Why do some trajectories stop at the edge of the window?

Integration halts when the state leaves a slightly expanded box around the visible axes. That keeps the numeric solution from running forever on unbounded orbits (for example along unstable manifolds of a saddle) and keeps the display responsive.

Why does the damped preset look like a center when ζ = 0?

With ζ = 0 the equations reduce to the same undamped harmonic form as the harmonic preset (up to the shared ω scaling). Any tiny ζ > 0 breaks the closed loops and produces a spiral in toward the origin for this linear model.

What do the yellow dots mean?

They mark equilibrium points where f = g = 0 at the resolution of the model. The origin is always shown for the linear presets; the pendulum preset also marks ±π on the angle axis where sin θ = 0.

Is this a substitute for a full ODE solver or bifurcation tool?

No. It is a visual sketchpad: fixed viewing window, fixed step RK4, no error control, no phase or eigenvalue readouts beyond what you infer from the picture. Use a dedicated solver for publication-quality plots or parameter scans.

2D Phase Portrait (ODE)

A phase portrait for a planar autonomous ordinary differential equation plots the state (x, y) and shows how it evolves under x′ = f(x, y) and y′ = g(x, y) without displaying time explicitly. This simulator draws a uniform grid of short arrows whose directions match the vector field (f, g), normalized for readability so you can read the flow at a glance. Clicking inside the plot seeds a fourth-order Runge–Kutta integrator that marches both forward and backward in time until the trajectory leaves an expanded viewport, building a single curve through the clicked point. Presets include the undamped harmonic oscillator (elliptic orbits around a center), a damped linear oscillator whose damping ratio ζ controls decay to a stable spiral, a hyperbolic saddle with separatrices, stable and unstable nodes, stable and unstable spirals from fixed 2×2 linearizations, and the ideal pendulum written as θ′ = ω, ω′ = −sin θ so x is angle and y is angular velocity. Equilibrium locations are marked with small dots. The tool is intentionally lightweight: no bifurcation scanning, no nullclines, and no detection of limit cycles—those are natural follow-ons after students grasp orbits in the plane.

Who it's for: Undergraduate ODE or dynamical systems students learning equilibria, linearization, and qualitative behavior before quantitative time-series plots.

Key terms

Phase plane
Autonomous ODE
Vector field
Trajectory
Equilibrium
Saddle
Spiral
Center
Runge–Kutta

How it works

The phase plane plots (x, y) for a planar autonomous system x′ = f(x, y), y′ = g(x, y). A direction field shows the tangent to trajectories; clicking seeds a high-order Runge–Kutta integration forward and backward in time. Presets include the harmonic oscillator (center), damped oscillator (stable spiral when ζ > 0), a saddle, stable and unstable nodes, stable and unstable spirals from a fixed linearization, and the undamped pendulum angle–angular velocity system.

Key equations

x′ = f(x, y), y′ = g(x, y)

Harmonic: x′ = y, y′ = −ω²x

Pendulum: θ′ = ω, ω′ = −sin θ

Frequently asked questions

Why do some trajectories stop at the edge of the window?: Integration halts when the state leaves a slightly expanded box around the visible axes. That keeps the numeric solution from running forever on unbounded orbits (for example along unstable manifolds of a saddle) and keeps the display responsive.
Why does the damped preset look like a center when ζ = 0?: With ζ = 0 the equations reduce to the same undamped harmonic form as the harmonic preset (up to the shared ω scaling). Any tiny ζ > 0 breaks the closed loops and produces a spiral in toward the origin for this linear model.
What do the yellow dots mean?: They mark equilibrium points where f = g = 0 at the resolution of the model. The origin is always shown for the linear presets; the pendulum preset also marks ±π on the angle axis where sin θ = 0.
Is this a substitute for a full ODE solver or bifurcation tool?: No. It is a visual sketchpad: fixed viewing window, fixed step RK4, no error control, no phase or eigenvalue readouts beyond what you infer from the picture. Use a dedicated solver for publication-quality plots or parameter scans.

Other simulators in this category — or see all 61.

View category →

School

Lissajous Curves

Beautiful patterns from two frequencies with adjustable ratio.

Launch Simulator

NewKids

Harmonograph

Two damped harmonic sums in x and y: decaying rosette trace vs Lissajous loops.

Launch Simulator

NewKids

Spirograph (Trochoids)

Hypo- or epitrochoid: fixed R, rolling r, pen d; hue trail and period hints.

Launch Simulator

NewKids

Sorting Algorithms (parallel)

Bubble, insertion, merge, quicksort, and heapsort on the same shuffled permutation — five bar rows advance in lockstep so you can compare how each method moves values.

Launch Simulator

NewSchool

SIR Epidemic Model

S + I + R = 1: βSI and γI; ℛ₀ ≈ β/γ, herd threshold 1 − 1/ℛ₀; RK4 time plot.

Launch Simulator

NewSchool

Sandpile (SOC)

BTW abelian model: add grains, ≥4 topples to neighbors; critical avalanches.

Launch Simulator

2D Phase Portrait (ODE)

How it works

Key equations

Frequently asked questions

More from Math Visualization

Lissajous Curves

Harmonograph

Spirograph (Trochoids)

Sorting Algorithms (parallel)

SIR Epidemic Model

Sandpile (SOC)

2D Phase Portrait (ODE)

How it works

Key equations

Frequently asked questions