Course: Nonlinear Dynamics. Laurette TUCKERMAN Maps, Period Doubling and Floquet Theory

Transcription

1 Course: Nonlinear Dynamics Laurette TUCKERMAN Maps, Period Doubling and Floquet Theory

2 Discrete Dynamical Systems or Mappings y n+1 = g(y n ) y,g R N Fixed point: ȳ = g(ȳ) 1D linear stability ofȳ: Sety n = ȳ + ǫ n ȳ + ǫ n+1 = g(ȳ + ǫ n ) = g(ȳ) + g (ȳ)ǫ n g (ȳ)ǫ 2 n ǫ n+1 g (ȳ)ǫ n g (ȳ) <> 1 ǫ ȳ stable / unstable Superstability: g (ȳ) = 0 = ǫ n g (ȳ)ǫ 2 n

3 Multidimensional system g (ȳ) = Dg(ȳ) (Jacobian) ȳ stable all eigenvaluesµof Dg(ȳ) inside unit circle µ exits at ( 1,0) µ exits ate ±iθ µ exits at (1,0)

4 Illustrate exit at (±1,0) via graphical construction (+1, 0) ( 1, 0)

5 Change of stability Bifurcations Saddle-node bifurcation: ẋ = µ x 2 x n+1 x n = µ x 2 n x n+1 = f(x n ) x n + µ x 2 n f (x n ) = 1 2x n µ < 0 µ > 0 no fixed point x = ± µ f (± µ) = 1 2 µ 1

6 Supercritical pitchfork bifurcation: ẋ = µx x 3 x n+1 x n = µx n x 3 n x n+1 = f(x n ) x n + µx n x 3 n f (x n ) = 1 + µ 3x 2 n µ < 0 µ > 0 x = 0 x = 0,± µ f (0) = 1 + µ < 1 f (0) = 1 + µ > 1 f (± µ) = 1 2µ < 1

7 Subcritical pitchfork bifurcation: ẋ = µx + x 3 x n+1 x n = µx n + x 3 n x n+1 = f(x n ) x n + µx n + x 3 n f (x n ) = 1 + µ + 3x 2 n µ < 0 µ > 0 x = 0,± µ x = 0 f (0) = 1 + µ < 1 f (0) = 1 + µ > 1 f (± µ) = 1 2µ < 1

8 Eig at Bifurcations Leads to (+1,0) saddle-node, pitchfork, transcritical other steady states e ±iθ secondary Hopf or Neimark-Sacker torus (next chapter) ( 1,0) flip or period-doubling two-cycle Period-doubling: impossible for continuous dynamical systems Illustrate via logistic map: x n+1 = ax n (1 x n ) { multiple of x n for x n small, but Next value x n+1 is reduced whenx n too large Mentioned in 1800s, popularized in 1970s: models population Famous period-doubling cascade discovered in 1970s by Feigenbaum in Los Alamos, U.S., and by Coullet and Tresser in Nice, France

9 Logistic Map x n+1 = f(x n ) ax n (1 x n ) for x n [0,1], 0 < a < 4 Fixed points: x = a x(1 x) { x = 0 or = 1 = a(1 x) = 1 x = 1/a = x = 1 1/a f(x) = ax(1 x) for x = 0 and x = 1 1/a a = 0.4, 1.2, 2.0, 2.8, 3.6 as function of a

10 Stability f(x) = ax(1 x) f (x) = a(1 x) ax = a(1 2x) f (0) = a = f (0) < 1 for a < 1 ( f 1 1 ) ( ( = a )) = a + 2 a ( a 1 < f 1 1 ) < 1 a 1 < a + 2 < 1 1 < a 2 < 1 transcrit: 1 < a < 3???

11 Graphical construction a = 0.8: x n 0 a = 2.0: x n 1 1/a a = 2.6: x n 1 1/a a = 3.04: x n 2 cycle

12 Seek two-cycle formed ata = 3, wheref ( x) = 1 f(x 1 ) = x 2 and f(x 2 ) = x 1 = f 2 (x 1 ) f(f(x 1 )) = x 1 f 2 (x) = af(x)[1 f(x)] = a(ax(1 x)) [1 ax(1 x)] = a 2 x(1 x) [ 1 ax + ax 2] = a 2 x [ 1 ax + ax 2 x + ax 2 ax 3] = a 2 x [ 1 (1 + a)x + 2ax 2 ax 3] Seek fixed points of f 2 : 0 = f 2 (x) x = x [ a 2 (1 (1 + a)x + 2ax 2 ax 3 ) 1 ] contains factors x and (x (1 1/a)) = x(a(x 1) + 1) [ ax 2 (a + 1)x + 4(a + 1) ] x 1,2 = a + 1 ± (a 3)(a + 1) 2a for a > 3

13 f 2 undergoes pitchfork bifurcation a Fixed Points 1.6 x = 0 unstable x = 1 1/a stable 2.3 x = 0 unstable x 1,2 stable x = 1 1/a unstable together comprise two-cycle for f

14 Stability of two-cycle d dx f2 (x 1 ) = f (f(x 1 ))f (x 1 ) = f (x 2 )f (x 1 ) x 1 + x 2 = a = a + 1 x 1 x 2 = a + 1 a a 2 f (x 1 )f (x 2 ) = a(1 2x 1 ) a(1 2x 2 ) = a 2 (1 2(x 1 + x 2 + 4x 1 x 2 ) ( ) ( )) a + 1 a + 1 = a ( a a 2 = a 2 2a(a + 1) + 4(a + 1) = a 2 + 2a = f (x 1 )f (x 2 ) 1 = a 2 + 2a = a 2 + 2a + 3 a = 2 ± = 1 ± 16 2 = 2 ± 4 2 = 3 pitchfork 2-cycle 0 = f (x 1 )f (x 2 ) + 1 = a 2 + 2a = a 2 + 2a + 5 a = 2 ± = = flip bif 4-cycle 2

15 Period-doubling cascade Successive period-doubling bifs occur at successively smaller intervals in a and accumulate at a = n a n n a n a n 1 δ n n 1 / n

16 Renormalization Feigenbaum (1979), Collet & Eckmann (1980), Lanford (1982) f(x) = 1 rx 2 on [ 1,1] T acts on mappingsf: (Tf)(x) 1 α f2 ( αx)

17 Seek fixed point of T : f(x) (Tf)(x) 1 rx 2 f 2 ( αx)/α = f(1 r(αx) 2 )/α = (1 r(1 r(αx) 2 ) 2 )/α = (1 r(1 2r(αx) 2 + r 2 (αx) 4 ))/α = (1 r + 2r 2 (αx) 2 r 3 (αx) 4 )/α Matching constant and quadratic terms: r 1 α = 1 2r 2 α 2 α = r r 1 = α 2rα = 1 = 2r(r 1) = 1 α = r = (1 + 3)/2 = 1.366

18 f(x) (Tf)(x) (T 2 f)(x)... φ(x) 2nd order 4th order 8th order... where limiting function φ(x) = x x x is fixed point oft (mapping-of-mappings), i.e. T(φ) = φ Single unstable direction with multiplierδ = (table) φ, δ, are universal for all map families with quadratic maxima, such asasinπx, 1 rx 2,ax(1 x).

19 If f has a stable 2-cycle, thenf 2 has a stable fixed point. More generally, if f has a stable 2 n cycle, thenf 2 has a stable 2 n 1 cycle. Since T involves taking f 2, then applying T takes maps with 2 n cycles to maps with2 n 1 cycles.

20 Other classes of unimodal maps, characterized by the nature of their extrema, undergo a period-doubling cascade. Each class has its own asymptotic value of δ and α. Examples are functions with a quartic maximum or the tent map { ax for x < 1/2 f(x) = a(1 x) for x > 1/2 a < 1: 0 is stable fixed point a > 1: 0 is unstable fixed point x = a/(1 + a) is stable

21 Periodic Windows f 3 has 3 saddle-node bifurcations at a 3 = = = two 3-cycles, one stable and one unstable ( (f 3 ) 1) Stable and unstablen-cycles created whenf n crosses diagonal

22 Bifurcation Diagram for Logistic Map Each n-cycle undergoes period-doubling cascade

23 Three-cycle: before and after a = a = 3.83 a = intermittency 3-cycle 6-cycle Stable three-cycle for = < r < 3.857

24 Intermittency Slow dynamics near a bifurcation Type I Type III near saddle-node bif near period-doubling bif near pitchfork bif g = x n x 2 n f = 1.2 x n x 3 n off 2 Slow dynamics near ghosts of not-yet-created or unstable fixed points Assume that another mechanism re-injects trajectories back tox 0 Type II intermittency associated with a subcritical Hopf bifurcation

25 Sharkovskii Ordering (odd numbers) (multiples of 2) (multiples of 2 2 ) (powers of 2) Sharkovskii s Theorem: f has a k-cycle = f has an l-cycle for any l k f has a 3-cycle = f has an l-cycle for anyl (Most cycles not stable) Logistic map for any r > r 3 has cycles of all lengths.

26 Period-doubling in Rayleigh-Bénard convection From Libchaber, Fauve & Laroche, Physica D (1983). A: freq f and f/2 = B:f/4 = C:f/8 = D:f/16

27 Rayleigh-Bénard convection in small-aspect-ratio container Type III intermittency Timeseries Poincaré map maxima (I k,i k+2 ) Subharmonic (period-doubled signal) grows til burst, followed by laminar phase From M. Dubois, M.A. Rubio & P. Bergé, Phys. Rev. Lett. 51, 1446 (1983).

28 Rayleigh-Bénard convection in small-aspect-ratio container Type I intermittency non-intermittent timeseries Ra = 280 Ra c intermittent timeseries Ra = 300 Ra c From P. Bergé, M. Dubois, P. Manneville & Y. Pomeau, J. Phys. (Paris) Lett. 41, 341 (1980).

29 Poincaré map of Lorenz system 3D trajectory of the Lorenz system Timeseries of (t) for standard chaotic value of r = 28 Timeseries ofz(t) From P. Manneville, Course notes

30 Successive pairs of maxima ofz resemble tent map: From P. Manneville, Course notes

31 From continuous flows to discrete maps limit cycle after pitchfork after period-doubling

32 Where do discrete dynamical systems come from? Hopf or global bifurcations= limit cycles Study these using Poincaré or first return map: Continuous dynamical system x(t) R d, x d (t) goes throughβ for some set of trajectories y n+1 = g(y n ) x(t) = (y n,β), ẋ d (t) > 0 x(t ) = (y n+1,β), ẋ d (t ) > 0 above not satisfied for any t (t,t ) From P. Manneville class notes, DEA Physique des Liquides From Moehlis, Josic, & Shea-Brown, Scholarpedia

33 Floquet theory Linear equations with constant coefficients: aẍ + bẋ + cx = 0 = x(t) = α 1 e λ1t + α 2 e λ 2t where aλ 2 + bλ + c = 0 ẋ = cx = x(t) = e ct x(0) N N c n x (n) = 0 = x(t) = α n e λ nt n=0 where n=1 N c n λ n = 0 n=0 Generalize to linear equations with periodic coefficients: a(t)ẍ + b(t)ẋ + c(t)x = 0 = x(t) = α 1 (t)e λ1t + α 2 (t)e λ 2t a(t),b(t),c(t) have period T = α 1 (t),α 2 (t) have period T

34 Floquet theory continued a(t)ẍ + b(t)ẋ + c(t)x = 0 = x(t) = α 1 (t)e λ 1t + α 2 (t)e λ 2t α 1 (t),α 2 (t) Floquet functions λ 1, λ 2 e λ1t, e λ 2T Floquet exponents Floquet multipliers λ 1,λ 2 not roots of polynomial = must calculate numerically or asymptotically ẋ = c(t)x = x(t) = e λt α(t) N N c n (t)x (n) = 0 = x(t) = e λnt α n (t) n=0 n=1

35 Floquet theory and linear stability analysis Dynamical system: Limit cycle solution: Stability of x(t) : ẋ = f(x) x(t + T) = x(t) with x(t) = f( x(t)) x(t) = x(t) + ǫ(t) x + ǫ = f( x(t)) + f ( x(t))ǫ(t) +... ǫ = f ( x(t))ǫ(t) Floquet form! ǫ(t) = e λt α(t) with α(t) of period T Re(λ) > 0 = x(t) unstable Re(λ) < 0 = x(t) stable λ complex = most unstable or least stable pert has period different from x(t)

36 Region of stability for exponent λ for multiplier e λt Imaginary part non-unique = choose Im(λ) ( πi/t, πi/t] In R N, x(t) stable iff real parts of all λ j are negative Monodromy matrix: Ṁ = Df( x(t))m withm(t = 0) = I Floquet multipliers/functions = eigenvalues/vectors ofm(t)

37 Faraday instability Faraday (1831): Vertical vibration of fluid layer = stripes, squares, hexagons In 1990s: first fluid-dynamical quasicrystals: Edwards & Fauve Kudrolli, Pier & Gollub J. Fluid Mech. (1994) Physica D (1998)

38 Oscillating frame of reference = oscillating gravity G(t) = g(1 acos(ωt)) G(t) = g(1 a[cos(mωt) + δcos(nωt + φ 0 )]) Flat surface becomes linearly unstable for sufficiently higha Consider domain to be horizontally infinite (homogeneous) = solutions exponential/trigonometric in x = (x,y) Seek bounded solutions = trigonometric: exp(ik x) Height ζ(x,y,t) = k e ik xˆζk (t) Oscillating gravity = temporal Floquet problem, T = 2π/ω ˆζ k (t) = j e λj k t f j k (t)

39 Height ζ(x,y,t) = k e ik xˆζ k (t) Ideal fluids (no viscosity), sinusoidal forcing= Mathieu eq. ω t ˆζ k + ω 2 0 [1 acos(ωt)] ˆζ k = 0 combinesg, densities, surface tension, wavenumber k 2 ˆζ k (t) = e λj k t f j k (t) j=1 Re(λ j k ) > 0 for somej,k= ˆζ k ր= flat surface unstable = Faraday waves with wavelength 2π/k Im(λ j k ) eλt waves period 0 1 harmonic same as forcing ω/2 1 subharmonic twice forcing period

40

41 Floquet functions λ (ζ-< ζ >)/ t / T λ (ζ-< ζ >)/ t / T (ζ-< ζ >)/ t / T within tongue 1 /2 within tongue 2 /2 within tongue 3 /2 subharmonic harmonic subharmonic µ = 1 µ = +1 µ = 1 λ From Périnet, Juric & Tuckerman, J. Fluid Mech. (2009)

42 Hexagonal patterns in Faraday instability From Périnet, Juric & Tuckerman, J. Fluid Mech. (2009)

43 Ideal flow Cylinder wake with downstream recirculation zone Von Kármán vortex street (Re 46) Laboratory experiment (Taneda, 1982) Off Chilean coast past Juan Fernandez islands

44 von Kárman vortex street: Re = U d/ν 46 spatially: temporally: two-dimensional (x,y) periodic,st = fd/u (homogeneous inz) appears spontaneously U 2D (x,y,t mod T)

45 Stability analysis of von Kármán vortex street 2D limit cycle U 2D (x,y,tmod T) obeys: t U 2D = (U 2D )U 2D P 2D + 1 Re U 2D Perturbation obeys: t u 3D = (U 2D (t) )u 3D (u 3D )U 2D (t) p 3D + 1 Re u 3D Equation homogeneous inz, periodic in t= u 3D (x,y,z,t) e iβz e λβt f β (x,y,t mod T) Fixβ, calculate largestµ = e λβt via linearized Navier-Stokes

46 From Barkley & Henderson, J. Fluid Mech. (1996) mode A:Re c = β c = = λ c 4 mode B:Re c = 259 β c = 7.64 = λ c 1 Temporally: µ = 1 = steady bifurcation to limit cycle with same periodicity asu 2D Spatially: circle pitchfork (any phase in z)

47 mode A at Re = 210 mode B at Re = 250 From M.C. Thompson, Monash University, Australia ( mct/mct/docs/cylinder.html)

48 transition to mode A initially faster than exp = subcritical transition to mode B initially slower than exp = supercritical

49 Describe via complex amplitudesa n,b n amplitude and phase of mode A, B at fixed instant of cycle A n+1 = ( µ A + α A A n 2 A n β A A n 4) A n B n+1 = ( µ B α B B n 2) B n

50 Numerical and experimental frequencies From Barkley & Henderson, JFM (1996) Solution to minimal model From Barkley, Tuckerman & Golubitsky, PRE (2000) Interaction between A and B: A n+1 = ( µ A + α A A n 2 + γ A B n 2 β A A n 4) A n B n+1 = ( µ B α B B n 2 + γ B A n 2) B n As Re is increased: mode A = A + B = modeb