ODE: Homogeneous Linear Systems with Complex Eigenvalues

Understanding homogeneous linear systems with complex eigenvalues is crucial in engineering for modeling dynamic systems that exhibit oscillation, such as mechanical vibrations, electrical circuits, and control systems. These systems often involve coupled differential equations where complex eigenvalues naturally arise, leading to solutions that describe sinusoidal motions. Mastering this topic enables you to predict system behavior, design stable controllers, and analyze real-world oscillatory phenomena.

Complex Eigenvalue Pairs and Eigenvectors

Consider a homogeneous linear system of ordinary differential equations given by $\dot{\mathbf{x}}=A\mathbf{x}$, where $\mathbf{x}$ is a vector of state variables and $A$ is a constant $n\times n$ coefficient matrix. When $A$ has real entries, its eigenvalues are the roots of the characteristic polynomial $\det (A-\lambda I)=0$. In many physical systems, these roots are complex numbers, and for real matrices, they always occur in complex conjugate pairs, meaning if $\lambda =\alpha +i\beta$ is an eigenvalue, then $\bar{\lambda }=\alpha -i\beta$ is also an eigenvalue, where $\alpha$ and $\beta$ are real numbers and $i$ is the imaginary unit.

For each complex eigenvalue, there is a corresponding complex eigenvector. To find an eigenvector $\mathbf{v}$ for $\lambda =\alpha +i\beta$, you solve $(A-\lambda I)\mathbf{v}=0$. Since the entries are complex, the eigenvector will also have complex entries. Crucially, the eigenvector for the conjugate eigenvalue $\bar{\lambda }$ is the conjugate of $\mathbf{v}$, denoted $\bar{\mathbf{v}}$. This pairing is fundamental because the general solution to the system will be built from these complex foundations, but we ultimately seek real-valued solutions that describe physical states.

For example, in a $2\times 2$ system, a matrix like $A=\left[\begin{array}{cc}a& -b\\ b& a\end{array}\right]$ has eigenvalues $\lambda =a\pm ib$. The eigenvector for $\lambda =a+ib$ can be found as $\mathbf{v}=\left[\begin{array}{c}1\\ -i\end{array}\right]$ or similar, highlighting the interplay between real and imaginary components.

Extracting Real-Valued Solutions Using Euler's Formula

The general solution for a system with complex eigenvalues involves complex exponential functions. For a pair of complex conjugate eigenvalues $\lambda =\alpha \pm i\beta$ with eigenvectors $\mathbf{v}$ and $\bar{\mathbf{v}}$, the complex-valued general solution is:

$$\mathbf{x}(t)=c_1{e}^{(\alpha +i\beta )t}\mathbf{v}+c_2{e}^{(\alpha -i\beta )t}\bar{\mathbf{v}},$$

where $c_1$ and $c_2$ are complex constants. However, since $\mathbf{x}(t)$ must be real for physical systems, we use Euler's formula, which states $e^{i\theta }=\cos \theta +i\sin \theta$, to extract real-valued solutions. By choosing constants such that $c_2={\bar{c}}_1$ to ensure reality, and after algebraic manipulation, the real general solution can be expressed as:

$$\mathbf{x}(t)=e^{\alpha t}\left({\mathbf{C}}_1\cos (\beta t)+{\mathbf{C}}_2\sin (\beta t)\right),$$

where ${\mathbf{C}}_1$ and ${\mathbf{C}}_2$ are real vector constants determined from initial conditions. This form reveals the oscillatory nature of the solution, modulated by the exponential term $e^{\alpha t}$.

Let's walk through a step-by-step example. Solve the system:

$$\dot{x}=-y,\dot{y}=x.$$

In matrix form, $A=\left[\begin{array}{cc}0& -1\\ 1& 0\end{array}\right]$. The characteristic equation is $\det (A-\lambda I)={\lambda }^2+1=0$, so eigenvalues are $\lambda =\pm i$ (here $\alpha =0$, $\beta =1$). For $\lambda =i$, solve $(A-iI)\mathbf{v}=0$:

$$\left[\begin{array}{cc}-i& -1\\ 1& -i\end{array}\right]\left[\begin{array}{c}{v}_1\\ v_2\end{array}\right]=\left[\begin{array}{c}0\\ 0\end{array}\right].$$

This yields $v_1=1$, $v_2=-i$, so $\mathbf{v}=\left[\begin{array}{c}1\\ -i\end{array}\right]$. The complex solution is $\mathbf{x}(t)=c_1{e}^{it}\left[\begin{array}{c}1\\ -i\end{array}\right]+c_2{e}^{-it}\left[\begin{array}{c}1\\ i\end{array}\right]$. Using Euler's formula, $e^{it}=\cos t+i\sin t$, and setting $c_1=\frac{1}{2}(C_1-i{C}_2)$ for real $C_1,C_2$, we derive the real general solution:

$$\mathbf{x}(t)=\left[\begin{array}{c}x(t)\\ y(t)\end{array}\right]=\left[\begin{array}{c}{C}_1\cos t+C_2\sin t\\ -C_1\sin t+C_2\cos t\end{array}\right].$$

This represents pure oscillations, as expected from eigenvalues with zero real part.

Phase Portraits: Spirals and Centers

The phase portrait of a linear system visually represents solution trajectories in the state space, such as the $xy$-plane for a 2D system. When eigenvalues are complex, the phase portrait depends critically on the real part $\alpha$. If $\alpha =0$, the solutions are purely sinusoidal, leading to closed elliptical orbits called centers. These systems are marginally stable, with trajectories circling equilibrium without decaying or growing.

If $\alpha \ne 0$, the exponential term $e^{\alpha t}$ causes amplitude change over time. When $\alpha <0$, the oscillations decay, and trajectories spiral inward toward the equilibrium, forming a stable spiral. Conversely, if $\alpha >0$, oscillations grow unbounded, resulting in an unstable spiral. The direction of rotation (clockwise or counterclockwise) is determined by the sign of the imaginary part $\beta$ or from the matrix entries.

For instance, in the previous example with $\lambda =\pm i$, the phase portrait is a center with circles centered at the origin. If we modify the system to $\dot{x}=-y$, $\dot{y}=x-y$, eigenvalues become $\lambda =-\frac{1}{2}\pm i\frac{\sqrt{3}}{2}$, so $\alpha =-\frac{1}{2}<0$, leading to a stable spiral where trajectories coil inward. Sketching these involves plotting a few key trajectories: from initial points like $(1,0)$, you'll see oscillatory motion that decays, forming a inward spiral pattern.

Physical Interpretation in Oscillatory Coupled Systems

In engineering, these mathematical concepts directly model oscillatory coupled system behavior. Consider a mass-spring system with damping or an RLC electrical circuit. Such systems are often described by second-order ODEs that can be rewritten as first-order linear systems. For example, a damped harmonic oscillator $m\ddot{x}+c\dot{x}+kx=0$ transforms into a system with state variables $x$ and $\dot{x}$, yielding a matrix with complex eigenvalues when damping is light ($c^2<4mk$).

The complex eigenvalue pair $\lambda =\alpha \pm i\beta$ encodes physical parameters: $\alpha$ (real part) relates to damping or growth rate, while $\beta$ (imaginary part) gives the oscillatory frequency. In the mass-spring case, $\alpha =-c/(2m)$ and $\beta =\sqrt{4mk-c^2}/(2m)$. The real-valued solutions extracted via Euler's formula describe the actual displacement and velocity over time, showing sinusoidal oscillations with exponential decay.

Coupled systems, like two masses connected by springs, can exhibit modes where complex eigenvalues represent superimposed oscillations. The phase portraits help visualize energy exchange between components: centers indicate conservative systems with constant energy, while spirals show energy dissipation or injection. This interpretation is vital for designing stable structures, tuning controllers in robotics, or analyzing signal processing in circuits.

Common Pitfalls

1. Ignoring the Conjugate Pair in Solutions: When writing the general solution, some learners omit the term with the conjugate eigenvalue, leading to incomplete solutions. Remember that for real matrices, complex eigenvalues always come in pairs, and both must be included to form a basis for the solution space. Correction: Always use $\mathbf{x}(t)=c_1{e}^{\lambda t}\mathbf{v}+c_2{e}^{\bar{\lambda }t}\bar{\mathbf{v}}$ before extracting real parts.

1. Failing to Extract Real-Valued Solutions Properly: Applying Euler's formula incorrectly can result in complex-valued outputs. A common error is forgetting to combine terms to eliminate imaginary units. Correction: Systematically separate real and imaginary parts after substituting $e^{i\beta t}=\cos (\beta t)+i\sin (\beta t)$, and choose constants to ensure the solution is real.

1. Misclassifying Phase Portraits: Confusing centers with spirals when the real part $\alpha$ is zero. For example, if $\alpha =0$, trajectories are closed loops (centers), not spirals. Misinterpreting this can lead to incorrect stability analysis. Correction: Always compute $\alpha$ exactly; if $\alpha =0$, it's a center; if $\alpha \ne 0$, it's a spiral.

1. Overlooking Physical Context in Applications: In engineering problems, neglecting to link eigenvalues to physical parameters like damping ratio or natural frequency can make solutions abstract. Correction: Explicitly relate $\alpha$ and $\beta$ to system constants, and interpret solutions in terms of measurable quantities like amplitude and phase shift.

Summary

• Complex eigenvalues for real matrices always occur in conjugate pairs $\lambda =\alpha \pm i\beta$, where $\alpha$ determines growth/decay and $\beta$ gives oscillation frequency.
• Complex eigenvectors are found by solving $(A-\lambda I)\mathbf{v}=0$, and their conjugates provide the basis for solutions.
• Euler's formula $e^{i\theta }=\cos \theta +i\sin \theta$ is essential for extracting real-valued solutions from complex exponentials, yielding terms like $e^{\alpha t}({\mathbf{C}}_1\cos (\beta t)+{\mathbf{C}}_2\sin (\beta t))$.
• Phase portraits classify behavior: centers for $\alpha =0$ (closed orbits) and spirals for $\alpha \ne 0$ (inward for $\alpha <0$, outward for $\alpha >0$).
• Physical interpretations in coupled systems, such as mechanical vibrations or circuits, use these concepts to model oscillatory dynamics, stability, and energy exchange.