A curve in a vertical half-plane generates a surface when rotated about the vertical axis. Its distance from that axis determines the circumference swept out, while its arc length determines the width of a narrow surface strip. Deriving both factors from a cross product avoids confusing profile length, projected area and actual surface area.
This section builds on the surface-integral definition, including its regularity, orientation and repeated-coverage warnings. The mass and surface-area applications provide the density and squared-distance models, while Stokes and divergence supplies the closed-boundary flux theorem. Here we develop profile-based rotation, a torus and an annular closing face.
Study route
This unit derives the two tangent vectors of a rotated profile, checks coverage and distinguishes the two possible normal directions, then sets up weighted scalar integrals and decides which pieces close an open surface. Required tools are the chain rule, cross products, one-variable arc length and elementary double or triple integration.
Start from a profile, not a remembered area formula
Definition 6.1.1
(Profile and surface of revolution)
A profile is a curve , , in the half-plane , where are on a neighborhood of and . Rotating it about the axis gives the surface of revolution parametrized by , .
Theorem 6.1.2
(Area of a surface of revolution)
Let the profile be simple, except that its endpoints may coincide when it is closed. Then
Proof
(Derivation from the cross product)
Let the profile be , where are on a neighborhood of a compact interval . Assume
The rotated parametrization is
Its tangent vectors are
Hence
and
The assumptions ensure regularity. To interpret the parameter integral as once-only geometric area, also require the profile to have no repeated points except possibly equal endpoints of a simple closed profile. With positive radius, a surface point determines its profile point and angle modulo the angular seam. The finitely many smooth seam curves do not add surface area. One may work with ordinary local patches across each seam rather than claiming that the closed rectangle itself is globally one-to-one.
Therefore
This is derived from the surface definition; no centroid theorem for surfaces of revolution is being assumed. If the profile or angular interval traces the same patch twice, the parameter integral counts it twice.
Proof X-ray
What each assumption buys
The non-routine step is to read the cross product as circumference times profile speed: , where comes from (the circle swept by one profile point) and the square root from (the profile's own speed). The hypothesis keeps and keeps ; together they make the norm positive, which is regularity. Simplicity of the profile is used only at the end, to turn the parameter integral into geometric area. If the profile reaches the axis, gives and the local argument fails; if the profile is traced twice, for example out and back along the same segment, the integral is double the area.
Graph versions and the role of the axis
For a profile written as , the formula becomes
For one written as , it becomes
Either representation can simplify a calculation. A derivative can become unbounded in one graph description even when another parameter gives a smooth profile.
If the profile meets the axis, the cross product above vanishes there. This does not automatically mean the geometric surface is singular: the sphere's poles are familiar coordinate exceptions. But the positive-radius regular-patch argument no longer applies at those points. Use another chart or a justified limiting argument. The examples below keep their profile radii positive, so this extra step is unnecessary.
Worked example 6.1.3
(A truncated cone and its normal)
Find the area of the surface generated by the profile , , , and a unit normal pointing out of the solid that lies inside the cone.
Here and , so the profile speed is and
As a check, a frustum with radii and and slant length has lateral area . The cross product is , which points toward the axis and upward, into the solid. The outward unit normal is therefore . The profile stays away from the axis because ; extending it to would reach the apex, where the cross product vanishes.
Common mistake
(Dropping the radius factor)
Using , circumference times profile length with no , gives for the frustum above instead of . The circumference varies along the profile and must stay inside the integral.
Quick check
(Use the area formula for the profile , , .)
Compute the profile speed first.
Solution · Answer
and , so the speed is and , the lateral area of a cylinder of radius and height .
Area does not select an outward normal
The cross product as ordered above has radial component . Reversing the order changes its sign. Whether either vector is outward depends on the solid being considered, not just on a desired positive coordinate component.
Worked example 6.1.4
(Area and outward normal of a torus)
For a torus with major radius and tube radius , where , set
The profile is a circle whose radius coordinate stays positive. Its speed is , and
The parenthesized vector is unit length. It points away from the center of the tube's circular cross-section, so it is outward for the solid torus. Near its inner equator the radial component is negative. Outward does not mean away from the vertical axis everywhere.
For , the scalar area is
The full torus is regular because . The limiting case is outside these assumptions and must not be inserted into a proof requiring strictly positive radius.
Quick check
(What is the outward unit normal of the torus at the top point and at the inner equator point ?)
Find the values of and for each point.
Solution · Answer
The top point has , , so . The inner equator point has , , so , which points toward the axis.
Concept lensGeometric
Length times the path of the profile's centroid
The area formula also says , where is the length of the profile and is the average distance of its points from the axis, so is the circle traced by the profile's centroid. This is Pappus's centroid theorem for surfaces. For the torus, and , giving ; for the frustum, and , giving .
A surface density is evaluated on the rotated point
For continuous area density ,
For a physical mass model assume nonnegative density and positive mass. Weighted first moments use with the same surface factor, then divide by mass. The moment about the vertical axis uses
If the density depends only on height and distance from the axis, angular integration makes both horizontal first moments zero. The vertical coordinate still needs its own weighted computation or a symmetry that preserves density.
A density per unit area differs from a density per unit profile length. Multiplying a profile integral by a circumference is correct only after the density and area model have been identified.
Worked example 6.1.5
(A torus with height-dependent surface density)
Let the torus surface above carry area density with . Find its mass.
On we have and , so
Since and , the mass is . For the dimension check, gives . The density vanishes on the outer and inner equators, where , and is largest on the top and bottom circles.
A rotated open profile may close with an annulus
A top half of a torus has two boundary circles in the plane . Its closing planar face is an annulus, not a disk: material lies between the inner and outer radii. Adding a full disk would introduce an extra face across the central hole.
For the upper half of the solid torus, the planar base has downward outward normal. If the field has a nonzero vertical component on that base, its flux must be included. The divergence theorem computes flux through the complete closed boundary; subtract the correctly oriented base contribution to obtain flux through the curved part.
Always check that the vector field is on a neighborhood of the whole solid and that all boundary pieces have compatible outward normals. A torus's central hole does not invalidate the divergence theorem on a regular solid torus; an undefined field inside the material would.
Worked example 6.1.6
(Flux through the upper half of a torus)
Let be the part of the torus surface with , oriented outward, and let . Find .
The solid has boundary together with the base in the plane , which is the annulus . On the outward normal is and , so the base flux is . Also , and : the meridian cross-section is a half-disk of radius , and , where the last integral vanishes by symmetry. The divergence theorem gives
For a direct check, on we have and , with . Then
Counterexample mode
Closing with a disk instead of an annulus
Suppose one closes with the full disk . The base flux would be and the claimed flux would be . For , this is , whereas the annulus and the direct integral both give . The discrepancy is the flux through the disk of radius spanning the hole, which is not part of . The correct closing face is the annulus.
Quick check
(The lower half of the solid torus () is closed by a planar face at . What is the outward normal of that face, and what is its area?)
Decide whether the face is a disk or an annulus.
Solution · Answer
The solid lies below the plane, so the outward normal is . The face is the annulus , with area .
Checklist
- Derive the profile speed and multiply by the positive radius in the area factor.
- Check regularity and once-only coverage, including both seams for a closed profile.
- Derive the outward direction from the solid. Do not replace the scalar area factor with a signed radial component.
- Evaluate density on the surface and use the mass denominator for center-of-mass coordinates.
- Identify the squared distance to the requested axis.
- When closing a surface, list every boundary circle and determine whether the planar face is a disk or an annulus.