Добавил:
Upload Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
intro_physics_1.pdf
Скачиваний:
52
Добавлен:
08.02.2016
Размер:
5.79 Mб
Скачать

Week 9: Oscillations

411

that was the key parameter that determined whether or not a damped oscillator was underdamped, critically damped or overdamped (for ζ < 1, ζ = 1, ζ > 1 respectively). This form makes it easier to write the resonance solutions “generically” so that they can be applied to di erent problems.

This is a second order, linear, inhomogeneous ordinary di erential equation – the wordk “inhomogeneous” basically means that there is a function of time instead of 0 on the right.

Although we will not treat it in this course, it is possible to write arbitrary functions of time as a superposition of harmonic functions of time using either fourier series or more generally, fourier analysis187 . This means that the solution we develop here (however crudely) can be transformed into part of a general solution, valid for an arbitrary driving force F (t)! One day, maybe, you188 might learn how, but obviously that would cause our brains to explode to attempt it today, unless you are a math super-genius or at least, have learned a lot more math than undergrads usually have learned by the time they take this course.

9.4.2: Solution to Damped, Driven, Simple Harmonic Oscillator

We will not, actually, fully develop this solution this semester. It is a bit easier to do so in the context of LRC circuits next semester, and you’ll have had the chance to “sleep on” all that you’ve learned about driven oscillation this semester and will discover that, magically, it is somehow easier and less intimidating the second time around. Math like this is best learned in several passes with a bit of time for our brains to “digest” in between.

First, let us really learn an important property of inhomogeneous ordinary di erential equations. Suppose we have both:

d2xH

+ 2ζω0

dxH

+ ω02xH = 0

(854)

dt2

dt

 

 

 

(the homogeneous ODE for a damped SHO) with solution xH (t) given in the previous section, and also have:

d2xI

+ 2ζω

dxI

+ ω2x

 

=

F0

cos(ωt)

(855)

 

 

 

 

 

 

dt2

0 dt

I

 

m

 

0

 

 

 

 

(the inhomogeneous ODE for a damped harmonically driven SHO) with solution xI (t) yet to be determined. Suppose also that we have found at least one solution to the inhomogenous ODE xp(t), usually referred to as a particular solution. Then it is easy to show from the linearity of the ODE that:

 

 

 

 

xI (t) = xH (t) + xp(t)

 

 

 

 

 

 

 

 

 

 

(856)

also solves the inhomogeneous equation:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

d2xI

+ 2ζω

dxI

+ ω2x

 

=

F0

cos(ωt)

 

 

 

 

 

 

 

 

 

 

dt2

 

dt

 

 

m

 

 

 

 

 

 

 

 

 

 

0

 

 

0

 

I

 

 

 

 

 

d2(xH + xp)

+ 2ζω

0

d(xH + xp)

+ ω2

(x

H

+ x

)

=

F0

cos(ωt)

 

 

 

 

 

 

 

 

 

 

 

 

 

dt2

 

 

 

 

dt

 

 

 

0

 

 

 

 

p

 

 

m

 

 

µ dt2

 

 

+

µ dt2

 

 

 

 

 

 

 

+ ω02xp

=

 

 

+ 2ζω0 dt + ω02xH

+ 2ζω0 dt

 

m cos(ωt)

 

d2xH

 

dxH

 

 

 

d2xp

 

 

dxp

 

 

 

 

 

F0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 +

F0

 

cos(ωt)

=

F0

cos(ωt)

(857)

 

 

 

 

 

 

 

 

 

 

 

 

 

m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m

 

 

 

 

 

 

 

which is true as an identity by hypothesis (that xp solves the inhomogeneous ODE).

That is, given one particular solution to the inhomogeneous linear ODE, we can generate an infinite family of solutions by adding to it any solution we like to the homogeneous linear ODE. This (homogeneous) solution has two free parameters (constants of integration) that can be adjusted to

187Wikipedia: http://www.wikipedia.org/wiki/Fourier Analysis.

188For a very small statistical value of “you”, one that pretty much looks like not you unless you are a math or

physics major or perhaps an electrical engineering student.

412

Week 9: Oscillations

satisfy arbitrary initial conditions, the inhomogeneous solution inherits these constants of integration and freedom, and the mathematical gods are thus pleased.

The solution to the homogeneous equation, however, is exponentially damped. If we wait long enough, no matter how we start the system o initially, the homogeneous solution will get as small as we like – basically, it will disappear and we’ll be left only with the particular solution. For that reason, in the context of driven harmonic oscillation the homogeneous solution is called the transient solution and depends on initial conditions, the particular solution is called the steady state solution and does not depend on the initial conditions!

We can then search for a steady state solution that does not contain any free parameters but is completely determined by the givens of the problem independent of the initial conditions!

Although it is not immediately obvious, given a harmonic driving force at a particular frequency ω, the steady state solution must also be a harmonic function at the same frequency ω. This is part of that deep math mojo that comes out of fourier transforming not only the driving force, but the entire ODE. The latter e ectively reduces the entire ODE to an algebraic problem. It is by far easiest to see how this goes by using a complex exponential driving force and solution and then taking the real part at the end, but we will skip doing this for now. Instead I will point out some true facts about the solution one obtains when one does it all right and completely. Interested parties can always look ahead into the companion volume for this course, Intro Physics 2, in the chapter on AC circuits to see how it really should be done.

a) The steady state solution to the driven SHO is:

 

 

xp(t) = A cos(ωt − δ)

(858)

where:

 

 

 

 

F0/m

 

 

 

A =

 

 

 

 

(859)

 

 

 

 

 

 

 

and the phase angle δ is:

 

p(ω2 − ω02)2 + (2ζω0ω)2

 

 

 

 

 

 

 

 

µ ω2 − ω02

 

 

δ = tan1

 

2ζω0ω

 

(860)

 

 

 

 

b)The average power delivered to the mass by the driving force is a quantity of great interest. This is the rate that work is being done on the mass by the driving force, and (because it is in steady state) is equal to the rate that the damping force removes the energy that is being added. Evaluating the instantaneous power is straightforward:

 

~

 

 

 

 

 

 

 

 

(861)

P (t) = F · ~v = F v = −F0cos(ωt) sin(ωt − δ)

We use a trig identity to rewrite this as:

 

P (t) =

−F0cos(ωt)(sin(ωt) sin(δ) cos(ωt) cos(δ))

 

=

F0¡cos2(ωt) sin(δ) cos(ωt) sin(ωt) cos(δ)¢

(862)

The time average of cos2(ωt) is 1/2. The time average of cos(ωt) sin(ωt) is 0. Hence:

 

 

Pavg =

F0

sin(δ)

(863)

 

 

 

 

 

2

 

 

 

 

c) If one sketches out a right triangle corresponding to:

 

 

 

tan(δ) =

2ζω0ω

(864)

 

 

ω2 − ω02

 

 

one can see that:

 

 

 

 

 

2ζω0ω

 

 

sin(δ) =

 

 

 

 

(865)

 

 

 

 

 

 

 

 

 

 

 

p(ω2 − ω02)2 + (2ζω0ω)2

 

 

 

 

Week 9: Oscillations

413

Finally, substituting this equation and the equation for A above into the expression for the average power, we get:

Pavg(ω) =

F02ω2ζω0

(866)

m ((ω2 − ω02)2 + (2ζω0ω)2)

This equation is maximum when ω = ω0, at resonance. At that point the peak average power

can be written:

F 2

 

 

2

 

 

 

F

 

Pavg(ω0) = Pmax =

0

=

 

0

(867)

4mζω0

2b

d)This shape of Pavg(ω) is very important to semi-quantitatively understand. It is (for weak damping) a sharply peaked curve with the peak centered on ω0, the natural frequency of the undamped oscillator. You can see from the expression given for Pavg why this would be so – at ω = ω0 the quantity in the denominator is minimum. This function is plotted in figure 125 below.

RESONANCE OF A DRIVEN HARMONIC OSCILLATOR

 

 

 

 

 

Q = 3

 

 

 

 

 

Q = 10

 

10

 

 

 

Q = 20

 

 

 

 

 

 

8

 

 

 

 

POWER

6

 

 

 

 

 

 

 

 

 

 

4

 

 

 

 

 

2

 

 

 

 

 

0

 

 

 

 

 

0

0.5

1

1.5

2

 

 

 

ω

 

 

Figure 125: This plots the average power P (ω) for three resonances. In these figures, F0 = 1, k = 1, m = 1 (so ω0 = 1) and hence b is the only variable. Since Q = ω0/ = 0/b we plot Q = 3, 10, 20 by selecting b = 0.333, 0.1, 0.05. Thus Pmax = F02/2b = 3/2, 5, 10 respectively (all in suitable units for the quantities involved). Note that the full width of e.g. the Q = 20 curve (with the sharpest/highest peak) is 1/20 at P = Pmax/2 = 5. P is the average power added to the oscillator by the driving force, which in turn in steady state motion must equal the average power lost to to dissipative drag forces.

e)The sharpness of the resonance is controlled by the dimensionless Q-factor, or “quality factor”, of the resonance. Q for the driven SHO is defined to be:

Q =

ω0

 

 

(868)

ω

 

 

where ω is the full width of Pavg(ω) at half-maximum! To find

ω, one must use:

Pavg(ω) =

Pmax

 

(869)

 

2

 

 

and solve for the two roots where the equality is satisfied. The di erence between them is the full width.