WEBVTT

00:00:00.000 --> 00:00:08.280
JOEL LEWIS: Hi.

00:00:08.280 --> 00:00:09.840
Welcome back to recitation.

00:00:09.840 --> 00:00:12.380
In lecture, you've been
learning about flux and surface

00:00:12.380 --> 00:00:14.230
integrals in the
divergence theorem,

00:00:14.230 --> 00:00:16.720
and I have a nice problem
about that for you here.

00:00:16.720 --> 00:00:21.620
So I've got this field F, and
it's a little bit ugly, right?

00:00:21.620 --> 00:00:22.560
All right.

00:00:22.560 --> 00:00:27.210
So its coordinates
are x to the fourth y,

00:00:27.210 --> 00:00:31.890
minus 2 x cubed y
squared, and z squared.

00:00:31.890 --> 00:00:35.750
And it's passing through the
surface of a solid that's

00:00:35.750 --> 00:00:39.075
bounded by the plane z
equals 0, by the plane

00:00:39.075 --> 00:00:43.990
z equals h, and by the surface
x squared plus y squared equals

00:00:43.990 --> 00:00:44.510
R squared.

00:00:44.510 --> 00:00:48.400
So often we call this
solid a cylinder.

00:00:48.400 --> 00:00:51.730
So it's got its bottom surface
in the plane z equals 0,

00:00:51.730 --> 00:00:53.800
and its top surface in
the plane z equals h,

00:00:53.800 --> 00:00:58.830
and it's got a circular
base with radius R there.

00:00:58.830 --> 00:01:02.920
So what I'd like you to do is to
compute the flux of this field

00:01:02.920 --> 00:01:06.040
F through this cylinder.

00:01:06.040 --> 00:01:10.170
So I'll point out
before I let you

00:01:10.170 --> 00:01:13.920
at it, that to compute
this as a surface integral,

00:01:13.920 --> 00:01:15.230
you could do it.

00:01:15.230 --> 00:01:15.980
You could do it.

00:01:15.980 --> 00:01:19.590
If you really want an
exercise in nasty arithmetic,

00:01:19.590 --> 00:01:21.090
I invite you to do it.

00:01:21.090 --> 00:01:23.040
But you might be able
to think of a way

00:01:23.040 --> 00:01:26.700
to do this that requires less
effort than parametrizing

00:01:26.700 --> 00:01:29.420
the three surfaces and
integrating and so on.

00:01:29.420 --> 00:01:30.620
So I'll leave you with that.

00:01:30.620 --> 00:01:35.129
Why don't you pause the video,
work this one out, come back,

00:01:35.129 --> 00:01:36.420
and we can work on it together.

00:01:44.930 --> 00:01:47.640
Hopefully, you had some luck
working on this problem.

00:01:47.640 --> 00:01:52.140
Right before I left, I mentioned
that you were computing a flux

00:01:52.140 --> 00:01:55.080
through a surface here, but that
doing it as a surface integral

00:01:55.080 --> 00:01:56.750
is maybe not the best way to go.

00:01:56.750 --> 00:02:00.500
And so, even without that
hint, probably many of you

00:02:00.500 --> 00:02:02.440
realized that really
the way that we

00:02:02.440 --> 00:02:05.680
want to go about this problem
is with the divergence theorem.

00:02:05.680 --> 00:02:06.190
OK.

00:02:06.190 --> 00:02:12.470
So in our case, the
divergence theorem--

00:02:12.470 --> 00:02:16.000
I'm just abbreviating
it div T-H-M here--

00:02:16.000 --> 00:02:22.920
says that the double integral
over the surface of F dot

00:02:22.920 --> 00:02:31.220
n d surface area-- so S here
is the surface of this solid.

00:02:31.220 --> 00:02:35.780
So the divergence theorem says
that this surface integral,

00:02:35.780 --> 00:02:38.020
which is the flux that
we're interested in,

00:02:38.020 --> 00:02:43.020
is equal to the triple integral
over the solid region D--

00:02:43.020 --> 00:02:44.640
so that's bounded
by the surface,

00:02:44.640 --> 00:02:46.930
and so that's the
solid cylinder here--

00:02:46.930 --> 00:02:54.620
is equal to the triple
integral over D of div F dV.

00:02:54.620 --> 00:02:55.370
OK.

00:02:55.370 --> 00:02:59.490
So in our case, this is
nice, because in fact,

00:02:59.490 --> 00:03:03.100
this solid region D is
an easier to understand,

00:03:03.100 --> 00:03:06.514
or easier to grapple with
region than the surface

00:03:06.514 --> 00:03:07.680
that we started with, right?

00:03:07.680 --> 00:03:10.200
It's just one solid piece.

00:03:10.200 --> 00:03:11.910
It's easy to
parametrize, in fact.

00:03:11.910 --> 00:03:14.350
It's easy to
describe, especially

00:03:14.350 --> 00:03:15.870
in cylindrical
coordinates, but also

00:03:15.870 --> 00:03:17.481
in rectangular coordinates.

00:03:17.481 --> 00:03:19.730
Whereas this surface S, if
we wanted to talk about it,

00:03:19.730 --> 00:03:21.580
we'd need to split it
up into three pieces,

00:03:21.580 --> 00:03:23.010
and we'd need to parametrize it.

00:03:23.010 --> 00:03:26.010
And it's kind of a hassle,
relatively speaking.

00:03:26.010 --> 00:03:29.180
Also, the divergence
of this field F

00:03:29.180 --> 00:03:31.470
is a lot simpler than
the field itself.

00:03:31.470 --> 00:03:35.240
If we go and look at this
field, all of its components

00:03:35.240 --> 00:03:36.930
are polynomials.

00:03:36.930 --> 00:03:38.690
To compute its
divergence, we take

00:03:38.690 --> 00:03:40.260
derivatives of all of them.

00:03:40.260 --> 00:03:42.370
And so that makes
their degrees lower,

00:03:42.370 --> 00:03:44.310
and then we just add them.

00:03:44.310 --> 00:03:47.330
Life is a little bit simpler.

00:03:47.330 --> 00:03:48.990
So OK.

00:03:48.990 --> 00:03:51.730
So this process, using
the divergence theorem,

00:03:51.730 --> 00:03:54.690
is going to make
our lives easier.

00:03:54.690 --> 00:03:57.590
It's going to make
this nasty surface

00:03:57.590 --> 00:04:00.860
integral into an easy to
compute triple integral.

00:04:00.860 --> 00:04:02.580
So let's see
actually how it does.

00:04:02.580 --> 00:04:05.210
So let's compute div F first.

00:04:05.210 --> 00:04:06.630
So we know what
the integrand is.

00:04:09.500 --> 00:04:10.490
All right.

00:04:10.490 --> 00:04:13.360
So we need to look at
the components of F,

00:04:13.360 --> 00:04:16.120
and so we need to take the
partial of the first one

00:04:16.120 --> 00:04:17.140
with respect to x.

00:04:17.140 --> 00:04:19.620
So that's x to the fourth
y with respect to x.

00:04:19.620 --> 00:04:21.240
So put that down over here.

00:04:21.240 --> 00:04:24.090
That's 4 x cubed y.

00:04:24.090 --> 00:04:25.766
We just treat y as a constant.

00:04:25.766 --> 00:04:28.390
OK, so now, we come back and we
need to look at the second one.

00:04:28.390 --> 00:04:31.000
So it's minus 2 x
cubed y squared.

00:04:31.000 --> 00:04:32.000
And it's the second one.

00:04:32.000 --> 00:04:34.260
We take its partial
with respect to y.

00:04:34.260 --> 00:04:34.780
So OK.

00:04:34.780 --> 00:04:38.300
So that's going to be
minus 2 x cubed times 2y.

00:04:38.300 --> 00:04:43.360
So that's going to
be minus 4 x cubed y.

00:04:43.360 --> 00:04:46.320
And then we come back and we
look at the last component.

00:04:46.320 --> 00:04:47.670
And that's z squared.

00:04:47.670 --> 00:04:50.090
And so we need to take its
partial with respect to z.

00:04:50.090 --> 00:04:53.000
So in this case, that's just 2z,
and so we add that on as well.

00:04:53.000 --> 00:04:55.110
Plus 2z.

00:04:55.110 --> 00:04:59.370
And in this case, not only
are these polynomials simpler

00:04:59.370 --> 00:05:03.512
than the coordinates of F
that we had, but in fact,

00:05:03.512 --> 00:05:04.970
we've got some
simplification here.

00:05:04.970 --> 00:05:07.600
Life gets really, really simple.

00:05:07.600 --> 00:05:10.720
So in fact, this is just
going to work out to 2z.

00:05:10.720 --> 00:05:13.390
So the divergence here
is very simple compared

00:05:13.390 --> 00:05:17.210
with the function F. More simple
than we have a right to expect,

00:05:17.210 --> 00:05:19.980
but in any case, good.

00:05:19.980 --> 00:05:22.830
It's nice to work with.

00:05:22.830 --> 00:05:23.369
OK.

00:05:23.369 --> 00:05:24.410
So that's the divergence.

00:05:27.370 --> 00:05:29.320
So I'm going to write,
this is the flux.

00:05:29.320 --> 00:05:31.670
These integrals that
we're interested in.

00:05:31.670 --> 00:05:34.667
This surface integral, and
then by the divergence theorem,

00:05:34.667 --> 00:05:36.250
it's the same as
this triple integral.

00:05:36.250 --> 00:05:39.130
So the divergence is this 2z.

00:05:39.130 --> 00:05:41.960
So the flux is what I get
when I just put that in here.

00:05:41.960 --> 00:05:50.580
So flux is equal to the
triple integral over our solid

00:05:50.580 --> 00:05:54.970
of 2z dV.

00:05:54.970 --> 00:05:57.810
OK, so I've left some
stuff out of this.

00:05:57.810 --> 00:06:01.630
Because I'm going to start
writing down the bounds

00:06:01.630 --> 00:06:05.570
and writing this down as
an iterated integral now.

00:06:05.570 --> 00:06:06.070
OK.

00:06:06.070 --> 00:06:09.520
So we have to choose some
coordinate system in which

00:06:09.520 --> 00:06:12.620
to integrate over this solid.

00:06:12.620 --> 00:06:15.630
And so we have three
kinds of natural choices

00:06:15.630 --> 00:06:17.230
that we always look back to.

00:06:17.230 --> 00:06:19.630
There are rectangular
coordinates and cylindrical

00:06:19.630 --> 00:06:21.970
coordinates and
spherical coordinates.

00:06:21.970 --> 00:06:25.700
So spherical coordinates seem
pretty clearly inappropriate.

00:06:25.700 --> 00:06:27.220
Rectangular and cylindrical?

00:06:27.220 --> 00:06:30.470
You know, you could try
and do it in rectangular.

00:06:30.470 --> 00:06:31.890
It's not horrible.

00:06:31.890 --> 00:06:33.950
But this is a cylinder, right?

00:06:33.950 --> 00:06:37.809
I mean, it's crying out for us
to use cylindrical coordinates.

00:06:37.809 --> 00:06:39.350
So let's use
cylindrical coordinates.

00:06:39.350 --> 00:06:41.266
So we're going to use
cylindrical coordinates.

00:06:41.266 --> 00:06:45.480
So to get dV we need a
z, an r, and a theta,

00:06:45.480 --> 00:06:47.800
but remember there's
this extra factor of r.

00:06:47.800 --> 00:06:55.980
So it's going to be 2z
times r dz dr d theta.

00:06:55.980 --> 00:06:56.480
Right?

00:06:56.480 --> 00:06:57.580
This is dV.

00:06:57.580 --> 00:07:00.950
This r dz dr d theta part.

00:07:00.950 --> 00:07:04.370
So that's what dV is when we
use cylindrical coordinates.

00:07:04.370 --> 00:07:06.840
OK, so now let's figure
out what the bounds are.

00:07:06.840 --> 00:07:10.635
So let's go look at the
cylinder that we had over here.

00:07:13.220 --> 00:07:17.375
So it's bounded between z
equals 0 at the bottom surface

00:07:17.375 --> 00:07:19.610
and z equals h at
the top surface.

00:07:19.610 --> 00:07:20.270
OK.

00:07:20.270 --> 00:07:22.380
So that's easy enough.

00:07:22.380 --> 00:07:23.980
That's what the bounds on z are.

00:07:23.980 --> 00:07:25.760
So let's put those in.

00:07:25.760 --> 00:07:33.151
So z is the innermost one,
so that's going from 0 to h.

00:07:33.151 --> 00:07:33.650
OK.

00:07:33.650 --> 00:07:34.790
How about the next one?

00:07:34.790 --> 00:07:36.000
So the next one is r.

00:07:36.000 --> 00:07:37.470
So let's go back over here.

00:07:37.470 --> 00:07:44.080
So r is the radius here
after we project it down.

00:07:44.080 --> 00:07:47.577
And we just get the circle
of radius big R centered

00:07:47.577 --> 00:07:48.160
at the origin.

00:07:48.160 --> 00:07:51.110
So little r is going
from 0 to big R.

00:07:51.110 --> 00:07:53.034
And theta is the circle.

00:07:53.034 --> 00:07:53.950
It's the whole circle.

00:07:53.950 --> 00:07:55.690
So theta is going
from 0 to 2*pi.

00:07:55.690 --> 00:07:58.640
So cylinders are
really easy to describe

00:07:58.640 --> 00:08:00.610
what they look like in
cylindrical coordinates.

00:08:00.610 --> 00:08:01.526
So let's put those in.

00:08:01.526 --> 00:08:06.000
So little r is going
from 0 to big R,

00:08:06.000 --> 00:08:09.990
and theta is going
from 0 to 2*pi.

00:08:12.720 --> 00:08:14.260
OK.

00:08:14.260 --> 00:08:15.270
Wonderful.

00:08:15.270 --> 00:08:17.480
Now we just have to
compute this, right?

00:08:17.480 --> 00:08:19.930
We've got our flux is
this triple integral.

00:08:19.930 --> 00:08:20.850
So let's compute it.

00:08:20.850 --> 00:08:23.990
Let's walk over to this little
bit of empty board space.

00:08:23.990 --> 00:08:25.830
OK, so we have an
iterated integral.

00:08:25.830 --> 00:08:28.280
So let's do it.

00:08:28.280 --> 00:08:34.680
So the inner integral is
the integral from 0 to h

00:08:34.680 --> 00:08:36.050
of 2*z*r*dz.

00:08:38.667 --> 00:08:39.750
Well, that's not that bad.

00:08:39.750 --> 00:08:42.690
That's equal to--
r is a constant.

00:08:42.690 --> 00:08:50.520
So it's equal to r z squared
as z goes between 0 and h.

00:08:50.520 --> 00:08:52.950
It's dz, so z is
going from 0 to h.

00:08:52.950 --> 00:08:59.190
So we plug in, and we just
get h squared r minus 0.

00:08:59.190 --> 00:09:00.650
So just h squared r.

00:09:00.650 --> 00:09:01.150
OK.

00:09:01.150 --> 00:09:04.740
So now let's do the
middle integral.

00:09:04.740 --> 00:09:09.980
So the middle integral is
the integral from 0 to big R

00:09:09.980 --> 00:09:13.550
d little r of the
inner integral.

00:09:13.550 --> 00:09:19.570
So this is the integral
from 0 to big R

00:09:19.570 --> 00:09:25.260
of the inner integral, which was
h squared little r, d little r.

00:09:25.260 --> 00:09:25.930
OK.

00:09:25.930 --> 00:09:27.221
And that's not that bad either.

00:09:27.221 --> 00:09:28.600
So h is just a constant.

00:09:28.600 --> 00:09:35.350
So this is equal to 1/2 h
squared r squared from r

00:09:35.350 --> 00:09:43.810
equals 0 to big R. And so that's
1/2 h squared big R squared.

00:09:43.810 --> 00:09:45.230
That's the middle integral.

00:09:45.230 --> 00:09:48.230
So the outer one now.

00:09:48.230 --> 00:09:48.982
OK.

00:09:48.982 --> 00:09:50.290
So let's go back and look.

00:09:50.290 --> 00:09:55.480
So we're doing d theta as theta
goes from 0 to 2*pi of whatever

00:09:55.480 --> 00:09:57.340
the middle integral was.

00:09:57.340 --> 00:10:03.225
So it's the integral from 0
to 2*pi of whatever the value

00:10:03.225 --> 00:10:04.350
of the middle integral was.

00:10:04.350 --> 00:10:10.110
So this is 1/2 h squared
big R squared d theta.

00:10:10.110 --> 00:10:12.670
And this is all just constant
with respect to theta.

00:10:12.670 --> 00:10:18.950
So that's going to be just
pi h squared r squared.

00:10:18.950 --> 00:10:20.970
You're just
multiplying it by 2*pi.

00:10:20.970 --> 00:10:21.470
All right.

00:10:21.470 --> 00:10:23.070
So pi h squared r squared.

00:10:23.070 --> 00:10:25.010
So this is our final answer.

00:10:25.010 --> 00:10:27.570
Let's just quickly
recap what we did.

00:10:27.570 --> 00:10:31.490
We had to compute the
flux of this field F

00:10:31.490 --> 00:10:37.190
through the surface
of a solid cylinder.

00:10:37.190 --> 00:10:39.260
And so we had options.

00:10:39.260 --> 00:10:42.520
We could do it directly by
trying to compute the surface

00:10:42.520 --> 00:10:44.450
integrals, but in
this case, life

00:10:44.450 --> 00:10:47.560
was a lot easier if we applied
the divergence theorem.

00:10:47.560 --> 00:10:50.280
So the divergence theorem
says that the flux-- which

00:10:50.280 --> 00:10:52.170
is equal to this
surface integral--

00:10:52.170 --> 00:10:55.120
can also be written as
the triple integral,

00:10:55.120 --> 00:10:59.020
over the solid region
bounded by the surface,

00:10:59.020 --> 00:11:01.490
of the divergence of the field.

00:11:01.490 --> 00:11:02.050
All right.

00:11:02.050 --> 00:11:03.540
And so in our case,
the divergence

00:11:03.540 --> 00:11:07.220
was very nice and simple,
and the solid region

00:11:07.220 --> 00:11:12.010
D was relatively simpler to
describe than its surface that

00:11:12.010 --> 00:11:15.790
bounds it, S. So this is why
we think of the divergence

00:11:15.790 --> 00:11:16.290
theorem.

00:11:16.290 --> 00:11:20.890
Because the divergence of the
field is easy to understand,

00:11:20.890 --> 00:11:24.184
and the solid is easier to
describe than its surface.

00:11:24.184 --> 00:11:25.600
So those are both
things that make

00:11:25.600 --> 00:11:27.500
us think to use the
divergence theorem

00:11:27.500 --> 00:11:29.090
for a problem like this.

00:11:29.090 --> 00:11:30.950
So then by the
divergence theorem,

00:11:30.950 --> 00:11:34.394
the flux is just
that triple integral,

00:11:34.394 --> 00:11:35.560
and so we wrote it out here.

00:11:35.560 --> 00:11:37.340
We were integrating
over a cylinder.

00:11:37.340 --> 00:11:40.530
So a natural thing to do is
use cylindrical coordinates.

00:11:40.530 --> 00:11:42.750
And then we computed
the triple integral

00:11:42.750 --> 00:11:44.340
just like we always do.

00:11:44.340 --> 00:11:46.697
I did it in three steps:
inner, middle, and outer.

00:11:46.697 --> 00:11:49.280
You don't have to do it exactly
this way if you don't want to.

00:11:49.280 --> 00:11:51.950
But it works for me.

00:11:51.950 --> 00:11:52.450
OK.

00:11:52.450 --> 00:11:56.220
And we got our final answer:
pi h squared r squared.

00:11:56.220 --> 00:11:57.953
I'll stop there.