09-723 Proximal Probe Techniques Harmonic Oscillator Summary Linear Harmonic oscillator Mechanical model: mass m on a spring characterized by a spring constant k Elastic restoring force Fs = − kx is balanced according to Newton's second law Fs = ma acceleration mx = − kx mx + kx = 0 2 x + ω0 x = 0 where ω0 = k m free natural angular frequency We will show that such system oscillates with amplitude A and angular frequency ω0 . Basic facts about second order linear differential equations: 1. Solutions x = x (t; C1 , C2 ) will have two constants dependent on initial conditions 2. If x1 (t ) is a solution then Cx1 (t ) is also a solution. 3. If x1 (t ) and x2 (t ) are solutions then x1 (t ) + x2 (t ) as well as any linear combination C1 x1 (t ) + C2 x2 (t ) is also a solution. To solve the equation of motion 2 x + ω0 x = 0 we multiply both sides by 2x = −2ω 0 2 xx 2 xx which allows us to immediately carry out the first integration: 2 x 2 = −ω 0 x 2 + C 1 09-723 Proximal Probe Techniques Harmonic Oscillator Summary We determine the first integration constant C by observing that at a full "swing", the oscillator position is equal to its amplitude, x=A, and its velocity is zero (turning point of 2 oscillation) x = 0 . Thus C = ω 0 A2 and the equation to be integrated further becomes: 2 x 2 = ω0 ( A2 − x 2 ) Separation of variables gives ∫ dx A2 − x 2 = ω0 ∫ dt and after integration: x sin −1 = ω0t + φ A or in a more familiar form: x = A sin(ω 0t + φ ) ω0 is the natural angular frequency. ω0 = 2π f 0 , where f0 is the natural frequency. Oscillation period T0 = 1 . f0 Alternative solution method uses trial exponential function x = exp( λ t ) , and our task is to determine the constant λ . Substituting into the equation to be solved we obtain λ 2 exp( λt ) + ω0 2 exp( λt ) = 0 from which it is immediately clear that λ 2 = −ω0 2 and thus λ = ±i ω0 where i = −1 2 09-723 Proximal Probe Techniques Harmonic Oscillator Summary Thus the general solution has to be of the form x = A1 exp( +iω0t ) + A2 exp( −iω 0t ) (remember: we need two constants). Recall Euler's formula e ± iθ = cos θ ± i sin θ Since A1 and A2 are complex quantities, they may be inconvenient to use, and it is more informative to write the solution as: x = A exp[i (ω0t + φ )] = A cos(ω0t + φ ) + iA sin(ω0t + φ ) Energy considerations Consider solution x = A sin(ω 0t + φ ) The velocity v is equal to x = Aω0 cos(ω0t + φ ) and thus the kinetic energy 1 2 1 2 mx = mA2ω 0 cos 2 (ω 0t + φ ) = K 0 cos 2 (ω0t + φ ) 2 2 where the maximum kinetic energy is equal to K= K0 = 1 1 2 mA2ω 0 = kA2 2 2 The potential energy – work done by applied force displacing the system from 0 to x x U ( x ) = ∫ kx dx = 0 1 2 kx 2 Substituting x 3 09-723 Proximal Probe Techniques Harmonic Oscillator Summary 1 2 2 kA sin (ω0t + φ ) = U 0 sin 2 (ω 0t + φ ) 2 U ( x) = where U0 is the maximum potential energy (for x=A) U0 = 1 2 kA 2 The average values over one oscillation period are calculated using the definition t 1 2 f = f (t )dt t2 − t1 ∫t1 Thus: T U = T ∫ Udt ∫ U 0 T = 0 sin 2 (ω0t + φ ) 0 ∫ dt T 1 1 = U 0 = kA2 2 2 0 and K = 1 1 K 0 = kA2 . 2 2 In conclusion, the average values of kinetic and potential energy per oscillation cycle are equal to U = K = 1 E 2 Damped harmonic oscillator This time, we introduce the additional force, which will dissipate the energy. Fd = −bv = −bx The equation of motion gains one more term: 4 09-723 Proximal Probe Techniques Harmonic Oscillator Summary mx + bx + kx = 0 Denote: γ = b k 2 , ω0 = 2m m 2 x + 2γ x + ω 0 x = 0 Use exponential trial function x = exp( λ t ) x = λ exp(λ t ) x = λ 2 exp( λt ) 2 exp( λt )[λ 2 + 2γλ + ω 0 ] = 0 since exp(λ t ) ≠ 0 λ 2 + 2γλ + ω0 2 = 0 λ1 = −γ + γ 2 + ω 02 λ2 = −γ − γ 2 + ω 02 x (t ) = A1 exp( λ1t ) + A2 exp(λ2t ) 2 2 x (t ) = exp( −γ t )[ A1 exp( +t γ 2 − ω0 ) + A2 exp( −t γ 2 − ω0 )] Depending on the sign of the expression under the root, there are three possible cases: Underdamped γ 2 − ω0 2 < 0 λ1 and λ2 are imaginary: oscillating solutions. Overdamped γ 2 − ω0 2 > 0 λ1 and λ2 are both real Critically damped γ 2 = ω02 Underdamped oscillator 5 09-723 Proximal Probe Techniques ω1 = ω0 2 − γ 2 = Harmonic Oscillator Summary k b2 − m 4m 2 x (t ) = exp( −γ t )[ A1 exp(iω1t ) + A2 exp(iω1t )] which recalling Euler's formula becomes x (t ) = exp( −γ t ) [i ( A1 − A2 )sin(ω1t ) + ( A1 + A2 ) cos(ω1t )] Substitute i ( A1 − A2 ) = B and A1 + A2 = C x (t ) = exp( −γ t ) [ B sin(ω1t ) + C cos(ω1t )] Introduce A = B 2 + C 2 and tan(φ ) = − C B x (t ) = A exp( −γ t ) cos(ω1t + φ ) Damped oscillator moves at "frequency" smaller than undamped: γ2 ω1 = ω0 − γ = ω0 1 − 2 ω0 For small damping γ expand in binomial series and retain only the first two terms γ2 γ2 + .....) ≈ (1 − ) ω1 = ω 0 (1 − ω 0 2 2 2ω0 2ω 0 and for small damping γ ω0 and ω1 ≈ ω0 2 2 Critically damped λ1 = λ2 = −γ Solution x (t ) = ( A1 + A2 ) exp( −γ t ) = ( B1 ) exp( −γ t ) This is not a general solution (it contains just one constant). We can show that if x (t ) = exp( −γ t ) than x (t ) = t exp( −γ t ) is also a solution. Substitute 6 09-723 Proximal Probe Techniques Harmonic Oscillator Summary x (t ) = exp( −γ t ) − γ t exp( −γ t ) = (1 − γ t ) exp( −γ t ) x (t ) = (1 − γ t )( −γ ) exp( −γ t ) − γ exp( −γ t ) = (γ 2t − 2γ ) exp( −γ t ) 2 [γ 2t − 2γ + 2γ (1 − γ t ) + ωo t ]exp( −γ t ) = 0 2 [(ωo − γ 2 )t ]exp( −γ t ) = 0 Always satisfied. Thus the general solution for a critically damped oscillator is: x (t ) = ( B1 + B2t ) exp( −γ t ) Overdamped γ 2 − ω0 2 < ω 2 x (t ) = exp( −γ t )[ A1 exp(ω 2t ) + A2 exp( −ω 2t )] Energy considerations Total energy: E (t ) = E (0) + W f work performed by friction Frictional force: f = −bx = −bv t Wf = ∫ 0 t t t dx fdx = ∫ f dt = ∫ fvdt = ∫ −bv 2dt dt 0 0 0 The rate of energy loss: dE dW f = = −bv 2 dt dt E (t ) = K (t ) + U (t ) = 1 2 1 2 mx + kx 2 2 recall that x (t ) = A exp( −γ t ) cos(ω1t + φ ) γ cos(ω1t + φ )] ω1 γ Assume that the system is lightly damped ( 1 ), so we can neglect the second term. ω1 x (t ) = −ω1 A exp( −γ t )[ sin(ω1t + φ ) + Then: E (t ) = 1 2 A exp( −2γ t )[m ω1 sin 2 (ω1t + φ ) + k cos 2 (ω1t + φ )] 2 7 09-723 Proximal Probe Techniques Harmonic Oscillator Summary 2 2 Since light damping was assumed, ω1 ≈ ω 0 = E (t ) = k and E(t) becomes m 1 2 kA exp( −2γ t ) 2 The initial energy 1 E0 = kA2 2 and thus E (t ) = E0 exp( −2γ t ) (notice that the energy decays twice as fast as amplitude!) The characteristic decay time (E decreases to E/e) E0 = E0 exp( −2γτ ) e 2γτ = 1 1 2m m τ= = = 2γ 2b b Quality factor Q = 2π energy stored in the oscillator energy dissipated in one time period Define P = power loss = rate of energy dissipation 2π since one time period T1 = ω1 the denominator can be written as 2π PT1 = P ω1 Thus Q = 2π E 2π P ω1 1 - time necessary to complete 1 radian of oscillation, we can redefine Q as ω1 energy stored in the oscillator Q= average energy dissipated per radian Since 8 09-723 Proximal Probe Techniques Harmonic Oscillator Summary For a lightly damped oscillator Q can be calculated as follows: E (t ) = E0 exp( −2γ t ) dE = −2γ E dt Thus the energy dissipated in time ∆t will be equal to dE ∆t = 2γ E ∆t dT 1 If we choose ∆t = (time necessary to complete 1 radian) ω1 ∆E = Q= ω E E = = 1 ∆E 2γ E / ω1 2γ For light damping ω1 ≈ ω0 and thus ω Q= 0 2γ Forced (driven) damped harmonic oscillator Net force Fnet = Fs + F f + Fd restoring elastic force dissipative friction force driving force where: Fs = − kx ; F f = −bx Fnet = mx mx + bx + kx = Fd assume harmonic driving force Fd = F0 cos(ω t + θ 0 ) mx + bx + kx = F0 cos(ω t + θ 0 ) This is an inhomogeneous 2-nd order linear differential equation. Its solution is the sum of two parts, according to the following theorem: If Xi is a particular solution of an inhomogeneous differential equation, and Xn is a solution of a complementary homogeneous equation, then X(t)=Xi(t)+Xh(t) is a general solution. 9 09-723 Proximal Probe Techniques Harmonic Oscillator Summary From our previous considerations, the solution of the complementary homogeneous equation is given by one of the three equivalent forms: xh (t ) = exp( −γ t ) [ A1 exp( +iω1t ) + A2 exp( −iω1t )] xh (t ) = exp( −γ t ) [ B sin(ω1t ) + C cos(ω1t )] xh (t ) = Ah exp( −γ t ) cos(ω1t + φh ) For an inhomogeneous equation, let's postulate the following particular solution xi (t ) = A cos(ω t ± φ ) and focus on the – sign solution. xi (t ) = A cos(ω t − φ ) xi (t ) = − Aω sin(ω t − φ ) xi (t ) = − Aω 2 cos(ω t − φ ) Upon substitution − mAω 2 cos(ω t − φ ) − bAω sin(ω t − φ ) + kA cos(ω t − φ ) = F0 cos(ω t ) (for simplicity we assumed that θ 0 = 0 ). Recall that: cos(ω t − φ ) = cos ω t cos φ + sin ω t sin φ sin(ω t − φ ) = sin ω t cos φ − cos ω t sin φ thus − mAω 2 [cos ω t cos φ + sin ω t sin φ ] −bAω [sin ω t cos φ − cos ω t sin φ ] + kA [cos ω t cos φ + sin ω t sin φ ] = = F0 cos(ω t ) This can be regrouped as cos ω t[− mAω 2 cos φ + bAω sin φ + kA cos φ ] − sin ω t[− mAω 2 sin φ − bAω cos φ + kA sin φ ] = F0 cos(ω t ) + 0 sin(ω t ) Since the cos(ω t ) and sin(ω t ) coefficients on both sides of the equation have to be equal, we obtain the system of two equations: 10 09-723 Proximal Probe Techniques Harmonic Oscillator Summary F0 A 2 ( k − mω )sin φ − bω cos φ = 0 ( k − mω 2 ) cos φ + bω sin φ = From the above it follows that: tan φ = b ω m sin φ bω 2γω = = = 2 2 2 k cos φ k − mω − ω 2 ω0 − ω m thus sin φ = cos φ = 2γω 2 (ω 0 − ω 2 ) 2 + 4γ 2ω 2 ω02 − ω 2 2 (ω 0 − ω 2 ) 2 + 4γ 2ω 2 Substistute these back to obtain F0 / m A= 2 ω 02 − ω 2 + 4γ 2ω 2 ( ) Thus a particular solution of the inhomogeneous equation is: xi (t ) = F0 / m (ω 2 0 −ω 2 ) 2 + 4γ ω 2 2 cos(ω t − φ ) where φ = tan −1 2γω ω0 2 − ω 2 The general solution is: x (t ) = xh (t ) + xi (t ) = Ah exp( −γ t ) cos(ω1t + φh ) + F0 / m (ω 2 0 −ω 2 ) 2 + 4γ ω transient term steady state term 11 2 2 cos(ω t − φ ) 09-723 Proximal Probe Techniques Harmonic Oscillator Summary Amplitude resonance The amplitude of the particular solution reaches maximum when the driving force is equal to ω = ω r = ω 02 − 2γ 2 On resonance, the phase shift π φ= 2 Far below resonance ω ωr , φ → 0 Far above resonance ω ωr , φ → π Energy resonance x (t ) = A cos(ω t − φ ) v = x (t ) = −ω A sin(ω t − φ ) 1 1 K (t ) = mv 2 = mω 2 A2 sin 2 (ω t − φ ) 2 2 1 1 U (t ) = kx 2 = kA2 cos2 (ω t − φ ) 2 2 1 2 A [mω 2 sin 2 (ω t − φ ) + k cos2 (ω t − φ )] 2 Recall that per period: 1 cos 2 (ω t − φ ) = sin 2 (ω t − φ ) = 2 Substitute A (ω ) to K(t) E (t ) = K (t ) + U ( t ) = 2 K (t ) = F0 / m 2 1 mω 2 sin 2 (ω t − φ ) 2 2 2 2 2 2 ω0 − ω + 4γ ω ( ) 2 K (t ) = 1 F0 ω2 4 m ω 2 − ω 2 2 + 4γ 2ω 2 0 ( ) Kinetic energy resonance: d K (t ) = 0 for ω = ω 0 dω 12 09-723 Proximal Probe Techniques Harmonic Oscillator Summary Potential energy resonance since U (t ) = 1 2 2 kx , occurs at the same frequency as amplitude resonance ω = ω0 − 2γ 2 2 Total energy resonance. Resonance peak width and Q E (t ) = 1 2 2 1 A mω sin 2 (ω t − φ ) + A2k cos2 (ω t − φ ) 2 2 and after substituting A 2 E = 2 1 F0 ω 2 + ω0 4 m ω 2 − ω 2 2 + 4γ 2ω 2 0 ( ) For very weak damping γ ω0 ω 2 + ω 02 ≈ 2ω0 2 ω 2 − ω 02 = (ω − ω 0 )(ω + ω 0 ) ≈ 2ω0 (ω − ω0 ) 2 E (ω ) = 1 F0 1 8 m (ω 0 − ω )2 + γ 2 Lorentzian Maximum at ω = ω 0 2 1 F0 1 8 m γ2 The energy is equal ½ of its value at (ω − ω 0 )2 = γ 2 or ω − ω 0 = ±γ Resonance peak width at half height E (ω 0 ) = ∆ω = 2γ Recall from previous considerations that ω Q= 0 2γ This is the basis for determining Q from the energy resonance peak width: Q= ω0 . ∆ω 13