Surface Area of Revolution — Surface Area of Solids of Revolution

The problem of measuring curved surfaces has fascinated mathematicians since antiquity. Archimedes was the first to rigorously compute the surface area of a sphere, showing it equals four times the area of its great circle — a result he considered his greatest achievement and had engraved on his tombstone. His method of exhaustion, which approximated curved surfaces with inscribed and circumscribed polyhedra, prefigured the integral calculus by nearly two millennia. The deeper question that Archimedes' work raised — how to systematically compute the area of any surface generated by revolving a curve — would require the development of infinitesimal calculus in the seventeenth century.

The central question this lesson addresses is deceptively simple: given a smooth curve y = f(x) on an interval [a, b], what is the exact surface area of the three-dimensional solid formed when this curve is revolved about an axis? The answer synthesizes two ideas you have already encountered — arc length and integration of infinitesimal elements — into a single powerful formula.

Before diving into the integral formula, it is essential to understand the geometric and analytic ingredients that underpin the surface-area-of-revolution computation. The entire derivation rests on decomposing a curved surface into infinitesimally thin frustum bands — the lateral surface of a truncated cone — and summing their areas via integration. Each frustum band has a slant height equal to a differential arc-length element ds and a circumference determined by the distance from the curve to the axis of revolution.

The following diagram illustrates how a curve y = f(x) rotated about the x-axis generates a surface of revolution, and how a small element of arc length ds produces a frustum band whose lateral surface area is the building block of the integral.

In the diagram, observe that each point on the curve traces out a circle of radius f(x) as it rotates about the x-axis. The key geometric insight is that the surface is not made up of flat rectangles (as with Riemann sums for area) but rather of conical frustum bands — thin strips whose slant height is the arc-length element ds rather than dx. Ignoring the slant (using dx instead of ds) would systematically undercount the surface area, just as using horizontal segments instead of the actual curve would undercount arc length.

We now derive the surface area integral rigorously. Consider a smooth curve y = f(x) ≥ 0 on [a, b] revolved about the x-axis. Partition [a, b] into n subintervals and approximate the curve on each subinterval by a line segment. When this line segment is revolved about the x-axis, it generates a conical frustum. The lateral surface area of a frustum with radii r₁ and r₂ and slant height ℓ is π(r₁ + r₂)ℓ. Taking the limit as n → ∞, the average radius (r₁ + r₂)/2 converges to f(x) at each point, and the slant height ℓ converges to the arc-length element ds. This yields the integral formula.

To solidify the derivation, let us examine the frustum approximation in detail and verify that the surface-area formula recovers known results. The lateral surface area of a right circular cone with base radius R and slant height L is πRL. A conical frustum with radii r₁ (top) and r₂ (bottom) and slant height ℓ has lateral area π(r₁ + r₂)ℓ. When r₁ and r₂ are nearly equal — as they are for an infinitesimal strip — this becomes 2πr̄ ℓ where r̄ is the average radius. The Mean Value Theorem guarantees that for a continuous function, this average equals f(x*) for some x* in the subinterval, and in the limit the sum of frustum areas converges to the Riemann integral.

As a verification, consider the sphere of radius R, which is generated by revolving the semicircle y = √(R² − x²) about the x-axis on [−R, R]. Computing f′(x) = −x/√(R² − x²), we get 1 + [f′(x)]² = R²/(R² − x²). Then √(1 + [f′(x)]²) = R/√(R² − x²), and f(x)√(1 + [f′(x)]²) = R. The integral becomes S = 2π∫₋ᴿᴿ R dx = 2πR × 2R = 4πR², confirming Archimedes' result. This elegant simplification — the integrand becoming a constant — is specific to the sphere and does not occur for general curves.

Let us compute the surface area generated by revolving the curve y = √x on the interval [0, 4] about the x-axis. This is a standard example that illustrates the full computational pipeline.

Students often conflate the surface-area and volume-of-revolution formulas. While both involve integrating 2π times a radius function, the structural difference is significant: volume integrals sum up cross-sectional areas (using disks or shells), whereas surface-area integrals sum up circumferential strips weighted by slant height. The table below highlights the key distinctions.

The surface-area-of-revolution formula is a special case of the more general theory of surface integrals encountered in multivariable calculus and differential geometry. In Calculus 3, you will compute the area of surfaces defined parametrically by r(u, v) = ⟨x(u,v), y(u,v), z(u,v)⟩ using the cross-product magnitude ‖rᵤ × rᵥ‖ dA. The revolution formula emerges as the case where one parameter is the angle of rotation θ and the other parametrizes the generating curve.

Another important extension involves Gabriel's Horn (also called Torricelli's trumpet), the solid formed by revolving y = 1/x on [1, ∞) about the x-axis. The volume integral π∫₁∞ 1/x² dx converges to π, but the surface-area integral 2π∫₁∞ (1/x)√(1 + 1/x⁴) dx diverges. This yields the famous paradox: a solid with finite volume but infinite surface area — you can fill it with paint, but you cannot paint its surface. This paradox is a vivid illustration of how volume and surface area integrals can have qualitatively different convergence behavior, and it foreshadows the study of improper integrals and the comparison test.