Differential Equations: Second-Order Linear

Second-order linear ordinary differential equations (ODEs) sit at the core of applied mathematics because they model systems where acceleration, curvature, or a second rate of change matters. They appear naturally in mechanical vibrations (mass-spring-damper), electrical circuits (RLC networks), and control systems (closed-loop dynamics). The practical goal is straightforward: given an equation involving $y$, $y'$, and $y''$, determine the function $y(x)$ that satisfies it, and interpret what that solution says about the system’s behavior.

A general second-order linear ODE has the form $$a(x)y'' + b(x)y' + c(x)y = g(x),$$ where $a(x)\neq 0$. When $g(x)=0$, the equation is homogeneous; otherwise it is non-homogeneous (or forced).

What “linear” really buys you

Linearity means $y$ and its derivatives appear only to the first power and are not multiplied together. This property gives two powerful consequences:

1. Superposition (homogeneous case): if $y_1$ and $y_2$ solve the homogeneous equation, then any linear combination $C_1y_1 + C_2y_2$ is also a solution.
2. Structured solution form (non-homogeneous case): every solution can be written as

$$y(x) = y_h(x) + y_p(x),$$ where $y_h$ solves the homogeneous equation and $y_p$ is any particular solution of the full equation.

In applications, $y_h$ typically describes the system’s natural response, while $y_p$ captures the steady-state response to an input or forcing term.

Homogeneous equations with constant coefficients

The most widely used class is $$ay'' + by' + cy = 0,$$ with constants $a,b,c$. The standard method is to try an exponential solution $y=e^{rx}$, which leads to the characteristic equation $$ar^2 + br + c = 0.$$ The roots of this quadratic determine the structure of $y_h$.

Case 1: two distinct real roots

If $r_1 \neq r_2$ are real, then $$y_h = C_1 e^{r_1 x} + C_2 e^{r_2 x}.$$ Interpretation in dynamics is immediate: negative roots imply decay, positive roots imply growth (instability), and differing magnitudes determine fast and slow modes.

Case 2: a repeated real root

If the characteristic equation has a repeated root $r$, then a second independent solution is $xe^{rx}$: $$y_h = (C_1 + C_2 x)e^{rx}.$$ Repeated roots often indicate a “critically damped” boundary in mechanical or electrical systems, where the system returns to equilibrium as quickly as possible without oscillating.

Case 3: complex conjugate roots

If the roots are $r=\alpha \pm i\beta$ with $\beta \neq 0$, then the homogeneous solution becomes $$y_h = e^{\alpha x}\left(C_1\cos(\beta x) + C_2\sin(\beta x)\right).$$ Here, $\beta$ sets the oscillation frequency and $\alpha$ sets exponential growth or decay. In vibration and circuit problems, $\alpha<0$ corresponds to damping.

Non-homogeneous equations: how forcing enters

For the non-homogeneous equation $$a(x)y'' + b(x)y' + c(x)y = g(x),$$ the general solution is $y=y_h+y_p$. The work is in finding a suitable $y_p$. Two broad strategies are common:

• When coefficients are constant and $g(x)$ has a standard form (polynomials, exponentials, sines/cosines), a trial function approach is often efficient.
• For variable coefficients or complicated forcing, variation of parameters is a systematic method that works under broad conditions.

This article focuses on variation of parameters because it is explicitly general and highlights the structure of second-order linear theory.

Variation of parameters (method outline)

Consider the standard form obtained by dividing through by $a(x)$: $$y'' + p(x)y' + q(x)y = r(x),$$ where $p=b/a$, $q=c/a$, and $r=g/a$.

Assume you already know two linearly independent solutions $y_1(x)$ and $y_2(x)$ of the associated homogeneous equation $$y'' + p(x)y' + q(x)y = 0.$$ Then $$y_h = C_1 y_1 + C_2 y_2.$$ Variation of parameters seeks a particular solution in the form $$y_p = u_1(x)y_1(x) + u_2(x)y_2(x),$$ where $u_1$ and $u_2$ are functions to be determined (instead of constants).

To make the algebra manageable, impose an auxiliary condition: $$u_1' y_1 + u_2' y_2 = 0.$$ Differentiating $y_p$ and substituting into the ODE yields a second equation: $$u_1' y_1' + u_2' y_2' = r(x).$$ These two equations form a linear system for $u_1'$ and $u_2'$. Its determinant is the Wronskian $$W(x) = y_1y_2' - y_2y_1'.$$ Assuming $W(x)\neq 0$, the derivatives are $$u_1' = -\frac{y_2\,r(x)}{W(x)}, \quad u_2' = \frac{y_1\,r(x)}{W(x)}.$$ Integrating gives $u_1$ and $u_2$, and thus $y_p$.

Why the Wronskian matters

The Wronskian is a quick test of linear independence: if $W(x)\neq 0$ on an interval, $y_1$ and $y_2$ form a valid fundamental set there. In physical terms, linear independence means the homogeneous system genuinely has two distinct modes; without it, you cannot span all natural behaviors with two solutions.

Practical tips for using variation of parameters

• Keep the ODE in standard form $y'' + p(x)y' + q(x)y = r(x)$ to apply formulas cleanly.
• Compute $W(x)$ carefully; small sign errors propagate.
• Choose antiderivatives strategically. If $r(x)$ is complicated, integrals may not have elementary closed forms, but the method still yields a correct integral representation.

Connecting the math to vibrations, circuits, and control

Second-order linear models often come from balance laws.

Mechanical vibrations (mass-spring-damper)

A common model is $$m y'' + c y' + k y = F(t),$$ where $y$ is displacement, $m$ mass, $c$ damping, $k$ stiffness, and $F(t)$ an external force. The homogeneous part describes how the system responds when the force is removed. The characteristic equation $mr^2 + cr + k=0$ classifies behavior into overdamped (two real roots), critically damped (repeated root), and underdamped (complex roots). A periodic $F(t)$ produces a particular solution representing steady oscillation driven by the input.

Electrical circuits (RLC)

For a series RLC circuit, a second-order ODE appears in the charge or current depending on formulation. The roles of $m,c,k$ are played by inductance, resistance, and capacitance. The same root structure determines whether the transient decays smoothly or rings.

Control systems (closed-loop dynamics)

In control, second-order linear ODEs model the time response of many systems and controllers. The roots of the characteristic polynomial are the system poles; their real parts determine stability and decay rate, and their imaginary parts determine oscillation frequency. Even when the forcing term is a step or ramp input, the decomposition $y=y_h+y_p$ matches the standard transient-plus-steady-state narrative used in design.

Choosing the right method in practice

Second-order linear problems rarely present a single “best” method. A good workflow is:

1. Identify homogeneous vs non-homogeneous: if $g(x)=0$, focus on the characteristic equation (constant coefficients) or known solution techniques (variable coefficients).
2. Solve the homogeneous equation first: you need $y_h$ regardless.
3. Select a particular-solution method:

• If coefficients are constant and $g(x)$ is simple, a targeted trial approach is often fastest.
• If coefficients vary with $x$ or $g(x)$ is awkward, use variation of parameters for a reliable, general construction.

Second-order linear ODEs are popular in engineering not because they are simple, but because they are tractable. Their solution structure exposes stability, resonance, damping, and steady-state behavior in a way that aligns closely with the real systems they model.