The shell method is your best choice when the axis of rotation is perpendicular to the natural variable of your function, because it lets you avoid solving for the other variable. In practical terms, this means: if you have y = f(x) and you’re rotating around a vertical axis, shells let you integrate with respect to x directly, while the disk/washer method would force you to rewrite everything in terms of y. That single difference saves enormous effort on many problems.
The Core Decision: Axis vs. Variable
The shell method and the disk/washer method use opposite relationships between the axis of rotation and the variable of integration. With the disk/washer method, rotating around a vertical axis means you integrate with respect to y, and rotating around a horizontal axis means you integrate with respect to x. The shell method flips this entirely: rotating around a vertical axis uses dx, and rotating around a horizontal axis uses dy.
This opposite behavior is the key to deciding which method to use. Ask yourself: which variable is my function naturally written in? If you have y as a function of x and you’re rotating around the y-axis, the shell method lets you keep everything in terms of x. The disk method would require you to solve for x in terms of y, which ranges from mildly annoying to genuinely impossible depending on the function.
When Shells Save You Real Work
The shell method shines in three specific situations.
Your function is hard (or impossible) to invert. Consider the region under y = √x from x = 0 to x = 4, rotated around the y-axis. Using the washer method, you’d need to rewrite √x as x = y², set up outer and inner radii, and deal with the geometry of two separate radius functions. With shells, you integrate 2πx·√x dx from 0 to 4. One integral, no algebra gymnastics.
The washer method requires two or more integrals. Sometimes the outer or inner radius of a washer changes partway through the region, splitting your integral into pieces. The shell method often collapses those multiple integrals into a single one. When a region is bounded by curves that intersect in ways that change which function defines the boundary, shells frequently give you one clean integral where disks would need two or three.
The integrand is simpler with shells. Even when both methods technically work, the shell integral can be easier to evaluate. For a region where the shell method produces 2πy(2 − 2y) dy while the washer method produces something requiring more simplification, fewer terms means fewer opportunities for errors and faster computation.
The Shell Method Formula
The volume of a solid of revolution using cylindrical shells is:
V = 2π ∫ (radius)(height)(thickness)
For rotation around a vertical axis with functions of x, this becomes V = 2π ∫ from a to b of x·f(x) dx. Each piece has a concrete meaning: the radius is the horizontal distance from the axis of rotation to your shell, the height is the vertical length of the shell (determined by your function), and the thickness is dx or dy depending on your variable of integration.
When your region is bounded by two curves, the height of each shell is the top function minus the bottom function: h(x) = f(x) − g(x). This works exactly like finding the area between two curves, just multiplied by 2π times the radius.
How to Identify the Radius and Height
Setting up the integral correctly comes down to identifying two things: how far each shell sits from the axis of rotation (the radius), and how tall each shell is (the height).
Draw a thin vertical slice through your region if you’re integrating with respect to x, or a thin horizontal slice if you’re integrating with respect to y. The radius is the distance from that slice to the axis of rotation. When rotating around the y-axis, the radius is simply x. When rotating around a different vertical line, say x = k, the radius becomes |x − k|.
The height is the length of that slice, measured along the direction parallel to the axis of rotation. For a vertical slice, the height runs from the bottom curve to the top curve. For a horizontal slice, it runs from the left curve to the right curve.
Rotation Around Non-Standard Axes
Many textbook problems rotate around a line like x = 5 or y = −2 instead of the coordinate axes. The shell method handles these well, but the radius term changes. If you’re rotating around x = 5 and your region sits to the left of that line, the radius of a shell at position x is (5 − x), not just x. Everything else stays the same.
The height function doesn’t change when you shift the axis of rotation. Only the radius is affected, because the radius measures distance to the axis. This makes adapting to non-standard axes straightforward with shells: swap the radius expression and leave everything else alone.
A Quick Decision Checklist
- Function given as y = f(x), rotating around a vertical axis: Use shells. You integrate with respect to x and avoid solving for x in terms of y.
- Function given as y = f(x), rotating around a horizontal axis: Try disks/washers first. They let you integrate with respect to x naturally. Shells would require rewriting in terms of y.
- Function given as x = g(y), rotating around a horizontal axis: Use shells. You integrate with respect to y and keep your original function.
- Multiple integrals needed for disks: Check whether shells reduce it to one integral. They often do.
- Function can’t be solved for the other variable: Use whichever method lets you keep the variable you already have.
Common Setup Mistakes
The most frequent error is mixing up the variable relationships. Students often set up shells the same way they’d set up disks, integrating with respect to y for a vertical axis of rotation. Remember that shells work opposite to disks: vertical axis means dx, horizontal axis means dy.
Another common mistake is using the wrong radius when the axis of rotation isn’t a coordinate axis. If you’re rotating around x = 3, writing the radius as just x gives you the distance from the y-axis, not from your actual axis. Always measure radius as the distance from the shell to the line you’re rotating around.
Finally, watch your limits of integration. They should correspond to the variable you’re integrating with respect to, not the other variable. If you’re integrating with respect to x, your limits are x-values marking where the region starts and ends horizontally. If a problem gives you y-bounds but you’re doing shells in x, you need to convert those bounds.