The length of a curve y = f(x) from x = a to x = b is L = integral from a to b of sqrt(1 + (f'(x))^2) dx. This formula comes from summing infinitesimal hypotenuses (sqrt(dx^2 + dy^2)) along the curve. Arc length integrals are often difficult or impossible to evaluate in closed form, making them good candidates for numerical methods.
Derive the formula from the Pythagorean theorem applied to infinitesimal segments. Compute arc length for functions where the integral simplifies nicely (e.g., y = x^(3/2), y = (x^2)/2 - ln(x)/4). Emphasize that most arc length integrals do not have neat answers.
You already know how to use the Fundamental Theorem of Calculus to compute accumulations, and you know u-substitution for handling composite integrands. Arc length applies integration to a new question: instead of asking "how much area is under this curve?" you ask "how long is this curve?" The setup uses a trick from a prerequisite you may not expect — the Pythagorean theorem.
Picture the curve y = f(x) from x = a to x = b cut into thousands of tiny segments. Each segment has a horizontal run of dx and a vertical rise of dy = f′(x) dx. The straight-line length of that tiny segment, by the Pythagorean theorem, is √(dx² + dy²). Factor out dx: √(dx² + [f′(x) dx]²) = √(1 + [f′(x)]²) dx. Now integrate — sum these infinitesimal hypotenuses — to get the total arc length: L = ∫ₐᵇ √(1 + [f′(x)]²) dx. The "1 +" inside the square root accounts for the horizontal component that is always present, even along a nearly flat curve.
The formula has a clean three-step setup: differentiate f(x) to get f′(x), square it, add 1, take the square root, and integrate. The challenge is in that last step. Most of the time, √(1 + [f′(x)]²) does not have a nice antiderivative. For polynomials like y = x^(3/2), the derivative f′(x) = (3/2)x^(1/2), so [f′(x)]² = (9/4)x, and 1 + (9/4)x has an elementary antiderivative. These "nice" examples are engineered specifically to work out. For a curve like y = sin(x) or y = x³, the arc length integral has no elementary closed form.
This is an important conceptual checkpoint: arc length teaches you that not every naturally-arising integral can be computed symbolically. The setup and formula are always the same; the evaluation may require numerical integration. When you move to parametric curves (arc length parametric) and surfaces of revolution, the same Pythagorean-theorem derivation extends in a natural way. Understanding the derivation — not just memorizing the formula — is what lets you adapt it to those new settings.