WEBVTT

00:00:00.499 --> 00:00:03.350
GILBERT STRANG: OK,
it's time to move on

00:00:03.350 --> 00:00:05.560
to second order equations.

00:00:05.560 --> 00:00:10.220
First order equations,
we've done pretty carefully.

00:00:10.220 --> 00:00:13.990
Second order equations
are a step harder.

00:00:13.990 --> 00:00:18.080
But they come up in nature,
they come in every application,

00:00:18.080 --> 00:00:21.196
because they include
an acceleration,

00:00:21.196 --> 00:00:23.360
a second derivative.

00:00:23.360 --> 00:00:29.670
OK, so this would be a second
order equation, because

00:00:29.670 --> 00:00:31.670
of that second derivative.

00:00:31.670 --> 00:00:35.430
I'm often going to have
constant a, b, and c.

00:00:35.430 --> 00:00:37.930
We have enough
difficulties to it

00:00:37.930 --> 00:00:41.390
without allowing
those to change.

00:00:41.390 --> 00:00:47.200
So a, b, c constants, and
doing the null solution

00:00:47.200 --> 00:00:49.310
to start with.

00:00:49.310 --> 00:00:53.270
Later, there'll be a forcing
term on the right-hand side.

00:00:53.270 --> 00:00:57.370
But today, for this
video, null solutions.

00:00:57.370 --> 00:01:00.810
And the point is, now,
what's new is that there

00:01:00.810 --> 00:01:02.890
are two null solutions.

00:01:02.890 --> 00:01:08.190
So y null will be a
combination of e--

00:01:08.190 --> 00:01:10.820
and they're both exponentials.

00:01:10.820 --> 00:01:13.090
Having constant
coefficients there

00:01:13.090 --> 00:01:16.810
means exponentials
in the solution.

00:01:16.810 --> 00:01:22.690
So e to the sum
exponent, and another one

00:01:22.690 --> 00:01:27.730
to a, hopefully,
different exponent.

00:01:27.730 --> 00:01:31.300
Sometimes if s1 equals s2,
that'll be a special case

00:01:31.300 --> 00:01:33.480
and we'll have a slight change.

00:01:33.480 --> 00:01:35.530
But this is typical.

00:01:35.530 --> 00:01:39.950
So we have two constants, c1
and c2, in the null solutions.

00:01:39.950 --> 00:01:43.480
And we need two
initial conditions

00:01:43.480 --> 00:01:45.710
to determine those constants.

00:01:45.710 --> 00:01:51.600
So previously, for a first order
equation, we were given y of 0.

00:01:51.600 --> 00:01:59.590
Now, when we have acceleration,
we give the initial velocity,

00:01:59.590 --> 00:02:00.865
y prime of 0.

00:02:05.900 --> 00:02:11.910
May I use prime as a
shorthand for derivative?

00:02:11.910 --> 00:02:16.680
y prime of 0 is dydt, at 0.

00:02:16.680 --> 00:02:20.610
So two condition will,
at the right time,

00:02:20.610 --> 00:02:22.850
determine those two constants.

00:02:22.850 --> 00:02:26.210
And just to say again,
so second derivative

00:02:26.210 --> 00:02:29.270
will be y double prime.

00:02:29.270 --> 00:02:32.500
And that represents,
in physics, it

00:02:32.500 --> 00:02:38.220
represents acceleration-- change
in velocity, change in y prime,

00:02:38.220 --> 00:02:40.110
is y double prime.

00:02:40.110 --> 00:02:45.270
And in a graph of a function,
y double prime shows up

00:02:45.270 --> 00:02:49.000
in bending of the graph.

00:02:49.000 --> 00:02:52.590
Because bending is
a change in slope.

00:02:52.590 --> 00:02:59.820
Bending is a change
in the slope.

00:02:59.820 --> 00:03:03.660
And the slope is y prime,
the first derivative.

00:03:03.660 --> 00:03:06.670
So to measure changes
in y prime, which

00:03:06.670 --> 00:03:12.300
will bend the graph, my chalk
would be a tangent line.

00:03:12.300 --> 00:03:15.400
But if that changes,
that gives us

00:03:15.400 --> 00:03:18.630
a y double prime, a
second derivative.

00:03:18.630 --> 00:03:24.530
OK, so I'm ready
for some examples.

00:03:24.530 --> 00:03:31.660
And the first example-- the
most basic equation of motion

00:03:31.660 --> 00:03:36.700
in physics and engineering,
I would say-- it's

00:03:36.700 --> 00:03:39.090
called harmonic motion.

00:03:39.090 --> 00:03:41.835
And b is 0, that's
the key point.

00:03:41.835 --> 00:03:44.580
b is 0.

00:03:44.580 --> 00:03:46.570
It's Newton's law.

00:03:46.570 --> 00:03:51.530
And so a will be the mass m.

00:03:51.530 --> 00:03:55.062
y double prime,
second derivative.

00:03:55.062 --> 00:03:57.900
b is 0.

00:03:57.900 --> 00:04:03.120
Later, b will be a damping term,
a friction term, a resistance

00:04:03.120 --> 00:04:03.620
term.

00:04:03.620 --> 00:04:05.280
But let's have that 0.

00:04:05.280 --> 00:04:08.500
So we're going to have
perpetual motion, here.

00:04:08.500 --> 00:04:13.080
Plus the force-- so
this is Newton's law.

00:04:13.080 --> 00:04:17.089
ma is f, f equal ma.

00:04:17.089 --> 00:04:20.610
The force is proportional--
with a minus sign,

00:04:20.610 --> 00:04:25.400
so it's going to come on this
side as a plus-- proportional

00:04:25.400 --> 00:04:27.280
to y.

00:04:27.280 --> 00:04:30.780
There's the equation.

00:04:30.780 --> 00:04:33.780
No y prime term.

00:04:33.780 --> 00:04:37.860
my double prime,
plus ky equals 0.

00:04:37.860 --> 00:04:41.190
Starting from an
initial position,

00:04:41.190 --> 00:04:44.340
and an initial velocity,
it's like a spring

00:04:44.340 --> 00:04:50.740
going up and down, or a clock
pendulum going back and forth.

00:04:50.740 --> 00:04:55.830
And we'll see it, so
it's going to be-- well,

00:04:55.830 --> 00:04:58.840
we want to solve that equation.

00:04:58.840 --> 00:05:03.260
So do we see solutions to that?

00:05:03.260 --> 00:05:08.140
So if m and k were 1,
suppose m and k were 1.

00:05:08.140 --> 00:05:10.510
I'm looking for a
second derivative plus

00:05:10.510 --> 00:05:11.930
the function is 0.

00:05:11.930 --> 00:05:14.900
Second derivative is
minus the function.

00:05:14.900 --> 00:05:21.440
I immediately think of
sine t, and cosine t.

00:05:21.440 --> 00:05:24.940
Sines and cosines, because
the second derivative--

00:05:24.940 --> 00:05:27.720
the first derivative
of a sine is cosine.

00:05:27.720 --> 00:05:30.140
The second derivative
is minus the sine.

00:05:30.140 --> 00:05:35.820
It gives us the minus sign
here, the plus sign there, 0.

00:05:35.820 --> 00:05:45.720
So the special solutions here
are y is-- this is the null.

00:05:45.720 --> 00:05:47.890
I'm finding the null solution.

00:05:47.890 --> 00:05:53.230
Is, let me call them
c1 times a cosine.

00:05:53.230 --> 00:05:57.510
And now I have to figure
out the cosine of what?

00:05:57.510 --> 00:06:02.630
I want the cosine to satisfy,
to be a null solution,

00:06:02.630 --> 00:06:04.460
satisfy my equation.

00:06:04.460 --> 00:06:05.940
Let me put it in.

00:06:05.940 --> 00:06:09.740
If there's a square
root of k over mt.

00:06:13.370 --> 00:06:16.760
And you have to
see that if I take

00:06:16.760 --> 00:06:20.120
two derivatives of the
cosine, that will produce

00:06:20.120 --> 00:06:23.170
minus the cosine, which I want.

00:06:23.170 --> 00:06:28.710
And because of that, the chain
will bring out this square root

00:06:28.710 --> 00:06:30.060
twice.

00:06:30.060 --> 00:06:36.550
So it will bring out the factor
k over m for y double prime.

00:06:36.550 --> 00:06:39.225
And that factor, k over
m, the m's will cancel

00:06:39.225 --> 00:06:41.990
and I'll have the k
that matches that k.

00:06:41.990 --> 00:06:44.350
It's a solution.

00:06:44.350 --> 00:06:47.346
And the other solution
is just like it.

00:06:47.346 --> 00:06:49.722
It's the sine.

00:06:49.722 --> 00:06:55.330
Sine of this square
root of k over m, t.

00:06:59.240 --> 00:07:02.550
That's worth putting
a box around.

00:07:02.550 --> 00:07:07.890
That's what I mean by
free harmonic motion.

00:07:07.890 --> 00:07:10.530
Something is just oscillating.

00:07:10.530 --> 00:07:12.620
In rotation problems,
something is just

00:07:12.620 --> 00:07:16.770
going around a circle
at a constant speed.

00:07:16.770 --> 00:07:23.325
And notice, these are
not the same as those.

00:07:25.940 --> 00:07:29.120
Cosines are related
to exponentials,

00:07:29.120 --> 00:07:30.810
but not identical.

00:07:30.810 --> 00:07:37.170
So I could write the answer
this way, using cosine and sine.

00:07:37.170 --> 00:07:41.420
Or as you'll see, I can
write the same formula

00:07:41.420 --> 00:07:45.530
using exponentials,
complex exponentials.

00:07:45.530 --> 00:07:48.480
Everybody remembers
the big formula

00:07:48.480 --> 00:07:51.000
that allows complex
numbers in here

00:07:51.000 --> 00:07:57.720
is Euler's formula, that
the exponential of i omega t

00:07:57.720 --> 00:08:03.200
is the cosine plus i
times the sine of omega t.

00:08:03.200 --> 00:08:05.310
I'll write it again.

00:08:05.310 --> 00:08:07.200
That's the solution.

00:08:07.200 --> 00:08:11.190
It's got two null solutions.

00:08:11.190 --> 00:08:12.440
They're independent.

00:08:12.440 --> 00:08:13.910
They're different.

00:08:13.910 --> 00:08:16.050
And we've got two constants.

00:08:16.050 --> 00:08:18.060
Because our equation
is linear, we

00:08:18.060 --> 00:08:22.360
can safely multiply by any
constant, and add solutions,

00:08:22.360 --> 00:08:26.920
and they stay solutions because
we have a linear equation,

00:08:26.920 --> 00:08:29.391
and 0 on the right-hand side.

00:08:29.391 --> 00:08:29.890
Good.

00:08:32.419 --> 00:08:37.360
Of course, we can't write
this square root of k over m

00:08:37.360 --> 00:08:38.419
forever.

00:08:38.419 --> 00:08:43.010
Let me do what everybody
does-- introduce omega.

00:08:43.010 --> 00:08:45.000
It's omega natural.

00:08:45.000 --> 00:08:48.540
The n here stands for
the natural frequency,

00:08:48.540 --> 00:08:52.330
the frequency that
that clock is going at.

00:08:52.330 --> 00:08:58.600
And that is the square
root of k over m.

00:08:58.600 --> 00:09:02.290
So our equation, we could
rewrite that equation.

00:09:02.290 --> 00:09:05.235
Let me rewrite that
equation to make it simple.

00:09:05.235 --> 00:09:07.230
I'll divide by m.

00:09:07.230 --> 00:09:09.620
No problem, to divide by m.

00:09:09.620 --> 00:09:13.320
So then I have y
double prime, plus k

00:09:13.320 --> 00:09:17.630
over m, which is
omega n squared,

00:09:17.630 --> 00:09:20.160
the natural frequency squared.

00:09:20.160 --> 00:09:23.140
y equals 0.

00:09:23.140 --> 00:09:27.850
Let's put a box around that one,
because you couldn't be better

00:09:27.850 --> 00:09:28.930
than that.

00:09:28.930 --> 00:09:30.650
The constant a is 1.

00:09:30.650 --> 00:09:32.470
The constant b is 0.

00:09:32.470 --> 00:09:37.850
The constant c is a known
omega n squared, depending

00:09:37.850 --> 00:09:42.210
on the pendulum itself.

00:09:42.210 --> 00:09:46.070
OK, and the solutions then,
I'll just copy this solution.

00:09:46.070 --> 00:09:53.370
y null is c1 cosine of omega nt.

00:09:53.370 --> 00:09:55.700
Of course, omega n
is that square rood.

00:09:55.700 --> 00:10:00.740
And c2 sine of omega nt.

00:10:00.740 --> 00:10:02.560
Oh, well, wait a minute.

00:10:02.560 --> 00:10:08.650
I can figure out
what c1 and c2 are ,

00:10:08.650 --> 00:10:10.840
coming from the initial
conditions, right?

00:10:10.840 --> 00:10:15.580
The initial conditions,
if I plug in t equals 0,

00:10:15.580 --> 00:10:18.920
then I want to get
the answer y of 0.

00:10:18.920 --> 00:10:22.260
The known initial
condition, the place

00:10:22.260 --> 00:10:27.960
the pendulum started swinging
from, the place the spring,

00:10:27.960 --> 00:10:30.530
you pull the spring
a distance y of 0.

00:10:30.530 --> 00:10:33.210
You let go.

00:10:33.210 --> 00:10:38.450
At t equals 0, so I'm
plugging in t equals 0, at t

00:10:38.450 --> 00:10:39.870
equals 0, that's 0.

00:10:39.870 --> 00:10:42.000
Forget the sine.

00:10:42.000 --> 00:10:43.810
This is 1.

00:10:43.810 --> 00:10:47.540
So I discover that
c1 should be y of 0.

00:10:47.540 --> 00:10:48.840
Simple.

00:10:48.840 --> 00:10:50.400
c1 is y of 0.

00:10:56.870 --> 00:11:01.030
Because that gives me the
right answer, at t equals 0.

00:11:01.030 --> 00:11:02.730
And what about c2?

00:11:02.730 --> 00:11:04.090
Can I figure out c2?

00:11:04.090 --> 00:11:08.520
Well, that's going to
involve the initial velocity,

00:11:08.520 --> 00:11:12.220
the derivative, at t equals 0.

00:11:12.220 --> 00:11:14.170
Because the
derivative of the sine

00:11:14.170 --> 00:11:19.190
is the cosine, which
equals 1 at t equals 0,

00:11:19.190 --> 00:11:21.970
the derivative of this
is the sine, which is 0.

00:11:21.970 --> 00:11:26.630
So that when I'm looking
at y prime, the derivative,

00:11:26.630 --> 00:11:29.470
I'm looking here at t equals 0.

00:11:29.470 --> 00:11:33.000
And I want y prime of 0.

00:11:33.000 --> 00:11:35.370
But I don't just
want y prime of 0.

00:11:35.370 --> 00:11:39.050
Do you see that that doesn't
have the right derivative,

00:11:39.050 --> 00:11:40.400
at t equals 0?

00:11:40.400 --> 00:11:42.530
Because when I take the
derivative and omega

00:11:42.530 --> 00:11:49.040
n-- that constant, you remember
that constant-- the derivative

00:11:49.040 --> 00:11:51.850
of this will bring
out an omega n.

00:11:51.850 --> 00:11:56.240
So I better have an omega
n down here to cancel it.

00:11:56.240 --> 00:11:58.680
And now I've got it.

00:11:58.680 --> 00:12:05.250
That tells me the
motion, forever and ever.

00:12:05.250 --> 00:12:07.070
Energy is constant.

00:12:07.070 --> 00:12:09.370
Potential energy
plus kinetic energy,

00:12:09.370 --> 00:12:10.940
I could speak about energy.

00:12:10.940 --> 00:12:13.770
But I won't.

00:12:13.770 --> 00:12:19.980
That motion continues
forever, free harmonic motion.

00:12:19.980 --> 00:12:22.740
OK, it goes on and on.

00:12:22.740 --> 00:12:23.750
OK.

00:12:23.750 --> 00:12:29.890
And again, I could write this in
terms of complex exponentials.

00:12:29.890 --> 00:12:33.180
But I'm pretty happy
with that form.

00:12:33.180 --> 00:12:35.120
It's hard to beat that form.

00:12:35.120 --> 00:12:40.030
OK, so what else to do here.

00:12:40.030 --> 00:12:44.950
First of all, we're going
to have cosine omega

00:12:44.950 --> 00:12:48.930
t's for quite awhile.

00:12:48.930 --> 00:12:54.160
I better draw the graph of
that simple, familiar function.

00:12:54.160 --> 00:12:59.500
And so here's a graph
of cosine omega t.

00:12:59.500 --> 00:13:01.930
So here's t.

00:13:01.930 --> 00:13:03.350
Here's cosine of omega.

00:13:03.350 --> 00:13:04.540
Here's 0.

00:13:04.540 --> 00:13:07.690
Here's 2-- well, let
me see what I get.

00:13:07.690 --> 00:13:12.130
So I go cosine of omega, cosine
of omega t is what I'm drawing,

00:13:12.130 --> 00:13:13.920
not cosine t.

00:13:13.920 --> 00:13:16.920
Cosine t would go 0 to 2 pi.

00:13:16.920 --> 00:13:20.110
But I have cosine of omega t.

00:13:20.110 --> 00:13:24.120
So cosine of omega t is
what I want to graph.

00:13:24.120 --> 00:13:27.725
So it starts at 1,
and it comes back.

00:13:27.725 --> 00:13:33.510
It drops, comes back
up, comes back to 1.

00:13:33.510 --> 00:13:37.290
But what is this t final?

00:13:37.290 --> 00:13:45.400
The period, this t is the
period of the oscillation.

00:13:45.400 --> 00:13:51.590
It's the time it takes for
the swing to go up and back.

00:13:51.590 --> 00:13:53.340
And what is that?

00:13:53.340 --> 00:13:59.910
That would be-- so I'm
graphing cosine of omega t,

00:13:59.910 --> 00:14:00.850
is what I'm graphing.

00:14:03.900 --> 00:14:08.070
So omega t starts out
at 0, where t is 0.

00:14:08.070 --> 00:14:14.230
And it goes up to-- so I want
omega t, when I get here,

00:14:14.230 --> 00:14:16.400
this omega t should be 2 pi.

00:14:19.150 --> 00:14:21.420
Right?

00:14:21.420 --> 00:14:31.310
Then I've completed the cosine,
one rotation around the circle,

00:14:31.310 --> 00:14:34.080
one movement of the
pendulum back and forth,

00:14:34.080 --> 00:14:36.140
is in that picture now.

00:14:36.140 --> 00:14:40.270
OK, so omega t is 2 pi, right?

00:14:40.270 --> 00:14:41.920
Right.

00:14:41.920 --> 00:14:43.130
The period is t.

00:14:43.130 --> 00:14:47.120
I think of omega as
the circular frequency.

00:14:47.120 --> 00:14:56.915
Omega is in radians per second.

00:15:01.180 --> 00:15:04.500
That's the units.

00:15:04.500 --> 00:15:08.810
And units truly are
important to keep track.

00:15:08.810 --> 00:15:12.810
Omega, this is omega,
radians per second,

00:15:12.810 --> 00:15:16.980
when I multiply by the
period, the t in seconds,

00:15:16.980 --> 00:15:19.290
I get 2 pi radians.

00:15:19.290 --> 00:15:24.200
OK now, in engineers
and in everyday use,

00:15:24.200 --> 00:15:29.970
there's another frequency called
f, for frequency probably.

00:15:29.970 --> 00:15:34.570
OK, so you should know about f.

00:15:34.570 --> 00:15:39.080
And its frequency is in hertz.

00:15:39.080 --> 00:15:49.550
So f is measured in hertz,
H-E-R-T-Z, named after the guy.

00:15:49.550 --> 00:15:55.770
Not the tomato ketchup guy,
but-- oh, that's Heinz anyway.

00:15:55.770 --> 00:15:57.480
Sorry about that.

00:15:57.480 --> 00:16:02.950
Hertz -- not the car, that's
what I was trying to say,

00:16:02.950 --> 00:16:10.100
but the German guy who
was involved, early,

00:16:10.100 --> 00:16:11.640
with this stuff.

00:16:11.640 --> 00:16:12.810
So what is f?

00:16:17.870 --> 00:16:20.630
f times t is 1.

00:16:20.630 --> 00:16:24.135
Instead of dealing with 2
pi, which count to radians,

00:16:24.135 --> 00:16:31.770
the 1 just counts complete
loops, complete oscillations.

00:16:31.770 --> 00:16:39.350
So f compared to know omega,
f is smaller by a factor 2 pi.

00:16:39.350 --> 00:16:43.750
So f times t is
1. f is 1 over t.

00:16:43.750 --> 00:16:46.290
Omega is 2 pi over t.

00:16:46.290 --> 00:16:48.770
So putting these
together, all I'm saying

00:16:48.770 --> 00:16:54.190
is, omega is 2 pi times f.

00:16:54.190 --> 00:16:58.960
So when we say we're
getting 60 cycles-- so

00:16:58.960 --> 00:17:02.520
that's what I would
measure tf in.

00:17:02.520 --> 00:17:10.495
f would be in cycles,
is cycles per second.

00:17:13.430 --> 00:17:16.209
One cycle, 2 pi radians.

00:17:19.290 --> 00:17:21.440
This isn't big, heavy math.

00:17:21.440 --> 00:17:24.470
But it's more important
than a lot of math,

00:17:24.470 --> 00:17:28.170
just to get these
letters straight.

00:17:28.170 --> 00:17:31.030
So there's capital
T, the period,

00:17:31.030 --> 00:17:32.990
and two measures of frequency.

00:17:32.990 --> 00:17:36.170
One is omega in
radians per second,

00:17:36.170 --> 00:17:40.830
and the other is f in
full cycles per second.

00:17:40.830 --> 00:17:42.740
So one is 2 pi times the other.

00:17:42.740 --> 00:17:43.300
Good.

00:17:43.300 --> 00:17:44.070
OK.

00:17:44.070 --> 00:17:45.730
And we have this.

00:17:45.730 --> 00:17:46.450
OK.

00:17:46.450 --> 00:17:54.440
I think we've got the
key ideas here, then,

00:17:54.440 --> 00:18:01.360
for unforced motion, pure
oscillation going on forever.

00:18:01.360 --> 00:18:07.960
And let me just write
what I already mentioned.

00:18:07.960 --> 00:18:15.700
A different way to express
yn would be with small c's, e

00:18:15.700 --> 00:18:20.210
to the i omega nt.

00:18:20.210 --> 00:18:26.400
And c2 e to the
minus i omega nt.

00:18:26.400 --> 00:18:33.280
All I'm saying is that this
form, with exponentials,

00:18:33.280 --> 00:18:40.100
is entirely equivalent to this
form, with cosine and sine.

00:18:40.100 --> 00:18:44.190
That's form allows me two
constants, capital C1 and C2.

00:18:44.190 --> 00:18:48.080
This form allows me two
constants, little one and c2.

00:18:48.080 --> 00:18:49.850
And from that, I have this.

00:18:49.850 --> 00:18:50.950
From this, I have this.

00:18:50.950 --> 00:18:55.150
So we really do have
exponentials here.

00:18:55.150 --> 00:18:59.680
And the key message is
that for pure oscillation,

00:18:59.680 --> 00:19:07.070
those exponentials are pure
imaginary exponent, i omega nt.

00:19:07.070 --> 00:19:11.790
OK, that's the best example,
the simplest example,

00:19:11.790 --> 00:19:13.500
the first example.

00:19:13.500 --> 00:19:15.050
Thanks.