Home / Courses / CSIR NET Physics / Practice / Nonlinear dynamics and chaos

Nonlinear dynamics and chaos

Classical mechanics (GATE and NET) · 84 questions and templates for CSIR NET Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

New to this topic? Read the lesson first: Nonlinear dynamics and chaos.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. Its fixed points are
    1. x=0 and x=1−1/r
    2. x=0 and x=1/r
    3. x=1/2 only
    4. x=0 and x=r−1
    Answer and solution

    Answer: A

    x=rx(1−x) gives x=0 or 1=r(1−x), i.e. x=1−1/r. The second lies in [0,1] only for r≥1.
  2. MCQWarm-up

    Sensitive dependence on initial conditions, the hallmark of chaos, means that the largest Lyapunov exponent is
    1. zero
    2. negative
    3. positive
    4. imaginary
    Answer and solution

    Answer: C

    Two nearby starting points separate as δn≈δ0eλn. A positive λ means exponential separation (chaos); negative means the orbits converge (a stable fixed point or cycle); zero is typical of periodic or quasi-periodic motion along the flow.
  3. NATWarm-up

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. For r=2.5, iterates from almost any x0 in (0,1) settle to the value ______.
    Answer and solution

    Answer: 0.6 (±0.001)

    For 1<r<3 the non-zero fixed point x∗=1−1/r is stable: x∗=1−0.4=0.6. Its multiplier f′(x∗)=2−r=−0.5 has magnitude below 1, so the iterates approach it, alternating sides.
  4. MCQWarm-up

    What is the smallest dimension of phase space in which an autonomous system of ODEs x→˙=f→(x→) can be chaotic?
    1. 1
    2. 2
    3. 4
    4. 3
    Answer and solution

    Answer: D

    In one dimension trajectories can only move monotonically to fixed points. In two, the Poincaré-Bendixson theorem limits bounded motion to fixed points and limit cycles. Three dimensions are needed (e.g. the Lorenz system). Maps are different: a one-dimensional non-invertible map such as the logistic map can be chaotic.
  5. NATWarm-up

    The middle-thirds Cantor set is made of 2 copies of itself, each scaled down by a factor 3. Its similarity dimension is ______.
    Answer and solution

    Answer: 0.6309 (±0.002)

    For N copies scaled by 1/s: N=sD, so D=ln⁡N/ln⁡s=ln⁡2/ln⁡3≈0.631. More than a set of points (dimension 0), less than a line (1).
  6. MCQWarm-up

    In the period-doubling route to chaos, the ratio of successive intervals between bifurcations, (rn−rn−1)/(rn+1−rn), tends to Feigenbaum's constant
    1. α≈2.503
    2. δ≈4.669
    3. r∞≈3.570
    4. ϕ≈1.618
    Answer and solution

    Answer: B

    δ=4.6692… is universal: the same for every smooth one-humped map (and seen in fluid and electronic experiments). α=2.5029 is the second Feigenbaum constant, for the scaling of the orbit widths; 3.5699 is where the logistic map's cascade ends.
  7. MCQWarm-up

    Which of these maps has a positive Lyapunov exponent for almost every starting point?
    1. x↦x/2
    2. x↦x+αmod1 with α irrational
    3. x↦2xmod1
    4. x↦x2 on [0,1)
    Answer and solution

    Answer: C

    The exponent is the average of ln⁡|f′(x)| along the orbit. x/2: ln⁡12<0. The rotation x+α: f′=1, exponent 0 (nearby points stay the same distance apart, even though the orbit never repeats). 2xmod1: ln⁡2>0, so nearby points separate by a factor 2 each step. x2: every orbit falls to 0, where f′→0, so the exponent is negative.

    Why people get this wrong: A never-repeating orbit (irrational rotation) is not chaotic: there is no stretching.

  8. NATWarm-up

    The Lyapunov exponent of the map xn+1=3xnmod1 is ______ per iteration (to three decimal places).
    Answer and solution

    Answer: 1.099 (±0.003)

    |f′(x)|=3 everywhere (the jumps of the "mod 1" do not affect the slope), so λ=lim⁡1n∑ln⁡|f′(xi)|=ln⁡3=1.099. A small error triples each step.

    Why people get this wrong: The exponent is ln⁡3, not 3: it is a rate of exponential growth.

  9. MCQStandard

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. The fixed point x∗=1−1/r is stable for
    1. 0<r<1
    2. 1<r<3
    3. 3<r<4
    4. r close to 4 only
    Answer and solution

    Answer: B

    f′(x)=r(1−2x), so f′(x∗)=r(1−2+2/r)=2−r. Stability needs |2−r|<1: 1<r<3. At r=1 it exchanges stability with x=0 (transcritical); at r=3 the multiplier reaches −1 and the first period doubling happens.
  10. NATStandard

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. For r=2.8, the multiplier f′(x∗) at the non-zero fixed point is ______.
    Answer and solution

    Answer: -0.8 (±0.001)

    f′(x∗)=2−r=−0.8. Its magnitude is below 1, so the point is stable, and the negative sign means the iterates approach it alternately from above and below. Each step shrinks the distance by 0.8.
  11. MCQStandard

    The doubling map xn+1=2xnmod1 has Lyapunov exponent
    1. ln⁡2
    2. 2
    3. 0
    4. −ln⁡2
    Answer and solution

    Answer: A

    |f′(x)|=2 everywhere, so λ=ln⁡2≈0.693 per step: errors double each iteration. In binary, the map shifts the digits of x one place left, so each step reveals one more digit of the initial condition: one bit of information per step.
  12. NATStandard

    Each step of building the Koch curve replaces every segment by 4 segments, each one-third as long. The similarity dimension of the Koch curve is ______.
    Answer and solution

    Answer: 1.2619 (±0.002)

    D=ln⁡4/ln⁡3≈1.262. The length grows by 4/3 at each stage and becomes infinite, but the curve does not fill an area: its dimension is between 1 and 2.
  13. NATStandard

    The Lorenz equations are x˙=σ(y−x), y˙=rx−y−xz, z˙=xy−bz, with σ=10, b=8/3, r=28. Phase-space volumes shrink as V(t)=V0e−kt. The value of k is ______.
    Answer and solution

    Answer: 13.667 (±0.01)

    The rate of change of volume is the divergence: ∂xx˙+∂yy˙+∂zz˙=−σ−1−b=−13.67, the same everywhere. So k=σ+1+b≈13.67; volumes shrink by e−13.67 per unit time, and the attractor has zero volume.
  14. MCQStandard

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. At r=3, what happens?
    1. The fixed point x=0 becomes stable
    2. The map becomes chaotic
    3. A period-3 orbit appears
    4. The fixed point 1−1/r loses stability as its multiplier passes −1, and a stable 2-cycle is born
    Answer and solution

    Answer: D

    At r=3, f′(x∗)=2−r=−1: a flip (period-doubling) bifurcation. Beyond it the iterates alternate between two values. The next doubling, to period 4, comes at r=1+6≈3.449, and chaos sets in near r∞≈3.5699.
  15. MSQStandard

    Which properties describe deterministic chaos?
    1. The rule of evolution is completely deterministic
    2. The long-term behaviour is aperiodic
    3. Nearby initial conditions separate exponentially fast
    4. It needs random noise to appear
    Answer and solution

    Answer: A, B, C

    Chaos is aperiodic long-term behaviour in a deterministic system that shows sensitive dependence on initial conditions. No randomness is put in: the apparent randomness comes from stretching and folding of phase space, which amplifies tiny uncertainties in the initial state.
  16. NATStandard

    A chaotic map has Lyapunov exponent 0.5 per iteration. An initial uncertainty of 10−6 grows to order 1 after about ______ iterations.
    Answer and solution

    Answer: 27.6 (±0.3)

    10−6e0.5n=1 gives n=ln⁡(106)/0.5=13.82/0.5≈27.6. Improving the initial accuracy a million times more (10−12) only doubles this horizon: predictability grows logarithmically with precision.
  17. MCQStandard

    A strange attractor differs from a limit cycle or a stable fixed point in that it
    1. repels nearby trajectories
    2. exists only in Hamiltonian systems
    3. has a fractal structure (non-integer dimension) and motion on it is sensitive to initial conditions
    4. always contains exactly one fixed point
    Answer and solution

    Answer: C

    Like any attractor it pulls in a set of initial conditions (dissipative systems only). Inside it, nearby orbits separate exponentially, while the repeated stretching and folding gives it a Cantor-like layered structure of non-integer dimension (about 2.06 for the Lorenz attractor).
  18. MCQStandard

    The Hénon map is xn+1=1−axn2+yn, yn+1=bxn. Each iteration multiplies areas in the plane by
    1. |b|
    2. a
    3. 1
    4. 2ax
    Answer and solution

    Answer: A

    The Jacobian matrix is (−2ax1b0), with determinant −b, the same everywhere. So areas scale by |b| per step (0.3 for the classic a=1.4, b=0.3): the map is dissipative and has a strange attractor.
  19. MCQStandard

    Which of these maps is invertible (each point has exactly one pre-image)?
    1. The logistic map x↦rx(1−x)
    2. The tent map
    3. The Hénon map with b≠0
    4. The doubling map x↦2xmod1
    Answer and solution

    Answer: C

    The logistic, tent and doubling maps are two-to-one: two values of x map to the same image (folding). The Hénon map can be inverted explicitly: xn=yn+1/b and yn=xn+1−1+axn2. That is why it can model a Poincaré section of a real (invertible) flow.
  20. MCQStandard

    The power spectrum of a signal recorded on a chaotic attractor typically shows
    1. a few sharp lines at one frequency and its harmonics
    2. sharp lines at two incommensurate frequencies and their combinations
    3. a single sharp line
    4. a broad continuous band, possibly with some peaks on top
    Answer and solution

    Answer: D

    A periodic signal has lines at f0,2f0,3f0,… A quasi-periodic one (motion on a torus) has lines at mf1+nf2. Chaos never repeats and loses memory of its past (correlations decay), which spreads power over a continuous band. Broadband spectra are one of the classic experimental signatures of the onset of chaos, for example in convection experiments.

    Why people get this wrong: Many lines from two incommensurate frequencies can look busy, but they are still sharp lines, not a continuum.

  21. NATStandard

    A driven, damped pendulum obeys θ¨+γθ˙+sin⁡θ=Acos⁡ωt with γ=0.5 s⁻¹. In its three-dimensional phase space (θ,θ˙,ωt), a small blob of initial conditions shrinks in volume by the factor ______ in 2 s (to three decimal places).
    Answer and solution

    Answer: 0.368 (±0.003)

    Write the flow as θ˙=v, v˙=−γv−sin⁡θ+Acos⁡ϕ, ϕ˙=ω. Its divergence is ∂θ˙/∂θ+∂v˙/∂v+∂ϕ˙/∂ϕ=0−γ+0=−γ, a constant. So volumes shrink as e−γt=e−0.5×2=e−1=0.368, whether the motion is regular or chaotic.

    Why people get this wrong: Only the damping term depends on its own variable; the forcing and sin⁡θ do not contribute to the divergence.

  22. NATExam level

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. The 2-cycle points are x±=(r+1)±(r−3)(r+1)2r. For r=3.2, the larger point of the cycle is ______.
    Answer and solution

    Answer: 0.7995 (±0.001)

    x+=4.2+0.2×4.26.4=4.2+0.91656.4=0.7995; the other is 0.5130. Check: 3.2×0.5130×0.4870=0.7995, and 3.2×0.7995×0.2005=0.5130.
  23. MCQExam level

    The logistic map is xn+1=rxn(1−xn), with 0≤x≤1 and 0<r≤4. The stable 2-cycle born at r=3 itself loses stability (giving period 4) at
    1. r=3.000
    2. r=1+6≈3.449
    3. r=3.570
    4. r=1+8≈3.828
    Answer and solution

    Answer: B

    The 2-cycle's multiplier is f′(x+)f′(x−)=r2(1−2x+)(1−2x−)=4+2r−r2. It reaches −1 when r2−2r−5=0: r=1+6. 3.570 is the end of the cascade and 1+8 the start of the period-3 window.
  24. NATExam level

    For a one-dimensional map the Lyapunov exponent is λ=limn→∞⁡1n∑i=0n−1ln⁡|f′(xi)|. The tent map x↦μmin⁡(x,1−x) with μ=1.8 has Lyapunov exponent ______.
    Answer and solution

    Answer: 0.5878 (±0.005)

    |f′(x)|=μ at every point (except the peak), so every term in the sum is ln⁡μ and λ=ln⁡1.8≈0.588. Positive, so chaotic. The tent map is chaotic for every 1<μ≤2.
  25. NATExam level

    The first three period-doubling points of the logistic map are r1=3, r2=3.4495 and r3=3.5441. The estimate of Feigenbaum's δ from them, (r2−r1)/(r3−r2), is ______.
    Answer and solution

    Answer: 4.75 (±0.03)

    0.44950.0946≈4.75, already close to the limit δ=4.669 that later ratios approach.
  26. NATExam level

    For the Lorenz system x˙=σ(y−x), y˙=rx−y−xz, z˙=xy−bz with r=28, the two non-trivial fixed points C± have z-coordinate ______.
    Answer and solution

    Answer: 27

    x˙=0 gives y=x. Then y˙=x(r−1−z)=0 gives z=r−1=27 (for x≠0), and z˙=0 gives x2=bz, so x=y=±b(r−1)=±8.49 for b=8/3. These are the centres of the two "wings" of the butterfly; they are unstable for r>24.74.
  27. MCQExam level

    For a continuous map of an interval to itself, the existence of an orbit of period 3 implies
    1. that the map is linear
    2. that there are no other periodic orbits
    3. that every orbit is periodic
    4. that there are periodic orbits of every period (and chaotic orbits)
    Answer and solution

    Answer: D

    Sharkovskii's theorem (and the Li-Yorke paper "Period three implies chaos"): period 3 comes first in Sharkovskii's ordering, so it forces every other period. For the logistic map the period-3 window opens at r=1+8, deep in the chaotic region.
  28. NATExam level

    The Sierpinski triangle consists of 3 copies of itself, each scaled by one half. Its similarity dimension is ______.
    Answer and solution

    Answer: 1.585 (±0.002)

    D=ln⁡3/ln⁡2≈1.585. Its area is zero (each step removes a quarter of the remainder) but it is more than a curve.
  29. MSQExam level

    For the Lorenz attractor at the classic parameters, which statements are correct?
    1. The sum of the three Lyapunov exponents equals −(σ+1+b)
    2. Exactly one Lyapunov exponent is positive
    3. Trajectories cross each other where the two wings meet
    4. Its fractal dimension is slightly above 2
    Answer and solution

    Answer: A, B, D

    The exponents sum to the average divergence, −(σ+1+b)≈−13.67; numerically they are about 0.91,0,−14.57. Trajectories never cross in phase space (uniqueness); the wings only look as if they meet in a 2-D projection. The Kaplan-Yorke dimension is 2+0.91/14.57≈2.06.
  30. MCQExam level

    The KAM (Kolmogorov-Arnold-Moser) theorem says that when an integrable Hamiltonian system is perturbed slightly,
    1. all invariant tori are destroyed at once
    2. most invariant tori, those with sufficiently irrational frequency ratios, survive slightly deformed
    3. the motion becomes dissipative
    4. only tori with rational frequency ratios survive
    Answer and solution

    Answer: B

    Tori whose frequency ratio is badly approximated by rationals survive small perturbations; resonant (rational) tori break up into chains of islands with chaotic layers between them. As the perturbation grows, more tori break; the last to go have frequency ratios related to the golden mean.

← All topics · Timed mock test →