01 / 21

Systems:
Complex Eigenvalues

Discover how complex eigenvalues produce spirals and centers in linear systems of differential equations.

Differential Equations

What Are Complex Eigenvalues?

When we solve a \(2 \times 2\) linear system \(\mathbf{x}' = A\mathbf{x}\), we find the eigenvalues of \(A\) by solving the characteristic equation:

These complex eigenvalues mean the system's solutions involve oscillation: sines and cosines, not just exponentials.

Why Do Complex Eigenvalues Matter?

Oscillation
The imaginary part \(\beta\) controls the frequency of rotation in the phase portrait.
Growth or Decay
The real part \(\alpha\) controls whether trajectories spiral inward, spiral outward, or form a closed orbit.
Physical Models
Damped oscillators, coupled circuits, and predator-prey models all exhibit complex eigenvalue behavior.

Three Types of Phase Portraits

Spiral Sink \(\alpha < 0\) Center \(\alpha = 0\) Spiral Source \(\alpha > 0\)

The General Solution Theorem

If \(A\) has complex eigenvalues \(\lambda = \alpha \pm \beta i\) with eigenvector \(\mathbf{v} = \mathbf{a} + \mathbf{b}i\), the general real solution is:

Here, \(\mathbf{a} = \operatorname{Re}(\mathbf{v})\) and \(\mathbf{b} = \operatorname{Im}(\mathbf{v})\) are the real and imaginary parts of the eigenvector. The factor \(e^{\alpha t}\) controls growth/decay, while \(\cos\beta t\) and \(\sin\beta t\) produce rotation.

Building Two Real Solutions

From the single complex eigenpair \(\lambda, \mathbf{v}\), we extract two linearly independent real solutions:

1
Compute one complex solution: \(\mathbf{x}_c(t) = e^{(\alpha + \beta i)t}\mathbf{v}\).
2
Expand using Euler's formula: \(e^{i\beta t} = \cos\beta t + i\sin\beta t\).
3
Separate real and imaginary parts: \(\mathbf{x}_1(t) = \operatorname{Re}(\mathbf{x}_c)\), \(\mathbf{x}_2(t) = \operatorname{Im}(\mathbf{x}_c)\).

Spirals vs. Centers

Spiral Center
Real part \(\alpha\) \(\alpha \neq 0\) \(\alpha = 0\)
Eigenvalues \(\alpha \pm \beta i\) \(\pm \beta i\) (purely imaginary)
Trajectories Spiral in or out Closed ellipses
Stability Asymptotically stable (\(\alpha < 0\)) or unstable (\(\alpha > 0\)) Stable but not asymptotically stable
\(e^{\alpha t}\) factor Causes amplitude to grow or shrink Absent (amplitude constant)

Step-by-Step Method

1
Find eigenvalues: Solve \(\det(A - \lambda I) = 0\). Confirm the discriminant is negative.
2
Pick one eigenvalue: Use \(\lambda = \alpha + \beta i\) (either conjugate works).
3
Find the eigenvector: Solve \((A - \lambda I)\mathbf{v} = \mathbf{0}\) with complex arithmetic.
4
Separate \(\mathbf{v} = \mathbf{a} + \mathbf{b}i\): Identify real part \(\mathbf{a}\) and imaginary part \(\mathbf{b}\).
5
Write the general solution: \(\mathbf{x}(t) = c_1\mathbf{x}_1(t) + c_2\mathbf{x}_2(t)\), where \(\mathbf{x}_1 = e^{\alpha t}(\mathbf{a}\cos\beta t - \mathbf{b}\sin\beta t)\) and \(\mathbf{x}_2 = e^{\alpha t}(\mathbf{a}\sin\beta t + \mathbf{b}\cos\beta t)\).

Example 1: Finding Eigenvalues

Solve \(\mathbf{x}' = A\mathbf{x}\) where \(A = \begin{pmatrix} 1 & -2 \\ 1 & 1 \end{pmatrix}\).

Example 1: Eigenvector and Solution

Example 2: A Pure Center

Solve \(\mathbf{x}' = A\mathbf{x}\) where \(A = \begin{pmatrix} 0 & -3 \\ 3 & 0 \end{pmatrix}\).

Example 3: Spiral Sink

Solve \(\mathbf{x}' = A\mathbf{x}\) where \(A = \begin{pmatrix} -1 & 1 \\ -4 & -1 \end{pmatrix}\).

Visualizing the Spiral Sink

\(x_1\) \(x_2\) origin

The Roles of \(\alpha\) and \(\beta\)

\(\alpha\) : Real Part
Controls the envelope \(e^{\alpha t}\).
\(\alpha < 0\): decay (spiral sink).
\(\alpha = 0\): constant (center).
\(\alpha > 0\): growth (spiral source).
\(\beta\) : Imaginary Part
Controls the angular frequency of rotation.
Larger \(|\beta|\): faster spiraling.
Period of one loop: \(T = \frac{2\pi}{\beta}\).

Common Mistakes to Avoid

✕ Solving for Both Conjugates
You only need one eigenvalue \(\alpha + \beta i\). The conjugate gives the same real solutions.
✕ Forgetting \(e^{\alpha t}\)
When \(\alpha \neq 0\), the solution is not just sines and cosines: the \(e^{\alpha t}\) factor is essential.
✕ Swapping \(\mathbf{a}\) and \(\mathbf{b}\)
Mixing up real and imaginary parts of the eigenvector leads to incorrect solutions. Label carefully.

Summary of Eigenvalue Types

Eigenvalue Type Phase Portrait Stability
Real, distinct, same sign Node (sink or source) Stable or unstable
Real, distinct, opposite signs Saddle point Unstable
Complex, \(\alpha < 0\) Spiral sink Asymptotically stable
Complex, \(\alpha = 0\) Center Stable (not asymptotic)
Complex, \(\alpha > 0\) Spiral source Unstable
Repeated (defective) Degenerate node / spiral Depends on sign

Question 1 of 5

The matrix \(A = \begin{pmatrix} 0 & -5 \\ 5 & 0 \end{pmatrix}\) has eigenvalues \(\lambda = \pm 5i\). What type of equilibrium is the origin?

Question 2 of 5

If a \(2 \times 2\) system has eigenvalues \(\lambda = -3 \pm 4i\), what does the factor \(e^{-3t}\) contribute to the solution?

Question 3 of 5

Given eigenvalue \(\lambda = 2 + 3i\) with eigenvector \(\mathbf{v} = \begin{pmatrix} 1 \\ i \end{pmatrix}\), what are \(\mathbf{a}\) and \(\mathbf{b}\)?

Question 4 of 5

A system has a spiral source at the origin. Which of the following must be true about its eigenvalues?

Question 5 of 5

For eigenvalues \(\lambda = \pm 7i\), the period of one complete orbit around the origin is: