Surface area of revolution: Visual walkthrough

Step 1 — What "spinning a curve" even means

WHAT. Take a curve sitting above a horizontal line (the axis). Grab that line like a skewer and spin the curve all the way around it, $36 0^{\circ}$ .

WHY start here. Before any formula, you must see the object we are measuring: not a solid blob, just its thin outer skin.

PICTURE. Below, the blue curve $y = f (x)$ lives above the black axis. Pick one point on it, height $y$ above the axis. As we spin, that single point sweeps a red circle. Every point sweeps its own circle — stack all those circles and you get the surface.

Step 2 — Chop the curve into tiny straight pieces

WHAT. Zoom into a microscopic stretch of the curve. Over a tiny horizontal step $d x$ , the curve climbs by a tiny vertical amount $d y$ . That short stretch is almost perfectly straight.

WHY. A curved thing is hard; a straight thing is easy. Curves are just infinitely many tiny straight pieces stitched together — that is the whole idea of calculus. We will find the surface of ONE tiny piece, then add them all up (integrate).

PICTURE. The orange segment is one tiny piece. Its horizontal shadow is $d x$ (gray, on the floor). Its true tilted length is longer — call it $d s$ . Notice: the piece leans, so $d s > d x$ .

Step 3 — Spin ONE tiny piece → a frustum band

WHAT. Take that one leaning orange segment and spin it $36 0^{\circ}$ . Because its two ends sit at slightly different heights, it does not make a flat ring — it makes a short slanted cone-band, called a frustum.

WHY a frustum and not a flat washer. A washer (disk with a hole) is flat and belongs to volume problems. Here we want surface, and the surface leans — so its band is tilted, with a small radius at one end and a slightly bigger radius at the other.

PICTURE. The orange segment sweeps the green band. Inner rim radius $r_{1}$ (lower end), outer rim radius $r_{2}$ (higher end). The slanted width of the band is exactly the $d s$ from Step 2.

See Frustum and cone geometry for where the next formula is born.

Step 4 — Unroll the frustum to read off its area

WHAT. Cut the band along one side and flatten it. It opens into a curved strip. Its area is:

$A_{band} = once around 2 π \cdot average radius \frac{r _{1} + r _{2}}{2} \cdot slant width ℓ$

WHY average radius. The band is fatter at the top than the bottom. Its "typical" distance from the axis is the average of the two rim radii, $\frac{r _{1} + r _{2}}{2}$ . Multiply the average circumference $2 π \cdot \frac{r _{1} + r _{2}}{2}$ by the slant width $ℓ$ — exactly like unrolling a cylinder into a rectangle: (distance around) $\times$ (height).

PICTURE. Left: the band. Right: it unrolled into a strip of length "average circumference" and height $ℓ$ .

Step 5 — Shrink the band: $r_{1} \approx r_{2} \approx y$ , and $ℓ \to d s$

WHAT. Make the piece infinitely thin. Then its two rim radii are so close together they both equal the curve's height $y$ at that spot. The average $\frac{r _{1} + r _{2}}{2}$ collapses to plain $y$ , and the slant width $ℓ$ is the arc element $d s$ .

WHY. In the limit, "average of two nearly-equal numbers" is just the number itself. This kills the clumsy $\frac{r _{1} + r _{2}}{2}$ and leaves something clean.

PICTURE. The fat band of Step 4 pinched down to a whisker-thin ring at height $y$ , slant $d s$ .

Step 6 — Turn $d s$ into something we can integrate in $x$

WHAT. We want to add up all rings from $x = a$ to $x = b$ , so everything must be written in $x$ . From the Step 2 triangle, factor a $d x$ out of the square root:

$d s = d x^{2} + d y^{2} = 1 + (\frac{d y}{d x})^{2} d x$

WHY. $\frac{d y}{d x}$ is the slope — how steep the curve is. A steep curve has big slope, so $1 + (slope)^{2}$ is big, so $d s$ is much longer than $d x$ . The picture and the algebra agree: steeper $\Rightarrow$ longer slanted edge. This is the same $d s$ you met in Arc length.

PICTURE. Two triangles side by side: a gentle slope (tiny stretch of $d s$ over $d x$ ) versus a steep slope (big stretch). The number $1 + (d y / d x)^{2}$ is that stretch factor.

Step 7 — Edge & degenerate cases (never get surprised)

WHAT / WHY. A formula you can't stress-test is a formula you don't trust. Here is what happens at the boundaries.

PICTURE. Four mini-panels, one per case, each with the ring it produces.

Definition Case-by-case

Flat curve, $\frac{d y}{d x} = 0$ (horizontal line at height $y$ ). Then $1 + 0 = 1$ , so $d s = d x$ : the band is a plain cylinder of area $2 π y (b - a)$ . No tilt, no stretch. Sanity ✅.
Point touches the axis, $y = 0$ . Then $2 π y = 0$ : that ring has zero radius, collapses to a point, adds nothing. (This is exactly the sphere's poles in Example 1 of the parent.)
Vertical-ish curve, slope $\to \infty$ . Then $1 + (slope)^{2}$ blows up: the tilted edge is enormous, so the surface is large. Only $d s$ captures this — $d x$ would lie. When the curve is truly vertical, switch to integrating in $y$ (radius becomes $x$ , use $1 + (d x / d y)^{2} d y$ ).
Rotate about the line $y = k$ instead of the axis. The radius is no longer $y$ but the distance to that line, $∣ y - k ∣$ . The stretch $d s$ is unchanged. Always ask: how far is the point from the axis?

The one-picture summary

Everything above, on one canvas: a curve, one tiny leaning piece, its swept ring, and the three factors $2 π$ , $y$ , $d s$ labelled right where they act.

$S = \int 2 π radius (dist to axis) arc length d s$

Compare with Pappus's theorem: $S = 2 π \overset{r}{ˉ} L$ — surface = (circumference of the centroid) $\times$ (arc length). Our integral is Pappus done one ring at a time.

Recall Feynman retelling (say it to a 12-year-old)

Picture a curve floating above a stick. Spin the stick — the curve smears out into a smooth 3D skin, like a vase on a potter's wheel. To measure that skin, chop the curve into teeny slanted steps. Spin one step and it makes a thin ring, like a slice of an ice-cream cone. A ring's area is easy: "how far around" (that's $2 π$ times how far the ring is from the stick) times "how wide the slanted band is." The one thing everybody forgets: use the leaning length of the step, not its flat floor-shadow — because the curve tilts, and the tilted edge is longer. Add up all the rings and you've painted the whole shape. That sum is the integral. Done.

Radius factor

2 π y

means

the circumference of the circle the point at height

y

traces when spun

Stretch factor

1 + (d y / d x)^{2}

means

how much longer the tilted arc

d s

is than its horizontal shadow

d x

Ring area when

y = 0

zero — the ring collapses to a point (e.g. the sphere's poles)

Radius when rotating about the line

y = k

∣ y - k ∣

, the distance from the point to that line

Connections

Surface area of revolution — the parent result this page derives.
Arc length — the identical $d s = 1 + (d y / d x)^{2} d x$ .
Frustum and cone geometry — source of the $2 π \cdot \frac{r _{1} + r _{2}}{2} \cdot ℓ$ band area.
Volume of revolution (disk & shell) — flat washers there, tilted bands here.
Parametric curves — when $d s$ comes from $(d x / d t)^{2} + (d y / d t)^{2}$ .
Pappus's theorem — the one-line summary $S = 2 π \overset{r}{ˉ} L$ .