23.2 Surface integrals of vector fields

Surface integrals of vector fields measure the total flow of a vector field through a surface. Think of fluid passing through a pipe cross-section or electric field lines penetrating a charged surface. You calculate them by integrating the dot product of the vector field with the surface normal over the entire surface.

This topic pulls together your knowledge of vector fields, parametrized surfaces, and dot/cross products, then applies them to flux calculations that show up constantly in physics and engineering.

Vector Fields and Surface Integrals

Vector Field Fundamentals

A vector field assigns a vector to each point in a region of space. In 3D, you write it as $\vec{F}(x,y,z) = (P(x,y,z),\, Q(x,y,z),\, R(x,y,z))$. Common physical examples include velocity fields in a fluid, gravitational force fields, and electric fields.

The flux integral measures the total flow of a vector field through a surface. It's computed by integrating the dot product of the vector field $\vec{F}$ and the surface normal vector $\vec{n}$ over the surface. The result tells you how much "stuff" (fluid, energy, field) passes through.

A few pieces you need:

• The surface normal vector is perpendicular to the surface at each point. For closed surfaces (spheres, cubes), it points outward by convention. The choice of normal direction is what gives the surface its orientation.
• The dot product $\vec{a} \cdot \vec{b} = a_1b_1 + a_2b_2 + a_3b_3$ measures how aligned two vectors are. It's maximized when the vectors are parallel and zero when they're perpendicular. This is why flux captures only the component of $\vec{F}$ that actually passes through the surface, not the component sliding along it.

Calculating Surface Integrals

Here's the step-by-step process for evaluating a flux integral $\iint_S \vec{F} \cdot d\vec{S}$:

1. Parametrize the surface. Express the surface using two parameters $(u,v)$: $\vec{r}(u,v) = (x(u,v),\, y(u,v),\, z(u,v))$ and identify the parameter domain $D$ in the $uv$-plane.

2. Compute the normal vector. Take the partial derivatives $\frac{\partial \vec{r}}{\partial u}$ and $\frac{\partial \vec{r}}{\partial v}$, then take their cross product: $\vec{n}\, du\, dv = \left(\frac{\partial \vec{r}}{\partial u} \times \frac{\partial \vec{r}}{\partial v}\right) du\, dv$ This cross product gives a vector normal to the surface whose magnitude accounts for the area element. You don't need to normalize it separately.

3. Evaluate the dot product. Substitute the parametrization into $\vec{F}$ and dot it with the cross product from step 2.

4. Integrate over the parameter domain: $\iint_S \vec{F} \cdot d\vec{S} = \iint_D \vec{F}(\vec{r}(u,v)) \cdot \left(\frac{\partial \vec{r}}{\partial u} \times \frac{\partial \vec{r}}{\partial v}\right) du\, dv$

Common shortcut for graphs: If the surface is given as $z = g(x,y)$, you can skip the full parametrization. The flux integral becomes:

$\iint_S \vec{F} \cdot d\vec{S} = \iint_D \vec{F}(x,y,g(x,y)) \cdot \left(-g_x,\, -g_y,\, 1\right) dx\, dy$

where the normal points in the positive $z$-direction. This saves a lot of work on homework problems.

Orientation and the Divergence Theorem

Surface Orientation

Every smooth surface has two possible orientations, corresponding to the two directions the normal vector can point. For a closed surface, the convention is outward-pointing normals. For an open surface (a piece of a plane, a paraboloid cap, etc.), you need to be told which orientation to use, or choose one and stay consistent.

Orientation directly affects the sign of the flux integral. Flipping the normal reverses the sign. If you compute a flux of $+5$ with outward normals, the same integral with inward normals gives $-5$. Getting orientation wrong is one of the most common mistakes on exams.

Divergence Theorem

The Divergence Theorem (also called Gauss's theorem) converts a flux integral over a closed surface into a volume integral over the region it encloses:

$\iint_S \vec{F} \cdot d\vec{S} = \iiint_V \nabla \cdot \vec{F}\, dV$

The divergence of $\vec{F} = (P, Q, R)$ is the scalar field:

$\nabla \cdot \vec{F} = \frac{\partial P}{\partial x} + \frac{\partial Q}{\partial y} + \frac{\partial R}{\partial z}$

Divergence measures the net rate of outward expansion of the field at a point. Positive divergence means the field is "spreading out" (a source); negative means it's "pulling in" (a sink).

Why is this theorem useful? Sometimes the surface integral is hard to compute directly (imagine flux through all six faces of a cube), but the volume integral of the divergence is straightforward. The theorem only applies to closed surfaces with outward orientation enclosing a bounded volume where $\vec{F}$ is smooth throughout.

Applications

Fluid Dynamics

A velocity field $\vec{v}(x,y,z)$ gives the velocity of a fluid at each point. The volumetric flow rate through a surface $S$ is the flux integral:

$Q = \iint_S \vec{v} \cdot d\vec{S}$

For example, if water flows through a circular pipe of radius $R$ with velocity profile $\vec{v} = (0, 0, v_0(1 - r^2/R^2))$, you'd integrate this over the pipe's cross-section to find the total flow rate. Positive flux means net outward flow; negative means net inward flow.

The Divergence Theorem connects to a key physical idea: for an incompressible fluid (like water at everyday conditions), $\nabla \cdot \vec{v} = 0$ everywhere. That means the net flux through any closed surface is zero, so whatever flows in must flow out. Non-zero divergence indicates sources (fluid being created) or sinks (fluid being removed) within the region.

Electromagnetism

Electric and magnetic fields are vector fields, and flux integrals appear throughout Maxwell's equations.

• Electric flux through a surface: $\Phi_E = \iint_S \vec{E} \cdot d\vec{S}$. This counts the net number of electric field lines passing through $S$.
• Magnetic flux through a surface: $\Phi_B = \iint_S \vec{B} \cdot d\vec{S}$. Changes in magnetic flux through a loop induce an electromotive force (Faraday's law).

Gauss's Law is the Divergence Theorem applied to the electric field. For any closed surface $S$:

$\iint_S \vec{E} \cdot d\vec{S} = \frac{Q_{\text{enc}}}{\epsilon_0}$

where $Q_{\text{enc}}$ is the total charge enclosed and $\epsilon_0$ is the permittivity of free space ($\approx 8.85 \times 10^{-12}\, \text{C}^2/\text{N·m}^2$). This means the net electric flux through a closed surface depends only on the charge inside, regardless of the surface's shape. It's one of the cleanest applications of the math you've been building all semester.