How To Draw Phase Portrait: A Complete Step-by-Step Guide
A phase portrait is a geometric representation of the trajectories of a dynamical system in the phase plane, mapping all possible states as functions of time. Mastering this sketching technique requires identifying equilibrium points, calculating eigenvalues and eigenvectors, and determining nullclines to accurately capture system stability and global behavior.
Mathematical Foundations and Preparation Checklist
Before attempting to sketch a phase portrait for a system of ordinary differential equations (ODEs), you must establish a structured mathematical workflow. Phase portraits translate abstract algebraic calculations into spatial geometry, where each axis represents a state variable of the system.
- Essential Tools and Materials: High-precision graph paper or digital tablet with vector graphing software, sharp graphite pencils (2H and HB), a metric ruler, and a scientific calculator capable of computing eigenvalues and eigenvectors.
- Mandatory Prerequisite Knowledge: Proficiency in multivariable calculus, linear algebra (specifically matrix diagonalization, determinants, and trace-determinant planes), and the qualitative theory of autonomous differential equations.
- Estimated Benchmark Duration: 30 to 45 minutes for a standard two-dimensional linear or weakly nonlinear autonomous system.
Step-by-Step Guide to Constructing a Phase Portrait
Step 1: Define the System and Locate Equilibrium Points
Begin by writing your two-dimensional autonomous system in standard form, setting the derivatives of the state variables equal to zero. Let the system be represented by dx/dt = f(x, y) and dy/dt = g(x, y). Solve the simultaneous algebraic equations f(x, y) = 0 and g(x, y) = 0 to find all critical points (equilibrium points) where the system is completely at rest.
Pro-Tip: Mark these equilibrium points clearly on your coordinate plane using solid dots for stable nodes and open circles for saddles or unstable nodes before sketching any directional curves.
Step 2: Calculate the Jacobian Matrix and Classify Equilibrium Points
Compute the Jacobian matrix J of partial derivatives for your system at each equilibrium point. Evaluate the trace (tr(J)) and determinant (det(J)) of the Jacobian matrix to classify the local behavior of each critical point using the standard trace-determinant plane.
- Calculate the partial derivatives: df/dx, df/dy, dg/dx, and dg/dy.
- Substitute the coordinates of a specific equilibrium point into the Jacobian matrix.
- Determine the eigenvalues (lambda) by solving the characteristic equation det(J - lambda * I) = 0.
- Categorize the point as a stable/unstable node, saddle point, center, or spiral based on the real and imaginary parts of the eigenvalues.
Step 3: Sketch Eigenvectors and Separatrices
For linear systems or near equilibrium points of nonlinear systems, calculate the eigenvectors associated with real eigenvalues. Draw these eigenvectors through the equilibrium point with dashed lines; they act as invariant manifolds or separatrices that bound the flow of trajectories.
Warning: Never allow trajectories to cross a separatrix line unless they originate or terminate directly at the saddle point itself.
Step 4: Determine Nullclines and Direction Fields
Find the nullclines of the system by setting dx/dt = 0 (horizontal nullclines) and dy/dt = 0 (vertical nullclines). Plot these curves lightly in pencil, as they divide the phase plane into regions where the signs of dx/dt and dy/dt remain constant. Compute directional arrows in each region to show whether trajectories are moving up, down, left, or right.
Step 5: Connect Trajectories and Refine the Global Portrait
Synthesize your nullclines, equilibrium points, and eigenvector directions to sketch continuous solution curves (trajectories). Ensure that arrows along the trajectories point in the direction of increasing time. Smoothly connect the curves, paying attention to asymptotic behavior near infinity or boundary limits.
How To Draw A Phase Portrait In Matlab
Analytical Comparison of Equilibrium Classifications
| Equilibrium Type | Trace (tr J) & Determinant (det J) Criteria | Eigenvalue Characteristics | Local Phase Portrait Geometry |
|---|---|---|---|
| Saddle Point | det(J) < 0 | Real, opposite signs | Trajectories approach along one axis and repel along the other. |
| Stable Node | tr(J) < 0, det(J) > 0, tr(J)^2 >= 4 det(J) | Real, negative | All trajectories converge directly into the equilibrium point. |
| Unstable Node | tr(J) > 0, det(J) > 0, tr(J)^2 >= 4 det(J) | Real, positive | All trajectories diverge outward away from the equilibrium point. |
| Center | tr(J) = 0, det(J) > 0 | Purely imaginary, conjugate pairs | Closed concentric orbits circling neutrally around the origin. |
| Stable Spiral | tr(J) < 0, det(J) > 0, tr(J)^2 < 4 det(J) | Complex with negative real parts | Inward-spiraling trajectories winding into the steady state. |
Troubleshooting Common Phase Portrait Sketching Errors
- Root Cause: Trajectories intersecting one another in the interior of the phase plane.
- Actionable Fix: Remember the uniqueness theorem for ordinary differential equations states that trajectories for autonomous systems cannot cross. Re-evaluate your sign checks in the nullcline regions to correct vector orientations.
- Root Cause: Misidentifying a nonlinear center due to linearization limitations.
- Actionable Fix: When the Jacobian yields pure imaginary eigenvalues (a linear center), apply Lyapunov direct methods or check for Hamiltonian symmetry to confirm whether the nonlinear system is a true center or an asymptotic spiral.
- Root Cause: Inaccurate curvature near saddle points.
- Actionable Fix: Always calculate the exact eigenvectors for saddle points and ensure your sketched trajectories are tangent to the stable eigenvector as they approach the node.
Frequently Asked Questions
What is the primary purpose of a phase portrait?
A phase portrait provides a qualitative overview of all possible solutions to a system of differential equations without requiring explicit analytical formulas for every trajectory. It instantly reveals system stability, cyclic behavior, and long-term asymptotic states.
How do I know which direction to draw the arrows on trajectories?
Arrowheads on phase portrait trajectories always point in the direction of increasing time (t going forward). You can determine the correct orientation by testing a single coordinate point within a region and evaluating the signs of dx/dt and dy/dt.
Can phase portraits be drawn for three-dimensional systems?
Yes, three-dimensional phase portraits exist, but they are significantly more complex to visualize and manually sketch on a two-dimensional page. They often require 3D vector software, projection slices, or advanced manifold analysis to interpret correctly.
What is the difference between a nullcline and a separatrix?
Nullclines are curves where either the horizontal or vertical derivative equals zero, dividing the plane into directional zones. Separatrices are specific trajectories, often lying along eigenvectors of saddle points, that separate qualitatively different types of global behavior.
How do nonlinear systems alter basic phase portraits?
Linear systems maintain consistent behavior across the entire plane, whereas nonlinear systems feature multiple equilibrium points and local behaviors that change depending on spatial location. Linearization techniques only accurately depict behavior in the immediate microscopic vicinity of each separate equilibrium point.
Mastering phase portraits transforms your ability to analyze complex systems in physics, engineering, and mathematics. Start practicing your sketching techniques with linear systems today to build the intuition needed for advanced nonlinear dynamics.
