Phase Space Topology Slides Phase Space Topology
Analytical Mechanics of Oscillatory Systems
Lesson 1: Geometry of Dynamic States
The Geometry of States
01 / Phase Space
In analytical mechanics, we shift from tracking a particle's position \( x(t) \) to its state in Phase Space \((q, p)\).
A single point represents the complete state of the system.
Dynamics are viewed as flow lines (trajectories) in this space.
Uniqueness theorem: Trajectories in phase space cannot cross .
Generalized Position (q)
Momentum (p)
Phase Portrait of an undamped SHO
Harmonic Oscillator Topology
02 / SHO Analysis
Energy Conservation
\[ E = \frac{p^2}{2m} + \frac{1}{2}m\omega^2 q^2 \]
The trajectory is an ellipse in \((q, p)\) space. If we scale coordinates such that \( P = p/\sqrt{m} \) and \( Q = q\sqrt{m\omega^2} \), the trajectory is a circle .
Constant energy surfaces in phase space represent the topology of the system's evolution.
Area \( A = \oint p \, dq \)
The area enclosed by a trajectory is proportional to the total energy \( E \).
\[ A = \frac{2\pi E}{\omega} \]
Fixed Points & Stability
03 / Stability
Center
Small perturbations lead to orbits around the point.
Stable, not Asymptotic
Stable Sink
Perturbations decay; trajectories spiral into the fixed point.
(Damped Oscillator)
Asymptotically Stable
Saddle Point
Some paths approach, but most diverge exponentially.
(Inverted Pendulum)
Unstable
The Geometry of History
"How can a simple geometric loop completely describe the energy and history of a dynamic system without explicit time dependence?"
Thought Experiment
Consider a system where the "flow" in phase space converges to a single loop rather than a point. What physical phenomenon does this represent?
Limit Cycles Non-linear Oscillators Van der Pol
Stability Analysis Workshop Phase Space Topology Workshop
Analytical Mechanics of Oscillatory Systems
NAME:
DATE:
1
Constructing Phase Portraits
Consider a simple harmonic oscillator with mass \( m \) and spring constant \( k \). The total energy is given by: \[ E = \frac{p^2}{2m} + \frac{1}{2}kx^2 \]
1.1. Derive the equation for the trajectory in the \((x, p)\) phase plane. What geometric shape does this represent? How do the axes scale with energy?
1.2. On the axes below, sketch three trajectories corresponding to energies \( E_1 < E_2 < E_3 \). Indicate the direction of flow with arrows.
Position (x)
Momentum (p)
2
Critical Points and Stability
Analyze the stability of the following non-linear system's fixed points: \[ \dot{x} = y, \quad \dot{y} = -x + x^3 \]
2.1. Identify all fixed points \((x^*, y^*)\) for this system.
2.2. Linearize the system around the fixed point at \((0, 0)\). Calculate the eigenvalues of the Jacobian matrix and determine the type of stability (Center, Saddle, Node, etc.).
2.3. Physical Interpretation: This system represents a "Double-Well Potential". Explain how the phase portrait at \(( \pm 1, 0 )\) differs from the portrait at the origin.
3
Damping and Topology
A damping term \(-b\dot{x}\) is added to a harmonic oscillator. Describe how the topology of the phase space trajectories changes (e.g., from closed loops to spirals). Does the "no-crossing" rule still hold? Explain.
Phase Space Teacher Guide Phase Space Instruction Guide
Teacher Resource | Lesson 1
Sequence: Analytical Mechanics
Instructional Intent
The goal of this lesson is to move students away from time-domain analysis (\(x\) vs \(t\)) and towards state-space geometry. By the end of this session, students should view a "system" not as a single particle moving in space, but as a "representative point" flowing through a higher-dimensional manifold.
Core Concepts
Phase Space Mapping
Energy Surfaces
Lyapunov Stability
Jacobian Linearization
Lesson Roadmap
1
The Hook (10 mins)
Discuss Slide 5. Challenge students to think about why trajectories can't cross. Answer: If they crossed, the future state would not be uniquely determined by the initial state (violates Determinism).
2
Visualizing SHO (20 mins)
Work through Slide 3. Emphasize that the area of the ellipse is constant. This is the first hint at Action Variables (Lesson 5) and Liouville's Theorem (Lesson 3).
3
The Workshop (60 mins)
Students complete the "Stability Analysis Workshop". Focus heavily on Section 2 (Linearization). Graduate students should be comfortable with the Jacobian but might struggle with interpreting the physical meaning of the saddle points in the double-well potential.
Common Pitfalls
Flow vs. Velocity: Students often confuse the "velocity" in phase space \(\vec{v} = (\dot{q}, \dot{p})\) with the physical velocity \(\dot{q}\).Stability vs. Equilibrium: A center is stable but not "asymptotically stable." Ensure they understand the difference: perturbations don't decay; they just don't grow.
Discussion Prompts
"If a system is dissipative (damped), what happens to the area of the phase space ellipses over time?"
"How would the phase portrait change if we added a driving force?"
Lagrangian Dynamics Slides Lagrangian Dynamics of
Complex Oscillators
Analytical Mechanics Sequence
Lesson 2: The Action Principle and Constraints
Hamilton's Principle
01 / Foundation
The trajectory of a system is the path that makes the Action Integral stationary:
\[ S = \int_{t_1}^{t_2} \mathcal{L}(q, \dot{q}, t) \, dt \]
Lagrangian: \( \mathcal{L} = T - V \)Works in generalized coordinates \( q_i \).
Euler-Lagrange Equation
\[ \frac{d}{dt} \left( \frac{\partial \mathcal{L}}{\partial \dot{q}_i} \right) - \frac{\partial \mathcal{L}}{\partial q_i} = 0 \]
One equation for each degree of freedom. Forces of constraint vanish automatically if coordinates are chosen wisely.
Multiple Degrees of Freedom
02 / Scaling Complexity
For small oscillations near equilibrium, the Lagrangian can be expanded as:
\[ \mathcal{L} = \frac{1}{2} \sum_{i,j} (M_{ij} \dot{q}_i \dot{q}_j - K_{ij} q_i q_j) \]
Mass Matrix (\(M\))
Kinetic energy coupling.
Stiffness Matrix (\(K\))
Potential energy coupling.
The Eigenvalue Problem
\[ \text{det}(K - \omega^2 M) = 0 \]
Solving this yields the Normal Modes and Characteristic Frequencies of the system.
Constraints & Penalties
03 / Advanced Modeling
Holonomic Constraints
Representable as \( f(q, t) = 0 \). Reduces DOF by 1.
Lagrange Multipliers
Used when constraint forces are needed explicitly.
\[ \mathcal{L}' = \mathcal{L} + \lambda f(q) \]
"Why does nature minimize action, and how does this principle simplify the calculation of a complex pendulum's period?"
How many generalized coordinates would you use for a double pendulum constrained to move on a sphere?
Transitioning to Hamilton
While the Lagrangian focuses on configurations (\(q\)), we will soon discover that momentum (\(p\)) is not just a secondary derivative, but a fundamental axis of existence.
L = T - V
H = T + V
Complex Oscillator Lagrangian Workshop Complex Oscillator Derivations
Lagrangian Formulation Workshop
STUDENT:
1
The Double Pendulum
A double pendulum consists of two masses \( m_1 \) and \( m_2 \) connected by rigid rods of lengths \( L_1 \) and \( L_2 \). Let \( \theta_1 \) and \( \theta_2 \) be the angles with the vertical.
1.1. Express the Cartesian coordinates \((x_1, y_1)\) and \((x_2, y_2)\) in terms of the generalized coordinates \( \theta_1, \theta_2 \).
1.2. Write the Lagrangian \( \mathcal{L} = T - V \) for the system. (Hint: Be careful with the kinetic energy of the second mass).
2
Coupled Mass-Spring System
Two identical masses \( m \) are on a frictionless table, connected to each other by a spring \( k \), and each connected to opposite walls by springs of the same constant \( k \).
2.1. Construct the Mass Matrix \( M \) and Stiffness Matrix \( K \) for this system using displacements \( x_1, x_2 \).
Mass Matrix M
Stiffness Matrix K
2.2. Solve the secular equation \( \text{det}(K - \omega^2 M) = 0 \) to find the normal mode frequencies.
2.3. Describe the physical motion associated with each normal mode. What happens to the center of mass in each mode?
3
Constraint Forces
A particle of mass \( m \) is constrained to move on the inner surface of a vertical cone with half-angle \( \alpha \) under gravity. Using a Lagrange multiplier \( \lambda \), derive the equations of motion and an expression for the normal force exerted by the cone on the particle.
Hamiltonian Dynamics Slides Hamiltonian Dynamics
Liouville's Theorem & Phase Fluid
"If we track a cloud of initial conditions, why does the 'fluid' of probability neither compress nor expand over time?"
From Lagrangian to Hamiltonian
01 / Symmetry Shift
We transition from configuration space \((q, \dot{q})\) to phase space \((q, p)\) via the Legendre Transformation .
Generalized Momentum
\[ p_i = \frac{\partial \mathcal{L}}{\partial \dot{q}_i} \]
The Hamiltonian
\[ \mathcal{H}(q, p, t) = \sum p_i \dot{q}_i - \mathcal{L} \]
Hamilton's Canonical Equations
\[ \dot{q}_i = \frac{\partial \mathcal{H}}{\partial p_i} \] Kinematics
\[ \dot{p}_i = -\frac{\partial \mathcal{H}}{\partial q_i} \] Dynamics
Conservation of Phase Volume
02 / Liouville
The Incompressibility of State
The phase-space distribution function \(\rho(q, p, t)\) is constant along the trajectories of the system.
\[ \frac{d\rho}{dt} = 0 \]
Equivalent to saying the divergence of the Hamiltonian flow vector field is zero: \(\nabla \cdot \vec{v} = 0\).
Incompressible Fluid Analogy
While the shape of a region in phase space may deform (shear, stretch), its volume remains invariant.
Bridge to Statistical Mechanics
03 / Ensembles
Ensemble Dynamics
Representing uncertainty in initial conditions as a cloud of points.
Poincaré Recurrence
A direct consequence: any state will eventually return arbitrarily close to its initial configuration.
Microcanonical Ensemble
Equiprobability on the energy surface shell.
Liouville's Theorem is the reason why Hamiltonian systems cannot have attractors .
Liouville Theorem Inquiry Worksheet Liouville's Theorem Inquiry
Conservation of Phase Volume
RESEARCHER:
1
Proving Incompressibility
Let \(\vec{v} = (\dot{q}_1, \dots, \dot{q}_n, \dot{p}_1, \dots, \dot{p}_n)\) be the velocity vector of a point in phase space. Liouville's Theorem states that the divergence of this flow is zero: \(\nabla \cdot \vec{v} = 0\).
1.1. Write the expression for \(\nabla \cdot \vec{v}\) in terms of partial derivatives of \(\dot{q}_i\) and \(\dot{p}_i\).
1.2. Substitute Hamilton's canonical equations into your expression from 1.1 to prove that the divergence vanishes. Show all steps.
2
The Deformation of Ensembles
Consider a rectangular region in phase space defined by \( \Delta q \times \Delta p \). As the system evolves, this region shears and stretches.
"Liouville's Theorem implies that Hamiltonian systems cannot have 'attractors' or 'repellers' in the sense of dissipative systems."
2.1. Explain why a stable fixed point (like a damped oscillator's origin) violates Liouville's Theorem. What happens to the volume of a cloud of points approaching a sink?
2.2. If a system is chaotic, trajectories diverge exponentially. How is this reconciled with Liouville's Theorem? Hint: Think about the shape of the volume over time.
3
Poincaré Recurrence Theorem
The Poincaré Recurrence Theorem states that for certain systems, almost all initial states will eventually return arbitrarily close to the initial state.
Required Condition: The motion must be confined to a finite volume of phase space (e.g., constant energy surface).
3.1. Outline a qualitative proof of Poincaré Recurrence using Liouville's Theorem. Why is the "finite volume" condition necessary?
Canonical Transformations Slides Canonical Transformations
Generating Functions & Simplification
"Can we mathematically transform a moving oscillator into a stationary point to solve its equations of motion trivially?"
Transformative Dynamics
01 / Motivation
We seek a transformation from \((q, p)\) to \((Q, P)\) such that the Hamiltonian form is preserved.
Preserving Canonical Form:
\[ \dot{Q} = \frac{\partial \mathcal{K}}{\partial P}, \quad \dot{P} = -\frac{\partial \mathcal{K}}{\partial Q} \]
where \(\mathcal{K}\) is the new Hamiltonian.
Why do this?
To find Cyclic Coordinates : If \(\mathcal{K}\) is independent of \(Q_i\), then \(P_i\) is a constant of motion.
To transform the problem into one that is trivial to solve (Equilibrium).
Generating Functions
02 / Mechanics
The transformation is "generated" by a function that bridges the old and new coordinates.
Type 2 Generator: \( F_2(q, P, t) \)
\[ p = \frac{\partial F_2}{\partial q} \]
\[ Q = \frac{\partial F_2}{\partial P} \]
\[ \mathcal{K} = \mathcal{H} + \frac{\partial F_2}{\partial t} \]
Symplectic Condition
A transformation is canonical if and only if it preserves the Poisson Brackets : \[ \{Q, P\}_{q,p} = 1 \]
This geometric structure ensures that phase volume is conserved throughout the transformation.
Harmonic Oscillator Case
03 / Implementation
For the SHO, we choose a transformation such that the new Hamiltonian \(\mathcal{K}\) is a constant .
Transformation Equations:
\[ q = \sqrt{\frac{2P}{m\omega}} \sin Q \]
\[ p = \sqrt{2m\omega P} \cos Q \]
The Result
\[ \mathcal{K} = \omega P \]
The new coordinate \(Q\) is cyclic !
\[ \dot{P} = -\frac{\partial \mathcal{K}}{\partial Q} = 0 \implies P = \text{const} \]
\[ \dot{Q} = \frac{\partial \mathcal{K}}{\partial P} = \omega \implies Q = \omega t + \beta \]
Canonical Transformation Workshop Worksheet Canonical Transformation Workshop
Generating Functions & Poisson Brackets
ANALYST:
1
Testing Canonicity
A transformation is canonical if it preserves the structure of Hamilton's equations. This is equivalent to checking the fundamental Poisson Brackets.
1.1. Consider the transformation \( Q = \ln(1 + \sqrt{q}\cos p) \) and \( P = 2(1 + \sqrt{q}\cos p)\sqrt{q}\sin p \). Use Poisson Brackets to determine if this transformation is canonical. Show your derivation.
2
Type 2 Generator Application
A system has the Hamiltonian \( \mathcal{H} = \frac{p^2}{2} + \frac{q^2}{2} \). We want to transform to a new coordinate system using the generating function \( F_2(q, P) = q P \).
2.1. Derive the transformation equations for \( Q(q, p) \) and \( P(q, p) \). What kind of transformation does this generator produce (e.g., identity, exchange, etc.)?
2.2. If we instead use \( F_1(q, Q) = q Q \), what are the resulting transformation equations? How does this swap the roles of position and momentum?
3
Simplifying the Harmonic Oscillator
Recall the SHO transformation: \( q = \sqrt{\frac{2P}{m\omega}} \sin Q, \quad p = \sqrt{2m\omega P} \cos Q \).
3.1. Solve for the new Hamiltonian \( \mathcal{K} = \mathcal{H}(q, p) \) in terms of the new coordinates. Confirm that \( Q \) is indeed cyclic.
3.2. Physical Insight: What physical quantity does the new momentum \( P \) represent? (Hint: Consider its units and its relation to the total energy \( E \)).
3.3. Write the solution for \( q(t) \) and \( p(t) \) using the solutions for \( Q(t) \) and \( P(t) \).
Action-Angle Variables Slides Action-Angle Variables
Adiabatic Invariants & Integrability
"What happens to the amplitude of a pendulum if the string length is shortened very slowly, and why is the energy not conserved?"
The Action Variable \( J \)
01 / Definitions
For periodic systems, the Action Variable is defined as the integral of momentum over a complete cycle:
\[ J = \oint p \, dq \]
Geometrically, \( J \) is the area enclosed by the orbit in phase space.
\( J \) is a constant of motion for an integrable system.
The Angle Variable \( w \)
The conjugate coordinate to \( J \). Its evolution is linear in time:
\[ \dot{w} = \frac{\partial \mathcal{H}}{\partial J} = \nu \]
Where \( \nu \) is the frequency of the oscillation.
Frequencies of Periodic Motion
02 / Power of AA Variables
The Great Simplification
The primary advantage of Action-Angle variables is that they allow us to calculate the oscillation frequency without solving the full equations of motion.
1. Express \( p \) as \( p(q, E) \).
2. Calculate \( J = \oint p(q, E) \, dq \).
3. Solve for \( E(J) \).
4. Identify frequency \( \nu = \frac{dE}{dJ} \).
Harmonic Oscillator Check
\[ J = \oint p \, dq = \frac{E}{\nu} \] \[ E = \nu J \] \[ \frac{dE}{dJ} = \nu \quad \text{(Self-consistent!)} \]
Adiabatic Invariants
03 / Slow Variation
If a parameter \(\lambda\) of the system (like spring constant or length) changes very slowly compared to the period \( T \):
\( J \) remains constant.
The energy \( E \) and frequency \( \omega \) may change, but the phase space area \( J \) is an invariant.
Example: Shortening a Pendulum
As length \( L \downarrow \), frequency \( \omega \uparrow \). Since \( J = E/\omega \) is invariant, the energy \( E \) must increase!
Work is done on the system by the external force shortening the string.
Adiabatic Invariance Case Study Worksheet Adiabatic Invariance Case Study
Action-Angle Variables Mastery
RESEARCHER:
1
Action Calculation for a Particle in a Box
Consider a particle of mass \( m \) moving freely in a 1D box of length \( L \) with perfectly elastic walls. The energy is purely kinetic \( E = p^2/2m \) between the walls.
1.1. Calculate the Action Variable \( J = \oint p \, dq \) for one complete cycle of the motion. Express \( J \) in terms of \( m, E, \) and \( L \).
1.2. Invert this relationship to find the energy \( E(J) \). Then, find the frequency of the oscillation \( \nu = dE/dJ \). Does this match the classical frequency of a particle in a box?
2
The Slowly Shrinking Box
Suppose the walls of the box are slowly moved together such that the length \( L(t) \) changes adiabatically.
2.1. Using the principle of adiabatic invariance, how does the energy \( E \) of the particle scale with the length \( L \)? (e.g., \( E \propto L^n \)?). Find \( n \).
2.2. Show that the work done on the particle by the moving walls during a displacement \( dL \) is consistent with this energy change. (Hint: Consider the average force exerted by the particle on the walls).
3
Synthesis: The Analytical Framework
Reflect on the sequence "Analytical Mechanics of Oscillatory Systems".
3.1. How does the transformation from Hamiltonian mechanics to Action-Angle variables encapsulate the "Integrability" of a system?
3.2. Essential Question Response: Why is the phase space formulation more powerful than Newtonian mechanics for analyzing the long-term stability and invariance of oscillatory systems?
