pizza spinning?

00:45:07.128 --> 00:45:07.920
PROFESSOR: Similar.

00:45:07.920 --> 00:45:08.820
Very similar to that.

00:45:08.820 --> 00:45:11.040
Yeah.

00:45:11.040 --> 00:45:12.860
You are doing it this way, OK?

00:45:12.860 --> 00:45:14.140
And you're just rotating it.

00:45:14.140 --> 00:45:15.320
Yeah.

00:45:15.320 --> 00:45:17.320
So in fact, I was looking
at some Instagrams.

00:45:17.320 --> 00:45:19.320
They have pictures of
this [INAUDIBLE] glassware

00:45:19.320 --> 00:45:20.897
of preparing for
a glass lecture.

00:45:20.897 --> 00:45:23.480
And people were commenting here,
like, whoa, this glass window

00:45:23.480 --> 00:45:24.340
looks so cute.

00:45:24.340 --> 00:45:27.910
Why you only see in
Europe and not in the US?

00:45:27.910 --> 00:45:29.660
Because we don't have
those old buildings.

00:45:29.660 --> 00:45:30.760
That's it.

00:45:30.760 --> 00:45:37.000
We all use a newer process
to actually make the windows.

00:45:37.000 --> 00:45:40.200
So one of the standard
process that people

00:45:40.200 --> 00:45:42.960
use today to make
large flat glass sheet,

00:45:42.960 --> 00:45:45.080
this is called
float glass process.

00:45:45.080 --> 00:45:50.180
So what I'm illustrating here
is a continuous process flow--

00:45:50.180 --> 00:45:53.320
how people actually manufacture
this kind of flat glasses here.

00:45:53.320 --> 00:45:54.940
It's all a continuous process.

00:45:54.940 --> 00:45:56.910
You started out from
feeding the raw material

00:45:56.910 --> 00:45:59.390
all the way to coming out,
having glass sheets coming out

00:45:59.390 --> 00:46:00.210
continuously.

00:46:00.210 --> 00:46:02.750
And non-stop.

00:46:02.750 --> 00:46:03.783
24/7.

00:46:03.783 --> 00:46:05.450
It's a continuous
manufacturing process.

00:46:05.450 --> 00:46:08.070
So basically, the
process involves

00:46:08.070 --> 00:46:10.630
starting with mixing different
kinds of ingredients.

00:46:10.630 --> 00:46:15.270
So for soda-lime glasses, the
ingredient would be sand, SiO2.

00:46:15.270 --> 00:46:22.470
You have soda, which is
sodium carbonate, for example.

00:46:22.470 --> 00:46:26.830
You also have some minerals that
actually contain calcium oxide,

00:46:26.830 --> 00:46:28.190
like dolomite.

00:46:28.190 --> 00:46:29.550
OK.

00:46:29.550 --> 00:46:33.710
That gives you the
rare earth element.

00:46:33.710 --> 00:46:38.410
And then finally, you also
add some broken glasses,

00:46:38.410 --> 00:46:40.030
used glasses.

00:46:40.030 --> 00:46:44.130
Those used glasses
are called cullets.

00:46:44.130 --> 00:46:48.310


00:46:48.310 --> 00:46:49.250
OK.

00:46:49.250 --> 00:46:50.590
This is a name to use--

00:46:50.590 --> 00:46:53.910
you put some weighted glasses
into your feeding material,

00:46:53.910 --> 00:46:55.110
into the furnace.

00:46:55.110 --> 00:46:56.730
Now, this is actually
quite important.

00:46:56.730 --> 00:46:58.350
Why do they need to
put cullets there?

00:46:58.350 --> 00:47:00.142
Because if you think
about it, if you don't

00:47:00.142 --> 00:47:01.558
need to have the
cullets, you just

00:47:01.558 --> 00:47:03.370
have all the raw
minerals coming in there,

00:47:03.370 --> 00:47:06.770
they're all solid, crystalline
solid-- in order to melt it,

00:47:06.770 --> 00:47:09.350
you need to go all the
way to the melting point,

00:47:09.350 --> 00:47:12.270
and then you start
forming the liquid phase.

00:47:12.270 --> 00:47:15.510
On the other hand, if you have
cullets, they are glasses.

00:47:15.510 --> 00:47:17.590
It means that they
start to soften,

00:47:17.590 --> 00:47:21.150
and you form an early liquid
phase well before you actually

00:47:21.150 --> 00:47:24.270
reach the melting point.

00:47:24.270 --> 00:47:27.430
Now, that's important
because for any kind

00:47:27.430 --> 00:47:30.270
of solid-state
reaction, diffusion

00:47:30.270 --> 00:47:31.830
or any kind of
transport phenomena

00:47:31.830 --> 00:47:33.870
always happens a lot
faster in the liquid phase

00:47:33.870 --> 00:47:35.930
compared to the solid phase,
which makes a lot of sense.

00:47:35.930 --> 00:47:37.490
Liquid phase
molecules move around.

00:47:37.490 --> 00:47:39.270
So of course, you
can get much higher

00:47:39.270 --> 00:47:42.650
molecular transport or
reaction compared to any solid.

00:47:42.650 --> 00:47:45.390
So having these
cullets allows you

00:47:45.390 --> 00:47:48.070
to form this early
liquid phase, and that

00:47:48.070 --> 00:47:51.972
helps you to reduce the energy
consumption in the process.

00:47:51.972 --> 00:47:53.430
You don't have to
melt the material

00:47:53.430 --> 00:47:56.610
all the way by going all the
way up to the melting point.

00:47:56.610 --> 00:47:59.110
So then you can significantly
cut down the power consumption

00:47:59.110 --> 00:48:01.750
by something like 20%
by just adding cullets

00:48:01.750 --> 00:48:02.930
into the process there.

00:48:02.930 --> 00:48:04.870
So that's why waste
glasses is actually

00:48:04.870 --> 00:48:09.230
quite important ingredient.

00:48:09.230 --> 00:48:10.710
In fact, I have
seen people doing--

00:48:10.710 --> 00:48:12.710
they do internship
at glass products.

00:48:12.710 --> 00:48:14.570
They make glass
bottles and stuff.

00:48:14.570 --> 00:48:16.528
If they see any glass
bottles that's defective,

00:48:16.528 --> 00:48:18.070
they don't try to
fix it or anything.

00:48:18.070 --> 00:48:20.110
Just take it out, throw
it into a cullet piece,

00:48:20.110 --> 00:48:21.490
because they need it anyway.

00:48:21.490 --> 00:48:21.990
Yeah.

00:48:21.990 --> 00:48:24.590


00:48:24.590 --> 00:48:25.090
OK.

00:48:25.090 --> 00:48:28.380
So then once you mix the
material, you then melt it.

00:48:28.380 --> 00:48:30.380
And then the melting
furnace has the liquid that

00:48:30.380 --> 00:48:32.150
then feed into a floating bath.

00:48:32.150 --> 00:48:35.990
So for the floating bath, you
basically have molten tin.

00:48:35.990 --> 00:48:39.010
So molten tin has a density
that's higher than glass melts.

00:48:39.010 --> 00:48:40.750
So then this molten
glass is then

00:48:40.750 --> 00:48:43.070
going to float on top of tin.

00:48:43.070 --> 00:48:50.190
And now, due to the action of
surface tension and gravity,

00:48:50.190 --> 00:48:53.030
then the liquid then spreads
on top of this molten tin

00:48:53.030 --> 00:48:55.630
to form a flat glass surface.

00:48:55.630 --> 00:48:56.310
OK.

00:48:56.310 --> 00:48:59.450
At the end, when it comes out
from this floating furnace

00:48:59.450 --> 00:49:04.870
there, then it enters
into the annealing Lehr.

00:49:04.870 --> 00:49:08.830
So annealing Lehr then helps
you to relax any internal stress

00:49:08.830 --> 00:49:11.630
inside this flat glass sheet
before it really comes out

00:49:11.630 --> 00:49:13.190
from here to get your--

00:49:13.190 --> 00:49:15.910
then it can be cut and packaged
into all kinds of flat glass

00:49:15.910 --> 00:49:16.610
products.

00:49:16.610 --> 00:49:20.470
So what I will do
next is I'm actually

00:49:20.470 --> 00:49:22.990
going to play a little YouTube
video, which probably talks

00:49:22.990 --> 00:49:24.870
all about this
process and gives you

00:49:24.870 --> 00:49:30.016
a visual feeling of how things
are done.

00:49:39.550 --> 00:49:42.175
OK, so basically you
see the visuals of the process,

00:49:42.175 --> 00:49:44.508
yeah-- it's that the temperature
monotonically decreases

00:49:44.508 --> 00:49:45.975
from one end to the other end.

00:49:45.975 --> 00:49:47.815
And then it's a
continuous process that

00:49:47.815 --> 00:49:50.455
manufactures glass products.

00:49:50.455 --> 00:49:52.135
So one thing I
also mentioned here

00:49:52.135 --> 00:49:54.177
is about the ingredient
they have mentioned here.

00:49:54.177 --> 00:49:56.095
They talk about dolomite
increasing weathering

00:49:56.095 --> 00:49:56.637
properties.

00:49:56.637 --> 00:49:59.095
That's one way to mention is
to say that they are improving

00:49:59.095 --> 00:50:00.553
the chemical
resistance or chemical

00:50:00.553 --> 00:50:01.970
stability of the material.

00:50:01.970 --> 00:50:03.595
So when we talk about
glass structures,

00:50:03.595 --> 00:50:05.455
we say that these
rare earth elements,

00:50:05.455 --> 00:50:07.155
they have a twofold
coordination.

00:50:07.155 --> 00:50:09.335
So they help to maintain
some network connectivity

00:50:09.335 --> 00:50:11.215
and therefore improving
the-- reducing

00:50:11.215 --> 00:50:12.555
the chemical sensitivity there.

00:50:12.555 --> 00:50:14.263
So that's why they
added dolomites there.

00:50:14.263 --> 00:50:17.815
The sodium sulfates, those
are to add the alkaline

00:50:17.815 --> 00:50:21.915
to decrease the glass
processing temperature.

00:50:21.915 --> 00:50:22.415
All right.

00:50:22.415 --> 00:50:26.295
So the video shows here--
this is basically talking

00:50:26.295 --> 00:50:29.055
about a standard process
for making things

00:50:29.055 --> 00:50:31.195
like windowpane glasses
or architectural glasses.

00:50:31.195 --> 00:50:34.895
I want to mention that this
float glass process has also

00:50:34.895 --> 00:50:36.365
been further advanced.

00:50:36.365 --> 00:50:38.615
You can now even actually
get high-end glass products,

00:50:38.615 --> 00:50:41.815
like consumer electronic
products for [? hover glasses ?]

00:50:41.815 --> 00:50:44.575
and things like that, also
using the float glass process.

00:50:44.575 --> 00:50:46.415
So this is actually one example.

00:50:46.415 --> 00:50:48.495
This paper basically
reports the manufacturing

00:50:48.495 --> 00:50:51.735
of ultra-thin glass
with a thickness that's

00:50:51.735 --> 00:50:54.195
down to about 0.1 millimeters.

00:50:54.195 --> 00:50:57.135
That's about a human
hair thickness, OK?

00:50:57.135 --> 00:51:00.855
And it's a four-meter-wide
flat sheet.

00:51:00.855 --> 00:51:02.535
So I actually had
the opportunity

00:51:02.535 --> 00:51:05.135
to visit one of these plants.

00:51:05.135 --> 00:51:09.335
And when this glass kind of
moves around on the rollers,

00:51:09.335 --> 00:51:11.307
you can basically see ripples.

00:51:11.307 --> 00:51:12.515
They are essentially flexing.

00:51:12.515 --> 00:51:13.515
They are flexible.

00:51:13.515 --> 00:51:15.260
So it's quite a
remarkable process.

00:51:15.260 --> 00:51:17.135
Now, if you look at how
they actually did it,

00:51:17.135 --> 00:51:20.495
they spent lots of time trying
to optimize the temperature

00:51:20.495 --> 00:51:23.175
distribution across zones here.

00:51:23.175 --> 00:51:25.015
They even spent lots
of time optimizing

00:51:25.015 --> 00:51:27.375
the location of these
rollers on the side

00:51:27.375 --> 00:51:30.535
to ensure it generates uniform
stretching force rather

00:51:30.535 --> 00:51:35.205
than having stress or
strain localization.

00:51:35.205 --> 00:51:40.355
So in the end, they were able
to get glass surface roughness--

00:51:40.355 --> 00:51:42.015
for example, on
the top surface--

00:51:42.015 --> 00:51:45.185
down to about 1.5
nanometers [? RMS ?] value,

00:51:45.185 --> 00:51:47.845
which is quite remarkable.

00:51:47.845 --> 00:51:50.325
This basically just
relies on surface tension.

00:51:50.325 --> 00:51:52.057
So that's quite good.

00:51:52.057 --> 00:51:53.765
Now, the downside,
though, is that if you

00:51:53.765 --> 00:51:57.145
use this kind of tin bath,
on the bottom surface,

00:51:57.145 --> 00:52:00.285
you still have the contact of
the glass with the tin surface.

00:52:00.285 --> 00:52:02.685
This increases the
surface roughness up

00:52:02.685 --> 00:52:05.045
to about 3 or 4 nanometers.

00:52:05.045 --> 00:52:07.365
Also, at the same
time, this may cause

00:52:07.365 --> 00:52:08.795
a little bit of contamination.

00:52:08.795 --> 00:52:11.045
There can be a little bit
of diffusion, interdiffusion

00:52:11.045 --> 00:52:12.765
between tin and the glass.

00:52:12.765 --> 00:52:15.085
Also, any kind of a
mechanical vibration

00:52:15.085 --> 00:52:17.245
or any kind of surface
unevenness of the tin bath

00:52:17.245 --> 00:52:19.205
is going to translate
into unevenness

00:52:19.205 --> 00:52:22.645
of the surface of the glass
as well at the bottom surface.

00:52:22.645 --> 00:52:28.165
So that's why, if you're looking
for very high-end, extremely

00:52:28.165 --> 00:52:32.045
high-quality surface glasses,
for example, for semiconductor

00:52:32.045 --> 00:52:35.965
manufacturing
processes, people tend

00:52:35.965 --> 00:52:39.925
to use a different
manufacturing process.

00:52:39.925 --> 00:52:43.525
Where the glass surface
is formed without contact

00:52:43.525 --> 00:52:46.245
into any substance.

00:52:46.245 --> 00:52:48.725
So there's a bunch of
these kinds of variants

00:52:48.725 --> 00:52:49.965
of this process there.

00:52:49.965 --> 00:52:53.085
They are typically
called down-draw process.

00:52:53.085 --> 00:52:55.765
I'm showing you two
variants of this process.

00:52:55.765 --> 00:52:58.125
On the left-hand side,
this is a process

00:52:58.125 --> 00:52:59.725
that's developed
by Schott, which

00:52:59.725 --> 00:53:01.465
is a German glass manufacturer.

00:53:01.465 --> 00:53:04.645
So the way it works is that
you have your molten glass,

00:53:04.645 --> 00:53:08.045
it comes out from an orifice
at the bottom of the furnace,

00:53:08.045 --> 00:53:10.425
and then this glass melt
essentially just drops.

00:53:10.425 --> 00:53:12.665
It comes down by
gravitational force.

00:53:12.665 --> 00:53:16.435


00:53:16.435 --> 00:53:19.005
And in the process
here, this glass surface

00:53:19.005 --> 00:53:21.365
is not coming into
contact with anything.

00:53:21.365 --> 00:53:23.485
So you maintain the
pristine glass quality.

00:53:23.485 --> 00:53:25.725
And in the end, you then go
through annealing furnace

00:53:25.725 --> 00:53:26.725
and then go to a roller.

00:53:26.725 --> 00:53:28.685
You can get continuous
glass ribbon coming out

00:53:28.685 --> 00:53:29.845
from the furnace that way.

00:53:29.845 --> 00:53:33.745
Now, Corning has its
own patented process.

00:53:33.745 --> 00:53:36.377
They call it the fusion process.

00:53:36.377 --> 00:53:38.085
I'm going to show you
a little video that

00:53:38.085 --> 00:53:40.645
shows how this is done.

00:53:40.645 --> 00:53:41.585
So same idea.

00:53:41.585 --> 00:53:44.065
You start by mixing them,
put them into a furnace.

00:53:44.065 --> 00:53:48.965


00:53:48.965 --> 00:53:50.373
It's also a continuous process.

00:53:50.373 --> 00:53:51.665
It can be a continuous process.

00:53:51.665 --> 00:53:53.645
So then, you keep on
adding these materials

00:53:53.645 --> 00:53:55.005
until it overflows.

00:53:55.005 --> 00:53:58.965


00:53:58.965 --> 00:54:02.525
So you see that the glass melts
comes down from the bottom,

00:54:02.525 --> 00:54:07.583
goes into a big crucible.

00:54:07.583 --> 00:54:09.625
Then because you continuously
add the glass melt,

00:54:09.625 --> 00:54:11.385
it starts to overflow
from the top.

00:54:11.385 --> 00:54:15.285
Then the glass doesn't come out
from two sides of the crucible.

00:54:15.285 --> 00:54:17.285
You engineer the crucible
so that the bottom

00:54:17.285 --> 00:54:18.885
has this kind of wedge shape.

00:54:18.885 --> 00:54:21.505
So when the liquid comes
down, it comes at the bottom,

00:54:21.505 --> 00:54:24.925
merges, and forms a
thin sheet of glass.

00:54:24.925 --> 00:54:27.385
And it cools down as it drops.

00:54:27.385 --> 00:54:31.085


00:54:31.085 --> 00:54:33.165
So that's why they call
it fusion process--

00:54:33.165 --> 00:54:36.445
because it involves a fusion of
two liquid streams at the bottom

00:54:36.445 --> 00:54:38.845
of the crucible there.

00:54:38.845 --> 00:54:39.805
OK.

00:54:39.805 --> 00:54:41.247
So same idea there.

00:54:41.247 --> 00:54:43.205
You want to make sure
that when you're actually

00:54:43.205 --> 00:54:45.085
trying to form
the glass surface,

00:54:45.085 --> 00:54:47.752
you don't have any contamination
or it doesn't come into contact

00:54:47.752 --> 00:54:48.845
with any other substances.

00:54:48.845 --> 00:54:53.165
So that's why this is mostly
used for high-end applications

00:54:53.165 --> 00:54:55.487
like flat-panel
displays, where you

00:54:55.487 --> 00:54:57.945
need to process all kinds of
electronic transistors on top.

00:54:57.945 --> 00:55:00.028
So it requires extremely
high surface quality free

00:55:00.028 --> 00:55:01.085
of contamination.

00:55:01.085 --> 00:55:01.845
Question.

00:55:01.845 --> 00:55:05.645
AUDIENCE: Is-- for
the Corning method,

00:55:05.645 --> 00:55:08.765
is the crucible hot
on the exterior?

00:55:08.765 --> 00:55:12.388
Are there binding issues
as the two glass faces--

00:55:12.388 --> 00:55:13.805
PROFESSOR: Oh, you
mean like-- ah.

00:55:13.805 --> 00:55:14.925
OK.

00:55:14.925 --> 00:55:16.485
Normally no, because
it's actually

00:55:16.485 --> 00:55:18.000
at a high enough temperature.

00:55:18.000 --> 00:55:19.625
So the glass is
basically just a fluid.

00:55:19.625 --> 00:55:22.358
So when they come together,
it's just basically two fluids

00:55:22.358 --> 00:55:23.025
mixing together.

00:55:23.025 --> 00:55:25.785
It's not two solid sheets
trying to bind together.

00:55:25.785 --> 00:55:27.425
Basically, it's a fluid stream.

00:55:27.425 --> 00:55:30.385
And it slowly cools down
and the viscosity increases.

00:55:30.385 --> 00:55:31.845
That's when it
becomes solid glass.

00:55:31.845 --> 00:55:32.165
OK.

00:55:32.165 --> 00:55:32.665
Yes?

00:55:32.665 --> 00:55:34.707
AUDIENCE: What is the
benefit of that versus just

00:55:34.707 --> 00:55:36.895
having a thicker [INAUDIBLE]?

00:55:36.895 --> 00:55:38.020
PROFESSOR: Like this one?

00:55:38.020 --> 00:55:38.645
AUDIENCE: Yeah.

00:55:38.645 --> 00:55:40.770
What is the benefit of the
Corning technique there?

00:55:40.770 --> 00:55:42.605
PROFESSOR: OK.

00:55:42.605 --> 00:55:43.345
Patent.

00:55:43.345 --> 00:55:45.745


00:55:45.745 --> 00:55:46.245
Yeah.

00:55:46.245 --> 00:55:48.245
Each of the glass company
have their own patents

00:55:48.245 --> 00:55:49.025
on the process.

00:55:49.025 --> 00:55:51.203
Schott holds patent
on this process,

00:55:51.203 --> 00:55:52.745
Corning holds patent
on that process.

00:55:52.745 --> 00:55:55.145
They cannot use the
other people's process.

00:55:55.145 --> 00:55:57.485
So that's why they
develop their own variants

00:55:57.485 --> 00:55:59.037
of the down-draw process.

00:55:59.037 --> 00:56:00.745
I mean, that's the
main reason, honestly.

00:56:00.745 --> 00:56:02.345
I don't see a big
difference, to be honest,

00:56:02.345 --> 00:56:03.985
between the fusion process
and the down-drawn process.

00:56:03.985 --> 00:56:05.277
They are more or less the same.

00:56:05.277 --> 00:56:07.445
If you look at other
glass manufacturers, AGC

00:56:07.445 --> 00:56:10.005
in Japan, Pilkington
in the UK, they also

00:56:10.005 --> 00:56:12.765
have their own variants of
this kind of down-draw process.

00:56:12.765 --> 00:56:14.848
AUDIENCE: I'm going to
create the up-draw process.

00:56:14.848 --> 00:56:15.717
[LAUGHTER]

00:56:15.717 --> 00:56:17.425
PROFESSOR: You are
going against gravity.

00:56:17.425 --> 00:56:19.665
You can patent it, but you
are going against gravity.

00:56:19.665 --> 00:56:21.285
That's the downside there.

00:56:21.285 --> 00:56:22.345
This is a little easier.

00:56:22.345 --> 00:56:23.285
You just let things flow.

00:56:23.285 --> 00:56:24.385
AUDIENCE: I can
corner the market.

00:56:24.385 --> 00:56:24.885
So.

00:56:24.885 --> 00:56:27.065
PROFESSOR: You could do
a horizontal draw maybe.

00:56:27.065 --> 00:56:32.725
[LAUGHS] Oh, by the way, do
you know when this process was

00:56:32.725 --> 00:56:34.565
first developed?

00:56:34.565 --> 00:56:36.695
AUDIENCE: Like any sort
of down-draw method?

00:56:36.695 --> 00:56:37.362
PROFESSOR: Yeah.

00:56:37.362 --> 00:56:42.035


00:56:42.035 --> 00:56:43.015
When, I mean.

00:56:43.015 --> 00:56:46.235
When was it-- yeah.

00:56:46.235 --> 00:56:47.815
AUDIENCE: Recently?

00:56:47.815 --> 00:56:50.315
PROFESSOR: I mean, the demand
really only comes in recently.

00:56:50.315 --> 00:56:52.915
It's basically when all the
consumer electronics companies

00:56:52.915 --> 00:56:55.855
started to want to make this
kind of high-quality glasses,

00:56:55.855 --> 00:56:57.755
that's actually when
the market matures.

00:56:57.755 --> 00:57:00.715
It was developed
back in the 1960s.

00:57:00.715 --> 00:57:02.075
Yeah.

00:57:02.075 --> 00:57:03.955
So people at the time,
they were thinking

00:57:03.955 --> 00:57:06.155
that they need a high-end
glass process to make

00:57:06.155 --> 00:57:08.915
the windshields for cars.

00:57:08.915 --> 00:57:10.215
The market doesn't exist.

00:57:10.215 --> 00:57:12.875
Nobody needs these high-end
glasses with a very high

00:57:12.875 --> 00:57:14.355
precision surface quality.

00:57:14.355 --> 00:57:17.308
So Corning actually goes
back to their dossier, right,

00:57:17.308 --> 00:57:18.475
looking into old literature.

00:57:18.475 --> 00:57:18.975
When?

00:57:18.975 --> 00:57:22.155
Like 20 years ago when Apple
was approaching them trying

00:57:22.155 --> 00:57:23.735
to make iPhone cover glasses.

00:57:23.735 --> 00:57:25.755
They actually had to go
back into their dossier,

00:57:25.755 --> 00:57:29.355
dig out this old process
and kind of scale it up

00:57:29.355 --> 00:57:30.535
to actually manufacturing.

00:57:30.535 --> 00:57:32.615
So yeah, I mean,
it's an old process,

00:57:32.615 --> 00:57:35.115
but it has a renewed interest
now because of the electronics

00:57:35.115 --> 00:57:35.735
industry.

00:57:35.735 --> 00:57:38.915


00:57:38.915 --> 00:57:40.075
OK.

00:57:40.075 --> 00:57:43.555
How do I go to the next slide?

00:57:43.555 --> 00:57:45.195
Oh, all right.

00:57:45.195 --> 00:57:49.135
So far we've been talking about
how to make flat-panel glasses.

00:57:49.135 --> 00:57:50.285
Big, flat glasses, right?

00:57:50.285 --> 00:57:52.035
So next, I'm going to
talk about something

00:57:52.035 --> 00:57:53.285
that's a little bit different.

00:57:53.285 --> 00:57:56.835
It's glassblowing-- using air
to help to conform glasses

00:57:56.835 --> 00:57:58.075
into different shapes.

00:57:58.075 --> 00:58:00.695
There are a few different
variants of glassblowing here.

00:58:00.695 --> 00:58:03.715
In this slide, I'm showing
you the manual version,

00:58:03.715 --> 00:58:06.325
where this is what people
do in a glass workshop.

00:58:06.325 --> 00:58:08.575
This is what you're going
to see in the MIT Glass Lab.

00:58:08.575 --> 00:58:10.075
We're going to try
to make pumpkins,

00:58:10.075 --> 00:58:12.335
for example, that involves
the glassblowing process.

00:58:12.335 --> 00:58:16.395
So in the glassblowing
process, you basically

00:58:16.395 --> 00:58:19.815
have this big blow pipe,
which is hollow in the middle,

00:58:19.815 --> 00:58:22.275
so you can blow air through it.

00:58:22.275 --> 00:58:27.215
You then collect a
big gob of glass.

00:58:27.215 --> 00:58:28.355
Big chunk of glass.

00:58:28.355 --> 00:58:31.195
They call it big "gob" of glass.

00:58:31.195 --> 00:58:34.075
You start to see these
specific terms that people use

00:58:34.075 --> 00:58:37.035
when it comes to glassmaking.

00:58:37.035 --> 00:58:40.155
This big gob of glass,
then you collect it

00:58:40.155 --> 00:58:42.595
at the end of the blowpipe.

00:58:42.595 --> 00:58:45.675
You can blow air bubble
into it, and you essentially

00:58:45.675 --> 00:58:48.635
push it against the mold
as you blow air into it.

00:58:48.635 --> 00:58:51.115
So then the outer
surface of the gob

00:58:51.115 --> 00:58:53.555
then conforms to the
shape of the mold.

00:58:53.555 --> 00:58:57.995
And this kind of structure is
essentially a specific shape,

00:58:57.995 --> 00:59:00.475
and the air bubble in
it is called a parison.

00:59:00.475 --> 00:59:03.975


00:59:03.975 --> 00:59:05.735
Sort of like pre-formed glass.

00:59:05.735 --> 00:59:08.995
But that's the name people use
to call this kind of thing.

00:59:08.995 --> 00:59:12.175
And then you can
then open it up,

00:59:12.175 --> 00:59:14.535
or you can take it to
different kinds of geometries.

00:59:14.535 --> 00:59:15.910
There are a ton
of different ways

00:59:15.910 --> 00:59:18.475
people can modify this
process to create all kinds

00:59:18.475 --> 00:59:19.875
of delicate glass products.

00:59:19.875 --> 00:59:21.475
You're going to see that
at the MIT Glass Lab.

00:59:21.475 --> 00:59:22.933
It's a lot easier
for you to see it

00:59:22.933 --> 00:59:26.035
than for me trying to
explain it, what happens.

00:59:26.035 --> 00:59:28.355
The person that actually
does the blowing job

00:59:28.355 --> 00:59:30.482
is called a gaffer.

00:59:30.482 --> 00:59:32.815
Depending on context, if you
go to different industries,

00:59:32.815 --> 00:59:34.275
"gaffer" means different things.

00:59:34.275 --> 00:59:36.155
If you go to a movie
industry, "gaffer"

00:59:36.155 --> 00:59:38.075
is the person in
charge of lighting.

00:59:38.075 --> 00:59:40.475
If you go to a
sports team, "gaffer"

00:59:40.475 --> 00:59:42.595
means the manager of the club.

00:59:42.595 --> 00:59:45.235
If you go to glass, "gaffer" is
the person actually in charge

00:59:45.235 --> 00:59:45.975
of glassblowing.

00:59:45.975 --> 00:59:47.395
So it's basically like the boss.

00:59:47.395 --> 00:59:48.155
OK?

00:59:48.155 --> 00:59:51.715
So now, this job for
gaffer, or glass worker,

00:59:51.715 --> 00:59:53.758
it's very much a
technical know-how

00:59:53.758 --> 00:59:55.675
that's usually passed
down from one generation

00:59:55.675 --> 00:59:57.135
to another of glass makers.

00:59:57.135 --> 00:59:59.935
You don't learn it
from a textbook.

00:59:59.935 --> 01:00:01.635
You learn it by
actually doing it.

01:00:01.635 --> 01:00:06.395
And even up till today,
there's still demand,

01:00:06.395 --> 01:00:09.775
a lot of demand into making
specialized glass products.

01:00:09.775 --> 01:00:11.755
These are some
scientific instruments

01:00:11.755 --> 01:00:14.715
that people make using this
kind of glassblowing process.

01:00:14.715 --> 01:00:16.895
And you just cannot
simply buy it.

01:00:16.895 --> 01:00:18.615
This is not
off-the-shelf products.

01:00:18.615 --> 01:00:21.715
You need to have a good glass
blower that can actually create

01:00:21.715 --> 01:00:23.215
these kinds of things for you.

01:00:23.215 --> 01:00:26.392
So it's a very skill.

01:00:26.392 --> 01:00:27.475
It requires lots of skill.

01:00:27.475 --> 01:00:30.035
It's a job that requires a
lot of skill and experience

01:00:30.035 --> 01:00:31.555
to design the process.

01:00:31.555 --> 01:00:33.155
OK.

01:00:33.155 --> 01:00:33.795
OK.

01:00:33.795 --> 01:00:37.075
So this is the more
manual process here.

01:00:37.075 --> 01:00:39.035
Now, at the industrial
scale, people also

01:00:39.035 --> 01:00:41.475
use a similar blowing
process to create,

01:00:41.475 --> 01:00:43.795
in this case, glass containers.

01:00:43.795 --> 01:00:46.260
So what I'm showing you
here, top and bottom, these

01:00:46.260 --> 01:00:47.635
are two kinds of
glass containers

01:00:47.635 --> 01:00:49.835
that are created using
this similar process here.

01:00:49.835 --> 01:00:52.595
On top, this is for
narrow-neck containers--

01:00:52.595 --> 01:00:55.415
like beverage bottles.

01:00:55.415 --> 01:00:59.635
So for beverage bottle, you do a
so-called blow-and-blow process.

01:00:59.635 --> 01:01:03.235
So what you do, you start
from a gob of glass.

01:01:03.235 --> 01:01:06.015
You put it into a mold.

01:01:06.015 --> 01:01:08.995
You have an opening at one end
where you use compressed air

01:01:08.995 --> 01:01:10.755
to blow this mold.

01:01:10.755 --> 01:01:14.635
And then it's going to conform
to this closed mold here.

01:01:14.635 --> 01:01:17.675
Then you get this parison.

01:01:17.675 --> 01:01:20.475
You then put this parison
into a second mold,

01:01:20.475 --> 01:01:24.515
you blow compressed air again
to go to final geometry.

01:01:24.515 --> 01:01:26.395
Now, you may wonder,
why bother doing

01:01:26.395 --> 01:01:27.655
this two-step blowing process?

01:01:27.655 --> 01:01:29.795
Why not just simply
using one-step blowing?

01:01:29.795 --> 01:01:32.435
The reason for that
is a couple of things.

01:01:32.435 --> 01:01:34.435
One is that people
want to minimize

01:01:34.435 --> 01:01:36.675
any kind of very large
local shear strain,

01:01:36.675 --> 01:01:39.955
because those can cause
strain accumulation

01:01:39.955 --> 01:01:42.395
and maybe leading to
cracking of the products.

01:01:42.395 --> 01:01:44.815
So by designing this
two-step process,

01:01:44.815 --> 01:01:49.035
they can separate the
engineer, the deformation

01:01:49.035 --> 01:01:50.515
in the two blowing steps there.

01:01:50.515 --> 01:01:52.878
And this minimizes local
strain accumulation.

01:01:52.878 --> 01:01:53.795
Makes it more uniform.

01:01:53.795 --> 01:01:58.155
Makes the deformation
more uniform, basically.

01:01:58.155 --> 01:02:00.195
Now, the other thing
also to consider here

01:02:00.195 --> 01:02:02.275
is that this blowing
process takes place

01:02:02.275 --> 01:02:04.795
at fairly high temperature,
where glass is actually liquid,

01:02:04.795 --> 01:02:05.295
right?

01:02:05.295 --> 01:02:09.795
It also involves quite a bit of
friction between this glass gob

01:02:09.795 --> 01:02:12.275
and the mold you're using there.

01:02:12.275 --> 01:02:14.675
So this friction is
harmful because it leads

01:02:14.675 --> 01:02:16.675
to wear and tear of the mold.

01:02:16.675 --> 01:02:18.635
So in order to
ensure longevity, you

01:02:18.635 --> 01:02:22.795
want to make sure that this
glass deformation takes

01:02:22.795 --> 01:02:25.815
place ideally along the surface
normal direction of the mold.

01:02:25.815 --> 01:02:27.985
You try to avoid any
kind of shear motion

01:02:27.985 --> 01:02:30.185
on the surface of the mold.

01:02:30.185 --> 01:02:33.925
So having a two-step process
also helps with that.

01:02:33.925 --> 01:02:35.465
It gives you more
degree of freedom

01:02:35.465 --> 01:02:39.065
to engineer the process to
minimize any kind of wear

01:02:39.065 --> 01:02:42.145
and tear due to the shearing
motion between the glass gob

01:02:42.145 --> 01:02:44.865
and the glass
parison and the mold.

01:02:44.865 --> 01:02:48.825
Now, at the bottom here,
this is for wide-open jars,

01:02:48.825 --> 01:02:51.305
like [? Mark's. ?]
And this is done

01:02:51.305 --> 01:02:53.445
using a so-called
press-and-blow process.

01:02:53.445 --> 01:02:54.405
It's quite similar.

01:02:54.405 --> 01:02:56.965
You replace the first
compressed air blowing process

01:02:56.965 --> 01:02:58.485
with just a mechanical mold.

01:02:58.485 --> 01:03:00.985
You use a mechanical
press to essentially form

01:03:00.985 --> 01:03:04.545
the parison shape rather
than using a blowing process.

01:03:04.545 --> 01:03:05.705
OK.

01:03:05.705 --> 01:03:06.225
OK.

01:03:06.225 --> 01:03:09.865
So what I will do
next is I will again

01:03:09.865 --> 01:03:12.505
show you another video
that actually illustrates

01:03:12.505 --> 01:03:14.045
this glassblowing process.

01:03:14.045 --> 01:03:16.425
Hopefully no ad this time.

01:03:16.425 --> 01:03:17.945
[VIDEO PLAYBACK]

01:03:17.945 --> 01:03:19.245
- Why choose glass?

01:03:19.245 --> 01:03:22.464


01:03:22.464 --> 01:03:25.384
The benefits of
glass [INAUDIBLE].

01:03:25.384 --> 01:03:27.545
It is sustainable.

01:03:27.545 --> 01:03:29.985
It is virtually inert.

01:03:29.985 --> 01:03:36.145
It is 100% recyclable,
reusable, and refillable.

01:03:36.145 --> 01:03:38.545
It is safe to store
food and drinks in it,

01:03:38.545 --> 01:03:42.945
and it is a premium way
to market your products.

01:03:42.945 --> 01:03:46.985
Glass containers do not contain
phthalates, or biphenyl-A,

01:03:46.985 --> 01:03:50.504
or any other potential hormone
disruptors or carcinogens.

01:03:50.504 --> 01:03:53.464
It offers long shelf life
as high levels of vacuum

01:03:53.464 --> 01:03:55.785
and carbonization
can be maintained,

01:03:55.785 --> 01:03:58.344
and it can protect the product
from the harmful effects

01:03:58.344 --> 01:03:59.865
of light.

01:03:59.865 --> 01:04:01.665
Manufacturing process.

01:04:01.665 --> 01:04:03.665
Where do we start?

01:04:03.665 --> 01:04:07.145
Stage one, raw materials.

01:04:07.145 --> 01:04:10.625
The major raw materials used
in glass container production--

01:04:10.625 --> 01:04:12.705
PROFESSOR: The same
as the flat glass.

01:04:12.705 --> 01:04:14.145
Mixing different elements.

01:04:14.145 --> 01:04:15.318
- --limestone.

01:04:15.318 --> 01:04:16.485
PROFESSOR: Batching process.

01:04:16.485 --> 01:04:17.818
This is called batching process.

01:04:17.818 --> 01:04:19.785
You mix different
elements together.

01:04:19.785 --> 01:04:20.304
- --cullet.

01:04:20.304 --> 01:04:22.504
Cullet is recycled
glass that helps

01:04:22.504 --> 01:04:24.945
to reduce the amount
of raw materials needed

01:04:24.945 --> 01:04:26.625
to make new glass.

01:04:26.625 --> 01:04:29.105
Cullet provides an
important energy savings

01:04:29.105 --> 01:04:30.865
by lowering the
temperature required

01:04:30.865 --> 01:04:34.065
for the melting process,
helping to reduce greenhouse gas

01:04:34.065 --> 01:04:35.825
emissions.

01:04:35.825 --> 01:04:38.464
Stage two, batching.

01:04:38.464 --> 01:04:42.424
One, raw materials are
delivered into the batch house,

01:04:42.424 --> 01:04:44.065
where they are
visually inspected

01:04:44.065 --> 01:04:48.025
prior to being stored in silos.

01:04:48.025 --> 01:04:52.785
Two, it is then weighed
and mixed to produce batch.

01:04:52.785 --> 01:04:55.565
Three, the batch is then
delivered and stored

01:04:55.565 --> 01:04:59.865
at the furnace, where it
is melted to form glass.

01:04:59.865 --> 01:05:02.085
Stage three, melting.

01:05:02.085 --> 01:05:05.785


01:05:05.785 --> 01:05:09.345
The batch is continuously fed
into one end of the furnace

01:05:09.345 --> 01:05:11.905
via a batch charger.

01:05:11.905 --> 01:05:15.265
A combustible mixture of
natural gas and air is forced

01:05:15.265 --> 01:05:18.585
into the furnace to generate
a temperature of approximately

01:05:18.585 --> 01:05:23.465
1,565 degrees Celsius.

01:05:23.465 --> 01:05:26.445
It is here that the batch
is melted to produce glass.

01:05:26.445 --> 01:05:33.145


01:05:33.145 --> 01:05:35.125
Stage four, feeders.

01:05:35.125 --> 01:05:38.265


01:05:38.265 --> 01:05:40.545
At the end of the
forearm, the glass

01:05:40.545 --> 01:05:42.105
is then forced
through an opening

01:05:42.105 --> 01:05:45.245
in the bottom of the spout
called an orifice ring.

01:05:45.245 --> 01:05:50.072
The glass is then cut
by blades called shears.

01:05:50.072 --> 01:05:52.405
PROFESSOR: Those are [INAUDIBLE]
the technologies that--

01:05:52.405 --> 01:05:56.705
- --elongated cylinders
of glass called gobs.

01:05:56.705 --> 01:05:59.025
Each gob has the
correct amount of glass

01:05:59.025 --> 01:06:00.545
to produce one container.

01:06:00.545 --> 01:06:05.705
At this stage, they are ready to
be formed into glass containers

01:06:05.705 --> 01:06:07.845
and are delivered to
the forming machine.

01:06:07.845 --> 01:06:10.385


01:06:10.385 --> 01:06:13.025
Stage five, forming.

01:06:13.025 --> 01:06:15.345
The forming machine
uses compressed air

01:06:15.345 --> 01:06:17.385
to transform the
gobs into containers

01:06:17.385 --> 01:06:19.505
in a two-step process.

01:06:19.505 --> 01:06:22.545
The gob is delivered to the
blank side of the machine, where

01:06:22.545 --> 01:06:24.465
it is formed into
a parison, which

01:06:24.465 --> 01:06:26.245
is a blank form of a container.

01:06:26.245 --> 01:06:28.265
It is then transferred
onto the mold side

01:06:28.265 --> 01:06:30.385
of the machine,
where it is blown

01:06:30.385 --> 01:06:32.445
into the final shape
of the container.

01:06:32.445 --> 01:06:35.705


01:06:35.705 --> 01:06:39.065
Stage six, hot end coating.

01:06:39.065 --> 01:06:42.545
Tin coating.

01:06:42.545 --> 01:06:46.105
Once formed, the containers are
sent through a hot end coating

01:06:46.105 --> 01:06:48.945
hood and are covered
with a thin layer of tin.

01:06:48.945 --> 01:06:52.585
This is part one of a
permanent coating process.

01:06:52.585 --> 01:06:54.685
Stage seven, annealing.

01:06:54.685 --> 01:06:58.145


01:06:58.145 --> 01:07:00.505
When a new glass
container is formed,

01:07:00.505 --> 01:07:04.465
the outside surface cools more
rapidly than the inside surface,

01:07:04.465 --> 01:07:06.325
causing stresses in the glass.

01:07:06.325 --> 01:07:09.345
The process to relieve these
stresses is called anneal

01:07:09.345 --> 01:07:11.425
and is done in a lehr.

01:07:11.425 --> 01:07:15.345
The lehr reheats the containers
and gradually and evenly cools

01:07:15.345 --> 01:07:16.805
them to room temperature.

01:07:16.805 --> 01:07:19.065
The time taken to
travel through the lehr

01:07:19.065 --> 01:07:21.695
depends on the container,
which, on average, is

01:07:21.695 --> 01:07:24.895
between 60 and 80 minutes.

01:07:24.895 --> 01:07:27.935
Stage eight, hold and spray.

01:07:27.935 --> 01:07:30.895
Duracote.

01:07:30.895 --> 01:07:33.375
At this stage, a
thin layer of polymer

01:07:33.375 --> 01:07:35.615
is applied to the
container to provide them

01:07:35.615 --> 01:07:39.055
with a scratch-resistant
coating.

01:07:39.055 --> 01:07:41.295
This coating will also
allow the containers

01:07:41.295 --> 01:07:45.655
to travel along the customer's
filling line without jamming up.

01:07:45.655 --> 01:07:48.195
PROFESSOR: It's like a
lubricant, reduce the friction.

01:07:48.195 --> 01:07:50.655
- --Inspection.

01:07:50.655 --> 01:07:53.215
Before packaging,
each glass container

01:07:53.215 --> 01:07:55.655
undergoes a rigorous
inspection process

01:07:55.655 --> 01:07:57.815
to ensure the highest quality.

01:07:57.815 --> 01:08:00.295
It is inspected at multiple
stages of the production

01:08:00.295 --> 01:08:01.335
process.

01:08:01.335 --> 01:08:05.615
Bottles that do not meet OIPJC's
stringent design specifications

01:08:05.615 --> 01:08:11.415
are rejected and
recycled as cullet.

01:08:11.415 --> 01:08:13.695
Finished containers are
prepared for shipment

01:08:13.695 --> 01:08:15.415
by our packaging department.

01:08:15.415 --> 01:08:17.335
From our warehouses,
our products

01:08:17.335 --> 01:08:20.551
are shipped all
over Southeast Asia.

01:08:20.551 --> 01:08:22.320
[END PLAYBACK]

01:08:22.320 --> 01:08:23.195
PROFESSOR: All right.

01:08:23.195 --> 01:08:24.903
So there are a few
things that's actually

01:08:24.903 --> 01:08:29.095
mentioned there in this process
that I just want to highlight.

01:08:29.095 --> 01:08:31.075
One is that they talk
about this tin coating.

01:08:31.075 --> 01:08:34.015
So the reason they
do a tin coating--

01:08:34.015 --> 01:08:35.375
there are two purposes there.

01:08:35.375 --> 01:08:40.235
One is that the tin oxide, when
the glass contains tin oxide,

01:08:40.235 --> 01:08:42.607
it has a higher hardness
compared to normal glass.

01:08:42.607 --> 01:08:44.815
And higher hardness is really
important in preventing

01:08:44.815 --> 01:08:46.427
any kind of scratch damage.

01:08:46.427 --> 01:08:48.635
So later on when we talk
about mechanical properties,

01:08:48.635 --> 01:08:50.895
you're going to find that
the Achilles heel of glass

01:08:50.895 --> 01:08:52.270
in terms of
mechanical properties

01:08:52.270 --> 01:08:53.500
is usually surface defects.

01:08:53.500 --> 01:08:54.875
If there are any
surface defects,

01:08:54.875 --> 01:08:55.958
they become the weak link.

01:08:55.958 --> 01:08:58.335
You can have cracks initiate
and propagate quickly

01:08:58.335 --> 01:08:59.435
from those defects.

01:08:59.435 --> 01:09:03.815
So that's why having a surface
treatment to empower glass

01:09:03.815 --> 01:09:05.995
with a higher scratch
resistance is quite important.

01:09:05.995 --> 01:09:07.370
That's why they
do a tin coating.

01:09:07.370 --> 01:09:08.935
One purpose.

01:09:08.935 --> 01:09:12.035
Now, second purpose is that
if you think about glass,

01:09:12.035 --> 01:09:14.915
its surface is-- if it's not
treated, it's hydrophilic.

01:09:14.915 --> 01:09:18.415
That's why you see,
during rain, raindrops

01:09:18.415 --> 01:09:20.455
can hang on glass windows.

01:09:20.455 --> 01:09:24.148
But after you have done
the hot end coding,

01:09:24.148 --> 01:09:25.815
we're trying to do
the cold end coating.

01:09:25.815 --> 01:09:27.415
That's a polymer
coating, duracote,

01:09:27.415 --> 01:09:32.215
to reduce friction and
also further improve

01:09:32.215 --> 01:09:33.455
scratch resistance.

01:09:33.455 --> 01:09:37.015
So the polymer coating
is hydrophobic,

01:09:37.015 --> 01:09:40.295
so it doesn't stick very
well to the glass surface.

01:09:40.295 --> 01:09:42.415
So by adding tin
there, tin actually

01:09:42.415 --> 01:09:44.815
renders a more
hydrophobic surface,

01:09:44.815 --> 01:09:48.495
which helps improve adhesion
between the subsequent polymer

01:09:48.495 --> 01:09:50.175
coating with the glass surface.

01:09:50.175 --> 01:09:53.295
So that's why, actually, people
do this tin coating there,

01:09:53.295 --> 01:09:54.735
just to serve their purpose.

01:09:54.735 --> 01:09:56.615
OK?

01:09:56.615 --> 01:09:57.115
Yes?

01:09:57.115 --> 01:09:59.095
AUDIENCE: Do they have
to remove the coating

01:09:59.095 --> 01:10:00.875
when they recycle the glass?

01:10:00.875 --> 01:10:04.015
Or do you just get--

01:10:04.015 --> 01:10:08.455
or maybe that concentration
of tin becomes negligible?

01:10:08.455 --> 01:10:10.610
PROFESSOR: Great question.

01:10:10.610 --> 01:10:12.235
There are actually
two coatings, right?

01:10:12.235 --> 01:10:14.655
There's also-- well, it depends.

01:10:14.655 --> 01:10:16.415
When you recycle
glass products, there

01:10:16.415 --> 01:10:18.135
is a thin layer of
polymer coatings.

01:10:18.135 --> 01:10:21.255
Those things just burn
away in the glass furnace.

01:10:21.255 --> 01:10:22.495
Becomes carbon dioxide.

01:10:22.495 --> 01:10:24.755
The tin, because it's
only on the surface,

01:10:24.755 --> 01:10:26.935
it's a very, very small
amount, so it barely

01:10:26.935 --> 01:10:27.915
impacts the process.

01:10:27.915 --> 01:10:29.775
So that's why they can
go back immediately.

01:10:29.775 --> 01:10:31.358
It's only a very
thin layer of coating

01:10:31.358 --> 01:10:33.995
there, like a few
hundred nanometers.

01:10:33.995 --> 01:10:37.615
So if you look at the actual
fraction, it's negligible.

01:10:37.615 --> 01:10:38.215
Yes?

01:10:38.215 --> 01:10:40.715
AUDIENCE: Is it
possible-- so I mean,

01:10:40.715 --> 01:10:43.455
I guess the primary way that
they kind of recycle glass

01:10:43.455 --> 01:10:45.535
is by using it as cullet,
which is basically,

01:10:45.535 --> 01:10:47.775
like, 30% of the [INAUDIBLE].

01:10:47.775 --> 01:10:53.215
So is it not possible to do the
full recycling of all produced

01:10:53.215 --> 01:10:53.935
glassware?

01:10:53.935 --> 01:10:54.975
PROFESSOR: You could.

01:10:54.975 --> 01:10:55.495
Yeah.

01:10:55.495 --> 01:10:57.593
If you have enough stream,
a stable supply stream

01:10:57.593 --> 01:10:59.135
of weight glasses,
then in principle,

01:10:59.135 --> 01:11:00.075
yeah, you can do that.

01:11:00.075 --> 01:11:00.575
Yeah.

01:11:00.575 --> 01:11:03.155
Yeah, it's a supply
chain problem.

01:11:03.155 --> 01:11:04.755
It's not about the
technology problem.

01:11:04.755 --> 01:11:04.975
Yeah.

01:11:04.975 --> 01:11:06.600
And that's one
advantage about glasses.

01:11:06.600 --> 01:11:09.495
Glasses, in principle,
they are fully recyclable,

01:11:09.495 --> 01:11:13.395
unlike other kind of products.

01:11:13.395 --> 01:11:15.197
The glass industry
is very power-hungry

01:11:15.197 --> 01:11:17.655
because it takes lots of energy
to actually melt the glass.

01:11:17.655 --> 01:11:19.335
So being able to
recycle is quite

01:11:19.335 --> 01:11:22.215
important in reducing carbon
footprint of the industry.

01:11:22.215 --> 01:11:22.853
Yeah.

01:11:22.853 --> 01:11:24.895
Now, the second thing I
also want to mention here

01:11:24.895 --> 01:11:26.603
is that for this kind
of container glass,

01:11:26.603 --> 01:11:29.295
they also have to go through a
process called dealkalization.

01:11:29.295 --> 01:11:33.455


01:11:33.455 --> 01:11:37.635
So these are soda-lime glasses.

01:11:37.635 --> 01:11:39.423
They have sodium.

01:11:39.423 --> 01:11:41.215
We have mentioned that
sodium-- the problem

01:11:41.215 --> 01:11:43.675
is that when it comes into
contact with aqueous solutions,

01:11:43.675 --> 01:11:46.415
sodium tends to
slowly leach out.

01:11:46.415 --> 01:11:49.895
And when they leach out,
they become sodium hydroxide.

01:11:49.895 --> 01:11:52.595
So it turns the
solution more basic.

01:11:52.595 --> 01:11:55.615
It also changes the taste
of the beverage over time.

01:11:55.615 --> 01:11:58.775
So that's why, for
these container glasses,

01:11:58.775 --> 01:12:06.975
they tend to add
basically SO2 or SO3

01:12:06.975 --> 01:12:08.595
at the end of the process.

01:12:08.595 --> 01:12:12.575
And then they let them react
with the sodium on the glass

01:12:12.575 --> 01:12:20.152
surface, which then turns them
basically into sodium sulfate.

01:12:20.152 --> 01:12:22.735
Sodium sulfate is water-soluble,
so you can just simply remove

01:12:22.735 --> 01:12:26.055
it using the rinsing process.

01:12:26.055 --> 01:12:28.055
Now, practically, the
way they implement it,

01:12:28.055 --> 01:12:31.135
instead of using glass--
sorry, using gas here,

01:12:31.135 --> 01:12:33.175
this gas is a pollutant,
so they don't want

01:12:33.175 --> 01:12:34.515
to really use the gas phase.

01:12:34.515 --> 01:12:40.215
They just put something like
ammonia sulfate, for example.

01:12:40.215 --> 01:12:41.475
So it's a solid material.

01:12:41.475 --> 01:12:44.058
You put a little bit into the
glass container, you heat it up,

01:12:44.058 --> 01:12:45.595
it decomposes,
forms these gases.

01:12:45.595 --> 01:12:47.555
In the process, you seal
the glass containers,

01:12:47.555 --> 01:12:49.847
so you don't have any kind
of leakage of the pollutants

01:12:49.847 --> 01:12:50.935
into the atmosphere.

01:12:50.935 --> 01:12:52.793
After you finish the
reaction, you rinse it

01:12:52.793 --> 01:12:54.335
and then you actually
get the surface

01:12:54.335 --> 01:12:56.815
that's depleted of sodium.

01:12:56.815 --> 01:13:00.415
So that's also how you can help
improve the chemical resistance

01:13:00.415 --> 01:13:04.455
for container glasses, OK?

01:13:04.455 --> 01:13:04.955
OK.

01:13:04.955 --> 01:13:06.855
We are right on time here.

01:13:06.855 --> 01:13:10.073
Any last-minute questions
before we conclude?

01:13:10.073 --> 01:13:10.615
AUDIENCE: OK.

01:13:10.615 --> 01:13:12.085
I wanted to ask.

01:13:12.085 --> 01:13:15.645
So since the cullets,
they could come

01:13:15.645 --> 01:13:18.405
from different types
of glasses, did

01:13:18.405 --> 01:13:20.285
they separate the different--

01:13:20.285 --> 01:13:21.623
PROFESSOR: They have to.

01:13:21.623 --> 01:13:22.165
They have to.

01:13:22.165 --> 01:13:24.133
I mean, if it's soda-lime--
most of soda-lime

01:13:24.133 --> 01:13:26.425
are more or less the same,
so it doesn't really matter.

01:13:26.425 --> 01:13:28.092
But you really cannot
mix-- for example,

01:13:28.092 --> 01:13:30.545
if you have borosilicate
glasses, no,

01:13:30.545 --> 01:13:32.225
it really cannot go
into the same batch.

01:13:32.225 --> 01:13:33.825
It has to be separated.

01:13:33.825 --> 01:13:34.325
Yeah.

01:13:34.325 --> 01:13:36.325
Or if you have
silica glass, that's

01:13:36.325 --> 01:13:37.985
creating more problems
than it solves.

01:13:37.985 --> 01:13:39.445
[LAUGHS] Yeah.

01:13:39.445 --> 01:13:41.385
So you do have to sort
glasses, essentially.

01:13:41.385 --> 01:13:41.885
Yeah.

01:13:41.885 --> 01:13:44.645
And if they also have any
kind of specialized coatings

01:13:44.645 --> 01:13:46.662
or any kind of
contaminants or whatever,

01:13:46.662 --> 01:13:49.245
they all have to be sorted out
before you go into that process

01:13:49.245 --> 01:13:49.785
there.

01:13:49.785 --> 01:13:50.765
Yeah.

01:13:50.765 --> 01:13:54.605
So that's why the best cullet
is the factory's own product.

01:13:54.605 --> 01:13:56.405
It's exactly the same.

01:13:56.405 --> 01:13:58.385
Any kind of a product
that doesn't qualify,

01:13:58.385 --> 01:14:00.365
it doesn't meet the
quality standard

01:14:00.365 --> 01:14:01.865
just goes back as cullets.

01:14:01.865 --> 01:14:02.365
Yeah.

01:14:02.365 --> 01:14:05.445


01:14:05.445 --> 01:14:06.245
OK.

01:14:06.245 --> 01:14:07.045
All right.

01:14:07.045 --> 01:14:09.035
Thank you very much.

01:14:09.035 --> 01:15:10.275