Volume Integrals

A volume integral is the three-dimensional Riemann integral: instead of accumulating $f(x)$ along a line or a density across a surface, we accumulate a field $f(\mathbf{r})$ throughout a solid region $V\subset\mathbb{R}^3$ . The recipe is the same one Riemann fixed for the line — partition $V$ into small cells, weight each cell's volume by the field inside it, and take the limit as the cells shrink to points. What comes out depends only on what you feed in: with the mass density $\rho$ it returns total mass, with charge density it returns total charge, with $f=1$ it returns the plain geometric volume, and with $\mathbf{r}\rho$ it returns the first moment that locates the centre of mass. One machine, many totals. Two skills carry the whole subject, both geometric rather than symbolic. The first is choosing coordinates matched to the region's symmetry, so a sphere's boundary collapses to the single equation $r=R$ instead of a thicket of square-root limits. The second is writing the volume element correctly: the factors $r^2\sin\theta$ (spherical) and $\rho$ (cylindrical) are not rules to memorise but the true volume of a curved coordinate cell, delivered by the Jacobian determinant of the coordinate map. Getting that Jacobian right is the single most common make-or-break step. Volume integrals are also the hinge of vector calculus. Through the divergence (Gauss) theorem, $\int_V(\nabla\cdot\mathbf{F})\,\mathrm{d}V=\oint_{\partial V}\mathbf{F}\cdot\mathrm{d}\mathbf{S}$ , they equate the source density summed inside a region to the net flux leaving its boundary — the identity behind Gauss's law, the continuity equation, and every Green's identity in potential theory. Master the volume integral and these deeper theorems become bookkeeping.

Key takeaways

A volume integral $\int_V f\,\mathrm{d}V$ is the limit of $\sum_i f(\mathbf{r}_i)\,\Delta V_i$ over a partition of a solid region, returning a total — mass, charge, moment, or the volume itself.
The volume element is $\mathrm{d}V=\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}z$ (Cartesian), $\rho\,\mathrm{d}\rho\,\mathrm{d}\phi\,\mathrm{d}z$ (cylindrical), and $r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ (spherical); the extra factors are the Jacobian of the coordinate change.
Any smooth change of variables rescales $\mathrm{d}V$ by the absolute Jacobian determinant $\left|\partial(x,y,z)/\partial(u,v,w)\right|$ — geometry, not a mnemonic.
Vector fields integrate componentwise in a fixed Cartesian basis, giving total force, dipole moment, and the centre of mass $\mathbf{R}_{\mathrm{cm}}=\tfrac1M\int_V\mathbf{r}\rho\,\mathrm{d}V$ .
The divergence theorem ties $\int_V(\nabla\cdot\mathbf{F})\,\mathrm{d}V$ to the boundary flux $\oint_{\partial V}\mathbf{F}\cdot\mathrm{d}\mathbf{S}$ , making the volume integral the bridge from local source to global total.
Key idea: a volume integral is the limit of a sum $\sum_i f(\mathbf{r}_i)\,\Delta V_i$ over a partition of the region; the entire craft is writing $\mathrm{d}V$ in symmetry-adapted coordinates with its Jacobian factor and choosing limits that sweep $V$ exactly once.

The core insight

A volume integral is the limit of a sum $\sum_i f(\mathbf{r}_i)\,\Delta V_i$ over a partition of a three-dimensional region; the whole craft lies in writing $\mathrm{d}V$ in symmetry-adapted coordinates (with its Jacobian factor) and choosing limits that sweep $V$ exactly once.

Where this lives today

Volume integrals are the quiet engine of computational physics and precision measurement. Every finite-element and finite-volume solver — the tools that design aircraft wings, predict weather, and model the heart — works by breaking a domain into cells and evaluating $\int_V f\,\mathrm{d}V$ cell by cell, which is the Riemann-sum definition turned into an algorithm. In geodesy, the twin GRACE and GRACE-FO satellites map Earth's gravity field to detect its shifting mass: the gravitational multipole moments they infer are volume integrals of the planet's density, $\int_V\rho\,r^n\,\mathrm{d}V$ , and tracking their monthly change has weighed the loss of the polar ice sheets to within gigatonnes. Magnetic-resonance imaging reconstructs a body as a volume integral of spin density. Wherever a "total" is extracted from a "density" — mass, charge, energy, or probability — a volume integral is doing the work, with the divergence theorem usually right behind it, converting an intractable interior into a manageable boundary.

The physical picture

Imagine chopping a solid body into a swarm of tiny cells. Each cell has a volume $\Delta V$ and carries a local field value $f$ ; multiply them, add over every cell, then refine the chopping until the cells shrink to points. The volume integral is what that sum converges to — the same story as the one-dimensional integral being the limit of a sum of rectangle areas, lifted into three dimensions. Nothing about the definition prefers a coordinate system; the coordinates only enter when we compute, and they enter through the shape of the cell.

Figure

Chop the region V into cells and sum the field over each — the Riemann sum ∑ᵢ f(rᵢ) ΔVᵢ converges to ∫_V f dV. A spherical cell has physical edges dr, r dθ and r sinθ dφ, so its volume is the Jacobian factor dV = r² sinθ dr dθ dφ, and the r² tells you cells far from the origin sweep more space

The Jacobian factor earns its place by pure geometry. In spherical coordinates a cell spanning the increments $\mathrm{d}r,\mathrm{d}\theta,\mathrm{d}\phi$ has physical edge lengths $\mathrm{d}r$ (radial), $r\,\mathrm{d}\theta$ (a polar arc), and $r\sin\theta\,\mathrm{d}\phi$ (an azimuthal arc, shrinking to zero at the poles where $\sin\theta\to0$ ). Their product $r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ is the honest volume of that curved brick — the $r^2$ says cells far from the origin subtend more space, the $\sin\theta$ that cells near the poles are pinched thin. Symmetry should pick the coordinates before a single limit is written: a ball begs for spherical, a pipe or disk for cylindrical, a slab for Cartesian. The second figure shows why volume integrals are more than bookkeeping. The divergence theorem converts the volume integral of a source density $\nabla\cdot\mathbf{F}$ into the flux of $\mathbf{F}$ through the bounding surface — so the total escaping a region is fixed entirely by what the region contains. Feeding it the special field $\mathbf{F}=\tfrac13\mathbf{r}$ , whose divergence is exactly $1$ , even turns the plain geometric volume into a boundary integral.

Figure

The divergence theorem makes a volume integral equal to a boundary flux: ∫_V (∇·F) dV = ∮_∂V F·dS. Sum the local source density inside, and it equals the net flux leaving the surface. Taking F = r/3 gives ∇·F = 1, so the volume itself is a surface integral — for the unit ball V = ∮ (r/3)·dS = 4π/3 ≈ 4.19

The core equations

Everything begins with the definition: the volume integral is the limit of a weighted sum over a shrinking partition, exactly mirroring the one-dimensional Riemann integral. The limit exists, and is independent of how the region is chopped, whenever $f$ is bounded and continuous almost everywhere on a region of finite volume.

\int_V f(\mathbf{r})\,\mathrm{d}V=\lim_{\Delta V_i\to0}\sum_i f(\mathbf{r}_i)\,\Delta V_i \qquad (1)

where: $V$ — the three-dimensional region of integration; $f(\mathbf{r})$ — the scalar field or density sampled at $\mathbf{r}$ ; $\Delta V_i$ — volume of the $i$ -th cell; $\mathbf{r}_i$ — a sample point inside cell $i$ . Reading: add up "how much field times how much room" over every cell, then refine — the value is a single number attached to the whole region. To compute, we replace the abstract cell by its size in a chosen coordinate system. The three standard volume elements are the Cartesian box and its curvilinear cousins, whose extra factors are Jacobians.

\mathrm{d}V=\underbrace{\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}z}_{\text{Cartesian}}=\underbrace{\rho\,\mathrm{d}\rho\,\mathrm{d}\phi\,\mathrm{d}z}_{\text{cylindrical}}=\underbrace{r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi}_{\text{spherical}} \qquad (2)

where: $\rho$ (cylindrical) — distance from the $z$ -axis; $r$ (spherical) — distance from the origin; $\theta$ — polar angle from the $+z$ axis; $\phi$ — azimuthal angle. Reading: the same physical cell, described three ways; the factors $\rho$ and $r^2\sin\theta$ convert the coordinate box into real volume. For any smooth invertible change of variables $\mathbf{r}=\mathbf{r}(u,v,w)$ , equation (2) is a special case of a single master formula: multiply by the absolute Jacobian determinant.

\int_V f\,\mathrm{d}V=\iiint f\,\left|\frac{\partial(x,y,z)}{\partial(u,v,w)}\right|\,\mathrm{d}u\,\mathrm{d}v\,\mathrm{d}w \qquad (3)

where: $(u,v,w)$ — the new curvilinear coordinates; $\partial(x,y,z)/\partial(u,v,w)$ — the $3\times3$ Jacobian matrix of the map; the vertical bars denote the absolute value of its determinant. Reading: a coordinate change stretches and skews the cell, and the Jacobian is exactly the local volume-stretch factor. The workhorse applications feed a density into (1). Total mass (or charge) is the integral of the density; the centre of mass is the density-weighted mean position.

M=\int_V\rho(\mathbf{r})\,\mathrm{d}V,\qquad \mathbf{R}_{\mathrm{cm}}=\frac1M\int_V\mathbf{r}\,\rho(\mathbf{r})\,\mathrm{d}V \qquad (4)

where: $M$ — total mass in $V$ ; $\rho(\mathbf{r})$ — mass density; $\mathbf{R}_{\mathrm{cm}}$ — centre-of-mass position; $\mathbf{r}$ — position of the volume element. Reading: the first integral is the zeroth moment of the density, the second its normalised first moment. A vector field is integrated by resolving it into a fixed Cartesian basis and integrating each component separately — the basis vectors $\hat{\mathbf{x}},\hat{\mathbf{y}},\hat{\mathbf{z}}$ come outside the integral because they are constant.

\int_V\mathbf{F}\,\mathrm{d}V=\hat{\mathbf{x}}\!\int_V F_x\,\mathrm{d}V+\hat{\mathbf{y}}\!\int_V F_y\,\mathrm{d}V+\hat{\mathbf{z}}\!\int_V F_z\,\mathrm{d}V \qquad (5)

where: $\mathbf{F}$ — the vector field over $V$ ; $F_x,F_y,F_z$ — its Cartesian components. Reading: three ordinary scalar volume integrals assembled into one vector — used for total force, electric and magnetic dipole moments, and the centroid. Finally, the identity that gives volume integrals their reach: the divergence theorem equates the volume integral of a divergence to the flux through the closed bounding surface.

\int_V(\nabla\cdot\mathbf{F})\,\mathrm{d}V=\oint_{\partial V}\mathbf{F}\cdot\mathrm{d}\mathbf{S} \qquad (6)

where: $\nabla\cdot\mathbf{F}$ — the divergence, a local source density; $\partial V$ — the closed surface bounding $V$ ; $\mathrm{d}\mathbf{S}$ — the outward-pointing area element. Reading: what a region emits equals what it contains — the statement that upgrades a volume integral into a boundary law.

Derivation AThe Jacobian volume element from a change of coordinates

The goal is to show that the factors $\rho$ and $r^2\sin\theta$ in equation (2) are not memorised rules but the exact volume of a distorted coordinate cell, and that for any smooth map $\mathbf{r}=\mathbf{r}(u,v,w)$ the abstract box $\mathrm{d}u\,\mathrm{d}v\,\mathrm{d}w$ is rescaled by the absolute value of the Jacobian determinant. We build the three physical edge vectors of a cell, take their scalar triple product to get its volume, recognise that triple product as a determinant, and then specialise to spherical coordinates to recover $r^2\sin\theta$ from first principles.

Step 1. Hold two coordinates fixed and vary the third: the increments produce three physical edge vectors $\mathrm{d}\mathbf{r}_u=\tfrac{\partial\mathbf{r}}{\partial u}\,\mathrm{d}u$ , $\mathrm{d}\mathbf{r}_v=\tfrac{\partial\mathbf{r}}{\partial v}\,\mathrm{d}v$ , $\mathrm{d}\mathbf{r}_w=\tfrac{\partial\mathbf{r}}{\partial w}\,\mathrm{d}w$ . Why: a coordinate increment is abstract, but $\partial\mathbf{r}/\partial u$ turns it into an actual displacement in space, so the cell is a small parallelepiped spanned by these three vectors.

Step 2. The volume of a parallelepiped is the magnitude of the scalar triple product of its edges, $\mathrm{d}V=\left|\mathrm{d}\mathbf{r}_u\cdot(\mathrm{d}\mathbf{r}_v\times\mathrm{d}\mathbf{r}_w)\right|$ . Why: the cross product gives the base area with the right orientation, and dotting with the third edge multiplies by the perpendicular height — the absolute value discards orientation, since volume is positive.

Step 3. The triple product of three vectors equals the determinant of the matrix whose columns are those vectors, so $\mathrm{d}V=\left|\det\tfrac{\partial(x,y,z)}{\partial(u,v,w)}\right|\,\mathrm{d}u\,\mathrm{d}v\,\mathrm{d}w$ , which is equation (3). Why: the $\partial\mathbf{r}/\partial u_i$ are exactly the columns of the Jacobian matrix, so the geometric volume factor and the analytic determinant are the same object.

Step 4. Specialise with $x=r\sin\theta\cos\phi$ , $y=r\sin\theta\sin\phi$ , $z=r\cos\theta$ ; expanding the $3\times3$ determinant gives $\det=r^2\sin\theta$ , so $\mathrm{d}V=r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ . Why: the abstract master formula, evaluated for the spherical map, reproduces the familiar element — confirming the factor is geometry, and by the same route cylindrical coordinates give $\det=\rho$ .

Go deeper · Derivation B · 1 minThe volume of a sphere in spherical coordinates

Derivation BThe volume of a sphere in spherical coordinates

The goal is the canonical sanity check that pins down the spherical volume element: set $f=1$ over the ball $r\le R$ and confirm the integral returns the classical $\tfrac43\pi R^3$ . Because the integrand and the volume element are both products of single-variable functions, the triple integral separates into three independent one-dimensional integrals, and their product must recover the elementary-geometry answer — if it does not, the volume element is wrong. This is the fastest way to catch a dropped Jacobian factor.

Step 1. Write the volume as $V=\int_0^{2\pi}\!\int_0^{\pi}\!\int_0^{R} r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ , with $r\in[0,R]$ , $\theta\in[0,\pi]$ , $\phi\in[0,2\pi]$ . Why: these three ranges sweep the entire ball exactly once with no overlap, which is the condition for the iterated integral to equal the volume integral.

Step 2. The radial integral is $\int_0^R r^2\,\mathrm{d}r=R^3/3$ . Why: the $r^2$ comes from the Jacobian, not from the constant integrand $f=1$ ; the power rule then does the rest.

Step 3. The polar integral is $\int_0^{\pi}\sin\theta\,\mathrm{d}\theta=[-\cos\theta]_0^{\pi}=2$ , and the azimuthal integral is $\int_0^{2\pi}\mathrm{d}\phi=2\pi$ . Why: the $\sin\theta$ is again the Jacobian factor doing its job, and the full turn in $\phi$ accounts for the whole revolution about the polar axis.

Step 4. Multiply the three separable factors: $V=\tfrac{R^3}{3}\cdot2\cdot2\pi=\tfrac43\pi R^3$ . Why: separability lets a genuinely three-dimensional integral collapse to a product, and the product reproduces the textbook volume — the volume element is validated.

Go deeper · Derivation C · 1 minFrom a charge density to the integral form of Gauss's law

Derivation CFrom a charge density to the integral form of Gauss's law

The goal is to show that a volume integral of charge density, combined with the divergence theorem, is precisely what turns Maxwell's differential first equation into the integral Gauss's law taught in introductory electromagnetism. Starting from $\nabla\cdot\mathbf{E}=\rho/\varepsilon_0$ , we integrate both sides over an arbitrary region $V$ , recognise the right side as the enclosed charge (a volume integral of density), and convert the left side to a surface flux with the divergence theorem — the volume integral is the load-bearing step that makes the two forms equivalent.

Step 1. Integrate the local law over $V$ : $\int_V(\nabla\cdot\mathbf{E})\,\mathrm{d}V=\int_V\tfrac{\rho}{\varepsilon_0}\,\mathrm{d}V$ . Why: the differential Maxwell equation holds at every point, so it holds when integrated over any region — this is what promotes a pointwise statement to a statement about a finite volume.

Step 2. The right side is the enclosed charge over $\varepsilon_0$ : $\int_V\tfrac{\rho}{\varepsilon_0}\,\mathrm{d}V=\tfrac1{\varepsilon_0}\int_V\rho\,\mathrm{d}V=\tfrac{Q_{\mathrm{enc}}}{\varepsilon_0}$ . Why: by definition $Q_{\mathrm{enc}}=\int_V\rho\,\mathrm{d}V$ is the volume integral of charge density — the very operation this article is about — so no new physics enters here, only the definition of total charge.

Step 3. Apply the divergence theorem to the left side: $\int_V(\nabla\cdot\mathbf{E})\,\mathrm{d}V=\oint_{\partial V}\mathbf{E}\cdot\mathrm{d}\mathbf{S}$ . Why: equation (6) trades the volume integral of a divergence for the flux through the boundary, converting an integral over the interior into one over the surface without changing its value.

Step 4. Equate the two evaluations to obtain $\oint_{\partial V}\mathbf{E}\cdot\mathrm{d}\mathbf{S}=\tfrac{Q_{\mathrm{enc}}}{\varepsilon_0}$ . Why: both sides equal the same integrated quantity, so the flux of $\mathbf{E}$ out of any closed surface is fixed by the charge it encloses — the integral Gauss's law, born from a volume integral.

Limiting cases

The general change-of-variables formula specialises to each standard coordinate system, and to the trivial constant-field case, through its Jacobian:

Case	Result	Meaning
Cartesian	$\mathrm{d}V=\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}z$	Jacobian is unity; best for boxes and planar-faced regions.
Cylindrical	$\mathrm{d}V=\rho\,\mathrm{d}\rho\,\mathrm{d}\phi\,\mathrm{d}z$	Jacobian $\rho$ ; matched to axial symmetry (pipes, disks, solenoids).
Spherical	$\mathrm{d}V=r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$	Jacobian $r^2\sin\theta$ ; matched to point/radial symmetry (balls, shells).
Constant field $f=1$	$\int_V\mathrm{d}V=\mathrm{vol}(V)$	the integral returns the plain geometric volume.
General coordinates	$\mathrm{d}V=\left\vert\det\tfrac{\partial(x,y,z)}{\partial(u,v,w)}\right\vert\,\mathrm{d}u\,\mathrm{d}v\,\mathrm{d}w$	the absolute Jacobian rescales the abstract cell to physical volume.

Common trapDropping the Jacobian, or pulling a curvilinear unit vector out of the integral

Two errors dominate. The first is writing $\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ instead of $r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ — omitting the Jacobian factor. It is easy to spot: without it, $\int_V\mathrm{d}V$ for a ball does not come out to $\tfrac43\pi R^3$ , so always sanity-check a new setup against a known volume. The second is subtler and bites when integrating a vector field. You may not write $\int_V\mathbf{F}\,\mathrm{d}V=\hat{\mathbf{r}}\int_V F_r\,\mathrm{d}V$ , because $\hat{\mathbf{r}}$ , $\hat{\boldsymbol\theta}$ , $\hat{\boldsymbol\phi}$ point in different directions at every point and cannot be taken outside the integral. A uniformly radial field over a full sphere integrates to zero by symmetry, not to something proportional to $\hat{\mathbf{r}}$ . Always resolve $\mathbf{F}$ into the constant Cartesian basis before integrating componentwise, as in equation (5).

Worked examples

1 · Moment of inertia of a uniform solid sphere (fully worked)

A uniform solid sphere has mass $M=2.0\ \mathrm{kg}$ and radius $R=0.10\ \mathrm{m}$ . Find its moment of inertia about a diameter.

Step 1. The moment of inertia about the $z$ -axis is the volume integral $I=\int_V\rho\,d_\perp^2\,\mathrm{d}V$ , where $d_\perp^2=x^2+y^2=r^2\sin^2\theta$ is the squared distance from the axis. With constant density $\rho$ and the spherical volume element, $I=\rho\int_0^{2\pi}\!\int_0^{\pi}\!\int_0^{R} (r^2\sin^2\theta)\,r^2\sin\theta\,\mathrm{d}r\,\mathrm{d}\theta\,\mathrm{d}\phi$ , so the integrand is $\rho\,r^4\sin^3\theta$ .

Step 2. Separate the three factors: $\int_0^R r^4\,\mathrm{d}r=\tfrac{R^5}{5}$ , $\int_0^{\pi}\sin^3\theta\,\mathrm{d}\theta=\tfrac43$ , and $\int_0^{2\pi}\mathrm{d}\phi=2\pi$ . Multiplying, $I=\rho\cdot\tfrac{R^5}{5}\cdot\tfrac43\cdot2\pi=\tfrac{8\pi\rho R^5}{15}$ .

Step 3. Eliminate $\rho$ using $M=\rho\cdot\tfrac43\pi R^3$ , i.e. $\rho=3M/(4\pi R^3)$ . Substituting, $I=\dfrac{8\pi R^5}{15}\cdot\dfrac{3M}{4\pi R^3}=\boldsymbol{\tfrac25 MR^2}$ , the classic result — independent of density and of the sphere's size relative to its mass.

Step 4. Put in numbers: $I=\tfrac25(2.0)(0.10)^2=0.4\times2.0\times0.01=\mathbf{8.0\times10^{-3}\ kg\,m^2}$ . The key intermediate $\int_0^{\pi}\sin^3\theta\,\mathrm{d}\theta=\tfrac43$ is worth memorising; it recurs in every spherical moment calculation.

2 · Total charge of a radially varying ball (fully worked)

A ball of radius $R=0.20\ \mathrm{m}$ carries charge density $\rho(r)=\rho_0\left(1-\tfrac{r}{R}\right)$ with $\rho_0=3.0\times10^{-6}\ \mathrm{C/m^3}$ — dense at the centre, vanishing at the surface. Find the total charge.

Step 1. The charge is the volume integral $Q=\int_V\rho(r)\,\mathrm{d}V$ . Since $\rho$ depends only on $r$ , the angular integrals give the full solid angle $4\pi$ and the problem reduces to a single radial integral: $Q=4\pi\int_0^R\rho_0\left(1-\tfrac{r}{R}\right)r^2\,\mathrm{d}r$ .

Step 2. Split and integrate: $\int_0^R r^2\,\mathrm{d}r=\tfrac{R^3}{3}$ and $\tfrac1R\int_0^R r^3\,\mathrm{d}r=\tfrac1R\cdot\tfrac{R^4}{4}=\tfrac{R^3}{4}$ . The bracket becomes $\tfrac{R^3}{3}-\tfrac{R^3}{4}=\tfrac{R^3}{12}$ .

Step 3. Assemble: $Q=4\pi\rho_0\cdot\tfrac{R^3}{12}=\boldsymbol{\tfrac{\pi\rho_0 R^3}{3}}$ . Numerically, with $R^3=8.0\times10^{-3}\ \mathrm{m^3}$ , $Q=\tfrac{\pi(3.0\times10^{-6})(8.0\times10^{-3})}{3}=\pi\times8.0\times10^{-9}=\mathbf{2.5\times10^{-8}\ C}\approx25\ \mathrm{nC}$ .

Step 4. Sanity check via the average density: $\bar\rho=Q/\big(\tfrac43\pi R^3\big)=\rho_0/4=7.5\times10^{-7}\ \mathrm{C/m^3}$ . A linear taper from $\rho_0$ at the centre to $0$ at the surface, weighted by the $r^2$ volume factor, averages to a quarter of the peak — not a half, because the outer shells (where $\rho$ is small) carry disproportionately more volume. The $r^2$ weighting is what makes the average $\rho_0/4$ rather than $\rho_0/2$ .

Physics of the portrait

The volume integral becomes a theorem in the hands of the man who tied the source inside a region to the flux across its boundary — and who, not coincidentally, gave electromagnetism its most-used law.

Origins · 1 minCarl Friedrich Gauss and how this was discovered

Carl Friedrich Gauss

1777–1855 · the divergence theorem, potential theory, and Gauss's law

Gauss is remembered as the "Prince of Mathematicians", but for the physicist his lasting gift is the machinery that connects a volume integral of a source to a surface integral of flux. Working on the theory of gravitation and terrestrial magnetism in the 1810s, he proved special cases of what is now the divergence theorem — that the integral of a divergence over a region equals the flux through its boundary — and used it to relate the potential of a mass or charge distribution to integrals over the enclosing surface. The general theorem was stated and proved by Mikhail Ostrogradsky in 1826, and Green's identities of 1828 are the same idea applied twice, which is why the result travels under the names Gauss, Ostrogradsky, and Green. Gauss's own law of electrostatics, $\oint\mathbf{E}\cdot\mathrm{d}\mathbf{S}=Q_{\mathrm{enc}}/\varepsilon_0$ , is exactly the volume integral of charge density made global by his theorem — the derivation reproduced above. He also founded the systematic study of the potential $\Phi(\mathbf{r})=\tfrac1{4\pi\varepsilon_0}\int_V\tfrac{\rho(\mathbf{r}')}{|\mathbf{r}-\mathbf{r}'|}\,\mathrm{d}V'$ , a volume integral over sources that remains the template for boundary-value problems throughout physics. When Bernhard Riemann, Gauss's own student, later formalised the integral itself as a limit of sums in 1854, both halves of this article — the definition and its deepest theorem — were, in a sense, Göttingen work.

History

1762

— Lagrange meets the volume-to-surface identity while studying fluid flow

1813

— Gauss proves special cases of the divergence theorem in his work on gravitation

1826

— Ostrogradsky states and proves the divergence theorem in general form

1828

— Green publishes the identities linking volume integrals to boundary integrals

1854

— Riemann's habilitation formalises the integral as a limit of sums

Going further

Texts. Arfken & Weber, Mathematical Methods for Physicists, 7th ed., ch. 3 (vector analysis, Jacobians, Gauss's theorem) · Riley, Hobson & Bence, Mathematical Methods for Physics and Engineering, 3rd ed., ch. 6 and 11 (multiple integrals and vector calculus) · Stone & Goldbart, Mathematics for Physics, ch. on vector calculus · Boas, Mathematical Methods in the Physical Sciences, 3rd ed., ch. 5 · Griffiths, Introduction to Electrodynamics, 4th ed., ch. 1 (for the divergence theorem and Gauss's law). Key results to keep in mind: the spherical Jacobian $\det=r^2\sin\theta$ ; the three separable integrals giving $\tfrac43\pi R^3$ ; the identity $\int_0^{\pi}\sin^3\theta\,\mathrm{d}\theta=\tfrac43$ used in the moment-of-inertia example ( $I=\tfrac25 MR^2\Rightarrow8.0\times10^{-3}\ \mathrm{kg\,m^2}$ ); and the radial-charge integral giving $Q=\tfrac{\pi\rho_0R^3}{3}\approx25\ \mathrm{nC}$ with volume-average density $\rho_0/4$ .

Prerequisites. Scalar and Vector Fields · Surface Integrals and Flux · Divergence of a Vector Field. Leads to. The Divergence (Gauss) Theorem · The Laplacian Operator · Green's Theorem in the Plane.

Ask the tutor: Bring me any solid — a cone, a shell, a lopsided charge cloud — and I will help you pick the coordinates, write $\mathrm{d}V$ with the right Jacobian, and set the limits so the integral sweeps it exactly once.

Is a volume integral just a triple integral?

In practice, yes — once you pick coordinates, $\int_V f\,\mathrm{d}V$ becomes an iterated triple integral like $\int\!\int\!\int f\,J\,\mathrm{d}u\,\mathrm{d}v\,\mathrm{d}w$ . But the two ideas are logically distinct. The volume integral is defined coordinate-free, as a limit of sums over a partition, and its value depends only on the region and the field, not on how you slice it. The "triple integral" is one way to compute that value after choosing coordinates. Fubini's theorem is the guarantee that the iterated integral equals the volume integral (and that the order of the three integrations does not matter) whenever $f$ is well behaved. So every volume integral can be evaluated as a triple integral, but thinking of it as coordinate-free first is what tells you which coordinates to choose.

How do I choose between Cartesian, cylindrical, and spherical coordinates?

Let the boundary of the region decide. If the region's surfaces are planes (a box, a wedge), Cartesian coordinates keep the limits constant and simple. If there is an axis of symmetry — a cylinder, a disk, a cone, a solenoid — use cylindrical coordinates so the boundary becomes $\rho=\text{const}$ or $z=\text{const}$ . If there is a centre of symmetry — a ball, a spherical shell, a point charge's field — use spherical coordinates so the boundary is simply $r=R$ . The payoff is twofold: the limits of integration become constants, and the integrand often separates into a product of single-variable functions. Choosing the wrong system is rarely fatal but usually forces ugly square-root limits that defeat the calculation by hand.

Why can't I pull r-hat out of a vector volume integral like I pull x-hat out?

Because $\hat{\mathbf{x}}$ is the same arrow everywhere in space, while $\hat{\mathbf{r}}$ points in a different direction at every point. A constant can come outside an integral; a position-dependent quantity cannot. So $\int_V\mathbf{F}\,\mathrm{d}V=\hat{\mathbf{x}}\int F_x\,\mathrm{d}V+\hat{\mathbf{y}}\int F_y\,\mathrm{d}V+\hat{\mathbf{z}}\int F_z\,\mathrm{d}V$ is valid, but $\hat{\mathbf{r}}\int F_r\,\mathrm{d}V$ is not. The classic trap is a purely radial field of constant magnitude over a full sphere: naively it looks nonzero, but resolving into Cartesian components and integrating shows every component cancels, giving exactly zero — the outward arrows on opposite sides of the sphere point in opposite directions and annihilate. Always integrate vector fields componentwise in the fixed Cartesian basis.

Does the order of the three integrations matter?

For the value, no — Fubini's theorem says that if $f$ is integrable (which it is for bounded, piecewise-continuous fields over finite regions, the physics norm), the iterated integral gives the same answer in any order. For the effort, it can matter enormously. When the region's limits couple the variables (say $z$ runs up to a surface depending on $\rho$ ), one order may give constant limits while another forces you to split the region into pieces. A good rule is to integrate first over the variable whose limits depend on the others, working outward to the variable with constant limits. When the integrand and element separate into single-variable functions with constant limits, the order is irrelevant and the triple integral factorises — the situation exploited in every worked example above.

What's the difference between ∫f dV, ∫F dV, and ∫(∇·F) dV?

They are three different objects with three different outputs. The first, $\int_V f\,\mathrm{d}V$ , integrates a scalar field and returns a scalar — a mass, a charge, a volume. The second, $\int_V\mathbf{F}\,\mathrm{d}V$ , integrates a vector field componentwise and returns a vector — a total force, a dipole moment, a centroid times mass. The third, $\int_V(\nabla\cdot\mathbf{F})\,\mathrm{d}V$ , integrates the scalar divergence of a vector field and returns a scalar that, by the divergence theorem, equals the net outward flux $\oint_{\partial V}\mathbf{F}\cdot\mathrm{d}\mathbf{S}$ . Only the third has a boundary-integral twin; the first two do not reduce to surface integrals in general. Keeping the type of the integrand straight — scalar, vector, or divergence-of-vector — is half the battle in setting up a problem correctly.

How does the divergence theorem actually help me compute?

It lets you trade a hard integral for an easier one in either direction. If a volume integral of a divergence is awkward but the flux through the boundary is simple (as when $\mathbf{F}$ is constant or radial and the surface is a sphere), convert to the surface integral. Conversely, a surface flux that would require patching together many faces can sometimes be replaced by a single volume integral of $\nabla\cdot\mathbf{F}$ . The most elegant trick is computing a volume itself: since $\nabla\cdot(\tfrac13\mathbf{r})=1$ , the divergence theorem gives $\mathrm{vol}(V)=\oint_{\partial V}\tfrac13\mathbf{r}\cdot\mathrm{d}\mathbf{S}$ , so a volume becomes a surface integral — the identity a CAD system uses to compute the volume of a mesh from its triangular faces alone.