ODE: Fourier Series Fundamentals

Fourier series are one of the most powerful tools in applied mathematics and engineering, transforming how we analyze periodic phenomena. They allow you to decompose complex, repeating signals—like an electrical waveform, a sound vibration, or a mechanical oscillation—into a sum of simple sine and cosine functions. This frequency-based perspective is fundamental to fields ranging from signal processing and acoustics to heat transfer and quantum mechanics.

What is a Fourier Series?

A Fourier series is an infinite sum of sine and cosine terms used to represent a periodic function. If a function $f(t)$ is periodic with period $T$ (meaning $f(t+T)=f(t)$ for all $t$), then its Fourier series representation on the interval $[-T/2,T/2]$ is given by:

$$f(t)\sim \frac{a_0}{2}+\sum _{n=1}^{\infty }\left[a_n\cos \left(\frac{2\pi nt}{T}\right)+b_n\sin \left(\frac{2\pi nt}{T}\right)\right]$$

The symbol "$\sim$" indicates association; the series converges to the function under specific conditions we will explore. The term $\frac{a_0}{2}$ represents the average value of the function over one period. The sine and cosine terms oscillate at integer multiples (harmonics) of the fundamental frequency ${\omega }_0=2\pi /T$. The coefficients $a_n$ and $b_n$ are weights that tell you how much of each harmonic frequency is present in the original signal $f(t)$.

Computing Fourier Coefficients

The magic of the Fourier series lies in calculating the coefficients $a_n$ and $b_n$. They are determined by exploiting the orthogonality properties of sine and cosine functions over a complete period. For a function with period $T$, the formulas are:

$$a_0=\frac{2}{T}{\int }_{-T/2}^{T/2}f(t)dt$$

$$a_n=\frac{2}{T}{\int }_{-T/2}^{T/2}f(t)\cos \left(\frac{2\pi nt}{T}\right)dt,n\ge 1$$

$$b_n=\frac{2}{T}{\int }_{-T/2}^{T/2}f(t)\sin \left(\frac{2\pi nt}{T}\right)dt,n\ge 1$$

Step-by-Step Example: The Square Wave Consider an odd square wave with period $T=2\pi$ and defined on $[-\pi ,\pi ]$ as: $f(t)=-1$ for $-\pi <t<0$ and $f(t)=1$ for $0<t<\pi$.

1. Find $a_0$ and $a_n$: Because $f(t)$ is an odd function, and cosine is even, their product is odd. The integral of an odd function over a symmetric interval is zero. Therefore, $a_0=0$ and $a_n=0$ for all $n$.
2. Find $b_n$: We compute the sine coefficients.

$$b_n=\frac{2}{T}{\int }_{-T/2}^{T/2}f(t)\sin (nt)dt=\frac{1}{\pi }\left[{\int }_{-\pi }^0(-1)\sin (nt)dt+{\int }_0^{\pi }(1)\sin (nt)dt\right]$$ Solving the integrals yields: $$b_n=\frac{2}{n\pi }\left(1-\cos (n\pi )\right)=\frac{2}{n\pi }\left(1-(-1{)}^n\right)$$ This means $b_n=0$ for even $n$, and $b_n=\frac{4}{n\pi }$ for odd $n$.

Thus, the Fourier series for this square wave is: $$f(t)\sim \frac{4}{\pi }\left(\sin (t)+\frac{\sin (3t)}{3}+\frac{\sin (5t)}{5}+\cdots \right)$$

Convergence and the Dirichlet Conditions

A Fourier series does not converge to its function for every possible periodic signal. The Dirichlet conditions specify a set of sufficient (but not strictly necessary) criteria for convergence. If a periodic function $f(t)$ satisfies these on one period, its Fourier series converges:

1. Absolutely Integrable: The integral of $|f(t)|$ over one period is finite.
2. Finite Number of Extrema: The function has a finite number of maxima and minima in one period.
3. Finite Number of Discontinuities: The function has a finite number of finite (jump) discontinuities in one period.

When these conditions hold, we can state the convergence behavior precisely:

• At any point $t$ where $f$ is continuous, the Fourier series converges exactly to $f(t)$.
• At a point $t_0$ where $f$ has a jump discontinuity, the series converges to the average of the left-hand and right-hand limits:

$$\frac{f(t_0^{-})+f(t_0^{+})}{2}$$ This is a crucial result. For our square wave example at $t=0$, the function jumps from -1 to 1. The Fourier series will converge to $(-1+1)/2=0$, which you can observe as the midpoint of the jump.

Parseval's Theorem: Conservation of Power

Parseval's theorem provides a profound link between the time domain and the frequency domain, often interpreted as a conservation of energy or power. For a Fourier series, it states that the average power of the signal over one period equals the sum of the powers of its individual frequency components.

Mathematically, for a function satisfying Dirichlet conditions: $$\frac{1}{T}{\int }_{-T/2}^{T/2}[f(t){]}^{2}dt=\frac{a_0^2}{4}+\frac{1}{2}\sum _{n=1}^{\infty }(a_n^2+b_n^2)$$

The left side is the mean-square value of $f(t)$. The right side is the sum of the squared magnitudes of each Fourier coefficient (the DC term $a_0/2$ and the amplitudes of each harmonic). This theorem is invaluable for calculating total harmonic distortion in electrical engineering or for verifying the accuracy of a partial series approximation.

Physical Interpretation and Frequency Decomposition

The most powerful application of Fourier series is the physical interpretation as frequency decomposition of signals. Think of a complex musical chord played by an orchestra. Your ear hears a single, rich sound, but it is physically composed of a fundamental tone (the note's pitch) and many higher-frequency harmonics (overtones) from different instruments. A Fourier series performs this exact decomposition mathematically.

In engineering terms:

• The fundamental frequency ${\omega }_0=2\pi /T$ is the base rhythm of the signal.
• The harmonics $n{\omega }_0$ are integer multiples of this base frequency.
• The coefficients $a_n$ and $b_n$ (often combined into a single magnitude and phase) form the frequency spectrum of the signal. This spectrum tells you precisely "how much" of each frequency is present.

This allows engineers to filter signals (remove unwanted frequencies), compress data, solve differential equations (like the heat equation), and analyze system responses by examining how individual frequency components are affected.

Common Pitfalls

1. Incorrect Period and Integration Limits: A frequent error is using the wrong period $T$ or integrating over an interval that is not one period long. Always identify the fundamental period correctly and use the corresponding limits in the coefficient formulas, typically $[-T/2,T/2]$ or $[0,T]$.

1. Misapplying Symmetry: Assuming all coefficients are zero for an even or odd function without verifying the function's definition over the chosen symmetric interval. A function is only odd/even with respect to the origin if its definition on $[-L,L]$ matches that property. If the interval is shifted, the symmetry argument may not hold.

1. Confusing Convergence with Equality: The Fourier series equals the function only where it is continuous. At discontinuities, it converges to the average. Expecting it to match the function value exactly at a jump point leads to misunderstanding the Gibbs phenomenon, where partial sums overshoot near the discontinuity.

1. Neglecting the $a_0/2$ Form: Writing the series with $a_0$ instead of $a_0/2$ is a common algebraic mistake. The factor of $1/2$ in the constant term makes the formula for $a_0$ consistent with the general formula for $a_n$ when $n=0$.

Summary

• A Fourier series represents a periodic function as an infinite sum of sine and cosine harmonics, providing a frequency-domain view of the signal.
• The Fourier coefficients $a_n$ and $b_n$ are computed using integral formulas that project the function onto each harmonic basis function.
• Under the Dirichlet conditions, the series converges to the function at points of continuity and to the average of the left and right limits at jump discontinuities.
• Parseval's theorem states that the total power in the signal is the sum of the powers in each of its frequency components.
• The primary physical interpretation is that of spectral or frequency decomposition, which is foundational for signal analysis, filtering, and solving boundary value problems in engineering and physics.