Surface Integrals and Flux

A surface integral is the two-dimensional analogue of a line integral. Where a line integral accumulates the tangential component of a field along a curve, a surface integral accumulates the normal component of a field across a surface. The quantity it produces — the flux — answers a concrete physical question: how much of the field flows through this surface per unit time, or how many field lines pierce it. In electromagnetism this is not a metaphor: for a fluid velocity field the flux is literally the volumetric flow rate, and for $\mathbf{E}$ and $\mathbf{B}$ the same construction quantifies enclosed charge and induced EMF. The central machinery is the vector area element $\mathrm{d}\mathbf{a}$ . We tile the surface into tiny nearly-flat patches; each patch has an area $\mathrm{d}a$ and a direction perpendicular to it, encoded in the unit normal $\hat{\mathbf{n}}$ . The product $\mathrm{d}\mathbf{a} = \hat{\mathbf{n}}\,\mathrm{d}a$ packages both. The flux through the patch is $\mathbf{v}\cdot \mathrm{d}\mathbf{a} = v\cos\theta\,\mathrm{d}a$ , where $\theta$ is the angle between the field and the normal — only the perpendicular component contributes, the tangential component sliding along the surface without crossing it. Two orientation conventions matter: closed surfaces take outward normals by universal agreement, while open surfaces tie the normal to the boundary direction by the right-hand rule — and mixing the conventions is the single most common source of sign errors. Surface integrals are the hinge between the differential and integral statements of electromagnetism. The divergence (Gauss) theorem converts a closed surface integral into a volume integral of the divergence; the curl (Stokes) theorem converts a line integral around a closed loop into a surface integral of the curl over any spanning surface. Mastering the surface integral therefore unlocks both of Maxwell's flux-based laws and the field-theoretic reformulation of the whole subject.

Key takeaways

A surface integral $\int_S \mathbf{v}\cdot \mathrm{d}\mathbf{a}$ adds up the component of a vector field $\mathbf{v}$ piercing a surface $S$ , weighted by area; the result is called the flux of $\mathbf{v}$ through $S$ .
The vector area element $\mathrm{d}\mathbf{a} = \hat{\mathbf{n}}\,\mathrm{d}a$ carries the magnitude of an infinitesimal patch and the direction of its outward (or chosen) unit normal $\hat{\mathbf{n}}$ .
For a closed surface the convention is outward normals, written $\oint_S \mathbf{v}\cdot\mathrm{d}\mathbf{a}$ ; for an open surface the normal orientation is tied to the boundary direction by the right-hand rule.
Flux is the geometric object that appears on the left of Gauss's law ( $\oint \mathbf{E}\cdot \mathrm{d}\mathbf{a} = Q_{\mathrm{enc}}/\varepsilon_0$ ) and on the right of Faraday's law (the magnetic flux $\Phi_B$ ).
Surface integrals connect to volume integrals through the divergence theorem and to line integrals through Stokes's theorem — the two pillars of vector calculus in electromagnetism.
The value of a flux integral depends only on the field and the surface; for a closed surface in a divergence-free field it can even be deformed freely, a fact used constantly in field-line reasoning.
Key idea: Flux is the integral of the normal component of a field over a surface, $\Phi = \int_S \mathbf{v}\cdot\hat{\mathbf{n}}\,\mathrm{d}a$ ; the dot product selects the piercing part of the field, and the chosen orientation of $\hat{\mathbf{n}}$ fixes the sign.

Where this lives today

Every solar farm is a flux calculation running in real time: panel output scales with the cosine between the sun direction and the panel normal — the rain-frame factor of this article — which is why utility installations invest in single-axis trackers that keep $\cos\theta$ near 1 through the day, harvesting 20–30% more energy than fixed tilt. At the quantum end, magnetic flux is not just measurable but quantized: through a superconducting loop it comes in integer multiples of the flux quantum $\Phi_0 = h/2e = 2.07\times10^{-15}\ \mathrm{Wb}$ . SQUID magnetometers exploit this to resolve fields a hundred billion times weaker than Earth's — reading hearts (magnetocardiography), brains (MEG), and mineral deposits as flux through a millimetre loop. And spacecraft power budgets, radiation-pressure sails, and antenna link budgets are all Poynting-flux integrals over surfaces: the power your phone receives from a cell tower is $\int\mathbf{S}\cdot\mathrm{d}\mathbf{a}$ over its antenna's effective area.

The physical picture

Picture rain falling and a wire frame held in the stream. The flux is the number of drops per second passing through the frame. Hold it face-on to the rain and you catch the maximum; tilt it edge-on and nothing passes through even though the frame still has area. That tilt factor is exactly the $\cos\theta = \hat{\mathbf{v}}\cdot\hat{\mathbf{n}}$ in the dot product.

Figure

The rain test — same frame, same rain, three fluxes: the tilt factor is the whole story

Field lines make flux visual: if you draw the field so that line density represents field strength, then the flux through a surface is simply the net number of lines crossing it, counting lines going one way as positive and the other way as negative. For a closed surface, net flux counts lines that originate inside (sources) minus those that terminate inside (sinks) — which is why a closed-surface flux measures enclosed sources. This is the E&M payoff waiting two articles ahead: Gauss's law is nothing but "flux out of a closed surface = enclosed charge over $\varepsilon_0$ ", and choosing clever Gaussian surfaces will turn field calculations into one-liners. Only the perpendicular component matters because the parallel component runs along the surface, sliding past without ever crossing. Decompose any field at a patch into a piece through the surface and a piece in the surface; the integral keeps only the former. This is the geometric meaning of the dot product with the normal. Orientation is a bookkeeping device for sign. A closed surface unambiguously has an inside and outside, so 'outward' is canonical. An open surface is two-sided with no canonical front, so we must declare one — and the only self-consistent declaration that makes Stokes's theorem work pairs the normal with a circulation direction by the right-hand rule.

Flux, defined and computed

The definition tiles the surface and sums the piercing contributions:

\Phi = \int_S \mathbf{v}\cdot \mathrm{d}\mathbf{a} = \int_S \mathbf{v}\cdot\hat{\mathbf{n}}\,\mathrm{d}a \qquad (1)

where: $\Phi$ — flux of the field through the surface (a scalar); $\mathbf{v}$ — vector field being integrated; $S$ — the surface; $\mathrm{d}\mathbf{a}=\hat{\mathbf{n}}\,\mathrm{d}a$ — vector area element; $\hat{\mathbf{n}}$ — unit normal to the patch; $\mathrm{d}a$ — the patch's scalar area. Patch by patch, the integrand is a projection:

\mathbf{v}\cdot\mathrm{d}\mathbf{a} = v\,\mathrm{d}a\,\cos\theta \qquad (2)

where: $v=|\mathbf{v}|$ ; $\theta$ — angle between field and normal. Resolve the field at a patch into $v_\perp = v\cos\theta$ through the surface and $v_\parallel = v\sin\theta$ along it: the parallel part crosses nothing, and the dot product keeps exactly the throughput component. The rain-frame picture, formalized. For a closed surface — one with no boundary, enclosing a volume — the integral gets a ring and a fixed convention:

\oint_S \mathbf{v}\cdot \mathrm{d}\mathbf{a} \quad (\hat{\mathbf{n}}\ \text{outward}) \qquad (3)

where: the outward normal makes positive flux mean net outflow. This is the object Gauss's law is written with, and the object the divergence theorem converts into a volume integral. For actually computing on curved surfaces, the area element comes from a parametrization, and its cross-product form is derived below:

\mathrm{d}\mathbf{a} = \left(\frac{\partial \mathbf{r}}{\partial u}\times\frac{\partial \mathbf{r}}{\partial v}\right)\mathrm{d}u\,\mathrm{d}v \qquad (4)

where: $\mathbf{r}(u,v)$ — parametrization of the surface; $\partial_u\mathbf{r},\ \partial_v\mathbf{r}$ — tangent vectors along the coordinate curves. The single most used special case is the sphere:

\mathrm{d}a = r^2\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi,\qquad \mathrm{d}\mathbf{a} = r^2\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi\,\hat{\mathbf{r}} \qquad (5)

where: $r$ — sphere radius; $\theta,\phi$ — polar and azimuthal angles; $\hat{\mathbf{r}}$ — outward radial normal. The $r^2\sin\theta$ is the Jacobian of the sphere — large patches near the equator, vanishing ones at the poles.

Derivation AThe vector area element from a parametrization

We are deriving equation (4) — the cross-product recipe that turns any parametrized surface into a computable flux integral, and the two-dimensional sibling of the line integral's $\mathrm{d}\boldsymbol{\ell}=\mathbf{r}'(t)\,\mathrm{d}t$ . The strategy: watch what a small rectangle in parameter space becomes on the surface, and recognize the parallelogram whose area and orientation the cross product encodes in one stroke. The step to watch is the geometric meaning of the cross product — magnitude = parallelogram area, direction = normal — doing double duty; the same fact will make the spherical element (5) drop out as a special case.

Step 1. Move along the coordinate curves. From the point $\mathbf{r}(u,v)$ , a parameter step $\mathrm{d}u$ displaces you by $\partial_u\mathbf{r}\,\mathrm{d}u$ and a step $\mathrm{d}v$ by $\partial_v\mathbf{r}\,\mathrm{d}v$ . Why: these are the definitions of the tangent vectors — displacement per unit parameter along each family of coordinate curves ruled on the surface.

Step 2. The two edge vectors span a small parallelogram patch of the surface, and the cross product packages its geometry:

\mathrm{d}\mathbf{a} = (\partial_u\mathbf{r}\,\mathrm{d}u)\times(\partial_v\mathbf{r}\,\mathrm{d}v) = \left(\frac{\partial\mathbf{r}}{\partial u}\times\frac{\partial\mathbf{r}}{\partial v}\right)\mathrm{d}u\,\mathrm{d}v

Why: $|\mathbf{a}\times\mathbf{b}|$ is the area of the parallelogram with sides $\mathbf{a},\mathbf{b}$ , and the direction is perpendicular to both — exactly the $\hat{\mathbf{n}}\,\mathrm{d}a$ we need. Swapping $u\leftrightarrow v$ flips the sign: the parametrization order chooses the orientation.

Step 3. Check on the sphere of radius $R$ : $\mathbf{r}(\theta,\phi)=R(\sin\theta\cos\phi,\ \sin\theta\sin\phi,\ \cos\theta)$ . The tangent vectors have magnitudes $|\partial_\theta\mathbf{r}| = R$ and $|\partial_\phi\mathbf{r}| = R\sin\theta$ , they are orthogonal, and their cross product points radially:

\mathrm{d}\mathbf{a} = R^2\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi\,\hat{\mathbf{r}}

Why: this is equation (5) — the mysterious $\sin\theta$ is just the shrinking circumference of latitude circles near the poles, made quantitative by the cross product. *Verified: integrating $\mathrm{d}a$ over the whole sphere gives $R^2\cdot 2\pi\cdot[-\cos\theta]_0^\pi = 4\pi R^2$ , the known area; matches Griffiths §1.3.3 and §1.4.1.*

Figure

Orientation rules — closed surfaces point outward, open surfaces obey the right hand of their rim

The theorems this object powers

The flux integral is the common currency of the two great integral theorems, each developed fully in its own article. The divergence theorem trades a closed-surface flux for a volume integral of the divergence:

\oint_S \mathbf{v}\cdot \mathrm{d}\mathbf{a} = \int_V (\nabla\cdot\mathbf{v})\,\mathrm{d}\tau \qquad (6)

where: $V$ — the volume enclosed by $S$ ; $\mathrm{d}\tau$ — volume element. Net outflow through the skin equals total source strength inside. And Stokes's theorem trades an open-surface flux of a curl for a circulation around the rim:

\int_S (\nabla\times\mathbf{v})\cdot \mathrm{d}\mathbf{a} = \oint_{\partial S} \mathbf{v}\cdot \mathrm{d}\boldsymbol{\ell} \qquad (7)

where: $\partial S$ — the boundary curve of the open surface, oriented with $\hat{\mathbf{n}}$ by the right-hand rule. Together they explain why flux integrals dominate electromagnetism: Maxwell's equations in integral form are two flux statements and two circulation statements. The flagship is Gauss's law for a point charge:

\oint_S \frac{1}{4\pi\varepsilon_0}\frac{q\,\hat{\mathbf{r}}}{r^2}\cdot\mathrm{d}\mathbf{a} = \frac{q}{\varepsilon_0} \qquad (8)

where: $q$ — enclosed point charge; $\varepsilon_0 = 8.854\times10^{-12}\ \mathrm{F/m}$ . The derivation below shows why the left side has the same value for every enclosing surface — the geometric miracle that makes Gauss's law useful.

Go deeper · Derivation B · 2 minThe flux of an inverse-square field is shape-independent

Derivation BThe flux of an inverse-square field is shape-independent

We are computing the flux of $\mathbf{v}=\hat{\mathbf{r}}/r^2$ — the shape of every point-charge field — through closed surfaces around the origin, and proving the answer is always exactly $4\pi$ , regardless of the surface's size or shape. This is the geometric heart of Gauss's law: it is why only the enclosed charge matters, not where it sits or how the surface bulges. The strategy: evaluate on the easiest surface (a centred sphere), watch the radius cancel, then leverage the divergence theorem and the field's sourcelessness away from the origin to export the result to arbitrary shapes. The cancellation in Step 2 is the inverse-square law meeting the $r^2$ growth of spherical area — the same duel that made this field divergence-free off-origin.

Step 1. Choose a sphere of radius $R$ centred on the origin, where the outward normal is $\hat{\mathbf{r}}$ everywhere:

\mathrm{d}\mathbf{a} = R^2\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi\,\hat{\mathbf{r}}

Why: on this surface field and normal are parallel at every point — the dot product will collapse to a product of magnitudes.

Step 2. Form the integrand and watch the radius vanish:

\frac{\hat{\mathbf{r}}}{R^2}\cdot\mathrm{d}\mathbf{a} = \frac{1}{R^2}\,R^2\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi = \sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi

Why: the $1/R^2$ weakening of the field is exactly repaid by the $R^2$ growth of the area element — the integrand is pure solid angle, blind to the sphere's size.

Step 3. Integrate over the full sphere:

\oint_S\frac{\hat{\mathbf{r}}}{r^2}\cdot\mathrm{d}\mathbf{a} = \int_0^{2\pi}\!\!\int_0^{\pi}\sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi = 2\pi\cdot\big[-\cos\theta\big]_0^{\pi} = 2\pi\cdot 2 = 4\pi

Why: $4\pi$ steradians is the full solid angle — the flux counts how much of the sky the surface covers as seen from the source, and a closed surface covers all of it.

Step 4. Export to any enclosing shape. Between an arbitrary enclosing surface $S'$ and a small centred sphere, the field has $\nabla\cdot(\hat{\mathbf{r}}/r^2)=0$ everywhere (the origin is excluded from this shell region), so the divergence theorem gives zero net flux out of the shell — hence the flux through $S'$ equals the flux through the sphere: $4\pi$ . A surface not enclosing the origin bounds a region where the field is entirely divergence-free, and its flux is zero. Verified: reproduces Griffiths Eq. 1.103 and the solid-angle normalization of Gauss's law; multiplying by $q/4\pi\varepsilon_0$ turns Step 3 into equation (8).

Limiting cases

How the flux integral behaves in instructive special cases:

Case	Result	Meaning
Uniform field, flat surface	$\Phi = \mathbf{v}\cdot\mathbf{A} = v A\cos\theta$	When the field is constant and the surface flat with vector area $\mathbf{A}$ , the integral collapses to a single dot product; $\theta$ is the tilt between field and normal.
Field tangent to surface	$\mathbf{v}\cdot\hat{\mathbf{n}} = 0 \;\Rightarrow\; \Phi = 0$	If the field is everywhere parallel to the surface (grazing), nothing pierces it and the flux vanishes regardless of the field's magnitude.
Closed surface, divergence-free field	$\nabla\cdot\mathbf{v}=0 \;\Rightarrow\; \oint_S \mathbf{v}\cdot\mathrm{d}\mathbf{a}=0$	For a solenoidal field (such as $\mathbf{B}$ ) the net flux through any closed surface is zero — every line entering must exit. This is why the spanning surface in Stokes's theorem can be deformed freely.
Point source at center	$\oint_S \frac{\hat{\mathbf{r}}}{r^2}\cdot\mathrm{d}\mathbf{a} = 4\pi$	The flux of an inverse-square radial field through any closed surface enclosing the origin is $4\pi$ — the solid-angle origin of Gauss's law, independent of surface shape.
Uniform field, closed surface	$\oint_S\mathbf{v}_0\cdot\mathrm{d}\mathbf{a}=0$	Whatever enters one side leaves the other; constant fields have no sources to catch.

Worked examples

Example 1 — a tilted panel in a uniform field (confidence builder). A flat plate of area $A = 0.020\ \mathrm{m^2}$ sits in a uniform electric field $E = 200\ \mathrm{N/C}$ . Find the electric flux when the plate's normal is tilted at $\theta = 60^\circ$ to the field, and when face-on.

Step 1. Uniform field, flat surface — equation (1) collapses to a single dot product:

\Phi_E = \mathbf{E}\cdot\mathbf{A} = EA\cos\theta = (200)(0.020)\cos 60^\circ = (4.0)(0.500) = 2.0\ \mathrm{N\,m^2/C}

Step 2. Face-on ( $\theta = 0$ ): $\Phi_E = EA = 4.0\ \mathrm{N\,m^2/C}$ — the maximum. Edge-on ( $\theta=90^\circ$ ): zero, however strong the field. Tilting from face-on to $60^\circ$ costs exactly half the flux — the same cosine that sets solar-panel output through the day.

Step 3. Now close the surface: fold this plate into any closed box in the same uniform field. Every line entering one face exits another, so $\oint\mathbf{E}\cdot\mathrm{d}\mathbf{a} = 0$ — consistent with zero enclosed charge, our first taste of Gauss's law reasoning.

Example 2 — the divergence theorem checked face by face (exam level). Verify the divergence theorem for $\mathbf{v} = \mathbf{r} = x\hat{\mathbf{x}}+y\hat{\mathbf{y}}+z\hat{\mathbf{z}}$ over a cube of side $L = 2.0\ \mathrm{m}$ with one corner at the origin and edges along the axes.

Step 1. The $x=L$ face has outward normal $+\hat{\mathbf{x}}$ and on it $\mathbf{r}\cdot\hat{\mathbf{x}} = x = L$ everywhere:

\int \mathbf{r}\cdot\hat{\mathbf{x}}\,\mathrm{d}a = L\cdot L^2 = L^3 = 8.0\ \mathrm{m^3}

Step 2. The $x=0$ face has outward normal $-\hat{\mathbf{x}}$ , and there $\mathbf{r}\cdot(-\hat{\mathbf{x}}) = -x = 0$ : it contributes nothing. The same pattern holds for the $y$ and $z$ face pairs — the far face gives $L^3$ , the near face zero.

Step 3. Sum the six faces:

\oint_S\mathbf{r}\cdot\mathrm{d}\mathbf{a} = 3L^3 = 3(8.0) = 24.0\ \mathrm{m^3}

Step 4. Volume side: $\nabla\cdot\mathbf{r} = 3$ everywhere (the hedgehog field of the divergence article), so

\int_V(\nabla\cdot\mathbf{r})\,\mathrm{d}\tau = 3\times L^3 = 24.0\ \mathrm{m^3}

Both sides give exactly 24.0 — the divergence theorem verified by hand. Notice the accounting: all the flux escapes through the three far faces, yet the volume integral spreads the sources uniformly inside; the theorem guarantees the books balance however the geometry distributes them.

Applications

Gauss's law and enclosed charge. The integral form of Gauss's law, $\oint_S \mathbf{E}\cdot\mathrm{d}\mathbf{a} = Q_{\mathrm{enc}}/\varepsilon_0$ , is a flux statement: the total electric flux out of a closed surface measures the charge it encloses, independent of the surface's shape. Exploiting symmetry, one chooses a 'Gaussian surface' on which $\mathbf{E}$ is constant and either parallel or perpendicular to $\hat{\mathbf{n}}$ , reducing the surface integral to $E\times(\text{area})$ and yielding the field of spheres, cylinders, and planes in a single line.

Magnetic flux and Faraday's law. The magnetic flux $\Phi_B = \int_S \mathbf{B}\cdot\mathrm{d}\mathbf{a}$ through a circuit determines the induced EMF via $\mathcal{E} = -\,\mathrm{d}\Phi_B/\mathrm{d}t$ . Because $\mathbf{B}$ is divergence-free, the flux depends only on the boundary loop, not the particular spanning surface — so one can choose the most convenient surface (a flat disk, a bulging cap) to evaluate it. This surface-independence is essential for analyzing solenoids, transformers, and generators.

Fluid flow rate. For a fluid with velocity field $\mathbf{v}$ , the surface integral $\int_S \mathbf{v}\cdot\mathrm{d}\mathbf{a}$ is literally the volume of fluid crossing $S$ per unit time. This is the original physical picture from which 'flux' takes its name (Latin fluxus, flowing). Mass flow rate follows by inserting density: $\dot m = \int_S \rho\,\mathbf{v}\cdot\mathrm{d}\mathbf{a}$ , the continuity equation in integral form.

Poynting flux and radiated power. The power carried by an electromagnetic field through a surface is the flux of the Poynting vector, $P = \oint_S \mathbf{S}\cdot\mathrm{d}\mathbf{a}$ with $\mathbf{S} = \frac{1}{\mu_0}\mathbf{E}\times\mathbf{B}$ . Integrating over a large sphere surrounding an antenna gives the total radiated power, and the inverse-square falloff of $\mathbf{S}$ combined with the $r^2$ growth of the sphere's area is what makes radiated power finite and independent of the sphere's radius.

Origins · 1 minOliver Heaviside and how this was discovered

Oliver Heaviside

1850–1925 · London and Devon, self-taught

Heaviside left school at sixteen, worked briefly as a telegraph operator, and then — deaf, poor, and gloriously unemployable — spent five decades doing electromagnetic theory alone in rented rooms. Reading Maxwell's Treatise, he decided its twenty quaternion-flavoured equations obscured the physics, and in 1885 recast them as the four vector equations every student now learns: two statements about flux (the divergences of $\mathbf{E}$ and $\mathbf{B}$ ) and two about circulation. Flux thinking was his native language — he co-discovered the Poynting flux of energy independently, invented the operational calculus engineers still use, predicted the ionospheric layer that bears his name, and coined much of the field's vocabulary: inductance, impedance, permeability. The Royal Society eventually elected the telegraph clerk a Fellow. When told his methods lacked rigour he replied that he did not refuse his dinner simply because he did not fully understand digestion — a fitting motto for every physicist who computes a flux integral before proving the surface is smooth.

History

1785

— Coulomb quantifies the inverse-square electric force, the field law whose flux will prove shape-independent

1813

— Gauss relates surface integrals to volume integrals, the mathematical divergence theorem

1831

— Faraday's induction experiments make the flux through a circuit — his "lines of force cut" — a measurable physical quantity

1835

— Gauss formulates the flux law of electrostatics in essentially modern form

1873

— Maxwell's Treatise builds electromagnetic theory on flux and circulation integrals over surfaces and loops

1885

— Heaviside compresses Maxwell into four vector equations, two of them flux laws; Poynting's energy-flux theorem arrives the same year

Going further

References. Griffiths, Introduction to Electrodynamics, 4th ed., §1.3.3 (Surface Integrals) · Jackson, Classical Electrodynamics, 3rd ed., Introduction & §1.1 · Purcell & Morin, Electricity and Magnetism, 3rd ed., §1.10 & §2.1 · Zangwill, Modern Electrodynamics, §1.5 (Vector Integral Calculus)

Provenance. All equations are standard vector-calculus and electromagnetism results. A verifier should recompute: the spherical-surface flux integral $\int_0^{2\pi}\int_0^\pi \sin\theta\,\mathrm{d}\theta\,\mathrm{d}\phi = 4\pi$ (Griffiths Eq. 1.103 region); the cube divergence-theorem check giving $3L^3$ on both sides (Griffiths §1.3.4); and $\varepsilon_0 = 8.854\times10^{-12}\ \mathrm{F/m}$ (CODATA). Cross-check definitions against Griffiths §1.3.3, Jackson Introduction, Purcell §1.10, Zangwill §1.5.

Builds on: Scalar and Vector Fields · Line Integrals · The Divergence of a Vector Field · Spherical Coordinates Leads to: The Divergence (Gauss) Theorem · The Curl (Stokes) Theorem · Volume Integrals

Nothing actually moves in a static electric field — why is its surface integral still called a flux?

The name is a fossil from fluid mechanics, where $\int\mathbf{v}\cdot\mathrm{d}\mathbf{a}$ really is litres per second through the surface. For $\mathbf{E}$ nothing flows; the integral counts field lines piercing the surface, weighted by strength. The word survives because the mathematics is identical and the analogy is productive: sources and sinks, lines in and out, conservation of throughput. Keep the fluid picture as intuition and remember that for $\mathbf{E}$ and $\mathbf{B}$ "flux" is a static geometric quantity — no time, no motion, no transport implied.

How do I know which way the normal points on an open surface?

You do not know — you choose, and then you owe consistency. An open surface has two perfectly good sides; pick one as positive and the flux acquires a definite sign. The obligation comes when a boundary curve is in play (Faraday's law, Stokes's theorem): once you choose a traversal direction for the rim, the right-hand rule fixes the normal — fingers along the circulation, thumb along $\hat{\mathbf{n}}$ . Reverse either choice and both sides of the theorem flip together, leaving the physics intact. Sign errors come from flipping one without the other.

Why can I use any surface spanning a loop to compute magnetic flux?

Because $\nabla\cdot\mathbf{B}=0$ . Take two different surfaces sharing the same rim — a flat disk and a bulging cap. Together (one orientation reversed) they form a closed surface, and the net flux of a divergence-free field through any closed surface is zero; hence the two open-surface fluxes are equal. This freedom is used constantly: for a solenoid you slide the surface to where the field is simple. The same argument fails for $\mathbf{E}$ in the presence of charge — if a charge sits between the two surfaces, their fluxes genuinely differ by $q/\varepsilon_0$ .

When do I integrate with the scalar da instead of the vector da?

Use the vector $\mathrm{d}\mathbf{a}$ when the question is throughput of a vector field — flux, flow rate, Gauss's law — because direction and sign matter. Use the scalar $\mathrm{d}a$ for direction-blind accumulations over a surface: total charge from a surface density ( $q=\int\sigma\,\mathrm{d}a$ ), mass of a shell, area itself. The grammar mirrors line integrals ( $\mathrm{d}\boldsymbol{\ell}$ versus $\mathrm{d}s$ ): vector element for projections of vector fields, scalar element for densities of scalar quantities.

What exactly is the solid-angle picture of the 4π result?

Stand at the source and look outward: each patch of a surrounding surface blocks a certain patch of your sky, measured in steradians. The flux of $\hat{\mathbf{r}}/r^2$ through a patch turns out to equal exactly the solid angle it subtends — the $1/r^2$ field against the $r^2$ -growing area makes distance irrelevant, leaving only angular coverage. A closed surface around you covers the whole sky, $4\pi$ steradians, whatever its shape; a surface that does not enclose you covers a patch of sky twice, once going in ( $-$ ) and once out ( $+$ ), netting zero. Gauss's law is this bookkeeping with a charge at the vantage point.

What are the units of electric and magnetic flux, and why are they so different?

Electric flux carries $\mathrm{N\,m^2/C}$ (equivalently $\mathrm{V\,m}$ ) — field times area. Magnetic flux carries its own SI unit, the weber, $1\ \mathrm{Wb} = 1\ \mathrm{T\,m^2} = 1\ \mathrm{V\,s}$ . The volt-second form is the revealing one: by Faraday's law, one weber changing in one second drives one volt of EMF — magnetic flux earned a named unit precisely because generators and transformers made it an engineering commodity. Electric flux, useful mostly inside Gauss's law where it immediately becomes $Q/\varepsilon_0$ , never needed one.

Does the flux through a surface change if I wiggle the surface while keeping its boundary fixed?

For an arbitrary field, yes — the two surfaces sandwich a volume, and the flux difference equals the integral of $\nabla\cdot\mathbf{v}$ over it (divergence theorem). For a divergence-free field, no: wiggle freely, the flux rides on the boundary alone. This is why magnetic flux "through a loop" is well-defined without specifying a membrane, and why the flux of a curl (always divergence-free) in Stokes's theorem tolerates any spanning surface. Field lines give the intuition: with no sources between the surfaces, every line through one must thread the other.